Poetryandpoeticsinthepresocratic Philosophers

POETRY AND P O E T I C S I N T HE P RE S O C RA T I C
P HI L O SO P HE R S
Of the Presocratic thinkers traditionally credited with the foundation

of Greek philosophy, Xenophanes, Parmenides and Empedocles are
exceptional for writing in verse. This is the first book-length literary-
critical study of their work. It locates the surviving fragments in their
performative and wider cultural contexts, applying intertextual and
intratextual analyses in order to reconstruct the significance and
impact they conveyed for ancient audiences and readers. Building
on insights from literary theory and the philosophy of literature, the
book sheds new light on these authors’ philosophical projects and
enriches our appreciation of their works as literary artefacts. It also
expands our knowledge of the genres in which they wrote, of the
literary culture of the Western Greek world, and of the development
of Greek poetics from the Archaic to the Classical periods, exposing
the influence of these thinkers on more famous Sophistic and Platonic
ideas about literature.
tom mackenzie is a Research Fellow in the Department of Greek

and Latin at University College London.
POETRY AND POETICS IN
THE PRESOCRATIC
PHILOSOPHERS
Reading Xenophanes, Parmenides and Empedocles as
Literature
TOM MACKENZIE
University College London
University Printing House, Cambridge cb2 8bs, United Kingdom
One Liberty Plaza, 20th Floor, New York, ny 10006, USA
477 Williamstown Road, Port Melbourne, vic 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
79 Anson Road, #06–04/06, Singapore 079906
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781108843935
doi: 10.1017/9781108921084
© Tom Mackenzie 2021
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2021
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
names: Mackenzie, Tom (Classicist), author.
title: Poetry and poetics in the Presocratic philosophers : reading Xenophanes,
Parmenides and Empedocles as literature / Tom Mackenzie.
other titles: Reading Xenophanes, Parmenides, and Empedocles as literature
description: Cambridge, United Kingdom : Cambridge University Press, 2021. | Includes
bibliographical references and index.
identifiers: lccn 2021000634 (print) | lccn 2021000635 (ebook) | ISBN 9781108843935
(hardcover) | isbn 9781108925846 (paperback) | isbn 9781108921084 (ebook)
subjects: lcsh: Xenophanes, approximately 570 B.C.-approximately 478 B.C. |
Parmenides. | Empedocles. | Didactic poetry, Greek – History and criticism. |
Greek poetry – History and criticism. | Pre-Socratic philosophers.
classification: lcc pa3022.d5 m33 2021 (print) | lcc pa3022.d5 (ebook) |
ddc 881/.0109–dc23
LC record available at https://lccn.loc.gov/2021000634
LC ebook record available at https://lccn.loc.gov/2021000635
isbn 978-1-108-84393-5 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
For my mother and the memory of my father
Contents
Acknowledgements page ix
List of Abbreviations xi
Note on Fragment Numbers xiii
Introduction
Philosophy and Poetry, Truth and Symbolism 1
1 Aims, Methods and Assumptions 1
2 Truth, Symbolism and Sublime Perspectives in Early Greek Song 9
3 Poetics 20
1 Xenophanes 24
1 Introduction 24
2 Xenophanes and the Hexameter Genre 27
a Epistemology and the Immersive World of the Text 29
b Xenophanes’ Theology and the Language of Hexameter Poetry 34
c The Attacks on Homer and Hesiod 41
3 Xenophanes’ Elegies 46
a Xenophanes as a ‘Wandering Poet’ 49
b D61=B2 52
c D59=B1 56
4 Xenophanean Elegies and Xenophanean Hexameters 61
5 Conclusion 63
2 Parmenides 65
1 Introduction 65
2 Poetry and Parmenides’ Philosophy 67
3 Poetry and Deception 70
a Deception and the Doxa 70
b The Sirens 73
c Deception and the Proem 76
4 Profound Visions and Sublime Emotions 78
a Seeing the Unseeable 78
b The Vision 80
c Perspective and Character 82
vii
viii Contents
d The Emotional Trajectory of the Poem 84
e Parmenides and the Sublime 87
f Mystic Initiation 90
g The Reception of Parmenides’ Poem 91
h The Emotions and Parmenides’ Philosophy 93
5 Poetry and Symbolism 93
a Intratextuality 93
b Intertextuality 94
c Allegory and Subtext 97
6 Conclusion 100
3 Empedocles 102
1 Introduction 102
2 Religion and Philosophy, One Poem or Two 104
3 Empedoclean Poetics 107
a Empedocles and His Predecessors on Poetry 108
b Love and Poetry in Empedocles’ Cosmology 116
c Love’s Craft and Empedocles’ Poetry 119
d Conclusion to Section 3 124
4 Empedocles’ Narratives 126
a The Narrative of the Daimon 127
b Initiation, D257=B110 and the Education of the Singular Addressee 142
c The Two Narratives Combined 152
5 The Effects of Empedocles’ Narratives 156
a Empedocles and Myth 157
b Emotions and Aesthetics 162
c Textual Visions 166
d Interpreting the Text, Understanding the World 170
6 Conclusions: Empedocles in the History of Greek Poetry and Poetics 175
Conclusion 177
Epilogue: The Legacy of Presocratic Poetics 183
1 Introduction 183
2 Anaxagoras, Allegory and Ainigmata 183
3 Gorgias 188
4 Democritus on Literature 192
5 Democritus and the Sophists on Language 196
6 Plato and Mimesis 199
Bibliography 207
Index 236
Acknowledgements
In the lengthy process of bringing this book to light, I have racked up debts
of gratitude that no honest man can pay. I am grateful to my undergradu-
ate alma mater, Trinity College, Oxford, both for awarding me
a scholarship to assist with funding my Master’s degree and for later
appointing me as a stipendiary lecturer from 2014–2015. The doctoral
thesis on which this book is loosely based was undertaken at Magdalen
College, Oxford from 2012–2015 and was made possible by a studentship
from the Arts and Humanities Research Council. I have been able to turn it
into a book as a result of being appointed, first, as a Teaching Associate at
Cambridge during the year 2015–2016 and then as the A. G. Leventis
Research Fellow in Ancient Greek Culture at UCL from 2016–2020. In
addition to these educational institutions, thanks are due to many libraries,
but to three in particular. I have spent more hours in the Bodleian Library,
the Institute of Classical Studies library and the British Library than any
therapist would recommend.
For my doctoral thesis, I had the great good fortune to be supervised by
Tobias Reinhardt and Bruno Currie, who set examples of scholarly rigour
and erudition that I have struggled to follow. My examiners, Gregory
Hutchinson and Malcolm Heath, provided a wealth of detailed, generous
and constructive comments, in light of which I realised that, in an attempt
to perform the due diligence required of a doctoral thesis, I had become too
entangled in the familiar exegetical debates surrounding these authors
(precisely the ones I had hoped to avoid) and had to approach the topic
anew.
Cambridge and UCL provided immensely stimulating environments in
which to refine my thinking. At Cambridge, I was especially made to feel
welcome by David Butterfield, James Clackson, Renaud Gagné, Richard
Hunter, Emily Kneebone and Tim Whitmarsh. My colleagues at UCL
have been supportive far beyond what could reasonably be expected. I am
particularly grateful to Peter Agócs, Nicolò Benzi, Elena Cagnoli Fiecconi,
ix
x Acknowledgements
Adam Lecznar, Fiachra Mac Góráin, Gesine Manuwald, Phiroze Vasunia
and Maria Wyke for mentorship, camaraderie and comments on earlier
drafts.
Outside of UCL, Stefan Sienkiewicz, Henry Spelman, Barney Taylor,
Rob Watt and Lizzie Wells were kind enough to read individual chapters,
always with helpful suggestions. For more general encouragement with this
project and my career, I am grateful to Felix Budelmann, Edith Hall,
Casper de Jonge, Beppe Pezzini, David Sider, Shaul Tor, Gail Trimble and
Victoria Wohl. I would also like to thank Michael Sharp and Cambridge
University Press for taking a punt on this typescript and for always
responding to my queries in a helpful and timely manner. The two
anonymous readers for the Press provided detailed, valid and immensely
helpful criticisms which are greatly appreciated.
Moving on to longer-standing and less formal debts, like any scholar,
I owe more to my teachers than can be put into words. I cannot list them
all, but from my childhood and adolescence, I should single out Jonathan
Katz, Andy Mylne and the late Hazel Beasley; and from undergraduate
days, Bob Cowan Gregory Hutchinson, Mike Inwood, Laura Swift,
Christopher Taylor (under whom I first studied the Presocratics) and,
above all, Peter Brown, who sadly passed away as this typescript was
under peer review. For friendship and eager discussion of all things
Classical, I thank, in addition to individuals already named, Anthony
Ellis, Dan Jolowicz, Scarlett Kingsley and Henry Mason. In this regard,
I must also mention again Fiachra Mac Góráin for unfailing assistance and
advice from the period of the conception of this project right through to its
completion.
My greatest debts of all are owed to my family and to my wife. My father,
Charlie, first inspired my interest in the ancient world with trips to the
British Museum in my early childhood. Through the difficulties of his long
illness and young death, my mother, Emma, moved mountains to the ends
of the earth and back again for the upbringing of me and my siblings. My
brother Alex and my sister Bella have been a constant support in matters
professional and personal. Finally, I would never have gotten remotely
close to completing this book were it not for the encouragement, love and
companionship of Lizzie, with whom it is my great joy and honour to share
my life.
Abbreviations
Abbreviations of ancient authors and works follow those of OCD4 or, in

the case of Greek texts for which it does not provide abbreviations, LSJ.
Abbreviations of journal titles in the bibliography follow those of L’Année
philologique. In addition, the following abbreviations are used:
BNJ I. Worthington (ed.) (2007–). Brill’s New Jacoby. Leiden.

Online at https://referenceworks.brillonline.com/browse/bril
l-s-new-jacoby (accessed 18 September 2019).
BNP H. Cancik, H. Schneider, M. Landfester, C. F. Salazar and
F. G. Gentry (eds.) (1996–). Brill’s New Pauly. Leiden. Online
at https://referenceworks.brillonline.com/browse/brill-s-new-
pauly (accessed 18 September 2019).
CEG Hansen, P. A. (1983). Carmina epigraphica graeca saeculorum
VIII–V a. Chr. N. Berlin.
DK H. Diels, and W. Kranz (1952). Die Fragmente der
Vorsokratiker. 6th ed. 3 vols. Zurich.
FGE D. L. Page (1982). Further Greek Epigrams. Cambridge.
FGrH F. Jacoby (ed.) (1923–). Fragmente der griechischen Historiker.
Berlin/Leiden.
KRS G. S. Kirk, J. E. Raven and M. Schofield (1983). The Presocratic
Philosophers. 2nd ed. Cambridge.
L.-P. E. Lobel and D. L. Page (1963). Poetarum Lesbiorum frag-
menta. Corr. ed. Oxford.
LfgrE B. Snell et al. (eds.) (1955–2010). Lexikon des frühgriechischen
Epos. 4 vols. Göttingen.
LM A. Laks and G. W. Most (2016). Early Greek Philosophy. 9 vols.
Cambridge, MA.
LSJ H. G. Liddell and R. Scott (1925–1940). Greek–English
Lexicon. 9th ed., rev. H. Stuart Jones. Oxford.
xi
xii List of Abbreviations
M.-W. R. Merkelbach and M. L. West (1967). Fragmenta Hesiodea.
Oxford.
OCD4 S. Hornblower, A. Spawforth and E. Eidinow (eds.). The
Oxford Classical Dictionary. 4th ed. Oxford.
OF A. Bernabé (2004–2007) Poetae epici graeci: Testimonia et
fragmenta Pars II. 3 vols. Munich, Leipzig and Berlin.
PCG R. Kassel and C. Austin (eds.). Poetae comici graeci. 8 vols.
(1983–2001).
PMG D. L. Page (1962). Poetae melici graeci. Oxford.
SH H. Lloyd-Jones and P. Parsons (1983). Supplementum
Hellenisticum. Berlin.
SVF H. von Arnim (ed.) (1903–1905). Stoicorum veterum fragmenta.
Stuttgart.
W M. L. West (1989–1992). Iambi et elegi graeci ante Alexandrum
cantata. 2 vols. 2nd ed. Oxford.
Note on Fragment Numbers
I generally refer to fragments and testimonia of the Presocratics and

Sophists using the numbering systems of both LM and DK, with the
LM number placed first. ‘P’, ‘D’ and ‘R’ numbers are from LM, whereas
‘A’ and ‘B’ numbers are from DK. Occasionally, it has not been possible to
include both, since some fragments and testimonia are only included in one
or other edition. A further complication is that the two editions do not
always divide the fragments in the same way: DK may include lines in
a fragment that are considered spurious by LM, and, in some cases, two
fragments that are kept separate in DK are joined together in LM (e.g.
Parmenides B7 and B8 in DK are treated in LM as a single passage, D8).
Where I only give the fragment numbers and a single line number, it can be
assumed that the extent of the fragment – and therefore the line number –
is the same in both editions. For instance, Xenophanes ‘D49=B34.4’ refers
to line 4 of both D49 and B34, and the extent of the two fragments is
identical. Where I give two separate line numbers, this is not the case. For
instance, Parmenides D8.2=B7.2 refers to line 2 of D8, which is the same as
line 2 of B7, but the two fragments are vastly different in extent, since D8
also includes the much longer B8.
xiii
Introduction
Philosophy and Poetry, Truth and Symbolism
1 Aims, Methods and Assumptions

Ἐμπεδοκλῆς δὲ σπασαμένου τὸ ξίφος ἤδη νεανίου τινὸς ἐπὶ τὸν
αὑτοῦ ξενοδόχον Ἄγχιτον, ἐπεὶ δικάσας δημοσίαι τὸν τοῦ νεανίου
πατέρα ἐθανάτωσε, καὶ ἀίξαντος, ὡς εἶχε συγχύσεως καὶ θυμοῦ,
ξιφήρους παῖσαι τὸν τοῦ πατρὸς καταδικαστὴν ὡσανεὶ φονέα
Ἄγχιτον, μεθαρμοσάμενος ὡς εἶχε τὴν λύραν καὶ πεπαντικόν τι
μέλος καὶ κατασταλτικὸν μεταχειρισάμενος εὐθὺς ἀνεκρούσατο τὸ
‘νηπενθὲς ἄχολόν τε, κακῶν ἐπίληθες ἁπάντων’ κατὰ τὸν ποιητήν [δ
221], καὶ τόν τε ἑαυτοῦ ξενοδόχον Ἄγχιτον θανάτου ἐρρύσατο καὶ
τὸν νεανίαν ἀνδροφονίας. ἱστορεῖται δ’ οὗτος τῶν Ἐμπεδοκλέους
γνωρίμων ὁ δοκιμώτατος ἔκτοτε γενέσθαι.
Once, a certain young man drew a sword on Empedocles’ host,
Anchitus, since, as a judge, he had sentenced the young man’s father
to death. The youth, in confusion and anger, and bearing a sword,
rushed to strike the man who had condemned his father as if he,
Anchitus, were a murderer. Empedocles, since he had the lyre, changed
its tuning, played a softening and soothing strain, and immediately
struck up the line, ‘free from sorrow and lacking anger, forgetful of all
ills’ [Od. 4.221], in the words of the poet, and he saved his own host,
Anchitus, from death, and the young man from homicide. And it is
reported that this man then became the most distinguished of
Empedocles’ pupils. (Iamblichus VP 113=Empedocles P17<A15)
Iamblichus’ wonderful story (probably deriving from the c. 100 ce

Neopythagorean, Nicomachus of Gerasa) has, understandably, been
given short shrift by most interpreters of Empedocles,1 but it is
a revealing document in the history of his ancient reception. Here,
there is little hint of his importance for the traditions of science and
1
For an exception, note Obbink (1993) 80–1.
1
2 Introduction: Philosophy and Poetry, Truth and Symbolism
philosophy. Instead, he is a poetic performer, striking a chord on the lyre
and reciting a line of Homer. As with other ancient accounts of poets’
lives, the story could be understood as an interpretation of Empedocles’
poetry.2 It raises the themes of blood guilt and purification so prominent
in the surviving fragments. The performance of an Odyssean line is
suggestive of his use of epic metre and diction. But perhaps most reveal-
ing is the effect that Empedocles’ performance is portrayed as having on
its listener. Some idea of this effect may be gleaned from a closer examin-
ation of the line he performs. Od. 4.221 describes the drug that Helen puts
into the wine for Telemachus and Peisistratus. Anyone who drinks it
would not shed a tear, ‘not even if his mother or father were to die’ (Od.
4.224), a comment that hints at Telemachus’ fears that his own father has
perished. At least since Plutarch (Mor. 614b), Helen’s drug has been
interpreted metapoetically, as symbolising the emotional function of
the story she tells in accompaniment to the drinking. This interpretation
is put into action in Iamblichus’ anecdote: the swift transition to the
outcome of the affair implies that the performance has the immediate,
affective impact of Helen’s drug, causing the young man, like
Telemachus, to forget his anguish at his father’s death. If Empedocles
here lived up to the ideals of philosophy familiar to modernity (and
already promoted in Plato’s dialogues), we might expect him to persuade
his addressee by means of rational argument. But the young man does not
change his mind after evaluating any claims presented by Empedocles or
the Homeric line quoted. Instead, he is affected emotionally, in real time,
whilst listening to a performance, an experience that alters his outlook on
life.
This sort of emotional, transformative effect is also implicit in the
language Empedocles uses of his own poetry, where the narrator has
a reputation for being able to provide a ‘healing utterance’ (εὐηκέα βάξιν,
D4=B112.11) and his voice is presented as a ‘pure stream’ (καθαρήν . . .
πηγήν, D44=B3.2), implying that the text itself cleanses its audience like
the holy springs used in ritual purifications.3 It is paralleled in several
famous metapoetic passages from early Greek hexameter. Helen’s drug is
remarkably similar to the song of Hesiod’s Muses, who impart ‘forgetful-
ness of ills and ceasing of cares’ (Theog. 55; note also Theog. 96–103), and it
is for the same purpose that Achilles sings of the glories of men, distracting
2
For ancient biographies of poets as a form of literary criticism, note Haubold (2010) 13–4 (on Hesiod)
and Graziosi (2006). On ancient biographies of poets more generally, see Lefkowitz (2013) (orig.
published 1981); Kivilo (2010); and, on Homer, Graziosi (2002); on Hesiod, Koning (2010).
3
See the discussion of this fragment in Chapter 3 Section 4c below.
Aims, Methods and Assumptions 3
himself from his anger and frustration (Il. 9.186–9). In these passages
(among several others),4 verse does not serve primarily as a mnemonic to
preserve important information or as a means of publicising grand claims;
rather, its function is to provide a particular experience. An ancient recep-
tion, the poetological language of the fragments themselves, and the
generic context all bespeak the affective qualities of Empedocles’ text.
This aspect is overlooked in the prevailing tendency to read Empedocles
in a philosophical manner, primarily for the purposes of reconstructing
and evaluating his propositional content.5 The same preoccupation has
predominated to an even greater extent in scholarship on Xenophanes and
Parmenides. This tendency has a long history. Already in the fifth century
bce, Hippias of Elis excerpted the cosmological claims of these Presocratic
poets and other authors, categorising them by subject in his lost Sunagoge,
a practice that suggests that he was more interested in their ideas than in
any aesthetic experience of their poetry. Gorgias seems to have provided
a similar resumé of previous opinions at the start of his On What Is Not
(D26a=[Arist.] MXG 5.979a12–16).6 For the most part, Plato and
Aristotle’s procedure when discussing the Presocratics is fundamentally
analogous to that of modern philosophical scholarship. Relying heavily on
Hippias’ and Gorgias’ collections, they take the claims made (or reportedly
made) by the figures in question and explore their entailments and impli-
cations, assessing their coherence, validity and veracity.7 In short, they treat
them as philosophers, and this categorisation has stood ever since. Even
such astute literary critics as Karl Reinhardt, Bruno Snell and Hermann
Fränkel, working at a time before sub-disciplinary boundaries within the
field of Classics had reached their current rigidity, were primarily occupied
with the thoughts and mentality of these authors, rather than aesthetic or
emotional matters, as part of a teleological ‘history of the mind’
(Geistesgeschichte).8 This prioritisation of propositional content is, of
course, a manifestation of the famous ‘ancient quarrel’ between philosophy
4
E.g. passages that depict the ‘pleasure’ of song: Il. 1.474, 9.186, 189, 18.526, 604; Od. 1.347; 8.91, 368;
12.188; Hes. Theog. 917; passages that associate song and storytelling with ‘longing’ (ἵμερος): note Od.
1.421, 17.518–21, 18.194, 304, 23.144. On these, see Ritoók (1989) 333–9; Halliwell (2011) 45–50.
5
For this as the prevailing method in the history of philosophy, see e.g. Williams (2006) 257; Barnes
(2011a), (2011b).
6
On Hippias’ collection, see Snell (1944); Patzer (1986); on Gorgias’, Mansfeld (1985); on both, Palmer
(2008).
7
See e.g. Aristotle’s critical overview of previous opinions at Metaph. 1.3–10, 983a24–993b26, the root
of later groupings of the Presocratics, on which see Cherniss (1935); Frede (2008). On Plato’s
treatment of the Presocratics, see Palmer (1999); McCabe (2000).
8
See especially Reinhardt (1916); Snell (1924), (1953); Fränkel (1955), (1975b).
and poetry mentioned by Plato (Resp. 10.607b–e), who crystallised the
notion that the two activities are fundamentally incompatible, presenting
arguments that poetry (or at least, mimetic poetry) seduces the irrational
part of the soul, hindering an audience’s rational evaluation of the matters
portrayed. Since Xenophanes, Parmenides and Empedocles are retrospect-
ively categorised as philosophers, there has often been an implicit assump-
tion that they cannot be poets in any more essential sense than that they
happen to use verse.9 Consequently, they are relatively neglected by
scholars of Greek poetry.10
In light of this neglect, the present study applies methods from modern
literary criticism of ancient poetry to the texts of these three authors, whilst
also remaining sensitive to their philosophical significance, in an attempt
to explain what sorts of experiences they could provide to the attentive
listener and by what methods. In the following chapters, I enlist a range of
historical and archaeological evidence that might seem superfluous for the
history of philosophy to help reconstruct the wider cultural norms that
could affect an audience’s response. Ancient literary criticism and the
ancient reception history of these authors, often of little value in the
reconstruction of their arguments, provide evidence for their emotional
and aesthetic impact: Iamblichus’ anecdote tells us virtually nothing about
Empedocles’ cosmology, but it does indicate something about the way in
which his poetry was received in antiquity. Some might question the value
of such late evidence,11 but even if the story arose some centuries after
Empedocles’ lifetime, it is still the product of a pre-modern, primarily
polytheistic society. It therefore reflects a reception context vastly more
similar than our own to that which he could have expected.
Of special importance for this investigation is our evidence for the
literary context, by which I mean the performance context and the wider
network of texts which determine an audience’s ‘horizon of expectation’
(Erwartungshorizont)12 concerning literary genres and topoi. The readings
offered here are intertextual, both in the sense that they identify and
interpret particular allusions – that is, instances where an author refers to
another text and expects such a reference to be identified by at least some of
9
Though Kranz (1916) 1163 makes this assumption explicit. On the history of the ‘ancient quarrel’ see
the scholarship cited at Chapter 1 n. 2 below.
10
Note that they are absent from Bernabé (1987–2007), an otherwise exhaustive edition of fragmen-
tary early Greek hexameter poetry.
11
See Feeney (1995) 303–4, criticising Cairns (1972) 32 on these grounds, but we should not overlook
the crucial respects in which Imperial literary culture was similar to that of the Archaic period and
different from our own.
12
The expression was influentially coined by Jauss (1970).
the audience –13 and in the sense that they posit typological resemblances
between texts where no such allusion can be plausibly hypothesised.14
Chapter 2 offers a reading of Parmenides’ proem within the context of
other ancient accounts of supernatural journeys to places beyond the usual
mortal realm, including some Near Eastern examples from outside the
Greek tradition. Although these examples were almost certainly unfamiliar
to Parmenides and his audiences, they originated from an ancient literary
culture which was broadly similar to, and had at least some points of
contact with, that of the Greeks.15 They are therefore taken as evidence for
the sorts of narratives which could have been familiar and so can elucidate
the connotations and particularities of Parmenides’ text. In Chapter 3, the
wider context of Greek stories about wandering blood-exiles supports an
interpretation of the function of Empedocles’ story of the transmigrating
daimon who has been exiled from the gods for committing bloody deeds.
Though this study does not aim to offer radically new interpretations or
detailed analyses of the arguments, its exploration of potential audience
responses cannot be fully divorced from an investigation of the philosophy.
As a growing body of work has demonstrated, the application of literary-
critical methods can provide important insights into the Presocratics’
claims and objectives. An early milestone in this approach was Alexander
Mourelatos’ The Route of Parmenides (first published 1970),16 which pro-
vided an analysis of Parmenides’ use of metre and narrative in support of
the contention that the poem essentially outlines a ‘quest’ for truth, a ‘way’
of thinking that might lead to such an entity. In a similar vein, more recent
works by Simon Trépanier on the unity of Empedocles’ thought, Jenny
Bryan on the concepts of likeness and likelihood in Xenophanes and
Parmenides, and Shaul Tor on early Greek epistemology have enhanced
13
It has been doubted whether such allusions are a feature of early Greek poetry (e.g. on Homer, Nagy
[1979] 42; Schein [1984] 28; on lyric, R. Fowler [1987] 20–33; on early Greek performed poetry more
generally, Calame [2005] 7–10), but they are present beyond reasonable doubt in Parmenides and
Empedocles (see, especially, Wright [1998]; Most [2007]) and are quite plausibly identified in
Xenophanes (e.g. Classen [1989]; Bryan [2012] 6–57). For a defence of the language of ‘allusion’ in
early Greek epic, see B. Currie (2016) 4–36; and, in lyric, Garner (1990) 1–20.
14
That is, Kristeva’s original sense of intertextuality, not as a shorthand for ‘allusion’, but as the quality
whereby texts exist in a matrix consisting of other texts (Kristeva [1980] 66; see also Culler [1981]
110–31). For particularly influential discussions of these issues in ancient literature, note Conte
(1986); D. Fowler (1997); Hinds (1998). ‘Intertextuality’ in this sense is applied to early Greek epic by
Pucci (1987) and Tsagalis (2008).
15
See Burkert (1992) and, with particular reference to the Presocratics, (2008); Murray (1993) 81–101;
and, most fully, West (1997) (superseding his earlier work on the topic, West [1971]), though West’s
thesis of pervasive influence is convincingly modified by Metcalf (2015), whose study suggests
influence only at certain points.
16
Cited here as Mourelatos (2008a), a reprint with additional material.
our understanding of the philosophical claims embodied in these frag-
ments by providing detailed intertextual readings that locate them within
their broader cultural and literary contexts.17 In its emphasis on form and
literary context, the present book continues this trend. Although it is
primarily intended as a work of literary history, it will also be relevant to
historians of philosophy. But in contrast to those valuable contributions, it
does not enlist literary aspects primarily in the service of elucidating the
arguments. Instead, the philosophical claims are used to shed light on how
these texts are designed to affect their audiences. Chapter 1 explores the
function of Xenophanes’ poetry in the context of his epistemological and
theological claims; Chapter 2 looks at Parmenides’ use of hexameters in
light of his comments on the deceptiveness of mortal terminology; and
Chapter 3 argues that Empedocles’ poetry is designed, as Iamblichus’ story
illustrates, to have a purificatory function explicable in terms of his wider
doctrines concerning the purity of Love and the pollution of Strife.
Underpinning this approach is an assumption that the texts in question
are designed to afford experiences that are dependent on their unique
phrasing. That is to say, they fulfil the criterion for poetry encapsulated
in the New Critic Cleanth Brooks’ ‘heresy of paraphrase’: an essential
function would be lost in any prose summary of their content or themes.18
In this respect, Xenophanes, Parmenides and Empedocles are poets, not
mere versifiers, and hence I shall occasionally refer to the three collectively
as the ‘Presocratic poets’. Of course, prose may also fulfil Brooks’ criterion.
As the fragments of Heraclitus amply demonstrate, literary prose, whilst
still in its infancy in the sixth century bce, could have a mystifying yet
seductive impact on its readers, demanding a penetrating interpretation
that is sensitive to details of phrasing.19 But these sorts of effects were more
typically assumed to be an essential aspiration of verse. It is telling that
when Gorgias comes to argue that prose rhetoric irresistibly affects its
audience in an emotional manner, he uses the paradigm of poetry to
illustrate his point (D24=B11 §8–10): affective or symbolic prose, such as
17
Trépanier (2004); Bryan (2012); Tor (2017). For Empedocles it has been harder for philosophical
interpreters to ignore literary matters, but for particularly important applications note Graham
(1988); Primavesi (2001), (2005), (2008a); Picot (2007), (2008a), (2008b); Picot and Berg (2015),
(2018); Rashed (2018) (collecting the author’s works on the topic), as well as n. 27 below.
18
Brooks (1947) 192–214. Lamarque (2009a) 46 applies the principle to distinguish poetry from
philosophy, on the grounds that no serious philosophical argument demands a unique phrasing.
Lamarque’s contrast is taken up by many contributors to a recent volume on the philosophy of
poetry, Gibson (2015). Richards draws roughly the same distinction between ‘emotive utterance’ and
‘scientific statement’ (1970) 58.
19
See the commentary of Kahn (1979) passim and, more recently, Vieira (2013).
the fragments of Heraclitus, is prose that is especially poetic. Poetry is
habitually distinguished from prose, but in the sense of ‘poetry’ I am using,
this is a false dichotomy. It would be more precise to speak of poeticity,20
a quality that can be present in varying degrees in prose or verse, but which
is more characteristically associated with the latter. In contending that
these authors were poets, I mean that their work has a high degree of
poeticity, a point that is not self-evident from the mere fact that they used
verse.
This claim would perhaps be an obvious point if we were dealing with
later poetry,21 but it differs in focus from standard discussions of the
Presocratic use of verse which explain the phenomenon in terms of wider
cultural factors. Thus, it has been held that they chose the medium simply
because prose did not exist at this stage as a viable means of widely
publicising content, or because Ionian prose had not yet reached the
Western Mediterranean, or because they wished to supplant the authority
of Homer and Hesiod.22 All these theories may well be valid. We know of
a few prose books in the sixth century that may have been around at the
time of Xenophanes and certainly would have been by the time of
Parmenides and Empedocles,23 but publicly performed verse-texts that
were relatively easily memorised and transmitted could reach a far wider
audience. The three poets did not operate independently of one another –
there are close intertextual echoes of Xenophanes in Parmenides and of
Parmenides in Empedocles – so that, in covering this subject matter in this
medium, they may have constituted a particular tradition, one that was
distinctive to the locale of Sicily and Southern Italy.24 And they all feature
allusions to Homer and Hesiod that seem to assert a correction of, or
a progression over, those canonical predecessors.25 But these explanations
20
I take the term from Attridge (2019) 3.
21
Cf. the abundant treatments of Lucretius as poetry, rather than purely as philosophy, of which West
(1969) and Gale (1994) are worth singling out for their discussion of the interaction between his
doctrinal content and unique form.
22
No prose alternative: Havelock (1963) 294–5, 308 n. 38, (1983); Long (1985) 245; Osborne (1998);
Western way: Wöhrle (1993) 179–80; supplanting authority: Jaeger (1947) 93; Detienne (1996) 130–7;
Wright (1998) 6; Nightingale (2007) 190; Lesher (2008) 475–6 (who, however, acknowledges
a plurality of possible functions). On the practicalities of literary production for the Presocratics,
see Bernabé (1979); Patzer (2006).
23
I.e., the prose works of Anaximander, Anaximenes, Pherecydes of Syros, Hecataeus and Heraclitus;
possibly also the prose theogony of Akousilaos of Argos (see R. Fowler [2013] 623–4). On the
Presocratics and early prose, see Laks (2001); Kahn (2003) 143–55; Patzer (2006) 95–158; Granger
(2007) 412–17; and on early Greek prose more generally, Lilja (1968). For the dates of the
Presocratics, see the relevant discussions in KRS.
24
See Chapter 1 n. 119 and Chapter 3 Section 3.1 below.
25
See the Presocratic scholarship cited at n. 13 above.
tell us little about the intended or actual responses of audiences. On the
other hand, one group of interpreters who have taken this aspect seriously
are those who, rejecting the philosophical label altogether, have argued that
Parmenides and Empedocles used verse in order to induce a ‘mystical
experience’ in their auditors.26 I share the ambition to historicise the
experiences of audiences, but as we shall see in Chapters 2 and 3, the
evidence suggests a more mainstream performance context for these poems
by rhapsodes at public festivals. The question of what responses these texts
were designed to elicit is one that has rarely been asked and, consequently,
has not been fully addressed.27
Moreover, the treatment of these authors as ‘poets’ and their work as
‘literature’ would appear to some to commit an egregious and problematic
anachronism.28 Certainly, as we are often reminded, what we habitually
refer to as poetry during this period is more aptly labelled ‘song’, desig-
nated with vocabulary such as aoide, humnos and their cognates and, at least
in many cases, vocalised melodically to an instrumental accompaniment.29
It is performed before an audience, often in particular, ritualised circum-
stances, a very different phenomenon from the later practice of the private
reading of poems.30 The distinction is conceptual as well as practical:
Greek songs, at least up until the latter part of the fifth century, tend to
be evaluated in terms of their social utility and truthfulness; ‘poetry’ and
‘literature’ are closely bound up with notions of autotelic art and of
fiction.31 Xenophanes, Parmenides and Empedocles, in using verse to
present truth claims about the nature of the universe, might appear,
26
Böhme (1986); Kingsley (1995), (1999), (2003); Gemelli Marciano (2008), (2013) relate the verse form
of Parmenides and Empedocles to esoteric cults. A key influence on their approach is the classic
article on Parmenides’ proem, Burkert (1969).
27
This is true even of Empedocles, the most recognised of the three for his literary merits, although
more attention has been devoted to this aspect of his work, especially in francophone scholarship,
starting with the learned but often impenetrable commentaries of Bollack (1965–9), (2003). See also
van Groningen (1971); Gagné (2006); Rosenfeld-Löffler (2006); Gheerbrant (2017) for discussions
of Empedocles’ audience’s experience. On Parmenides, Robbiano (2006) addresses the audience’s
experience but, to my mind, assumes too narrow a view of the hexameter genre. See n. 36 below.
Regrettably, Leopoldo Irribaren’s Fabriquer le monde: technique et cosmogonie dans la poésie grecque
archaïque (Paris, 2018) reached me too late to be taken into consideration in the present study.
28
The suitability of such terms is denied by Goldhill (1999); Ford (2003); and Gemelli Marciano
(2008) 21–7; but for a defence of their usage, see Laird (2006) 25–30 and Maslov (2015) 9–22.
29
It is only with Herodotus, in the later fifth century, that poietes and poiesis come into play. See Ford
(2002) 131–57. On the singing of early Greek hexameter, see West (1981). On the ‘song culture’ of
this period, see the influential studies of Herington (1985) and Gentili (1988).
30
On this distinction, see Ford (2003).
31
See esp. Ford (2002). The assumption of such a process underlies the influential studies of Herington
(1985); Gentili (1988); and Nagy (1989).
Truth, Symbolism and Sublime Perspectives in Early Greek Song 9
above all, to exemplify the differences between early Greek song and later
poetry.
Important though these distinctions are, they should not occlude the
significant points of overlap between Archaic and Classical Greek concep-
tions of song on the one hand and later conceptions of the poetic and the
literary on the other.32 The Presocratic poets may conform to the tendency
of early Greek song to advertise its truth status, but the truthfulness of song
is always complicated by its emotive and immersive qualities,33 features that
would later be conceived as hallmarks of poetry and literature.34 What is
even more challenging to the traditional, philosophical approach to the
Presocratic poets is the possibility that the ‘truthfulness’ of early Greek
song is not restricted to its overt statements: the medium implicitly conveys
a symbolic significance that lurks beneath the surface meaning. My under-
standing of this symbolic quality of early verse and the role it plays in an
audience’s experience warrants some elaboration.
2 Truth, Symbolism and Sublime Perspectives in Early Greek Song

In an oral culture, verse can serve as both a mnemonic and a means of
publicity, so that poet-bards may be ‘masters of truth’, entrusted with
preserving valued information about the past.35 Whether the prominent
association between verse and truth in surviving Greek examples represents
a living continuation of this practice or a fossilised survival from the pre-
literate period, it is one reason for the Presocratics’ use of the medium,
though it does not exclude the possibility that they also chose it for its
affective qualities.36 Some scholars have insisted, on the grounds that
‘fiction’ is an alien concept to this sort of culture, that early Greek verse
32
For recent attempts to identify features of poetry that endure from the earliest surviving Greek texts
to modernity, see Culler (2015) (specifically on lyric) and Attridge (2019) (more generally).
33
As Halliwell (2011) 36–92 argues at length of Homer.
34
For immersiveness in both ancient Greek poetics and modern conceptions of the literary, see Allan
et al. (2017); note also Halliwell (2002) on the related concept of mimesis. For emotional affectivity,
see e.g. Verdenius (1983) 46–53; Laird (2006) 26–30; Halliwell (2011) passim; Attridge (2015) 259–79,
(2019) 11–54.
35
See Detienne (1996); Rösler (1980) 289–93; Puelma (1989) 72–3; Thomas (1992) 113–17; Finkelberg
(1998) 73–88; and, more generally, Finnegan (1977) 208.
36
Puelma (1989) 76–7 sees Parmenides and Empedocles as developing the Hesiodic tradition of
conveying physical and ethical truths in poetry, as opposed to Homer’s historical truth claims.
Robbiano (2006) 35–60 argues that Parmenides’ audience would have expected to hear ‘something
true and of great importance’ (42) of a hexameter poem, but not all passages of hexameter poetry fit
this description (cf. Demodocus’ song of Ares and Aphrodite at Od. 8.266–366; the Homeric Hymn
to Hermes) and, in any case, the association with truth does not rule out stereotypically pleasurable,
emotional and engrossing aspects of the medium. See further Halliwell (2011) 55 and n. 37 below.
can only mean exactly what it states overtly,37 but this is too restrictive. The
‘truth’ of a narrative need not be limited to its bare factual details.38 This
principle is already recognised in Aristotle’s argument that poetry is more
‘philosophical’ (φιλοσοφώτερον) than history since it does not simply report
what happened but tells a story which ideally conforms to the rules of
probability or necessity (Poet. 9. 1451a36–b7). On this view, a poetic narrative
should follow real patterns of human behaviour, such as how a certain type
of person would react to a certain type of situation, and it may achieve this
irrespective of whether it portrays events which actually happened.39 In other
words, the truth of a narrative may reside not exclusively – or even not at all –
in its correspondence to actual events, but may also encompass the accuracy
of the patterns of behaviour it instantiates.
There are good reasons for thinking that this sort of conception of poetic
truth was familiar to the Archaic period. In the Works and Days, the narrator
promises to declaim ἐτήτυμα, ‘true things’, to Perses (Op. 10) then proceeds to
recount two myths – of Prometheus and Pandora and of the Races – which are
ostensibly incompatible with one another, but which contain deep intercon-
nections on a thematic level: both are occupied with work and justice, both
describe a fall from a better state to a worse and warn of the human potential
for disaster.40 It is left for Perses and the audience to work out how these
stories fit together, a process that mirrors on an intellectual level the physical
labour that must be performed to access the means of life that is so promin-
ently ‘hidden’ in the poem (Op. 42). Even if we assume that, on some level,
Hesiod believed in the historicity of these stories, their ‘truth’ extends, beyond
their factual content, to the wider themes they convey.41
37
Most (1999) 339, 342–3; Gemelli Marciano (2008) 25–6. Finkelberg (1998) recognises that truth and
fiction are not necessarily mutually exclusive (21) but argues that early Greek hexameter poetry was
overwhelmingly valued for its historicity. For criticisms of this view, see Pratt (1993); Bowie (1993);
and Halliwell (2011) 36–92.
38
Pratt (1993) 1–9, 36–7; Halliwell (2011) 7. In modern philosophy, Lewis (1978) influentially argued
for the applicability of the concept of truth to aspects of fictional worlds (such as its themes or
morals) that go beyond the strict statements of the text and stimulated a range of responses which,
though critical of the details of his argument, still broadly accept this principle. See e.g. Pavel (1986)
esp. 144–5; G. Currie (1986), (1990) 52–91; Gibson (2009) contra e.g. Doležel (1980), (2010) 41–4.
39
This point is unaffected by the dispute over whether Aristotle here assumes that poetry should
produce new knowledge of these ‘universals’ in the audience (thus Halliwell [2002] 193–206, partly
on the basis of 4.1448b4–1449a31) or whether it should appeal to knowledge the audience already
possesses (thus Heath [1991] 399; see also Heath [2009]).
40
B. Currie (2012) 53; Canevaro (2015) 151. Ledbetter (2003) 51 makes a similar point on the didactic
function of the Theogony. Krischer (1965) 172–3 and Halliwell (2011) 18 n. 37 argue that ἐτήτυμος
differs from ἀληθής in referring to normative or didactic truth rather than narrative accuracy.
41
On such ‘abstracted thematic content’ as inherent to poetry, see Lamarque (2009a) 39–50.
Theorists have sometimes complained that such ‘thematic’ truths pur-
portedly conveyed by literature are often trivial or banal when spelled
out.42 One waggish commentator suggests that this approach reduces
Pride and Prejudice to the simple statement that ‘Stubborn pride and
ignorant prejudice keep attractive people apart.’43 An analogous complaint
would be applicable to Hesiod if we were to suggest that the essential
purpose of the myths is just to claim that life is tough and we should be just
and reverent to the gods. But of course, this is not all that they do. Whilst
some of these ‘thematic truths’ may be uncontroversial, it is hard to boil
them down to a definite set. Hesiod’s myths say something about justice
and man’s relationship to the gods, but also something about gender
relations, about the belatedness of mortals’ current situation, about
where the responsibility lies for mortal tribulations and so on. The myths
serve a specific rhetorical purpose (to persuade Perses to act justly and
work), but they do not fulfil that purpose simply by encoding a singular
‘moral’ that the addressee is to identify. Perses is not supposed to extract
a finite number of thematic truths after which he may safely forget the
myths. Their indeterminacy is part of the point: he is to mull them over,
exploring the various possible senses they may convey and their relevance
to his situation. The same could be said of the parenetic myths in the Iliad.
The story of the Litai and Meleager (9.434–605) is not reducible to the
simple point that Achilles should accept Agamemnon’s reparations, and
the myth of Niobe (24.699–720) does not only make the banal point that
Priam, though he grieves, still needs to eat; it invites reflection on the
vicissitudes of mortal blessings and the endurance of human resilience in
the face of overwhelming calamity, in light of which Priam is to change his
behaviour. In this sense, the myths are symbolic rather than allegorical.
They do not have a definite tenor to be decoded; their symbolic signifi-
cance is to be explored and discovered.44
These specific examples of embedded myths are instances of the way in
which Muse-inspired hexameter poetry in general was often assumed to
harbour such symbolic meanings. Its characteristic content is mytho-
logical, and myths occupy a space outside the limits of contemporary
mortal experience, which is why their mortal narrators typically require
the assistance of divine Muses. They take place in the remote past, prior to
42 43
See Stolnitz (1992) and the overview of Lamarque (2009b) 238–9. Stolnitz (1992) 193.
44
For this distinction between allegory and symbol, see Norman Friedman’s entry for ‘Symbol’ in the
Princeton Encyclopedia of Poetry and Poetics (4th ed. 2012), and the discussions of Gadamer (2004)
61–70, (1986) 31–9 and de Man (1983) 187–208. For ancient practices of symbolic interpretation,
which go back at least as far the sixth century bce, see Struck (2004).
recorded memory, or they concern the behaviour of the gods who inhabit
the inaccessible locations of Mount Olympus or Tartarus and who interact
with the familiar world in a manner that is normally invisible to mortals.
The details of these stories were consequently up for debate. Early hexam-
eter poets could baldly deny rival mythological accounts (e.g. hHymn 1.1–7)
or reject them on a priori grounds (e.g. Hes. Op. 11–12).45
Equally, interpreters of hexameter poems could debate the relevance of
the myths to present-day realities, applying elaborate explanations as to
their true significance.46 Pherecydes of Syros (fl. c. 544 bce), in a testimony
that derives from the second-century ce philosopher Celsus, is said to have
understood Zeus’s threatening words to Hera (Il. 1.590–1, 15.18–24) as the
creator-god imposing order on matter (R27<B5).47 According to Porphyry
(ap. Schol. B ad Il. 20.67), Theagenes of Rhegium defended Homer against
accusations of impiety on the grounds that the Theomachia does not in fact
represent a battle of anthropomorphic deities, but rather a cosmology of
opposing elements, in which the dry battles the wet, the hot the cold and
the light the heavy.48 Late though these sources are, they present a picture
of sixth-century bce interpretative practices which roughly coheres with
the impression we get of the better-evidenced fifth and fourth centuries.
Aristophanes, Plato and Xenophon attest to conceptions of poetry as
ainigmata, ‘riddles’, that contain hypo- or dianoiai, ‘hidden meanings’.49
One of the most famous such interpreters, Metrodorus of Lampsacus, is
reported as having understood Homer’s characters as representing the
celestial bodies: Achilles is the sun, Helen the earth and Hector the
moon (Tatian 21=DK61A3), so that Achilles’ pursuit of Hector around
the walls of Troy becomes the orbit of the Sun and the Moon around the
Earth.50 But our most extensive Classical-period evidence for the methods
involved in this sort of interpretation comes from the fourth-century bce
Derveni papyrus. This charred scroll preserves a fragmentary commentary,
45
See Feeney (1991) 13. This point undermines the view of Detienne (1996) 49, 52–67 and González
(2013) 201–17, 235–66 that Archaic Greek audiences would have regarded hexameter performances as
‘tautologically true, necessarily accurate’ (González [2013] p. 211). For an argument that epic is
always inherently allegorical (in a broad sense), see Laird (2003) 165–73.
46
For this and the subsequent two paragraphs, I am heavily indebted to the important study of Struck
(2004).
47
Tate (1927) identified the significance of this testimony for Pherecydes as an allegorist. Schibli (1990)
99–100 is more sceptical of that possibility. Struck (2004) 26–7 n. 14 sees Schibli as ‘overly cautious’
on this point.
48
On Theagenes, see also the Epilogue below.
49
E.g. Aristophanes, Pax 38–51; Plato Ion 530d, Resp. II 378d; Xenophon Symp. III. 6.
50
Under the assumption, of course, of a geocentric universe. For this interpretation of Metrodorus, see
Richardson (1975) 69, with further references.
thought to date from the fifth century bce, on an earlier theogony ascribed
to Orpheus.51 The commentator describes the poem as ‘riddling, even if
Orpheus himself intended by means of it to say not undeterminable
riddles, but rather great things in the form of riddles’ (Col. 7.5–6, text
and trans. LM).52 He then proceeds to identify Olympus with Time (Col.
12), and Zeus with the air (Col. 17).
In all these examples, the significance of a hexameter poem is taken to go
far beyond its strict, semantic meaning. Zeus is not just Zeus, and the
battle of the gods is not just a battle. Though these interpreters are
standardly referred to as ‘allegorists’, their conception of poems as ainig-
mata is broader than the later allegorical label implies: ‘riddling’ texts can
convey their hidden meanings through strategies other than allegory in the
strict sense of an extended metaphor. Thus, the Derveni commentator
understands the word ἄδυτον – which ostensibly means ‘shrine’ in the
context of the poem – in its homonymous sense of ‘unsetting’ (Col. 11.1–4;
see LSJ s.v. I 2). The term, by means of a pun, reveals a cosmological point
about Night, who, in the theogony, inhabits the shrine. The commentator
also ingeniously unpicks the name ‘Kronos’ as indicating the ‘mind that
strikes’ or ‘makes things collide’ (Col. 14. 7, κρούοντα τὸν Νοῦν) during
the formation of the cosmos.53 Though these particular interpretations are
hardly convincing to modern readers, the strategy of exploring the deeper
significance of particular terms, through identifying resemblances to other
vocabulary and analysing instances of wordplay, would today be regarded
as a legitimate hermeneutical procedure.54 Hexameter poetry, with its
typically mythological subject matter, seems to invite this sort of attempt
to identify an underlying significance through close analysis of particular
details.
Figures such as Theagenes, Metrodorus and the Derveni commentator
may be extreme cases, whose interpretations are ridiculed or regarded as
implausible by near-contemporary sources, but the general idea that
51
On the background and interpretation of this difficult text, see especially Betegh (2004) and Piano
(2016), along with the collections Laks and Most (1997), Papadopoulou and Muellner (2014) and
Santamaría (2018). The papyrus was first discovered in 1962 and was made available to scholarship at
large in an anonymous, unofficial edition on separately numbered pages in ZPE 47 (1982) (for the
story of this peculiar development, see Burkert [2014] 113–14), but its official editio princeps did not
appear until Kouremenos, Parássoglou and Tsantsanoglou (2006).
52
αἰνι̣ [γμ]ατώδης, [κε]ἰ [Ὀρφεὺ]ς̣ αὐτ[ὸ]ς̣ [ἀό]ριστ’ αἰν[ίγμα]τα οὐ̣ κ ἤ̣ θελε λέγειν, [ἐν αἰν]ίγμασ̣[ι]ν δὲ
[μεγ]άλα.
53
On this cosmological process, see Betegh (2004) 252–7.
54
On how the practices of the commentator compare with later literary criticism, see Henry (1986);
Struck (2004) 29–39; Betegh (2004) 132–5; Bierl (2014).
hexameter poems communicate profound truths in a manner that goes
beyond their straightforward semantic meaning was widespread in the
Classical period and already implicit in the poems themselves.55 In the
much-discussed Muse-inspiration scene of Hesiod’s Theogony, an episode
that would be adapted by Xenophanes and Parmenides,56 the Muses
‘inspire’ (ἐνέπνευσαν) a ‘divine voice’ (αὐδὴν/ θέσπιν) in the singer so
that he may glorify ‘the things that shall be and the things that were before’
(τά τ᾽ἐσσόμενα πρό τ᾽ἐόντα), and order him to sing of the ‘race of blessed
ones who always are’ (μακάρων γένος αἰὲν ἐόντων, 31–3). The expression
here translated as ‘divine voice’ implies that they do not merely affect the
content, but also the sound and timbre of the utterance,57 in such a manner
as to render it more appropriate for the contemplation of the eternal
entities that will form Hesiod’s main subject matter.58 Regardless of
whether we should imagine the Muses as telling Hesiod truths, lies like
the truth, or some combination of the two (Theog. 27–8),59 the mortal
narrator presents his speech as prophetic – he is to glorify not only events of
the past but also ‘things that shall be’, an expression elsewhere used of the
Iliadic prophet Calchas (Il. 1.70)60 – and divinatory speech, by such
strategies as ambiguity, vagueness and metaphor, characteristically
demands a penetrating interpretation to uncover its truths.61 In this
respect, the deeper truths of divinatory speech resemble the gods them-
selves, who can never really be understood with certainty by mortals or
adequately encapsulated in everyday language.62 The ‘divine’ quality of the
55
The passages cited at n. 49 are all critical or satirical of this type of interpretation, but see Struck
(2004) 39–50, who cites several Socratic dialogues that indicate how widespread it was.
56
See the discussion of Xenophanes D50=B35 at Chapter 1, Section 2a and of Parmenides
D8.65=B8.60 at Chapter 2, Section 3a below.
57
On these lines see Lanata (1963) 26–8. On the vocabulary see LfgrE s.v.; West (1966) ad loc.; and Ford
(1992) 184–5. The expression is also used of the bards in the Odyssey (1.328, 8.498).
58
See Ford (1992) 185 and Halliwell (2011) 67.
59
On this perennial interpretative crux, see the recent and thorough discussion of Tor (2017) 61–103.
60
Cf. also Pind. Pae. 8.83–5=fr. 52i. Nagy (1989) 23–9 argues, largely on the basis of the Hesiod passage,
that poetry and prophecy were indistinguishable in a period prior to our surviving texts, an idea that
has some ancient precedents (e.g. Paus. 10.5.7; cf. Pl. Ion 534b). See, similarly, Dodds (1951) 81;
González (2013) 284–8. Finnegan (1977) 207–10 criticises the once-popular view that all oral poets
are essentially seers or prophets.
61
On these strategies, see Manetti (1993) 24–9. This feature is more famous from oracular, rather than
prophetic, speech, above all in the hexameter oracles recorded by Herodotus, but as McNelis and
Sens (2016) 22 point out, prophecy and oracle are regularly conflated (e.g. of Cassandra’s prophecies
at Aesch. Ag. 1178 and Lycoph. Alex. 6).
62
On inscrutability of gods as a fundamental feature of religion (in spite of confident assertions of
insiders to the contrary), see Burkert (1996) 4–8. For neat expressions on the inscrutability of the
gods in antiquity, note Protagoras D10=B4; Ptolemy Syntaxis 1.1. See also Parker (1999) on
Sophocles. On the difficulties of encapsulating the divine in mortal language, note Price (1984)
80. The imperfection of mortal language may also be implicit in the traditional notion that the gods
singer’s voice thus points to its capacity to connect the audience with that
which would normally be ineffable. Hesiod’s song implicitly opens
a channel to the divine, eternal entities it concerns, in a society where
gods, though latently ubiquitous, were rarely thought to be encountered
directly.
I shall argue that Xenophanes, Parmenides and Empedocles conform to
this conception of the ‘truth’ of Archaic Greek poetry as something that
extends to its range of symbolic significance and is embodied in its form as
well as its semantic content.63 If this is correct, the use of verse performs at
least two uncontroversially philosophical functions that would be over-
looked by an exclusive emphasis on explicit propositional content. First,
verse form, in combination with the other assorted resources of poetry,
implies claims that are not made overtly; and secondly, it serves to draw the
audience into a process of enquiry, for instance, by inviting reflection on
the true significance of the words uttered and their relationship to their
referents.
More provocatively, there is a further way of thinking of the ‘truth’ of
Archaic Greek poetry that transcends its propositional content, whether
overt or implied. It is sometimes held that poetry or literature, as a form of
art, is essentially concerned with a kind of truth that is sui generis and
cannot be expressed in propositional form. This idea is familiar from the
Romantic notion that ‘Beauty is truth, truth beauty.’64 Poetry may stem
from and convey ‘true feeling’, that is, an experience that is ‘true’ in the
sense of ‘sincere’ or ‘authentic’.65 It is worth noting that one influential
version of this view is indebted to the Presocratic poets themselves. Martin
Heidegger argued that art is ‘a becoming and happening of truth’ and the
‘essence of art is poetry’.66 He conceived of truth, following the etymology
of aletheia, as ‘unconcealment’, a conception that was inspired in large part
by his reading of Parmenides.67 Heidegger’s symbolic interpretations of
the Presocratics are of limited value for a study such as this one which aims
speak a separate language from humans, on which see Versnel (2011) 389 n. 26 and Chapter 2, n. 19
below.
63
This argument is given more support if a recent suggestion by Richard Janko (2016) is correct, that
the Derveni papyrus quotes Parmenides D4=B1.1. The suggestion, however, is highly speculative,
since it rests on the identification of only seven letters.
64
Cf. Wordsworth’s ‘poetry is the first and last of all knowledge’ in the 1802 preface to Lyrical Ballads,
Wordsworth and Coleridge (2005) 302.
65
Thus, recently, Scruton (2015) 157–8, who also identifies this view in the work of T. S. Eliot and
F. R. Leavis. Cf. Eliot’s famous criticism of Hamlet on the grounds that it lacks an objective
correlative that authentically conveys Hamlet’s emotions (1920).
66
Heidegger (2002) 44, 47. 67 See Heidegger (1962) 256–7, (1992).
at a historical understanding of the ancient philosophers,68 though it
should be acknowledged that the approach he inaugurated was a key
influence on Mourelatos’ Route, one of the few interpretations to take
Parmenides’ forms and imagery seriously. Furthermore, his conception of
truth as a mysterious revelation seems closer than the truths of propos-
itional logic to the extraordinary, emotive utterances of divinely inspired
Archaic Greek song.69
Now, some may object that experiences of literature are neither true nor
untrue in any useful sense of the terms – Frege, for instance, famously
argued that poetry and fiction lack truth value on the grounds that their
function does not depend on their accurate reference to actual facts.70 Or it
may be the case that authentic feelings can in fact be adequately expressed
as propositions after all.71 It is far beyond my scope to pronounce a verdict
on these issues. But however we ought to conceive of truth in general, there
is one common conception of sui generis literary truth that seems particu-
larly applicable to early Greek poetry. This is the idea that truth is an
essential value of literature, in the way in which it exposes a particular
vision of the world.72 According to this theory, we do not necessarily learn
that the factual details or thematic truths of a literary text are valid, but we
learn what it is like to see the world in a certain way. As Hilary Putnam has
written of Celine’s Journey to the End of the Night, ‘I do not learn that love
does not exist, that all human beings are hateful and hating . . . What
I learn is to see the world as it looks to someone who is sure that hypothesis
is correct.’73 Following this line of thought, in the Works and Days, Perses is
shown that it is possible to view the world in such a manner that his laziness
and deceptiveness constitute a lack of foresight that is likely to be punished
in the long term. As a result of being granted this experience, his repertoire
of ways of looking at things is expanded and he may end up changing his
68
See Detienne (1996) 26–8; Most (2002); and now Lecznar (2020) 117–29, who charts the relationship
between Heidegger’s use of the Presocratics and his Nazism.
69
See Cole (1983) on the relationship between Heidegger’s understanding of aletheia and its sense in
Archaic Greek texts. On the ethical character of Greek poetic ‘truth’, see Detienne (1996) 49, 52–67;
Robbiano (2006) 43–8; González (2013) 201–17, 235–66. Though I am sceptical of their assumption
that Archaic Greek audiences would necessarily have granted the authority arrogated by poets in
Muse-inspiration scenes (see n. 45 above), such scenes, at least within the fiction of the poetry,
clearly imbue the statements uttered with a symbolic and religious character.
70
Frege (1960) 62–4. For criticisms of this argument and an overview of its legacy in literary
scholarship, see Doležel (1998) 3–5.
71
For comments along these lines, note Lamarque (2009b) 239–48.
72
E.g. Murdoch (1970) 34; Putnam (1978) 83–94; Novitz (1987) 135; Gibson (2009); Koethe (2009) 58;
G. Currie (2010) 88–93. See Lamarque (2009b) 239–48 for an overview (though I conflate here two
categories he distinguishes as ‘vision’ and ‘learning what it is like’).
73
Putnam (1978) 89–90.
own habitual attitudes. Of course, this visual language of ‘seeing’ or
‘viewing’ from a certain ‘perspective’ is metaphorical for the essentially
cognitive processes of ‘interpreting’ or ‘understanding’, but such concep-
tual metaphors are as prevalent in Greek as they are in English, for instance,
in the fact that οἶδα is in origin a perfect tense of a verb of seeing and in the
fact that νοέω, in Homer, usually refers to a perceptual process of ‘noticing’
or ‘recognising’ a character or situation.74
This perspectival conception of literary truth fits neatly with the Archaic
Greek emphasis on the vivid, visual quality of verse. Within this context,
vision and truth are intimately related. Before the lengthy catalogue of
ships, the Iliadic poet requests the Muses’ assistance on the grounds that
they ‘are goddesses and present at and know all things’ (ὑμεῖς γὰρ θεαί ἐστε
πάρεστέ τε ἴστέ τε πάντα, 2.485). As is often observed, ‘to know’ some-
thing in Homer is typically (though not always) to have seen it,75 so that
the Muses’ divine knowledge (arising as a consequence of their ‘presence’)
is empirical. The narrator’s aim here, then, is to replicate their knowledge-
able divine perspective on the innumerable Greek ships and warriors
through song. A further example is the Odyssey’s account of the bard,
Demodocus, who implicitly enjoys god-given powers of vision when it is
mentioned that the ‘Muse loved him above all, and she gave him both good
and bad;/ she deprived him of his eyes, but she gave him sweet song’ (Od.
8.63–4): song replaces his ability to see, so that he ‘shows his song’ (φαῖνε δ’
ἀοιδήν, Od. 8.499) and is able to recount, in an uncannily accurate
manner, events of which he has no experience, as Odysseus’ praise confirms
(Od. 8.487–91). A wealth of ancient and modern Homeric criticism regards
this visual quality as extending to the audience’s experience of the song:
they are made to feel like spectators.76 It is because of this immersive
quality that song can cause its audience to forget their immediate concerns
(Theog. 80–7, 98–103).
74
See studies of von Fritz (1943), (1945), (1946) and Lesher (1981) 8–16. The metaphor is also implicit in
the tendency for blindness to be compensated for by enhanced insight in Greek myth (most notably
in Teiresias and Oedipus), on which see Buxton (1980). On ‘Understanding/Knowing is Seeing’ as
a pervasive conceptual metaphor, at least in English, see Lakoff and Johnson (1980) 48, 103; Kövecses
(2010) 256–7.
75
The key study is Snell (1924), though he overstates his case. See Lesher (1981) 8–15, (2008) 459–64
and Hussey (1990) 11–17.
76
See Ford (1992) 53–4; Bakker (2005) 62–8; Clay (2011b); Graziosi (2013); Allan et al. (2017);
Grethlein and Huitink (2017); Gazis (2018) 1–13. For the term in ancient criticism, see Zanker
(1981); Meijering (1987) 29–52; Nünlist (2009) 194–8. A good example is ΣbT Il.23. 363–72,
discussed by Grethlein and Huitink.
When the best performances of Greek songs enable audiences to ‘see’ the
content, this is an assumption of a ‘perspective’ not only in the visual sense
of the term but also in the sense of seeing the world ‘in a certain way’. So
much is clear from the emotional dimension to this experience, as
described in our most detailed testimony for rhapsodic performance,
Plato’s Ion. The rhapsode Ion assents to Socrates’ suggestion that his soul
believes it is present at the actions it describes (535b–c), resulting in the
display of appropriate emotions: ‘whenever I say something pitiable
(ἐλεινόν), my eyes are full of tears; whenever it is something frightening
or awful (φοβερὸν ἢ δεινόν), my hair stands on end with fear and my heart
leaps’ (535c). The same emotions are then replicated in the audience
(535e).77 The perspective that Ion adopts and exposes to his audience is
not value-neutral, but entails certain attitudes, for instance, that a certain
event is worthy of fear.78 This axiological dimension can shed light on the
rhetorical function of parenetic myths: after hearing Hesiod’s stories of
Prometheus and Pandora and of the Races, Perses will ideally be able to
evaluate his actions from a wider, more global perspective, according to
which work is necessary and dishonest gains count as a lack of foresight
with potentially disastrous consequences. The Presocratic poets, I shall
suggest, deploy verse to impart ‘truths’ in this manner, that is, not explicitly
articulated propositional claims, but ways of experiencing the world which,
from their perspective, are more appropriate to its true nature.
One aesthetic concept which, in some formulations, involves this per-
spectival understanding of literary truth and which, I shall argue, applies to
the intended effects of Parmenides’ and Empedocles’ poetry is the sublime.
Longinus – a name I use here for convenience while remaining agnostic
over the authorship of the literary treatise preserved under it –79 uses visual
language for the experience of both composers and recipients of literature.
He discusses at length the poet’s powers of ‘visualisation’ (φαντασία)
which are conducive to sublimity (§15). More generally, in §35, the sublime
is made possible by the fact that ‘godlike’ (ἰσόθεοι) writers have ‘seen’
(εἶδον) something: namely, that nature has ‘judged’ mortals to be ‘specta-
tors’ (θεαταί) as well as competitors in her games, with a passion for
whatever is great and more divine than ourselves, so that the whole
universe is insufficient for the ‘contemplation’ – or, more literally, the
‘viewing’ (θεωρία) – of human thought and our ideas often transcend the
77
Note similarly Gorgias D24=B11.9.
78
On literature’s perspective as never value-neutral, see Gibson (2009) 481.
79
Heath (1999) argues that On the Sublime is in fact by the third-century ce philosopher Cassius
Longinus, against the general consensus of recent scholarship.
boundaries of the cosmos (§35.2–3).80 In sublime communication, the
listeners share in the emotions felt by the authors who have achieved
a perspective that goes beyond normal mortal limitations (§39.3). For
Longinus, this experience has a veridical dimension: there is no suggestion
that the perspective acquired by great authors is in any way invalid or
fallacious. Indeed, in the introduction to the treatise, he endorses the
saying that we have ‘beneficence’ (εὐεργεσία) and ‘truth’ (ἀλήθεια) in
common with the gods (§1.2–3) and, given that he frequently uses divine
language for great writers, we may assume that they are the ones who most
of all possess these two qualities.81 As Stephen Halliwell has argued at
length, the truth of the Longinian sublime is not, or at least not exclusively,
propositional, since it encompasses direct contact with greatness of mind
(§9), authenticity of the emotions conveyed (cf. §22.1) and the grand
cosmic spectatorship described in §35.82
Longinus’ treatise is of Imperial date, but its conception of divine,
transcendent poetic truth has its roots in much earlier poetic and rhetorical
practice.83 It is arguably an expansion and elaboration of the powers of song
already vaunted in Hesiod’s Theogony where, as we saw, the Muses do not
simply affect the factual content of the verse but also its form and emo-
tional impact. A neat expression of the grandeur of contemplation that
would come to be associated with the sublime is made by the Hellenistic
philosopher-poet, Cleanthes:
τοῦ λ̣[ό]γ̣ου τοῦ τῆς φιλοσοφίας ἱκανῶς̣ μὲν ἐξαγγέλλειν δυναμένου τὰ θεῖα
καὶ ἀνθ[ρ]ώ[πι]ν̣[α], μὴ ἔχοντ̣ος δὲ ψειλοῦ τῶν θείων μεγεθῶν λέξεις οἰκείας,
τὰ μέτρα καὶ τὰ μέλη καὶ τοὺς ῥυθμοὺς ὡς μάλιστα προσικνεῖσθαι πρὸς τὴν
ἀλήθειαν τῆς τῶν θείων θεωρίας
Though the discourse of philosophy has a sufficient capacity to make known
divine and human affairs, since it is stripped down, it does not possess the
aspects of speech appropriate for divine grandeurs, that is, epic and lyric verses,
as well as the rhythms which above all approach the truth of contemplation of
divine matters. (Cleanthes SVF 1.486 ap. Philodemus de musica 4 Col. 142)84
80
For τοῦ περιέχοντος here as ‘the boundary’ of the cosmos, see Halliwell (2011) 343 n. 35.
81
The saying is variously attributed to Pythagoras (Ael. VH 12.59), Demosthenes, Aristotle and others
(Sent. Vat. P.26 Sternbach). For divine language of great writers, in addition to the example at §35.2
note §4.6; §18.1; §32.5; §33.5; and the comparison with the Pythian oracle at §13.2. For truth as an
aesthetic quality in Longinus, note also the approving quotation of Pl. Resp. 9.586A at §13.1.
82
See Halliwell (2011) 343–67. Porter (2016) 76 argues that that this view ‘naively’ overlooks Longinus’
emphasis on concealed literary artifice, but the two seem to me to be compatible: literary artifice may
bring about the three effects Halliwell outlines.
83
On the long prehistory of the features Longinus identifies as Sublime, see Porter (2016).
84
Translation adapted from the French of Delattre (2007), who defends these interpretations of
παραδείγματα and τὰ μέτρα καὶ τὰ μέλη ad loc.
For Cleanthes, it is not merely the semantic content of the verse that is
meaningful, but also its metres and rhythms; it does not simply provide an
accurate account of the gods, but somehow reaches towards a ‘truth’ which
here seems to refer not so much to the particular theological claims made as
to the experience of ‘contemplating’ or ‘viewing’ them, something which
could not be replicated by philosophical prose.85 As I hope to show in
subsequent chapters, this description would be peculiarly apt for
Xenophanes, Parmenides and Empedocles. They do not use verse simply
to communicate a finite number of doctrinal claims, but to provide an
experience of what it feels like to view the world from a particular, value-
laden perspective, one that transcends normal mortal limitations and
exposes an otherwise unrecognised spectrum of truths.
3 Poetics
The process of bringing these texts further into the fold of Greek literary
studies helps to generate new interpretations of particular details but can
also affect our understanding of the broader history of Greek poetics.
Xenophanes’ criticisms of Homer and Hesiod (D8–9=B11–12) occupy an
important place in any account of the topic, but studies tend to treat them
briefly and in isolation from his wider cosmological and epistemological
claims.86 Chapter 1, accordingly, offers a reassessment of these criticisms in
the context of Xenophanes’ wider philosophy, arguing against the view
that they stem from a literal-minded failure to appreciate any value in
made-up stories. Rather, when read within the context of his comments on
the gods, human knowledge and human progress, it appears that it is not so
much Homer and Hesiod’s inaccuracy as their morally pernicious influ-
ence on society that is the problem. But in addition to this well-known
example, the Presocratic poets’ cosmological, theological and epistemo-
logical doctrines have further implications for the theory and practice of
poetry. The goddess of Parmenides’ poem apparently espouses a strict
monism which renders inaccurate almost all human language.87 These
linguistic consequences are underlined by her emphasis on the mortals’
act of naming in forming their erroneous opinions (D8.58=B8.52;
D13=B9.1; D62=B19.3). If mortal language is erroneous in this way, then,
85
Cleanthes’ comment is strikingly similar to those of modern scholars cited at n. 58 above.
86
E.g. Atkins (1934) 13–15; Grube (1965) 8–9; Pfeiffer (1968) 8–9; Russell (1981) 19, 87–8; Nagy (1989)
34–5; Too (1998) 119–20; Halliwell (2011) 10. Svenbro (1984) 79–100 and Ford (2002) 46–66 are
exceptions, with more substantial treatments of Xenophanes.
87
See Graeser (1977) 362–5; Kraus (1987) 81–97; Mason (1988).
Poetics 21
a fortiori, so is mortal-authored earlier Greek poetry. Empedocles, on the
other hand, valorises poets, placing them at a high grade in the cycle of
reincarnations (D39=B146). The fact that these authors themselves employ
the medium of verse should encourage us to explore these consequences of
their philosophy for poetry in general.
More subtly, Xenophanes, Parmenides and Empedocles imply views
about the nature and function of poetry through their own poetic practice.
‘One function of the poet at any time is to discover by his own thought and
feeling what seems to him to be the poetry at that time. Ordinarily he will
disclose what he finds in his own poetry by way of the poetry itself.’88 We
need not exclude the Presocratic poets from this principle simply because
they also belong to the history of philosophy. A nuanced exploration of the
significance of the forms, images and vocabulary deployed, within the
context of the Greek literary tradition, can reveal the ‘immanent’ poetics
of these texts.89 When Parmenides’ goddess refers to her account of mortal
opinions as a ‘deceptive kosmos of my words’ (κόσμον ἐμῶν ἐπέων
ἀπατηλόν, D8.57=B8.52), she uses a conventional expression that is else-
where used to mark speech as sung verse.90 These parallels suggest that
Parmenides’ use of the phrase has poetological implications: the world of
deceptive mortal opinions is presented as a world of poetic artistry; con-
versely, poetry itself is implicitly painted as the product of opinion.
Empedocles’ poetry is a rich source of poetological imagery.91 We have
already seen that his song is presented as a purifying liquid, but there is also
further evidence for the sort of response that his poetry is designed to elicit.
His narrator responds to the progress of the cosmic cycle in aesthetic terms,
when describing the creation of mortal things as a ‘wonder to behold’
(θαῦμα ἰδέσθαι, D75=B35.17), a formulaic expression which elsewhere is
almost always used of products of craftsmanship.92 The expression there-
fore illustrates that in Empedocles’ poem, the perceptible world itself is the
product of craftsmanship, designed by Love and Strife at different stages in
the cycle.93 The simile of the painters (D60=B23), in which the formation
of a pluralistic world from only four elements is likened to the process of
painting a variety of objects with only a limited number of paints, draws an
88
Stevens (1951) vii, also quoted by Ford (1992) 3–4.
89
This expression in this sense was, I believe, coined by Blumenberg (1966); it has been applied to early
Greek literature by Nünlist (1998) 10, and specifically to the poetry of Parmenides and Empedocles
by Most (1999) 335–6.
90
Esp. Solon fr. 1.2 W and Democritus D221=B21, but cf. also Od. 8.489, Bacch. 59, Simon. 11.23 W,
22.6 W, Pind. Ol. 11.13, fr. 194.3, OF 25.
91
Note the frequent references in Nünlist (1998).
92 93
See the discussion at Chapter 3, Section 3 below. See Sedley (2007) 31–74.
analogy between the perceptible world and artistic production. This ana-
logy between artwork and the cosmos, coupled with the fact that the
cosmos itself is Empedocles’ main topic, encourages us to see his text as
the product of craftsmanship, and as having the same wondrous effect as
the phenomena it describes.
It is surprising, then, that Parmenides and Empedocles are omitted in
most accounts of the history of Greek poetics, whilst Xenophanes is given
only cursory treatment.94 The Epilogue of this book sketches out in brief
some striking instances of direct influence by the Presocratic poets on the
literary ideas of certain crucial fifth- and fourth-century bce thinkers on
the subject, in particular, Metrodorus of Lampsacus, Antisthenes, Gorgias,
Democritus and Plato. In view of these instances, I shall argue that the
Presocratic poets were instrumental in the movement towards an explicit
articulation of what may be termed the Classical conception of poetry,
according to which poems are self-contained artefacts, crafted by mortals,
that present true claims only symbolically. The seeds of such a notion are
already traceable in Archaic song, in examples of hexameter ecphrasis and
in Simonides’ famous comparison between poetry and painting (ap. Plut.
Mor. 346 f). But it is only with Plato that we find an explicit and systematic
argument that both verbal texts and visual arts are united in offering
a depiction of the world. Similarly, the notion that the poet can encode
true claims symbolically in the text is implicit in several passages of early
Greek poetry, such as the paradigmatic myths in Homer. But it is only
later, in the context of allegorical commentators, that (as far as we can
reconstruct their practices from surviving fragments and the charred
remains of the Derveni papyrus) we find an explicit statement of the idea
that the poet encodes fundamental truths throughout the text and it is the
task of the interpreter to recover them.95 The Eleatic claim that the objects
of secure knowledge are fundamentally unchanging facilitated the sugges-
tion that poetry symbolically represents, and should be judged in relation
to, such objects.
94
Of the studies cited at n. 86 above, only Ford (2002) briefly mentions Parmenides and Empedocles,
in respect of the role played in the movement towards a conception of materialism, that paved the
way for Democritus and Gorgias’ materialist conceptions of song and poetic inspiration (161–5).
Maehler (1963) and Ledbetter (2003) make no mention of the Presocratics. Heath (2013), a concise
survey of ancient philosophical poetics, features brief references to Xenophanes and Parmenides
(6–8), but largely begins with Plato. The selection, with commentary, of Lanata (1963), and the
introductory discussion of Harriott (1969) 65–7, are exceptions to this trend.
95
Cf. Col. 7.5–7 of the Derveni papyrus, where Orpheus is said to have ‘wanted to say great things in
riddles’. On the emergence of this type of interpretation during this period, see Struck (2004) 21–53.
Poetics 23
In investigating the role of these texts in the history of poetics, then, the
narrative presented here is, like that of Geistesgeschichte, a diachronic,
developmental one, examining Presocratic ideas of poetry that would
lead up to the famous theories articulated by Plato. The term
‘Presocratic’ has its faults, in implying a teleological narrative culminating
in the triumphant achievement of Socrates, in being used, misleadingly, to
refer to figures who were contemporaneous with or younger than Socrates,
and in grouping together disparate thinkers who may not have been aware
of one another’s work whilst excluding others (such as Hesiod or the
Hippocratic authors) who may have equally legitimate claims to the title
of philosophy.96 I use it here primarily in reference to these three authors,
since, as we have noted, there are good grounds for treating them together
aside from the fact that they all wrote in verse. The term is also fitting as
I shall argue that the ideas about verse embodied in these texts can be seen
as important influences on those that are put into the mouth of Socrates by
Plato. But that is not to say that this story is a teleological one. The
Presocratics are not considered as points on a steady trajectory from
mythological to rational explanations, nor is it assumed that developments
in Greek poetics or philosophical expression were always necessarily super-
ior to what came before. It is outside of my scope to explain why matters
developed in the way that they did: my focus is on the ‘what’ and the
‘how’.97 The ‘Poetry and Poetics’ of my title is also mildly polemical, since
it presupposes that these words are applicable to authors who were active in
the song culture of Archaic and Classical Greece, before terms such as
poietes and poiesis caught on.98 Though the Presocratic poets may seem to
be the authors who epitomise the otherness of Greek song to later concep-
tions of poetry, the present study argues that, in their use of this medium,
they are closer to us than they may appear.
96
For an excellent discussion of the history of this label, its drawbacks, and its merits, see Laks (2018)
19–34; for further problems with it, see Lloyd (2006).
97
My interests therefore differ from those who seek to explain the rise of Greek philosophy or
‘rational’ thought primarily as a consequence of one or other wider historical development, be it
literacy (thus, Havelock [1963]; Goody and Watt [1963]; Goody [1977]), the rise of the polis (Vernant
[1982]; Lloyd [1987] esp. 78–83), architecture (Hahn [2001]), or monetisation (Seaford [2004]).
98
See n. 29 above.
chapter 1
Xenophanes
1 Introduction
Xenophanes is a pivotal figure in our understanding of ancient Greek
intellectual culture. On any account, he occupies an important place
in the history of poetics. In his works, we find probably the earliest
surviving mention of the names of Homer and Hesiod outside of the
latter’s self-reference.1 His criticisms of the Homeric and Hesiodic
depictions of the gods have traditionally been regarded as the ‘first
shot’ in the ‘ancient quarrel between poetry and philosophy’ made
famous by Plato (Resp. 10.607b–c).2 Such criticisms have often been
seen as provoking Theagenes of Rhegium to write the first book on
Homer, outlining allegorical interpretations in defence of the poet.3
Moreover, Xenophanes’ ‘attack’ on the poets, and in particular on
anthropomorphic depictions of gods, has been read as a philosophical
rejection of myth, and thus a crucial instance in the historiographical
narrative of a Greek progression from ‘mythological’ to ‘rational’ forms
1
Other contenders for the earliest mention of Homer are Callinus fr. 6 W=Paus. 9.9.5 (not a verbatim
fragment), ps.-Hesiod fr. 357, Simonides fr. 564 PMG and Heraclitus D21=B42, D22=B56,
D24=B105. The former pair are hard to date because of their possibly spurious nature.
Xenophanes’ apparently lengthy life means that we cannot rule out that his references to Homer post-
dated those of his younger contemporaries Simonides and Heraclitus, especially given that
Xenophanes seems to refer to the former in D68<B21. For Hesiod, the only other contender
(discounting Hesiod’s reference to himself) is Heraclitus D25a=B57.
2
‘First shot’: Kahn (2003) 154; see similarly Babut (1974b) 116–17; Heath (2013) 6–8. Kannicht (1988)
takes a broader perspective, regarding the quarrel already implicit in Hesiod’s Theogony. Nightingale
(1995) 60–7 and Murray (1996) 231 argue that the quarrel was invented by Plato, a view criticised by
Halliwell (1997) 455–6. See Edmundson (1995) and Barfield (2011) (both of whom begin with Plato)
for wide-ranging studies of the ‘quarrel’ throughout the Western tradition.
3
Pfeiffer (1968) 8–11; Russell (1981) 18–19; Heitsch (1983) 127; Detienne (1986) 64–7; Kannicht (1988)
20. Porphyry’s account of Theagenes’ interpretation already implies that it is a defence of epic
accounts of the gods against the accusation that they are ‘inappropriate’ (ἀπρεπής), cf. Xenophanes
D19=B26.2.
24
Introduction 25
of explanation.4 On the other hand, more recent commentators have
tended to stress that Xenophanes can hardly have intended to attack
poetry from the perspective of philosophy, since he was himself a poet
and his criticisms of Homer and Hesiod conform to the traditional
phenomenon of poetic agonism.5 Indeed, in many respects,
Xenophanes’ use of verse has seemed less problematic than that of
Parmenides or Empedocles: his elegiac fragments on the appropriate
behaviour at the symposium are of similar content and function to the
didactic poetry of Solon and Theognis. Thus, ‘as an elegiac poet, he was
a typical figure of his age’.6 Additionally, the narrative of a steady
progression from ‘mythological’ to ‘rational’ forms of explanation has
been criticised partly on the grounds that it assumes a teleological
development whereby polytheism must give way to Judaeo-Christian
monotheism. Once we relinquish such an assumption, Xenophanes’
‘single greatest god’ may seem no more ‘rational’ or ‘philosophical’
than the traditional Greek pantheon.7
In spite of Xenophanes’ immense importance for these fundamental
topics, recent scholarship has seldom engaged in a holistic discussion that
encompasses both the cultural context and the philosophical content of the
surviving fragments. Literary critics and cultural historians tend to occupy
themselves exclusively with the elegies, in isolation from Xenophanes’
epistemology and cosmology, which are largely in hexameters.
Philosophers, on the other hand, aim to reconstruct the content of
Xenophanes’ thought, whilst paying less attention to the rhetorical and
aesthetic function of the texts.8 Comments on the issue of Xenophanes’ use
of verse have tended to focus on the medium’s cultural authority and its
practical benefits.9 The present chapter adopts a more fine-grained
approach. I shall investigate the sorts of responses and interpretations
4
Rejection of myth: Detienne (1986) 63–9; Schäfer (1996) esp. 146–62. Key player in development
from ‘myth to reason’: Nestle (1942) 87–95; Snell (1953) 139–43, 195; Fränkel (1955) 335–49, (1975b)
325–37; Heitsch (1966) 216–17; Lloyd (1987) 176–7; Most (1999) 352–3. Morgan (2000) 47–53
provides a brief but nuanced discussion of the issue.
5
Nightingale (1995) 64; Ford (2002) 46–66; Most (2011) 5; Pozdnev (2016). On Archaic Greek poetic
agonism more generally, see Griffith (1990); Collins (2004) (who discusses Xenophanes on pp.
147–51); Burton (2011).
6
Kahn (2003) 156. Bernabé (1979) 370–1 makes similar remarks.
7
Gemelli Marciano (2005a); note also R. Fowler (2011) 56, ‘If Xenophanes has what we call “myths” in
view, then only some myths (the impious ones) are fictional; other myths are good-omened, and the
myth itself is not at stake.’ The ‘myth to reason’ narrative is challenged in Buxton (1999).
8
As examples of the former: Ford (2002) 46–66; Hobden (2013) 22–34; the collection Bugno (2005).
Exceptions to this trend are the commentaries of Heitsch (1983) and Lesher (1992).
9
Most (1999) 352; Patzer (2006) 75; Granger (2007) 430; Sassi (2018) 96–7.
26 Xenophanes
that Xenophanes’ verse elicits by drawing upon their literary and performa-
tive contexts as well as the wider context of his philosophy.
Unlike Parmenides, who conveniently wrote only a single poem, with
Xenophanes we cannot be certain that any two fragments come from the
same text. Three or four poems are attested. From Diogenes, in a report
which goes back to the disreputable Hellenistic biographer, Lobon of
Argos, we learn that ‘he composed both on the foundation of Colophon
and on the colonisation of Elea 2,000 lines’.10 Most commentators take
this to be a single work, perhaps divided into two books on each of the
topics, but the expression leaves open the possibility that they were two
separate poems, and that only the latter amounted to 2,000 lines. In any
case, these works are likely to have been in elegiacs rather than hexam-
eters: there are some parallels for such lengthy historical elegies, and
whilst none of Xenophanes’ surviving hexameter fragments would easily
fit such poems, at least one of his elegiac fragments (D62=B3) would suit
a poem on the history of Colophon.11 A number of sources mention
Xenophanes’ Silloi – a term usually translated as ‘satires’ – written in
a variety of metres against other philosophers and poets and imitated by
the Hellenistic Sceptic philosopher-poet Timon of Phlius.12 It is not
clear to what extent this latter work was a single continuous poem, or
a disjointed collection of short vignettes on particular figures. The plural
title, Timon’s imitation and the fact that it seems to have contained
a variety of metres suggest the latter. In either case, one testimony
presupposes that the work was of some length, stretching to at least
10
Diog. Laert. 9.20<D1<A1=Lobon fr. 17, ἐποίησε δὲ καὶ Κολοφῶνος κτίσιν καὶ τὸν εἰς Ἐλέαν τῆς
Ἰταλίας ἀποικισμὸν ἔπη δισχίλια. On Lobon as a source for this sort of information, see Trépanier
(2004) 27, who notes the biographer’s proclivity to provide long, descriptive titles of particular
works, and Cerri (2001) 181–2, who disputes the view that he was an unscrupulous falsifier.
11
See Bowie (1986) 31–2. Although Lobon/Diogenes use ἔπη of this work, the term can refer to
elegiacs, as Solon (fr. 1.2 W) and Theognis (20, 22, 755) demonstrate. KRS 166 doubt the existence of
this poem, and are followed by Tor (2017) 314, but they expressed this view before Bowie’s important
arguments in favour of the existence of ‘historical elegy’, which were in some part confirmed by the
publication in 1992 (in The Oxyrhynchus Papyri 59 and West [1992]) of new papyri of Simonides’
elegies on the battles of Platea and Artemisium. Bowie’s notion of ‘historical elegy’ has been criticised
by Sider (2006) and Grethlein (2010) 291–6, but Sider does not deny the existence of lengthy elegies
that reported details of a city’s legendary past, whilst Grethlein’s arguments only suggest that
Tyrtaeus’ Eunomia and Mimnermus’ Smyrneis could have been sufficiently short to have been
performed at symposia. Some positive evidence (i.e. Diogenes/Lobon on Xenophanes’ lengthy
poem) confirms Bowie’s thesis, whilst none is incompatible with it.
12
The title is attested by Strab.14.1.28 (=DKA20); Procl. in Hes. Op. 286=D3=A22; schol. ABT ad
2.212b (D4=A23). The fragments of Timon are collected as SH 775–840, and edited with commen-
tary by di Marco (1989); see also Clayman (2009).
Xenophanes and the Hexameter Genre 27
five books.13 Later sources mention a Peri Phuseōs,14 and Diels-Kranz
print fragments under that title, but it should be regarded with caution,
as it appears to be a standard Alexandrian label given to the works of
those labelled phusikoi by Aristotle.15
Despite the variety of rhetorical, didactic and aesthetic purposes these
various texts must have served, I shall assume that the fact that they all
circulated under the same name justifies reading them ‘intratextually’, that
is to say, alongside one another as parts of a coherent whole.16 But I shall
also argue that, within this coherent corpus, the division between the two
main metres that Xenophanes uses, hexameters and elegies – which prob-
ably also reflects different performance contexts – is a meaningful one.
Metre and performance context are the most obvious criteria by which to
identify Archaic Greek poetic genres, but genres distinguished by these
objective criteria also clearly accumulate distinctive ideological and the-
matic associations.17 Xenophanes uses hexameters and elegies for different
purposes, which can be elucidated by the divergent traditions of the two
genres. I shall consider the hexameter fragments first, since it is largely on
the basis of the hexameters that we can reconstruct an epistemological and
theological framework within which to read Xenophanes’ poetry as
a whole.
2 Xenophanes and the Hexameter Genre

Xenophanes has often been regarded as an itinerant rhapsode, a performer
of earlier hexameters, who travelled from Colophon to Elea, and who
rebelliously took to composing his own iconoclastic poetry.18 The story
offers the romantic image of a performer who one day grew frustrated with
his material and abandoned it in favour of his own compositions, setting
13
D56=B21a, from a Homeric scholiast (P.Oxy.1087.40), claims that the word ‘Erykos’ occurred in the
fifth book of Xenophanes’ Silloi.
14
The Geneva scholiast on the Iliad (D6a<B30); Pollux (<D6bB39); and Stobaeus (D6c<A36).
15
Cf. KRS 102–3 and 166: ‘That Xenophanes wrote a formal work on physical matters is highly
improbable.’ Contrast Marcovich (1978) 1. On the title Peri Phuseōs, see Schmalzriedt (1970).
16
For the term, see Sharrock and Morales (2000). On the principle that we should read the different
fragments of as part of a coherent oeuvre, cf. Morrison (2007) 57–67 who points out that, in the case
of most Archaic Greek poets, there is a consistent narratorial persona across the surviving corpus.
17
For Archaic poetic genres as determined by occasion, see Harvey (1955); Dover (1964) 189; Calame
(1974) 124–6 (critiquing the still-influential idea of Rossi [1971] 75–7 that Archaic poets followed
unwritten rules); Gentili (1988) 32–7; Nagy (1994); Ford (2002) 10–13. Cf. Plato Leg. 3, 700a–701b.
On the principle that genres are never purely formal, see Conte (1994) 105–28, and, with respect to
early Greek poetry, Swift (2010) 6–34.
18
This is assumed by the works cited at n. 3 above.
28 Xenophanes
the ball rolling for later attacks on poetry. This biography, however, rests
primarily on the flimsy evidence of Diogenes’ statement that he ‘rhapsod-
ized’ his own verses (Diog. Laert. 9.18=D1=A1, ἀλλὰ καὶ αὐτὸς ἐρραψώιδει
τὰ ἑαυτοῦ). The key term ‘rhapsodized’ (ἐρραψώιδει) has led some to infer
that Xenophanes was a rhapsodic performer of Homer and Hesiod,19 but
detailed studies of its usage suggest that it merely denotes a reciter of non-
lyric verses.20 Although Xenophanes seems to have been active during the
period, in the late sixth century bce, in which specialised rhapsodic guilds
dedicated to Homeric performance started to emerge,21 it is not clear that
there was yet a clear and impenetrable division between rhapsodes who, like
Plato’s Ion, performed the works of others and aoidoi, bards who composed
their own work. This was the period in which, according to a well-known
Pindar scholion, the ‘rhapsode’ Cynaethus composed a hymn to Apollo and
passed it off as Homer’s (FGrH 568 F 5). Some late sixth- and early fifth-
century bce sources present the poets Homer, Hesiod and Archilochus
themselves as ‘rhapsodes’.22 Most importantly, Diogenes’ testimony does
not even mention performance of Homer or Hesiod: its main point is that
Xenophanes performed his own verses, so it is rash to assume anything more
than that Xenophanes performed poetry that he composed.
Partly on the basis of this romantic narrative, commentators have often
stressed the polemical nature of Xenophanes’ relationship with the epic
tradition,23 as is evident from the incompatibility of many of his doc-
trines with epic as traditionally practised,24 the critical tone of certain
fragments (D8=B11, D9=B12) and the satirical nature of the Silloi.
Furthermore, it is the hexameters, rather than the elegies, which offer
new explanations on the theological and aetiological topics previously
covered so influentially by Homer and Hesiod, a detail which could be
taken to suggest that Xenophanes chose this medium primarily for
19
E.g. Gentili (1988) 158; Morgan (2000) 47. KRS 164 state, ‘He was certainly not . . . a Homeric
rhapsode.’ Cf. also Ford (2002) 50.
20
Ford (1988), (2002) 50 for this definition. Plato uses the term for performers of Archilochus (Ion
530b–531a) and Solon (Timaeus 21b). See also West (2011).
21
See, influentially, Burkert (1987b), although he reaches overconfident conclusions about the sixth-
century usage on the basis of fifth-century evidence, as Graziosi (2002) 18–40 points out. Hdt. 5.67.1
claims that Cleisthenes, at war with the Argives, banned the rhapsodes from competing at Sicyon in
around 570 because the Homeric poems glorified Argos. See further West (1999), (2010).
22
Heraclit. D21=B42; [Hes.] fr. 357; Pind. Isthm. 3.55–6=4.37–8, which, though not mentioning
rhapsodes explicitly, all feature wordplay on the etymology. On Cynaethus, see West (1999)
368–72, and, on the blurredness of the distinction between rhapsodes and aoidoi, West (2010) 2–3.
23
In addition to the scholarship cited at n. 3 above, note in particular Classen (1989); Granger (2007)
430; Sassi (2018) 93–8.
24
I use ‘epic’ here as a term of convenience to refer to hexameter poems of some length (i.e. including
Hesiod and the Homeric Hymns, but excluding hexameter epigrams).
agonistic reasons.25 But this polemicism, I shall argue, tells only half the
story. For there are also more positive dimensions to Xenophanes’ par-
ticipation in this tradition. Though he rejects some traditional aspects, he
repurposes others. In particular, he appropriates the immersive and
symbolic qualities of early Greek song to provide a particular experience,
one that guides his audience towards a new perspective on the world.
a Epistemology and the Immersive World of the Text

Earlier hexameter narrators, at least on the most literal reading, based their
ability to speak about the gods and the remote past on the grounds that they
had been divinely inspired, typically using an imperative to bid the Muse or
Muses to sing themselves, thereby creating some flexibility over whether the
audience is supposed to hear a mortal or a divine voice in the subsequent
account.26 Xenophanes, by contrast, rules out the possibility of direct divine-to-
mortal communication altogether (D53=B18, D15=A52).27 The narrator even
appears to undermine his own authority to comment on the grand matters of
his poem. D49=B34 and D50=B35 present crucial – if challengingly obscure –
evidence for the epistemological framework within which to read the surviving
fragments:
καὶ τὸ μὲν οὖν σαφὲς οὔτις ἀνὴρ ἴδεν οὐδέ τις ἔσται
εἰδὼς ἀμφὶ θεῶν τε καὶ ἅσσα λέγω περὶ πάντων·
εἰ γὰρ καὶ τὰ μάλιστα τύχοι τετελεσμένον εἰπών,
αὐτὸς ὅμως οὐκ οἶδε· δόκος δ’ ἐπὶ πᾶσι τέτυκται.
And so no man has seen nor will know that which is clear
About gods and the things which I say about all things;
For even if he happened especially to say something which has come to fulfilment,
Nevertheless he himself does not know; but opinion is wrought
upon all things.
(D49=B34)
25
Cf. Hdt. 2.53, a statement generally thought to reflect the pervasive influence of Homer and Hesiod
on notions of the gods throughout the Greek world from the Archaic period to the end of Classical
paganism, as Asheri, Lloyd and Corcella (2007) ad loc. note. See also Burkert (1985) 119–25. Schäfer
(1996) 147–8 points out the relevance of this testimony to Xenophanes.
26
Il. 1.1; Od. 1.1; Hes. Theog. 114–15, Op. 1–2, fr. 1.1–2; Herm. 1.1; Venus 1.1; Hymn. Hom. 14.2, 19.1, 20.1, 31.1,
32.1, 33.1. Some have argued that we are to imagine the Muses as the narrators (e.g. Ahl and Roisman
[1996]; Rabel [1997] 19; Palmer [2013] 318), but if that is the case, the voice can slip back to that of the
mortal narrator without warning (e.g. Il. 2.484–93; Theog. 963–8). For a good overview of this issue, see
Morrison (2007) 73–89.
27
Tor (2013), (2017) 104–54 argues that Xenophanes does not rule out the possibility of divine
disclosure of information to mortals in toto – they leave signs for us to understand the world (cf.
D51=B36); but this still seems to rule out Muse-inspiration.
30 Xenophanes
ταῦτα δεδοξάσθω μὲν ἐοικότα τοῖς ἐτύμοισι . . .
Let these things be thought to be like true things . . .
(D50=B35)
Some have confidently asserted that these lines go together and come at an
early point in the physical poem, but there is little supporting evidence for
either of these claims.28 All the same, wherever we place them, they are
programmatic of how we are to receive the text. The expression ‘about gods
and the things which I say about all things’ is a zeugma implying some
overlap between the two categories.29 Lesher is surely right to situate these
fragments within the context of ‘poetic pessimism’, such as Hesiod’s
Theogony proem (which D50=B35 seems to adapt) and Homer’s re-
invocation of the Muses before the Iliadic Catalogue of Ships (Il. 2.485–6),
both of which stress the ignorance of the mortal poet in contrast to the
knowledge of the Muses.30 This context highlights how Xenophanes’
example, at least as it survives, conspicuously lacks the knowledge of the
Muses as a counterpoint. As we noted in the Introduction,31 a number of
passages from early hexameter imply that the mortal poet can somehow
access the Muses’ superhuman eyewitness knowledge of events beyond the
perspective of current mortals, whereas, here, ‘no man has seen that which
is clear’ about the subject matter, nor, we may infer, is he granted
a superhuman vision by divine dispensation. Xenophanes subversively
adapts a traditional feature of hexameter poetry in a manner which under-
mines the authority of earlier poets:32 in Hesiod and Homer, the ignorance
of mortals is emphasised by contrast with the knowledge of the Muses,
a feature which underlines the privileged perspective of the Muse-inspired
singer; for Xenophanes, mortals are in an equally ignorant state, but there is
no possibility of such privileged status.
The epistemological foundation for this pessimism can be fleshed out by
some of the other hexameter fragments.33 Human beliefs tend to be
determined by the limited scope of their empirical experience, hence
28
Fränkel (1993) 128 and Classen (1989) 101, but the only evidence for these claims is that the two
fragments could coherently go together, that they could function as introductory to a work, and that
they bear resemblance to Hes. Theog. 27 and Alcmaeon fr. 1.
29
For this interpretation, see similarly Barnes (1982) 136–40, and Lesher (1992) 167–8. By contrast,
Babut (1974b) 438 n. 1, followed by Reibaud (2012) 55 n. 124, sees the two items ‘the gods’ and ‘what
I say’ as contrasted, but this possibility is undermined by the fact that Xenophanes does actually talk
about the gods.
30 31
Lesher (1992) 160. See Introduction, Section 2 above. 32 Thus Morgan (2000) 51.
33
Lesher (1992) 161–6 usefully summarises earlier views. For more recent interpretations, see Ioli
(2003), Mogyoródi (2006), Lesher (2008) and Tor (2013).
Thracians believe their gods to be red-haired and blue-eyed like themselves
(D13=B16, cf. D14=B15), and humans, if they had never tasted honey,
would think that figs were much sweeter than they do (D52=B38). As
Emese Mogyoródi points out, this implies that humans can never really
know how sweet a thing is, as there is always the possibility that there is
something sweeter that they have never tasted which would alter their
perception of sweetness. If humans could taste all sweet things, they might
be able to have knowledge of how sweet a thing is, as they would know how
sweet it is relative to all other sweet things; but that is impossible in
Xenophanes’ universe, as it appears to be infinite in extent (D41=B28).34
Gods may have such knowledge, but they do not communicate it to
mortals (D53=B18). The best mortals can do, it seems, is to amass as
many and as varied experiences as possible, in order to judge qualities
such as sweetness according to the widest possible scale.
This epistemological pessimism has marked linguistic implications.35
The hypothetical subject of the conditional clause of D49=B34.4–5, who
does not know whether he has spoken ‘something which has come to
fulfilment’, might just as well be Xenophanes himself: neither we, nor he,
can know whether his account happens to be telling the truth. The verb
τύχοι, as Fränkel argued, is best translated as ‘succeeded in’, rather than
‘happened to’, for it evokes the original usage of the verb as ‘to hit the
mark’, thereby activating the common metaphor of archery for speech.36
The speaker would have no means of knowing whether the ‘arrows’ of his
speech have ‘hit the mark’, in the sense of reporting what ‘has been
accomplished’ (τετελεσμένον). τετελεσμένον εἰπών adapts the traditional
hexameter expression σοὶ δ’ ἐγὼ ἐξερέω ὡς καὶ τετελεσμένον ἔσται, used of
confident assertions about the future, such as threats.37 Its usage here may
then imply that statements about the matters in question have the same
status as statements about the future, of which the truthfulness cannot be
determined in the present. The range of topics over which one cannot be
certain whether one’s speech has ‘hit the mark’ seems to be a wide one. If
34
Mogyoródi (2006) esp. 135–41. On Xenophanes’ infinite universe, see also Mourelatos (2002) 332–6
and (2008a) 138–9.
35
As Morgan (2000) 52–3 argues, Xenophanes raises wider problems about the relationship between
knowledge and its linguistic representation.
36
Fränkel (1993); for speech as an ‘arrow’ that might ‘hit the target’, note Aesch. Ag. 1194, Pind. Ol.
2.83–5.
37
Lesher (1983) 29ff. and Tor (2013) 260 regard the expression as having its provenance in Homeric
scenes of divination, but it is never used by any of the Homeric seers or prophets and is frequently
used of threats by both mortal and divine characters (See Il. 1.212, 2.257, 8.286, 401, 454, 9.310,
14.196, 18.427, 23.410, 672; Od. 2.187, 5.90, 15.536, 16.440, 17.163, 18.82, 19.309, 487, 547).
32 Xenophanes
the above sketch of Xenophanes’ epistemology is along the right lines, then
the uncertainty described in D49=B34.3–4 refers to large parts of language
in general, given that one cannot even have knowledge of such a mundane
detail as the sweetness of figs. In sum, we lack knowledge, and conse-
quently do not know whether our statements, even those concerning
mundane qualities, ‘hit the target’ and reflect objective reality.
These claims have been seen to entail a denial of the authority of earlier
Greek poets concerning the affairs of the gods and the events of the remote
past.38 But Xenophanes’ relationship with earlier poetry can be regarded as
more complex than a straightforward denial of their veracity. Audiences were
certainly aware that poets made up large parts of their tales and that the
effects of their poems do not depend upon their factual accuracy.39 For
Xenophanes, I suggest, many if not most claims about the perceptible world
assume the problematic status of earlier Greek poetry.40 Early Greek poetry
often presents an epistemological distinction between, on the one hand, the
affairs to which poets could only have access with the aid of the Muses and,
on the other, the everyday affairs for which Muse-inspiration is unnecessary.
Hesiod is able to teach Perses about seafaring, despite the fact that he has
hardly any experience in the skill, because the Muses have taught him (Op.
646–62), a detail which implies that he would not need the Muses if he had
had such experience. For Xenophanes, however, there is no such stark
distinction between the topics for which one could rely on the Muses and
those in our immediate expertise: we lack knowledge even over how sweet
a particular food is. In this respect, the familiar world we inhabit occupies the
same epistemic status as the world of the remote, epic past of which Homer
and Hesiod spoke: it is equally uncertain and open to contestation.
This analogy between the everyday world of experience and the remote
world of most surviving hexameter poetry41 is supported by the adaptation of
an earlier hexameter expression in D50=B35. The line conspicuously resembles
the locution, ‘Lies like true things’ (ψεύδεα . . . ἐτύμοισιν ὁμοῖα), used most
famously by Hesiod’s Muses (Theog. 27), but also of Odysseus’ deceptive tales
on Ithaca (Od. 19.203). In the most extensive study of this relationship, Jenny
38
Morgan (2000) 51–3.
39
See, on this topic, Pratt (1993) and Bowie (1993). Some poetic content (such as Demodocus’ lay of
Aphrodite and Ares, and Hesiod’s ainos of the hawk and the nightingale) makes no pretence of
factual truthfulness, and the claim, attributed to Solon (fr. 29 W), that ‘poets tell many lies’ became
proverbial. Cf. also Pind. Ol. 1.28–9, Nem. 7.20–3; Gorgias D35=B23; Dissoi Logoi 3.10.
40
Absolute claims of the type ‘this honey is very sweet’ would occupy such a status, but Xenophanes
might accept more limited claims of the type ‘this honey is sweeter than these figs’ as deriving
unproblematically from the knowledge of the speaker.
41
But not all, given the present-day setting of Hesiod’s Works and Days.
Bryan argues that the allusion draws attention to the replacement of ὁμοῖα
with ἐοικότα: the former denotes equivalence, the latter, ‘contingent and
potentially specious similarity’.42 Xenophanes’ account, then, is not marked
out as something which has the same effect as a true account, but rather as
something which has an uncertain relationship with the truth, since he cannot
know whether his speech has ‘hit the mark’ concerning the nature of god-
hood. I would add that the poetological significance of the earlier expression is
also relevant. ‘Lies like true things’ (ψεύδεα . . . ἐτύμοισιν ὁμοῖα) indicates the
capacity of Greek hexameter poetry to be fictionally verisimilitudinous:43
Odysseus, both liar and storyteller par excellence, is able to ‘fashion many
lies like the truth’ (Od. 19.203) in telling a story that seems real, even though it
is factually inaccurate and bears little resemblance to events which actually
occurred.44 Xenophanes, here, exploits his audience’s familiarity with this
capacity of earlier poetry to illustrate the epistemological status of the world
we inhabit: just as we cannot trust the engrossing tales of earlier poets, so we
cannot trust any account of the natural world, even an account concerning
mundane, familiar matters. An ancient reception corroborates this interpret-
ation. Plutarch quotes D49=B34.1–2 as verses that the young should bear in
mind so that they remember the precept that poets tell many lies (De aud.
poet. 17E), and places Xenophanes’ epistemological comments within the
context of longer-standing discussions over the truth status of poetry. Thus,
at least one perceptive ancient reader saw D49=B34 as a statement concerning
the potential fictionality of poetry.
D41=B28 further supports this understanding of Xenophanes’ use of
hexameters:
γαίης μὲν τόδε πεῖρας ἄνω παρὰ ποσσὶν ὁρᾶται
ἠέρι προσπλάζον, τὸ κάτω δ’ ἐς ἄπειρον ἱκνεῖται.
This limit of the earth is seen from on top at our feet
touching the air, but the underside stretches infinitely.
The fragment expands upon the earlier formulaic phrase ἐπ’ ἀπείρονα γαῖαν,
‘on the limitless earth’, which is especially used to label the earth as the residence
of mortals and demigods.45 The traditional expression for Xenophanes is only
42
Bryan (2012) 16–48.
43
Thus Rutherford (1992) ad Od. 19.203 and, on the much-discussed line from the Theogony, Stroh
(1976), Pratt (1993) 106–13 and Heiden (2007).
44
See Collobert (2004) for the idea that the Odyssey’s poetics is founded upon the notion of
vraisemblance or ‘verisimilitude’. See similarly, on Odysseus’ ‘lies like the truth’, Williams (2002)
175–7.
45
E.g. Il. 7.446; Theog. 187; Op. 160 etc. See LfgrE s.v. ἀπείρων 1aα.
34 Xenophanes
half-right: earth is limitless as far as it descends – ‘the underside stretches
infinitely’ or ‘indefinitely’ (ἐς ἄπειρον) – but the visible surface is a limit.
The fragment shows an expansion and adjustment of the traditional expression:
as was the case in the traditional formula, Xenophanes’ earth is specifically the
residence of humans, is presented from a human perspective (παρὰ ποσσὶν
ὁρᾶται) and is limitless; but, by contrast with the formula, Xenophanes
specifies that the earth is limitless only in one direction. If Xenophanes’ physical
poetry is about what ‘no man has seen’ (D49=B34.1), D41=B28 marks out the
limits to our visual experience, describing what goes on beneath the earth and
beyond our field of vision. The deictic τόδε implies that the audience are agents
of the passive ὁρᾶται: it is they who see ‘this’ limit of earth to which the
performer refers. They are thereby invited to regard themselves as characters in
Xenophanes’ poetry. The stylised world of hexameter becomes the world of our
immediate sensory experience. This strategy, I suggest, serves to illustrate an
epistemological point: just as epic is unverifiable, so, for the most part, we
cannot make verifiable, positive claims about our sensory experience. If this is
correct, then our experience of the poetry functions as a synecdoche for our
experience of the world at large: we are to apply the same sort of scrutiny to our
experiences as we might do to the questionable tales of the earlier poets.
It might be objected that this is an overinterpretation of a very brief
fragment, but it would be reductive to deny any connection between the
linguistic comment of D49=B34 and Xenophanes’ own use of a highly
defamiliarised form of language, the Kunstsprache or ‘artificial language’ of
hexameter poetry. The linguistic implications of his epistemological pessim-
ism encourage us to reflect on the triangular relationship between his
language, the language of hexameter tradition and the world to which he
refers. Through immersing his audience in the world of the text, challenging
them to penetrate its details and ponder its correspondence to reality,
Xenophanes exploits some of the traditional functions of hexameter poetry.
b Xenophanes’ Theology and the Language of Hexameter Poetry

In addition to the exclusion of Muse-inspiration, Xenophanes’ theology is
inimical to traditional epic in further respects. The epodic D12=B14
declares that mortals ‘suppose that gods are born (γεννᾶσθαι), wear their
clothes and have a voice and body’.46 The clear implication is that mortals
are mistaken in this supposition, especially since, in D16=B23, we learn that
46
D12=B14 is a couplet comprising an iambic followed by a hexameter. The first line has sometimes
been regarded as a corruption (e.g. Wright [1985] 47). However, Diogenes Laertius’ testimony that
the single greatest god is ‘not like mortals in body or thought’. The
principle that gods are so unlike humans is incompatible with a literal
understanding of much earlier hexameter poetry, and indeed, much earlier
Greek myth in general, with its anthropomorphic deities. Most pointedly,
theogony is invalidated, and there may be a denial of Hesiod’s Theogony in
particular if we read D50=B35 as alluding to its Muses.47
Nevertheless, in D26=B33, Xenophanes uses language that is redolent of
theogony, evoking the very genre to which his theology seems most hostile:
πάντες γὰρ γαίης τε καὶ ὕδατος ἐκγενόμεσθα.
For we were all born of earth and water.
Although Sextus Empiricus, who quotes the fragment, seems to read
D26=B33 as an adaptation from Homer,48 the closest parallel is Hesiod,
Theog. 154:49
ὅσσοι γὰρ Γαίης τε καὶ Οὐρανοῦ ἐξεγένοντο,
As many as were born of Earth and Heaven,
The second and third feet of Hesiod and Xenophanes’ lines are identical
and both lines refer to collections of beings who are born of two elemental
deities. The exact scope of Hesiod’s ὅσσοι, ‘as many as’, is unclear, as it
could refer to all of the children of Gaia and Ouranos or just the ones
recently mentioned, the Cyclopes and the Hundred-handers. The use of
a similar expression suggests a generic alignment with the Hesiodic text
which flags Xenophanes’ departure from the Hesiodic aetiology: Hesiod
explains the origins of these beings by means of sexual reproduction of the
earliest deities, whilst Xenophanes probably explained the origins of
humans by means of physical processes, given his non-anthropomorphic
theology and his physical explanation of other phenomena. The claim that
earth and water were the first principles of all things appears to have been
supported by empirical observations, such as the appearance of shells
inland and of fish-like fossils (Hippol. Haer. 1.14.5<D22<A33). The
Xenophanes wrote in iambics as well as elegiacs and hexameters (9.20<D1<A1) speaks in its favour.
Such an epodic combination is paralleled in the eighth-century bce ‘Nestor Cup’ inscription (CEG
454) and in the Margites. Wilamowitz (1962) 602–3 defended the authenticity of the fragment on the
basis of this evidence and even argued that Xenophanes composed the Margites, as the opening
fragment of the poem describes an old singer coming to Colophon.
47
On Xenophanes’ intertextual relationship with the Theogony more generally, see Classen (1989).
48
Math. 10.314, probably referring to Menelaus’ exclamation to the Achaeans, ἀλλ’ ὑμεῖς μὲν πάντες
ὕδωρ καὶ γαῖα γένοισθε (Il. 7.99), a curse meaning, effectively, ‘may you rot!’ See Kirk (1990) ad loc.
49
Cf. also Zeus’s expression at Theog. 648: Τιτῆνές τε θεοὶ καὶ ὅσοι Κρόνου ἐκγενόμεσθα.
36 Xenophanes
occurrence of the phrase, then, within the context of a different cosmology,
which purports to derive from human enquiry rather than divine inspir-
ation, is a function of how a traditional hexameter topic, the creation of
humans, is given new explanation. But the wider context of Xenophanes’
philosophy means that ‘born of’ in D26=B33 can no longer refer to a literal
process of sexual reproduction, as it does in Hesiod’s Theogony: it must be
understood as a metaphor. A further example occurs at D46=B30, where
the sea is referred to as the γενέτωρ (line 5), or ‘procreator’ of the clouds
and winds, a personifying term first attested here that evokes theogonical
explanation. As in D26=B33, within the context of Xenophanes’ anti-
anthropomorphism, the term becomes conspicuously metaphorical: the
sea is not really the procreator of clouds and winds but is its source.
One difference, then, between Xenophanes and his hexameter predeces-
sors is that he sharpens the distinction between literal and metaphorical
depictions of the gods by imposing such strict and explicitly stated condi-
tions on what counts as truly divine: where his language contradicts his
explicit doctrines, we must understand it as metaphorical.50 By contrast, in
Homer or Hesiod, it is not always clear whether this sort of personifying
language is meant literally or metaphorically, partly because an essential
function of their poetry does not depend on its factual accuracy.51 Ancient
audiences were divided on this issue: already by the fifth century, some
sources treat the canonical hexameter poets as representing the gods
literally, whilst others see their works as providing ainigmata, ‘allegories’
or ‘riddles’, which must be decoded in order to access the true underlying
meanings.52
Xenophanes’ more conspicuously metaphorical use of language might,
then, be taken to presuppose the possibility of a more ‘rational’, non-
metaphorical way of speaking about this sort of subject matter, one that
transparently corresponds to its referents. But the misgivings about mortal
language expressed in D49=B34 seem to rule out this possibility and the
claim that mortals wrongly suppose that gods have ‘their [i.e. human]
voice’ (τὴν σφετέρην . . . φωνήν, D12=B14.2) implies an imperfection in
mortal language.53 I would therefore prefer to see Xenophanes, not as
pointing towards some ‘rational’ ideal of non-metaphorical speech, but as
50
Cf. Lloyd (1987) 178: ‘the important difference [between Xenophanes and earlier Greek hexameter
poets] is that Xenophanes guards himself against too literal interpretation of his religious positions’.
51
Thus e.g. Macleod (1982) 7; Heath (1985) 259–62; Bowie (1993) 8–23; Pratt (1993) passim; Halliwell
(2011) 36–92.
52
See Introduction, Section 2 above.
53
See Chapter 2 n. 19 below on the possibility that divine language is different from that of mortals.
manipulating for didactic purposes the tendency of hexameters to invite
a metaphorical interpretation. Our interpretation of Xenophanes’ metaphors
serves to prepare us for an interpretation of the world. The occurrence of
gignomai and its cognates in new contexts such as D26=B33 and D46=B30
provokes reflection on the mechanisms by which currently observable phe-
nomena were produced, and on what it really means to ‘come to be’. In this
way, the characteristic tendency of Greek song to invite a symbolic interpret-
ation is enlisted to guide the audience to a new way of looking at things.
A further strategy that serves this end is the evocation of certain trad-
itional, anthropomorphic gods, especially Zeus. Consider D46=B30:
πηγὴ δ’ ἐστὶ θάλασσ(α) ὕδατος, πηγὴ δ’ ἀνέμοιο·
οὔτε γὰρ ἐν νέφεσιν <γίνοιτό κε ἲς ἀνέμοιο
ἐκπνείοντος> ἔσωθεν ἄνευ πόντου μεγάλοιο
οὔτε ῥοαὶ ποταμῶν οὔτ’ αἰ<θέρος> ὄμβριον ὕδωρ,
ἀλλὰ μέγας πόντος γενέτωρ νεφέων ἀνέμων τε
καὶ ποταμῶν.
And the sea is the source of water, the source of the wind;
for there would not within the clouds <come to be the force of wind
blowing> from inside without the great sea
nor the currents of rivers nor the rainy water of the air,
but the great sea is the procreator of clouds and winds
and the rivers.
The fragment is quoted by the Geneva scholiast to Il. 21.195–7, where
Okeanos is the source of all rivers, seas, springs and wells. Xenophanes’
notion of the sea as the γενέτωρ, not only of waters but also of winds
and clouds, may be a development from the occasionally attested idea
that Poseidon controls winds and storms (e.g. at Od. 5.291–9).54 It is
equally an adaptation of the Hesiodic depiction of Pontus as the father
of the anthropomorphic Nereus, Thaumas, Phorcys, Ceto and Eurybia
(Theog. 233–9).55 But Zeus is also strongly evoked. ‘The great sea, the
procreator of clouds and winds’ recalls the common hexameter formula
for the king of Olympus, ‘father of men and gods’ (πατὴρ ἀνδρῶν τε
θεῶν τε). The fact that the great sea here usurps Zeus’s long-standing,
traditional role as the god of wind and rains confirms this connection.56
In fact, Xenophanes presents the sea in a grandiloquent manner that
54
Classen (1989) 97.
55
Cf. Lesher (1992) 137: ‘for Xenophanes pontos “begets” only clouds, wind, and water’.
56
As Classen (1989) 98 points out, cf. Od. 5.303–5, 9.67–9, 12.313–15. For Zeus as responsible for wind
and rain in both literature and lived religion, see Parker (1996) 29–32.
38 Xenophanes
would befit Zeus, manifest in the anaphora of ‘spring’, and in the
emphasis on γενέτωρ, as an affirmative predicate after a tricolon of
negated ones.
Zeus is still more strongly evoked in the fragments that describe
Xenophanes’ single greatest god:
εἷς θεός, ἔν τε θεοῖσι καὶ ἀνθρώποισι μέγιστος,
οὔτι δέμας θνητοῖσιν ὁμοίιος οὐδὲ νόημα.
One god, both among gods and mortals the greatest,
Not like mortals in his body or his thought.
(D16=B23)
οὖλος ὁρᾶι, οὖλος δὲ νοεῖ, οὖλος δέ τ’ ἀκούει.

Whole he sees, whole he thinks, and whole he hears (D17=B24)
ἀλλ’ ἀπάνευθε πόνοιο νόου φρενὶ πάντα κραδαίνει

But without labour he shakes all things with the organ of his thought.
(D18=B25)
αἰεὶ δ’ ἐν ταὐτῶι μίμνει κινούμενος οὐδέν

οὐδὲ μετέρχεσθαί μιν ἐπιπρέπει ἄλλοτε ἄλληι
But always he remains in the same place, not moving at all
Nor is it fitting for him to travel now to one place, now to another
(D19=B26)
D18=B25 reworks the earlier conception of Zeus’s ability to deter-

mine all things simply by giving the nod. The detail that
Xenophanes’ god ‘shakes’ all things (πάντα κραδαίνει) recalls the
image of Zeus ‘shaking’ Olympus when he nods assent to Thetis’
request at Il. 1.528–30.57 In Xenophanes’ version, the physical act is
diminished: whilst Zeus nods with his head, causing his hair to flow
from his scalp and Olympus to shake, Xenophanes’ god does this all
with his mind, without any motion (cf. D19=B26). D16=B23 similarly
invites comparison with Zeus. Although the expression, ‘one god,
among gods’ seems paradoxical, the fragment does not present
a contradiction or a monotheistic theory, as some have supposed,
for there is no contradiction in the henotheistic claim that there is
one god who is greatest among gods and mortals but more than one
57
Note also Il. 8.199. See Lesher (1992) 109–10 for this and further parallels.
god exists.58 The notion of one god being ‘the greatest’ (μέγιστος) is a further
detail familiar from Greek literary depictions of Zeus. In similar phrasing to
D16=B23, the short Homeric hymn to Zeus describes him as ‘the best and
greatest of gods’ (θεῶν τὸν ἄριστον . . . ἠδὲ μέγιστον, Hymn. Hom. 23.1).
The detail that this is without toil conforms to the tendency of early Greek
poetry to describe gods as doing things ‘easily’, but this seems especially
associated with Zeus.59 The comparison, however, throws into relief the non-
anthropomorphic nature of Xenophanes’ god: D16=B23.1, taken on its own,
could in fact describe Zeus; the following line would then come as an
aprosdoketon, an unexpected twist that this god is in fact not Zeus at all for
it is not like humans in body or appearance.
Indeed, D16=B23.2 is an adaptation of the hexameter formula οὐ δέμας
οὐδὲ φυήν, ‘not in respect of body or stature’, which is repeatedly used of
comparisons between mortals and gods. Calypso compares herself favour-
ably to Penelope on the grounds that she is not like her ‘in respect of body
or stature’ (Od. 5.212–13). Odysseus himself uses the formula to deny that
he is similar to the gods (Od. 7.210), and the Cretan leader in the Homeric
Hymn to Apollo deploys it to identify the god as unlike mortals (463–5).60
Comparison with the original highlights, by contrast, the mental aspect of
Xenophanes’ god – it is unlike in respect of its mind rather than its
appearance – whilst at the same time illustrating that Xenophanes’ non-
anthropomorphic deity is simply an extension of the traditional belief that
gods are ‘unlike’ mortals in certain respects.
On one level, then, Xenophanes appropriates traditional forms and
imagery to throw into relief the points of contrast between his own and
previous explanations. But there is a further, positive dimension to this
implicit comparison. Rather than seeing the allusions to Zeus as wholly
negative, we might also regard them as a means of revealing the nature of
Xenophanes’ deity not only by contrast but also by analogy. This god, in
being the greatest and most powerful being, fills a role once occupied by
Zeus. The image of Zeus thus serves as a symbol for the unfamiliar god. In
58
On this question, see recently Versnel (2011) 244–66, who argues that Xenophanes exhibits cognitive
dissonance, in being both a monotheist (in conceiving of a god as encompassing the whole universe,
a view well attested in the testimonia e.g. [Arist.] MXG 3.977a114ff.=R6+R14=A28, Simpl. Phys.
22.22ff.=R4<A31, Cic. Acad. Pr. 2.18, Nat. D. 1.11.28=R23=A34) and polytheist (as conceiving of
multiple lesser gods within that singular cosmic god).
59
Of Zeus: Il. 15.490, 16.690=17.178, Hes. Op. 5–7, 379, Aesch. Eum. 651; of an anonymous god at
Thgn. 406; of gods in general: Od. 16.211, possibly Archil. 130; Aesch. Supp. 100. See West (1978)
139–40.
60
Note also the use at Il. 1.115 (Agamemnon’s comparison of Chryseis to Clytemnestra) and Ap. Rhod.
2.37 (the narrator compares the appearance of Amyclus and Polydeuces).
40 Xenophanes
fact, the restrictions over what one could accurately say about this god seem
so tight that it is virtually only in symbols that one could talk about him.
His unmoving (D19=B26), pointedly non-anthropomorphic nature
(D16=B23) almost rules out the possibility of a narrative about him:
narrative texts represent events and have a linear progression.61 The fact
that this god is unchanging (D19=B26) limits the sorts of events in which
he could be an agent: he participates in mental events (D17=B24; D18=B25)
but his physical immobility means that he could not be the subject of the
sorts of physical narratives of Greek myth. A narrative about the mental
behaviour of Xenophanes’ god would be quite different from any known
Greek narrative. The evocations of Zeus serve as a means of saying the
unsayable about this god who could not be the subject of any conventional
hexameter narrative.
This end is also served by the formal aspects. Xenophanes’ grandilo-
quent diction in D46=B30 has the trappings of thespis aude, the divine
speech that opens a channel between mortals and the divine, but it is
directed towards the sea rather than any anthropomorphic god, inviting
the audience to view the natural world with the respect that typified
attitudes towards the divine. Form also has a religious significance in the
fragments describing the single deity, for they appear distinctively
hymnic.62 In addition to the resemblance between D16=B23.1 and Hymn.
Hom. 23.1, these fragments, when taken together, constitute an aretalogy,
a form characteristic of hymns.63 We might compare the description of
Zeus’s attributes at the hymnic opening to the Works and Days (5–8):64
ῥέα μὲν γὰρ βριάει, ῥέα δὲ βριάοντα χαλέπτει,
ῥεῖα δ’ ἀρίζηλον μινύθει καὶ ἄδηλον ἀέξει,
ῥεῖα δέ τ’ ἰθύνει σκολιὸν καὶ ἀγήνορα κάρφει
Ζεὺς ὑψιβρεμέτης, ὃς ὑπέρτατα δώματα ναίει.
For easily he strengthens, and easily he crushes the strong,
And easily he diminishes the famous and increases the obscure,
And easily he straightens the crooked and withers the bold
Zeus, high-thundering, who dwells in the highest palace.
61
For these as two necessary conditions for narrative, see Lowe (2000) 20 and similarly Cobley
(2014) 5.
62 63
Deichgräber (1933) 350–1, 360; Granger (2007) 424. Janko (1981) 11–12.
64
For this opening as a hymn note e.g. West (1978) 136–7 (this section as a prooimion, like the Homeric
Hymns); Janko (1981) 22. Other hymnic uses of anaphora, noted by Deichgräber (1933) 360, include
OF 243.1–7, 245.1–7.
As appears to be the case in the Xenophanean lines, there is a list of present-
tense verbs outlining the god’s attributes and the repeated ‘easily’ corres-
ponds to the repetitions of Xenophanes’ ‘whole’ (D17=B24). His singular
god is thus presented in a stylised manner that implies that he is worthy of
the sort of veneration previously reserved for the familiar pantheon. The
hymnic quality of these fragments implies that their performance would be
an act of piety towards the god, and piety is attested elsewhere as
a distinctive concern of Xenophanes (D59=B1.14–15; Aristotle, Rhet.
2.23.1400b5–8=P17=A13). It is possible that, like other hymns, there is
a hope that this poetry will please the single greatest god and induce him
to dispense good fortune to the performer.65 But a further function must
surely be to promote reverence for this unfamiliar deity among the audi-
ence by redirecting the traditional associations of hymnic form towards
this strange new singular deity, a tactic that would be difficult if not
impossible in a prose treatise. The hymnic form, then, like the use of
symbolism, functions as a means by which to lead the audience to
a contemplation of the ineffable divine: through the form, the text com-
municates more than it states literally.
c The Attacks on Homer and Hesiod

This concern for piety also underlies what are arguably Xenophanes’ most
famous fragments, the attacks on Homer and Hesiod, though scholars have
tended to see the butt of these attacks as being inaccuracy.66 For Lloyd,
these criticisms presuppose a ‘literalist interpretation of representations of
the gods . . . [Xenophanes was] starkly literal-minded’.67 Certainly, as we
have noted, Xenophanes’ theological doctrines entail that most earlier
Greek hexameter poetry – with its anthropomorphic depictions of the
divine – is factually incorrect,68 and, as we shall see below, he describes
Titans, Giants and Centaurs, the subject matter of some earlier verse
(although not the Homeric epics), as πλάσματα, ‘fabrications’
(D59=B1.23). But again, this only tells half of the story, as a closer examin-
ation of the relevant fragments shall reveal:
65
Pace Warren (2013) 307–10, who argues that Xenophanes’ gods do not intervene directly inhuman
affairs, but this overlooks the fact that the god ‘steers all things’ (D18=B25), and so any good fortune
we experience must be divinely dispensed. Cf. Tor (2017) 139–42 on Xenophanean piety and divine
disclosure. On this function of Greek hymns, see Furley (1995).
66
Rösler (1980a) 286–9; Verdenius (1983) 28; Lloyd (1987) 177–9; Halliwell (1998) 10–13; Finkelberg
(1998) 18–27; Most (1999) 337–8; Cf. Verdenius (1970) on Homer as ‘the educator of the Greeks’.
67
Lloyd (1987) 177–8 (original italics).
68
Especially D12=B14, but also D14=B15 and D13=B16.
42 Xenophanes
ἐξ ἀρχῆς καθ’ Ὅμηρον ἐπεὶ μεμαθήκασι πάντες . . .
Since, from the beginning, all people have learned according to
Homer . . . (D10=B10)
πάντα θεοῖσ’ ἀνέθηκαν Ὅμηρός θ’ Ἡσίοδός τε,

ὅσσα παρ’ ἀνθρώποισιν ὀνείδεα καὶ ψόγος ἐστίν,
κλέπτειν μοιχεύειν τε καὶ ἀλλήλους ἀπατεύειν.
Homer and Hesiod have attributed all things to gods
Which are a cause for reproach and censure among mortals,
Stealing and committing adultery and deceiving one another. (D8=B11)
ὡς πλεῖστ(α) ἐφθέγξαντο θεῶν ἀθεμίστια ἔργα,

κλέπτειν μοιχεύειν τε καὶ ἀλλήλους ἀπατεύειν.
For the most part, they uttered profane things about gods,
Stealing and committing adultery and deceiving one another. (D9=B12)
D8=B11 and D9=B12 concern the impiety of the poets, as Sextus assumes
when quoting D8=B11 (Math. 9.192–3) to corroborate the point that poetry
‘is full of all kinds of impiety’ (πάσης γὰρ ἀσεβείας ἐστὶ πλήρης).69 The
repeated line has led some scholars to regard D9=B12 as a corruption of the
more extensive D8=B11, but since repeated lines are a distinctive feature of
Greek epic style this inference is unwarranted.70 The repetition is better
seen as a deliberate emphasis. As commentators have often pointed out, the
repeated line recalls events in the Homeric poems, such as the Dios Apate
scene, where Hera deceived Zeus, and Demodocus’ song of Aphrodite’s
adultery with Ares, as well as the Homeric Hymn to Hermes, which so
memorably paints the god as a shameless thief.71 There is not much more to
go on, but the vocabulary further implies that the poets’ depictions of the
gods commit a deleterious impiety. I have followed the standard transla-
tion of ἀνέθηκαν as ‘attributed’, but, as Heitsch has suggested, the verb may
here also connote the sense of ‘set up as set up as a votive gift, dedicate’ (LSJ
s.v. II), a possibility recommended by the juxtaposition with θεοῖσι, ‘to the
gods’.72 Songs are themselves often offerings,73 so the possible implication
is that Homer and Hesiod composed songs as gifts to the gods which are in
69
Translation here from Bury (1936).
70
As Babut (1974a) 85–6 points out. Cf. the self-conscious repetitions in Empedocles D45=B25.
71
For more extensive discussion of the possible episodes here referenced, see Babut (1974a) 87–9;
Heitsch (1983) 125–6; Classen (1989) 92–3.
72
See Heitsch (1983) 125.
73
For this principle of hymns, see Bremer (1998) 134–7. Note also the use metaphor of agalma,
‘pleasing offering’ for song in Pindar (Nem. 3.13, cf. Nem. 5.1) and Bacchylides (5.4, 10.11), on which
see Steiner (1993).
fact a source of revulsion even among mortals. This would be a satirical
gesture: the poets were foolish enough to think that they would actually
please the gods by portraying them as engaging in such deceptive, adulter-
ous and kleptomaniac behaviour. One might object that the final line of
the fragment refers to particular content, not to the songs themselves, but
this detail need not rule out the possibility of such an implicit joke and, in
any case, in an era before distinct, generally accepted song-titles are
attested, a summary of the content is one means of referring to
a particular story or text.74
Be that as it may, the impiety of these portrayals is made more explicit in
D9=B12. The adjective ἀθεμίστος is typically used in connection with
actions that are unspeakably perverse or transgress the rules of religion.
Homer uses it of the man who loves civil war (Il. 9.63) and the Cyclopes
(Od. 9.106), while Herodotus mentions the unspecified ἀθέμιστα deeds that
the Persian governor Artayctes does to women in the temple of Protesilaus,
for which he is eventually crucified (7.33). Parmenides and Empedocles
would cast their poems in the tradition of mystic speech that was character-
ised by θέμις, religious ‘right’ or ‘correctness’.75 For Xenophanes, the
Homeric and Hesiodic poems are the opposite of pious speech.
Possibly, the underlying idea is that this impious speech brings displeas-
ure to the single greatest god, just as Xenophanes’ own poetry seems, like
other hymns, to be designed to please him. This motivation would be
paralleled in the roughly contemporary denials of certain myths by
Stesichorus (frr. 90–1 Finglass) and Pindar (Ol. 1.23–35, 52–3), although
unlike Xenophanes, they do not rule out the whole anthropomorphic
framework for traditional depictions of the gods and, of course, it is not
clear how serious Stesichorus is when recanting his earlier account of
Helen. Less tentatively, D10=B10 supports the notion that Xenophanes
deemed this kind of speech wrongful on the grounds that it promotes
impiety and generally bad behaviour among mortals. The fragment is
agonisingly incomplete: it is quoted by Herodian to illustrate the contrac-
tion of the penultimate syllable of the verb, so he offers no clue as to
context or the apodosis of the sentence. ‘From the beginning’ (ἐξ ἀρχῆς)
may denote an early stage either of humankind in general or of an
individual lifetime (i.e. ‘from youth’).76 In either case, the fact that the
74
Note the way in which, before the book-divisions of the Homeric poems, references to particular
episodes take the form of summary descriptions (e.g. Hdt. 2.116, Thuc. 1.9.4).
75
See Chapter 2 n. 86 and Chapter 3, Section 4c below.
76
At D53=B18.1 Xenophanes uses ἀπ’ ἀρχῆς in the former sense, but note Gorgias B11a=D25.29 for the
latter.
44 Xenophanes
early influence is significant suggests that this education ingrains habits
that will affect subsequent behaviour and outlook. Studies of the early
usage of μανθάνειν suggest that the verb here refers not only to the learning
of information but also to the development of habits and customs.77 The
fragment leaves open the possibility that Homer’s factual inaccuracy is not
objectionable in and of itself – the tentative nature of Xenophanes’ own
claims about the gods makes it unlikely that he would have confidently
denied the factual content of claims made by others – rather, the problem
seems to be Homer’s undesirable influence on mortal customs. Indeed, this
is the problem with certain types of speech in D59=B1, where talk of battles
with mythical monsters or of civic strife is excluded from the ideal sympo-
sium (D59=B1.20–4), probably for the reason that they disrupt the har-
mony of the gathering. Strictly speaking, none of this entails that the
problem with Homeric and Hesiodic depictions of the gods was their
inaccuracy, since one could imagine depictions of the gods which would
be inaccurate from Xenophanes’ perspective without exhibiting these
boorish behaviours.
The parallels with Plato’s reservations about poetry in the Republic have
sometimes been noted and can be instructive.78 As scholars have pointed
out in the case of Plato, the sort of texts to be censored are not books that
are read privately but are communal public performances, more analogous
to modern mass media such as popular television programmes and block-
buster films than to modern poetry.79 Just as twentieth- and twenty-first-
century commentators sometimes criticise graphic depictions of sex and
violence on television for gradually moulding societal norms, Xenophanes
criticises Homer and Hesiod’s depictions of the gods as normalising
deception and adultery. In Xenophanes’ case, the issue is more salient
since it concerns the depiction of gods who are the object of worship.80 But
his criticism does not entail that he lacked a conception of fiction. The
fragments are concerned with poetry’s social function, and its veridical
status plays a role in this – it ought not to present falsehoods about the gods
lest audiences should believe false things about them – but the main
emphasis concerns the types of behaviour that the poems encourage,
a quality that is not exclusively determined by their veracity. It is not the
77
See Lesher (1992) 81–2, following Snell (1973). Cf. Il. 6.444–5.
78
Babut (1974a) 105–9; Lesher (1992) 82; Cerri (1996) 55–66, whose discussion of Xenophanes is
limited to D59=B1.
79
Nehamas (1988); Burnyeat (1997) esp. 249–55.
80
Cf. Aristotle’s anecdote about Xenophanes and the citizens of Elea, mentioned in Section 3a below.
veridical status of poetry (as opposed to other speech-genres) that is at issue
so much as its effect on its audience.
The attacks on poets, along with the denial of direct divine-to-mortal
communication and the non-anthropomorphic explanations for natural
phenomena, are grist to the mill of scholars who grant Xenophanes a major
role in the development towards a distinction between ‘mythical’ and
‘rational’ forms of explanation.81 Of course, it is unlikely that he would
have drawn the distinction in these terms, since the words muthos and logos
were yet to attain such opposing connotations.82 His recommendation at
D59=B1.15 to hymn the god with ‘auspicious muthoi and pure logoi’
(εὐφήμοις μύθοις καὶ καθαροῖσι λόγοις) presupposes that muthoi are not
all bad. Though he ‘demythologises’ hexameters in the sense that he strips
them of supernatural content such as monsters and anthropomorphic
gods, the stereotypical image of Xenophanes as a myth-busting, rational
philosopher overlooks the ways in which he recognises and manipulates
what might be regarded as song’s irrational aspects. Most egregiously, the
fragments hint at the possibility that songs can please or displease the single
greatest deity. As far as we can tell, Homer and Hesiod are criticised for
promoting impious beliefs as well as immoral behaviour. Xenophanes’
own, positive hexameter statements seem designed as a benevolent coun-
terpart to this influence. He uses formal features to imbue the natural
world and his own god with the numinous grandeur that would have
previously been reserved for anthropomorphic deities, in fragments that
would hardly constitute a logos in the later sense of a ‘rational argument’.
Certainly, later philosophers would regard such formal features as repeti-
tion and word order (where the syntax is unaffected) as ‘irrational’ charac-
teristics of song even if they are also prevalent in ‘rational’, philosophical
prose.83 In the Hellenistic period, this assumption could support the view
that verse and poetry aimed at psuchagogia, ‘enthralment’ or literally ‘soul-
leading’, prose and philosophy at didaskalia, ‘teaching’.84 But Xenophanes
81
See n. 4 above.
82
For a recent and learned overview of this topic, see R. Fowler (2011). Note how Hecataeus of Miletus
fr. 1, in the sixth century, uses μυθέομαι and λόγος to refer, respectively, to his own purportedly
plausible account and fanciful traditional stories, in the exact reverse of later usages.
83
Philodemus, in On Poems Book 1, disputes the view that the excellence of poetry resides in its
irrational sound, as opposed to its rational content (esp. Col. 175); his On Music 4 passim treats the
non-linguistic components of music as irrational (ἄλογος). For these texts, see, respectively, the
excellent editions of Janko (2000) and Delattre (2007). A roughly similar distinction between
rational content and irrational form occurs at Demetrius On Style 183–6.
84
For the clearest statement of this distinction, see Eratosthenes ap. Strabo 1.1.10; note also Polybius
1.56.10–12. For a good overview, see Gutzwiller (2010).
46 Xenophanes
operated in a period before such a distinction and, if viewed in these terms,
his hexameters are both didactic and psychagogic.
He may have desired to supplant certain explanations that had been
provided by Homer and Hesiod, but this was not the only reason why
hexameters were an appropriate medium, for he plays upon their trad-
itional association with the divine. He alludes to the tradition of Muse-
inspiration in order to illustrate the epistemological limitations of mortals.
While thespis aude, as traditionally conceived, opened a perspective on the
normally inaccessible magnitudes of the past, Xenophanes redirects this
perspective towards the natural world and the single greatest god. The
elegies, as we shall see, further Xenophanes’ religious and ethical projects
on the level of the more familiar, mortal realm.
3 Xenophanes’ Elegies
Early Greek elegy appears to us as a more protean genre than early
hexameter. In the case of the latter, Homer and Hesiod, from an early
stage, seem to have been regarded as the key exemplars in the tradition,
whereas for elegy, we lack such paradigmatic figures. The form spanned
shorter poems, designed to be sung to the aulos at symposia, as well as
longer narratives on historical and mythological matters, perhaps designed
for performance at public festivals.85 The surviving fragments cover diverse
subject matter, from love and drinking, as we might expect from sympotic
poetry (e.g. Mimnermus fr. 1 W; Xenophanes D59=B1; much of the
Theognidean corpus), to the appropriate behaviour on the battlefield
(Tyrtaeus frr. 10, 11 W), to the narrator’s political achievements (e.g.
Solon fr. 5). In recent decades, we have also come to know more about
the longer historical/mythological type, since the publication in 1992 of
a papyrus of Simonides’ elegies on the battles of Platea and Artemisium (P.
Oxy.LIX 3695) and in 2005 of a papyrus containing an elegy (P.Oxy.LXIX
4708), probably to be attributed to Archilochus, which describes Telephus’
rout of the Achaeans.86 Xenophanes (like Mimnermus, Tyrtaeus, and
Simonides) was credited with composing elegies of both the shorter and
85
West (1974) 10–13 identified eight performance contexts, but as Bowie (1986) showed, only two can
be posited with confidence, the symposium and public festivals. All the same, West’s list is a function
of the wide variety of imagined scenarios in the surviving corpus. Bowie (2016) refines his earlier
notion of the longer, public elegies, locating them more specifically in cultic contexts. For the
content and function of early elegiac poetry more generally, see the overview of Aloni (2009) (largely
summarising the more extensive Aloni and Iannucci [2007]).
86
The ‘new’ Simonides was contained in West (1992) and discussed in Boedeker and Sider (2001). On
the Archilochus poem, see Obbink (2006) and, on its relationship to epic, Swift (2012).
Xenophanes’ Elegies 47
the longer type, if indeed the poem(s) on Colophon and Elea was in
elegiacs.
Through such a diverse elegiac corpus one common thread, relevant to
Xenophanes, is the mortal perspective of the narrator. We noted above that
the hexameter performer commands the Muse(s) to sing, creating the
possibility that his voice encompasses a divine perspective. By contrast,
the elegiac singer ‘is not a channel of direct divine vision’.87 He may be
a servant of the Muses (Theognis 769) and present his songs as their gifts
(Archilochus fr. 1.2 W; Theognis 250; Solon fr. 13.51 W; Anacreon eleg. fr.
2.3 W), but he does not ask them for information. Instead, Solon prays to
them for ‘happiness from the blessed gods’ (ὄλβον . . . πρὸς θεῶν μακάρων,
fr. 13.2 W) and ‘always to have good reputation among all men’ (πρὸς
ἁπάντων/ ἀνθρώπων αἰεὶ δόξαν ἔχειν ἀγαθήν, fr. 13.2–3 W). Their role
seems to be to render elegies successful and enduring. The difference
between epic and elegiac Muse-functions is best illustrated by Simonides’
elegy in celebration of the battle of Platea (fr. 11 W), where the poet sets up
a comparison between the glorious Danaans of the Iliad and the more
recent Spartan victors of his own poem. Assuming the most widely
accepted supplements to this fragmentary text, Simonides describes the
Iliadic poet as having ‘received the whole truth from the dark-haired
Pierians’ (ὅς παρ’ ἰοπ]λοκάμων δέξατο Πιερίδ[ων/ πᾶσαν ἀλη]θείην,
lines 16–17) before summoning the Muse as an ‘ally’ (κικλήισκω]
σ’ἐπίκουρον ἐμοί, line 21) and bidding her to ‘equip this sweet array of
song’ (ἔντυνο]ν καὶ τόνδ[ε μελ]ίφρονα κ[όσμον ἀο]ιδῆς, line 23) so that
later ‘someone will recall’ (ἵνα τις [μνή]σεται, line 24) the men who fought
for Sparta.88 As Stehle points out, ‘array’ (kosmos), along with the words
here translated as ‘ally’ and ‘equip’, is a militaristic image, implying that the
verses are lined up like ranks of soldiers.89 Homer required the Muse for
the ‘truth’ of his account, but the elegist Simonides needs her not for the
content but for the ‘arrangement’ of his song, and to ensure it will be
remembered.
As we should expect, given his exclusion of direct divine-to-mortal
communication, Xenophanes, like the other elegists, does not call upon
the Muses to provide his subject matter, going so far as to relinquish them
altogether. D66=B8 proffers an alternative basis for his narratorial
87
Gagné (2013) 232. On this difference see also Edmunds (1985) and Ford (1985). On the divine
perspective of the narrator of the Iliad, see Graziosi (2013).
88
Translation here based on Sider (2001). 89 Stehle (2001) 108–14.
48 Xenophanes
authority, one that is firmly restricted to the mortal perspective character-
istic of elegy:
ἤδη δ’ ἑπτά τ’ ἔασι καὶ ἑξήκοντ’ ἐνιαυτοὶ
βληστρίζοντες ἐμὴν φροντίδ’ ἀν’ Ἑλλάδα γῆν·
ἐκ γενετῆς δὲ τότ’ ἦσαν ἐείκοσι πέντε τε πρὸς τοῖς,
εἴπερ ἐγὼ περὶ τῶνδ’ οἶδα λέγειν ἐτύμως.
Already seven and sixty years
Have been tossing my thought about the Greek land;
And from my birth there were then twenty years and five in addition
to these,
If I know how to speak truly about these things.
Although the conditional clause at the end of the fragment may undermine
the speaker’s credibility, the assertion of old age seems to be a positive,
competitive gesture. Xenophanes is reported elsewhere as having commented
on the alleged 154 years reached by the rival sage Epimenides (D5=B20),
perhaps an incredulous or sarcastic remark given the satirical stance of the
Silloi. The notion that ‘old age’ brings wisdom is conventional (cf. Il. 19.218–-
19), but the specific feature of a primary narrator announcing his age as a basis
for his own authority is paralleled in particular among the other elegists.90 The
implication is that we should listen to what he says because it is based on vast
and varied experience, gained through his longevity and his travels around the
Greek world. Such an authority finds a basis in Xenophanes’ epistemology
(see Section 2a above): since there is no direct divine-to-mortal communica-
tion, the best we can do is to amass as many and as varied experiences as
possible, for instance, tasting myriad foods in order to reach a more precise
understanding of how sweet a thing is. Accordingly, Xenophanes here adver-
tises the breadth of his experience as he does when reporting the beliefs of
distant peoples, the Ethiopeans and Thracians (D13=B16).
If travel is the best way for mortals to acquire something approaching
knowledge, Xenophanes’ poetry in a sense acts as a substitute for this
process. The poetic utterance is itself a journey in D64=B7.1: ‘Now,
then, I will proceed to another argument, and I will show the way’ (νῦν
αὖτ’ ἄλλον ἔπειμι λόγον, δείξω δὲ κέλευθον). As Lesher writes, ‘Both αὖτε
(“again”, “next”) and ἄλλον (“another”) suggest that this is but one of
several stories being told, hence “yet another”.’91 Such a transitional
90
Solon fr. 18; Simonides fr. 89 W and fr. 28 FGE. Even if these fragments are not genuinely
Simonidean, they attest to the fact that he was thought of as being an aged figure within the
broader tradition of antiquity.
91
Lesher (1992) 78.
expression finds a number of parallels in early hexameter,92 but perhaps
closest of all is the formulaic line found at the end of several Homeric
Hymns, σεῦ δ’ ἐγὼ ἀρξάμενος μεταβήσομαι ἄλλον ἐς ὕμνον (e.g. Ven. 293),
‘and I having started from you will move over to another song’, which, like
the Xenophanes fragment, employs the familiar metaphor of the ‘path of
song’:93 both the Xenophanean and the hymnic narrator present them-
selves as ‘embarking on’ or ‘moving onto’ another utterance. Xenophanes’
line may be an adaptation specifically of the hymnic formula.94 But in the
Homeric οἴμη and the hymnic formula, the song itself is the journey on
which the performer embarks, whereas here there is the sense that the
audience are also travellers on the journey: the hymnic line addresses
a deity, and announces a transition to a separate topic, but Xenophanes
seems to address us, the mortal audience, and will ‘show’ us the ‘way’,
implying that it is a way upon which we ourselves should embark: our
experience of receiving this text is thus figured as a journey in which the
well-travelled Xenophanes is our guide.
While his epistemology had undermined Muse-inspiration, in this case
a further traditional means of asserting poetic authority, more characteris-
tic of elegy than of hexameters, is vindicated, that is, by claiming to be
experienced and well travelled. Specifically, Xenophanes in D66=B8 and
D64=B7.1 conforms to the more widely attested type of the ‘wandering
poet’, a type which also included the elegists, Mimnermus, Semonides and
Tyrtaeus. I shall argue that this type can inform our understanding of the
function of the two longest surviving passages of Xenophanes, D59=B1 and
D61=B2, as well as our reading of the imagery of space and travel in some of
the other elegiac fragments.
a Xenophanes as a ‘Wandering Poet’

The subject of ‘wandering poets’ in Greek antiquity has been the focus of
an important volume edited by Richard Hunter and Ian Rutherford, who
take their cue from an observation by the anthropologist Mary Helms,
92
Note Epigoni fr. 1 West, Νῦν αὖθ’ ὁπλοτέρων ἀνδρῶν ἀρχώμεθα, Μοῦσαι, which West (2003b)
9–10 suggests is a transitional expression from the Thebaid; Hesiod Op. 106 Εἰ δ’ ἐθέλεις, ἕτερόν τοι
ἐγὼ λόγον ἐκκορυφώσω (where West [1978] ad loc. cites the Xenophanes line as a parallel).
93
Note the etymology of οἴμη, ‘song’ from οἶμος, ‘path’ or ‘way’ (thus LSJ, Beekes [2010], Chantraine
[1968] s.v.) and the exhaustive examples of journeying as a poetological metaphor in Nünlist (1998)
228–83 and as a metaphor more generally in Becker (1937).
94
With Collins (2004) 147–51.
50 Xenophanes
‘that things, information, and experience acquired from distant places,
being strange and different, have great potency, great supernatural power,
and if attainable, increase the ideological power and political prestige of
those who acquire them. Such attitudes underlie the activities of travellers
and the influences accorded those who, as shaman-curers/scholars/priests/
traders, may arrive at a given locale as learned and experienced “wise
strangers from afar.”’95
According to this principle, the outsider status of a traveller such as
Xenophanes could contribute to his ‘ideological power and social prestige’.
Helms’ description offers an explanation for the ubiquity of the metaphor
of the journey,96 and could moreover apply to the stories surrounding the
lives of several early Greek poets, legendary wise men and mythical heroes.
Orpheus came from Thrace, participated in the Argonautic expedition
and, most famously, went as far as Hades and back. Aristeas of
Proconnesus visited the Issedonians, the one-eyed Arimaspeans and the
Hyperboreans in the far north. Epimenides came from Crete and purified
Athens of the pollution from the massacre of the supporters of Cylon.97
Such figures will provide a context against which to understand the
characterisation of Parmenides and Empedocles’ itinerant primary narra-
tors. Especially pertinent to Xenophanes, however, are the travelling
elegists who composed poetry recounting the foundation of, or offering
constitutional advice to, particular poleis.98 Mimnermus, from Ephesus,
seems to have travelled (albeit not very far) to Smyrna and composed
a poem on its foundation. Semonides of Amorgos is credited with
a poem on the foundation of Samos. A well-attested tradition held that
Tyrtaeus was originally from Athens, but the Spartans summoned him on
the advice of an oracle to provide them with counsel in their war with the
Messenians. Solon advised the Cypriot king Philocyprus to move the city
of Aepeia, upon which the latter promptly followed his advice and
95
Helms (1988) 263, quoted by Bachvarova (2009) 37 in Hunter and Rutherford (2009).
96
See also D. Fowler (2000) 208–15 on this metaphor in ancient didactic poetry. Note the famous
example of Hes. Op. 286–92, a passage whose rich ancient reception history is well discussed by
Hunter (2014) 92–100. Further examples are collected by Becker (1937). See also, more generally,
Montiglio (2000) (which surprisingly does not mention Xenophanes) and (2005), where it is argued
(pp. 100–1) that Xenophanes restricted his ‘wandering’ to Sicily, but D66=B8 suggests he wandered
further afield around the Greek world.
97
Orpheus: e.g. Eur. Alc. 357–9, 962–71 (as Thracian who underwent a katabasis); Simon. 567; Pind
Pyth. 4.176ff. (as Argonaut). Aristeas: Hdt. 4.13–16. Epimenides: e.g. Aristotle, Athenaion Politeia 1.1;
Pl. Leg. 1.642d. For such figures, see West (1983) 3–7, 39–61 and, on Aristeas and Epimenides, the
entries in BNJ.
98
For the genre of historical elegy, see n. 11 above. See also d’Alessio (2009) for the phenomenon of
local identities being defined by foreign poets and, more generally, Martin (1993) for the tendency of
Greek sources to present sages as offering practical, political advice.
renamed it Soloi in the Athenian’s honour. This is the context for Solon fr.
19 W, which wishes Philocyprus well whilst at the same time advertising
the author’s alien status by praying for a safe nostos.99 Whether or not we
accept the historicity of all these details, their recurrence, both in the
internal evidence of certain surviving fragments and in the ancient bio-
graphical traditions, attests to the familiarity of this pattern – of wise men
in general, but elegists in particular, travelling to an alien polis to provide
advice and praise – within the Greek imagination.100
The stories about Xenophanes clearly conform to this pattern: having
travelled westwards across the Greek world, he composed a lengthy elegy on
the foundation of Elea and, according to the testimonia, travelled to Zancle in
Sicily and spent time at the court of Hieron of Syracuse (Plut. Reg. et imp.
apophth. 4 175C=P13=A11). Diogenes Laertius even claims that he was thought
to have been sold as a slave (9.20), a detail which further contributes to the
picture of him as an authoritative outsider, just as Aesop and Plato (Diog.
Laert. 3.19) were at times said to be slaves.101 Aristotle preserves an anecdote
which shows that Xenophanes was not only thought of as a celebrant of these
destinations, but also a source of civic counsel (Rhet. 2.23.1399b6–8=P17=A13).
The citizens of Elea reportedly asked him whether they should sacrifice to
Leucothea, to which he responded in a manner consistent with his celebrated
rejection of anthropomorphic deities: ‘if they thought she was a goddess they
should not mourn her and if they thought she was human they should not
sacrifice to her’.102 While these anecdotes may be of little value for piecing
together Xenophanes’ actual biography, they nevertheless draw upon details
found in the fragments themselves and attest to how his narratorial persona
was received later in antiquity. His travels are most conspicuous in D66=B8
but are also implicit when we consider the poem on the foundation of Elea, at
the extreme West of the Greek world, in the context of his Eastern,
Colophonian identity.103 The poet’s travels are not merely a potentially
99
Mimnermus’ Smyrneis: frr. 13, 13a, with West (1974) 74 and Bowie (1986) 28–30. Tyrtaeus as an
Athenian: Plato Leg. 1.692a–b, with schol. ad loc.; Lycurg. in Leocr. 106; Philodemus de Mus. 4 Col.
73 Delattre; Diod. Sic. 8.27.1–2; Paus. 4.15.6; Plut. apophth. Lac. 230d; with d’Alessio (2009) 154–5.
On Solon fr. 19 and the tradition of itinerant sages, see Irwin (2005) 147–54.
100
On the methodological principle of using ancient biography as a source of ancient interpretations
of authors, even though their factual content may be incredible, see Introduction n. 2 above.
101
For an investigation of the figure of Aesop within the context of itinerant sages, see Kurke (2010)
95–158. See also Martin (1992) for depictions of the outsider, or metanastes, as a figure of didactic
authority.
102
For an analysis of the argument of this anecdote, see Warren (2013).
103
His Colophonian identity underlies D62=B3, which criticises the Colophonians for acquiring an
indulgent taste for luxuries from the Lydians, and also D54=B22, which gives advice to an Ionian,
probably Colophonian, addressee on appropriate sympotic conversation.
52 Xenophanes
spurious detail from later biographies of the poet but a cultivated feature of the
persona loquens.
One feature of the wandering sage-poet is that he provides civic advice to
ensure a harmonious community. Foreign poets are often represented as
acting as mediators in times of civic strife.104 Such a mediator from outside of
the polis had the benefit of impartiality. Furthermore, sage-poets tend to
offer advice on the foundation of new colonies. Thus, Solon helped with the
foundation of Soloi, and Xenophanes, in Aristotle’s anecdote, is presented as
helping with the establishment of Elea. Such new colonies required the
institution of new cults and festivals.105 Xenophanes may have been thought
to have assisted in this process at Elea precisely because of his poem on the
foundation of the colony. D61=B2 and D59=B1 can be regarded as part of the
civic advice delivered by the wandering sage-poet, designed to entrench
particular values that help to forge a harmonious community.
b D61=B2
ἀλλ’ εἰ μὲν ταχυτῆτι ποδῶν νίκην τις ἄροιτο

ἢ πενταθλεύων, ἔνθα Διὸς τέμενος
πὰρ Πίσαο ῥοῆισ’ ἐν Ὀλυμπίηι, εἴτε παλαίων
ἢ καὶ πυκτοσύνην ἀλγινόεσσαν ἔχων,
εἴτε τι δεινὸν ἄεθλον ὃ παγκράτιον καλέουσιν, 5
ἀστοῖσίν κ’ εἴη κυδρότερος προσορᾶν
καί κε προεδρίην φανερὴν ἐν ἀγῶσιν ἄροιτο
καί κεν σῖτ’ εἴη δημοσίων κτεάνων
ἐκ πόλεως καὶ δῶρον ὅ οἱ κειμήλιον εἴη·
εἴτε καὶ ἵπποισιν, ταῦτά κε πάντα λάχοι – 10
οὐκ ἐὼν ἄξιος ὥσπερ ἐγώ. ῥώμης γὰρ ἀμείνων
ἀνδρῶν ἠδ’ ἵππων ἡμετέρη σοφίη.
ἀλλ’ εἰκῆι μάλα τοῦτο νομίζεται, οὐδὲ δίκαιον
προκρίνειν ῥώμην τῆς ἀγαθῆς σοφίης.
οὔτε γὰρ εἰ πύκτης ἀγαθὸς λαοῖσι μετείη 15
104
See d’Alessio (2009) 155–6, who cites anecdotes about Tyrtaeus, Terpander, Thaletas, Stesichorus
and Pindar.
105
See Dougherty (1993) 21–4, BNP s.v. apoikia and, most extensively, Malkin (1987) 135–86. The
oikistes, the founder of the colony, is often depicted as determining some of these matters, such as
the establishment of temples and cults. Note the examples of the story of Nausithous’ foundation of
Phaeacia at Od. 6.7–10 and Pindar’s account of Battus’ establishment of altars and processions as
part of the foundation of Cyrene (Pyth. 5.89–93). More generally, we hear reports of the legendary
law-givers, Zaleucus and Charondas, establishing constitutions for the colonies of Magna Graecia
(Pl. Resp. 599d–e; Arist. Pol. 2.12.1274aff.). Charondas seems to have been roughly contemporary
with Xenophanes (see OCD4 s.v.).
οὔτ’ εἰ πενταθλεῖν οὔτε παλαισμοσύνην,
οὐδὲ μὲν εἰ ταχυτῆτι ποδῶν, τόπερ ἐστὶ πρότιμον
ῥώμης ὅσσ’ ἀνδρῶν ἔργ’ ἐν ἀγῶνι πέλει,
τοὔνεκεν ἂν δὴ μᾶλλον ἐν εὐνομίηι πόλις εἴη.
σμικρὸν δ’ ἄν τι πόλει χάρμα γένοιτ’ ἐπὶ τῶι, 20
εἴ τις ἀεθλεύων νικῶι Πίσαο παρ’ ὄχθας·
οὐ γὰρ πιαίνει ταῦτα μυχοὺς πόλεως.
But if someone were to win a victory through swiftness of feet
Or through competing in the pentathlon, where there is the precinct of Zeus
By the streams of Pisa in Olympia, or wrestling
Or, further, participating in the painful art of boxing,
Or even in the terrible competition which they call the pankration, 5
Then, for the citizens, he would be more glorious to look upon
And he would win a conspicuous, privileged seat in the contests
And there would be food out of the people’s property
From the city and a gift which would be an heirloom for him;
Or also with horses, he would attain all these things – 10
Not being worthy as I am. For better than the strength
Of men and horses is our wisdom.
But this custom is completely random, nor is it just
To prefer strength to good wisdom.
For not if there were a good boxer among the people 15
Nor if he were good at the pentathlon, or in wrestling,
Nor indeed with swiftness of foot – which is the most honoured
Of all the deeds in the contest of men’s strength –
Would the city therefore be more in good order.
The city would gain little pleasure from him, 20
If someone competing wins a victory by the banks of Pisa
For these things do not fatten the city’s cellars.
It is unclear whether we have the start of the poem. The adversative ἀλλ’
appears to assume that something has come before the first sentence.
However, other Archaic elegiac fragments that may contain the openings
of the poem begin with such connectives,106 and the fragment, as Faraone
has shown, is roughly symmetrical: lines 13–22 act as a counterpoint to 1–12
and form a ring structure with the list of sports recurring in reverse order in
the latter section, albeit with the absence of the pankration.107 Moreover,
the opening conditional clause, listing the rewards of athletic victors and
106
Cf. e.g. Tyrtaeus frr. 10, 11 W; Solon fr. 15 W.
107
A detail which leads Faraone (2008) 124–5, following Weil (1862), to posit a lacuna before line 15.
Such a lacuna would be paralleled in Athenaeus’ other quotations and is supported by the fact that
that the pankration, unlike the other sports that are mentioned in the first stanza, does not recur in
ring structure in the second.
54 Xenophanes
capped with the claim that they are not so worthy as the narrator’s wisdom,
functions as a priamel, a standard means of opening an Archaic Greek
song.108 The fragment as we have it, then, may be a complete (or nearly
complete) song.
The priamel can be a means of presenting an idea as running counter to
cultural expectations: a thesis is presented in contrast to popular assump-
tions, which then has to be demonstrated in the following passage. In the
most famous example, Sappho fr. 16, the narrator’s thesis, that ‘what one
loves’ is the finest thing, is presented in contrast to opinions of others who
believe the finest thing to be cavalry or infantry or a naval fleet; this then
has to be demonstrated in the subsequent stanza with the example of
Helen.109 But a closer parallel to Xenophanes, both metrically and themat-
ically, is the opening of Tyrtaeus’ lengthy fr. 12 W, where the narrator
‘would not mention or take account of a man’ (οὔτ’ ἂν μνησαίμην οὔτ’ ἐν
λόγωι ἄνδρα τιθείην, line 1) for prowess in running or wrestling, or for size,
strength, speed, beauty, richness, regality, eloquence or for reputation for
anything (lines 1–9) except furious valour (θούριδος ἀλκῆς, line 9); for
a man is not good in war unless he can bear to see bloody slaughter and
lunge at the enemy whilst standing close in the ranks (lines 10–12).110 This,
says Tyrtaeus, is virtue and the best and finest prize for a young man (ἥδ’
ἀρετή, τόδ’ ἄεθλον ἐν ἀνθρώποισιν ἄριστον/ κάλλιστόν τε φέρειν γίνεται
ἀνδρὶ νέωι, lines 13–14), and the brave man is a common good to the city
and to all the people (ξυνὸν δ’ ἐσθλὸν τοῦτο πόληΐ τε παντί τε δήμωι, line
15). Like Xenophanes, Tyrtaeus delineates the particular virtue that he
claims to be most valuable to the polis in opposition to commonly praised
attributes. Indeed, some of those attributes (strength, speed, appearance;
the wrestling mentioned by Tyrtaeus is a similar sport to Xenophanes’
pankration) are the same as those of Xenophanes’ Olympic victors. It is
possible that Tyrtaeus’ passage came from the song that would become
known, from at least the time of Aristotle, as the Eunomia (frr. 1–4 W),
given that its apparent function is in part to establish good conventions for
a polis.111 Xenophanes’ song, too, ostensibly provides principles under
which a polis can be in a situation of eunomia or ‘good order’ (line 19).
108
On the form of the priamel, see Race (1982) (who discusses this fragment as an example, pp. 59–62)
and Bundy (1962) passim.
109
The contrast is marked with the extended μὲν . . . δὲ clause (lines 1–3 ο]ἱ μὲν . . . οἱ δὲ . . . οἱ δὲ and
ἔγω δὲ). See Page (1955) 55–6. Note the other similar example of Pind. Ol. 1.1ff.
110
The similarity was noted by Jaeger (1932), who argued that Xenophanes was clearly influenced by
Tyrtaeus; Marcovich (1978) 24–5 is more sceptical of the possibility.
111
On this function of the Eunomia, see d’Alessio (2009) 150–6.
The parallels are such that direct allusion is possible.112 Xenophanes
D61=B2 could be seen as a ‘correction’ or updating of Tyrtaeus’ position:
the erroneous popular praise does not go to the more general attributes
listed by Tyrtaeus but to the victors of the Olympic games. Even if we rule
out the possibility of direct allusion and explain the similarities as due to
shared usage of a common form, the Tyrtaeus passage is a close typological
parallel and can be taken as evidence of the way in which Xenophanes
could have used this form but chose not to. The virtue that is most
beneficial to the polis, that provides eunomia, is not bravery but wisdom.
It is notable in this regard that the exempla that Tyrtaeus uses come from
myth – the strong man ‘has the size and strength of the Cyclopes’; the swift
one ‘beats Thracian Boreas’ – whilst those of Xenophanes are Olympic
victors. His less elevated tone thus imbues his sentiments with a greater
sense of realism and is consistent with the idea that mythical monsters are
mere ‘fabrications’ (D59=B1.23).
The typological parallel with Tyrtaeus can be taken further. As we have
seen, both poets were regarded as foreign arrivals to the cities they
advised. Tyrtaeus’ poems seem to have played a role in the development
of Sparta’s reputation as a city that prided itself on its martial success.113
Within this context, Xenophanes’ poem can be read as performing
a similar role in the entrenchment of a particular ideology in the polis
to which he arrived. The first-person singular of D59=B1.11 has been
taken as evidence that this fragment presents a distinctive Xenophanean
voice, whilst the first-person plural at 1.12 is often translated as a plural for
singular.114 However, under re-performance, perhaps at Eleatic symposia,
any singer could assume the identity of the narrator.115 The lack of
a sphragis, or any specific autobiographical details, would facilitate this
possibility.116 The shift from singular ἐγώ to plural ἡμετέρη σοφίη seems
112
For a defence of such intertextual readings in early Greek elegy, see Irwin (2005) 19–29, contra
R. Fowler (1987) 20–33.
113
On this issue, see Irwin (2005) 19–34, who stresses the generic, widely applicable nature of Tyrtaeus’
sentiments, but even so, from the perspective of his ancient reception, his martial exhortations
certainly facilitated Sparta’s distinctive reputation as a warrior society.
114
It is translated as singular, e.g. by Gerber’s Loeb edition of early Greek elegy (1999). Ford (2002)
50–2 sees the fragment as an advertisement for Xenophanes’ wisdom.
115
On this feature of the performance of sympotic poetry, see, most recently, Bakker (2017a). On the
question of the ‘I’ in Archaic lyric more generally, note the nuanced discussion of Slings (1990).
Xenophanes B2 fits into the third type of personal ‘I’ distinguished by Slings, in which the ‘I’
expresses an opinion, but where no appeal to action is expressed, in which cases, Slings argues, the ‘I’
is supposed to be representative of a group.
116
Contrast Theognis’ Sphragis (19–30); Solon’s autobiographical statements at frr. 1, 5, 19, and his
geographical references that mark the narrator as Athenian at frr. 2 and 4.
56 Xenophanes
significant: ‘our wisdom’ could assert the collective wisdom of the sym-
potic group, or even the polis as a whole. Such a significant shift would
appear to be paralleled elsewhere in elegy.117
Given the Panhellenic significance of the Olympic games, the fragment
could serve to define the identity of a particular polis against that of other
Greek states. One might even argue that the poem achieved some measure
of success in the promotion of the value of wisdom at Elea, since the city
would become most famous for the ‘Eleatic school’ of philosophy, of which
Xenophanes was known as the founder.118 Modern scholars deny that this
should be taken to mean that Xenophanes founded a formal institution,
but he certainly influenced the famous philosophers from the city.119 At
any rate, the poem concerns both the narrator’s status as a wise man and
promotes particular civic values in the very literal sense of what a particular
polis should value (lines 6–9). In this respect, it conforms to the type of the
authoritative wanderers identified above.
c D59=B1
D59=B1 presents an image of sympotic song, but although the sympo-
sium is often thought of as an inherently confined locus for an elite
hetaireia (‘brotherhood’ or ‘male clique’),120 Xenophanes’ poem can also
be read as offering civic advice more generally, in particular on the
avoidance of stasis.
νῦν γὰρ δὴ ζάπεδον καθαρὸν καὶ χεῖρες ἁπάντων
καὶ κύλικες· πλεκτοὺς δ’ ἀμφιτιθεῖ στεφάνους,
ἄλλος δ’ εὐῶδες μύρον ἐν φιάληι παρατείνει·
κρατὴρ δ’ ἕστηκεν μεστὸς ἐυφροσύνης,
ἄλλος δ’ οἶνος ἕτοιμος, ὃς οὔποτέ φησι προδώσειν, 5
μείλιχος ἐν κεράμοισ’ ἄνθεος ὀζόμενος·
117
Cf. Gagné (2013) 226–49, who argues that the shift from singular to plural first-person statements in
Solon fr. 13 W (the Elegy to the Muses) marks a shift from individual to group sentiments. On the
significance of σοφίη as ‘wisdom’, rather than more specifically poetic skill, see Marcovich
(1978) 21–2.
118
Pl. Soph. 242d=R1<A29, Clem. Al. Strom. 1.64.2=R3<DKA8, Theod. Cur. 4.5=R11<DKA36.
119
To list some well-known correspondences: in Parmenides D8>B8, what-is is described as singular
(D8.11=B8.6 ἕν), just as Xenophanes’ god is εἷς (D16=B23); it is οὐλομελές (D8.9=B8.4), as
Xenophanes’ god is οὖλος (D17=B24); Xenophanes’ god does not move (D19=B26), just as
Parmenides’ entity is ἀκίνητον (D8.32=B8.26); like Xenophanes’ god, it is defined largely by what
it is not: in D8>B8, nine lines begin with negatives that negate attributes of what-is, similarly to the
negatives in D16=B23.2 and D19=B26.2. For scepticism over Xenophanes’ ‘Eleatic school’, see e.g.
KRS 165–6; Graham (2010) 95.
120
Thus, Ford (2002) 65; Murray (1993) 207–13; Stehle (1997) 213–62.
ἐν δὲ μέσοισ’ ἁγνὴν ὀδμὴν λιβανωτὸς ἵησι·
ψυχρὸν δ’ ἔστιν ὕδωρ καὶ γλυκὺ καὶ καθαρόν·
πάρκεινται δ’ ἄρτοι ξανθοὶ γεραρή τε τράπεζα
τυροῦ καὶ μέλιτος πίονος ἀχθομένη· 10
βωμὸς δ’ ἄνθεσιν ἀν τὸ μέσον πάντηι πεπύκασται,
μολπὴ δ’ ἀμφὶς ἔχει δώματα καὶ θαλίη.
χρὴ δὲ πρῶτον μὲν θεὸν ὑμνεῖν εὔφρονας ἄνδρας
εὐφήμοις μύθοις καὶ καθαροῖσι λόγοις·
σπείσαντας δὲ καὶ εὐξαμένους τὰ δίκαια δύνασθαι 15
πρήσσειν – ταῦτα γὰρ ὦν ἐστι προχειρότερον –
οὐχ ὕβρις πίνειν ὁπόσον κεν ἔχων ἀφίκοιο
οἴκαδ’ ἄνευ προπόλου μὴ πάνυ γηραλέος.
ἀνδρῶν δ’ αἰνεῖν τοῦτον ὃς ἐσθλὰ πιὼν ἀναφαίνει,
ὥς οἱ μνημοσύνη καὶ τὸν ὃς ἀμφ’ ἀρετῆς. 20
οὔτι μάχας διέπειν Τιτήνων οὐδὲ Γιγάντων
οὐδέ <τε> Κενταύρων, πλάσματα τῶν προτέρων,
ἢ στάσιας σφεδανάς, τοῖσ’ οὐδὲν χρηστὸν ἔνεστι·
θεῶν <δὲ> προμηθείην αἰὲν ἔχειν ἀγαθόν.
For now indeed the floor is pure and the hands of all
And the cups; one person puts on woven garlands,
And another proffers the sweet-smelling unguent into the saucer;
The mixing-bowl stands full of good cheer,
And another wine is ready, which says it will never betray us, 5
A soothing one in the clay jars, smelling of flowers;
And in the midst incense sends forth its holy smell;
And the water is cool and sweet and pure;
And yellow loaves lie nearby and the table is majestic
Loaded with cheese and fat honey; 10
And the altar, in the middle, is covered thickly with flowers,
And song and celebration surround the house.
And it is necessary first for well-minded men to sing of the god
With auspicious speeches and pure words;
And having poured libations and having prayed to be able to accomplish
just things, 15
(For these are more appropriate)
It is not insolent to drink so much that, being in such a state, you can
still come
Home without a guide, if you are not completely old.
And, of men, praise the one who, when drinking, reveals noble thoughts
According to his memory, and the one who [sc. speaks] about virtue. 20
And do not go through the fights of Titans or Giants
Or Centaurs, fabrications of earlier generations,
Or violent civic strife, in which there is no use at all.
But always have good forethought about the gods.
58 Xenophanes
The fact that this fragment begins in medias res – with the causal γάρ
lacking a protasis – has led some to suggest that we lack the opening of the
poem. Such a beginning, however, seems paralleled elsewhere in elegy,121
and could be rhetorically effective as the opening of a poem: the initial
deictic νῦν has the inclusive effect of inviting the listeners to imagine
themselves as already part of the sympotic group once the song has
begun. The effect would be enhanced if the audience has in fact recently
cleansed their hands in preparation for the symposium.122 As in the case of
B2, Faraone has demonstrated that this fragment is divided into two
roughly equal parts: the first half is descriptive in form whilst the second
is parenetic; the linguistic purity called for in the latter acts as
a counterpoint to the physical and ritual purity described in the
former.123 Whether or not we accept Faraone’s wider thesis that early
Greek elegy was composed in stanzas, the fragment displays a neat, sym-
metrical structure.
Now, the voices of sympotic narrators are often distinctive: the poetic
ego of Alcaeus or Solon is a specific figure in particular historical and
geographical circumstances; Anacreon and Theognis may be more generic
and imitable, but the former is strongly associated with lustful and bibu-
lous behaviour, the latter with aristocratic ideology.124 This song, however,
lacks any first-person statements or descriptions to specify its narrator. As
with other sympotic songs, the rhetoric seems designed to enhance the
sense of solidarity among the symposiasts,125 but the positive details speci-
fied are generic. The mixing cup, a potential source of disharmonious
drunkenness, is full of ‘good cheer’. An anonymous participant, probably
a slave, accommodatingly offers perfume in a small vase and even the wine
seems amicable, saying, in its personification, that it will not ‘give up’ –
προδώσειν, a verb which may also evoke its other sense of ‘betray’,126
a possibility which would further emphasise the harmony of the group.
121
See n. 106 above.
122
See Hobden (2013) 22–65 for a study of the ‘metasympotic’ nature of the symposium, that is, the
way in which, ‘representations of the symposion at the symposion in song and image construct
visions of the event that set out a persuasive view of sympotic performance and communicate ideas’
(22). As she observes, ‘metasympotic songs afforded symposiasts opportunities for self-styling, and
for the styling of others too’.
123
Faraone (2008) 116–20, contra Bowra (1938) 353 and Marcovich (1978) 4.
124
Cf. similarly, d’Alessio (2009) 151 on Tyrtaeus. On the principle that the personae of Archaic Greek
poetic narrators are typically distinctive, see Morrison (2007) 57–67.
125
This truism of the symposium and its poetry has been studied in detail by e.g. Rösler (1980b); part 1
of Schmitt Pantel (1992); Stehle (1997) 213–62; Hobden (2013) 117–56.
126
Ford (2002) 57 identifies such a pun, citing the Attic skolia 907 and 908 PMG as parallels, and
pointing out that ἕτοιμος, which can mean ‘trusty’ (see LSJ s.v.) also has such a personifying force.
When read within a civic context, ἐν δὲ μέσοισ’ takes on a particular
significance: it can refer not only to the literal, spatial centre of the
group, but also to the symbolic ‘middle ground’ where different armies
or political parties meet, frequently the location for rhetorical or martial
contest.127 That such an area should be suffused with a ‘holy odour’ may
therefore symbolise civic harmony. The description appeals to the different
senses, drawing on smell (εὐῶδες . . . ἁγνὴν ὀδμήν), taste (γλυκύ . . .
μέλιτος), touch (ψυχρόν) and sound (μολπή) as well as setting out
a visual image of the scene. The andrōn (‘male banqueting-hall’) almost
becomes a locus amoenus with cool water and flowers. There is nothing in
this evocative opening description to distinguish the group with any
particularity: by contrast with many other sympotic poems, it does not
define the sympotic group against any other faction.128 Potentially any free
male could picture himself participating in the scene.
By contrast, the second half of the fragment introduces the iconoclastic
criticism of certain traditional subjects for song. The form reflects that this
is a novel sentiment, for the gnomic instructions (χρή with the infinitives)
leave open the possibility that they are not yet being followed: Xenophanes
tells us that it is necessary to hymn the god; he does not describe this as
already being done. Although the stanza features the conventional senti-
ment that symposiasts ought not to drink too much, its main emphasis
concerns the appropriate subject matter for sympotic song. This emphasis
is created first by the way in which the details outlined in the first stanza
conclude with μολπή, as the first word of the final line and then by the fact
that the first gnomic statement that opens the second half concerns
hymning the god (line 14) and finally by the fact that song, and positive
speech in general, occupies the majority of the second half of the fragment.
Whilst the descriptions of the first half and the instructions concerning
justice and drinking (18–19) appeal to conventional sympotic sentiments,
the exclusion of stories of titans, giants and centaurs – occasional subject
matter for hexameters – seems novel. This novelty is illustrated by the
expression πλάσματα τῶν προτέρων (D59=B1.23), which, as Ford writes,
‘invidiously adapts a traditional epic expression for oral traditions about
127
Cf. LSJ s.v. μέσος IIIb; see further Barker (2009) 17–19. The political connotations of Xenophanes’
use of this term are well discussed by Svenbro (1984) 83–91.
128
Many of Alcaeus’ sympotic fragments seem designed to unite the sympotic group against a political
opponent (frr. 50, 70, 72, 332; perhaps also 170, 206, 207, 335, 338 L.-P.), whilst the elegies of the
Theognidean corpus characteristically define the sympotic group as aristocratic against others, often
those who may be wealthy but ignoble (e.g. 27–38, 39–52, 53–68, 69–72, 101–12, 149–50, 153–4,
161–4, 183–92, 193–6, 227–32, 283–6, 287–92, 333–4, 341–50, 367–70, 373–400, 523–6 etc.).
60 Xenophanes
the glorious deeds of heroes, “fames of men of long ago” (κλεῖα προτέρων
ἀνθρώπων)’.129 The rhetoric of this fragment, however, functions by
showing that the novel idea is in fact an entailment of conventionally
accepted sentiments. The recurrence of the vocabulary of purity
(καθαρόν . . . καθαροῖσι λόγοις) and ‘good-thinking’ (ἐυφροσύνης . . .
εὔφρονας ἄνδρας) implies that the pure speech is entailed by the conven-
tional understanding of sympotic purity outlined in the first half.
The fragment thus appeals to and manipulates accepted cultural
schemata to advance a novel precept concerning the subject matter of
song, a feature which is encapsulated by a pun in the final line. As
Collins points out, προμηθείην, whilst ostensibly used simply in its
literal sense of ‘forethought’, can be read as alluding to the name of
Prometheus.130 The allusion harnesses the myth of Prometheus to
illustrate the poem’s call to piety. Prometheus, at least in the
Hesiodic versions, was named ironically: he may have had greater
forethought than his brother Epimetheus, but he still could not
predict the deception of Zeus (Theog. 507–34; Op. 47–105). He may
therefore appear here as a negative paradigm: we must have the
forethought for the gods which Prometheus lacked and shun anything
that may be impious, including potentially impious speech.
The mythical figures invoked (Titans, Giants, Centaurs, Prometheus)
are those of Panhellenic myth and seemingly not of significance to a more
particular locale as we sometimes find elsewhere in elegy.131 Indeed, we
might wonder why these specific subjects should be excluded from the
symposium when, at least on the basis of the surviving evidence, they do
not seem to have been typical subjects for sympotic song. When read
within the wider context of Xenophanes’ criticism of the canonical
poems of Homer and Hesiod, the exclusion of these myths from the
Symposium seems to serve as a synecdoche for the wider-scale exclusion
of such narratives from the ideal polis, or even the Greek world as a whole.
The political vocabulary of betrayal (προδώσειν) and civic strife (στάσιας
σφεδανάς) and the middle ground (ἐν δὲ μέσοισ’) refer in the first instance
to the internal harmony of the sympotic group, but also evoke the risk of
wider-scale political unrest in the polis as a whole. The language of purity
also contributes to this picture, especially if we read this fragment alongside
D66=B8, since itinerant sages could provide purification for bloodshed
129 130
Ford (2002) 57–8. Cf. Theog. 100; Il. 9.189; hApollo 160. Collins (2004) 150.
131
Note Tyrtaeus’ references to the Spartans as descendants of Herakles (frr. 2, 11.1 W), and Solon’s
references to Athena as the patron goddess of Athens (fr. 4.1–4 W), and to Cypris in a poem on the
foundation of Soloi in Cyprus (fr. 19.4 W).
Xenophanean Elegies and Xenophanean Hexameters 61
arising from factional strife. Although the symposium is often the location
for an elite group that defines itself in opposition to a rival faction or
ideology, Xenophanes’ poem appropriates the political and ideological
connotations of sympotic elegy to advance a message of relevance to
a broader citizen body. A key aspect of that advice involves the avoidance
of stasis-inducing song.
4 Xenophanean Elegies and Xenophanean Hexameters

The longest elegiac fragments are inextricably embedded in the immediate
world of the polis: D59=B1 concerns the appropriate behaviour for avoiding
stasis in the symposium and, by extension, the polis as a whole, whilst
D59=B1 more explicitly pronounces a statement on what practices the polis
should value. In doing so, these fragments conform to the image of
Xenophanes’ elegiac persona as a ‘wandering sage-poet’ made most explicit
in D66=B8. When read within the wider context of Xenophanes’ epistem-
ology, this traditional model of authority gains some support: his travels
and longevity have provided varied experiences that allow him to make
judgements based on a wide range of data and to realise the cultural
contingency of certain Greek beliefs. But the epistemological implications
of what might be termed Xenophanes’ cultural relativism, at least in the
surviving fragments, seem confined to the hexameter fragments: it is only
in hexameters that Xenophanes expresses the impossibility of knowing
when one has ‘hit the mark’ with one’s speech. The assertions made in
the elegies, with the exception of the aged narrator’s self-doubt at
D66=B8.4, seem more confident.
This contrast between the hexameter and the elegiac perspectives is
evident from the elegiac D69=B6:
πέμψας γὰρ κωλῆν ἐρίφου σκέλος ἤραο πῖον
ταύρου λαρινοῦ, τίμιον ἀνδρὶ λαχεῖν,
τοῦ κλέος Ἑλλάδα πᾶσαν ἐφίξεται οὐδ’ ἀπολήξει,
ἔστ’ ἂν ἀοιδάων ἦι γένος Ἑλλαδικῶν.
For having sent the thighbone of a kid you received the rich leg
Of a fatted bull, a valuable thing for a man to attain,
Whose glory will reach the whole of Greece and will not cease,
For as long as there is a tribe of Greek singers.
The fragment seems to be addressed to a patron. In sending only the
thighbone of a kid, the patron has received good value for money in
attaining the leg of a fatted bull, most likely a metaphor for Xenophanes’
62 Xenophanes
song.132 Its kleos or ‘reputation’ will match Xenophanes’ own wanderings,
in travelling via the network of itinerant poets to ‘reach all Greece’ and to
persist ‘as long as there is a tribe of Greek singers’. This sentiment is closely
paralleled by Theognis 237–54 and may be conventional. In the Theognis
passage, the narrator’s song is said to have given Cyrnus wings, so that he
shall never lose his kleos (245) but shall ‘roam across the land of Greece and
among the islands’ (καθ’ Ἑλλάδα γῆν στρωφώμενος, ἠδ’ ἀνὰ νήσους, 247),
and will be a subject ‘for all who care about song, even for future gener-
ations, for as long as earth and sun exist’ (πᾶσι δ’, ὅσοισι μέμηλε, καὶ
ἐσσομένοισιν ἀοιδή/ ἔσσηι ὁμῶς, ὄφρ’ ἂν γῆ τε καὶ ἠέλιος, 251–2). In both
cases, the spatial extent of the song’s kleos travels specifically over the Greek
land and, chronologically, will endure for as long as there are Greek singers.
In comparison with Theognis, Xenophanes’ claims for the kleos of his song
are relatively humble. The relative humility of Xenophanes’ claim is even
more stark by comparison with an epic passage that may underlie both
passages, Od. 9.19–20, where Odysseus reveals his name to the conspicu-
ously non-Greek Phaeacians and is able to boast that his κλέος reaches
heaven (Od. 9.19–20), emphatically the only point in the poem at which
a character refers to their own κλέος in the first person and a claim that is
vindicated by the wider narrative, since the gods do indeed talk about
him.133 The kleos of Xenophanes’ poem, by contrast, seems limited to the
Greek world. Of course, the rhetorical function of this fragment depends
upon an emphasis on the magnitude of these chronological and spatial
limits. But what counts as a vast extent in this elegiac fragment (i.e. the
Greek world) is not so vast from the perspective of the world of epic, which
often adopts a perspective that goes beyond the boundaries of the Greek
world. More pertinently, it is not so vast from the perspective of
Xenophanes’ own hexameters (D14=B15; D13=B16), which reveal the con-
tingency of Greek beliefs. This tension between D69=B6 and Xenophanes’
cultural relativism is explicable in terms of the different functions of the
fragments – the hexameters are didactic and also satirical of certain beliefs;
this elegy is encomiastic – but the tension also derives from the different
perspectives typically assumed by epic and elegiac narrators.
132
Martin (2009) 97 sees Xenophanes’ ‘thighbone of a kid’ as abuse of the stinginess of his addressee,
but the emphasis falls less on the meagreness of the thighbone than on the greatness of the ‘fatted
bull’. I therefore prefer to see this as stressing the good value for money the patron has received.
133
On these lines see Segal (1983) and Goldhill (1991) 25–6. The relevance of this passage to Theognis is
pointed out by Campbell (1982) 363. Note the similarity of expression: both Odysseus and Cyrnus
have kleos and are (or shall be) a ‘concern’ (μέλω/μελήσεις) for men.
Conclusion 63
It is not only in their metre and performance context, then, that
Xenophanes’ hexameters and elegies participate in different genres. Epic
and elegiac perspectives occupy different epistemic statuses. Hexameter
poetry, in requiring the knowledge of the Muses to recount events prior to
any living or documented memory, demanded either a leap of faith or
a suspension of disbelief on the part of the bard and the audience, unlike
the elegies which, for the most part, concerned the familiar, present-day
world of mortals. The audience of Xenophanes’ D59=B1 could tell from
personal experience whether its descriptions conformed to typical or
possible sympotic practices. Poets of the mythological material typical of
hexameter poetry were restricted by clear and present realities to a far lesser
degree. This is one reason why hexameter accounts of gods and heroes
attracted allegorical interpretations throughout antiquity, whereas no such
extravagant interpretations are attested for elegies.134 The distance of epic’s
objects of reference allowed interpreters to posit a gap between what was
said and what was meant. This difference of epistemic status also explains
why the Xenophanean fragments classified by scholars as the most ‘philo-
sophical’ are the ones that are in hexameters: they inhabit the more
speculative realm of myth, even if their explanations no longer depend
on the arbitrary actions of anthropomorphic agents. In this respect, the
hexameter genre provided Xenophanes with a way of thinking about – and
encouraging his audience to think about – the big questions concerning the
origins of the universe, the nature of the divine and the limitations of
mortal understanding, whereas elegies provided a way to implement the
ethical consequences of those thoughts on a practical level.
5 Conclusion
A recent and extensive discussion of the issue has argued that ‘[v]erse . . .
was not essential to Xenophanes’ message’.135 This may be trivially true in
the sense that, like any verse text, it has semantic content that can be
paraphrased in prose. But Xenophanes’ metres and other formal features
participate in particular generic traditions through which they convey
a ‘message’ in at least three respects. First, they imply certain claims,
such as that the single god is worthy of hymnic veneration. Second, they
contribute to emotive effects, for instance, in D61=B2, by conjuring a sense
134
Though occasionally they use metaphors and fables that could be counted as ainigmata, such as the
ship of state (e.g. Theognis 667–82).
135
Granger (2007) 430.
64 Xenophanes
of injustice that athletic prowess is valued over wisdom. Third, they invite
certain types of interpretation. Whilst both sets of fragments present
distinctive ways of seeing the world, the elegies provoke imaginative
participation in the community described, whereas the hexameters invite
pensive reflection on the relationship between the statements made and the
real state of affairs. Unlike elegies, hexameters, partly due to the purport-
edly divine origins of their content, were treated as a genre that required
a symbolic interpretation through which an audience could confront the
divine truths underlying the literal text. Xenophanes capitalises on this
tendency by inviting us to reflect on the true reference of the verb γίγνομαι
and its cognates as well as on the ways in which a perfect god is or is not like
the anthropomorphic Zeus. This understanding of the function of
Xenophanes’ poetry counts against the view that he was ‘starkly literal-
minded’ or that for poets of this period, ‘the only validation of their poetry
is that it tells the truth, conforming veridically to a real past or present state
of affairs,’ unless ‘conforming to’ is taken in so broad a sense as to
encompass symbolic signification.136
As I shall argue in the subsequent chapters, the capacity of hexameters to
prompt questions concerning the relationship between the text, its trad-
itionally divine source of authority and its potentially symbolic referents
provides an explanation for why the genre was adopted by Parmenides and
Empedocles, over against prose or elegies that had been used by other
claimants to wisdom. But their choice must also have been influenced
more specifically by the example of Xenophanes, who expanded the
resources of the genre by using it to present physical explanations of divine
and natural affairs that were radically different from those he could have
expected of his audience, thereby inaugurating a tradition that would go on
to include some of the most influential poems of antiquity.
136
Most (1999) 343, commenting on Homer and Hesiod, but he sees this as a feature that they
bequeathed to the early Greek philosophers.
chapter 2
Parmenides
1 Introduction
While Xenophanes can be partly assimilated to the familiar category of
‘didactic elegists’, Parmenides’ use of verse has seemed more anomalous.
His single hexameter poem, depicting a fantastical journey in the first
person through the gate of the paths of Night and Day, where an
anonymous goddess reveals the true nature of reality, is quite unlike
any other substantially surviving work of early Greek literature.
Commentators have often been perplexed by Parmenides’ choice of
verse on the assumption that his poetry is obscure and of low literary
merit, betraying a wish that the reputed forefather of logic had written in
lucid prose.1 As we shall see, to some degree, Parmenides’ use of verse, like
that of Xenophanes, is indeed paradoxical: he presents overt claims that
are inimical to poetry as traditionally practised. But of course, any
complaint that he did not write in a manner characteristic of later
philosophical prose imports anachronistic expectations. A more charit-
able procedure is to treat the difficulties of Parmenides’ poetry as func-
tional, that is, as designed to elicit certain responses from the audience.2
The literary context in which Parmenides operated, as well as the overt
doctrinal claims presented in his poem, provide evidence for such
functions.
1
This wish is most clear in Guthrie (1965) 12, (‘the main content . . . would have been much better
conveyed in the plain prose of an Anaximenes or an Anaxagoras’) and Barnes (1982) 155. For negative
comments on Parmenides’ literary merits, see also Diels (1897) 7; Kranz (1916) 1163; Hussey (1972) 79;
Barnes (1982) 155; KRS 241; Kennedy (1989) 81.
2
Mackenzie (1982); Floyd (1992); Wöhrle (1993); Morgan (2000) 67–87; Barrett (2004); Robbiano
(2006); Mourelatos (2008a) 1–46; and Bryan (2012) 58–113 all adopt approaches along these lines.
I differ from them here in my emphasis on the immersive, aesthetic and emotional qualities of early
hexameter.
65
66 Parmenides
This approach differs from that of the most influential study of
Parmenides’ literary form, Alexander Mourelatos’ The Route of
Parmenides,3 which argued that the traditional structures of ‘questing’
and ‘journeying’ provided a basis for Parmenides’ concept of knowing:
‘straying off-course’ amounts to intellectual error and the traditional
‘fate-constraint’ of the epic hero becomes the newly discovered con-
cept of logical necessity. In suggesting that Parmenides used meta-
phors, not to encode preconceived ideas, but to think through new
concepts, Mourelatos’ 1970 study was remarkably ahead of his time,
anticipating, in some degree, later discussions of conceptual
metaphor.4 Nevertheless, my focus differs from that of Mourelatos,
primarily in two respects. First, his analysis relies heavily on
a comparison with the language and imagery of journeying in
Odyssey, but an expanded range of comparanda, both from within
the Odyssey and elsewhere, can help to identify the sort of journey
this is with greater precision.5 Second, and more fundamentally,
Mourelatos’ explanation focuses largely on the way in which trad-
itional patterns influenced Parmenides’ thought, saying relatively little
about how such patterns might affect an audience. Rather than seeing
Parmenides as simply constrained by tradition, we can instead regard
him as manipulating traditional features to provide certain experiences
and to provoke certain responses.
In this regard, it is worth mentioning a further, if somewhat eccentric,
trend in modern scholarship on the poem. The experience of Parmenides’
audience has been most extensively discussed by those who regard it as an
esoteric text. On this view, the poem ‘described a mystical experience and
above all aims through the power of language to induce this same experi-
ence in its listeners’.6 There are, however, a number of objections to the
3
Mourelatos (2008a).
4
I am of course thinking of Lakoff and Johnson (1980). For a useful introduction to this topic,
summarising earlier work, see Kövecses (2010). For literary-critical studies of metaphor as a means to
think through new concepts, see Lyne (2011) and Cave (2016) 83–105. Mourelatos himself drew on
Max Black’s influential work on metaphor.
5
Mourelatos (2008a) 41–4 does discuss ‘traditions other than epic’, but this section is mostly devoted
to the thorny issue of Parmenides’ relationship to shamanism and plays little role in his wider
argument. Meuli (1935) and Dodds (1951) 135–78 influentially argued that this phenomenon was
practised in Archaic Greece, but their theory is convincingly refuted by Bremmer (1983) 14–52 and
(2001) 27–40.
6
Gemelli Marciano (2008) 26, in summary of the views of Kingsley (1999) and (2003). Her interpret-
ation is presented more expansively in Gemelli Marciano (2013). West (1971) 222–3 and Ustinova
(2009) 191–209 similarly regard the poem as deriving from a religious experience, rather than rational
deduction. Tarán (1965) 26–30 discusses (and dismisses) earlier interpretations along these lines.
Poetry and Parmenides’ Philosophy 67
‘esoteric’ reading. The early reception history counts against such
a possibility: if the poem were really composed only for a select, esoteric,
group of followers, it seems unlikely that it would have swiftly achieved
such a widespread reputation across the Greek world as to be excerpted by
the fifth-century Sophist Hippias of Elis in his Sunagoge and then have
survived in the manuscript tradition to be quoted first-hand by Simplicius
in the sixth century ce.7 For what it is worth, Plato presents Parmenides as
an attendant at the Great Panathenaea, whose poem is already familiar to
the young Socrates (Parm. 127a7–b6, 128a–e).8 A more plausible means of
publication, then, would be rhapsodic performance at public festivals, as is
attested for Xenophanes’ performance of his poetry and for Empedocles’
Katharmoi, and as was common for hexameter poems longer than epigrams
or verse oracles.9 It was not only a self-defined, select group that would
have heard Parmenides’ poem, but a wider, festival-going public. Given
this context, Parmenides seems prima facie more likely to have envisioned
the sort of imaginative and emotional response to his poem that is evinced
for rhapsodic performances than the mystical hallucinations of the esoteric
reading.
A closer analysis of Parmenides’ narratives and imagery will support this
suggestion. But before going into this in detail, I want to explore some of
the implications of the goddess’s claims for poetry, for her arguments
provide some clues as to the purposes of Parmenides’ use of epic verse. In
this sense, we can understand the function of Parmenides’ poetry in light of
his philosophy.
2 Poetry and Parmenides’ Philosophy

The poem is divided structurally into three sections. The proem (D4=B1)
describes the youth’s chariot journey to a location beyond ‘the gate of the
7
On Hippias’ Sunagoge, see Introduction Section 1 above. Simplicius quoted extensive passages of
Parmenides precisely because he had access to a rare copy of the poem (In Phys. 144.25–28=R73=A21).
8
Plato’s Parmenides, however, should be treated as a historical source with extreme caution, not least
since its primary narrator is far removed from the events reported: the dialogue is narrated by
Cephalus, who has heard the story of the conversation from a reluctant Antiphon, who himself has
heard it from Pythodorus, who had been present on the occasion.
9
Similarly, Patzer (2006) 85. On the rhapsodic performance of Xenophanes and Empedocles, see,
respectively, Chapter 1 Section 2 above and Chapter 3 Section 2 below. Schwabl (1963) argued, on the
basis of certain structural features, that Parmenides’ poem was rhapsodically performed. On the
character of Greek public festivals, featuring contests of wisdom as well as athletic and poetic
competitions, see Richardson (1992) 225.
68 Parmenides
paths of Night and Day’ (D4=B1.11), where the goddess greets him and
speaks, as far as we can tell, for the rest of the poem.10 She announces that
she will reveal ‘both the unshaken heart of well-rounded truth / and the
opinions of mortals in which there is no true trustworthiness’
(D4=B1.29–30), descriptions that correspond respectively to the second
and third sections of the poem, the so-called Aletheia (D6>B2–B8.48) and
the Doxa (D8.54=B8.49–D62=B19). The goddess distinguishes two ‘paths
of enquiry’ (ὁδοί . . . διζήσιός, D6>B2.2), one ‘that it is and that it is not
possible that it is not’ (ὅπως ἔστιν τε καὶ ὡς οὐκ ἔστι μὴ εἶναι, D6>B2.3);
the other, ‘that it is not and that it is necessary that it is not’ (ἡ δ’ ὡς οὐκ
ἔστιν τε καὶ ὡς χρεών ἐστι μὴ εἶναι, D6>B2.5). The latter ‘path of enquiry’
turns out to be a dead end, for ‘you could not recognise or point out what is
not (for it is impossible)’ (οὔτε γὰρ ἂν γνοίης τό γε μὴ ἐὸν [οὐ γὰρ
ἀνυστόν] / οὔτε φράσαις, D6>B2.7–8). Most mortals, however, falsely
suppose that they do recognise such a thing as ‘what is not’, and they mix it
up with ‘what is’ (D7=B6). In the most extended fragment (D8>B8), the
goddess argues for certain properties of ‘what is’: that it is uncreated,
indestructible, singular, unmoving and undifferentiated. However, for
what appears to be the final and longest section of the poem,11 she recounts
the Doxa, an elaborate cosmology in which all things are formed out of the
two elements ‘Night’ and ‘Light’. This section in some sense reflects the
‘opinions’ of mortals (D8.55=B8.50) and constitutes a ‘deceptive arrange-
ment of words’ (D8.56=B8.51).
The interpretation even of the goddess’s most basic propositional claims
is a matter of great scholarly controversy.12 It is agonisingly unclear what
she means by ‘(it) is’ and in what respects the Doxa is ‘deceptive’. Since my
aim is not to offer new solutions to these problems, I shall outline
somewhat dogmatically the interpretation here adopted, though it would
not be accepted by all specialists. The items in the Doxa violate the
conditions for ‘what-is’ outlined in the Aletheia: they move and mingle
with one another (e.g. D14=B12). But this does not entail that they do not
in some sense exist, that the Doxa is purely there as some kind of thought
experiment, or as an exemplar of previous, mistaken cosmologies. Rather,
10
Simplicius’ introduction to D62=B19 (where the goddess is still talking) implies that it is the end of
the Doxa, the third section of the poem. There does not appear to have been a fourth section,
although this is an argument ex silentio.
11
Palmer (2009) 160 estimates that this occupied 80 per cent of the original poem, on the basis of the
testimonies of Plut. Col. 1114B–C and Simplicius in Cael. 559.26–7, who stress the exhaustiveness of
the Doxa’s detail.
12
For a recent summary of different views on the matter, see Kraus (2013).
Poetry and Parmenides’ Philosophy 69
the Doxa describes ‘Doxastic things’, objects over which we can have
opinion, but which do not live up to the standards of Aletheia, a term
better translated as ‘true reality’ than ‘truth’.13 The Doxa describes accur-
ately the world of mortal opinion, which is inherently deceptive, whilst the
Aletheia reveals to us the unchanging object of divine knowledge.14
These arguments have some troubling implications for poetry, at least as it
was traditionally practised. We saw in Chapter 1 that Xenophanes’ singular
god could not be an agent in a narrative that was anything like the traditional
narratives surrounding Greek deities. With Parmenides, the same issue arises
to an even greater degree: the entity outlined in the Aletheia could not be an
agent in any narrative whatsoever. According to Lowe’s definition, the
medium of a narrative text needs to be able to represent events, and it must
be linear, in the sense that it must be received in a particular order, just as
literary texts offer a string of words designed to be read in one order rather
than another.15 What-is is unchanging and singular, and so can hardly be
involved in ‘events’. Even the second of Lowe’s criteria, linearity, is frustrated
by the goddess’s statement that ‘it is indifferent to me, from whence I begin,
for I shall return there again later’ (D5=B5).16 Parmenides’ use of narrative, and
in particular the dynamic narrative of the journey, is rendered paradoxical by
the nature of being as outlined in the goddess’s speech. The story of the
journey is inherently one of movement and change, both features that are
denied of what-is (D8.43=B8.38).
More problematic still, language as a whole seemingly fails to give an
accurate representation of reality. Mortal error consists in establishing mis-
leading names for what-is (σήματ’ ἔθεντο, D8.60=B8.55; πάντα φάος καὶ νὺξ
ὀνόμασται, D13=B9.1; and emphatically, in the closing line, τοῖς δ’ ὄνομ’
ἄνθρωποι κατέθεντ’ ἐπίσημον ἑκάστωι, D62=B19.3). The plurality of mor-
tal onomata thus erroneously divides what truly is into multiple entities.17 The
only ‘true’ thing one can say that is not misleading in this way appears to be
13
See especially Cole (1983) and Palmer (2009) 89–93: aletheia for Parmenides is not a logical property;
rather, it is an ontological and epistemological one.
14
Most recent commentators, stressing the length and detail of the Doxa, have tended to regard it as
intended as an accurate account of doxastic things, but not of true reality. See e.g. Nehamas (2002)
56–63; Palmer (2009) 163; Bredlow (2011); Johansen (2016); Tor (2015) and (2017) 155–308 (esp.
163–221). As Tor points out, the notion that the Doxa is intended as an accurate account of the
doxastic world need not rule out the possibility that it also reveals how such a world fails to live up to
true reality. For a detailed summary of earlier scholarly views, see Kraus (2013) 481–96.
15
Lowe (2000) 20.
16
ξυνὸν δέ μοί ἐστιν / ὁππόθεν ἄρξωμαι · τόθι γὰρ πάλιν ἵξομαι αὖτις. See Coxon (2009) 286–7,
following Diels (1897), for the translation of ξυνόν as ‘indifferent’.
17
Thus Graeser (1977) 362; Kraus (1987) 94–7; and Mason (1988) 149–51. Note the principle of
language ‘dividing’ its objects at Pl. Crat. 338b.
70 Parmenides
ἐστί. The term is repeated in an almost incantatory manner by the goddess.18
The notion that mortal language differs from divine communication is already
implicit in earlier Greek hexameter, where certain terms are only ever used of
divine speech and the gods have special words for particular items.19
Xenophanes developed this distinction between divine and mortal communi-
cation by having the single greatest god communicate without speech.
Parmenides continues this trend by portraying mortal language as deficient.
All the same, the goddess must use mortal terminology, inadequate though it
may be, since she is communicating with a mortal addressee. As I argued was
the case for Xenophanes, the consequences of this deficiency for poetry are
double-edged: on the one hand, the fictional potential of poetry illustrates the
deceptive nature of mortal language and experience; on the other, poetry acts
as a means of saying the unsayable, of thinking through new concepts and
transporting the audience, beyond the limitations of ordinary speech, to
a contemplation of unchanging reality.
3 Poetry and Deception
a Deception and the Doxa

The first of these functions is most evident in the transition to the Doxa,
where Parmenides inscribes the self-consciously ‘deceptive’ part of the
poem within a tradition that stresses the tendency of poetry to tell made-
up stories (D8.55–66=B8.50–61):
ἐν τῶι σοι παύω πιστὸν λόγον ἠδὲ νόημα 50
ἀμφὶς ἀληθείης· δόξας δ’ ἀπὸ τοῦδε βροτείας
μάνθανε κόσμον ἐμῶν ἐπέων ἀπατηλὸν ἀκούων.
μορφὰς γὰρ κατέθεντο δύο γνώμας ὀνομάζειν·
τῶν μίαν οὐ χρεών ἐστιν – ἐν ὧι πεπλανημένοι εἰσίν –
τἀντία δ’ ἐκρίναντο δέμας καὶ σήματ’ ἔθεντο 55
χωρὶς ἀπ’ ἀλλήλων, τῆι μὲν φλογὸς αἰθέριον πῦρ,
ἤπιον ὄν, μέγ’ ἐλαφρόν, ἑωυτῶι πάντοσε τωὐτόν,
τῶι δ’ ἑτέρωι μὴ τωὐτόν· ἀτὰρ κἀκεῖνο κατ’ αὐτό
τἀντία νύκτ’ ἀδαῆ, πυκινὸν δέμας ἐμβριθές τε.
18
For the incantatory repetition in Parmenides, note Kingsley (1999) 118–21 and Gemelli Marciano
(2008) 32–4; on repetition as typical of incantation in ancient magical spells, see Versnel (2002)
130–5.
19
In Homer, terms such as ὄσσα, θέσπις and θεσπέσιος are only used of divine speech (see Ford [1992]
180–97), implying that it has a different status from mortal speech, and certain objects are given
different names by humans and the gods (Il. 1.403, 2.813–14, 14.291, 20.74; at Od. 10.305 and 12.62,
divine names, but no human names are given). See Gera (2003) 47–57.
Poetry and Deception 71
τόν σοι ἐγὼ διάκοσμον ἐοικότα πάντα φατίζω, 60
ὡς οὐ μή ποτέ τίς σε βροτῶν γνώμη παρελάσσηι.
At this point for you I end the trustworthy speech and thought 50
About truth; but, from here, the mortal opinions
Learn by hearing the deceptive arrangement of my words.
For they established shapes to name their two judgements;20
Of which they should not name one – on that point they have erred –
And they have judged body in opposite ways and placed signs 55
Apart from one another, to one of them the ethereal fire of flame,
Being gentle, very light, in every way the same as itself,
But not the same as the other; but that one, by itself,
The opposite, unknowing night, a thick body and weighty.
I tell you, this whole, likely arrangement, 60
So that no judgement of mortals shall ever surpass you.
A number of details advertise the status of this section as a poetic con-
struction. First, the goddess introduces the Doxa as a ‘kosmos of words’
(κόσμον ἐπέων) a punning reference both to the arrangement of the words
and to the cosmos it depicts. More importantly for our present purposes,
the expression has specifically poetological connotations.21 Solon pro-
claims that he has put together a ‘Song, a kosmos of words, instead of
a speech’ (fr. 1.2 W, κόσμον ἐπέων †ὠιδὴν ἀντ’ ἀγορῆς θέμενος). As
Noussia-Fantuzzi notes, this expression ‘designates an ordered and met-
rically defined sequence of ἔπεα᾽.22 A further relevant example is
Democritus’ statement that ‘Homer, who received as his share a nature
that was divine, has constructed a kosmos of all kinds of words’ (Ὅμηρος
φύσεως λαχὼν θεαζούσης ἐπέων κόσμον ἐτεκτήνατο παντοίων,
D221=B21). Given that Democritus’ atomism seems in large part to have
been formulated as a defence of pluralism against Eleatic monism, this
fragment could well be a direct allusion to Parmenides’ line, putting
a positive spin on something which had been negatively characterised in
20
For this interpretation of this line, see Mackenzie (2017) 46–7.
21
In favour of identifying the pun, see e.g. Kraus (1987) 80–1; Nehamas (2002) 60. On the aesthetic
and specifically poetological connotations of kosmos, see Nünlist (1998) 90–7, who exhaustively cites
parallels, including most closely, Od. 8.489, 8.492; hHom. 7.58–9; Simonides eleg. 11.23 W, 22.5–6 W;
OF 25; Pind. fr. 194.2–3. Lanata (1963) 12–13 and Walsh (1984) 8–9 discuss poetic kosmos in the
Odyssey but argue that it indicates the truthfulness of Demodocus’ account when it seems more
vaguely to indicate that it is ‘appropriate’ and ‘well-ordered’. For more a detailed, nuanced reading
of the Homeric passage, see Halliwell (2011) 84–8.
22
Noussia-Fantuzzi (2010) 211. Note also the roughly contemporary parallel of Simonides fr. 11.23 W κ
[όσμον ἀοι]δῆς, and the later instance of Philetas fr. 10.3 Powell=8.3 Lightfoot, where the narrator
‘knows the arrangement of words’ (ἐπέων εἰδὼς κόσμον), in lines which ‘contain an image, perhaps
self-image, of the refined, learned, and dedicated poet’ (Lightfoot [2009] 43). On this parallel, see
also Floyd (1992) 261–3.
72 Parmenides
the earlier poem.23 The parallels suggest, then, that ‘kosmos of words’
marks out the Doxa as an arrangement of verse.
Second, the fact that the account is ‘deceptive’ (ἀπατηλόν) also recalls
early Greek discussions of poetry. A proverb, apparently mentioned by
Solon, stated that ‘poets tell many lies’ (πολλὰ ψεύδονται ἀοιδοί, Solon
fr. 29 W). Famously, Gorgias would define tragedy as a ‘deception in which
the one who has deceived is more just than the one who has not deceived and
the one who has been deceived is wiser than the one who has not been
deceived’ (D35=B23).24 An association between poets and lying or deceiving
is already implicit in Hesiod’s Theogony, in the Odyssey’s combination of
singer and deceiver in its depictions of Odysseus, Helen and the Sirens, and
in the figure of Hermes in the Homeric Hymn.25 This context, coupled with
the connotations of the expression, ‘kosmos of words’, invites us to regard the
Doxa as deceptive in a manner analogous to the incredible tales of the poets.
Third, the goddess’s description of the Doxa as διάκοσμον ἐοικότα, ‘a
likely arrangement’ (D8.65=B8.60), has been seen as a ‘window-allusion’ to
the ‘lies like (ὁμοῖα) true things’ of Hesiod’s Muses (Theog. 27) via
Xenophanes D50=B35, ‘Let these things be thought to be like true things’
(ἐοικότα τοῖς ἐτύμοισι).26 Such an allusion is supported by the fact that, in
the proem, the goddess’s distinction between the Aletheia and the Doxa
corresponds to the distinction made by Hesiod’s Muses between truthful
and verisimilitudinous speech. Discussions of this relationship have
focused on its consequences for our understanding of Parmenides’
epistemology,27 but most notable for our purposes is the fact that the
allusion implicates the poem within discussions over the veridical status of
poetry in particular: the Doxa is not merely ‘likely’ but likely in a manner
that is comparable to potentially deceptive, Muse-inspired song.
Xenophanes had adapted the line, apparently, to mark the speculative
nature of his own cosmological claims, in opposition to the truth claims
of earlier poets (see Chapter 1 Section 2a above). Within the context of
Parmenides’ poem, the unstable nature of the Doxa’s cosmology contrasts
with the more secure knowledge attainable of what-is. But, like the Muses’
lies, the cosmology will appear convincing to mortals. The resonance of the
23
On the atomists’ response to Parmenides, see Sedley (1982) and (2008).
24
ἀπάτην ἣν ὅ τ’ ἀπατήσας δικαιότερος τοῦ μὴ ἀπατήσαντος καὶ ὁ ἀπατηθεὶς σοφώτερος τοῦ μὴ
ἀπατηθέντος. A similar claim is also made (perhaps under the influence of Gorgias) at Dissoi
Logoi 3.10.
25
On this topic, see esp. Pratt (1993) and Introduction Section 2 above.
26
Thus, most extensively, Bryan (2012) 58–113 (esp. 64–6, 93–100).
27
Coxon (2009) 349–51; Bryan (2012) 58–113.
Hesiodic scene thus marks the status of the Doxa as fulfilling the capacity of
poetry to be versimilitudinous.
b The Sirens
This section, then, like the most effective earlier hexameter poetry, is so
comprehensive (D4=B1.31–2) and plausible (a quality implied by ἐοικότα)
that we are tempted to immerse ourselves uncritically within the imagina-
tive world it describes, but the goddess protects us from this temptation by
marking it out as a world of poetic deception. A further feature serves to
illustrate this state of affairs, a reference to the paradigmatically alluring
singers of Greek myth, the Sirens. Scholars have observed that Parmenides’
reference to necessity holding ‘what-is’ ‘in the bonds of a limit’ (πείρατος
ἐν δεσμοῖσιν ἔχει, D8.36=B8.31) is modelled on Odysseus’ description of
being tied to the mast at Od. 12.196, πλείοσί μ’ ἐν δεσμοῖσι δέον, ‘they
bound me with more bonds’.28 The expression ἐν δεσμοῖσιν occurs in the
same metrical position in both lines. Whilst this parallel alone may not
seem particularly compelling, we can add a further one, briefly mentioned
by Havelock.29 The goddess teaches the youth this deceptive account so
that no ‘judgement of mortals shall surpass’ him (D8.66=B8.61). The verb
that Parmenides uses (παρελάσσηι) occurs several times in reference to the
journey of Odysseus’ ship past the Sirens (παρὲξ ἐλάαν, Od. 12.47, 55;
παρελαύνειν, 186, 197). Closest to the Parmenidean line is Od. 12.186, sung
by the Sirens themselves, ‘For no one has yet gone past in his black ship’
(οὐ γάρ πώ τις τῇδε παρήλασε νηῒ μελαίνῃ). Both Parmenides
D8.66=B8.61 and Od. 12.186 have ‘no-one’ (οὐ τίς) as the subject of
‘surpass’ (παρελαύνειν), qualified by a temporal adverb (πώ/ποτέ). The
verbal echoes are bolstered by similarities of circumstance: both lines are
spoken by a female deity to a mortal addressee, and in both cases divine
knowledge is emphasised. Further, both speakers are telling a deceptive
tale: we know that the Sirens’ listeners will not in fact ‘return’ (νεῖται) and
that they are saying this in an attempt to lure Odysseus as their next victim.
Parmenides’ goddess is also about to tell us a deceptive story
(D8.57=B8.52).
The background of the Sirens episode illuminates Parmenides’ use of
chain imagery as well as the deceptive nature of the Doxa. The goddess has
described ‘what-is’ as being held ‘in chains’ whilst most mortals, in their
beliefs, erroneously split it into separate things. Odysseus, on hearing this
28 29
Cassin (1987); Wright (1998) 18. Havelock (1958) 143 n. 59.
74 Parmenides
song, begs for his bonds to be loosened (Od. 12.193), at which point his
companions, Perimedes and Eurylochus, heap on further chains (195–6).
With this context in mind, the ‘Way of Opinion’ appears as an alluring
account, like the Sirens’ song, which tempts us to strive, like Odysseus, to
loosen the ‘bonds’ of what really is. In the Odyssey, the Sirens mix the truth
with lies. They claim to know all that happened at Troy – indeed, all that
happens on earth (Od. 12.189–91) – and they demonstrate their supernat-
ural knowledge when addressing Odysseus by name (Od. 12.184) before he
has provided any clue as to his identity.30 But the Sirens also fallaciously
claim that their listeners leave safely after hearing their song (Od. 12.188).
The Doxa similarly contains a mixture of qualities that do and do not apply
to true reality, at least as it is set out in the Aletheia. Fire is ‘in every way the
same as itself’ (ἑωυτῶι πάντοσε τωὐτόν, D8.62=B8.57) and Night is ‘by
itself’ (κατ’ αὐτό, D8.63=B8.58), features that resemble the entity described
in D8>B8, which is also ‘by itself’ (καθ’ ἑαυτό, D8.34=B8.29) and ‘equal to
itself in every direction’ (πάντοθεν ἶσον, D8.54=B8.49). At D13=B9.3, ‘all is
full together of light and invisible night’ (πᾶν πλέον ἐστὶν ὁμοῦ φάεος καὶ
νυκτὸς ἀφάντου), whilst in the Aletheia, ‘everything is full of what-is’ (πᾶν
δ’ ἔμπλεόν ἐστιν ἐόντος, D8.29=B8.24). The objects in the Doxa, like true
reality, are ‘bound’ within ‘limits’ (D12=B10.6–7). These qualities, in the
Aletheia, are ‘markers’, semata, of what-is (D8.7=B8.2). By following some,
but not all of them, we might be led astray into the ‘likely’ story of the
Doxa. In other respects, the objects of the Doxa clearly contrast with the
features of true being set out in the Aletheia since, in the Doxa, things are
generated (D12=B10.6) and move (D12=B10.4, D27=B14).31 The mytho-
logical reference points to the fact that the Doxa, like the Sirens’ song,
seductively combines elements that are deceptive and those that are
alēthēs – accurate and in conformity to true reality.
In this respect, both the Doxa and the Sirens’ episode also resemble the
Hesiodic ‘lies like the truth’. The three passages can be read in dialogue, as
contributions to a discussion surrounding the deceptive potential of song.
The Sirens and Hesiod’s Muses, using similar expressions, claim to ‘know’
two sets of things and both are able to speak both deceptions and truths,32
but the nature of the deception differs in each case. Hesiod’s Muses know
30
A particularly striking detail, given the significance of naming, and especially Odysseus’ name,
within the poem. See especially Austin (1972).
31
On the respects in which the cosmology of the Doxa is ‘like’ or resembles the entity of the Aletheia,
see Johansen (2016).
32
Note the anaphora ‘ἴδμεν . . . ἴδμεν’, occurring at both Od. 12.189, and Hes. Theog. 27–8. On the
theory of poetics underlying the Sirens passage, see Ledbetter (2003) 27–34.
how to speak lies that are equivalent to the truth, so that, from a mortal
perspective, the effect of their song is the same whether or not they choose
to tell truths.33 Their falsehoods have the property of verisimilitude: they
seem plausible even though, for all we know, all their details may be
invented. The Sirens, by contrast combine their lies with details that
their listener knows to be factually accurate, such as Odysseus’ name and
the fact that he toiled at Troy.34 The Doxa presents a further variation on
this theme. This section offers a cosmology that is ‘deceptive’ and in which
there is no ‘true trustworthiness’, but it is not necessarily false and in the
surviving fragments is nowhere labelled as pseudea. The word translated as
‘true’, alēthēs, encompasses factual accuracy, but, in its earliest usages,
seems to refer specifically to speech that neither omits significant details
nor adds extraneous information to its listener.35 In other words, a speech,
such as the Doxa, may be factually accurate, and refer to actual states of
affairs, whilst still failing to meet the conditions for aletheia. Rather, like
Xenophanes’ cosmology, it seems to refer to items over which true cer-
tainty is impossible.
The fact that, in the introduction to the Doxa, Parmenides alludes both
to the Sirens’ episode and to the lying Muses of Hesiod’s Theogony
encourages us to explore this comparison further. A mortal audience of
the Doxa, who had not heard the Aletheia, would be in the same position as
the addressees of Hesiod’s Muses: they are unable to tell whether it is true
or not, and it has the appearance of that over which there is certainty.
Having learned the Aletheia, however, we are ideally in a position to
identify the ways in which it does or does not conform to the characteristics
of that over which we can have certainty, just as an audience of the Odyssey
is able to tell the true from the false parts of the Sirens’ speech. We might
recall that the Sirens caused forgetfulness of homecoming (cf. Od. 12.41–6)
because of their song that appeared to be truthful; our knowledge of
aletheia – literally, ‘un-forgetfulness’ – will prevent us from being seduced
by the Doxa into thinking it is entirely alēthēs. Whether or not we would
expect an audience to have picked up on all these details in performance,
the Doxa draws upon well-known earlier episodes that had stressed some of
the troubling implications of poetry’s capacity to lure us into a vividly
33
Thus, Heiden (2007). For a thorough, recent discussion of these much-discussed lines see Tor (2017)
61–103, who emphasises their problematic nature.
34
On this distinction, with respect to Odysseus’ ‘lies like the true things’ (Od. 19.203), cf. Williams
(2002) 204–5, who points out that the expression refers to vraisemblance rather than resemblance to
things that actually occurred.
35
Cole (1983).
76 Parmenides
imagined world. Through alluding to these highly meta-literary episodes
and through marking this section out as a poetic construction, Parmenides
warns us that we should treat it with the epistemological caution appropri-
ate to the fanciful tales of the poets.
c Deception and the Proem

In light of the goddess’s rejection of sensory experience and her presenta-
tion of the Doxa as a world of poetic artifice, the extraordinary journey of
the proem appears as a deception in two senses. First, it is deceptive because
it is an instance of vivid literary mimesis. We imagine a chariot journey
without actually observing one. Though the precise topography of the
proem may be ‘blurred beyond recognition’,36 it presents an especially
immersive description. There is a selective reference to particular details in
an order that seems to follow the path of the youth’s gaze: we can imagine
him looking in amazement at the horses straining at the chariot
(D4=B1.1, 5), then at the maidens leading them (D4=B1.5), then turning
towards the source of the sound and looking at the wheels on either side
(D4=B1.6–7), then finally, again, turning to the maidens as they remove
their veils (D4=B1.10). The proem’s ‘blurredness’ and the uncertainty of
some of its syntax further contribute to this focalisation,37 generating
a sense of bewilderment that reflects the youth’s experience on this strange
trip. These features facilitate an audience’s assumption of the youth’s
perspective and their immersion into the world of the narrative.38
The fact that the journey is a fantastical one that stretches our credulity
underscores the deceptiveness of the mimetic world it conjures. As
Burkert’s classic study showed, the area described in the proem resembles
Hesiod’s Tartarus, and the youth’s claim to have gone to this place
conforms to a pattern recognisable in the stories of Archaic sages, such as
Aristeas, Epimenides, Salmoxis and Pythagoras, who were said to have
gone on extraordinary journeys and, in many cases, to have died and come
36
Mourelatos (2008a) 15.
37
The phrase εἰς φάος (line 10) may be been taken to refer to the direction of πέμπειν, that is, the
ultimate destination of where the Heliades are taking the youth (thus, e.g. in DK, and more recently
espoused by Kahn [2002] and [2005]) or to the direction of the participle προλιποῦσαι, that is, to
the place at which they arrive to collect the youth (thus, Morrison [1955] 60; Burkert [1969] 7–9;
Palmer [2009] 57; Cosgrove [2011] 37–8; Primavesi [2013a] 53–6). However, as Miller (2006) 20
argues, we may read here a functional ambiguity.
38
For selectivity of details (as opposed to exhaustive, panoramic description) and embedded focalisa-
tion as especially immersive features that facilitate an audience’s visualisation of the narrative, see,
respectively, Grethlein and Huitink (2017) and Allan, de Jong and de Jonge (2017) 42–3.
back to life.39 But it should be noted that these stories were met with
considerable scepticism in the fifth century bce.40 Sophocles’ Orestes
cynically states that he has ‘often seen wise men dying in false report.
Then, when they come home again, they are honoured more’ (El. 62–4).41
Herodotus reports a rationalising account of Salmoxis in which he hood-
winks the Thracians into believing he has died by living in an underground
chamber for three years (4.95). Democritus even wrote a treatise on such
stories, presumably providing rational explanations (D143<B1). Scepticism
of this kind over the proem is triggered by certain details. The goddess, on
greeting the youth, distinguishes ‘the unshaken heart of well-rounded
truth’ from ‘the opinions of mortals in which there is no true trustworthi-
ness’, a detail which may lead us to question whether what we have just
heard corresponds to the truth or to untrustworthy opinion. Her word for
‘trustworthiness’, πίστις, etymologically related to the verb for persuasion,
πείθω, recalls the Heliades’ careful persuasion of Justice (D4=B1.15–16) to
open the gates. The fact that they must persuade Justice ‘with gentle words’
(μαλακοῖσι λόγοισιν, D4=B1.15) invites speculation over whether their
persuasion is truthful or deceptive, especially since ‘much-punishing
Justice’ might impose a severe penalty on the latter. The expression may
have deceptive connotations: its only Homeric occurrence is in reference to
the crooked words by which Calypso causes Odysseus to forget Ithaka (Od.
1.56). This thematisation of the trustworthiness and deceptiveness of
speech raises the issue of the trustworthiness of the youth’s account,
which has been so vivid, yet has depicted an event which is so far from
the realm of everyday experience.
The second sense in which the proem is deceptive is more fundamental,
for the world it describes conspicuously fails to meet the goddess’s condi-
tions for true being.42 The salient images of movement, duality (in the
wheels ‘on both sides’) and opposition (in the image of Night and Day) are
pointedly incompatible with the true reality of the Aletheia. The image of
the paths of Night and Day – the two elemental components of the
cosmology – confirms that the proem, at least up until the youth crosses
39
Burkert (1969); cf. (1972) 283–5. The main details that recall Hesiod’s Tartarus are the place of Night
(D4=B1.9; cf. Theog. 744), the gates which straddle the paths of Night and Day (D4=B1.11; cf. Theog.
732–3, 748–50), and the chasm created when the gates open (D4=B1.18; cf. Theog. 740).
40
Possibly, there is already scepticism of this sort evident in the Odyssey if we are to detect an ironic
tone in Alcinous’ comment that Odysseus does not look like a liar and that he spoke expertly, like
a singer (Od. 11.363–9), just as he is in the midst of recounting the Nekuia, the scene that most of all
stretches our credulity. On this sort of irony in the epic, see Goldhill (1991) 36–56 and Segal (1983).
41
For the connection between this passage and the sages mentioned, see Finglass (2007) ad loc.
42
On this aspect of the proem, see Miller (2006) 12–39. Note also Coxon (2009) 276–7 ad D4=B1.12.
78 Parmenides
the gate, describes the deceptive world of the Doxa. According to
this second sense, all of our sensory experiences are deceptive. But the
poetological language that introduces the Doxa suggests that the first type
of deception is intended to illustrate the second: the familiar example of
poetic deception allows us to see that the whole world of our sensory
experience is deceptive, albeit in a different respect. Poetry deceives us by
generating an imagined world that need not correspond to any actual state
of affairs; our senses deceive us by presenting a shifting world that does not
live up to the demands of secure knowledge. In both cases, we should
refrain from granting the status of certain reality.
4 Profound Visions and Sublime Emotions
a Seeing the Unseeable

Whilst Parmenides’ poetic mimesis illustrates sensory deception, it also
serves the more positive didactic function of leading us to contemplate an
entity that could not be encapsulated in ordinary speech. We noted in the
Introduction that, already in our earliest sources, poetry characteristically
enables its audience to imagine a world separate from the hic et nunc and
that this process is often presented in visual terms. As a later ancient critic
would imply, when discussing Homer’s enargeia, poetry can turn its
listeners into spectators.43 Parmenides harnesses this particular characteris-
tic, inviting a visual response to his text, quite explicitly, when the goddess
commands the youth (and, by extension, the audience) to ‘see things which
though absent are present to the mind’ (D10=B4.1).44 Other details
entrench the impression that the goddess’s speech affords a visual experi-
ence. Those that do not recognise what-is and who have not been granted
this divine perspective are ‘blind’ (D7=B6.7) for trusting their senses. The
surprisingly intimate gesture, from a deity, of the handclasp (D4=B1.22) is
a sign of good faith but may also indicate that she is leading the youth by
the hand to show him the correct path.45 We might compare Il. 4.539–44,
43
ΣbT Il. 23.363–72, on which, see Grethlein and Huitink (2017) 67–9, as well as Meijering (1987)
29–52 and Nünlist (2009) 194–8. On the history of the term enargeia, see Zanker (1981).
44
LM assign this fragment to the Doxa, but I follow most editors in treating it as part of the Aletheia,
given that it seems to articulate a principle elaborated at greater length in D8>B8.
45
On the handclasp and its significance, note Floyd (1992) 254–5; Coxon (2009) 10; and especially
Mansfeld (2005) 554–5, who draws on the parallels such as Il. 6.223, 14.232 and Od. 1.120–3, 2.302 to
argue that it is a mark of good faith. Floyd on the other hand argues that the two closest parallels to
Parmenides’ line (Il. 21.286; Od. 2.321) involve some degree of deception, but his interpretation of
the Iliadic example is questionable.
Profound Visions and Sublime Emotions 79
where Athene protectively leads by the hand (χειρὸς ἑλοῦσ᾽, 542)
a hypothetical warrior across the battlefield, or a later statement, by
Dionysius of Halicarnassus (Ant. Rom. 11.1.3), that ‘the mind of every
man takes pleasure in being led by the hand (χειραγωγουμένη) through
words to deeds, and not only hearing the things said but seeing what is
done’.46 Our experience of ‘seeing’ what the goddess describes is, then,
a mise en abyme, a visualisation within the wider visualisation of the story as
a whole, that had begun with the proem. Whilst, in the proem, we were
invited to imagine something that was, in practice, inaccessible to mortal
experience, in the Aletheia, we are told to visualise an entity that is not only
practically but theoretically inaccessible (cf. D8.1–6=B7).47 The visualisa-
tion of the fantastical world of the proem prepares us for the visualisation
of something even further removed from our everyday experience, what-is.
Though the evidence is scanty, we can tell from D12=B10 that the Doxa
developed this visual language:
εἴσηι δ’ αἰθερίαν τε φύσιν τά τ’ ἐν αἰθέρι πάντα
σήματα καὶ καθαρᾶς εὐαγέος ἠελίοιο
λαμπάδος ἔργ’ ἀίδηλα καὶ ὁππόθεν ἐξεγένοντο,
ἔργα τε κύκλωπος πεύσηι περίφοιτα σελήνης
καὶ φύσιν, εἰδήσεις δὲ καὶ οὐρανὸν ἀμφὶς ἔχοντα 5
ἔνθεν ἔφυ τε καὶ ὥς μιν ἄγουσ(α) ἐπέδησεν Ἀνάγκη
πείρατ’ ἔχειν ἄστρων.
And you will know the aethereal nature and all the signs
That are in the aether, and of the pure, bright torch of the sun
The blinding [or hidden] works, and from whence they came to be,
And you will learn the wandering works of the round-eyed moon
And its nature, and you will know the surrounding sky, 5
From whence it was born and how Necessity, taking it, shackled it
To have the limits of the stars.
46
ἥδεται γὰρ ἡ διάνοια παντὸς ἀνθρώπου χειραγωγουμένη διὰ τῶν λόγων ἐπὶ τὰ ἔργα καὶ μὴ μόνον
ἀκούουσα τῶν λεγομένων, ἀλλὰ καὶ τὰ πραττόμενα ὁρῶσα. Cf. also Aelius Aristides, Smyrnaeus
politicus 17.8. On these passages, see Clay (2011b) 25–6.
47
D8.1–6=B7 is usually taken as rejecting completely the value of sensory experience as a means of
accessing reality (e.g. by Lesher [1984] 1–2; Coxon [2009] 308), although Laks (1999) 262–3 argues
that it only rejects misuse of the senses. Similarly, de Rijk (1983) argues that Parmenides does not
reject the sensible world, only the mortal understanding of it, as, in this fragment, the goddess is
instructing us not to direct our senses down the way of not-being, but rather, down the way of being.
He has been followed by Curd (1998) 60–2. Even if the target of D8.1–6=B7 is not sensory
experience in toto but simply sensory experience as normally used, it does not necessarily follow
that the entity described in the Aletheia is something that can be seen, and the adversative δέ at
D8.5=B7.5 suggests that we are to grasp it by some faculty other than our senses.
80 Parmenides
εἴσηι and εἰδήσεις are usually translated as ‘you will know’, and these forms
occur elsewhere in this sense.48 But this lexeme is in origin a verb of
seeing,49 and the emphasis on the brightness of the objects described (the
‘pure bright torch of the sun’ and the starry sky) activates the implicit sense
of, ‘you shall see’. ἀίδηλα, translated here as ‘blinding’, supports this
suggestion. It is a tricky word, sometimes used to mean ‘hidden’ and
elsewhere, ‘destructive’, probably from an active sense of ‘making
invisible’.50 Its etymology is explicable as an ἀ-privative + ἰδεῖν, a verb
from the same root as εἴσηι and εἰδήσεις. Here, it may either mean the
‘hidden’ works of the sun, in reference to the Sun’s journey during the
night when it is invisible, or to the ‘destructive’ nature of the sun, in that it
blinds those who look directly at it.51 Whatever we take the reference of
ἀίδηλα to be – and we may be justified in reading here a functional
ambiguity – it develops, paradoxically, the visual theme introduced by
εἴσηι: the youth will ‘see’ that which is either ‘unseen’ or that which is
‘blinding’.
We are, then, to ‘see’ the items in the Doxa, many of which (the sun, the
stars) are familiar from our normal sensory experience, but by that stage we
have been warned that this is a world of poetic artifice, one that does not
meet the conditions for true being. Through ‘seeing’ the unseeable, in the
Aletheia, we come to ‘see’ the content of our sensory experience in a new
light, as failing to conform to true being. Our visualisation in the Doxa
thus prepares us to reinterpret our actual visual experience once the poem is
over.
b The Vision
The poem could be said to provide a visual experience, not only in the
sense that it enables us to visualise certain entities, but also in the broader,
metaphorical sense that it offers the opportunity to see the world in
a different way, from a new perspective. The wider narrative tradition in
which the poem participates highlights this function. The story neatly fits
an identifiable genre of myth that Bruce Louden has termed ‘the vision’, ‘in
which the protagonist is removed from the mortal plane, accompanied by
48
εἴσομαι: Il. 1.548, Hp. VM 20, Ar. Ach. 332; εἰδήσω: Od. 7.327, Hdt. 7.234, Isoc. 1.44, Aen. Tact. 31.5.
49
They are future tenses of the putative verb *εἴδω, ‘I see’, of which οἶδα is in origin a perfect tense.
The same verb is used to form the aorist, εἶδον, of ὁράω. See LSJ s.v.
50
See Chantraine s.v.
51
The former sense is argued by Coxon (2009) 354; LM translate as ‘blinding’ (which I have followed),
but also note the possible meanings of ‘destructive’, ‘completely visible’ and ‘hidden’.
an otherworldly guide, who reveals to him a large truth, the “big picture”,
previously hidden from his view’.52 Other ancient examples include the
katabasis narratives of the Odyssey’s Nekuia scene, Bilgames/Gilgamesh and
the Netherworld (an episode from the Gilgamesh cycle), Aeneid book 6, as
well as a story from the Old Testament in which Saul uses a female
necromancer to speak with the dead prophet Samuel (1 Sam. 28), Plato’s
allegory of the cave, and Cicero’s Somnium Scipionis. In all these examples,
visual language plays a prominent role in the revelation granted.53 Louden
does not mention Parmenides’ poem, but it fits the scheme remarkably
well: the youth leaves the mortal plane, accompanied by the Heliades as
‘otherworldly guides’, before the goddess fulfils this role in revealing to
him, using visual terms, the ‘big picture’ that had been previously hidden
from view. Furthermore, several of the examples of the genre are katabasis
narratives, whilst Parmenides’ chariot journey certainly evokes such a trip to
the realm of the dead, even if its precise trajectory cannot ultimately be
determined with certainty.54
Through such visions, the protagonist does not only learn new informa-
tion, but must reassess his values in life. Odysseus learns from Teiresias the
details of how to achieve his homecoming (Od. 11.100–37), but he also sees
the punishment of the notorious wrongdoers, Tityos, Tantalus and
Sisyphus (Od. 11.576–600) and learns from Achilles that it is better to be
a serf to a pauper among the living than to rule over all the dead (Od.
11.489–91). These moralising details reveal to Odysseus part of the overall
ethical framework of the world of the poem, one that has been seen to
contrast pointedly with that of the Iliad.55 Bilgames/Gilgamesh, after
Enkidu does not return from the underworld, enlists the help of a young
hero (Utu in the Sumerian version; Shamash in the Akkadian) to summon
the phantom of his companion, who then reveals the posthumous fates of
various types of individual, the happiest of whom appear to be men with
many sons (Netherworld 255–306; Gilgamesh xii 96–153). Saul learns from
52
Louden (2011) 197–221, quotation from 198.
53
Odysseus repeatedly refers to whom he saw in the catalogue of heroines (Od. 11.235, 260, 271, 281,
298, 306, 321, 326), and Alcinous asks him if he saw any of his Greek comrades (371). Saul asks the
necromancer what she saw, which she then relates to him (1 Sam. 13). In Bilgames and the
Netherworld, Bilgames/Gilgamesh repeatedly asks the phantom of the dead Enkidu (in Andrew
George’s translation), ‘Did you see . . . [x type of person]’, and on each occasion, Enkidu responds, ‘I
saw him’, George (1999) 187–90 (Bilgames and the Netherworld, lines 246–303) 194–5 (tablet xii of
the ‘Series of Gilgamesh’, lines 102–53).
54
On the controversy surrounding this issue, see n. 104 below.
55
See Kullmann (1985). Achilles’ statement to Odysseus contrasts with his choice for a short but
glorious life at Il. 9.410–16, as commentators since the scholiasts have noted (e.g. Goldhill [1991]
105–6, who cites further scholarship). For a criticism of Kullmann’s view, see Allan (2006).
82 Parmenides
the shade of the deceased prophet Samuel that Yahweh has turned away
from him for his disobedience (1 Sam. 28.16–20).
Correspondingly, Parmenides’ youth, in the Doxa, will learn about the
origins of humanity (D46=B17; D49=B18) and the nature of human
cognition (D51=B16), but he will also learn about the futility and fleeting-
ness of mortal attitudes in contrast with the constancy of what-is. Most
emphatically, D7=B6 criticises the ignorance of mortals (they are ‘know-
nothings’, line 4), but it also hints at their limited lifespans, in referring to
them contemptuously as ἄκριτα φῦλα (line 7), ‘undiscerning [or ‘indistin-
guishable’] tribes’. As Peter Kingsley has argued, the adjective ἄκριτα,
though normally translated as active (‘undiscerning’) could also be under-
stood passively (‘indistinguishable’) and the expression as a whole is
modelled on the Homeric hapax, ἀκριτόφυλλος, ‘of undistinguishable
leafage’ (LSJ s.v.), a term used to refer to a wooded mountain (Il.
2.868).56 Parmenides’ expression thus recalls the famous comparison
between the fleeting generations of mortals and the seasonal regeneration
of leaves mortals and leaves (Il. 6.146–9; Simon. Eleg. 19.2 W; Mimnermus,
2.1–5 W). This echo aligns the goddess’s teachings with traditional wisdom
concerning the ephemerality and insignificance of mortal experience of
a kind that is amply paralleled in other katabasis narratives. Like the visions
of the Nekuia or Bilgames and the Underworld, the goddess’s claims are not
merely ontological or theoretical, but also have implications for how we are
to evaluate our status in the universe.
c Perspective and Character

Parmenides, then, fulfils the conception of literary truth that we met
in the Introduction, in that his poem exposes his audience to
a particular way of looking at the world. Of course, such a function
could also be performed in prose, but it is characteristic of early Greek
conceptions of song, a type of discourse that stereotypically transports
its audience into an imagined world. This new perspective, as Plato
recognised, has an axiological dimension: it assumes certain values, for
instance, in the way in which the Iliad treats the death of a great
warrior as worthy of lamentation, a principle that is rejected by Plato’s
Socrates (Resp. 3.387d–388d).57 Songs or poems do not expose
56
Kingsley (2003) 95–8.
57
My understanding of Socrates’ much-discussed views on poetry here is based on the studies of
Burnyeat (1999) and Moss (2007).
objective, value-neutral perspectives, but those of particular individ-
uals, whether the primary narrators or the characters of the narrative.58
In this way, perspective is closely related to characterisation. For most
of Parmenides’ poem, we are granted the perspective of a vividly
characterised individual. Unlike the generic youth, the goddess’s lively
use of language gives an indication of her appropriately self-assured
personality. She makes universalising, omnitemporal claims (e.g.
D6>B3; D7=B6.1–2; D8.27–35=B8.22–30) and uses intensifying expres-
sions (e.g. ‘That one indeed, I tell you is an entirely undiscernible path’,
D6>B2.6; ‘I bid you to reflect on these things’, D7=B6.2; ‘I tell you,
this is an entirely likely arrangement’, D8.65=B8.60). Her awareness of
her superior status is marked by frequent imperatives and imperatival
infinitives (D6>B2.1; D10=B4.1; D8.2=B7.2; D8.3=B7.3; D8.57=B8.52),
as well as future and optative second-person verbs that outline the
limits of her addressee’s possible actions (D6>B2.7–8; D10=B4.2;
D8.11=B8.6; D8.41=B8.36; cf. D12=B10.1; D12=B10.5). Her use of
anaphora, at D10=B4.2–4 and D8.27–9=B8.22–4, is paralleled nowhere
more so than in Hesiod’s Works and Days (5–7,182–4, 317–9, 578–80)59
and further serves to indicate her didactic authority. Consistent with
this high opinion of herself is a contempt for mortals, evident from
her almost sarcastic rhetorical questions in D8>B8. These questions
seem to imply that, for her, the nature of true reality is obvious, in
spite of the profound uncertainties over her meaning of the repeatedly
made assertion, ‘[it?] is’, ἔστιν.60 Through this characterisation, which
would be accentuated in performance through gesture and
intonation,61 we are exposed to a distinctive perspective on the
world in which mortals are supremely ignorant, and worthy of
contempt.
58
Cf. here Griffin (1986) 45–6, who shows that evaluative terms in Homer are far more frequent in
direct speech than in that of the primary narrator, although note also de Jong (1988) on the primary
narrator’s value-judgements.
59
Other examples from early hexameter (listed by West [1978] 139): Theog. 833–5, Il. 2.382–4, 23.315–16,
318, Od. 3.109–11.
60
It is beyond my scope to pronounce a confident verdict on this gargantuan problem of Parmenidean
interpretation, but I follow Mackenzie (1982) and Barrett (2004) in regarding this particular
difficulty as functional and designed to draw us into considering precisely what it is to ‘be’
something. For summaries of earlier interpretations, see Kraus (2013) 461–3 and Palmer (2016)
(though his summary is, naturally, partisan towards his own interpretation, without giving a fair
impression that this is still an open question).
61
For rhapsodic performance of direct speech as a kind of acting or impersonation, note Plato Ion
535b–e; Resp. 3.392d–395a. On the dramatic aspects of rhapsodic performance, see Herington (1985)
10–15, 51–4.
84 Parmenides
d The Emotional Trajectory of the Poem
Exposure to such a distinct perspective affects an audience’s emotional
response, though, of course, it does not necessarily prompt them to feel the
same emotions as the character in question. Probably, most audience
members would not themselves feel the contempt for humanity that the
goddess expresses, nor should we assume that Parmenides would have
intended such a mirrored reaction. It is notable that the youth, and not
the goddess, is the primary narrator, a feature that creates an extra layer of
distance between the author and her claims. But a number of features seem
designed to make an audience feel the emotions that would be appropriate to
the youth’s extraordinary situation. The focalisation of the proem, his generic
nature, and the fact that he effectively drops out of the narrative once the
goddess starts speaking, all encourage an audience to imagine themselves in
his shoes. Though the poem has been dismissed as not especially ‘moving’,62
the goddess’s characterisation, as well as certain details of the proem and her
speech, seem designed to have an emotional effect.
The hints that this is a journey to the realm of the dead create an ominous,
uncanny tone. I have mentioned that certain details of the proem’s topog-
raphy recall Hesiod’s description of Tartarus.63 The choice of paths that the
goddess offers the youth in later fragments (D6>B2, D7=B6, D8.1–6=B7)
resembles the eschatology of the ‘Orphic’ gold tablets, and of Plato’s eschato-
logical myths, in which the deceased is offered a choice of which ‘path’ to
take.64 The youth’s state of perturbation can be inferred from the fact that the
goddess must reassure him that it is not an ‘evil fate’ (μοῖρα κακή) that has
sent him, nor has he performed anything illicit (D4=B1.26–8). Indeed, this
may be a reassurance that he has not died in spite of reaching such
a destination, since the expression μοῖρα κακή, in Homer, occurs only in
reference to the death of Peisander (Il. 13.602).65 But in any case, her reassur-
ances are somewhat undermined by her utter contempt for mortals (D7=B6).
A further factor that contributes to this sinister mood is the goddess’s
anonymity. Various possibilities have been suggested for her identity.66
62
Even by one of the poem’s more sympathetic commentators, Mourelatos (2008a) 35.
63
See n. 39 above. On this relationship, see Palmer (2009) 54–5, who provides further bibliography.
On the topography in particular, see Dolin (1962), Burkert (1969) 8–13 and Pellikaan-Engel (1974).
64
On the connections between Parmenides and the gold tablets, see Feyerabend (1984); Sassi (1988);
Ranzato (2015) 59–71; Tor (2017) 237, 244–6, 267–74. For Plato’s adaptations of this eschatology, see
Edmonds (2004) 159–220.
65
Burkert (1969) 14, 25; Kingsley (1999) 61–2, 240.
66
E.g. Dike: Deichgräber (1958) 6, 7, 37; Peitho: Mourelatos (2008a); Aither personified: Coxon
(2009) 15 (although at 280–1 Coxon sees her as also named Dike); Mnemosyne: Pugliese Cratelli
The most plausible candidate is that she is Night, since the Heliades have
come from the ‘palace of Night’ (D4=B1.9), and the goddess welcomes the
youth to ‘my home’ (D4=B1.25).67 But this possibility is problematised by
the fact that Night is a fundamental feature of the world of the Doxa,
a world which the goddess presents in negative terms in contrast with the
Aletheia that she espouses. An alternative suggestion is that the goddess is
left unnamed as she is a chthonic deity, given the superstition against
uttering the names of the gods of the underworld.68 The hints at katabasis
support this possibility, as does the fact that the goddess, by giving the
youth a choice of ‘paths’, places him, figuratively, at a triodos, a fork in the
road. Triodoi were where the polluted remains of purification rituals were
cast out, as well as the corpses of executed malefactors.69 Here, travellers
could invoke the protection of Hekate or Persephone, both chthonic
deities who could be referred to as ‘god of the wayside’ (ἐνοδία θεός).70
The considerations in favour of identifying the goddess as Night preclude
us from labelling her as Persephone or Hecate with any degree of certainty,
but her anonymity – or, if she is Night, the elliptical manner in which she is
named – imbue her with the ominous tone appropriate for the frightening
spirits of the underworld.
The unveiling of the Heliades (D4=B1.10), while foreshadowing the
goddess’s unveiling of true reality, is another disturbing detail. Veils were
widely worn by women in Archaic and Classical Greece, especially among
elites, but also among lower status groups. Lack of the veil or unveiling was
taken as a sign of immodesty.71 In seeing the unveiled faces of the Heliades,
(1988); Persephone: Kingsley (1999) 94; deliberately anonymous: Tarán (1965) 16; Floyd (1992)
255–6; Conche (1996) 56.
67
As argued by Morrison (1955) 60; Mansfeld (1964): 244–7; Burkert (1969) 17 n. 37 (now asserted with
greater confidence in the reprint of that article for his Kleine Schriften [2008]); West (1983) 109;
Palmer (2009) 51–62; Primavesi (2013a). That Night should serve such an instructive character is
paralleled by her role, in the Derveni and later Orphic theogonies, in instructing Zeus on how to
overcome the first-born god (OF 6, 237, 238).
68
Kingsley (1999) 94; see also Burkert (1969) 13–14, who sees the anonymity as a further initiatory
aspect of the text. On the taboo over referring to certain gods by name, cf. Burkert (1985) 200, 428
n. 12; Parker (2017) 1–2, 5–6. Note e.g. Soph. OC 129–33; Pl. Crat. 403A.
69
Eupol. Fr. 132 PCG; Pl. Leg. 9.873ab. See Halliwell (1986).
70
See Halliwell (1986) 187. On the numinous quality of the crossroads, note also Theophr. Char. 16.14,
with Diggle (2004) 358, 371.
71
The relevant study is Llewellyn-Jones (2003) (esp. 121–88); see also Cairns (2002). Veiling seems to
be a sign of αἰδώς (who herself, personified, is veiled at Hes. Op. 198–200) Cf. Il. 3.139–45; Eur. Hipp.
243–6, IT 372–6, Ph. 1485–92. Blank (1982) 169–70 rightly comments on the intimacy of the
Heliades’ gesture, regarding it as establishing a relationship of trust and citing as parallel Ap.
Rhod. 4.1314–17. Whilst it may be designed to establish a relationship of trust in this way, it certainly
also inspires fear, just as, in the example Blank cites, Jason turns his eyes away from the ἡρῶσσαι out
of reverence for the goddesses (δαίμονας αἰδεσθείς).
86 Parmenides
the youth is granted an illicit perspective on these elite, divine (D4=B1.24)
maidens. We may be reminded of how, in myth, such an unmediated
vision of a goddess typically leads to the downfall of the viewer.72 As Hera
points out in the Iliad, ‘gods are hard to look upon when they are manifest’
(χαλεποὶ δὲ θεοὶ φαίνεσθαι ἐναργεῖς, Il. 20.131)73 and in epiphany scenes,
the appearance of a deity typically inspires awe and terror.74 The image of
the unveiled Heliades, then, is alluring but also potentially terrifying and
prepares us for the face-to-face encounter with the fearsome goddess.
Yet after the unsettling proem and the fearsome encounter with the
goddess, the Doxa seems to warrant quite a different response. More vividly
than in the Aletheia, we are granted a god’s-eye perspective on the world, in
which we see not only the constellations, the sun and the sky but their
origins (D12=B10.3; D12=B10.6; D11=B11) and limits (D12=B10.6–7), as if
we are taken to the very chronological and spatial ends of the universe itself.
In D14=B12, we view the cosmos on a wider scale than would be possible
for mortals to see literally, since we ‘see’ the cosmic rings that make up the
universe (and which presumably surround the Earth) with the daimon at
their centre, who steers all things and rules over sexual reproduction.75
Greek sources consistently present such images of the celestial bodies as
a source of ‘awe’, ‘wonder’ or ‘amazement’.76
A pertinent example is the Iliadic shield of Achilles, which depicts the
Earth, the sky (ouranos), the sea, and the constellations (Il. 18.483–9), as
well as more detailed scenes from the world of mortals.77 Hephaestus
72
Most famously in the myth of Actaeon as it is told by Callimachus (Hymn 5.107–18) and Ovid
(Met. 3.238–52), where he is punished for encountering the bathing Artemis, but this version
probably existed already before the Hellenistic period (see Gantz [1993] 480). Also relevant are
Ixion, punished for his passion for Hera (e.g. Pind. Pyth. 2.35–40; see Gantz [1993] 718–19), and
Semele, who perishes after wanting to see Zeus as he appears to Hera (this detail is made clear at
Diod. Sic. 3.64.3–4; 4.2.2–3, but may already have occurred in Aeschylus’ Semele, see Gantz [1993]
472–7).
73
Cf. also Od. 16.161, 3.420 and 7.201, where the extraordinary status of the Phaeacians is indicated by
the fact that the gods did use to appear manifest to them; Od. 13.312–13, where Odysseus claims it is
‘painful’, even for a very expert man, to recognise Athena.
74
Especially analogous is the apparent scorn of Hesiod’s Muses (Theog. 26). Also comparable is
Anchises’ frightened response to Aphrodite when she reveals her identity to him at hVen. 181–3,
noted as a standard element by Faulkner (2008) 234–5. Cf. Od. 16.179; h.Ap. 440–7 (note esp. 447,
μέγα γὰρ δέος ἔμβαλ’ ἑκάστῳ); h.Cer. 188–90 (note 190, τὴν δ’ αἰδώς τε σέβας τε ἰδὲ χλωρὸν δέος
εἷλεν); h.Hom. 7.32ff. (note 37, ναύτας δὲ τάφος λάβε πάντας ἰδόντας). See further Richardson
(1974) 208–9, 252.
75
The precise reconstruction of the cosmology is tricky, but for a plausible attempt, see Morrison
(1955) 60–2, with a useful diagram on p. 63.
76
E.g. Pl. Tim. 47a–c; Arist. Metaph. 1.3.983b; [Longinus] Subl. §35.4.
77
Reinhardt (1993) 297 also briefly compares Parmenides’ composition of the Doxa to Homer’s
description of the shield, on the basis that both authors were drawn not to ‘things in themselves
professes to make arms for the hero in such a manner that ‘whoever sees it
will be amazed (θαυμάσσεται)’ (Il. 18.467), yet the Myrmidons cannot bear
to look at the shield’s amazing level of detail through fear (Il. 19.14–15).
Achilles on the other hand, as a mark of his quasi-divine status, takes
pleasure in its appearance (Il. 19.18–19), enraged though he is.78 Of course,
there are reasons why the shield is terrifying that would not apply to the
Doxa, and the divine skill required to create such a supernatural physical
artefact would not be necessary for the goddess’s verbal description, but
ecphrasis invites a comparison between the verbal and the visual. The
‘amazement’ that Hephaestus envisages for the arms is programmatic for
our response to the Iliadic poet’s description of the shield. ‘Wonder’ seems
a fitting response both to the poet’s level of detail and to the cosmological
content of the image. Parmenides’ cosmological Doxa appears to have been
similarly detailed and could have bestowed an equivalent feeling of divine
elevation.79
e Parmenides and the Sublime

We are, then, moved from a state of fear, and awareness of our mortal
insignificance, to an elevated, divine position from which we view the
cosmos and its operations in their entirety. A description from
a distinguished Latin literary critic of the sublime qualities of Lucretius
seems peculiarly apt:
first [the Lucretian sublime] generates a sense of inadequacy in whoever
happens to contemplate it; then it elevates the reader’s mind to a higher level
of understanding and fills him with enthusiasm.80
In two substantial, recent studies, James Porter has argued that the
Presocratics discovered matter and therefore the sublime, since they take
the concept of matter to its limits.81 The aesthetic approach to these
authors is to be welcomed, but it is not clear, at least until Anaxagoras,
but an image’. The shield would later be allegorised as depicting various philosophical cosmologies
(including that of Empedocles), on which see Hardie (1985).
78
On these characters’ response to the shield of Achilles, see especially Purves (2010) 46–55 and Scully
(2003).
79
Note Plut. Col. 1114B–C; Simpl. in Cael. 559.26–7 on this section’s level of detail, and D62=B19,
indicating that the account is eternally applicable.
80
Conte (1994) 28. His comment, on Lucr. 1.271–6, is similarly applicable to Parmenides: ‘the listener
has had to confront an image that has carried him to the very limits of exaltation and made him
a “great spectator”’. For possible allusions to Parmenides in Lucretius, see Rumpf (2005).
81
Porter (2010) 138–75, (2016) 415–31, 547–54, though he does not mention Parmenides in this context.
88 Parmenides
that they would have articulated a distinction between ‘matter’ or the
‘material’ from the immaterial in any more explicit way than is found in
earlier texts.82 Be that as it may, although the ‘sublime’ would not have
been recognised as a distinct critical concept for Parmenides or his audi-
ence, the poem displays some striking resemblances to Longinus’ examples
of the quality. As an instance of phantasia, an author’s visualisation, he cites
the chariot journey from Euripides’ Phaethon (Subl. §15.3–4), a myth that is
also evoked in Parmenides’ proem, with its chariot led by the Heliades.
According to Longinus, Euripides creates the impression that ‘the soul of
the writer was travelling with him on the chariot, and, sharing the danger
with the horses, had taken wing with them’ (15.4). Parmenides’ proem
invites a similar reading: we might assume that Parmenides’ ‘soul was
travelling with him on the chariot’, especially since the journey is narrated
in the first person and was read as a journey of the soul in antiquity
(R8=Sext. Emp. Math. 7.112–14). Longinus expands on this notion that
the sublime author’s soul ‘takes wing’ when outlining the principle of
literary ‘genius’ (megalophuia), in a passage which bears further resem-
blance to Parmenides’ poem:
Τί ποτ’ οὖν εἶδον οἱ ἰσόθεοι ἐκεῖνοι καὶ τῶν μεγίστων ἐπορεξάμενοι τῆς
συγγραφῆς, τῆς δ’ ἐν ἅπασιν ἀκριβείας ὑπερφρονήσαντες; πρὸς πολλοῖς
ἄλλοις ἐκεῖνο, ὅτι ἡ φύσις οὐ ταπεινὸν ἡμᾶς ζῷον οὐδ’ ἀγεννὲς †ἐ..κρινε τὸν
ἄνθρωπον, ἀλλ’ ὡς εἰς μεγάλην τινὰ πανήγυριν εἰς τὸν βίον καὶ εἰς τὸν
σύμπαντα κόσμον ἐπάγουσα, θεατάς τινας τῶν ἄθλων αὐτῆς ἐσομένους
καὶ φιλοτιμοτάτους ἀγωνιστάς, εὐθὺς ἄμαχον ἔρωτα ἐνέφυσεν ἡμῶν ταῖς
ψυχαῖς παντὸς ἀεὶ τοῦ μεγάλου καὶ ὡς πρὸς ἡμᾶς δαιμονιωτέρου. (3)
διόπερ τῇ θεωρίας καὶ διανοίας τῆς ἀνθρωπίνης ἐπιβολῇ οὐδ’ ὁ σύμπας
κόσμος ἀρκεῖ, ἀλλὰ καὶ τοὺς τοῦ περιέχοντος πολλάκις ὅρους ἐκβαίνουιν αἱ
ἐπίνοιαι, καὶ εἴ τις περιβλέψαιτο ἐν κύκλῳ τὸν βίον, ὅσῳ πλέον ἔχει τὸ
περιττὸν ἐν πᾶσι καὶ μέγα καὶ καλόν, ταχέως εἴσεται πρὸς ἃ γεγόναμεν.
What, then did those godlike men see, who aimed at the greatest achieve-
ments of writing, but disdained exactitude in all areas? Besides many other
things, it was that nature did not [make?] us to be a low animal, nor man to
be an ignoble thing, but led us into life and into the whole cosmos as if to
a great festival, to be spectators at the contests and also the most glory-loving
contestants. She engendered in our souls an irresistible desire always for
everything that is great and, in relation to ourselves, supernatural. (3) For
82
On this topic see, recently, Betegh (2016) 415–20, who points out that all allegedly ‘incorporeal’
items in Presocratic cosmologies are divine or godlike cosmic agents, and so conform to the
traditional notion that divine bodies do not behave like ordinary objects. Contrast Curd (2013a),
(2013b) (who identifies a theory of incorporeality) against Renehan (1980) (for whom the theory only
arises with Plato).
that reason, the whole cosmos is not sufficient for the range of human
speculation and intellect, but thoughts often go beyond even the limits of
the boundary [sc. of the cosmos], and if anyone were to look around at his
life in circumspect, how much the extraordinary, and the great and the
beautiful abounds in all things, he will swiftly know that for which we were
born. (Subl. 35.2–3)83
The great writers are isotheoi, godlike; Parmenides presents the youth’s
education as one of homoiōsis theōi, in which he is granted a divine
perspective. Man’s capacity to transcend his immediate experience is
something that these authors have ‘seen’; the content of Parmenides’
poem, as we have seen, is presented in highly visual terms. The whole
kosmos is said to be insufficient for mortal speculation and thought;
Parmenides’ goddess similarly denigrates the kosmos as ‘deceptive’ and
offers a vision of true reality separate from it, but accessible to thought.
Most strikingly of all, mortal thoughts are capable of escaping its limits;
Parmenides’ proem, like Euripides’ Phaethon, depicts a journey beyond the
normal limits of the kosmos.
These resemblances are not coincidental. Whether or not we take the
author of On the Sublime to be the third-century ce Platonist philosopher
Cassius Longinus, the treatise is part of a Greek philosophical tradition to
which Parmenides is forefather. The idea that mortals can transcend the
limits of the perceptible cosmos through thought, and that by this means
they can most closely assimilate to the divine, seems distinctly Platonist
and stemming from those parts of Plato that are most heavily indebted to
Parmenides.84 Longinus’ more specific image of departure from the cos-
mos seems indebted to the myth of Plato’s Phaedrus, of the chariot journey
of the soul (246a3–b4), which itself has been seen to be influenced by
Parmenides’ journey.85 With this lengthy philosophical heritage, there are,
of course, divergences from Parmenides. Longinus seems more optimistic
about the cognitive faculties of humans in general than the goddess. But in
spite of the chronological distance between On Sublimity and Parmenides,
83
My translation here is in part adapted from Russell’s in Russell and Winterbottom (1972). In spite of
the textual uncertainty in the second sentence, the overall sense is clear enough. For τοῦ περιέχοντος
as ‘the boundary’ of the cosmos, rather than Russell’s ‘our surroundings’, see Halliwell (2011)
343 n. 35.
84
Especially the Phaedrus myth (see n. 85 below), but note also Resp. 5.467e–477b with Palmer (1999)
56–87; and Tim. 90a–d, in the context of a ‘likely account’, that is, one framed in pointedly similar
terms to Parmenides’ Doxa (see Bryan [2012] 161–95; Johansen [2016] 18–19). On the Platonist
aspects of On the Sublime, see Heath (2013) 169–79.
85
Halliwell (2011) 343, 352 argues for the influence of the Phaedrus here. On the relationship between
Parmenides and the Phaedrus, see Palmer (1999) 17–30 and Slaveva-Griffin (2003).
90 Parmenides
the treatise can, like other works of ancient literary criticism, be seen as
a crystallisation of trends that are already in evidence in much earlier
texts.86 The Longinus passage gives an apt impression of the sort of
enthralling and enthusing response envisioned for Parmenides’ audience,
once they have grasped the goddess’s teachings. The responses to the
Longinian sublime, of ‘ecstasy’ (ἔκστασις), along with ‘wonder’ and ‘aston-
ishment’ (σὺν ἐκπλήξει . . . τὸ θαυμάσιον, Subl. §1.3–4) seem appropriate
for Parmenides’ imagery. The goddess is more pessimistic than Longinus
about the mental capacity of mortals, but the notion that we can, at least
temporarily, transcend our mortal limitations is implicit in her education
of the mortal youth.
f Mystic Initiation
This emotional trajectory has religious associations and is an aspect of the
poem’s evocation of Mystic initiation.87 Just as Parmenides’ poem culmin-
ates in the sublime vision of the Doxa, initiation into the highest grade of
the Eleusinian mysteries reached its climax, after a journey on the ‘sacred
way’, in a ‘viewing’ of a certain spectacle in the Anaktoron/anaktora, the
innermost chapel of the telestērion.88 Though nominally secret, our sources
tell us that what was revealed was an ear of wheat, and perhaps, a statue of
Demeter, and a phallus, amid a shining light.89 In any case, this process is
referred to as epopteia, a ‘viewing’, and those who have been initiated are
said to have ‘seen’ the mysteries.90 What was seen was ineffable, both in the
sense that it was forbidden to speak of it and, at least according to one
prominent historian of religion, in the sense that it was hard to put into
86
On this topic, see Hunter (2009), who discusses On the Sublime at 128–68.
87
As Burkert (1969) 14, 25 pointed out, mystic initiation is suggested by the objectless use of εἰδότα,
‘knowing man’ (D4=B1.3), for which the main parallels refer to initiates into the Eleusinian
mysteries (Andoc. de Myst. 30, [Eur.] Rhes. 973, and Ar. Nub. 1241). The concordances with the so-
called ‘Orphic’ gold tablets also support this notion (see n. 67 above). Tor (2017) 254–7, 267–74
discusses Parmenides’ relationship to mystic initiation thoroughly. Cf. also D4=B1.28 with OF 3.
88
Paus. 1.36.2 refers to the road to Eleusis as the ‘sacred way’ (ὁδὸν ἱεράν). For reconstructions of the
process of initiation, from the complex variety of sources, see Burkert (1983) 248–97, Clinton (2003)
and Bremmer (2014) 1–20.
89
Hippol. Haer. 5.8.39 (for the ear of wheat); Tert. Adv. Valent. 1 (for the phallus); Plato Phdr. 250b–c,
Plut. fr. 178 Sandbach, Dio Chrys. Or. 12.33 (for the light). See Burkert (1983) 275–7, Clinton (2003)
65–6 and Bremmer (2014) 11–15.
90
As Bremmer (2014) 15 writes, ‘the importance of “seeing” and “showing” is continuously stressed by
our sources as a fundamental component of the highest degree of initiation’, citing hDem. 480;
Pindar, fr. 137.1; Soph. fr. 753.2; Eur. Her. 613, Hipp. 25; Andoc. de Myst. 31; Ael. Arist. Or. 22.2. Note
also Plato Phdr. 250b–c. Cf. also Burkert (1983) 274, ‘The goal of initiation is the path to Eleusis and
to seeing what occurred in the great chamber of initiation on the sacred night.’
words.91 There is some evidence to suggest that, in the period leading up to
this ‘viewing’, the first-time participants, or mustai, had to wander, blind-
folded, in a search for Kore/Persephone (in imitation of Demeter’s search for
Persephone), aided by their mystagogues, friends who had already been
initiated, whilst other epoptai (those who had been initiated in previous
years) looked on.92 This would explain why the process leading up to the
viewing is frequently described by ancient sources as a fearsome
experience.93 Like the initiates, the youth has undergone a fearsome jour-
ney along a sacred path, at the end of which there is a divine revelation of
a truth that could not be expressed in ordinary mortal speech, resulting in
a superior status. The goddess is herself similar to a mystagogue in guiding
him. As a consequence of her teaching, he is able to ‘look’ on at the mortals
who, in the manner of the mustai, are wandering blindly (D7=B6).
These parallels should not lead us to conclude, as some have,94 that the
poem is really a mystic text, intended to represent an actual process of
initiation and to induce a similar experience in its audience. There is no
mention of ritual,95 and, as we have seen, the early reception history of the
poem counts against the possibility that it was designed for a group of
initiates (see Section 1 above). Rather than offering a model for a literal
initiation into a particular mystery cult, the poem offers the possibility of
an emotional experience which in some respects overlaps with the familiar
(even if nominally secret) experience of mystery-initiates. The hints of
initiation illustrate how the youth’s acquisition of knowledge grants him
a privileged status over other mortals.
g The Reception of Parmenides’ Poem

Some might object that the poem fails to affect its recipients in such
a manner. Many readers, ancient and modern, have experienced puzzle-
ment and frustration rather than any life-changing transformation.96
But certain moments in its reception suggest that it inspired awe in at
91
Burkert (1987a) 69, discussing Procl. In Remp. 2.108.17–30 Kroll and Aristotle fr. 15=Synesius Dio 10
p. 48a.
92
See Clinton (2003) 65–6. Plut. fr. 178 attests to the fearsome ‘wandering’; Lactant. Div. inst. 23 to the
search for Kore.
93
Plut. fr. 178; Proclus, Theol. Plat. 3.18; Aesch. fr. 387; Plut. Ages. 24.7; Ael. Arist. 22.2; Luc. Cataplus
22. See Bremmer (2014) 13–14, who suggests that the fear may have been caused not by blindfolded
wandering (pace Clinton) but by the image of a female monster with snaky hair.
94
See n. 6 above.
95
Contrast e.g. the allusions to ritual in the Homeric Hymn to Demeter 472–82.
96
See the scholarship cited at n. 1 above.
92 Parmenides
least some of its ancient readers.97 The Platonic description of
Parmenides as ‘venerable and awesome’ (αἰδοῖός . . . δεινός τε, Tht.
183e)98 could be taken, like other ancient biographical comments on
authors, as a response to Parmenides’ poetry. The goddess is certainly
‘venerable and awesome’ in her powers of deduction and her contempt
for mortals. More explicitly, Plutarch includes Parmenides in a list of
authors (also including Archilochus, Phocylides, Euripides and
Sophocles) with whom one might find a particular fault, in spite of
the fact that ‘each one of them is praised for the special faculty by which
he is endowed by nature to move us and draw us on’ (De aud. poet. 13
45A–B=R2b=A16).99 The comment that the poem ‘moves’ us refers in
particular to its ability to move us emotionally.100 Elsewhere, Plutarch
describes Parmenides and Empedocles as ‘borrowing from poetry its
weight and metre, like a chariot, in order to escape the pedestrian’
(κεχρημένοι παρὰ ποιητικῆς ὥσπερ ὄχημα τὸν ὄγκον καὶ τὸ μέτρον,
ἵνα τὸ πεζὸν διαφύγωσιν, De aud. poet. 2 16C=A15). The expression
deploys what appears to have been a standard Imperial-era image in
figuring verse as a ‘chariot’ (ὥσπερ ὄχημα) and prose as ‘pedestrian’ or
‘on the floor’ (τὸ πεζὸν),101 but it is unlikely to be a coincidence that the
image of the chariot is also used by both Parmenides and Empedocles
(Parmenides D4=B1; Empedocles D44=B3.5). The Parmenidean version
is evoked in particular because the chariot mentioned by Plutarch soars
away from the ground level (‘fleeing the pedestrian’), just as Parmenides’
chariot, in following the path of the sun, appears to travel across the sky.
Plutarch’s comment identifies an elevated tone in Parmenides’ verse,
which corresponds to the elevating effect that we have identified in the
Doxa, in granting us a divine perspective on the cosmos. But even if, in
practice, a transformative experience of the poem has rarely been real-
ised, the narrative tempts us with its possibility, creating the impression
that, if only we could understand the goddess’s teachings, our outlook
on the world would be radically altered.
97
For an overview of the reception of Parmenides’ poetry in later ancient Greek texts, see Popa
(2000).
98
Adapting an expression used of Priam at Il. 3.172.
99
ἕκαστός γε μὴν ἐπαινεῖται κατὰ τὸ ἴδιον τῆς δυνάμεως, ᾧ κινεῖν καὶ ἄγειν πέφυκεν (45B). The
passage comes in a section in which Plutarch argues that a good listener should find something to
praise in any scholarly lecture, in spite of its faults.
100
Emotions are said to ‘move’ their subjects in Greek from at least as early as Aesch. Cho. 289.
101
See Hunter and Russell (2011) 88. Cf. Strabo 1.2.6.
Poetry and Symbolism 93
h The Emotions and Parmenides’ Philosophy
If this interpretation is accepted, we can offer an explanation for
Parmenides’ use of hexameter narrative that differs both from that of
Mourelatos and from the ‘esoteric’ interpretations. Traditional patterns
provide a basis, not only for Parmenides’ conceptions of intellectual error
and logical necessity, but also for the cognitive and emotional experience
that intellectual inquiry involves. As in Odysseus’ Nekuia, Bilgames/
Gilgamesh’s vision of the netherworld, and the process of mystic initiation,
a frightening experience leads up to a transformative revelation that alters
the protagonist’s perspective on the world. In Parmenides’ poem, the
protagonist slips out of the narrative so that the audience themselves can
feel imaginatively the emotions that are elsewhere experienced by heroes in
the third person. Through presenting a model of this process with such
emotive associations, Parmenides gives an impression of what it would be
like to comprehend the object of true reality, an entity that seems so hard to
capture in ordinary language.
5 Poetry and Symbolism

As a heuristic, I have distinguished the destructive from the positive
functions of Parmenides’ use of poetry. On the one hand, it illustrates
the deceptiveness and malleability of mortal senses, whilst on the other, it
leads us to an imagined encounter with true reality. But these two func-
tions are closely intertwined. For, as we saw in the case of Xenophanes, our
immersion into the imagined world of the poem invites critical reflection
on the triangular relationship between that world, the world of our senses,
and the world of true reality. One means by which Parmenides prompts
this kind of reflection is in phrasing his language and imagery in such
a manner as to provoke an exploration of their deeper, symbolic
significance.
a Intratextuality
This feature is evident in the intratextual reverberations of particular words
and images across the poem. The opening fragment prefigures certain
details of the Doxa. The peculiar focus on the chariot’s whirling wheels
(D4=B1.7–8) is echoed in the fiery cosmic rings in the Doxa (D14=B12;
D15<A37). Both sections feature a daimon who is not clearly identified
(D4=B1.3; D14=B12). Most obviously, the paths of ‘Night and Day’
94 Parmenides
foreshadow the component elements of the cosmology. There are also
connections between these entities and the Aletheia: the ‘paths’ of Night
and Day, once the goddess starts speaking, become the paths of is and is-
not, while the language that is used of the two elements in the Doxa recalls
the outline of what-is.102 These correspondences have led some commen-
tators, starting with Aristotle (Gen. corr. 1.3, 318b7–8), to regard Light/Day
and Night as allegorical labels for, respectively, Being and Not-Being,103
but if they are allegorical in this manner then they are conspicuously
imperfect representations of those two possibilities. The contrasts are
significant: Day and Night alternate, whereas what-is is unchanging. The
bright objects of the Doxa move, but what-is is unmoving. Night, unlike
what-is-not, is an entity that can be pointed out and cognised (cf.
D6>B2.5–8; D8.13–14=B8.8–9). Night and Day/Light mix with one
another (D51=B16), whereas what-is is unmixed (cf. D8.47–54=B8.42–9).
Whether or not we are to regard Day/Light and Night as ‘allegories’, the
fact that they share the path imagery and certain linguistic similarities with
the Aletheia encourages us to reflect upon how the sensible world, with its
alternations of Night and Day, might relate to true reality.
b Intertextuality
In addition to this intratextual approach, sources from outside of the poem
itself enable us to extrapolate the symbolic meanings that Parmenides’
language and imagery could have conveyed for his earliest audiences. Of
the surviving fragments, the allusive proem most clearly draws upon
a variety of different traditions with different possible connotations. The
journey has variously been regarded as a successful re-enactment of
Phaethon’s chariot ride, an ascent to Olympus, a katabasis to the land of
the dead, a process of mystic initiation, or an adaptation of Odysseus’
travels.104 The references to these various traditions can be interpreted in
102
See n. 31 above.
103
This view has been endorsed by Vlastos (1946) 73–4; Primavesi (2005); and myself in Mackenzie
(2017), though I would no longer subscribe to it so strongly. Cf. also Metaph. 1.5, 986b18=R12=A24,
which expresses a similar, but subtly different interpretation.
104
E.g. Phaethon: Kranz (1916); Bowra (1937) 103–4; Palmer (2009) 56 with n. 16; katabasis: Morrison
(1955) 59–60; Kingsley (1999) esp. 50–3; Palmer (2009) 55–6; cf. Burkert (1969) 1–15 (who sees it as
a journey to a place ‘beyond’, rather than, strictly speaking, a katabasis); Odysseus: Havelock (1958);
anabasis: Diels (1897) 7–8; Bowra (1937); Fränkel (1975a) 1–6; Kahn (2009) 210–25; mystic initi-
ation: Burkert (1969) 5; Kingsley (1999) 62; Palmer (2009) 58. Mourelatos (2008a) 11–41 argues that
there are shared motifs, but not direct allusions to other journeys. For an overview of this debate
with further references, see Tor (2017) 347–59.
light of Parmenides’ doctrinal claims. To succeed where Phaethon failed
would be to surpass mortal limitations and to enjoy the traditionally all-
seeing perspective of the Sun.105 The passage through the gates (D4=B1.20–1)
adapts the Iliadic description of Hera driving her chariot through the gates of
Olympus (Il. 5.749–52=8.393–6),106 thereby hinting that the youth’s journey
is one of homoiōsis theōi. A process of divinisation is also implicit in the
evocations of katabasis and initiation, since initiation in various sources is
depicted as a symbolic death and rebirth and, in at least some instances,
promised the attainment of divine status.107 The associations of these images
therefore hint at the Doxa’s theory of epistemology, in which the acquisition
of knowledge amounts to an assimilation to the divine.108
In assessing the relationships with the Phaethon myth, katabasis poetry and
mystic initiation, we have to make do with fragmentary or later evidence. With
the Odyssey, we are on firmer ground, but if we can analyse Parmenides’
response to this epic in greater detail than the other references it is also because
his allusions to it are particularly rife. In a much-cited article, Havelock first
identified the full extent of these allusions, concluding that the Odyssey
provided a model for a journey outside of the ordinary realm, with
Odysseus’s reckless crew standing for the ignorant mortals.109 Mourelatos
refined Havelock’s argument with a closer examination of the relationship
between the Odyssey’s language of journeying and the arguments of the
Aletheia.110 With a greater focus on the epistemological presuppositions of
early epic, James Lesher has shown that the goddess guides the youth through
an inferential process that has much in common with the recognition scenes of
the Odyssey.111 Yet Parmenides’ engagement with the Odyssey can be explored
105
E.g. Il. 3.277=Od. 11.109, 12.323.
106
The expression in Parmenides τῆι ῥα δι’ αὐτέων (D4=B1.20) is paralleled nowhere else but in the
near-identical phrase in these Iliadic passages, τῇ ῥα δι’ αὐτάων (Il. 5.752, 8.396), describing the
movement of Hera’s chariot. The gates in Homer are guarded by the Horai, one of whom was Dike
(along with Eunomia and Eirene, see Hes. Theog. 901–2), who guards the door in Parmenides’
proem.
107
On initiation as symbolic death and rebirth (in a pagan context) see Burkert (1987a) 99–101 and
note the key sources, Plut. fr. 178 Sandbach and Apuleius Met. 11.21.1. Bernabé (2014) argues that
the ritual described in the Derveni papyrus is most likely to be an initiation that involves an imitatio
mortis. For initiation and divinisation, see Tor (2017) 270–3, and note especially OF 487.4, 488.9.
108
Tor (2015) 22–32 and Tor (2017) 277–84. See similarly Long (1996) 146.
109
Havelock (1958). For further parallels and possible allusions, see Floyd (1992) and the commentary
of Coxon (2009).
110
Mourelatos (2008a) 17–25, 31–4.
111
Lesher (1994) 27–34, (1999) 239. Most conspicuously, both Eurykleia and Penelope recognise the
truth of the disguised Odysseus’ identity through his semata (Od. 19.217, 23.110, 206), just as the
youth is to recognise the semata of the goddess (D8.7=B8.2). On this term in the Odyssey and its
relationship to the vocabulary of noos and aletheia, see Nagy (1990) 202–22.
96 Parmenides
still further. We saw above that the imagery of the Sirens hints at the seductive
and deceptive nature of the Doxa. More generally, the Odyssey persistently plays
on the themes of the truthfulness of song and the accuracy of names, themes
that recur prominently in Parmenides’ poem. Famously, the first word of the
Odyssey is the unspecified ‘man’ rather than ‘Odysseus’, and the hero himself is
not named until line 21. As Norman Austin argued, this forms part of the wider
theme of ‘name-magic’ in the poem: Odysseus’ main mistake will be in
revealing his name to Polyphemus (Od. 9.502–36), whilst the climactic recog-
nition scene, in which the nurse Eurykleia recognises the disguised the
Odysseus for who he is, is also the point at which we learn the true origins
of Odysseus’ name (19.386–475).112 Parmenides borrows this technique in the
proem, refraining from identifying the horses, the path, the daimon, the youth
or the goddess. Although the path seems most plausibly to be that of the Sun,
the narrator is notably elliptical in his reference to it. The ‘path of the daimon,
which carries the knowing mortal to every city’ –113 an expression that may be
modelled on the Odyssey’s ‘man who . . . saw the cities and got to know the
minds of many men’ (Od. 1.1, 3) – tantalises us with the possibility that it is
a familiar road whilst refraining from making its identity explicit. This
reticence assumes a particular significance within the wider context of the
poem, in the stress placed on the act of ‘naming’ in the later fragments. The
anonymity of these key figures in the proem invites speculation as to their
identity but also prefigures the point that names themselves are misleading and
superficial. The recollection of the Odyssey, then, raises the themes of naming
and identity that will be so important in the goddess’s arguments.
In the face of this variety of associations, scholars have sometimes
attempted to pin down singular referents to the proem’s language and
imagery, as if all audience members would have shared assumptions about
the topography of the extraordinary locations mentioned,114 but this
notion is highly implausible, since what little evidence we have reveals
the existence of various different accounts of the cosmic geography outside
of the mortal realm.115 On the other hand, Mourelatos has argued that
Parmenides’ use of ostensibly incompatible symbols is a function of his
112
Austin (1972).
113
See Lapini (2013) 37–86 in support of reading here κατὰ πάντ’ ἄστη as a conjectural emendation for
the manuscripts’ incomprehensible πάντ’ ἄτη (even though it was falsely thought by Mutschmann
to occur in the oldest MS of Sextus Empiricus, as Coxon [1968] discovered).
114
The methodological principle that the proem could only have one frame of reference is espoused
explicitly by Havelock (1958) 135–6, but even more recent commentators (e.g. Palmer [2009] 54–61)
have been similarly reductive in approach.
115
Compare the various accounts in Hesiod’s Theogony, Mimnermus fr. 12 W and Stesichorus fr. 8
Finglass. See the study of Romm (1992), esp. 9–31.
‘archaic mentality’,116 but a poet may make use conflicting images as
a result of a deliberate choice in order to create certain effects and not
necessarily as a consequence of a pre-logical or mythical way of thinking. In
fact, Parmenides’ choice of multiple frames of reference seems quite
deliberate. He self-consciously inscribes the poem within a multifarious
poetic tradition when referring to the road as πολύφημος (D4=B1.2),
a word usually translated in the passive sense of ‘far-fabled’ or ‘far-
famed’. The adjective thus indicates that the path has already been the
subject of many songs.117 Yet, in its two Homeric occurrences, both in the
Odyssey, it seems to have a more active meaning, in describing the agora
(Od. 2.150, ‘Of many voices’, cf. Hdt. 5.79) and the singer Phemius
(πολύφημος ἀοιδός, ‘singer of many songs’ Od. 22.376). Pindar uses it of
the choral singing of a dirge (‘Of many voices’, Isthm. 8.58). In Parmenides,
it may simply be left undetermined whether the voice is active or passive,
but the active usage leaves open the possibility that the road is ‘of many
songs’, not only in the sense that the particular path has been sung of many
times, but also in the meta-literary sense that this particular poetic journey
alludes to multiple previous ones. The self-consciousness of this variety
prompts us to ponder the motivation for these different references and, in
particular, to consider their possible relationship to the claims put forth so
forcefully by the goddess.
c Allegory and Subtext

Rather than following any single familiar route, the proem is suggestive,
hinting, through underdetermined details, through language that will be
echoed later in the poem and through allusions to other texts and tradi-
tions, that this journey bears a significance that goes beyond its overt
meaning. It is this quality that has led commentators, since at least as
early as Sextus Empiricus, to regard the proem as an allegory of enlighten-
ment. Recent scholars have rejected the label of ‘allegory’ on grounds of
anachronism.118 Certainly, Sextus’ interpretation of the proem (Math.
7.112–14=R8) as the departure of the rational soul is both internally
116
Mourelatos (2008a) 29–30. Cf. similarly Gheerbrant (2017) 647. This type of explanation was
influentially promoted in the early twentieth century by Lévy-Bruhl (especially in La Mentalité
primitive, Paris, 1922), but for cogent criticisms (with particular respect to the symbolic/scientific
dichotomy), see Lloyd (1990) 14–38.
117
Tarán (1965) 10 argues for an active sense of ‘“uttering many things,” i.e. bestowing knowledge’.
118
Gemelli Marciano (2008) 26; Graham (2010) 234. For a measured defence of the allegorical
interpretation, see Granger (2010) 33–4.
98 Parmenides
inconsistent and suspiciously Platonising.119 But the impulse to seek an
underlying significance to the journey in terms of the fundamental philo-
sophical principles it promotes is not anachronistic. As we saw in the
Introduction, some of our earliest surviving texts presuppose such
a mode of interpretation.
If the allegorical approach is objectionable, it is because it assumes that
there is only one possible correct interpretation that the proem encodes,
whereas the multiplicity of its references suggests otherwise. Most com-
mentators would agree that the passage, in depicting a journey outside
the familiar realm, symbolises an intellectual process that treats the
familiar world of mortal experience as deceptive, yet the symbolic signifi-
cance can be explored further, into more subjective territory. The wheels
may foreshadow the rings of the Doxa, and the allusions to the Odyssey
may recall the epic’s themes of naming and identity, but not all readers or
audience members are likely to pick up on these details. With symbolic
texts, one can never be quite sure how far the interpretation can legitim-
ately be taken.120 But what may be more important, with regard to the
poem’s didactic purpose, is the fact that it invites this sort of subtextual
reading to begin with. Such a hermeneutic process mirrors the way in
which we are to look beneath our sensory experience in order to identify
an underlying true reality. In this way, Parmenides’ use of a symbolic
form of discourse can be regarded as a strategy to circumvent the prob-
lems with mortal language identified by the goddess: though language
reflects the deceptive, pluralistic world of mortal experience, with
Parmenides’ language, we are to look beyond its overt meaning if we
are to understand true, unchanging reality.
This is not only the case with the proem and the Doxa, the sections of
the poem that reflect the deceptive world. The Aletheia develops the
symbolism of the proem and resembles certain stereotypically enigmatic
forms of speech. The lengthy argument of D8>B8 continues the katabasis
119
For Sextus, the κοῦραι are τὰς αἰσθήσεις, yet the Heliades are described as τὰς δὲ ὁράσεις, the eyes
or the vision, and the wheels of the wagon are the ‘round parts of the ears’ (δοιοῖς . . . κύκλοις’,
τουτέστι τοῖς τῶν ὤτων). The κοῦραι of D4=B1.5 and the Ἡλιάδες κοῦραι of D4=B1.9 are one and
the same, so either Sextus does not recognise this point, or the maidens represent simultaneously
eyesight and the senses in general, whilst the wheels represent hearing. In either case, the corres-
pondence drawn is implausible. For further discussion of Sextus’ interpretation, see Tarán (1965)
17–22 and Palmer (2009) 51–3.
120
For this axiom of literary hermeneutics, see (e.g.) Jakobson (1960) 370–1; Barthes (1974) 1–21,
216–17; Culler (1975) 224–30, (1981) 52–6; and, specifically within the discipline of Classics, Hinds
(1998) 17–51 (with reference to intertextuality); Laird (2003) 153. All these studies are in
a Structuralist vein, but for a pluralistic approach to literary interpretation that does not accept
Structuralist tenets, see Heath (2002) 39–57.
theme, with its Tartarean imagery of binding and punishment (D8.19,
35–6=B8.14, 30–1). The fragment’s abundant use of repetition, especially of
ou/oude and esti, creates assonance and alliteration, resembling incantation
and magical spells.121 The fragment also recalls prophecy. The divine
speaker refers to the arguments she makes as semata, ‘signs’, a term which
is elsewhere used of divine portents and of the linguistic signs of Delphic
oracles.122 D8.10=B8.5 and D8.25=B8.20 adapt the formulaic line used by
Homer of the knowledge of the prophet Calchas, and by Hesiod of the
song of the Muses:
ὃς ᾔδη τά τ’ ἐόντα τά τ’ ἐσσόμενα πρό τ’ ἐόντα
[sc. Calchas] who knew the things that are, the things that
shall be and the things that came before
(Il. 1.70≈Hes. Theog. 38)
οὐδέ ποτ’ ἦν οὐδ’ ἔσται, ἐπεὶ νῦν ἔστιν ὁμοῦ πᾶν

nor was it nor shall it be, since it is now, together, whole (D8.10=B8.5)
εἰ γὰρ ἔγεντ’, οὐκ ἔστ(ι), οὐδ’ εἴ ποτε μέλλει ἔσεσθαι
for if it was born, it is not, nor if it is ever going to be (D8.25=B8.20)
Parmenides’ adaptation of the formula illustrates the contrast between the

goddess’s world view and that of convention. The Homeric line’s pro-
phetic wisdom of past, present and future concerns the world of becoming,
which, from the goddess’s perspective, cannot be the object of authoritative
knowledge.
Oracles are enigmatic and typically misunderstood, while incantations
and magical spells ‘display a propensity verging on obsession to create
abnormal words’.123 The evocation of these speech-genres therefore marks
up the way in which the goddess’s speech is, like all hexameter poetry,
highly defamiliarised from ordinary speech. Indeed, the formalist concept
of defamiliarisation is appropriate for the goddess’s speech in at least two
further respects. It defamiliarises traditional epic language by adapting and
altering formulaic lines and, in a sense closer to Shklovsky’s original
121
See Gemelli Marciano (2008) 44–5. Note especially the anaphora at D8.27–9=B8.22–4. For
repetition, assonance and alliteration as features of incantations and voces magicae, see the evidence
gathered by Versnel (2002) 130–5, and also Segal (1974) esp. 146–51. A pertinent parallel is the highly
alliterative and assonant Empedocles D43=B111.
122
LSJ s.v. 1. For the Delphic oracle, note, famously, Heraclitus D41=B93, on which see Tor (2016) for
detailed discussion. Cf. also Hes. Theog. 498–500.
123
Versnel (2002) 141.
100 Parmenides
definition of the term,124 it defamiliarises the world, presenting familiar
entities differently from how they are usually regarded. Through defami-
liarisation in this latter sense, texts prompt the reader (or audience mem-
ber) to reflect, in a more prolonged manner, on phenomena that have
previously been taken for granted. The markedly mysterious and enigmatic
qualities of the goddess’s speech demand this sort of response, leading an
audience to reconsider the meaning of one of the most prosaic and
seemingly unremarkable items in the Greek lexicon, esti. She even instructs
us to interpret her language in a critical and penetrating manner, telling us
to ‘judge by reason the much-disputed refutation spoken [by her]’
(D8.5–6=B7.5–6).125
6 Conclusion
Parmenides combines the characteristically deceptive, immersive and sym-
bolic aspects of poetry, three functions that are closely intertwined: we are
granted a perspective on a world separate from our immediate experience
by means of the poem’s self-consciously deceptive mimesis. This process
has axiological and emotive dimensions – we see what it is like to view the
world from a divine but contemptuous perspective and this appears as
a frightening and enthralling experience – but it also invites reflection on
the nature of deception, that is, on the ways in which the world of the
poem and the world of our sensory experience relate to unchanging reality.
They do so (from the goddess’s perspective) in an oblique manner, so that,
in order to comprehend true reality, we cannot take their details at face
value. The poem marks up this quality by recalling characteristically
symbolic forms of discourse, such as incantations and oracles and even
epic poetry itself. Through interpreting the poem in a symbolic manner,
we are to look beyond its deceptions in order to recognise true reality,
a procedure that prepares us to apply a similar interpretation to our sensory
experience.
124
See Shklovsky (1990). Although ‘estrangement’ may be etymologically closer to the original Russian
term ostranenie, I use ‘defamiliarisation’ as it has entered the anglophone critical vernacular.
125
The notion that we are to judge for ourselves, that is, by our own reasoning, the validity of the
goddess’s account, is stressed e.g. by Curd (2002) 134–5 and Lesher (2008) 473–4. Kingsley (2003)
126–40 emends the dative λόγωι to the genitive λόγου on the grounds that it is too early for the
word to have the sense of ‘reason’ or ‘reasoning’, but the term comes close to this sense in Heraclitus
(esp. D46=B50, see Kahn [1979] 94–5, 97–102, 130–2), and Kingsley’s reasoning seems problemat-
ically circular: why could Parmenides not be the earliest attested example of this usage? On this
issue, see the detailed discussion of Granger (2010) 30–3.
Conclusion 101
If this interpretation is accepted, some more general conclusions can be
drawn concerning Parmenides and Greek poetics. While it is undoubtedly
true that performed verse, during a period of incipient literacy, served both
as a mnemonic and as a means of reaching a wider public than prose, these
practical benefits provide only a limited explanation for the significance
and effects of the medium and Parmenides’ use of it. Equally limited is the
theory that he wrote in verse because this was the culturally sanctioned
medium by which to pronounce authoritatively true claims. For he also
composed verses in order to provoke certain types of response from his
audience. In this respect, Parmenides is a true poet and not a mere versifier.
His use of verse is not simply an unthinking conformity to tradition but
a self-conscious manipulation of a particular type of discourse.
What is more, through alluding to meta-literary episodes such as
Hesiod’s Theogony proem and the Sirens episode of the Odyssey, he engages
in an ongoing discussion concerning the nature and function of song. This
contribution can be regarded as an important moment in the emergence of
the Classical conception of poetry. With the Doxa, the goddess distin-
guishes a section of her account as both verse and as a mortal fabrication to
be held up to critical scrutiny. In distinguishing this section, she operates
with a conception of both language in general, and poetry in particular, as
something that can be isolated and analysed as distinct entities. In fact, in
presenting the Doxa as a poetic world deriving from mortal opinion, she
comes close to Platonic conceptions of literary mimesis. Her stark epis-
temological distinction between divine knowledge and mortal opinion is
a prerequisite to the Platonic idea that poetry is a misleading imitation of
reality based on opinion rather than knowledge. These issues will be
discussed in greater detail in the Epilogue. First, however, let us turn
from Elea to Acragas, and Empedocles’ dazzling response to Eleatic anxie-
ties over mortal language.
chapter 3
Empedocles
1 Introduction
Unlike Xenophanes and Parmenides, Empedocles’ merits as a poet have
been almost unanimously acknowledged in both antiquity and modernity.1
His dynamic cosmology, in which the myriad phenomena of the percep-
tible world are formed out of just four elements and governed by the
competing forces of Love and Strife, provided ample opportunity for lush
description and vivid narrative. Love brings the different elements together
until they are eventually blended to form a single cosmic Sphere; Strife
causes them to separate from one another until they are four pure cosmic
masses. The influence of Love and Strife alternates over a cosmic cycle of
many thousands of years. The familiar world we inhabit is at some
intermittent period between the time of Love’s Sphere and the absolute
separation of the elements under Strife. It seems to occur at two points over
the course of the cycle: under increasing Love and under increasing Strife.2
The cosmology is thus a divine drama in which the six permanent entities
(Love, Strife, and the four elements), referred to at times by the names of
gods, are the protagonists.3 In addition to this cosmology, certain frag-
ments also present a doctrine of metempyschosis: the souls of the wise
1
Praise of Empedocles’ poetry in antiquity: Arist. fr. 70 Rose=F73 Janko (ap. Diog. Laert. 8.57); Cic.
De or. 1.50.217; Lucr. 1.729–33; Plut. Quaest. Conv. 683E; in modernity: e.g. KRS 283; Most (1999) 356;
Kahn (2003) 158; Warren (2007) 135; Graham (2010) 328.
2
These basic details of the cosmic cycle are now relatively uncontroversial. For a more detailed
overview, see Primavesi (2016), building largely on the important work of O’Brien (1969) (developed
further in O’Brien [1995]). The orthodox view that there is a ‘double-cosmogony’ was disputed in the
sixties by Bollack (1965–9) i 97–124, Hölscher (1965) and Solmsen (1965) (for an overview of this
debate, see Long [1993]), but their ‘single-cosmogony’ view has since been rebutted by Graham (1988)
and Trépanier (2003). Sedley (2007) 31–62 proposes a complex (non-symmetrical) alternative which
does not seem to have found much support.
3
On the cosmology as a divine drama, with Empedocles as ‘l’Homère de la nature’ see Bollack (1965–
9) i 277–310.
102
Introduction 103
become ‘prophets, poets, singers and chiefs’ (D39=B146). The narrator
himself tells the story of how he, a daimon or ‘spirit’, has been exiled from
the gods (D10=B115) and reincarnated as ‘a boy, a girl, a bush, a bird,
a leaping fish from the sea’ (D13=B117). The doctrine of metempsychosis is
used as a basis for injunctions against animal sacrifice, the consumption of
meat (D28=B136) and even eating certain vegetables (D32=B140,
D31=B141) since they may contain the souls of our kin (D29=B137).
It will be clear even from this brief summary that Empedocles’ vision of
the world is an inherently more aesthetic one than that of Parmenides. He
grants the status of true reality to phenomena that mortals experience as
beautiful and is more accommodating to poetry (as traditionally con-
ceived) than Xenophanes or Parmenides had been, according a high status
to poets (D39=B146) and allowing for the validity of Muse-inspiration and
of certain mythical tales. Yet despite his value as a poet and his significance
for the history of philosophical poetics, Empedocles has been the subject of
relatively little scholarship that could be characterised as literary criticism.4
The vast majority of work has been devoted to textual criticism and to the
detailed exegesis of his philosophy, especially the question of how the
‘religious’ aspects of his thought might cohere with the cosmology,
a question that has become more pressing since the publication of the
Strasbourg papyrus.5 This important issue cannot be ignored, but it will
not form the main focus of the present chapter, which aims to enhance our
appreciation of Empedocles as a poet through an investigation of the
responses that the fragments are designed to elicit. The question of the
relationship between the two sides to Empedocles’ thought (and the related
one of whether the surviving fragments come from one or two poems) will
be touched upon further in the following section, but it shall be assumed
(along with most recent scholarship) that the ‘religious’ and ‘cosmological’
aspects form part of the same world view and that all surviving fragments
form part of a singular didactic project, even if they come from two poems.
4
On specific, literary-critical issues, see: van Groningen (1971); Rösler (1983); Sider (1984); Obbink
(1993); Wright (1998); Gagné (2006); Most (2007); Palumbo (2007); and Hardie (2013). Note also
the larger-scale studies of Rosenfeld-Löffler (2006) (marred by a failure to take sufficient account of
the cosmology) and Gheerbrant (2017). The technical linguistic studies of Traglia (1952); Gemelli
Marciano (1990); and Willi (2008) 193–263 are also valuable in this regard, as are the commentaries of
Bollack (1965–9) (only on the fragments attributed to the Phusika) and Wright (1995), though they
have been rendered somewhat outdated by the Strasbourg papyrus (of which Bollack [2003], on the
Katharmoi, fails to take sufficient account).
5
For a bibliography and concise overview of the history of Empedoclean scholarship, see Primavesi
(2013b). For an annotated bibliography on items published on Empedocles up to 1965, see O’Brien
(1969) 337–98; for an exhaustive, continuously updated bibliography, see https://sites.google.com/s
ite/empedoclesacragas/bibliography-a-z.
104 Empedocles
The validity of these assumptions will in part be tested by whether the
readings here offered are convincing. Such an approach is circular, but this
sort of circularity is inevitable in the interpretation of fragmentary texts.
In what follows, Section 2 addresses important first-order questions,
which should be preliminary to any understanding of the function of
Empedocles’ poetry, concerning the number of poems from which the
fragments derive and their performance context. Section 3 assesses the
position of poetry within Empedocles’ cosmology, in comparison with
Xenophanean and Parmenidean poetics. Section 4 then argues that the
poem is structured by two particular types of narrative and investigates
their significance. Section 5 explores various effects, characteristic of Greek
poetry, that these narratives help to generate.
2 Religion and Philosophy, One Poem or Two

As in the case of Parmenides, there has been disagreement over whether to
assimilate Empedocles to philosophy or to religion: according to the
former approach, he is a cosmologist who dresses his philosophical claims
in poetico-religious language in imitation of Parmenides; according to the
latter, he is primarily a religious figure, who composed poetry analogous to
the verses in the ‘Orphic’ gold tablets, designed for ritual and incantatory
purposes.6 This debate has been fuelled (again, similarly to the case of
Parmenides) by the fact that Empedocles presents two different types of
account, two different ways of describing the world, that are distinguish-
able by criteria that would be used in later distinctions between poetry and
philosophy. Thus, some fragments have traditionally been ascribed to the
Katharmoi, a religious poem describing the cycle of reincarnations of
a daimon, an exile from the gods who has committed bloodshed, whilst
others are thought to come from the Phusika or Peri Phuseōs, a poem
detailing the cosmology in which all things are composed of the four
elements and governed by the two forces of Love and Strife. These
‘physical’ aspects of Empedocles’ thought have been seen as more ‘philo-
sophical’, whilst the religious aspects have been felt to be vestiges of
6
For the former approach, see e.g. KRS 283; O’Brien (2001) (and, concisely, [2005]); Primavesi
(2008a). The latter approach is championed most vociferously by Kingsley (1995), (2002), (2003)
315–559; but note also Dodds (1951) 145–7 and Gemelli Marciano (2001) 205–9, (2005b). A nuanced
discussion of the issue is provided by Laks (2013), who ultimately regards Empedocles’ use of religious
language as a philosophical ‘reference’ rather than a representation of an actual ‘phenomenon’. For
studies of the religious background to Empedocles that do not take such an avowedly anti-
philosophical stance, see Zuntz (1971) 179–274; Riedweg (1995); Rodriguez (2005); Petrovic and
Petrovic (2016) 78–100.
Religion and Philosophy, One Poem or Two 105
primitive and distinctively non-philosophical religion. It was this distinc-
tion that led to Jaeger’s memorable description of Empedocles as
a ‘philosophical centaur’.7
Of course, as commentators have pointed out, there is no clear-cut
division between ‘philosophers’ and religious figures at this period; we
can read Empedocles productively and on good historical grounds within
both contexts: the fragments display the conspicuous influence of both
Parmenides and Pythagoreanism.8 Accordingly, scholarship in more recent
decades has tended to treat the two aspects as designed to be construed
together.9 This unifying approach is supported by the recurrence of certain
key terms across the different parts of Empedocles’ work,10 and has been
corroborated by two major developments in Empedoclean studies: first,
Catherine Rowett’s cogent, if ultimately conjectural, argument, published
in 1987, that the surviving fragments come from a single poem, referred to
variously by different titles in different sources;11 and second, the publica-
tion in 1999 of the Strasbourg papyrus, which showed beyond any reason-
able doubt that a single poem both outlined the cosmology and recounted,
in the first person, experiences of the daimon.12 Although the papyrus has
not ruled out the possibility that the fragments come from two poems –
one of which predominately contained the cosmology, the other, the
narrative of the daimon and the injunctions against meat-eating – it does
at the very least provide strong justification for interpreting the ‘daimonol-
ogy’ alongside the cosmology as part of a singular project.
It shall be assumed here, on the strength of arguments provided by
Rowett and, more recently, in light of the papyrus, by Simon Trépanier,
that the fragments do indeed come from a single poem and, moreover, on
the principle of parsimony, that the narrator remains the same across the
surviving fragments.13 My readings should, for the most part, be unaffected
by the two-poem hypothesis, so long as it is assumed that two such poems
7 8
Jaeger (1939) i.295; cf. Jaeger (1947) 128–54. A point made e.g. by Graham (2010) 327.
9
Thus, e.g. (prior to the publication of the Strasbourg papyrus) Kahn (1960); Barnes (1982) 391–6;
Wright (1995) 57–76.
10
Most conspicuously, the daimon’s crime of bloodshed is motivated by Neikos (D10=B115.14), ‘Strife’,
which, along with ‘Love’, Philia, is one of the two cosmic forces that govern the movements of the
four elements in the cosmology (e.g. D73>B17).
11
Published as Osborne (1987). Diogenes Laertius’ report, which stems from the notoriously unreliable
Hellenistic biographer Lobon, that ‘his work On Nature and the Katharmoi (Τὰ μὲν οὖν Περὶ φύσεως
αὐτῷ καὶ οἱ Καθαρμοὶ) stretches to five thousand verses’ (8.77), may be either a mistaken amalgam-
ation of the different titles used by different authors or a reference to different parts of the same poem
analogous to Hesiod’s Works and Days. See Trépanier (2004) 27.
12
Martin and Primavesi (1999).
13
Osborne (1987) (Rowett’s previous name); Trépanier (2004) 1–30.
106 Empedocles
warrant a unified reading. Even if we do take the fragments to come from
two separate poems, the very fact that they are attributed to the same
author should be sufficient justification for this approach.14 Narratorial
personae in early Greek poetry are typically consistent across different
works attributed to the same author, thus forming the continuous story
of a particular character.15 For Empedocles, similarities of language and
theme across the fragments should equally warrant such a unifying reading.
One significant entailment of the single-poem hypothesis is that the
same audiences would have received (notwithstanding textual variations)
the same text. This runs counter to the view, under the assumption of the
two-poem hypothesis, that the daimonological narrative is an ‘exoteric’
text, addressed to a collective audience and designed to illustrate symbolic-
ally the principles of the cosmology which is itself an ‘esoteric’ text,
addressed to Pausanias alone.16 Indeed, some use the distinction between
singular and plural addressee as a criterion for distinguishing the two
poems.17 This criterion, however, is not compelling. Such a variety of
singular and plural addressees is paralleled in the comparable didactic
poems of Hesiod, Lucretius and Virgil.18 One fragment even contains
a shift of addressee from the Muse to Pausanias (D44=B3), so the existence
of multiple addressees alone cannot be taken as secure evidence for mul-
tiple poems. More importantly, even if we do suppose two poems, it is
highly unlikely that, in any literal sense, one would have been ‘esoteric’ and
the other ‘exoteric’. The fragments with a singular addressee formed part of
a poem that became known to Plato and Aristotle and, eventually, was
translated into Latin.19 It is improbable that this transmission would have
occurred had the poem been designed only for the ears of an esoteric-
religious sect. A much more likely context for the circulation of the parts of
the poem thought to be ‘esoteric’ is public, rhapsodic performance, as is
14
For a defence of reading the literary oeuvre of a particular author as a meaningfully unified whole, see
A. Fowler (1982) 128–9.
15
Note such examples as Archilochus, Hipponax and Solon. See Morrison (2007) 45, 61–7.
16
Thus, Bollack (2001), (2005) (who, with problematic circularity, sees the occurrence of B139 in the
Strasbourg papyrus as a quotation from the religious poem in the physical one – but we have no
grounds for supposing that the fragment occurred anywhere else in Empedocles’ work than in the
part of the Phusika preserved in the papyrus); Kingsley (1995) 363–6, (2002) 344–350; Primavesi
(2008a).
17
This criterion was influentially applied by Stein (1852) (whose division of the fragments is followed,
more or less, by DK and Wright [1995]), and has more recently been advocated by Cerri (2001) 182–3
and Primavesi (2007) 216–19.
18
On Empedocles’ addressees, see further Obbink (1993).
19
Assuming the Empedoclea of Sallustius, mentioned by Cic. QFr. 2.9.3=R36, is a translation. On the
Latin reception of Empedocles, see Hardie (1995); Sedley (1998) 1–34; Nelis (2004); Garani (2007);
and articles by Farrell, O’Rourke, Fabre-Serris and Nelis in Dictynna 11 (2014).
Empedoclean Poetics 107
attested for the Katharmoi.20 On the other hand, if the narrative of the
daimon were intended as an ‘exoteric’ text, then it would be a very strange
one indeed, since it advances beliefs and practices that were associated with
Orphism and Pythagoreanism, religious movements that, in Classical
sources, are always depicted as peripheral.21 Even if we do suppose two
poems, then, they are likely to have had similar performance contexts. The
esoteric feel of some of the fragments and their distinction of a singular
chosen addressee are better understood as rhetorical features. After all, the
poem was not composed for the ears of Pausanias alone, if indeed he was
ever a real person. The peculiarly appropriate etymology of Pausanias’
name, ‘one who stops pain’, for the student of this text (cf.
D4=B112.10–11) may betoken his fictional nature.22 As in the case of
Parmenides, the distinction of a singular addressee challenges any potential
audience member to step into Pausanias’ role.23
3 Empedoclean Poetics
Whilst immense efforts have been deployed towards a reconstruction of the
details of Empedocles’ cosmology, the role of poetry within this system has
received little attention.24 The present section argues that Empedocles’
philosophy, in addition to responding to physical claims made by earlier
authors, also repositions the role of poets and poetry. Against the criticisms
of poets by Xenophanes and Heraclitus and against the negative implica-
tions for poetry of Parmenides’ arguments, Empedocles’ cosmology pro-
vides a systematic defence of the source of poetry. This section deals with
20
Dicaearchus fr. 87 Wehrli (ap. Athen. 14, 620d=A12=D3) and Favorinus (ap. D.L. 8.63=Empedocles
A1) report that the Katharmoi was performed at Olympia by the rhapsode Cleomenes. Obbink
(1993) 77–8 accepts the validity of the story and dates Cleomenes, on the basis of the dating of the
other rhapsodes Athenaeus mentions, to the late fifth or early fourth centuries bce, citing West
(1981) 125.
21
On this see, concisely, Burkert (1982) and Parker (1995). On the marginality or marginalisation of
Pythagoreans and Orphics, cf. Xenophanes D64=B7 (lampooning Pythagoras); Eur. Hipp. 952–5
with Barrett (1964) ad loc.; Pl. Resp. 364b–365a; Thphr. Char. 16.12. Edmonds (2013) argues at length
that the unifying feature of references to Orphics in ancient sources is precisely that they are out of
the ordinary.
22
The etymology is pointed out by Clay (2015) 132, who has also pointed out a similarly appropriate
etymology of Hesiod’s Perses as deriving from πέρθω, ‘to lay waste’, ‘to plunder’ (Clay [2003] 38).
23
Contrast the way in which Hesiod’s Perses and Lucretius’ Memmius are sometimes seen as foolish
foils against whom we are to distinguish ourselves (see Clay [1993], [2003] 33–6 on the former; Mitsis
[1993] on the latter). Obbink (1993) 52–3, 80–9 argues that Empedocles’ Pausanias similarly serves as
a foil, a view I shall reject in Section 4b below.
24
Although see Gagné (2006) and, now, Gheerbrant (2017), esp. 633–705. Hardie (2013) also touches
on this matter.
108 Empedocles
the status and source of poetry within Empedocles’ cosmos. Later on,
Section 5 will deal with its ideal effects.
a Empedocles and His Predecessors on Poetry

Xenophanes criticised Homer and Hesiod’s depictions of anthropomorphic,
morally fallible gods and referred to certain mythical creatures – the Titans
and the centaurs – as ‘fabrications’ (plasmata). His conception of divine
communication ruled out the validity of Muse-inspiration. Sometime later,
Heraclitus denied that Hesiod and Xenophanes (along with Pythagoras and
Hecataeus) had ‘understanding’ (noos, D20=B40; cf. D25a=B57; D26=B129)
and stated that Homer and Archilochus ‘deserve to be thrown out of the
contests and beaten with a stick’ (D21=B42; cf. D22=B56; D24=B105). By the
time of Empedocles’ composition, then, there were attacks against (1)
the status of poets (they deserve to be thrown out of the contests), (2) the
content of their poetry (the fabricated, morally corrupting stories) and (3) their
claims to divine inspiration. Empedocles denies all three of these criticisms. (1)
He elevates the status of hymnopoloi – ‘singers’ or ‘minstrels’ –25 along with
chieftains (promoi) and prophets (manteis) as incarnations of the souls of the
wise (D39=B146, with Clement’s context for the fragment).26 (2) His cosmol-
ogy allows for the historical existence of fantastical, mythical beasts. During
the onset of Love, individual limbs are initially formed and come together into
various combinations, but only the combinations that we would recognise as
animals in the present stage are capable of survival.27 Thus, at an earlier point
in the cosmic cycle, there did indeed exist Minotaur-like ‘ox-faced men’ (as
well as ‘man-faced oxen’, D156=B61). Such a process could also easily have
accounted for other composite mythical creatures, such as the Centaurs or the
Chimaera. (3) Empedocles assumes the validity of the traditional forms of
25
The term is, for instance, used of Anacreon at [Simon.] 67 FGE. It might be thought that it
specifically denotes a singer of ‘hymns’, but humnos does not yet seem to have the specific meaning of
a ‘song to a god’ in Archaic Greek poetry (it is used, for instance, of the song of Demodocus at Od.
8.429, and Pindar uses it repeatedly of his own work, e.g. Ol. 1.105, 2.1, 3.3 etc.). The term seems
sufficiently broad at this stage to encompass more or less all poetry. See Maslov (2015) 286–94 for
a recent discussion, although his conclusion, that the term was originally used specifically for choral
singing, seems to me to go beyond the implications of the surviving evidence.
26
Picot and Berg (2015) argue suggestively that the figures of D39=B146 have achieved a high status
through engagement with strife, mainly on the grounds that πρόμοι primarily means ‘battle-chief’
(an abridgement of πρόμαχοι) in Homer, but as Gheerbrant (2017) 701–2 argues, by the time of
Greek tragedy, the word can simply mean ‘chief of the land’ (e.g. Soph. OC 884) and here,
Empedocles is reinterpreting traditionally prestigious societal roles in terms of his cosmology.
27
Cf. D154=B57, D149=B59, with Simplicius’ context for the two fragments. For the most detailed
reconstruction of this process, see O’Brien (1969) 196–236.
divine-to-mortal communication, prophecy and Muse-inspiration. Prophets
have achieved a high grade of incarnation (D39=B146), and countless mortals
come to the narrator in search of ‘prophecies’ (D4=B112.9). In the surviving
fragments, the narrator addresses the Muse twice, and refers to her a third time
(D44=B3; D47=B4; D7=B131).
It is quite possible that Empedocles was responding directly to these earlier
critical comments. Chronology permits that he was familiar with Heraclitus’
work: he was active no earlier than the second quarter of the fifth century;28
Heraclitus’ book became sufficiently famous to have raised a group of
followers, Heracliteans, whom Plato depicts as known to Socrates and his
interlocutors in the second half of the century (R16–26). This possibility is
further supported if, as some have argued, Parmenides D7=B6 alludes to
Heraclitus D49=B51,29 since Empedocles certainly knew Parmenides’ poem.
In any case, the connection with Parmenides makes it more than likely that
Empedocles was aware of Xenophanes’ criticisms of certain poets. The ancient
testimonies that Parmenides was a student of Xenophanes and Empedocles an
associate of Parmenides may be questionable,30 but an intellectual genealogy
linking the three authors is supported by a common geographical association
with Southern Italy and confirmed by internal evidence. Among Empedocles’
copious allusions to Parmenides are the descriptions of the Sphere that occurs
during the period of Love’s supremacy (D90=B28; D92=B29), which indicate
that it is modelled on the Sphere-like entity outlined in Parmenides’ Aletheia
(D8.47=B8.42).31 Parmenides’ Sphere-like entity is itself modelled on
Xenophanes’ singular unmoving god.32 This intellectual lineage supports
a reading of Empedocles’ accommodation of poetic inspiration and authority
against Xenophanes’ criticisms: the contrast may well have been intended by
Empedocles and would have been recognisable to some ancient audiences.
28
Empedocles was born later than Anaxagoras (Arist. Metaph. 1.3.984a11=A6; Theophr. ap. Simpl. In
Phys. 25.19=A7) but was apparently Gorgias’ teacher (D.L. 8.58<A1), and so probably older than him,
details which suggest a birth date between around 500 and 485 bce. Diogenes Laertius places his
floruit in 444–441 bce (8.74=P1<A1), which seems a decade or two too late. See further KRS 280–1
and O’Brien (2005) 318–21.
29
There is a long-standing debate on this issue, which goes back to Bernays (1850). For a cogent and
relatively recent defence of the identification of such an allusion, see Graham (2002).
30
Theophrastus reported that Parmenides was Xenophanes’ student (ap. D.L. 9.21=Parmenides
P6<A1), a notion that probably ultimately derives from Plato Sophist 242D; and that Empedocles
was Parmenides’ ‘emulator and associate’ (ap. Diog. Laert. 8.55=P13<A1). Note also Hermippus’
report that Empedocles spent time with and imitated not Parmenides but Xenophanes (fr. 26
Wehrli ap. Diog. Laert. 8.56=P15<A1).
31
Cf. especially Empedocles D90.1=B28.1 ἀλλ’ ὅ γε πάντοθεν ἶσος <ἑοῖ> καὶ πάμπαν ἀπείρων with
Parmenides D8.54=B8.49 οἷ γὰρ πάντοθεν ἶσον, ὁμῶς ἐν πείρασι κύρει.
32
Cf. Parmenides D8.9, 43=B8.4, 38 with Xenophanes D17=B24, D19=B26.
110 Empedocles
Most conspicuous, however, is the relationship with Parmenides. As
has long been acknowledged, Empedocles alludes copiously to
Parmenides throughout the fragments.33 What I propose here is that
this engagement has implications for poetics. It was argued in Chapter
2 that Parmenides, in the introduction to the Doxa, presents the world of
mortal appearance as one of poetic artifice, enlisting the fictional poten-
tial of poetry to illustrate the deceptive nature of the world of our mortal
experience. Empedocles alludes precisely to that part of Parmenides’
poem, rewriting it in a more optimistic manner. In the lengthy outline
of the cosmic cycle that was preserved by Simplicius and then augmented
by the Strasbourg papyrus, the narrator bids his addressee to ‘hear the
undeceptive expedition of my account’, an expression that reverses the
‘deceptive kosmos of words’ of Parmenides’ goddess (D8.57=B8.52). To
clarify the poetical significance of this allusion to Parmenides, the line is
worth quoting in context (D73.245–66=B17.14–35) and will require
a detailed analysis:
ἀλλ’ ἄγε μύθων κλῦθι· μάθη γάρ τοι φρένας αὔξει·
ὡς γὰρ καὶ πρὶν ἔειπα πιφαύσκων πείρατα μύθων, (D73.246=B17.15)
δίπλ’ ἐρέω· τοτὲ μὲν γὰρ ἓν ηὐξήθη μόνον εἶναι
ἐκ πλεόνων, τοτὲ δ’ αὖ διέφυ πλέον’ ἐξ ἑνὸς εἶναι,
πῦρ καὶ ὕδωρ καὶ γαῖα καὶ ἠέρος ἄπλετον ὕψος,
Νεῖκός τ’ οὐλόμενον δίχα τῶν, ἀτάλαντον ἁπάντηι,
καὶ Φιλότης ἐν τοῖσιν, ἴση μῆκός τε πλάτος τε (D73.251=B17.20)
τὴν σὺ νόωι δέρκευ, μηδ’ ὄμμασιν ἧσο τεθηπώς·
ἥτις καὶ θνητοῖσι νομίζεται ἔμφυτος ἄρθροις,
τῆι τε φίλα φρονέουσι καὶ ἄρθμια ἔργα τελοῦσι,
Γηθοσύνην καλέοντες ἐπώνυμον ἠδ’ Ἀφροδίτην·
τὴν οὔ τις μετὰ τοῖσιν ἑλισσομένην δεδάηκε (D73.251=B17.20)
θνητὸς ἀνήρ· σὺ δ’ ἄκουε λόγου στόλον οὐκ ἀπατηλόν.
ταῦτα γὰρ ἶσά τε πάντα καὶ ἥλικα γένναν ἔασι,
τιμῆς δ’ ἄλλης ἄλλο μέδει, πάρα δ’ ἦθος ἑκάστωι,
ἐν δὲ μέρει κρατέουσι περιπλομένοιο χρόνοιο.
καὶ πρὸς τοῖς οὔτ’ ἄρ τι34 ἐπιγίνεται οὐδ’ ἀπολήγει (D73.261=B17.30)
εἴτε γὰρ ἐφθείροντο διαμπερές, οὐκέτ’ ἂν ἦσαν·
τοῦτο δ’ ἐπαυξήσειε τὸ πᾶν τί κε; καὶ πόθεν ἐλθόν;
33
The instances are too copious to list here, but the commentary of Wright (1995) notes them passim.
34
Wright (1995) obelises here on metrical grounds, but synecphonesis (running together of two
adjacent vowels to form one syllable) of this kind (τι ἐπιγίνεται) is paralleled at D76.5=B139.1 (ὅτι
οὐ). See further Gallavotti (1975) 187, 348.
πῆι δέ κε κἠξαπόλοιτο, ἐπεὶ τῶνδ’ οὐδὲν ἔρημον;
ἀλλ’ αὐτ(ὰ) ἔστιν ταῦτα, δι’ ἀλλήλων δὲ θέοντα
γίγνεται ἄλλοτε ἄλλα καὶ ἠνεκὲς αἰὲν ὁμοῖα. (D73.266=B17.35)
But come, listen to my words; for learning will increase your wits;
For just as I said before, showing the limits of my words, (D73.246=B17.15)
I will speak twofold; for at a time one thing grew to be alone
From many, and at another, it grew apart to be many things from one,
Fire and water and earth and the boundless height of air,
And destructive Strife apart from them, balanced in all ways,
And Love among them, equal in length and breadth. (D73.251=B17.20)
See her with your mind and do not sit stunned by your eyes;
She who is also considered to be inborn in mortal limbs,
By whom they think thoughts of love and accomplish fitting deeds,
Calling her Joy by name and Aphrodite;
Her whom no one has known among them (D73.256=B17.25)
No mortal man. But you, listen to the undeceptive expedition of my account.
For these are all equal and the same age in respect of their birth,
But each presides over a different honour, and each one has its character,
And they rule in turn within the cycle of time.
And in addition to these, not a thing is created nor ceases to be; (D73.261=B17.30)
For if they perished forever, they would no longer be.
For what might increase this whole? And coming from where?
And how could it be destroyed, since nothing is devoid of these things?
But these things are themselves, and passing through one another
Different things come to be at different times and continuously always similar.
(D73.266=B17.35)
As commentators have noted, the allusion can be read as a ‘correction’ of

Parmenides’ poem.35 The sort of pluralistic cosmology that was deceptive for
Parmenides’ goddess is deceptive no longer. The allusion flags a difference
between the two epistemologies: for Parmenides, sensory experience led to
a world of deception, but Empedocles vindicates the use of the senses as
a means of accessing the true nature of the world (D44=B3.9–13).36 Palmer
denies that the line is a challenge to Parmenides on the grounds that the
significance of ‘not deceptive’ is restricted to Love’s cosmological activity,37 but
such a restriction is unnecessary: that the ‘not deceptive expedition’ should
refer, in the first instance, to Love’s activity does not prevent it from marking
35
Thus e.g. Wright (1995) 170; Trépanier (2004) 49; O’Brien (2005) 323–4. Contrast Nünlist (2005)
74–5, who sees the polemic against humans in general, rather than Parmenides (but the evocation
surely suggests that Parmenides is one of those humans).
36
This particular contrast is marked by allusions to Parmenides’ rejection of the senses as a means of
accessing the truth at D8.1–6=B7. See n. 224 below.
37
Palmer (2009) 273–6.
112 Empedocles
a more general relationship between Empedocles’ cosmology and that of the
Doxa.
Furthermore, this reading is supported by the manner in which
Empedocles combines elements from both the Doxa and the Aletheia. The
commands in line 21 (‘See her with your mind and do not sit stunned by your
eyes’) recall the instruction of Parmenides’ goddess to ‘see (λεῦσσε) things
which though absent are present to the mind (νόωι)’ (D10=B4.1), and her
description of ignorant mortals as ‘stunned’ (τεθηπότες, D7=B6.7). The
recollection illustrates how Empedocles’ cosmological Love is, like the entity
in the Aletheia, accessible to the mind though not evident to the senses, at least
as usually deployed by mortals.38 The argument against the perishability of the
four elements (D73.262; 264=B17.31, 33) and against the possibility of anything
being created in addition to them (D73.263=B17.32) is especially reminiscent
of the goddess’ contentions in Parmenides D8>B8 against the possibility that
what-is was created or is perishable (D8.8, 11–17, 24–6=B8.3, 6–12, 19–21), or
that anything could be created in addition to it (D8.18=B8.13). The phrasing is
similar: like the goddess, Empedocles makes his argument using an unreal
conditional clause (cf. D73.262=B17.31 with Parmenides D8.38=B8.33) and
impassioned rhetorical questions (cf. D73.263–4=B17.32–3 with Parmenides
D8.11, 12, 14–15, 24=B8.6, 7, 9–10, 19). Given these evocations of Parmenides –
and specifically the Aletheia – the seemingly tautological expression, ‘these
things are themselves’ (αὐτ[ὰ] ἔστιν ταῦτα, D73.265=B17.34) appears as
a response to Parmenides’ ἔστιν: Empedocles specifies what Parmenides had
left notoriously unspecified (the subject of ἔστιν), intimating that the four
elements have the ‘trustworthiness’ that Parmenides ascribed to what-is.39
From one perspective, they ‘always are, unmoving in the cycle’ (αἰὲν ἔασιν
ἀκίνητοι κατὰ κύκλον, D73.244=B17.13), just as what-is is unmoving (D8.31,
42=B8.26, 37). There has been some debate over whether or not Empedocles’
elements are meant as a ‘plurality of Parmenidean Beings’,40 but this dichot-
omy fails to appreciate the subtlety of the engagement: the elements are not
simply Parmenidean Beings, but, in certain respects, assume the role that
‘Being’ or ‘what-is’ occupied within Parmenides’ philosophy, in being the
uncreated and indestructible objects of secure knowledge. Accordingly, the
account also contains conspicuous resemblances to the Doxa, in violation of
38
On this connection, see further Palmer (2009) 275.
39
πίστις is specified as a quality of both Parmenides’ Aletheia (D4=B1.30) and Empedocles’ account at
D61=B71.1. Note also the similar epistemological vocabulary, reminiscent of Parmenides, used in
D44=B3.9–13 and D47=B4.
40
Palmer (2009) 279–312 vociferously denies that the elements are Parmenidean Beings, against the
orthodox view represented by e.g. Guthrie (1965) 146; Curd (2005) 155–64.
the strictures in the Aletheia. The elements both ‘are’ (D73.265=B17.34) and
‘come to be’ (D73.266=B17.35), since they are ever-present in all compounds,
but only ‘come to be’ manifest as pure stuffs at occasional points in the cycle.
‘Coming to be’ is a process that, in Parmenides’ poem, was confined to the
deceptive world of the Doxa. The phrasing here, with γίγνεται positioned
emphatically at the start of the line, seems designed to underline the co-
existence of these two processes against the Parmenidean division. The narra-
tor has just mentioned the creation (γένεσις) and destruction (ἀπόλειψις) of
mortal things (D73.235=B17.3) and will proceed to describe how ‘we come
together into one kosmos’ (D73.267) as well as, in detail, the perceptible world
that had been explained in the Doxa (D73.279–308, D74 from the same
papyrus).
In sum, Empedocles’ allusions here illustrate his defence of the
pluralistic world of mortal experience against Parmenides’ misgivings.
He thus provides grounds for a defence of the validity of the dynamic,
colourful world of Greek poetry against the Parmenidean claims that
only singular ‘what-is’ lives up to the standards of true reality, claims
which entailed that the world of traditional poetry was a deception.
Empedocles here also invites a reading against the wider tradition of
‘poetic pessimism’ in which Parmenides’ introduction to the Doxa
participated. The mortal Homeric and Hesiodic narrators bore an
uncertain relationship with their Muses: in the Iliad, we only hear
kleos, and lack the perfect knowledge of the Muses (Il. 2.484–93); in
the Theogony, it is left unclear whether the Muses inspired Hesiod with
‘true things’ or ‘lies like true things’ (Theog. 27–8).41 Xenophanes drew
on this passage to highlight the provisional nature of his cosmology, and
Parmenides’ goddess did so when distinguishing the deceptive and
trustworthy parts of her speech. Empedocles presents a further step
within this tradition: he adapts the line of Parmenides’ poem which
itself had been a window-allusion to both Xenophanes D50=B35 and
Hesiod’s Theogony.
The connection with this wider tradition is corroborated by further
thematic parallels: the Empedoclean narrator is, like the Hesiodic and
Homeric Muses, and like Parmenides’ goddess, a deity (D4=B112.4;42
D60=B23.11) who has knowledge of the truth (D6=B114) in contrast with
41
On the Muses in Hesiod’s Theogony, see the scholarship cited at Introduction n. 59 and Chapter 1
n. 43 above.
42
This problematic line is discussed in greater detail in Section 4a below.
114 Empedocles
the general ignorance of mortals (D42=B2).43 The insult to mortals in
D42=B2 recalls Hesiod’s Muses: the narrator lambasts the ignorance of
mortals – they have ‘narrow resources’ (στεινωποί . . . παλάμαι, D42=B2.1)
and ‘many vile things . . . blunt their wits’ (πολλὰ δὲ δείλα . . . ἀμβλύνουσι
μέριμνας, D42=B2.2) – before proceeding to single out the addressee, presum-
ably Pausanias: ‘you, since you have come here, shall learn’ (σὺ (δ’) οὖν, ἐπεὶ
ὧδ’ ἐλιάσθης / πεύσεαι, D42=B2.8–9). This discursive pattern of an insult to
the masses,44 combined with a singling out of a particular favoured student,
also occurs in the Theogony, where the Muses insult the farmers in general
(κάκ᾽ ἐλέγχεα, γαστέρες οἶον, ‘wicked shameful things, mere bellies’, 26)
before they privilege Hesiod by giving him the laurel staff and inspiring him
with wondrous song (30–4).
The thematic parallels (divine knowledge; mortal ignorance; mediation via
poetry) and the more specific parallels with Hesiod in D42=B2 support our
reading of the narrator’s assertion of the undeceptiveness of his words not only
as an allusion to Parmenides, but as a participation in this long-standing
discussion concerning the poet’s access to divine truths. Indeed, we are
reminded of the contrast between divine narrator and mortal addressee at
D73.256–7=B17.25–6, with the statement that no ‘mortal man’ (an expression
emphasised by line position) has known of Love among them and it is because
of this ignorance that the narrator must reassure his mortal addressee of the
undeceptiveness of his account. Comparison with this tradition reveals that
the situation in Empedocles is much less problematic than the earlier
examples: Homer, Hesiod, Xenophanes and Parmenides all had emphatically
mortal primary narrators, and contrasted divine knowledge with the mortal
perspective. Of these forebears Parmenides presents arguably the most opti-
mistic picture, since the youth is assured by the goddess that she will teach him
true reality and she clearly distinguishes that content from the deceptive part
of his speech, but the youth is nonetheless mortal.45 Empedocles presents
a still more positive scenario: the primary narrator himself appears as a god
(D4=B112.4) who assures us of the validity (D6=B114) and undeceptiveness of
his claims. Of course, as we shall see below, the divine status of Empedocles’
narrator is a difficult crux, but the key point is that, unlike Parmenides’ poem,
43
Although he also invokes the Muse, it is notable that he does so not to make use of any superhuman
knowledge she may possess, but to ensure that his speech is pious, a ‘good account of the gods’
(D7=B131.4) and no more than what is ‘right for ephemeral beings to hear’ (D44=B3.4).
44
Arguably a Presocratic topos, given the similar scorn expressed by Heraclitus (D1=B1) and
Parmenides’ goddess (D7=B6).
45
Notwithstanding the possibility that he achieved a homoiōsis theōi through learning the goddess’
lessons. See Chapter 2 Section 5b above.
the divine authority does not communicate to the audience via a mortal
intermediary.46
It might be objected that Empedocles’ narrator still invokes the Muse.
However, it is notable that he does not obviously do so to learn new
information, but to ensure that his speech is pious, a ‘good account of
the gods’ (D7=B131.4) and something that is ‘right for ephemeral beings to
hear’ (D44=B3.4).47 When Empedocles refers to the ‘assurances of my
Muse’ (D47=B4.2 ἡμετέρης . . . πιστώματα Μούσης) it is likely, given the
parallels for the expression, that he is referring metonymically to his song,
and not to any external authority.48 The possessive pronoun is suggestive of
the poet’s mastery. Other hexameter poets invoke the Muse with an
imperative verb of singing or speaking, implying that the Muse herself
knows the content to be performed, but Empedocles’ Muse is only an
accomplice who is asked to ‘stand by’ him (D7=B131.3).49 Instead, the
knowledge of Empedocles’ narrator derives from his recollection of differ-
ent reincarnations (D6=B114; D13=B117) and coincides with his divine
status.50 He demands that his student ‘know these things having heard
them from a god’ (θεοῦ πάρα μῦθον ἀκούσας, D60=B23.11). Tor has
argued that this ‘god’ is the Muse,51 but the fact that the line occurs at
the end of a lengthy fragment in which no Muse is mentioned, as well as
the male gender of θεοῦ, make it more probable that the god is the narrator
himself.
If it is true that ‘One function of the poet at any time is to discover by his
own thought and feeling what seems to him to be the poetry at that time’
and that ‘ordinarily he will disclose what he finds in his own poetry by way
of the poetry itself’,52 Empedocles’ ‘discoveries’ concerning the nature of
46
Though the narrator himself may be both mortal and divine, from different perspectives, as Tor
(2017) 333–5 argues.
47
Pace Tor (2017) 333–5. Palmer (2013) argues that the narrator for most of the Physica (including the
details of the cosmology) is in fact the Muse, with Empedocles as mortal addressee; fatal for his
thesis, however, is that the narrator of the section of the poem preserved in the Strasbourg papyrus
(i.e. the most detailed account of the cosmic cycle) laments his crimes of bloodshed (D76), and so is
almost certainly to be identified with the exile from the gods who narrates D10=B115.
48
Cf. Pl. Symp. 189b; Callim. frr. 75.77, 112.1 with Harder (2012) ad loc.; Theocr. 16.107 plays upon this
sense, on which see Hunter (1996) 92–3. For Empedocles’ Muse as metonymic of his song, see
Hardie (2013) 218–20.
49
Contrast Homer Il. 1.1, 2.484; Od.1.1; hMerc. 1.; hHom. 9.1, 14.2, 17.1, 19.1, 20.1, 31.1, 32.1, 33.1; Hes.
Theog. 105; Op. 2. On the form of early Greek Muse-invocations see Morrison (2007) 73–89. On
Empedocles’ Muses more generally, see Hardie (2013) and Gheerbrant (2017) 99–213.
50
The fact that the truth is painful for mortals (D6=B114) may suggest that knowledge of the truth
comes through experiencing the cycle of reincarnations. See also n. 47 above.
51
Tor (2017) 335 n. 51. See also n. 47 above.
52
Stevens (1951) vii. Quoted in the Introduction Section 3 above.
116 Empedocles
poetry are disclosed by his relationship with earlier poetry. Along with the
defence against Parmenides of the trustworthiness of the sensible world,
Empedocles’ narrator enacts a more positive relationship between the
singer and the truth. This contrast is marked by the allusions to
Parmenides and by the overall enunciative scenario involving
a knowledgeable divine narrator who addresses mortals. Of course, there
is no reason to take the narrator himself as typifying the status of singers or
poets within Empedocles’ cosmos: there is every reason to think that he is
an extraordinary case. But his more positive status coheres with the claim
that poets are reincarnated souls of the wise (D39=B146) and, unlike
Homer or Hesiod, presupposes that it is at least possible for a singer to
have an unmediated access to the truth. Furthermore, the defence of the
perceptible world permits a greater degree of veracity to the stories of the
poets than Xenophanes or Parmenides had allowed: Xenophanes had
stressed the fallaciousness of Homeric and Hesiodic depictions of the
gods, and Parmenides’ ontology dictated that no previous Greek narrative
accurately reflected true reality. Within the confines of Empedocles’ cos-
mology, the mythical stories of the earlier Greek poets could still be true.
Even certain fantastical stories – in particular, that of the Minotaur, but
also other composite monsters – are supported by the cosmology.
b Love and Poetry in Empedocles’ Cosmology

The cosmology provides a systematic basis for the defence of poetry in
a further respect. Empedocles has often been lauded for developing
a system that provides a basis for ethics, with Love responsible for good
and Strife for evil.53 But it has less often been appreciated that these ethical
functions extend to the realm of aesthetics. As we shall see, Love is
responsible for poetry that is both beautiful and socially beneficial, with
Strife the source of the converse. To the extent that Empedocles provides
a systematic ethics, he also provides a systematic aesthetics.
Beauty arises in Empedocles’ cosmology in D21=B122:
ἔνθ’ ἦσαν Χθονίη τε καὶ Ἡλιόπη ταναῶπις,
Δῆρίς θ’ αἱματόεσσα καὶ Ἁρμονίη θεμερῶπις,
Καλλιστώ τ’ Αἰσχρή τε, Θόωσά τε Δηναίη τε,
Νημερτής τ’ ἐρόεσσα μελάγκαρπός τ’ Ἀσάφεια.
53
E.g. Barnes (1982) 122; O’Brien (2005) 341–2. This tradition goes back to Aristotle (Metaph.
1.4.985a4–7).
The Earthly was there, and the far-sighted Sun-faced,
And bloody combat and calm-looking Harmony,
And Beauty and Ugliness and Swift and Slow,54
And Lovely infallibility and dark-fruited Obscurity.
The list is modelled on the Homeric and Hesiodic catalogues of Nereids and
probably occurred in Empedocles’ ‘theogony’, presumably an account of the
origin of certain superhuman ‘gods’.55 Plutarch tells us that this is a list of
good and evil ‘fates and spirits’ that accompany men throughout life (De
tranq. anim. 474b). He identifies the second pair as Love (Philia) and Strife
(Neikos, De Is. et Os. 370d), and with good reason, since Ἁρμονίη is
repeatedly associated with Love and the force of attraction throughout the
fragments (D60=B23.4; D89=B27.2; D192=B96.4).56 But it would be strange
for Love and Strife to arise in the course of a theogony when they are
supposed to be permanent entities. Possibly, then, these are two deities or
spirits that are strongly associated with, or predominately formed of, Love
and Strife, in the same way that the Land and the Sea are predominately
formed of the elements Earth and Water.57 Alternatively, these may be
names for different aspects of the two entities, just as ‘Joy’ and ‘Aphrodite’
are different names for Love at D73.255=B17.24. This possibility would be
contiguous with traditional Greek religious practice, whereby different
epithets, when attached to a particular divine name, can refer to different
forms of the deity, which are sometimes depicted in wildly different ways.58
In any case, the pair of D21=B122.2 are clearly strongly associated with
the two cosmic forces and the same could be said of the other opposing
pairs in this fragment. Infallibility is ‘lovely’ (ἐρόεσσα), thereby associated
with cosmic Love; we can plausibly assume that ‘dark-fruited obscurity’ is
conversely associated with Strife. The positive and negative associations of
the two terms cohere with the respectively positive and negative depictions
of Love and Strife throughout the fragments.59 On the same grounds, Love
54
Δηναίη should properly mean ‘long-lasting’, but, as Wright (1995) 281 argues, probably here means
something like ‘slow old age’ (cf. its usage as ‘aged’ at Aesch. PV 794).
55
Porphyry Abst. 2.21. 56 See further Sider (1984).
57
At D122=B38, the Sun appears to be called a Titan, that is, another quasi-divine character formed of
one particular element. Cf. D126=A56. See further Long (2017) 6–7 on these bodies in Empedocles.
58
E.g. Zeus Meilichios was sometimes depicted as a giant snake, and it is not clear whether ‘Zeus
Chthonios’ referred to Zeus in his capacity as lord of the underworld, or the underworld equivalent
to Zeus (i.e. Hades). On this issue, see Parker (2011) 67–70 and Versnel (2011) 60–87. The question
of whether such various labels referred to the same god may not usually have arisen in popular
religion, but it is directly addressed in some literary and philosophical texts (Hdt. 1.44; Xen. Symp.
8.9; Cic. Nat. D. 3.21–4).
59
Love is ‘Joy’ D73.255=B17.24 and worshipped in the ideal community of D25=B128; Strife is
‘grievous’ (D207=B109.3); ‘mad’ (D10=B115.14); ‘destructive’, (D73.250=B17.19); ‘wrath’
118 Empedocles
may be associated with ‘Swiftness’ and Beauty; Strife, with ugliness and
Slowness. Love’s association with Swiftness is supported by the ‘swift
thoughts’ (φροντίσι . . . θοῆισιν) of the ‘holy mind’ of D93=B134, since
the mind (given the similarities with D92=B29) seems to be another
product of Love. The ‘Earthly’ and the ‘far-sighted Sun-faced’ less obvi-
ously fit with this scheme, but this pairing has normative value if we regard
it as an opposition between the Chthonic and the Heavenly. χθόνιος is
most often used in reference to deities of the netherworld (see LSJ s.v.),
whilst the Sun is said by Empedocles to ‘shine back towards Olympus with
a fearless face’ (D123=B44). Recently, Simon Trépanier has argued that the
gods who stand at the highest point in the cycle of reincarnations
(D39=B146; D40=B147) are in fact the stars, with the Sun acting as the
‘hearth’ around which they are gathered (D40=B147.1; cf. Philolaus
D15=B7).60 If, then, we take Love to be the cause (whether material or
efficient, in Aristotelian terms) of the positive spirits in this fragment, then
beauty is given cosmological explanation as resulting from Love.
That this is beauty as applicable to poetry is clear from the reference to
the Muse as Kalliopeia (D7=B131):
εἰ γὰρ ἐφημερίων ἕνεκέν τινος, ἄμβροτε Μοῦσα,
ἡμετέρας μελέτας <ἅδε τοι> διὰ φροντίδος ἐλθεῖν,
εὐχομένωι νῦν αὖτε παρίστασο, Καλλιόπεια,
ἀμφὶ θεῶν μακάρων ἀγαθὸν λόγον ἐμφαίνοντι.
If, for the sake of one of the ephemerals, immortal Muse,
It pleased you that my concerns should come through your thoughts,
Now, again, stand by me as I pray, Kalliopeia,
Revealing a good account about the blessed gods.
‘One of the ephemerals’ may be Pausanias, or it may be the narrator
himself, perhaps in some earlier, mortal guise, if the first two lines refer
to previous assistance, as we would expect in a kletic prayer.61 In either case,
the scenario here presented contrasts with other hexameter invocations to
the Muse, in that she is presented not as the source of the message or as
(D77A=B21.7 cf. D101=B22.9). As Kamtekar (2009) 225 points out, Strife’s one ‘blameless’ act is to
move out to the furthest limits of the circle (D75=B35.9). Cf. Arist. Metaph. 1.4.985a4–7.
60
Trépanier (2017) 176–7.
61
Cf. esp. Sappho fr. 1.5–24; also Il. 5.116; Pind. Isthm. 6.42; Soph. OT. 165. Tor (2017) 334–5 assumes
that this fragment marks the narrator as an ephemeral creature, but it need not refer to his current
incarnation. Gheerbrant (2017) 193–5 rightly stresses how, even if this indefinite figure is the
narrator, the expression hints at the ephemerality of the mortal audience. I find less persuasive his
argument that the absence of a first-person personal pronoun in the extant text constitutes a marked
‘effacement’ of the primary narrator (199–201, 211).
singing herself, but as the accomplice to the narrator, who himself ‘reveals’
the account of the gods.62 This detail, when taken with the name
Kalliopeia, suggests that she will be responsible for the beauty of the
verse rather than its content. The name Kalliopeia, as Alex Hardie has
recently argued, differs conspicuously from its Hesiodic spelling of
Καλλιόπη (Theog. 79), and so may be etymologised as καλὰ ἔπεα, ‘beauti-
ful words’ or ‘verses’, a reading supported by Empedocles’ fondness for
such paretymologies.63 If our analysis of D21=B122 is correct and Love is
indeed the cause of Kallisto, we can infer that Kalliopeia, ‘fineness in
words’, also arises as a consequence of Love.64 The description of the
Muse in D44=B3 (line 3) as a ‘much-wooed, white-armed maiden’ further
suggests an erotic quality.65 A Love-based Muse would be an appropriate
accomplice for the revelation of a ‘good account about the blessed gods’ if
those gods are the four elements: Love brings the four elements together
harmoniously in the cosmos just as she does in the poem.66
c Love’s Craft and Empedocles’ Poetry

Love’s responsibility for poetry is an aspect of her role as divine crafts-
woman. There are some indications that the crafting of the poem itself is in
imitation of Love’s divine craftsmanship. As Sedley notes, she is presented
as a carpenter, ‘fitting’ things together with ‘glue’ (D192=B96) and ‘dowels’
(D214=B87).67 At D75=B35.17, during the description of the onset of Love,
the narrator exclaims that the process is ‘a wonder to behold’ (θαῦμα
ἰδέσθαι), a formulaic expression that bespeaks the aesthetic quality of
Love’s activity, and which in Homer is only ever used of the products of
handicraft.68 The poem itself is presented as the product of woodwork in
D77A=B21:
62
See n. 49 above.
63
Hardie (2013) 215–16; similarly, Gheerbrant (2017) 206. On Empedocles’ paretymologies, see
especially Willi (2008) 224–7 and 243–7.
64
Gheerbrant (2017) 205–9 reaches the same conclusion, though on the different grounds that
Kalliopeia reprises the harmonising role of Calliope in Hesiod’s Theogony (87). An association
between Empedocles’ Muse and Philia is also supported by Hippolytus’ testimony (Haer.
7.31.3<R88).
65
On the erotic connotations of these epithets, see further Gheerbrant (2017) 115–20.
66
On the harmonising function of Empedocles’ Muse, see further Hardie (2013) 227–32.
67
Sedley (2007) 52.
68
A point made by Ritoók (1989) 344, who argues that ‘magic’ is never mentioned in the context of
early products of handicrafts. It is notable, however, that the expression is often used of particularly
special objects or the products of divine craftsmanship, e.g. the wheels of Hera’s chariot, Il. 5.725;
Rhesus’ golden armour (Il. 10.439); Achilles’ first set of armour (Il. 18.83); Hephaestus’ robots (Il.
18.377); the cloak woven by the nymphs (Od. 13.108).
120 Empedocles
ἀλλ’ ἄγε, τόνδ’ ὀάρων προτέρων ἐπιμάρτυρα δέρκευ,
εἴ τι καὶ ἐν προτέροισι λιπόξυλον ἔπλετο μορφῆι,
ἠέλιον μὲν λευκὸν ὁρᾶν καὶ θερμὸν ἁπάντηι,
ἄμβροτα δ’ ὅσσ’ εἴδει τε καὶ ἀργέτι δεύεται αὐγῆι,
ὄμβρον δ’ ἐν πᾶσι δνοφόεντά τε ῥιγαλέον τε· 5
ἐκ δ’ αἴης προρέουσι θελεμνά τε καὶ στερεωπά.
ἐν δὲ Κότωι διάμορφα καὶ ἄνδιχα πάντα πέλονται,
σὺν δ’ ἔβη ἐν Φιλότητι καὶ ἀλλήλοισι ποθεῖται.
ἐκ τούτων γὰρ πάνθ’ ὅσα τ’ ἦν ὅσα τ’ ἔστι καὶ ἔσται,
δένδρεά τ’ ἐβλάστησε καὶ ἀνέρες ἠδὲ γυναῖκες, 10
θῆρές τ’ οἰωνοί τε καὶ ὑδατοθρέμμονες ἰχθῦς,
καί τε θεοὶ δολιχαίωνες τιμῆισι φέριστοι.
αὐτὰ γὰρ ἔστιν ταῦτα, δι’ ἀλλήλων δὲ θέοντα
γίγνεται ἀλλοιωπά· τόσον διὰ κρῆσις ἀμείβει.
But come, look at this, the witness of my earlier words,
And see if anything in my earlier words was lacking-wood in form,
First, the sun, white to look at and hot in every dimension,
Then the immortal things which are drenched in its form and its bright
gleam, 5
Then the storm, in everything, dark and chill;
And then from the Earth flow forth things rooted and solid.
And in Anger all things are of different forms and split up,
But they come together in Love and have desire for one another.
For from these things all the things that were and are and shall be,
The trees shot up and men and women, 10
And beasts and birds and water-nourished fish,
And the long-lived gods, highest in honours,
For these things are themselves, but going through one another,
There come to be different shapes, so much does the mixture exchange them.
The adjective ‘lacking-wood’ (λιπόξυλος) only occurs in Empedocles and
seems to be a metaphor with the sense of ‘incomplete’, a sense that is
supported by the fact that the ensuing passage, which appears to be a recap
of the cosmology that has just been covered,69 is exhaustive in scope,
encompassing everything from the celestial bodies, to plants, animals,
humans and gods, to ‘as many things as were, are, and shall be’. It is as if
the poem were a wooden model of the entire history of the cosmos
fashioned by the poet, himself a master-carpenter, like Love.
The complete poem, then, one that does not ‘lack wood’, is crafted on
the model of Love’s crafting of the perceptible cosmos. This connection
69
For the placement of this fragment, see Trépanier (2004) 34–5, who draws on Simplicius’ introduc-
tion to it at in Phys. 1.159.10, which makes clear that it follows B17, a fragment for which, thanks to
a stichometric line in the Strasbourg papyrus, we now have line numbers.
between Love and the text that is not ‘lacking-wood’ is supported by the
word’s other occurrence, in D61=B71:
εἰ δέ τί σοι περὶ τῶνδε λιπόξυλος ἔπλετο πίστις,
πῶς ὕδατος γαίης τε καὶ αἰθέρος ἠελίου τε
κιρναμένων εἴδη τε γενοίατο χροῖά τε θνητῶν
τόσσ’, ὅσα νῦν γεγάασι συναρμοσθέντ’ Ἀφροδίτηι
And if your conviction is at all lacking-wood concerning these things,
How, when water and earth and air and the sun
Are mixed together, the forms and colours of mortal things come to be
As many as have now come to be, joined together by Aphrodite
Here, the wood imagery is applied to the student’s belief, as if Pausanias
must form his own wooden model on the basis of the text. Again, this
process is associated, this time more explicitly, with Love’s agency in
creating the objects of the perceptible universe, the things that have been
‘joined together by Love’, an expression that further develops Love’s role as
carpenter.70
Her craftsmanship extends specifically to artistic production in the well-
known painters simile, of D60=B23:
ὡς δ’ ὁπόταν γραφέες ἀναθήματα ποικίλλωσιν
ἀνέρες ἀμφὶ τέχνης ὑπὸ μήτιος εὖ δεδαῶτε,
οἵτ’ ἐπεὶ οὖν μάρψωσι πολύχροα φάρμακα χερσίν,
ἁρμονίηι μείξαντε τὰ μὲν πλέω, ἄλλα δ’ ἐλάσσω,
ἐκ τῶν εἴδεα πᾶσιν ἀλίγκια πορσύνουσι, 5
δένδρεά τε κτίζοντε καὶ ἀνέρας ἠδὲ γυναῖκας
θῆράς τ’ οἰωνούς τε καὶ ὑδατοθρέμμονας ἰχθῦς
καί τε θεοὺς δολιχαίωνας τιμῆισι φερίστους·
οὕτω μή σ’ ἀπάτη φρένα καινύτω ἄλλοθεν εἶναι
θνητῶν, ὅσσα γε δῆλα γεγάκασιν ἄσπετα, πηγήν, 10
ἀλλὰ τορῶς ταῦτ’ ἴσθι, θεοῦ πάρα μῦθον ἀκούσας.
And just as whenever painters decorate offerings,
Two men expert in skill, because of their wit,
So when these men take hold of many-coloured paints in their hands,
In harmony mixing some more, others less,
From them they make forms like all things, 5
Creating trees and men and women
And beasts and birds and water-nourished fish
And long-lived gods, best in honours;
In this way do not let deception overcome your mind that from elsewhere
70
For ἁρμόζω and its cognates in the context of carpentry, cf. Od. 5.162, 247; Eur. Cyc. 460.
122 Empedocles
Is the source of mortal things, all of the limitless things that have come to
be manifest, 10
But know these things clearly, having heard the speech from a god.
Although the dual form of the participles (δεδαῶτε, μείξαντε) may hint at the
fact that both Love and Strife have a role in the production of the sensible
cosmos out of the four elements,71 the detail that they are mixing ‘in harmony’
(ἁρμονίηι), a term associated with Love (D60=B23.4; D89=B27.2;
D192=B96.4), stresses Love’s role in particular. The fragment could be seen
as an example of ecphrasis (in the modern sense), like the Homeric shield of
Achilles or the Hesiodic Scutum:72 it describes the visual content of works of art.
The painters elaborately depict all the different phenomena of the world.
Generally, ecphrasis is seen to invite comparison between the work of art
described and the literary text in which the description is embedded. Fowler
and Elsner, for instance, have stressed, in general, the uneasiness of the
relationship between the ecphrastic description and the overall narrative, and
how precisely this uneasiness opens possibilities for interpretation.73 In this
instance, however, it is the similarities between the embedded content and the
wider cosmological narrative that are strongly emphasised: the content of the
paintings is identical to that of Empedocles’ text. Lines 6–8 also occur, with
only minor variations, at D77a=B21.10–12 (quoted above) where, as we saw,
Empedocles’ poem was presented as a product of woodwork, in the manner of
Love’s carpentry of the perceptible cosmos. That the painters’ work here is
analogous to Empedocles’ work of poetry is further implied by the visual
language which Empedocles repeatedly uses of his verbal account (D44=B3.9;
D77A=B21.1; D257=B110.2; D74.326=B76.3; D73.292–3, 300). In fact, the
fragment seems emblematic of Empedocles’ poetics: the painters
ποικίλλωσιν, ‘decorate’ offerings, a verbal cognate of the noun poikilia
which would become a technical term among the scholiasts and other literary
critics for ‘variation’ in poetry.74 This sort of ‘variation’ seems especially apt for
the variety of phenomena that Empedocles describes, in detail, in the physical
fragments.
Since Love plays a role in the painters’ work, we are entitled to infer that she
is also instrumental in the creation of Empedocles’ poetry. The finer details of
this process are obscure. It is not clear precisely how divine inspiration in
71
A point made by Sedley (2007) 59.
72
On the ancient use of the term, as a detailed description of any kind (not just of works of art), see
Webb (2009).
73
D. Fowler (1991); Elsner (2002) 3–9.
74
On the term in Homeric scholia, see Nünlist (2009) 198–202. Porter (2016) 419 identifies a poetics of
poikilia in Empedocles.
Empedocles’ cosmology works or, more specifically, how Love interacts with
the poet to produce poetry.75 Possibly, this is a modification of the traditional
model of ‘double-’ or ‘over-determination’, whereby the same action can be
attributed both to a mortal agent and to the god who inspired it:76 in this
instance, the same action (the performance of the poem) is simultaneously
attributable to a theos (the narrator) and to the cosmic principle (Love). That
we should infer some sort of process along these lines is implied by the
language he uses, and by the principles of his cosmology.
There are, however, two important and related objections to the notion
that Love is responsible for beauty. First, at least according to most
scholars, Empedocles’ perceptible cosmos arises as a result of both Love
and Strife, so that Strife must also play a role in the creation of the myriad
phenomena that exhibit this poikilia and might usually be taken as
beautiful.77 Second, the function of Love is ultimately to unify all entities
into the cosmic Sphere, an item that could hardly be described as ‘beauti-
ful’ in any conventional sense and which conspicuously lacks the poikilia of
Empedocles’ poetry.
In response to the first objection it could be maintained that, even
though the presence of Strife is required for the myriad phenomena,
Love’s agency is always stressed in the creation of the different objects we
perceive, perhaps since she works against the natural inclination of like
portions of elements to separate from mixtures and merge with like.78
Thus, at D77A=B21.14, it is ‘mixture’ (κρῆσις), the characteristic activity of
75
The key study of Delatte (1934) 21–7 sees Empedoclean Muse-inspiration as an example of
‘enthusiasm’ (enthousiasme), of ‘creative madness’ (délire créateur) which is ‘the manifestation of
the god who lives in us’ (la manifestation du dieu qui habite en nous). Delatte relied largely on the late
testimony of Aurelianus De mysteriis 5.15=A98; but as Murray (1981) showed, the notion of Muse-
inspiration as divine madness is not attested before Plato.
76
On which the classic treatment is Dodds (1951) 1–18. See also more recently (with some modifica-
tions of Dodds’ view, whilst still allowing for this basic principle of overdetermination that seems
inconsistent to modern readers) Versnel (2011) 163–79. The most explicit examples of the feature
occur at Il. 16.849, 19.86–96 and Od. 22.347–8.
77
See n. 2 above. Gheerbrant (2017) 643–705 arrives at this conclusion (concisely stated at 704).
78
For the principle that Empedoclean elements are intrinsically attracted to like elements but are
attracted towards unlike ones by the influence of Love, see O’Brien (1969) esp. 301–13, whose view
has been taken up more recently by Primavesi (2016). On the view above, Love is the agent of
creative acts and works against the increasing tide of Strife when, say, she forms individual limbs but
can no longer bring them together into complete animals due to Strife’s influence. Sedley (2007)
48–9 objects to this view on the grounds that it faces ‘the task of explaining what could possibly
motivate Love to continue with zoogony in the age of growing Strife, during which . . . the species
generated during her earlier ascendancy continue to exist’, but this objection is not compelling:
Love, for Empedocles, seems always to want to bring things together (even when she has already
formed some animals, and even when, under increasing Strife, the organs she forms have no chance
of coming together into complete organisms), just as Aphrodite will always want to bring together
lovers and needs no further motivation.
124 Empedocles
Love, rather than separation, that causes the variety of forms to appear.
Even though these objects could not exist at the point of maximum Love,
Love is, for Empedocles, responsible for their creation and operates in spite
of the force of Strife which works to disintegrate them.
The second objection is more challenging. Indeed, the desirability of
the divine Sphere poses a rhetorical challenge for Empedocles since it
involves the obliteration of everything that we know and care about. That
the Sphere is desirable is implied by the positive account of Love in the
cosmos and by the fact that the Sphere ‘rejoices in its surrounding
solitude’ (D89.3=D90.2=B28.2).79 The fragments do not furnish a clear
answer to the question of why we should want the formation of the
Sphere, but one can be provided on the basis of the distinction between
divine and mortal perspectives: the beauty of the phenomenal world is
beautiful from a mortal perspective, but once we relinquish such
a perspective, we realise that the Sphere itself is the most beautiful object
in the history of the cosmos. If this interpretation is along the right lines,
then the poem acts as a ‘a ladder which must be thrown away when one
has climbed it’:80 we are to identify the Love underlying the instances of
beauty in the current cosmos and thus be led to the realisation that the
Sphere itself is the most desirable state of affairs. This scheme may lie
behind the command to ‘see [Love] with your mind, and do not sit
stunned by your eyes’: rather than being dazed by the poikilia of the
sensible world, we are to contemplate the beneficial force of cosmic Love.
Such an understanding of the function of Empedocles’ text, tentative
though it is, would be consistent with the sort of subtextual interpretation
that he, at times invites us to apply (see further Section 5d below). At any
rate, the responsibility of Love for beauty is strongly implied by the
repeatedly positive descriptions of her and her work, and by the origins
of Kallisto in Empedocles’ theogony: these considerations should weigh
in favour of reading Empedocles’ beauty as a manifestation of the influ-
ence of Love, as opposed to that of Strife.
d Conclusion to Section 3
Empedocles has long been seen to defend the variegated perceptible world
against Parmenides, whilst accepting his principles that nothing can be created
79
See n. 59 above.
80
As Owen (1960) 100 understands Parmenides’ argument, borrowing an analogy from Wittgenstein
(Tractatus Logico-Physicus 6.54) and Sext. Emp. Math. 8.481.
ex nihilo or destroyed in nihil.81 In addition to this metaphysical response to
Parmenides, the cosmology can also be seen to encompass an aesthetic one:
poetry which depicts the wondrous variety of sensible phenomena need not be
‘deceptive’ (cf. D60=B23.9; D73.257); beauty in poetry derives from the same
source as harmony and moral rectitude in the perceptible world, that is,
cosmic Love. Empedocles’ theology is sometimes regarded as more systematic
than pre-philosophical, poetic conceptions of the gods, on the grounds that
his cosmic forces, whilst still being divine agents, follow regular, predictable
patterns of behaviour.82 This systematic theology encompasses a systematic
conception of beauty and specifically the beauty of poetry.
If the above reconstruction is accepted, then Empedocles’ poem repre-
sents a significant moment in the development of the Classical conception of
poetry. For some scholars, the idea of poetry as a mortally crafted product,
attributable to a particular creator-poet and explicable in terms of regular
natural processes, arises only later in the fifth century at the hands of prose
authors who were readers of poetry rather than performers or composers. In
support of this view, some have observed that Homer never refers to the
poet’s activity in terms of ‘skill or ‘craft’,83 and argued that, when epinician
poets use such imagery in the early to mid-fifth century, they do so in order
to assert the superiority of poetry over plastic arts as a means of bestowing
kleos on the laudandus.84 Later, Aristophanes and Plato used such imagery
poetologically, in pejorative contexts, once poetry had begun to be thought of
in terms of written texts as well as performances.85 This scholarly narrative can
be modified in light of Empedocles’ poetics. He stands out as a poet who,
probably in the first half of the fifth century, presented his work as a product of
craftsmanship. Although the rise of the written text undoubtedly facilitated
increasing adherence to the artisanal conception of poetry, that conception, in
81
On Empedocles’ cosmology as a response to Parmenides in this way, see e.g. Graham (1999); Laks
(2004); Curd (2005) 155–71; O’Brien (2005) 323–4. Contrast Palmer (2009) 260–317, whose views are
discussed in Section 3a above.
82
Contrast the arbitrariness of the behaviour of Homeric and Hesiodic gods. On this sort of regularity
as a distinctive achievement of the Presocratics, see e.g. Long (1999) 13–14. On Presocratic ‘theology’
as being more rational than that of Homer or Hesiod, see Trépanier (2010); cf. also Broadie (1999),
for whom what distinguishes certain Presocratics (including Empedocles) from Homer or Hesiod as
theologians is that they question, and investigate, what constitutes godhood.
83
Svenbro (1984) 156–79; Finkelberg (1998) 101–5; Ford (2002) 113–14.
84
Thus Ford (2002) 113–30, endorsing the verdict of Bowra (1964) 4. Contrast Maehler (1963) 25.
Grethlein (2008) argues for a similar interpretation of the references to visual arts in Homer.
85
E.g. Ar. Thesm. 52–7, where the slave describes the poet Agathon’s creative process. As Austin and
Olson (2004) ad loc. comment, ‘the overall effect . . . is to present Agathon not as a divinely inspired
poet . . . but as a mere wordsmith’; Pl. Phdr. 278d–e, where, as Yunis (2011) ad loc. comments, the
metaphor describes ‘the non-philosophical writer’s protracted, excessively fussy manner of compos-
ition’. On these passages, see Ford (2002) 156.
126 Empedocles
his work, is tied to the notion of a divine creator of the phenomenal world and
its contents: Empedocles’ poetry imitates the activity of Love in creating
a cosmos of sorts. On the other hand, he retains the older conception of poetry
as the product of divine inspiration, albeit in a systematised form. Moreover,
the narrator’s utterance, in the address to the Acragantines in D4=B112, as we
shall see, is presented in a specific social context, to the end of instituting new
cult practices. Poetry, then, at least on the evidence of his own poetic practice, is
not conceived by Empedocles as the self-contained, autotelic objet d’art of
(some) modern criticism and, to some degree, presaged by Aristotle’s Poetics.86
But in presenting his poetry as an object of craftmanship and as analogous to
painting in its representation of the world, Empedocles can be seen as a liminal
figure, with one foot in the earlier tradition of Muse-inspired Archaic song and
the other in the world of finely crafted Classical poetry.
4 Empedocles’ Narratives
A more detailed investigation of the function of Empedocles’ poetry will
require a consideration of its wider architecture. It shall be argued here
that two narratives in particular underpin the text and would have
conditioned a contemporary audience’s response to it. Glenn Most has
argued that narrative is a feature that early Greek poetry bequeathed to
early Greek philosophy: the divine drama of Empedocles’ cosmology
thus takes the form of a dynamic narrative like the lays of Homer and
Hesiod.87 But it would be fair to say that the account of the cosmology is
less conventionally a narrative than the story of the daimon’s exile. The
latter conforms to the familiar plot-type of exile and return, detailing the
actions of an anthropomorphic agent and the consequences of those
actions. In Aristotelian terms, the daimonology is a muthos, a ‘plot-
structure’,88 but the cosmology is not. When Aristotle denies the title of
poet to Empedocles on grounds of mimesis, it is clear that he is thinking
of the cosmology rather than the daimonology (Poet. 1.1447a–b): the
86
For the conception of literature as autotelic, note Maslov (2015) 9–10, who claims that this notion
was ‘made possible by the autonomization of the aesthetic realm in the second half of the nineteenth
century’. This may be true with respect to the specific term ‘autotelic’ (first attested, according to the
OED, in 1864), but the idea that literature is an end in itself is usually traced to eighteenth-century
Romanticism (e.g. Eagleton (2008) 15–19). It is for this reason that Goethe rules out the possibility
that poetry can be didactic in Über das Lehrgedicht. For the idea that Aristotle sees literature as
having autonomous value, note e.g. Finkelberg (1998) 10–18; Ford (2015) 18–20.
87
Most (1999) 348–9.
88
For this translation of Aristotle’s use of muthos, see Else (1957) 242–4; Halliwell (1998) 5, 12, 141–2 et
passim; Gudeman (1934) 76–7 sees it as synonymous with pragmata, ‘actions’ or ‘deeds’.
Empedocles’ Narratives 127
story of the daimon on its own would have fulfilled Aristotle’s criteria for
poetry since its subject matter consists of the actions of a particular
character (cf. Poet. 1.1448a, ‘those who engage in mimesis portray
[mimountai] people in action [prattontas]’).89 The first-person narrative
of the daimon, then, is a narrative in a manner in which the cosmology is
not. In addition, we can also distinguish, in Don Fowler’s term, ‘a
didactic plot’ in the story of the narrator’s education of the internal
addressee, Pausanias, another story involving ‘people in action’.90 This
process, as commentators have pointed out and as we shall explore in
greater detail, conforms to the model of mystic initiation. These two
narratives – the first-person narrative of the daimon and the initiation of
the student – will be the central concern of this section. I shall here
consider the two narratives separately and in turn but shall argue that
they are interrelated and complementary in important respects.
a The Narrative of the Daimon

D4=B112 appears to be the opening of the work, and introduces the
narrator as an arrival at the city of Acragas:91
ὦ φίλοι, οἳ μέγα ἄστυ κατὰ ξανθοῦ Ἀκράγαντος
ναίετ’ ἀν’ ἄκρα πόλεος, ἀγαθῶν μελεδήμονες ἔργων,
ξείνων αἰδοῖοι λιμένες, κακότητος ἄπειροι,
χαίρετ’· ἐγὼ δ’ ὑμῖν θεὸς ἄμβροτος, οὐκέτι θνητός
πωλεῦμαι μετὰ πᾶσι τετιμένος, ὥσπερ ἔοικα, 5
ταινίαις τε περίστεπτος στέφεσίν τε θαλείοις.
τοῖσιν ἅμ’ εὖτ’ ἂν ἵκωμαι ἐς ἄστεα τηλεθάοντα,
ἀνδράσιν ἠδὲ γυναιξί, σεβίζομαι· οἱ δ’ ἅμ’ ἕπονται
μυρίοι ἐξερέοντες, ὅπηι πρὸς κέρδος ἀταρπός,
οἱ μὲν μαντοσυνέων κεχρημένοι, οἱ δ’ ἐπὶ νούσων 10
παντοίων ἐπύθοντο κλυεῖν εὐηκέα βάξιν,
δηρὸν δὴ χαλεπῆισι πεπαρμένοι <ἀμφ’ ὀδύνηισιν>.
89
A point made by Gudeman (1934) 91 (writing at a time when there was a consensus that the
fragments came from two separate poems concerning separate topics). On mimesis and action, see
e.g. Halliwell (1998) 128, (2002) 177–9. On Aristotle’s literary reception of Empedocles, see Bremer
(1980) and Palumbo (2007) (both of whom focus on Empedocles’ use of poetic language, rather than
plot).
90
D. Fowler (2000).
91
Diogenes Laertius describes Empedocles as ‘starting the Katharmoi’ with the fragment (Diog. Laert.
8.62 ἐναρχόμενος τῶν Καθαρμῶν φησιν). On the order of the surviving fragments that can be placed
in the first book, see Trépanier (2004) 32–72. For a possible reconstruction of the opening lines of
the work, incorporating both D4=B112 and D10=B115, see Rashed (2008).
128 Empedocles
O friends, who, in the great city of yellow Acragas
Dwell in the citadel, concerned with good deeds,
Respectful harbours for visitors, inexperienced in ill,
Greetings: I come to you as an immortal god, no longer mortal
Honoured among all people, just as I appear, 5
Crowned with ribbons and with fresh garlands.
Whenever I arrive with them in flourishing cities,
I am honoured by men and women; and they follow me
In thousands, asking me where the path is to gain,
Some in need of prophecies, others, for diseases 10
Of all kinds, seeking to hear a healing utterance,
Pierced for a long time by difficult pains.
We should not assume that this fragment reflects a real historical scenario,
with Empedocles, now a celebrated sage, addressing the citizens of his
home town.92 The use of a fictional narrator would be paralleled in other
Archaic didactic texts, most famously the Hesiodic Precepts of Cheiron.93 In
this instance, the testimony that the poem was performed at Olympia by
the rhapsode Cleomenes assumes a performance outside of Acragas by
a performer who is unlikely to have believed that he was a god. This
testimony should caution against the assumption that the narrator is to
be identified with the author, since the connection would be severed in re-
performance.94 Dirk Obbink has argued that the reference to Acragas is
a sphragis indicating that the narrator is Empedocles,95 but if we assume the
narrator to be identical to that of D10=B115 (as Obbink does), then it seems
92
Palmer (2009) 260–1, for instance, treats this as evidence for ‘how Empedocles saw himself’; Guthrie
(1965) 132–4 similarly uses this as evidence for Empedocles’ personality, followed by Lloyd (1987)
100–1 (who cites the biographical anecdotes at Diog. Laert. 8.73 which themselves probably derive
from this passage); most extensively, Bidez (1894) 105–76 is a biography of Empedocles that takes the
first-person statements in the fragments at face value.
93
Kurke (1990) 104–7 catalogues evidence for a genre, traceable to the fifth century, of hypothekai
poems with mythological narrators. In addition to the Precepts of Cheiron, she identifies collections
of advice spoken by Amphiaraos (to Amphilochus) Apollo (to Admetus), Nereus, Rhadamanthys,
Pittheus, and Sisyphus. An older Nestor-poem may have stood behind Hippias’ composition of
a ‘Logos of Nestor to Neoptolemus’ (Pl. Hp. Ma. 286A5-B4; Philostr. V. S. 1.11.4). The best-attested
of these poems (aside from the Precepts of Cheiron) is the Admetou Logos, advice given by Apollo to
Admetus (Bacchyl. 3.76–83; Praxilla 3=Carm. Conv. 14 PMG; Cratinus fr. 254 PCG; Aristophanes fr.
444 PCG; Zenobius 1.18).
94
On the connection between author and narrator in Archaic Greek poetry, see Morrison (2007)
57–67, who argues that although, in most cases, the identity of the primary narrator seems grounded
in that of the author, in some examples this is clearly not the case (cf. female speakers in Alcaeus fr.
10, Anacreon PMG 385 and Theognis 257–60; Charon in Archilochus fr. 19; the self-deprecating,
debauched persona of Hipponax’s narrator). The extreme nature of Empedocles’ narrator (a god)
supports the notion that his narrator is another such example.
95
Obbink (1993) 78.
significant that he is not from Acragas; rather, he is at home in the
community of the gods from which he has been exiled.
Whatever the actual original performance context, it is significant that,
within the fictional world of the poem, the narrator is situated within
a specific real-world polis. The greeting (χαίρετ[ε], line 4) suggests that he
has just arrived, a detail further supported by the fact that he has been
wandering around various cities (lines 5–7). This scenario contrasts not-
ably with the Homeric epics – where the narrator’s location is nowhere
mentioned – and with Parmenides’ poem, where the narrator is a generic
kouros who speaks from an unspecified location. The specific setting of
the narrative within Acragas, then, highlights the relevance of the narra-
tor’s speech to a particular, contemporary political community.
The characterisation of the narrator concords with this political
theme. The description of his appearance (lines 5–6) – which would
remain constant regardless of what the performer was wearing –96
and of the honour he receives marks him out, in modern sociological
terms, as a charismatic, someone distinguished from the ordinary run
of mortals, who commands authority through personality rather than
by law or tradition.97 A pertinent parallel is Pythagoras, the leader of
an untraditional religious movement who is reported to have worn
the unusual clothing of a white robe with trousers and a golden
wreath.98 In Richard Martin’s phrase, the narrator is a ‘performer of
wisdom’, an itinerant sage who demonstrates his sagacity by public
enactments, including but not limited to the artistic performance of
poetry.99 He has a reputation for expertise in the areas of moneymaking,
prophecy and medicine, presumably because he has performed relevant
feats in a public context before. It may be ironic that some people come
seeking κέρδος, material ‘gain’ or ‘profit’:100 the fragments hardly offer
help with moneymaking, but we discover that the ‘gain’ to be achieved is
really a spiritual one, through purification of our souls and the reward of
96
It is of course possible that a rhapsodic performer would be garishly dressed (as suggested by Pl. Ion
530b), but this does not always seem to have been the case (note performer depicted on ARV2 183.15,
1632 wearing a simple himation – I see no reason for the scepticism of Herington [1985] 10–15 over
this figure’s identity as a rhapsodic performer). Whatever the actual performer happens to be
wearing, the fictional narrator is dressed in this elaborate manner.
97
The term stems from Weber (1947) 358–63. The sociologist Eileen Barker has argued that new
religious movements frequently have charismatic leaders. See Barker (2015). On this sort of figure
in Archaic Greece as a prototype for the figure of the philosopher, see Gernet (1981) 352–64.
98
Ael. VH 12.32. For Pythagoras as a Weberian charismatic, see Riedweg (2008) 1–41.
99
Martin (1993).
100
See LSJ s.v. Cf. e.g. Pind. Pyth. 3.54 (used of Asclepius’ acceptance of gold to bring the deceased
back to life); Soph. Ant. 222 (Creon sees profit as the motivation for burying Polyneices).
130 Empedocles
a superior reincarnation. The parallel with the dialogue of Solon and
Croesus supports this possibility (Hdt. 1.30–3). Also comparable is the
story that Thales, in order to prove how easy it is to acquire wealth,
rented several oil-mills on foreseeing that it would be a good season for
olives (Arist. Pol. 1.11.1259a6). The point of the story is the showiness of
the activity: Thales did not do this for the intrinsic benefits of wealth,
but in order to demonstrate a particular point to an audience.
Similar tales of practical or political wisdom, often with high-status
onlookers, are repeatedly told of the seven sages.101 Thales, in addition to his
reputation as a speculative philosopher, was also reputed to have performed
the more practical feat of enabling Croesus to cross the river Halys, without
the aid of a bridge (by diverting it upstream), so that he could attack the Medes
(Hdt. 1.75). Solon is perhaps most famous for his Athenian political achieve-
ments. Indeed, one well-known story, recorded in Plutarch (Sol. 1–3) and
Diogenes Laertius (1.46), bears a particular resemblance to Empedocles’
narrator: in order to convince the Athenians to renew the war with Megara
over Salamis, Solon feigned madness and then leapt into the agora, wearing
either a traveller’s cap (pilidion, in Plutarch’s version) or a garland (according
to Diogenes), and when a large crowd assembled, performed his poem on
Salamis.102 Empedocles’ narrator similarly arrives at a public part of the city
and performs poetry to an amassed crowd whilst wearing distinctive clothing,
garlanded, as Solon was if Diogenes is to be believed. Whilst such anecdotes
about the sages are rarely credible, they evince a cultural stereotype to which
Empedocles presents his narrator as conforming. The addressees of D4=B112
seem similarly high-status. In occupying the acropolis, they are presumably the
elites of the city, perhaps the chieftains who are incarnations of the souls of the
wise (D39=B146).103 In depicting his narrator as a sage in this manner,
Empedocles implies that he will be able to dispense practical or political advice.
Indeed, Empedocles himself would be credited with similar political
achievements to those of Solon: according to Timaeus, he put an end to an
‘oligarchy of the 1,000’ which had existed for three years (FGrH 566
F 2=Diog. Laert. 8.66) and then refused the offer of kingship (basileia;
FGrH 765 F 33; Arist. fr. 865 Rose), as a result of which he put an end to
101
As Martin (1993) 115–16 points out.
102
On the significance of the pilidion and this story in general, see Irwin (2005) 134–53, who cites
ancient depictions of Odysseus wearing the same cap (e.g. Il. 10.265 with the scholiasts ad loc.) and
suggests that the detail paints Solon as a cunning but transgressive trickster.
103
Stehle (2005) argues that these addresses are the gods of the citadel, but if that is the case, the
reference is an oblique one. As I shall argue below, the praise heaped on the Acragantines is better
explained as a captatio benevolentiae to ensure the success of the narrator’s supplication as a polluted
exile in need of refuge.
civic stasis (FGrH 84 F 28).104 Solon was similarly said to have refused the
tyranny of Athens (frr. 32–4 W; Plut. Solon 14–15). Whether or not we
believe these accounts of Empedocles’ career, they support a political
reading of D4=B112: if the biographical stories were invented, then they
most likely derive, like other ancient lives of poets, from the poetry itself.105
Even if they were not invented, the characterisation of the narrator
unquestionably contributed to Empedocles’ reputation in later antiquity,
and so D4=B112 could have facilitated the spread of these particular stories.
Ancient lives of poets are a form of literary criticism,106 and, in this case,
they support a reading of the Empedoclean narrator as a polis-helping sage
who fosters civic harmony and vanquishes internal strife (cf. D91=B27a).
The most notorious feature of this fragment, however, is the narrator’s claim
to be ‘to you an immortal god no longer mortal’. This is a further feature that
conforms to the type of the charismatic sage: Pythagoras is, again, a relevant
parallel since he was said to have been an incarnation of Apollo (Arist. fr. 191
Rose; Iambl. VP 30). But commentators are divided as to whether we should
take Empedocles’ expression as an actual claim to godhood or merely as an
indication of the narrator’s status from the perspective of his addressees. This
issue depends largely on whether we take ὑμῖν as an ‘ethic’ dative, intensifying
the expression (as it were, ‘I am a god, you see’; cf. Hdt. 5.30) or a ‘limiting’ one
(‘I am a god to you’, but not necessarily a god simpliciter).107 The latter
interpretation is supported by the possibility that the rest of the sentence
qualifies that he is only a god in appearance (ὥσπερ ἔοικα, ‘just as I appear’),
but ancient readers certainly tended to treat this simply as a claim to be a god,
and their interpretation is supported by D60=B23.11 (‘But know these things
clearly, having heard the speech from a god’).108 As Alex Long points out, in
a parallel instance (Od. 22.348), ὥσπερ ἔοικα indicates what the speaker
deserves, not how he is perceived.109 To have such a divine narrator in
a didactic text of this period is also paralleled in the Admetou Logos,
a collection of sayings that Apollo supposedly delivered to Admetus.110 It
104
One the history of Acragas, see Fischer-Hansen, Nielsen, and Ampolo (2004) 186–9, who treat
these stories of Empedocles as accurate. On the ancient sources for the life of Empedocles, see Bidez
(1894) 1–104.
105
That ancient lives of poets are extrapolated from the works themselves has been shown by Lefkowitz
(2013) and Kivilo (2010). For the application of the same principle to Empedocles, especially the
stories about his death, see Chitwood (1986).
106
For this principle, see Introduction n. 2 above.
107
Most commentators (e.g. Zuntz [1971] 189–90; KRS 313–14; Wright [1995] 266; Trépanier [2004]
80–2; Long [2017] 12) assume the former interpretation, but van der Ben (1975), 23–4; Rösler (1983)
172–5; Graham (2010) 404–5, 430; and Palmer (2013) 311 adopt the latter.
108
A point made by Willi (2008) 193. On D60=B23.11, see Section 3a above.
109
Long (2017) 12. 110 See n. 93 above.
132 Empedocles
seems preferable, then, to take this as a claim actually to be a theos, but the
description of his appearance and reputation may already lead us to question in
precisely what respect he is a god (just because of his reputation and appear-
ance, or because of his abilities?). This question prepares us for the later
revelation that a theos for Empedocles is rather different from a theos as
conventionally understood.111 But in any case, the expression is designed to
be striking, even shockingly hubristic, perhaps calling to mind the mythical
Salmoneus who was punished for imitating Zeus.112 It is, then, a further feature
that distinguishes the narrator as an extraordinary charismatic.
When viewed within the context of other Archaic Greek charismatics
and performers of wisdom, the self-characterisation of the narrator, as well
as the detail of his arrival at a particular polis, intimate that he will offer
political advice and may bring about political reform. But this success story
has been regarded by some as incompatible with the other major source for
the narrator’s biography, D10=B115:
ἔστιν Ἀνάγκης χρῆμα, θεῶν ψήφισμα παλαιόν,
ἀίδιον, πλατέεσσι κατεσφρηγισμένον ὅρκοις·
εὖτέ τις ἀμπλακίηισι φόνωι113 φίλα γυῖα μιήνηι,
< > ὅς κ(ε) ἐπίορκον ἁμαρτήσας ἐπομόσσηι,
δαίμονες οἵτε μακραίωνος λελάχασι βίοιο, 5
τρίς μιν μυρίας ὧρας ἀπὸ μακάρων ἀλάλησθαι,
φυομένους παντοῖα διὰ χρόνου εἴδεα θνητῶν
ἀργαλέας βιότοιο μεταλλάσσοντα κελεύθους.
αἰθέριον μὲν γάρ σφε μένος πόντονδε διώκει,
πόντος δ’ ἐς χθονὸς οὖδας ἀπέπτυσε, γαῖα δ’ ἐς αὐγὰς 10
ἠελίου φαέθοντος, ὁ δ’ αἰθέρος ἔμβαλε δίναις·
ἄλλος δ’ ἐξ ἄλλου δέχεται, στυγέουσι δὲ πάντες.
τῶν καὶ ἐγὼ νῦν εἰμι, φυγὰς θεόθεν καὶ ἀλήτης,
νείκεϊ μαινομένωι πίσυνος.
There is an oracle of Necessity, an ancient decree of the gods,
Eternal, sealed with broad oaths;
111
Thus Trépanier (2004) 73–82. On Empedocles’ different senses of theos, see Rangos (2012).
112
Salmoneus, who was struck by a thunderbolt and consigned to Tartarus for imitating Zeus, a story
covered in the Catalogue of Women (frr. 15, 30 M.-W.) and most famously told by Virgil at Aen. 6.
585–94. See Gantz (1993) 171–3.
113
Wright (1995) 272–3; Gagné (2006) 83–7; Picot (2007); and Rashed (2008) 10 here all prefer the
reading φόβωι, with the sense ‘in exile’ or ‘because of fear’, as found unanimously in the manuscripts
of Plutarch De exilio (which quotes the line). However, I follow most editions (including DK and
LM) in accepting Stephanus’ emendation to φόνωι (found in his 1572 edition of Plutarch).
Primavesi (2001) 35–7 and Gheerbrant (2017) 657–66 argue in favour of the emendation, in support
of which is the fact that pollution through bloodshed is the narrator’s main concern at B139=D73,
D28=B136 and D29=B137, whilst lack of bloodshed marks the realm of the blessed at D25=B128.8.
Whenever someone in error pollutes his dear limbs with bloodshed
< . . . > whoever has committed a fault by breaking an oath,
The daimones who have been allotted a long life, 5
He must wander for thirty thousand seasons away from the blessed ones
Being born during this time in all kinds of forms of mortals
Changing the painful paths of life
For the force of the upper air drives him to the sea,
And the sea spits him out onto the ground of the land, and the earth into
the rays 10
Of the shining sun, who throws him into the eddies of the upper air;
And one receives him from another, but they all hate him.
And I am now one of them, an exile from the gods and a wanderer,
Relying on mad Strife.
The narrator’s claim, at D4=B112.4 ‘to be an immortal god no longer
mortal’ seems to jar with the statement, at D10=B115.13, that he is ‘now
exile from the gods and a wanderer’. Consequently, some have thought
D10=B115 to come from an earlier poem than D4=B112, in which the
narrator was still an exile, before he regained his godhood.114 If that is
indeed the case, then it should be noted that the two poems form two
parts of a single narrative of the fall and then resurgence of a particular
daimon. This sort of narrative development across an author’s two poems
would be paralleled in Hesiod’s correction of the Theogony’s account of
Strife (Theog. 225) in the Works and Days (Op. 11–12).115 But we need not
suppose that D4=B112 and D10=B115 come from two distinct poems.116
As Trépanier has argued, the apparent inconsistency could serve
a rhetorical purpose.117 He draws upon Mansfeld’s observation that
Presocratic texts often begin with strikingly obscure openings which are
only elucidated later on in the text, thus affording an ‘insight by
hindsight’.118 The two fragments can be read as conforming to this
pattern. The narrator’s claims to godhood in D4=B112 are shocking,
but we later learn, in D10=B115, that the narrator is in fact a daimon
who is exiled from the real gods, in some respects conforming to popular
conceptions of gods, but no longer at the level of the highest status of
being that can be achieved. In retrospect, we can see that the narrator
114 115
Kahn (1960) 29; Sedley (1998) 9–10; Long (2017) 16. On which see Most (1993a) esp. 77.
116
Rashed (2008) offers a plausible, if ultimately highly speculative, reconstruction of the first thirty-
three lines of the Katharmoi that incorporates both D4=B112 and D10=B115, proving that the
fragments can be taken together in a manner that makes sense. It seems most likely that D10=B115
came from a poem entitled the Katharmoi, since Hippolytus mentions the Katharmoi shortly after
quoting from the fragment (Haer. 7.29.14–25; 7.30.3–4).
117 118
Trépanier (2004) 73–107. Mansfeld (1995).
134 Empedocles
claims to be a god, not merely because of his dress and honours, but in
a stronger sense: he is what they take to be a god, an anthropomorphic
being with superhuman powers, who will outlive mortals. The expression
‘just as I appear’ (ὥσπερ ἔοικα) hints at this typically Presocratic gap
between appearance and reality.
Against this rhetorical reading it might still be objected that the celebratory
tenor of D4=B112 is incompatible with the lamentations of D10=B115. The
exile seems distinct from the sage: as Silvia Montiglio has argued, the narrator
of D4=B112 is like the triumphant Oedipus at the start of the Oedipus
Tyrannus, whereas the speaker of D10=B115 is more like the exiled
Orestes.119 These two fragments, however, are in fact more compatible than
her contrast implies, and when taken together, conform to a plot-type more
widely attested in Greek myth and legend of the blood-polluted exile who
arrives at a new polis and plays an important, beneficial, role in its history. 120
As Petrovic and Petrovic argue, D4=B112 can be read, in light of D10=B115, as
an exile’s supplication on arriving at a new destination. The expression
‘respectful harbours for strangers’ suggests that the location is a sanctuary
for suppliant exiles such as the narrator, just as the Areopagus acts as
a sanctuary for the polluted Orestes.121 The praise of the Acragantines
(D4=B112.2–3) serves as a captatio benevolentiae to ensure that the narrator’s
supplication is granted. But we can go further than this. The status of an exiled
arriviste not only presents the narrator as a suppliant, but also imbues him
with a special authority. As Richard Martin has shown, the story of the
outsider or metanastes who moves to a new community and is able to speak
with an insight not found among the insiders was a narrative topos.122 Perhaps
119
Montiglio (2005) 100–5.
120
I use ‘myth’ here to refer to stories which are unlikely to be true and ‘legend’ to refer to stories which
could be true, but for my present purposes the difference is inconsequential: these stories, whether
they were true or not, were cultural narratives and evince patterns of thought. Cf. Dougherty (1993)
3–11 on such narratives concerning the foundation of Greek colonies.
121
Petrovic and Petrovic (2016) 82–5. Note LSJ s.v. λιμήν II ‘haven, retreat, refuge’. Cf. Aesch. Supp.
471, where king Pelasgus, in a state of aporia as to whether to receive the suppliant Danaids, claims
that he has ‘nowhere a refuge from ills’ (κοὐδαμοῦ λιμὴν κακῶν), an expression that is clearly
designed to recall the situation of the Danaids themselves. Aeschylus’ Danaid trilogy (on which see
Garvie [1968] 163–233) presents a further example of a myth that involved the arrival of strangers,
murder, the purification of blood guilt and the institution of a new political order. Purification only
occurs explicitly in the version of the myth told by Apollod. Bibl. 2.1.5, where Athena and Hermes
purify the sisters of the murder at the command of Zeus, but as Garvie notes (p. 213), the mass
murder would pollute the city as a whole (cf. Suppl. 455ff.), which itself would then require
purification. See also Gould (1973) 85–90 and Cairns (1993) 113–9, 183–5, 189–93, 221–7 and
276–87 for the role of the aidōs/adoios in supplication scenes.
122
Martin (1992). In addition to the examples above, note also the legendary Scythian wise man
Anacharsis, who travelled to Greece and learned Greek ways (Hdt. 4.76ff.), as well as Thucydides,
the best-known example is the Iliadic Phoenix who, like Empedocles, has been
exiled as a consequence of neikos (Il. 9.448) and is then able to speak effectively
to the otherwise intransigent Achilles (Il. 9.434–603). Patroclus also fits the
model: he has had to leave his native country because he has killed another
boy (Il. 23.85–90) and is then given refuge by Pelias. When he speaks, he
makes more of an impression on Achilles than anyone else (esp. Il. 16.21–45).
More similar to Empedocles is the Odyssean Theoclymenus, a prophet
(mantis), who has had to flee Argos for killing one of his kinsmen (Od.
15.222–78). He is granted refuge by Telemachus at Pylos and, when they
return to Ithaca, correctly interprets omens that Odysseus is already on the
island (Od. 17.152–61) and that the suitors will all be slaughtered (Od.
20.351–70). Oedipus himself was an exile at Thebes who had been polluted
through bloodshed. Martin’s article argues that the Hesiodic persona in the
Works and Days is also a metanastes, but the pattern is much more applicable
to Empedocles.123
The figure of the authoritative arriviste evoked in D4=B112, then, seems
closely associated in the Greek imagination with the polluted exile. I want
to suggest that one particular type of this figure is relevant here: the
murderous oikist, the founder of a new colony who has had to leave his
native land due to blood-pollution. Thus, in the Iliad (2.661–70),
Tlepolemus fled Argos after murdering his uncle to colonise the island of
Rhodes, a story that also occurs in Pindar’s Olympian 7 (20–33, 77–80).124
Similar stories are told, for instance, of Archias’ foundation of Syracuse
after being exiled from Corinth for killing the boy Actaeon (Plut. Mor.
772e–773b); of Orestes’ foundation, when in exile, of a city called Argos
Oresticum (Strabo 7.7.8); and of Perseus’ foundation of Mycenae after
accidentally killing his grandfather Acrisius at Argos (3 FGrH fr. 12; Paus.
2.15.4, 2.16.2–3; Apollod. 2.4.4).125
whose exiled status enables him to acquire a broader perspective on events (5.26.5). Cf. also the
authority of ‘wandering poets’, discussed at Chapter 1 Section 3a above.
123
Hesiod’s outsider status is hardly emphasised in the poems, with only a brief mention of his father’s
travels in the Works and Days (635–40). Martin (1992) 19, 31 n. 31 strengthens his case by arguing
that second-generation immigrants were perceived as being immigrants as much as their parents
and by referring to biographies that stress Hesiod’s immigrant status. By contrast, the distinction
between the Empedoclean narrator and his addressees is clearly stressed, both in the fragments
traditionally assigned to the Katharmoi (D4=B112) and in those assigned to the Phusika (e.g.
D42=B2).
124
The story first occurs at Il. 2.661–70 and is also told by Pindar when celebrating a victory for
Diagoras of Rhodes (Ol. 7.20–33, 77–80). In Pindar’s version, Tlepolemus, like Empedocles, is
honoured ‘like a god’ (ὥσπερ θεῷ) by the citizens of Rhodes.
125
On this topic, see esp. Dougherty (1993) 31–44; note also Parker (1983) 392: ‘The list of [mythical
characters] who, but for the need to transfer them from one mythological homeland to another (or
136 Empedocles
Carol Dougherty argues that the reason for the preponderance of such
stories concerning oikists is in the connection between blood-pollution
and purification:
Rituals of purification . . . provide the Greeks with a conceptual model with
which to describe colonization in terms of the expulsion of part of its
population, its galvanization of individuals into a unified group, and its
creative role in founding a new city. A city’s need to colonize appears to be
very much like its need to be purified.126
This role of purification in the galvanisation of new unified groups is not
restricted to the foundation-myths of colonies but is also evident in certain
stories concerning watershed moments in the histories of particular poleis.
Oedipus, in Sophocles’ Oedipus at Colonus arrives a polluted murderer at
the Athenian hill, his supplication is granted and his death establishes
a hero-cult that will protect Athens from Sparta (1518–39).127 A further
pertinent example is Epimenides, who, like Empedocles, was credited with
composing a Katharmoi.128 In Epimenides’ case, the title is clearly con-
nected to his reported achievement in purifying Athens, in the early sixth
century bce, after the Alcmaeonid massacre of the supporters of Cylon.129
Gagné writes of the role of this story in Athenian cultural memory, that ‘his
purification is portrayed as a watershed separating then and now’; and that,
in Plutarch’s account, ‘The reform of Epimenides . . . is not presented as
a mere reaction to the agos, but nothing less than a step in the creation of
Athens.’130 Commentators have sometimes noted the parallels between the
Epimenides story and Aeschylus’ Eumenides.131 In both cases, an earlier
other aetiological reasons), need never have killed is a long one . . . The motif of the killer who,
perhaps after consulting Delphi, founded a foreign colony, was a natural development.’
126
Dougherty (1993) 37, applying the framework of Douglas (1966). Parker (1983) 375–92 lists killers in
Greek myth who are exiled and purified.
127
Note also that Oedipus’ speech at 84–110 in address to the Eumenides shares a number of elements
with Empedocles D4=B112 (arrival; plural address to residents of a holy place; appeal to xenia; praise
of the city). Cf. also 469–70 with D44=B3.1–2.
128
Strabo 10.4.14 claims Epimenides wrote a Katharmoi in verse, while Suda s.v. Epimenides claims it
was in prose. See Obbink (1993) 56–7 n. 15. Much later, Pythagoras is also presented as having
composed a poem of that title (Carmen aureum 67f.). Musaeus (schol. ap. Ar. Ran. 1033=2A6 DK,
Plato Resp. 2.364e) and Orpheus (Plato Resp. 2.364e) are also credited with coming up with
‘purifications’ (katharmoi), but in their cases it is not clear that the word refers to the title of
a particular work.
129
Hdt. 5.71 tells the story of the massacre; Epimenides’ purification is mentioned by Plato, Leg. 1.642d
and Aristotle, Athenaion politeia 1.1, and told in detail by Plutarch (Sol. 12–14). An alternative
tradition held that Epimenides’ purification was of a plague (Diog. Laert. 1.110; Max. Tyr. 38.3
Trapp cites both stories). On the role of this episode, and Epimenides, in later Athenian cultural
memory, see Gagné (2013) 312–20.
130 131
Gagné (2013) 319, 315. Samons (1999); Federico (2001); Gagné (2013) 320.
blood crime causes pollution and conflict across generations until it is
purified through the institution of new cults. The story of Epimenides’
purification of the Alcmaeonid massacre seems only to have arisen in the
fourth century bce, perhaps as part of Athens’ renegotiation of its past in
the wake of the defeats by the Spartans and Macedonians,132 but the
Eumenides attests to the existence of this particular pattern as early as 458
bce, around the time of Empedocles’ floruit. Purification of bloodshed,
then, at this stage, can act as a conceptual model with which to describe
watershed moments in the history of a polis, not just the foundation of new
colonies. This may, at least partly, explain the authority accorded to
murderous suppliants in Martin’s metanastic topos.
The story of Empedocles’ narrator evokes this type: he is an exile who is
out to purify his bloodshed (D10=B115), through the institution of new cult
practices and the creation of a new type of community out of the city of
Acragas. Indeed, Primavesi has argued that the narrator’s story is modelled
on a further myth which in some respects conforms to this pattern: Apollo
was exiled to the mortal plane for killing the Cyclopes and forced to serve
Admetus for a year in penance, a story already attested in the Catalogue of
Women (Hes. fr. 54b–c; cf. Eur. Alc. 1–7, 3 FGrH fr. 35a).133 The narrator,
like Apollo, is a god in exile, but it would be reductive to claim that this is
the only myth that his story evokes. The scenario of D4=B112 and the
injunctions against blood-sacrifice (D28=B136; D29=B137) – injunctions
which demand, effectively, a change in fundamentally communal
institutions –134 warrant comparison with the versions of this topos that
are more obviously embedded in histories of particular poleis. This more
specific context is suggested by the title of Katharmoi: that this generic title
is used of both Empedocles and Epimenides’ poetry suggests that later
readers saw the two authors as performing analogous functions.
132
Thus Gagné (2013) 320–1, who sees the story as influenced by Aeschylus.
133
Primavesi (2008a) argues that this myth is the model for the exile of Empedocles’ narrator, partly on
the basis of Ammonius’ identification of the god of D93=B134 with Apollo. Whilst I accept that this
myth – and the broader type it represents – is certainly relevant, there is no particularly good reason
to think, with Primavesi, that the exiled narrator is identical to the god of D93=B134 or was referred
to in the poem as Apollo. There was also a tradition that Apollo’s exile was for killing the Delphic
dragon (404 FGrH fr. 5) and that after the killing he went to Tempe or Crete for purification (Paus.
2.7.7). Cf. Dougherty (1993) 31–44, who sees Apollo’s role in purification as the reason for his
association with the foundation of colonies (cf. Callim. Hymn to Apollo 55–7).
134
On the role of animal sacrifice as a means of forming bonds in groups, partly through communal
feasting, see Burkert (1983) 35–48; Detienne (1989); Parker (2011) 305–9; Bakker (2013) 36–52 (who
points out the relative rarity of meat consumption in the ancient world, a phenomenon that usually
only occurred at festivals and other special occasions). This explanation, indebted to Durkheim’s
sociological account of religion (1912), is supported by such primary evidence such as Isoc. 19.10;
Dem. 19.128, 190, 58.40; Aeschin. 3.52; Din. 2.9.
138 Empedocles
Empedocles’ poem, I suggest, is designed to effect a similar watershed to
that with which Epimenides was credited: the poem offers instructions for
purifying a pollution that affects the community but, in the process, also
presents the possibility of the creation of a new community, based on new
principles.
This point finds further support in the fact that D25=B128 provides the
model of an ideal community to which we should aspire:
οὐδέ τις ἦν κείνοισιν Ἄρης θεὸς οὐδὲ Κυδοιμός
οὐδὲ Ζεὺς βασιλεὺς οὐδὲ Κρόνος οὐδὲ Ποσειδῶν,
ἀλλὰ Κύπρις βασίλεια.< >
τὴν οἵ γ’ εὐσεβέεσσιν ἀγάλμασιν ἱλάσκοντο
γραπτοῖς τε ζώιοισι μύροισί τε δαιδαλεόδμοις 5
σμύρνης τ’ ἀκρήτου θυσίαις λιβάνου τε θυώδους,
ξανθῶν τε σπονδὰς μελίτων ῥίπτοντες ἐς οὖδας·
ταύρων δ’ ἀκρήτοισι φόνοις οὐ δεύετο βωμός,
ἀλλὰ μύσος τοῦτ’ ἔσκεν ἐν ἀνθρώποισι μέγιστον,
θυμὸν ἀπορραίσαντας ἐ<ν>έδμεναι ἠέα γυῖα. 10
Nor was Ares a god for them, nor was Uproar
Nor was Zeus a king nor Kronos nor Poseidon,
But Kypros was queen < . . . >
Her they appeased with pious images
And with painted animals and with elaborate-smelling perfume 5
And with sacrifices of unmixed myrrh and of fragrant incense,
And casting libations of yellow honey onto the floor.
And the altar was not wetted with the unmixed blood of bulls,
But this was the greatest defilement among men,
To rip out the soul and eat the noble limbs. 10
Here, the contrast between the idealised community and that of a fifth-
century Greek polis is marked by the paradoxical expression, ‘sacrifices of
unmixed myrrh and of fragrant incense’135 and by the delayed negative of
line 8, where we learn that the altar was not wetted with the unmixed blood
of bulls. What is normal for the Acragantines – sacrificial slaughter – is in
fact the greatest source of defilement (9–10). This detail inverts what seems
to have been the central act in the purification for homicide: the cleansing
of the killer’s hands with the blood of sacrificial victims.136 Although the
scenario is set in the past and may reflect the community from which the
daimon has been exiled, its idealised quality – its lack of Ares or ‘Uproar’,
and the lush opulence of the offerings to Kypris – entices us to hope that it
135
Θυσία normally refers to burnt offerings or blood sacrifices (LSJ s.v.).
136
Heraclitus D15=B5; Eur. IT 1224, 1338; Aesch. Eum. 281; Ap. Rh. 4.685–717. See Parker (1983) 370–4.
can again be realised. Comparable is Hesiod’s myth of the races, where the
plenteous land of the earlier golden race (Op. 109–20) can be recovered by
the community that adheres to Justice (cf. Op. 225–36).137 The narrator,
like Orestes, is a polluted incomer whose arrival is associated with the
institution of new cult practices, but those practices themselves are radic-
ally different from traditional Greek ritual, requiring, not purification of
blood with (sacrificial) blood, but purification through the abstinence from
bloodshed.
Although the reference to the Acragantines in D4=B112 establishes
a context within one particular polis, D25=B128 offers a model that is
more generally applicable. The stories of Orestes and Epimenides were
told as watershed moments in the history of Athens, but Empedocles has
geographically wider aspirations. His narrator, like these legendary
figures, is an exile associated with purification who is to play a role in
galvanising individuals into a unified group, but that group is not
restricted to a particular polis, nor to the Greek world, nor even to the
human species. Rather, various details suggest that the poem is designed
to induce a sense of community with all other living things, by virtue of
the fact that they are composed of the same elements as humans and are
species into which we may be (and may have been) incarnated (cf.
D13=B117). As part of this wider perspective, the narrator has not been
exiled from a particular polis but from the gods. It is not exactly clear
whether this divine community only exists at some prior time in the
cosmic cycle or whether it still exists in some location geographically
remote from humans,138 but in either case, the limits of the traditional
mythical pattern, involving exile from one polis followed by arrival and
purification at another, have been expanded so that they take on cosmic
proportions. The area to which he has been exiled seems to consist of the
entirety of the familiar, mortal world.139 Empedocles’ narrator has not
137
I follow, here, the influential interpretation of Vernant (1983) 3–32. On the significance of this
passage, see more recently Currie (2012), who further discusses (among other issues) the respects in
which Hesiod creates the impression that the golden age may yet be recreated.
138
E.g. Picot (2008a) argues that the community of the blessed is on the moon; Trépanier (2017)
147–78 provides an elaborate argument for the location being the celestial vault; Primavesi (2008a),
on the other hand, sees it as a metaphor for an earlier stage in the cycle.
139
Assuming the most widely accepted interpretation of D16=B120, ‘We came under this roof of
a cave’, that it refers to world of mortals to which the daimon has been banished (as implied by
Porphyry’s introduction to the fragment; thus, Wright [1995] 280 and Bollack [2003] 73–4; contrast
Zuntz [1971] 255). D19=B126 may describe the estrangement from the divine community as being
wrapped ‘in a foreign cloak of flesh’, although, there, both the object and the grammatically
feminine subject of the participle ‘wrapping’ (περιστέλλουσα) are unclear.
140 Empedocles
been punished according to the law of a particular polis, but according to
a universal one:
ἀλλὰ τὸ μὲν πάντων νόμιμον διά τ’ εὐρυμέδοντος
αἰθέρος ἠνεκέως τέταται διά τ’ ἀπλέτου αὐγῆς
But that which is lawful for all things
Is stretched through the wide-ruling Aether and through the boundless light
(D27a=B135)
Aristotle, who quotes this fragment, tells us that it comes in support of the
injunction against killing living things (Rhet. 1.13.1373b14–16).140 We can
infer that this is the law that the narrator himself has broken in polluting
his limbs through bloodshed (D10=B115), a law that holds sway over the
cosmic realms (the Aether, the light) through which he has passed in the
wanderings of his exile (D10=B115.9–11). Whilst sages were often painted as
lawmakers,141 Empedocles does not present his narrator as instituting new
laws but as revealing the universal laws that have always existed. He is thus
like Sophocles’ Antigone, whom Aristotle also quotes in this context, in
appealing to an objective law that overrides the mortal customs accepted by
his addressees (Cf. Soph. Ant. 456–7).
Most clearly, however, our kinship with other species is illustrated by
D29=B137, another fragment with affinities to Attic tragedy. Here, we are
presented with the terrifying scenario of family members unwittingly
sacrificing and eating one another:
μορφὴν δ’ ἀλλάξαντα πατὴρ φίλον υἱὸν ἀείρας
σφάζει ἐπευχόμενος μέγα νήπιος· οἱ δ’ ἀπορεῦνται
λισσόμενον θύοντες· ὁ δ’ αὖ νήκουστος ὁμοκλέων
σφάξας ἐν μεγάροισι κακὴν ἀλεγύνατο δαῖτα.
ὡς δ’ αὔτως πατέρ’ υἱὸς ἑλὼν καὶ μητέρα παῖδες 5
θυμὸν ἀπορραίσαντε φίλας κατὰ σάρκας ἔδουσιν.
The father, lifting up his own son who has changed shape,
Cuts his throat, with a prayer – fool that he is! The others are at a loss
While they sacrifice the suppliant; but he, deaf to the shouts,
Has cut the throat and prepared an evil meal in his house,
In the same way, a son seizes his father and children their mother, 5
And ripping out their life they devour the flesh of their dear ones.
(Trans. LM)
140
Cf. similarly Cicero Rep. 3.11.19, who also attributes to Empedocles this specifically legal claim.
141
Most famously Solon, but also Pittacus (Arist. Pol. 2.12.1274b18–20) and Chion (Diog. Laert. 1.68).
Relevant, too, is Parmenides’ later reputation as a lawgiver (Diog. Laert. 9.23; Strabo 6.1.1; Plut.
Adv. Col. 32 1126A–B).
The example of the father sacrificing his beseeching child is reminis-
cent of Agamemnon’s sacrifice of Iphigenia, whilst the ‘evil meal’
recalls the banquet of Thyestes. The gruesome imagery elicits pity
and fear. But if we consider this example literally, within
Empedocles’ cosmology, then such instances will presumably be
relatively rare. If you were to select at random a sheep for sacrifice,
you would have to be particularly unlucky to choose one that
happens to be a reincarnation of a deceased parent. Yet the instance
is emblematic of a more general kinship in Empedocles’ cosmology
between all living things: what matters is not merely that a non-
human animal might happen to be a reincarnated family member,
but that non-human animals are creatures like us, formed of the
same elements, and harbouring (at least in many cases) transmigrat-
ing daimones.142 Our kinship with any potential sacrificial victim is
elevated to a status analogous to that typically felt towards one’s
family members. The fragment thus appeals to our emotions and to
a conventional notion of kinship to illustrate our wider kinship with
all living things. The summary of the act at line 6 uses the same
language as describes the ‘greatest source of defilement’ in the ideal-
ised community of D25=B128. This tragic and horrific image of
society is thus the polar opposite of the idealised community under
Love and, in failing to obey the universal law, is more representative
of the generally carnivorous Greek world. The world we inhabit
becomes the world of Attic tragedy, a world normally confined to
the remote past or the geographically distant.143
To recapitulate: Empedocles’ narrator conforms to a traditional pattern in
Greek myth and legend of exiles, often killers, who arrive at a new location
where they achieve purification, are accorded a particular authority and often
play a role in the creation of new institutions such as laws and cult practices.
D4=B112 and D10=B115 both cohere with this pattern. This, of course, cannot
prove that the two fragments came from the same poem, but it does suggest
that they are parts of the same narrative. The contrasting tones of the two
fragments and the differences in the self-presentation of the narrator can be
142
Most commentators assume that all humans and animals have such a daimon, but note Sedley
(2007) 33–52, who argues that only certain animals (and not all humans) will enjoy reincarnation.
I discuss this fragment, and my understanding of the Empedoclean daimon, in Mackenzie
(forthcoming).
143
That Attic tragedy typically represented such a remote world is demonstrated by the well-known
case of Phrynichus’ Fall of Miletus, a play that was banned from being re-performed, and resulted in
its author being fined, because it reminded audiences ‘of their familiar misfortunes’ (οἰκεῖα κακά,
Hdt. 6.21.10). In other words, it was banned partly because it did not depict such a remote world.
142 Empedocles
read as rhetorically effective: an initial captatio benevolentiae with praise of the
polis assumes a new significance in light of the narrator’s status as an exile from
the gods; the localised perspective of the address to the Acragantines expands
to encompass the cosmos as a whole. From this wider, cosmic perspective, we
learn of an ideal society, from which the narrator has been exiled, that is the
polar opposite of the world we inhabit. The text offers some guidance on how
to achieve such an ideal society in its dietary restrictions. In this respect, the
narrator is an aspiring reformer: he wants us to change our meat-eating ways
and worship Love alone with vegetarian sacrifices. Such a process, it seems,
would amount to a ‘purification’, since bloodshed itself is the cause of
pollution. This sort of reformation would further conform to the traditional
pattern identified, but, since this is a formally didactic text,144 the pattern is
not yet complete: the narrator has not yet achieved the purification of his
audience through new institutions, because such a process is dependent on the
audience’s behaviour. A traditional pattern is evoked but not its closure: the
end of the story that we find elsewhere – the establishment of new laws and
cults – is omitted, and it is up to the audience themselves to determine how
this story will end. The lack of closure to this pattern, then, acts as a prompt
for an audience to complete this process of purification.
b Initiation, D257=B110 and the Education of the Singular Addressee

Empedocles evokes the ‘didactic plot’ of mystic initiation in a number of
respects. As commentators have long pointed out, the doctrines of reincar-
nation and vegetarianism are paralleled in Pythagoreanism and Orphism,
groups that practised initiatory rituals.145 This connection is strengthened
by the fact that the unnamed sage of D38=B129 is most plausibly identified as
Pythagoras.146 The fragments distinguish between speech that is themis for
an audience to hear and that which is not (D44=B3.4; D54=B9.5), recalling
the distinction made in the formulaic Orphic expression between those for
144
That is, it is ‘purporting to be intended to instruct’, following the terminology of Heath (1985) 253.
145
On Empedocles’ connection with these groups, note already Kern (1888), but especially Riedweg (1995)
and more recently Rodriguez (2005). On initiation in early Orphism, see e.g. Parker (1995) 484–6 and
Bremmer (2014) 55–80; for Pythagoreanism, see Burkert (1972) 120–65 (on the view that, in the earliest
sources, the Pythagorean and Orphic doctrines are connected or even identical, e.g. Hdt. 2.81); Kahn
(2001) 8–9; and similarly, Riedweg (2008) 98–105 (although he does not use the term ‘initiation’, he
regards the Pythagoreans as a ‘sect’ requiring members to take a vow of silence, cf. Iambl. VP 72).
146
That this fragment is a reference to Pythagoras is claimed by Porphyry VP 30 and Iamblichus VP 67
(both reliant on the same source) and by Timaeus as reported by Diogenes Laertius 8.54. Diogenes
adds that some say that Parmenides is referred to here. As Trépanier (2004) 124–5 points out,
Pythagoras is the more likely subject, as he was more prominently associated with the theory of
reincarnation (e.g. at Xenophanes D64=B7).
whom it is themis to hear and those who are uninitiated.147 D8=B132 takes
the form of a makarismos (‘blessed is he who . . .’) celebrating the man ‘who
has attained a wealth of divine thoughts’, a type of expression especially
associated with initiates into Eleusinian and other mysteries.148 The acquisi-
tion of knowledge in Empedocles’ cosmos is thus presented in terms
appropriate for mystery-initiates; initiates themselves were often referred to
as those who ‘know’.149 Furthermore, the purificatory function of the poem,
implied by certain internal details (D44=B3.2; D257=B110.2) and by the title
Katharmoi,150 evokes this sort of initiation: the Eleusinian mysteries required
initiates to be pure of bloodshed and, in addition, to undergo a ritual
cleansing before the procession to the sanctuary of Demeter and
Persephone (in 339 bce one unlucky prospective initiate, while cleansing
himself and his sacrificial piglet in the sea, was eaten by a shark).151 Possessors
of some of the ‘Orphic’ gold tablets are described as the ‘pure from the pure’
(katharos, OF 488.1; 489.1; 490.1; 491.1), whilst initiation into Orphic mystery
cults is reported as the expiation of a previous crime.152
The distinction of a singular addressee further conforms to this pattern
of initiation. In the most detailed study of Empedocles’ addressees,
Obbink argues that the author exploits ‘a common archaic phenotype of
a master (ritual-) craftsman indoctrinating his (occasionally dull-witted)
apprentice’, of which Hesiod’s relationship with Perses in the Works and
Days is the best preserved Greek example.153 Thus, the student Pausanias is
to learn the extraordinary skills of controlling the weather and bringing
back the force of the dead (D43=B111) from the master-narrator. For
Obbink, the narrator’s use of insistent imperatives and his appeal to ‘the
simplest phenomenal entities, the most obvious features of the physical
world’ is an indication of the addressee’s dull-wittedness.154 However, the
phenomena appealed to – such as the wider course of the cosmic cycle
147
See Section 3c below.
148
Cf. esp. hDem. 480 with Richardson (1974) ad loc. Other examples in an initiatory context include
Pind. fr. 137a; Soph. fr. 837.; Eur. Ba. 72ff. Apul. M. 11.16.2ff. For a comprehensive list of examples,
see Norden (1913) 99–100, 100 n. 1.
149
See Chapter 2 n. 86 above.
150
Cf. the examples cited at n. 128 above. Note in particular that Plato Resp. 2.364e mentions itinerant
priests who ‘persuade not only individuals but even cities’ that they can achieve ‘release and
purifications (katharmoi)’ through performing certain rituals.
151
See Bremmer (2014) 4–5. For the requirement to be pure of bloodshed, note Ar. Ran. 354–71 with
the scholion on 369; Isoc. 4.157. For the ritual cleansing, note Ephorus FGrH 70 F 80; for the shark
story, Aeschin. 3.130 and schol. ad loc.; Plut. Phoc. 28.3.
152
Plato Resp. 2.364b2–365a3; Crat. 400c. Cf. Pind. fr. 133, Ol. 2.58. This crime may be explained by the
myth of Dionysus Zagreus, on which see Section 4c below.
153
Obbink (1993) 52. 154 Obbink (1993) 83.
144 Empedocles
(D73), the deeper origins of humans, animals and gods (D77A=B21), or the
presence of cosmic Love (D73.252) – are far from obvious.
Rather than seeing Pausanias as a dull-witted apprentice who is being
taught a particular craft, I would suggest that he undergoes a more trans-
formative experience than the apprentice-image implies. A notable point of
contrast with Hesiod’s addresses to Perses is that Empedocles uses, in addition
to commands (in the form of imperatives or imperatival infinites, e.g. D41=B1,
D258=B5, D28=B136, D74.245 etc.) and conditional clauses (D257=B110),
future indicatives: ‘You shall learn’ (D42=B2.9); ‘you shall cease the force of
the tireless winds . . . you will make for mortals an opportune dryness out of
a dark storm . . . and you shall lead from Hades the force of a dead man’
(D43=B111). This grammatical form signifies a more confident prediction of
the addressee’s fate than is found in the Works and Days and is reminiscent
both of prophecy – a context we might especially recall given the narrator’s
reputation as a source of oracles (D4=B112.10) – and the predictions made of
the addressee in the ‘Orphic’ gold tablets.155 This suggests a less mundane
scenario than that of a master addressing an apprentice. The abilities to be
acquired in D43=B111 equate to the assumption of divine status that the ideal
student will achieve (D39=B146). Pausanias’ education thus amounts to an
acquisition of godhood rather than merely of certain skills, a scenario that
recalls the transformative experiences of mystic initiates, which, at least in
some cases, was conceived of as an apotheosis.156
The narrator distinguishes Pausanias as the addressee at D41=B1, which,
according to Diogenes Laertius, ‘prefaced the material On Nature’.157 This
testimony, along with the evidence that D4=B112 opened the poem and
D10=B115 occurred a little later, can be taken to suggest that, on the single-
poem hypothesis, after an initial address to the Acragantines and the story
of the narrator’s exile, the narrator shifts to a singular addressee before
delivering the details of the cosmology.158 If this is indeed the structure,
then, within the fiction of the poem, it is a singular, distinguished pupil
155
For the use of future indicatives in prophecy cf. e.g. Od. 11.118; Pind. Isthm. 51–2 (where the eagle ‘spoke
like a prophet’, εἶπέν τε φωνήσαις ἅτε μάντις ἀνήρ); Soph. Ant. 998, 1067; OT 453, 456, 457; OC 1533.
Second-person singular future indicatives in some of the Orphic gold tablets: OF 474, 475, 476, 477.
156
See Tor (2017) 270–3, who criticises earlier deniers of this view. Note especially the gold tablets, OF
488.9 and 487.4.
157
Diog. Laert. 8.60 ὧι δὴ καὶ τὰ Περὶ φύσεως προσπεφώνηκεν οὕτως. As Trépanier (2004) 26–9
argues, this need not be taken as the title of a single poem but may only be a particular section.
158
Obbink (1993) 70–6 argues, similarly, that the poem began with an initial hymn to a deity followed
by the Muse-invocations and then the address to Pausanias, but there is no evidence for such
a hymn (pace Sedley [1998] 1–34, who argues that we can reconstruct Empedocles’ proem from
Lucretius, but this approach seems methodologically unsound: Lucretius creatively adapts rather
than slavishly follows his models, so that we cannot reconstruct the structure of the latter from the
who is granted the divine narrator’s god’s-eye view of the cosmos. This
individual is most clearly distinguished from the masses at D42=B2:
στεινωποὶ μὲν γὰρ παλάμαι κατὰ γυῖα κέχυνται·
πολλὰ δὲ δείλ’ ἔμπαια, τά τ’ ἀμβλύνουσι μέριμνας.
παῦρον δ’ ἐν ζωῆισι βίου μέρος ἀθρήσαντες
ὠκύμοροι καπνοῖο δίκην ἀρθέντες ἀπέπταν
αὐτὸ μόνον πεισθέντες, ὅτωι προσέκυρσεν ἕκαστος 5
πάντοσ’ ἐλαυνόμενοι, τὸ δ’ ὅλον <πᾶς> εὔχεται εὑρεῖν·
οὕτως οὔτ’ ἐπιδερκτὰ τάδ’ ἀνδράσιν οὔτ’ ἐπακουστά
οὔτε νόῳ περιληπτά. σὺ <δ’> οὖν, ἐπεὶ ὧδ’ ἐλιάσθης,
πεύσεαι· οὐ πλεῖόν γε159 βροτείη μῆτις ὄρωρεν.
For narrow resources are spread out across their limbs;
And many wretched things are in their way, and things which blunt their wits.
And having seen a small part of life during their lifetimes
Swift-fated, they fly off, lifted up in the manner of smoke
Persuaded of that thing alone, whatever it is that each one has encountered 5
Driven in every direction, but everyone claims to have found the whole;
Thus, these things are neither seen nor heard by men
Nor grasped with the mind. But you, since you have deviated here,
Shall learn. Mortal intelligence has never soared further.
The image is of one who, perhaps by chance, has ‘deviated’160 from the
normal run of mortals and will learn more than mortal wit can generally
reach. The rest are collectively indistinguishable, flitting about like smoke
in their ephemeral concerns. They have ‘seen’ only a small part of life and
this limited vision will contrast with the broad, divine perspective that
Pausanias will enjoy, for which visual language is repeatedly used
(D44=B3.9; D77a=B21.1; D257=B110.2; D74.326=B76.3; D73.292–3, 300).
Here, then, Pausanias, pace Obbink, is not a dull-witted foil; rather, an
audience is encouraged to identify with this distinguished individual,
against the ignorant masses. Such a distinction, consisting in being granted
former). Trépanier (2004) 31–72 argues in greater detail for the structure of the poem that I sketch
above.
159
Along with Bollack and more recent editors (e.g. Mansfeld and Primavesi [2012]; LM), I see no
reason to emend here. DK adopt Karsten’s emendation οὐ πλέον ἠέ, ‘no more than’, but it would
be strange to place such a limitation on what Pausanias shall learn, especially since he will attain
superhuman skills (D43=B111). Wright (1995) obelises, but expresses a preference for the more
modest sort of sense provided by Karsten’s emendation on the grounds that it coheres with Sextus’
introduction to the fragment, in which he mentions the limitations of ‘mortal reason’ (anthrōpinos
logos) – but Sextus could easily be referring to lines 1–6 of the fragment.
160
διάζομαι often seems to refer to a departure from a particular trajectory. E.g. Il. 15.520 (shrinking
beneath an attack); Il. 23.879 (the drooping of a bird’s wings, a sign that it has died); Od. 5.462
(Odysseus deviating from the path of the river).
146 Empedocles
access to a vision inaccessible to most, further recalls accounts of mystery-
initiates, as we saw in the previous chapter with respect to Parmenides (see
Chapter 2 Section 4f above).
Most conspicuously, however, initiatory language occurs at
D257=B110:
εἰ γάρ κέν σφ’ ἀδινῇσιν ὑπὸ πραπίδεσσιν ἐρείσας
εὐμενέως καθαρῇσιν ἐποπτεύσῃς μελέτῃσιν,
ταῦτά τέ σοι μάλα πάντα δι’ αἰῶνος παρέσονται,
ἄλλα τε πόλλ’ ἀπὸ τῶνδε κτήσεαι· αὐτὰ γὰρ αὔξει
ταῦτ’ εἰς ἦθος ἕκαστον, ὅπῃ φύσις ἐστὶν ἑκάστῳ. 5
εἰ δὲ σύ γ’ ἀλλοίων ἐπορέξεαι οἷα κατ’ ἄνδρας
μυρία δειλὰ πέλονται ἅ τ’ ἀμβλύνουσι μερίμνας,
ἦ σ’ ἄφαρ ἐκλείψουσι περιπλομένοιο χρόνοιο
σφῶν αὐτῶν ποθέοντα φίλην ἐπὶ γένναν ἱκέσθαι·
πάντα γὰρ ἴσθι φρόνησιν ἔχειν καὶ νώματος αἶσαν. 10
For if, planting them under your firm organs of thought
You gaze on them benevolently with pure efforts,
These things will all be present to you throughout your lifetime,
And you will obtain many other good things from them. For these things
Will increase themselves in regard to each character, how each thing has its
nature 5
But if you reach out for the other kinds of things, the sort of things which, among
men
Are countless and wretched and which blunt their wits,
Certainly, they will suddenly leave you in the course of time
Longing to come to the dear race of their own
For know that all things have intellect and a share in thought. 10
The verb ἐποπτεύω can mean ‘become an ἐπόπτης, be admitted to the
highest grade at the mysteries’ (LSJ s.v. II.) and this sense is activated by the
mention of ‘pure efforts’, since the word for ‘pure’, katharos, often refers to
ritual purity.161 The ‘blunting’ of wits (line 7) recalls the distinction of the
addressee from the masses at D42=B2.2. The process of education here,
then, is modelled on that of initiation, with the narrator performing the
role as mystagogue.
For some scholars, these details indicate that the poem is designed
to perform a role in an actual process of initiation into an esoteric
161
See Willi (2008) 236–7. See, e.g. Pl. Phd. 250c for ἐποπτεύω as ‘viewing’ in the mysteries (as one
who has become an ἐπόπτης, the highest grade of initiation) and for καθαρός as ritual purity in
a mystic context. εὐμενής is often an epithet of gods (h.Hom. 22.7, Pind. Pyth. 2.25, Aesch. Supp.
686) and in Attic is not normally used of human beings but is applied to a goddess in a prayer at Ar.
Lys. 204 (thus, Willi [2003] 28–9).
group.162 The use of a theogonical poem to this end would be paralleled
in the Derveni papyrus, where the commentator seems to think that the
right process of initiation depends upon the correct interpretation of the
poem he quotes.163 The context in which the Strasbourg papyrus was
preserved further supports the notion the poem was (eventually) used in
a ritualistic context: the papyrus formed part of a funerary wreath for
a corpse, a detail that may indicate that the deceased hoped to bear the
master’s instructions into the next life.164 However, the internal evidence
of the poem scarcely supports the notion that it was composed for an
esoteric context in this manner. At no point does the narrator mention
initiatory rites (τελεταί or μυστήρια) or use the verb for the process of
initiation (μυέω). As we have seen, the relationship depicted between
teacher and ‘initiate’ is unlikely to represent the actual circumstances of
performance: the poem was composed not only for the ears of Pausanias,
if indeed he was ever a real person. Any member of the audience could
theoretically choose to play the role of Pausanias and distinguish them-
selves from the masses by refraining from meat-eating and receiving the
wisdom of the poem in the correct way.
If we are to think of the education of Empedocles’ student as an
initiation (as the initiatory language of D257=B110 encourages us to do),
then this is a new kind of initiation, one that differs fundamentally from
the sort of rituals required for membership of the Eleusinian mysteries or
the Orphic or Pythagorean mystery cults. A closer analysis of D257=B110
will suggest that, rather than offering an actual initiation into a particular
cult, the fragment refigures the process of initiation as one of understand-
ing the nature of the cosmos in terms of Empedocles’ elemental cosmology.
According to the most influential interpretation, that of A. A. Long,
Empedocles here conceives of his teachings in physical terms, as actually
composed of the same elements that they depict. Thus, when Pausanias
is to plant ‘them’ in his ‘organs of thought’, this physical language is
meant literally.165 Long sees this hypothesis as making the best sense of
lines 4–5, glossing them as, ‘the elements (of my teaching) cause each
162
Kingsley (1995) 367–8, (2002) 352–3. Similarly, on Empedocles as an initiator (but without reference
to this particular fragment), Gemelli Marciano (2001), (2005b).
163
At least, according to some scholars of this highly problematic text. See e.g. Betegh (2004) 349–72;
Bernabé (2014). This view is lent support by Adeimantus’ description of wandering initiators who
use Orphic texts at Pl. Resp. 2. 364b2–365a3.
164
As Most (1999) 356 notes; also argued by Clay (2012).
165
Long (1966) 270, followed by (e.g.) Wright (1995) 258–9 and Wolfsdorf (2009) 50–1. Similarly,
Schwabl (1956) and Bollack (1965–9) iii.2 577 take the antecedent of sphe as the elements and the six
permanent entities respectively.
148 Empedocles
thing to grow in its own way, according to its nature’. Whilst this
fragment most plausibly concerns how Pausanias is to receive the narra-
tor’s teachings, I find it preferable to take the physical language figura-
tively. A number of considerations count against the notion that
Empedocles’ teachings are literally conceived of as being composed of
the four elements in this way. It may seem prima facie implausible that
the process of education should be thought of as simply physically
increasing the elements within oneself, for there is no reason to think
that, for Empedocles, an increase in mass accompanies an increase in
intelligence; rather, what seems to be more important is the proportion
of the elements and the extent to which they are blended.166 Instead, it is
more economical to suppose that he understood the teachings to be
instances of sound and so composed of air, since Aëtius
(4.16.1=D227=A93) tells us that he explained hearing as occurring
when ‘breath’ falls against the cartilaginous part of the ear.167 The use
of physical language to describe speech or teachings is traditional.
Particularly comparable are hexameter expressions where addressees are
told to ‘cast’ instructions into their thoughts or spirit.168 Elsewhere,
Empedocles also presents his utterances in physical terms (D44=B3,
D47=B4.3, D75=B35.1–2), but in ways which seem incompatible with
the view that they are formed out of the four elements. In D44=B3 and
D75=B35.1–2, his poetry is presented as water, in a manner which may
evoke specifically the water used in purificatory rituals, but no one would
suggest that his speech is somehow literally conceived of as watery. Here,
the physical language is similarly figurative.
Rather than denoting physical increase within the body of Pausanias,
I propose that lines 4–5 illustrate the increase in his understanding once he
has grasped the basic principles of the elemental theory. The lines can thus be
understood roughly as follows: ‘these (teachings) will increase themselves
(αὐτὰ γὰρ αὔξει ταῦτ’) with regard to each character (εἰς ἦθος ἕκαστον),
how the nature of each of them is’. εἰς in the sense of ‘with regard to’ is well
166
As mentioned above, we think especially with the blood (D24=B105), a substance comprising
a harmonious blend of the four elements that comes about through the agency of Love (D190=B98;
cf. A78; A85). Kamtekar (2009) 223–5 raises further objections to Long’s view.
167
A process discussed by Long (1966) 265. Moreover D73.291–3 indicates that the student’s education
will not depend on absorbing effluences of teachings.
168
E.g. σὺ δ’ ἐνὶ φρεσὶ βάλλεο σῇσι (Il. 1.297, 4.39, 5.259 etc.; Od. 11.454, 16.281, 16.299 etc.; Hes. Op.
107, 274); σὺ δὲ ταῦτα τεῷ ἐνικάτθεο θυμῷ (Hes. Op. 27; cf. Od. 23.233). Wright (1995) 258–9 also
notes the Homeric description of words as ‘winged’ (Il. 1.201, 2.7, 4.69 etc.), as going past the
‘barrier of the teeth’ (Il. 4.350, 14.83 etc.) and put by the listener into the thumos or spirit (Od. 1.361,
21.355).
attested in both verse and prose texts of the fifth century (LSJ s.v. IV). I take
the phrase ὅπῃ φύσις ἐστὶν ἑκάστωι, ‘how the nature of each of them is’ as an
epexegetical indirect question that denotes the information that the teachings
will explain once they have ‘increased’.169 The teachings will expand in the
sense that, if Pausanias has grasped the basic idea that all things are composed
of the four elements, he will be able to explain the ‘character’ and ‘nature’ of
each thing. This will involve identifying the elements that are present in
particular compounds, just as, in D74, the narrator promises that Pausanias
will ‘see’ the ‘earth’ (ὄψει χθόνα) in the shells and horns of certain animals.
Whilst Pausanias requires the physical presence of the elements within him to
perform such cognitive tasks (as D207=B109 indicates), there is nothing to
suggest that the elements are literally conferred by the narrator’s utterance.
Instead, the presence of the elements in our bodies seems to arise, as we might
expect, through nutrition (Simpl. in Phys. 372.5).
On this view, the fragment proposes that a grasp of the basic
components of Empedocles’ cosmology will help us to understand
anything we might encounter. By contrast, most mortals vainly
attempt to understand the ‘countless things’ they perceive without
the aid of the elemental theory. The final line of the preserved text is
probably to be explained as reassuring Pausanias that the ideas will stay
with him regardless of whatever incarnation he will subsequently
experience:170 even if he is a non-human animal, he will still have
a share in thought and so these ideas will be relevant. The situation is,
then, similar to that at the end of book 1 of the De rerum natura,
a passage which may imitate these very lines:
Haec sic pernosces parva perductus opella;
namque alid ex alio clarescet nec tibi caeca 1115
nox iter eripiet, quin ultima naturai
pervideas: ita res accendent lumina rebus.
Thus you will learn these things when led by a little effort;
for one thing will become clear from another, nor will dark 1115
night break off your journey to prevent you from
seeing the limits of nature: thus, things kindle fires for things
169
For similar indirect interrogative uses of ὅπῃ in hexameter, note Od. 8.573, 9.279, 9.457, 10.190–2.
170
Hippolytus Haer. 7.29.25 takes it as claiming that there is, in addition to Love and Strife, a νοητὴ
τρίτη τις δύναμις. However, such a ‘third power’ is nowhere else attested for Empedocles, and
Mansfeld (1992) 221–6 convincingly argues that Hippolytus attributes such a view to him to find
precedent for his contemporary Prepon’s theory of a third and ‘just’ principle intermediary between
good and evil (Haer. 7.31).
150 Empedocles
As Kranz pointed out long ago, the repetitions (alid ex alio . . . res rebus)
seem stylistically similar to the Empedocles passage (ταῦτά . . . αὐτά . . .
ταῦτά . . . ἕκαστον . . . ἑκάστωι) and direct influence is possible.171 In the
Lucretius passage, the idea of res will enable the student to understand
further res in the universe. A similar process occurs in the Empedocles
passage, where the basic idea of the four elements will enable the student to
identify instances of the elements in the world at large (cf. D207=B109).
A further consideration in favour of this interpretation is the fact that it
coheres with some of the fragment’s imagery. As Petrovic and Petrovic
suggest, the terms for ‘planting’ (ἐρείσας) and ‘growing’ (αὔξει) have
vegetal connotations that recall the reference to the elements as rhizomata
or ‘roots’ (D57=B6).172 Whilst this imagery might be taken to support
Long’s view that the teachings are literally composed of the four elements,
I would suggest instead that it reveals how the process by which we
understand the cosmos is analogous to the composition of the cosmos as
a whole. For just as complex phenomena are constituted by the four basic
elements, so our understanding of those phenomena is built up from the
basic idea of the six permanent entities. Such an analogy between our
understanding of the world, and the world at large, also underlies
the second-person singular command of D47=B4.3:
γνῶθι διατμηθέντος ἐνὶ σπλάγχοισι λόγοιο.
Know this in your innards when the account has been cut up.
It is as if the account were itself a compound that Pausanias is to dissect
into its constituent elements so that he can reassemble them into a complex
picture of the world.
This meta-didactic comment, addressed to the singular addressee, has,
I suggest, been misunderstood. Kamtekar takes ‘cut up’ to be a metaphor for
analytical reasoning, which, she tentatively proposes, is the result of Strife in
Empedocles’ cosmology.173 Whilst her reading is supported by the verb
διατμηθέντος, ‘cut up’, which elsewhere refers to the activity of Strife
(D73.305=B20.4), it misses a crucial irony in the line. For this is a further
instance of the reappropriation of religious imagery in Empedocles’ poem. The
171
Kranz (1944) 77, 104. On Lucretius’ reception of Empedocles more generally, see Sedley (1998) 1–34
and especially Garani (2007).
172
Petrovic and Petrovic (2016), 98–100, although I have not been able to find an instance where
ἐρείδειν means ‘to plant a seed’ or ‘crop’.
173
Kamtekar (2009) 224–5, followed by LM ad loc. and Gheerbrant (2017) 177–9, who identifies the
wordplay I discuss here but does not explore its implications in the context of Empedocles’
injunctions against animal sacrifice.
language of ‘cutting up’ and ‘innards’ evokes animal sacrifice. The σπλάγχνα
were especially ‘the heart, lungs, liver, kidneys, which in sacrifices were reserved
to be eaten by the sacrificers at the beginning of their feast’ (LSJ s.v. I, citing Il.
1.464; Od. 3.9; Ar. Pax 1105; Eq. 410 etc.).174 We have, then, a case of
‘interaction’ in the imagery: the language, in the metaphor, of ‘cutting up’,
should properly be used of an item outside of the metaphor, the ‘innards’.175
Furthermore, the σπλάγχνα were used in the most common form of proph-
ecy during Empedocles’ time, extispicy, the inspection of the sacrificial victim’s
entrails, especially the liver, by the seer (Aesch. PV 493; Eur. Supp. 212; El.
828,838; Aeschin. 3.160).176 Empedocles’ use of the word σπλάγχοισι, in the
context of ‘knowing’ something and spoken by a speaker who has appeared as
a mantis or ‘prophet’ (D4=B112.10), recalls this particular practice. The point is
precisely that the speaker’s prophetic utterance does not derive from the killing
of an animal and the dismantling of its entrails; rather, what is to be ‘cut up’ is
the logos itself, which is then to be placed within the ‘entrails’ of Pausanias, the
seat of his thought (cf. D24=B105.3). Far from indicating the agency of Strife,
the language hints at the contrast between the purity of the speaker’s mantic
utterance and the Strife-inspired pollution caused by animal sacrifice.
D47=B4.3 thus coheres with the instruction in D257=B110.2 that we should
gaze upon the teachings with ‘pure efforts’.
The ‘initiation’ in D257=B110, then, refers to a process by which the student
will distinguish himself from the common run of mortals and attain a superior
understanding of the world through grasping the basic principles of
Empedocles’ teaching. Although this process shares certain similarities with
initiation as it is attested for mystery cults – access to knowledge inaccessible to
most people; the attainment of a quasi-divine status – it is grounded in
Empedocles’ distinctive cosmology. As well as illustrating Empedocles’ epis-
temology, the initiatory imagery entices us to become a member of an elite
group, but it is an elite group defined primarily in terms of an understanding of
the cosmos that would be accessible to any recipient of the text. The initiatory
vocabulary of D257=B110, rather than indicating an actual esoteric-religious
context for the fragment, is better seen as contributing to the wider rhetorical
purpose of the poem, in encouraging the audience to accept Empedocles’
cosmology.
174
See also Gould (2001) 218–21 for discussion of this element of animal sacrifice.
175
For this sort of ‘interaction’ throughout Archaic Greek poetry, see Silk (1974). A good example is
Pind. fr. 75.5 ἰοδέτων λάχετε στεφάνων τᾶν τ’ ἐαριδρόπων ἀοιδᾶν, ‘receive violet-bound garlands
and songs that are plucked in spring’, where really it is the violets that should be ‘plucked’.
176
Flower (2008) 25.
152 Empedocles
c The Two Narratives Combined
Two distinct narratives, then, are identifiable in the fragments. The story
of the blood-polluted exile who arrives to play an important role in the
history of a particular polis is apparently addressed to a plural audience. The
‘didactic plot’ of mystic initiation involves a select, singular student. The
former structures the story of the narrator, whilst the latter structures the
story of the internal addressee, Pausanias, and implicitly that of the
individual audience member. But the two narratives are related, primarily
in the notion of purification: blood-exiles are often depicted as oikists,
colony-founders, precisely because the foundation of a new polis equates to
an act of purification for a past crime. Initiation, too, is closely associated
with purification, in at least two respects: Eleusinian initiates had to
undergo purification in the form of a ritual cleansing before the procession
to Eleusis, whilst in other cases, the process of initiation itself was con-
ceived of as a means of purification.
There is also a more specific connection between the daimon’s purification
and that involved in mystic initiation. In some cases, initiation was con-
ceived as a purification for an earlier crime. In a well-known passage, Plato’s
Adeimantus complains of ‘begging priests and prophets’ (ἀγύρται and
μάντεις) who go to the rich and claim that through ‘rituals and incantations’
(θυσίαις τε καὶ ἐπῳδαῖς) they can provide ‘purifications’ (καθαρμοί) for any
past crimes, including crimes committed by the deceased. They make use of
books ascribed to Musaeus or Orpheus and call these rituals ‘initiations’
(τελεταί; Resp. 2.364b2–365a3). It is possible that one such itinerant initiator
was the author of the Derveni papyrus, who makes use of an Orphic
theogony and criticises rival ‘mages’ who ‘perform sacrifices as if paying
a penalty’ (Col. 6.5; 20).177 Later sources would explain initiation into
Orphic mystery cults as an expiation for a primordial blood-pollution in
terms of the myth of Dionysus Zagreus. In this myth, Dionysus is the son
of Zeus and Persephone/Kore (rather than Semele). He is murdered,
dismembered and eaten by the Titans, whom Zeus then destroys with the
thunderbolt, and humans are formed from the soot that remains.
A second Dionysus is then formed from the heart. Consequently,
humans have both a guilty, Titanic element and a divine one (from the
digested chunks of Dionysus). Initiation is therefore explained as
a purification of the original crime of the Titans.178 Pindar may have
177
On this connection and the identity of the commentator, see, recently, Piano (2016) 277–308.
178
Olymp. In Phd. 1.3=OF 304 i; 313 ii; 318 iii; 320 i. See West (1983) 140–75 for the occurrence of the
myth in Orphic theogonies. Earlier sources connect the myth to the mysteries (e.g. Diod. Sic.
been familiar with this explanation for certain kinds of initiation, as fr. 133
attests:
οἷσι δὲ Φερσεφόνα ποινὰν παλαιοῦ πένθεος
δέξεται, ἐς τὸν ὕπερθεν ἅλιον κείνων ἐνάτῳ ἔτεϊ
ἀνδιδοῖ ψυχὰς πάλιν, ἐκ τᾶν βασιλῆες ἀγαυοί
καὶ σθένει κραιπνοὶ σοφίᾳ τε μέγιστοι
ἄνδρες αὔξοντ’·ἐς δὲ τὸν λοιπὸν χρόνον ἥροες ἁγνοὶ πρὸς ἀνθρώπων
καλέονται 5
But for those from whom Persephone shall receive
Penalty for her ancient grief, in the ninth year, into the upper sunlight
She will return their souls again and from them proud kings
And those swift in strength and greatest in wisdom
Arise; and for the rest of time they are called sacred heroes by men 5
(trans. Race [1997], adapted)
A select group thus enjoy benefits after reincarnation through paying the
penalty for Persephone’s ‘ancient grief’, an expression which probably
refers to her grief at the murder of her son, Dionysus, by the Titans. As
commentators have observed, the expression mirrors Empedocles
D39=B146, ‘at the end, they are prophets and singers and doctors / and
chieftains / from whence gods sprout up, greatest in honours’.179 Pindar
operated in a similar cultural context to Empedocles: he was his older
contemporary and composed epinicians for Sicilian victors, including the
Acragantine brothers Theron (Ol. 2, 3) and Xenokrates (Pyth. 6; Isthm. 2).
These examples of initiation as the expiation of a past crime indicate
a specific point of overlap between the narrative of Pausanias’ initiation
and the story of the daimon: like the daimon’s cycle, Pausanias’ initiation
involves the purification of a past crime. Certainly, if Pausanias is a meat-
eater, he will be in need of such purification according to the principles of
Empedocles’ cosmology.
3.62.8ff.; the earliest source that certainly mentions the myth is Callim. fr. 643). Walter Burkert has
argued that Pind. fr. 133.1 and Hdt. 2.62, 132, 179 allude to it ([1983] 225, [1987a] 73, 155 n. 38).
Certain ‘Orphic’ gold tablets (OF 488, 489, 490) may also allude to the myth (thus, Bernabé and
Jiménez San Cristóbal [2008] 40–4, 105–9). Edmonds (2013) 296–391 approaches the whole topic
with a salutary scepticism, pointing to the fact that the dismemberment of Dionysus, the punish-
ment of the Titans, the creation of humans, and human inherited guilt are four components that
are never found together in any extant source. However, absence of evidence is not evidence of
absence, and I find highly plausible the hypothesis that the myth combined these four components
from the time of Pindar.
179
E.g. Wright (1995) 291; Riedweg (1995) 44–5; Rodriguez (2005) 65–7. On the eschatological
background to the Pindaric fragment, see Lloyd-Jones (1985).
154 Empedocles
I suggest, then, that Empedocles’ initiatory language hints that the
student will achieve a purification of the crime of bloodshed. The story
of the daimon is thus an example demonstrating that we, too, are in need of
purification. This process finds explanation in Empedocles’ cosmology,
where ritual ‘purification’ is rationalised as the physical purification of Love
from the stain of Strife. Thus, in D10=B115, the guilty daimon ‘stains’ or
‘pollutes’ (μιήνηι)180 his ‘dear limbs’ (φίλα γυῖα)’ (line 3) through commit-
ting an act of bloodshed, an action that equates to ‘trusting in mad Strife’
(νείκεϊ μαινομένωι πίσυνος, line 14).181 On the other hand, we perform our
most effective thinking with the blood (D24=B105), a substance compris-
ing a harmonious blend of the four elements that can only come about
through the agency of Love (D190=B98; cf. D189<A78).182 Accordingly,
the sage of D91=B27a (if the line is to be attributed to Empedocles)183 lacks
strife (stasis, ‘civic strife’ and dēris, ‘warfare’) within his limbs. An individ-
ual’s purification, in Empedocles’ cosmos, thus has a cognitive dimension:
purified Love is responsible for the most effective thought and – if we take
the student’s ‘initiation’ to involve a purification – a deeper understanding
of the cosmos.
Whilst initiation is a practice that stereotypically focusses on the indi-
vidual and the distinction of that individual from the masses, in this case,
the purification of the individual serves as a synecdoche for the purification
of the community as a whole: one purpose of the text is to persuade its
audience to relinquish animal sacrifice and worship Love, effectively estab-
lishing a new community; it will achieve this end by ‘initiating’ its individ-
ual recipients, leading them to recognise the common origins of all living
things. Expiation of guilt, the acquisition of knowledge and the transform-
ation of society’s institutions are commingled in the figuration of the
response to the text as one of initiation and purification.
The Muse-invocation at D44=B3 seems programmatic of these
functions:
ἀλλὰ θεοὶ τῶν μὲν μανίην ἀποτρέψατε γλώσσης,
ἐκ δ’ ὁσίων στομάτων καθαρὴν ὀχετεύσατε πηγήν
180
Following, here, Stephanus’ universally accepted emendation for the manuscripts’ incomprehen-
sible μιν.
181
The connection is even more explicit if we accept Diels’ supplement for the lacuna at the start of line
4, νείκεΐ, so that the daimon ‘stains his dear limbs in Strife’.
182
On these fragments as evidence for the principle that, the greater the influence of Love, the more
effective the thinking, note Kahn (1960) 15–16; Long (1966) 267–7; Broadie (1999) 219–20; Laks
(1999) 267.
183
The line is transmitted anonymously by Plut. Princ. phil. 777C, and was attributed to Empedocles
by Wilamowitz (1902) 326.
καὶ σέ, πολυμνήστη λευκώλενε παρθένε Μοῦσα,
ἄντομαι, ὧν θέμις ἐστὶν ἐφημερίοισιν ἀκούειν,
πέμπε παρ’ Εὐσεβίης ἐλάουσ’ εὐήνιον ἅρμα. (D44=B3.1–5)184
But, gods, turn away their madness from my tongue

And channel a pure stream from holy mouths,
And you, much-wooed, white-armed maiden,
I entreat, the things which it is right for ephemerals to hear,
Send, driving your well-built chariot from Piety.
The identity of the ‘gods’ in the first line is unclear – they may be
specifically the six permanent entities, or this may be a prayer to the gods
in general.185 Similarly unclear is the antecedent of ‘their’, which Hardie
takes to refer specifically to other poets, but since ‘madness’ is elsewhere
associated with Strife and bloodshed (D10=B115.14), it more obviously
refers to meat-eating mortals in general.186 In any case, the ‘pure spring’
and the ‘holy mouths’ evoke ritual purity. Water imagery in general is
commonly used as a metaphor for poetry,187 but Empedocles’ ‘pure spring’
has a more specific significance, for it recalls in particular the springs from
which water was drawn to ‘purify’, either medicinally – the sick would
sometimes bathe in such springs in the belief that they remove diseases – or
ritually, in which case water from particular springs was used to ‘purify’
participants of evils at the start of certain rituals.188 The metaphor thus
implies that the poem has the same purifying function as purificatory
water, of cleansing the addressees of the Strife within them. If, as seems
prima facie likely, the fragment occurred early in the poem, such imagery
184
Along with LM, I take lines 6–13 of this fragment as addressed to Pausanias. I see no need to posit,
with Wright (1995) 157, a substantial lacuna between the two sections on the basis of the shift in
addressee; a somewhat abrupt shift between divine and mortal addressee is paralleled at Hes.
Op. 9–10.
185
Bollack (1965–9) took them to be the six permanent entities; Hardie (2013) 231 tentatively takes
them as a reference to the four elements (A32=Aët. 1.7.28); Gheerbrant (2017) 114, 119–20 sees them
as the four elements plus Love. However, the heightened emotion suggested by the content and the
imperatives of the fragment would make an exclamation to the gods in general at this point
appropriate (cf. Aesch. Supp. 77).
186
Hardie (2013) 210. Wright (1995) 158 argues for a more general reference. Cf. the general criticism of
mortals in D42=B2.
187
See Nünlist (1998) 178–205. Some pertinent parallels are Hes. Theog. 39–40 and Pind. Nem. 7.62–3.
188
Especially comparable is Soph. OC 469–70 πρῶτον μὲν ἱερὰς ἐξ ἀειρύτου χοὰς/ κρήνης ἐνεγκοῦ,
δι’ ὁσίων χειρῶν θιγών, ‘First bring holy libations from an ever-flowing spring, taking them in
unblemished hands.’ See further Parker (1983) 212–13 for medicinal pure springs and 226–7 for ritual
ones. Healing springs are attested by e.g. Hdt. 4.90 and Hippoc. Epid. 5.9. Springs used for ritual
purification are attested at Theophr. Char. 16.2, Ap. Rhod. 3.860, Paus. 2.17.1. Empedocles himself
may allude to this kind of purification at D31=B141, where he describes an action which Theon of
Smyrna labels a καθαρμόν; although there it is likely that the springs are metaphorical springs of
blood, and the fragment is referring to animal sacrifice (see Wright [1995] 289–90).
156 Empedocles
would serve as an appropriate overture for a text that will ‘initiate’ us: this
invocation takes the place of the ritual cleansing that was a prerequisite to
such an experience.
The address to the Muse in lines 3–5 is further suggestive of a ritual
context. In contrast with other Muse-invocations where the Muse is told to
speak or sing, the command here to send a chariot is a striking metaphor,
which may allude to Parmenides’ chariot journey.189 The unusual quality
of this invocation underlines that she will govern ‘what it is right for
ephemerals to hear’ and that her chariot (unlike Parmenides’ one) comes
from ‘piety’. Line 4 recalls the conventional Orphic line, ‘I speak to those
for whom it is right (οἷς θέμις ἐστί); shut your doors, uninitiated!’ (OF 3),
which, was also echoed in Parmenides (D4=B1.28).190 The scenario, then,
is that we will hear a pure speech akin to the sort of speech that is normally
preserved only for ritually purified initiates. If the ‘madness’ at the start of
D44=B3 is, like the madness at D10=B115.14, a consequence of Strife, then
we are further justified in treating the ‘purity’ to which it is opposed as
a purity of Love from Strife’s stain.191 This understanding of the Muse’s
‘pure spring’ is corroborated by the fact that ‘piety’ (Eusebie) only occurs
elsewhere in the fragments specifically as piety towards Love, in the
idealised community that knows no Strife (D25=B128.4). This invocation
implies, then, that the narrator’s speech has a purificatory function, in
washing away the strife from our bodies, something that it will achieve
through initiating us into his understanding of the world.
5 The Effects of Empedocles’ Narratives

If the ideal function of Empedocles’ text is ultimately to purify an audience
by convincing them to worship Love and refrain from acts of Strife, it is
189
On the contrast with other Muse-invocations, see n. 49 above.
190
The recollection of this line in Empedocles is observed by Riedweg (1995) 54.
191
This consideration may weigh in favour of the view, first originating from Cornford, that the
daimon is effectively a shard of Love who has been contaminated by Strife. This view was first
suggested by Cornford (1912) 237–40, who was followed by Long (1949); Kirk and Raven (1957)
358–61; Kahn (1960) 21–3; and O’Brien (1969) 325–36 (who mentions unpublished lecture notes in
which Cornford elaborated the theory), (1995) 442–3, 452–6, (2006) 54–5. Martin and Primavesi
(1999) 90–5 express support for it. However, alternative possibilities are espoused by (among others)
Long (1966) 275–6, who contends that the question of the physical basis of the daimon simply
would not have occurred to Empedocles; Barnes (1982) 495–501, who argues that the δαίμονες are
compounds of elements, and on this point is followed by Sedley (2007) 51 n. 62; Trépanier (2014)
and (2017); and Shaw (2014), who argues that he is a portion of Aither. Whilst I find Cornford’s
theory the most plausible, for the purposes of the present chapter, I am agnostic on this issue, but
see further Mackenzie (forthcoming).
The Effects of Empedocles’ Narratives 157
designed to achieve this end through means that are characteristic of both
Archaic conceptions of song and later conceptions of the poetic. In the
present section, I shall explore four respects in which Empedocles’ use of
these narratives exploits traditional features of Greek verse to achieve his
didactic and transformative ends: (a) its connection with myth, (b) its
emotionally engaging qualities, (c) its ability to make an audience visualise
phenomena they cannot literally see and (d) its capacity to invite penetrat-
ing, subtextual readings.
a Empedocles and Myth

André Laks and Oliver Primavesi have characterised the story of the
daimon as a myth, in presenting an alternative type of account to the
cosmology, one that illustrates different facets of Empedocles’ philosophy.
For Laks, the tension between the two accounts draws us into a dialectic
between different ways of understanding the world. In particular, we are to
reflect on whether Love and Strife are primarily principles of cosmic
processes, or of human/daimonic action; whether their onset is inevitable
(as in the cosmology), or whether it derives from our choices (as in the
daimonology).192 For Primavesi, the daimonology functions as an allegory
for the cosmology: the disintegration of the community of the daimones
symbolises the disintegration of the Sphere.193 This is an extreme approach,
one which is much less economical than supposing that such a community
is actually meant to exist somewhere (or at some point) in the cosmos.194
Laks’ understanding is attractively subtle. But the claim that daimonology
is in some sense a myth requires further precision.
As is well-known, muthos only acquires its later sense of ‘myth’ in the
Classical period, probably with the Sophists, in contradistinction to logos
or ‘rational explanation’.195 The Empedoclean narrator’s choice of the term
muthos for his speech (D60=B23.11; D46=B24.2; D157=B62.3; D6=B114.1;
D73.245–6) does not yet have this significance and, on the basis of earlier
usage, may instead indicate that it is an authoritative, weighty, public
pronouncement, in keeping with his characterisation as a performer of
wisdom.196 Nevertheless, it is clear that, from the point of our earliest
sources onwards, the Greeks had a distinct category of stories, mostly
192 193
Laks (2004) 18–32; Laks (2005). Primavesi (2008a).
194
See n. 133 above for further objections to Primavesi’s allegorical interpretation.
195
On this development, see R. Fowler (2011), who cites further bibliography.
196
For muthos as an authoritative, weighty pronouncement, see Martin (1989) 1–42; Lincoln (1999);
R. Fowler (2011) 53.
158 Empedocles
traditional,197 that were empirically unverifiable because they took place
prior to living memory or outside of the mortal realm and that could serve
a variety of functions, from pure entertainment, to the demonstration of
moral principles, to the aetiological explanation of present-day natural
phenomena and human customs.198 The type of story that would later be
referred to as ‘myth’ would have been distinctive even before it acquired
a particular label and was closely associated with (if not the exclusive
property of) Muse-inspired poetry.199 Empedocles’ poetry occupies the
space of myth in this sense: both the cosmology and the daimonology
concern events outside the realm of normal mortal experience, and both
are told in Muse-inspired hexameters.
If Greek poetic ‘myth’ concerns events outside of the empirical scope of
the audience, then Empedocles’ narrator himself is a mythical character: he
has seen first-hand the phenomena he describes (cf. D6=B114; D13=B117)
that are outside his audience’s experience and comes to the familiar world
of Acragas (D4=B112) as an exile from the gods who has committed divine
perjury in a pointedly similar manner to the gods of Hesiod’s Theogony
who forswear by the river Styx.200 Although he invokes the Muse
(D44=B3; D7=B131), he does not do so in order to draw on the resources
of their divine knowledge but to ensure that his poetry is pious and
beautiful.201 Unlike the narrators of Hesiod’s and Parmenides’ poems, he
does not have to rely on the Muses, or depart from the familiar realm of
mortals, to speak authoritatively about this extraordinary world; rather, he
comes to us from it. He is a god who interacts openly with the mortal
Acragantines, a situation that contrasts starkly with, for instance, the
opening of the Catalogue of Women, where such divine–mortal interactions
197
Traditionality is often regarded as central to myth (e.g. in the oft-cited definition, Burkert [1979] 23:
‘myth is a traditional tale with secondary, partial reference to something of collective importance’),
but to make traditionality a necessary condition would be to exclude several stories that are usually
considered myths, such as Plato’s ‘myths’ and, probably, Hesiod’s version of the myth of the races.
This point is well-made by Bremmer (1987) and Csapo (2005) 8–9. On the development of the
muthos/logos distinction, see Detienne (1986) and especially R. Fowler (2011).
198
See similarly Buxton (1994) 12–17. For the category of ‘myth’ in the context of early Greek poetry,
see also Detienne (1986) 22–62; Martin (1989); Morgan (2000) 20–4.
199
Contrast (e.g.) Bremmer (1987) 7, ‘in archaic Greece myths were the exclusive territory of poets’,
a statement which overlooks material culture. Nevertheless, as Hdt. 2.53 attests, the Homeric and
Hesiodic poems became particularly authoritative sources for some myths. Dowden (1992) 6–8
explains the peculiar authority of Hesiodic and Homeric poetry as a source for myth as due to their
Panhellenic qualities. Cf. also Morrison (2007) 78–9 on the lack of Muse-inspiration for non-
mythic subject matter in Archaic poetry.
200
On this relationship, see especially Most (2007). Note in particular the allusion to Theog. 793 in
D10=B115.4, and to 800 in D10=B115.12.
201
See n. 49 above.
are confined to the remote past (fr. 1; cf. also Theog. 535–7). By confronting
us with a primary narrator from the world of myth, Empedocles breaks
down the barrier between the world of the audience and the extraordinary
world of mythical epic.
This end is also served by the redeployment of epic vocabulary to the
familiar, natural world. The Homeric hapax legomenon ὀξυβελής, used
significantly of the arrow with which Pandarus breaks the truce by shoot-
ing Menelaus (Il. 4.126),202 is now applied to the sun (D125=B40, probably
referring to its rays) and to the spines of hedgehogs (D197=B83 ὀξυβελεῖς
χαῖται). The epithet ‘late-born’ is used by Homer for future generations of
men, thus marking the chronological distance between characters and
audience (Il. 3.353; Od. 1.302), but Empedocles uses it of pomegranates
(ὀψίγονοι, D255=B80) because their fruit appears late in the year. The
antlers of deer are μελίαι (D74.6), a word that properly refers to the ‘ashen
spears’ that are the weapons of elite Homeric warriors. Furthermore, as we
have noted, the cosmology accommodates the fantastical creatures of
myth, such as the Minotaur. There is thus a continuum between the
period of such fantastical creatures and that of the contemporary audience.
This situation contrasts with the traditional notion that there has been an
irrevocable break since the world of myth, when monsters of this kind were
all slayed by Herakles and other heroes.203 By eliding the normal barrier
between the world of myth and the world of the audience, Empedocles
mythologises and defamiliarises the familiar world, inviting us to apply the
sort of reflective attention to his text and to the world that is normally
reserved for the tales of epic singers (cf. Theog. 55).
Yet in addition to this respect in which Empedocles’ poetry as a whole
assumes mythical status, there is a further respect in which Laks and
Primavesi are right to identify the story of the daimon more specifically
as a myth, which would not apply to the cosmology. The narrator’s story
conforms to a mythical plot-type of exile and purification and concerns the
actions and choices of anthropomorphic agents, whereas the cosmology
concerns the inevitable movements of non-anthropomorphic (albeit per-
sonalised) divine forces. As I argued above (Section 4), the daimonology
would count as a muthos in Aristotelian terms, but the cosmology would
not. It is then, in a sense, a ‘myth’ within the wider muthos of the poem.
Given this ‘mythical’ quality, I want to suggest that two traditional
202
And thus re-establishing Trojan guilt at an early point in the poem. On the significance of this
episode, see Taplin (1992) 103–9.
203
Thus, explicitly, Eur. HF 696–700, a passage adapted at Seneca HF 882–92.
160 Empedocles
functions of Greek myth are relevant to the story in particular: as an
aetiology for ritual practice and as an exemplum for moral lessons.
Whilst it would be an exaggeration to claim (as some have) that myth is
always connected to ritual or vice versa, it is undeniable that, in our
sources, myths are often used as explanations for ritual practices.204 Two
examples in particular bear close comparison to the story of Empedocles’
narrator. First, Orestes’ trial in the Eumenides serves as the model for trials
at the Areopagus.205 Like the Empedoclean narrator, Orestes needs to be
purified of blood guilt and this purification involves a watershed moment
in the history of the polis. Second, Prometheus’ deception of Zeus, as told
by Hesiod (Theog. 535–64), is the first instance of the sacrificial practice
whereby the meat, wrapped in an ox-stomach, is eaten by the humans and
the bones wrapped in fat are left for the gods.206 The narrator, like
Prometheus, commits a treacherous primordial crime involving the con-
sumption of meat, one that indicates a reason to perform divine offerings
in a certain way; but whereas, in Prometheus’ case, the sacrifice sets
a precedent to be followed, the story of the daimon is an example of what
not to do: the crime of killing the animal demonstrates that offerings
should be bloodless. Like these other examples, the story of the daimon
concerns the choices of an anthropomorphic agent and is tied to
a justification for a particular ritual practice, but unlike them it does not
provide the origins for the custom. It is not as a result of the daimon’s
actions that we should refrain from blood-sacrifice, since the law pre-
existed his actions (D27a=B135). Explanation for this proposed custom is
instead to be found in the cosmology: animals contain daimones just as we
do and so should not be killed. By outlining particular rituals in
a hexameter poem alongside a diachronic narrative that resembles certain
aetiological myths, Empedocles draws attention to the manner in which
they are explained, not by a mythological aetiology, but by a systematic
cosmology.
But although the story of the daimon is not strictly aetiological, it serves
a purpose typical of aetiological myths in functioning as an exemplum on
which to model subsequent human behaviour. Aetiological myths are
especially associated with this function through the ‘genetic fallacy’ (e.g.
‘Prometheus originally made sacrifices in a certain way, so we should too’),
204
On this much-discussed topic see, with respect to tragedy, Kowalzig (2006) and Romano (2012);
with respect to Pindar, Rutherford (2011); and, more generally, Kowalzig (2007) 24–32.
205
See Sommerstein (1989) 13–17.
206
For other hexameter examples, note Dem. 265–7 with Richardson (1974) ad loc.; Hermes 105ff. with
West (2003a) 13–14.
but of course, myths often illustrate the right or wrong course of action
even when there is no such aetiological connection, as, for instance, when
Achilles tells the myth of Niobe to encourage Priam to eat in spite of his
grief (Il. 24.599–620). In this example, and in several other instances
(aetiological or otherwise), a diachronic narrative, set in a past time,
illustrates an ethical point of synchronic relevance to the speaker’s
addressee. Thus, in the Works and Days, Hesiod uses the myths of
Prometheus and of the different races of men to illustrate particular ethical
points. This latter example is of particular relevance to Empedocles, since
he seems to have adapted it in his own golden age imagery. According to
Vernant’s celebrated analysis, Hesiod’s diachronically successive races in
fact represent types that are still present, a possibility supported by the fact
that some of the imagery and vocabulary used in the myth recur later on in
the poem to describe contemporary phenomena.207 Likewise, the ideal
society that Empedocles describes in D25=B128 functions, like Hesiod’s
golden race, as a description of a previous community that can again be
realised, if only we behave in the correct way. More generally, Empedocles’
story of the daimon in exile performs a paradigmatic function along the
lines of these earlier examples: it is a particular diachronic narrative that
illustrates ethical points of synchronic relevance to the addressee. The story
justifies the commands against committing bloodshed and eating beans
(D27a=B135; D29=B137; D30=B145) by providing an emotionally engaging
example of the negative consequences of failing to follow these injunctions.
Both the cosmology and the daimonology, then, combine to promote
the institution of new ritual practices by taking over the aetiological and
exemplary roles traditionally performed by myth. Pace Primavesi, the story
of the daimon is not an allegorical representation of the cosmic cycle, but
illustrates a moral principle that is of direct importance to the citizens of
Acragas (and, by implication, the citizens of other poleis), whilst at the same
time prompting reflection (as Laks argues) on the nature of necessity and
human agency. The fact that the story conforms to a recognisable pattern
and is fashioned to serve these purposes may cast doubt over whether its
author believed in its factual veracity, that is, over whether Empedocles
actually believed himself to be a god in exile who had undergone this
extraordinary experience.208 Of course, this consideration could not pro-
vide conclusive proof one way or the other, but it is notable that the
daimonic story may serve its illustrative and dialectical purposes regardless
of the author’s commitments to its factual accuracy, just as many of the
207 208
See n. 137 above. Cf. Griffith (1983) on the conventionality of Hesiod’s persona.
162 Empedocles
embedded myths in Homer, Hesiod and Epinician do not seem to demand
that we treat them as historical.209 By contrast, the same could not be said
of the cosmology: Empedocles seems to stake a stronger claim to the factual
accuracy of this component, since his ethical instructions depend upon it
and since it competes with other truth-candidates such as the cosmologies
of the earlier Presocratics. With the daimonic narrative, what is more
important is that the story illustrates a moral point in a manner that is
emotionally relatable.
b Emotions and Aesthetics

The ‘myth’ of the daimon illustrates the value of Empedocles’ ethical
injunctions precisely because it is an engrossing and emotionally affecting
narrative. We observed above that D29=B137, involving the image of
a father sacrificing his son and children eating their parents, recalls espe-
cially the plots of Attic tragedy. In the daimon’s urgent exclamation at
D76.5–16, he retrospectively wishes that he had been destroyed prior to
having committed an act of bloodshed (‘Alas that the pitiless day did not
destroy me earlier / Before I contrived terrible deeds about food with my
claws!’, 5–6), crying, in true tragic fashion, oimoi! These and further details
can be taken to suggest that narrative of the daimon is designed to elicit the
characteristically tragic emotions of pity and fear that would become
ratified in Aristotle’s Poetics (1449b27–8). Although Aristotle is a later
source and although he famously excludes Empedocles from the category
of poetry, he can nevertheless be taken as evidence for the sorts of responses
that Empedocles’ verses could have elicited roughly a century earlier. This
is because Aristotle articulates and systematises traditional notions that are
evident in earlier Greek texts.210 The notion that pity and fear are appro-
priate to tragedy is already implicit in the practice of the tragedians
themselves and may ultimately derive from Homer.211 By the time of
209
On this topic in Hesiod’s Works and Days, see Currie (2012).
210
Aristotle makes explicit his ‘endoxic’ method of selecting the best widely held views at Eth. Nic.
7.5.1145b2–7, on which see e.g. Shields (2013) 27–32. For a defence of the (critical) use of Aristotle as
a guide to interpreting fifth-century texts, see Heath (1987) 2–4, and (specifically with respect to the
tragic emotions) 11–17.
211
For earlier conceptions of pity and fear as appropriate responses to poetry, note Gorg. (Helena)
D24=B11.8–9; Pl. Ion 535b–e, Phdr. 268c; Andocides 4.23; Isocrates 4.168 (with Halliwell [2002]
212–14 on these latter two). The lineage of this idea is already evident in Homer, Od. 14.388–9;
Shankman (1983) persuasively argues that it is a mark of Antinous’ moral deficiency that he does not
respond with such emotions to Odysseus’ story at Od. 17.415–44. See also Rutherford (1982) for pity
in the Iliad.
Empedocles’ composition, they were well established as the responses
appropriate to scenarios such as D29=B137.
The story of the daimon conforms to Aristotelian principles of tragedy in
certain further respects. The narrator’s exclamation at D76.5–16 is reminiscent
in particular of scenes of tragic anagnorisis where the protagonist has realised
too late that he has committed a monstrous action.212 The narrator has fallen
from good to bad fortune through his own error (cf. Poet. 13.1452b28–
1452b39). For both Empedocles and Aristotle, the reception of the text is
tied to the process of katharsis or ‘purification’ (cf. D44=B3.2, D257=B110.2;
Arist. Poet. 1.1449b27–8). This conceptualisation of literary response as kathar-
sis probably stems, in the case of both authors, from an association with ritual
purification.213 Aristotle even attributed tragedies to Empedocles (fr. 73
Janko=D.L.8.57=Empedocles P25=A1). Diogenes Laertius reports that this
claim was already disputed in antiquity and, as Janko notes, since the Suda
ascribes twenty-four tragedies to the philosopher’s namesake grandson, the
latter’s work was probably mistakenly attributed to the former.214 But the
tragic features of Empedocles’ own poetry (especially the characteristically
tragic scenario of D29=B137) could have facilitated this error. Indeed, one
reason why Aristotle singles out Empedocles (rather than Xenophanes or
Parmenides) for exclusion from the category of poetry (Poet. 1.1447b) may
have been precisely because otherwise, on the basis of Aristotle’s account of
tragedy, we might have assumed that Empedocles’ work fitted the bill.
Aristotle articulates a further dimension to tragic pity and fear that is of
particular relevance to Empedocles: they provoke an awareness of shared
vulnerability between character and spectator.215 For Aristotle, pity
depends on the sense that we too might suffer what afflicts those pitied
(Rhet. 2.8.1385b13–33, 1386a25–9); tragic fear also derives from the feeling
that similar things to the events onstage might happen to us (Poetics
13.1453a4–6; cf. Rhetoric 1386b27–9). Again, these ideas have been traced
to much earlier poetic texts.216 If we accept Aristotle’s definitions of pity
212
This point is made by Gagné (2006) 95–6. Cf. esp. Soph. OT 1307–8; 1316–18 (with Cairns [1993]
218, ‘his pain, which is so vividly presented, lies less in fear of disgrace than in his knowledge of what
he has done’); Ant. 1271–6; OC 1254–66; Eur. HF 1146–52.
213
For this interpretation of Aristotelian katharsis, see Ford (2016) and, more briefly, Halliwell (1998)
194. Aristotle uses religious ecstasy (enthousiasmos) to illustrate musical katharsis at Politics
8.7.1341b32–1342a18.
214
Janko (2010) 538.
215
For Aristotelian pity as based upon a notion of shared vulnerability, see Halliwell (1998) 175–84.
216
See Rutherford (1982) 158–60; Stanford (1983) 21–9; Halliwell (2002) 207–33; Munteanu (2012) esp.
181–207; Cairns (2015) 86–7. For earlier examples where shared vulnerability is implicit, note e.g. Il.
24. 485–551; Soph. Ajax 121–6, OT 1194, 1211, 1216–21, 1286, 1296, 1299, 1303, 1347.
164 Empedocles
and fear, Empedocles’ incitement of these emotions invites comparison
between the circumstances of the audience and those of the narrator.
Initially, this comparison is one of contrast, with the narrator (at least
seemingly) a god speaking to mortal addressees (D4=B112), but the contrast
is inverted when we learn that that narrator is an exile supplicating those
who are ‘inexperienced in ill’ (D10=B115; D4=B112.3), a situation that
contrasts pointedly with his own. Ultimately, however, we must come to
the chilling realisation that we ourselves (if, like most ancient Greeks, we
are meat-eaters) are in the same predicament as the narrator and have
committed crimes that contravene the immutable laws of the cosmos.
Thus, it is not only the case that we might suffer what the narrator has
experienced; we ourselves undergo an anagnorisis similar to that of
Sophocles’ Oedipus (of the Oedipus Tyrannus), that we are in fact desper-
ately polluted and in need of purification. The world we inhabit is one of
‘anger’ and ‘slaughter’ (D24=B121); we are a ‘wretched’ and ‘utterly
unblessed tribe of mortals’ who are ‘born of strife and wailing’
(D17=B124). As with Oedipus, it is not that our circumstances have
changed, but that we realise that those circumstances are far worse than
we ever suspected them to be. The ideal experience of the audience, then,
replicates a further tragic topos, that to learn is to suffer.217 Indeed, the
narrator’s statement that the truth ‘is very painful for men, and the onrush
of conviction is hard for the mind’ (D6=B114) is a variant of this topos.
If this interpretation of the emotionally affecting quality of Empedocles’
story is accepted, it can help us to understand its exemplary function.
Douglas Cairns draws on the resources of cognitive science to explain the
emotive and exemplary aspects of certain Greek narratives: stories, by
facilitating an audience’s imaginative engagement with the lives of others,
build empathy and can crystallise norms in the patterns of narratives and
the emotional responses they warrant.218 Empedocles’ story of the daimon
can be read in these terms: by conforming to certain familiar patterns (exile
and purification; human sacrifice and unintentional cannibalism) it serves
to crystallise those patterns and the emotional responses they elicit (pity,
fear), but it does so in the service of building empathy with non-human
beings (sacrificial animals and even plants into which the narrator has been
incarnated). More simply put, Empedocles presents the killing of animals
as an instance of the traditional mythical patterns of human sacrifice and
cannibalism and so worthy of the emotions of pity and fear warranted by
217
E.g. Aesch. Ag. 177; Soph. OT 544–5, Ant. 1272; Eur. HF 1115, Bacch. 1113. Cf. Hdt. 1.207.1.
218
Cairns (2014) 104–5.
such narratives. Traditional narrative patterns thus illustrate novel precepts
(we should not eat meat or beans), which are presented as an entailment of
traditional principles (cannibalism is taboo; murderers should be ritually
purified; gods deserve piety). Empedocles uses hexameter poetry for its
traditional purpose of eliciting emotions, but this purpose serves the
further didactic end of grafting new customs onto well-established ethical
principles.
Yet to focus only on the tragic qualities of the narrative of the daimon
would be to omit a large part of the surviving evidence of the poem. It
seems likely that this narrative of the daimon functioned as a preface to the
cosmology (just as Parmenides’ proem prefaces the goddess’ speech).219 If
this structure is correct, then through the narrative of the daimon we are
made bitterly aware of our mortal pollutedness before we are granted an
awe-inspiring, god’s-eye perspective on the cosmos. As was the case with
Parmenides’ Doxa, in the physical fragments, we are granted a view that
would normally be inaccessible to mortals. A good example occurs at
D75=B35.14–17:
αἶψα δὲ θνήτ’ ἐφύοντο, τὰ πρὶν μάθον ἀθάνατ’ εἶναι,
ζωρά τε τὰ πρὶν ἄκρητα διαλλάξαντα κελεύθους. 15
τῶν δέ τε μισγομένων χεῖτ’ ἔθνεα μυρία θνητῶν,
παντοίαις ἰδέηισιν ἀρηρότα, θαῦμα ἰδέσθαι.
Swiftly they became mortal, the things which before learned to be immortal
And mixed, things which before were unmixed, exchanging their paths. 15
And from them, as they mixed, poured forth countless tribes of mortals,
Fitted with all kinds of forms, a wonder to behold.
The passage is a formal tour de force, featuring artful antitheses (‘mortal . . .
immortal’ . . . ‘mixed . . . unmixed’) and repetitions that illustrate the sheer
number of new forms produced (τὰ πρίν . . . τὰ πρίν), culminating in the
chiasmus of line 16 (μισγομένων χεῖτ’ ἔθνεα μυρία θνητῶν). It is as if the
poem itself were ‘fitted with all kinds of forms, a wonder to behold’. But
although these phenomena may be a ‘wonder to behold’, no member of
Empedocles’ audience could actually have beheld them. To recall such
a wondrous sight (according to Empedoclean principles) would require
either a memory of one’s past incarnations or the status of being a ‘long-
lived god’. Throughout the physical fragments, we are granted this god’s-eye
perspective on the cosmos, in all its details, both on a macroscopic and on
a microscopic level. Porter has justifiably described this level of detail as
219
See n. 158 above.
166 Empedocles
a sublime ‘poetics of excessiveness’.220 As we observed in the case of
Parmenides’ Doxa, the perspective is comparable to the Iliadic shield of
Achilles – which itself would later be allegorised as representing Empedocles’
cosmology –221 where the level of detail was so astounding that only gods or
the godlike Achilles could bear to look at it. Again, the experience of initiates
into the Eleusinian mysteries seems comparable: an initially terrifying
experience ultimately culminates in an awe-inspiring, sublime vision.222
The emotional trajectory, and the imagery of initiation, tender the possibil-
ity that we can purify ourselves of our mortal pollution and acquire a divine
status.
Again, Conte’s observation of the De rerum natura seems applicable,
that the text ‘generates a sense of inadequacy in whoever happens to
contemplate it; then it elevates the reader’s mind to a higher level of
understanding and fills him with enthusiasm’.223 Empedocles’ poem gen-
erates a sense of our inadequacy (both ethically and epistemologically, cf.
D17=B124, D42=B2) as fallen mortals, but then places us on a divine
vantage point from which we can imagine the benefits of regaining our
purified godhood. But whereas, in Lucretius, the attainment of this quasi-
divine status is a psychological shift, an alteration of our beliefs and
attitudes, for Empedocles, through changing our ritual habits, we may
actually become, ultimately, theoi. The divine perspective on the cosmos
that Empedocles provides thus offers a foretaste of the sort of status we
might attain if we follow the instructions and regain our godhood so that
we lack the limitations and miseries of mortals.
c Textual Visions
As with Parmenides, there is a marked visual dimension to this fabulous
experience that can be regarded as an aspect of the poem’s evocation of
mystic initiation, but also takes on a more specific significance within the
context of author’s philosophy. We noted in the Introduction that
a frequently observed characteristic of early Greek song is its capacity to
turn its listeners into spectators, exposing them to a particular perspective
220
Porter (2016) 419.
221
See Hardie (1985) 15. This interpretation stems from the Hellenistic allegorical interpretations
preserved in Heraclitus’ Homeric Problems and in Eustathius.
222
Gagné (2006) 98–102 relates the fear elicited by Empedocles’ poetry to the fear experienced in
mystic initiations, but without commenting on the further detail that both culminate in a sublime
vision.
223
Conte (1994) 28, quoted in Chapter 2 Section 4e above with respect to Parmenides.
on the world. Empedocles stresses that his poetry exercises this capacity by
repeatedly using visual language of his verbal account (D44=B3.9;
D77a=B21.1; D257=B110.2; D74.326=B76.3; D73.292–3, 300). This visual
language forms part of the figuration of the response to the text as a process
of initiation: Pausanias’ ‘initiation’ consists in being granted a special
‘vision’ (cf. D257=B110.2) of the broad-scale workings of the cosmos (e.g.
D75=B35.17), just as initiation into the mysteries at Eleusis culminated in
the ‘viewing’ of a certain spectacle in the Anaktoron/anaktora (see Chapter 2
Section 4f above). But this visual language also carries a more specific
significance, given the nature of Empedocles’ cosmology. By contrast with
Parmenides, Empedocles accords a greater role to the senses as a means of
accessing the underlying nature of reality, a contrast that is marked by
pointed allusions.224 Yet the evidence of our senses alone is still not
sufficient to grasp a true understanding of the world: the subject matter
of the poem is ‘neither seen nor heard by men / nor grasped with the mind’
(D42=B2.7–8). At D73.252=B17.21, the narrator commands the student to
‘See [Love] with your mind, and do not sit stunned by your eyes’, that is, to
visualise an item that is literally invisible. The poem reveals to us an image
of the wider workings of the cosmos but will ideally also grant us the ability
to see in a more penetrating manner than conventional seeing, so that we
can identify the underlying elemental constitution of manifest
phenomena.
This sort of penetrating vision is made more explicit at D73.291–300:
σπεῦ]δ̣ε δ᾽ ὅπως μὴ μοῦνον ἀν᾽ οὔατα [μῦθος ἵκηται,
ἠδέ] μευ ἀμφὶς ἐό̣ ντα κλύων [ν]η̣μερτ[έα δέρκευ·
δεί]ξω σοι καὶ ἀν᾽ ὄσσε ἵνα μείζονι σώμ̣[ατι κύρει,
π]ρῶτον μὲν ξύνοδόν τε διάπτυξίν τ̣⌊ε γενέθλης,⌋
ὅ̣ σ̣[σ]α τε νῦν ἔτι λοιπὰ πέλει τούτοιο τ[όκοιο, 295
τοῦ̣ το μὲν [ἂν] θηρῶν ὀριπλάγκτων ἀγ̣[ρότερ’ εἴδη,
τοῦτο δ᾽ ἀν᾽ ἀ[νθρώ]πω̣ν δίδυμον φύμα, [τοῦτο δ᾽ ἀν᾽ ἀνθέων
224
In D44=B3 to Parmenides D8.1–6=B7. Both fragments mention the three senses of vision, hearing
and taste as potential means to provide a ‘path’ (πόρος ἐστὶ νοῆσαι, D44=B3.12; ὁδοῦ, ὁδὸν,
D8.2–3=B7.2–3), and in each case there is a list of negative commands followed by a concise positive
command connected with an adversative δέ. Empedocles’ command contrasts pointedly with that
of Parmenides: Pausanias is not told to reject the senses but, instead, to use them in a balanced way;
for Parmenides, the senses head down the road of what-is-not, a road from which the youth must
withhold his ‘understanding’ (νόημα), whereas Empedocles’ senses are all implicitly ‘routes for
understanding’ (πόροι νοῆσαι). The command νόει δ’ ἧι δῆλον ἕκαστον, ‘understand how each
thing is clear’, bids the student to perform an action that would be impossible to perform correctly
according to Parmenides’ epistemology, where only one thing is properly there to be the object of
νοεῖν. The allusion to Parmenides, then, underlines Empedocles’ defence of the senses as a means of
achieving understanding.
168 Empedocles
ῥιζοφόρων γέ̣ ννημα καὶ ἀμπελοβάμ[ονα βότρυν.
ἐκ τῶν ἀψευδῆ κόμισαι φρενὶ δείγματα μ[ύθων·
ὄψει γὰρ ξύνοδόν τε διάπτυξίν τε γενέθλη⌊ς⌋ 300
And make an effort so that [my speech?] does not only [reach?] your ears,
[and] hearing from me the things around us [see] unerring things;
I will show you even by your eyes where they meet a larger body,
First the coming-together and the unfolding of a generation,
And as many things as now remain from this [brood]225 295
Firstly, the one among the wild [forms] of mountain-wandering beasts,
Then the one among the double race of human beings, [and then among the]
Race of root-bearing [flowers] and vine-climbing [grapes.]
From these bear to your mind the unlying revelations of my words;
For you will see the coming-together and the unfolding of a generation. 300
(Text from LM, based largely on Martin and Primavesi [1999])
These lines have sometimes been taken as an appeal to empirical evidence:

Pausanias, it is held, should not merely accept the word of the narrator, but
judge by the evidence of his senses.226 I suspect, however, that the situation
is not quite so similar to modern scientific methods. Our verdict on this
point depends partly upon our supplementation of the gaps in the papyrus,
especially line 292. If we accept Martin and Primavesi’s suggestions for the
line (reprinted above), the power of vision afforded to Pausanias will be
gained by hearing the words of the poem. It is not the case that Pausanias is
to see certain things and then find confirmation of Empedocles’ words;
rather, Empedocles’ words will enable Pausanias to see in a certain way,
that is, to see the objects of his experience for what they really are,
compounds of the four elements, governed by the forces of Love and
Strife, at some stage in the long-term cosmic cycle. Against this view,
Sedley has proposed the following supplement and translation:227
εἰ δέ] μευ ἀμφὶς ἐό̣ ντα κλύων [ν]η̣μερτ[ὲς ἐπόψει
If, as you listen to me, you look unerringly upon the things around you
Even this, however, need not suggest an appeal to empirical evidence, since
it could equally be translated as ‘if, hearing about the things around you
from me, you look unerringly’, thus still presenting the text as the instru-
ment by which the student will achieve this unerring vision. However we
choose to supplement 292, the fact that the visual language in this passage is
not a proto-scientific appeal to empirical evidence is confirmed by the
225
Probably, the brood that has come to be since the last Sphere and has not yet disintegrated in this
period of increasing Strife (contra Sedley [2007] 36–7).
226 227
Sedley (2007) 35–8. Sedley (2007) 35–6.
narrator’s promise to ‘show’ Pausanias the ‘coming-together and unfolding
of a generation’, for this phenomenon would normally be impossible for
a mortal to ‘see’ in any simple, literal sense. The speech itself affords the
relevant visual experience, as the expression (which depends on a more
secure supplementation) ‘revelations of my words’ (299) implies.228
Through hearing Empedocles’ words, Pausanias will ‘see’ the cosmos, in
a way in which mortals do not typically see.
A specific example of one of these ‘revelations’ is preserved in D74>B76,
a fragment now expanded thanks to the Strasbourg papyrus:
τοῦτο μὲν ἐν κόγχαισι θαλασσονόμοις βαρυνώτοις,
ἠδ᾽ ἐν πε]τ̣ρ̣αίοισι κα[
ἔνθ᾽ ὄψει χθόνα χρωτὸς ὑπέρτατα ναιετάουσαν·
θώρηξ δ᾽ αὖ]τ̣ε κραταιν[ώ]των α[
ναὶ μὴν κηρύκων τε λιθορρίνων χελύων τε
ὄστρακα κα]ὶ μ̣ε̣ λ̣ί̣ α̣ι̣ κ̣εραῶν ἐλά[φων ὀριπλάγκτων.
ἀλλὰ οὐκ ἂν τελέσαι̣ μ]ι̣ λέγων σύμ̣[παντα γένεθλα
This, in sea-grazing, heavy-backed shells,
And in r]ocky[
There you will see earth inhabiting the uppermost parts of the flesh;
And, once more, the breast-plate] of strong-backed [
Yes indeed, and of the stone-skinned trumpet-shell, and of turtles
The shell and] the ashen spears of the horns of [mountain-wandering] stags
But I would not finish] telling of [all the species]
Pausanias will ‘see’ the ‘earth’ on the back of shelled animals, such as
seashells and turtles. Whilst these animals can of course be part of normal
visual experience, it is only because of Empedocles’ instruction that
Pausanias will be able to see the ‘earth’ on them, a detail that, presumably,
would have escaped the mortals dismissed in D42=B2. The vivid experi-
ence of the poem enables us to ‘see’ the elemental constitution underlying
the phenomena surrounding us.
In this way, Empedocles, like traditional hexameter poetry, transports his
audience away from the hic et nunc, a process that is facilitated by the fact that
he portrays the world in the defamiliarised language of epic. In D74 the stags’
horns are ‘ashen spears’ and, possibly, the animals in line 4 have a ‘breast-
plate’. But rather than transporting us into the imagined distant past of myth,
the poem leads us to a vision of the regular structures that govern present-day
228
Note also how the sage of D38=B129 (probably Pythagoras) was easily able to ‘see’ (λεύσσεσκεν) the
things across ten and twenty lifetimes: this special power of seeing here is the result, rather than the
cause, of a correct understanding of the universe.
170 Empedocles
sensible phenomena. The fact that Pausanias is to learn to ‘see’ in a special way
suggests that the audience, too, should acquire this power, once they have
heard the poem. Through our immersive experience of Empedocles’ vivid
poetic world, we are to attune our senses so that we are then able to see the
elements which underlie the various phenomena of our sensory experience.
We no longer merely see tortoise- or snail-shells or antlers, but different
portions of indestructible earth. Indeed, like the student of Parmenides’
poem, we are taught to identify what is constant in an ostensibly changing
world (D73.240–4; D19=B26.12). The defamiliarising, emotionally affective
and vividly visual aspects of the poem thus co-operate to modify the audi-
ence’s attitudes and perceptions. As with Parmenides, a further component to
this process is the manner in which the poem invites a critical reflection on the
way in which certain textual details relate to reality.
d Interpreting the Text, Understanding the World

In Chapter 2, I argued that Parmenides adopts a conspicuously symbolic
form of discourse as a strategy to prepare his audience to apply
a penetrating interpretation to their sensory experience: the poem enig-
matically represents a deeper meaning, while the shifting variety of our
sensory experience is an imperfect likeness of unchanging true reality.
Empedocles deploys the same tactic: just as we are to identify a deeper
symbolic meaning beneath the obvious sense of his language, so we are to
look beneath the surface of the manifold phenomena presented to us by
our senses in order to identify the six permanent entities. He most obvi-
ously demands this mode of interpretation in his use of divine names:
τέσσαρα γὰρ πάντων ῥιζώματα πρῶτον ἄκουε·
Ζεὺς ἀργὴς Ἥρη τε φερέσβιος ἠδ’ Ἀιδωνεύς
Νῆστίς θ’, ἣ δακρύοις τέγγει κρούνωμα βρότειον.
Hear first the four roots of all things;
Bright Zeus and life-giving Hera and Aidoneus
And Nestis who wets the mortal spring with her tears.
(D57=B6)
The four names conventionally refer to anthropomorphic deities and the

anthropomorphism of these figures is accentuated by the fact that they are
paired as two married couples: Zeus and Hera; Aidoneus/Hades and Nestis,
who appears to have been a Sicilian goddess akin to Persephone.229 With the
229
Thus Kingsley (1995) 348–56; Picot (2008b).
exception of Nestis’ identification with water, there is disagreement over
precisely which deity corresponds to which element,230 but this question
does not concern me so much as the precise respect in which these names
relate to their referents. Scholars have sometimes described the fragment as
an allegory on the assumption that the anthropomorphism is intended
figuratively.231 On this view, Empedocles uses Nestis to symbolise the cosmic
mass of water, but without positing the existence of an actual, human-
shaped deity of that name. This approach is supported by the parallel of
D93=B134, where the pointedly non-anthropomorphic ‘holy mind’ was
probably referred to as Apollo. Certainly, Empedocles uses traditional divine
names in an untraditional manner. We are to read beneath the anthropo-
morphic connotations of Zeus, Hera and the rest to identify a cosmic mass,
just as allegorical interpreters such as Theagenes and the Derveni commen-
tator do with the divine names in Homer and the Orphic Theogony.232 But
rather than seeing this unconventional usage as intended figuratively, it is
perhaps better thought of as a redefinition: Zeus and Hera still exist in
Empedocles’ cosmos, but they are actually identified with, rather than
figuratively representative of, the cosmic masses to which they correspond.
The apparent rationale behind Empedocles’ enigmatic language con-
firms this understanding of his divine names. The anthropomorphic names
are the terms used to refer to the elements from a mortal perspective. So
much is evinced by D73.251–5=B17.20–4:
καὶ Φιλότης ἐν τοῖσιν, ἴση μῆκός τε πλάτος τε·
τὴν σὺ νόωι δέρκευ, μηδ’ ὄμμασιν ἧσο τεθηπώς·
ἥτις καὶ θνητοῖσι νομίζεται ἔμφυτος ἄρθροις,
τῆι τε φίλα φρονέουσι καὶ ἄρθμια ἔργα τελοῦσι,
Γηθοσύνην καλέοντες ἐπώνυμον ἠδ’ Ἀφροδίτην·
And Love among them, equal in length and breadth.
See her with your mind, and do not sit stunned by your eyes;
She who is also considered to be inborn in mortal limbs,
By whom they think thoughts of love and accomplish fitting deeds,
230
I follow Kingsley (1995) 13–48 in seeing Zeus as the air, Hera as the earth and Hades as fire (with
Nestis more obviously standing for water); an interpretation also followed by Willi (2008) 227–8.
Cerri (2001) agrees that Hera is the earth but argues that Zeus is fire and Hades is the air. The
ancient evidence is divided: Diogenes Laertius (8.76) and Hippolytus (Haer. 7.29.4) claim that
Aidoneus (Hades) is Air, Hera Earth and Zeus Fire; an alternative identification going back to
Theophrastus and adopted by the Stoics treats Aidoneus as Earth and Hera as Air (see Aët. 1.3.20;
Philodem. Piet. 2, 63; Cic. Nat. D. 2.66, Plut. De Is. et Os. 363d; Men. Rhet. 1.5.2).
231
The fragment is understood as allegorical in this sense by Primavesi (2005), (2008a).
232
On Theagenes and the Derveni commentator, see Introduction Section 2 above.
172 Empedocles
Calling her Joy by name and Aphrodite.
Mortals refer to the feeling in their bodies as ‘Joy’ or ‘Aphrodite’, when in
fact, it is the non-anthropomorphic cosmic Love.233 Nevertheless,
Empedocles will repeatedly refer to cosmic Love using the mortal name
‘Aphrodite’. An explanation of this sort of practice, this time with reference
to the mortal terms for ‘birth’ and ‘death’, occurs in D53=B8 and D54=B9:
ἄλλο δέ τοι ἐρέω· φύσις οὐδενὸς ἔστιν ἁπάντων
θνητῶν, οὐδέ τις οὐλομένου θανάτοιο τελευτή,
ἀλλὰ μόνον μίξις τε διάλλαξίς τε μιγέντων
ἔστι, φύσις δ’ ἐπὶ τοῖς ὀνομάζεται ἀνθρώποισιν.
And I will tell you another thing: there is no creation of any at all
of mortal things, nor is there any end of accursed death,
but there is only mixture and exchange of the mixed things
but creation is the name given to them by mortals. (D53=B8)
οἱ δ’ ὅτε μὲν κατὰ φῶτα μιγέντ’ εἰς αἰθέρ’ ἵ<κωνται>

ἢ κατὰ θηρῶν ἀγροτέρων γένος ἢ κατὰ θάμνων
ἠὲ κατ’ οἰωνῶν, τότε μὲν τὸ <λέγουσι> γενέσθαι,
εὖτε δ’ ἀποκρινθῶσι, τὸ δ’ αὖ δυσδαίμονα πότμον·
ἣ θέμις <οὐ>234 καλέουσι, νόμωι δ’ ἐπίφημι καὶ αὐτός.
And they, when these things into the air mixed together in the form of men,
or in the race of wild beasts or of bushes
or of birds, then that being born,
but when they disintegrate, they call that ill-starred fate.
They do speak rightly, but I myself assent to the custom. (D54=B9)
In D53=B8 mortals typically use φύσις, ‘birth’ or ‘creation’ for what is in

fact μίξις, or ‘mixture’. We can also assume that the ‘end of death’
(θανάτοιο τελευτή) is a conventional ‘name’ applied by mortals for what
is really ‘exchange’ of the elements that have been mixed. The difference
between the mortal ‘name’ and what these phenomena actually ‘are’
implies that the mortal terminology is in some way inadequate, but this
situation is further complicated by D54=B9. I reproduce above the lacu-
nose fragment with its most commonly accepted supplements, according
to which the mortal naming practices are ‘not right’. This negative charac-
terisation is far from certain, but in any case, the narrator ‘assents’ to the
233
Cf. Xenophanes D39=B32. That the shape of cosmic Love is unlike humans is suggested by
D75=B35.
234
This is Wyttenbach’s supplement, which has been accepted by most editors (including Diels-
Kranz), but Bollack (1965–9) supplements γε, rendering a positive sense, and has now been
followed by LM.
convention of referring to death as δυσδαίμονα πότμον, ‘ill-starred fate’,
even though the process is actually a more specific one of ‘disintegration’.
As we saw with Parmenides’ poem, the difference in terms reflects
a distinction between divine and mortal perspectives. The Empedoclean
narrator is himself a god of sorts (D4=B112.4; D60=B23.11) who must
contend with a mortal audience (D42=B2; D44=B3.6–7), and the conven-
tional terms are specifically those applied by mortals (ἀνθρώποισιν,
D53=B8.4).235 These fragments thus develop the traditional distinction
between divine and mortal language.236 The divine narrator uses conven-
tional, mortal terms in spite of the fact that they may obscure the actual
nature of the phenomena to which they refer: ‘death’, for Empedocles,
given his doctrine of metempsychosis, is not really the end of conscious
existence as the term may imply. But by interpreting the text in the right
sort of way and identifying the processes that underlie these terms, we can
acquire a divine understanding of the world.
Given this explanation of his use of mortal terms and given that
‘Aphrodite’ is specified as such a term in D73, we can reasonably infer
that the names of the gods in D57=B6 are used in a similar manner: as
misleading words that mortals conventionally use to refer to certain
phenomena, but which Empedocles will nevertheless use himself. To
regard this usage as allegorical or figurative is somewhat misleading.
Mortal names contain an element of truth that can be accessed through
correct interpretation, but it is one of which the mortals themselves are
unaware. In Empedocles’ cosmos, the traditional, anthropomorphic names
for the six permanent entities are inaccurate insofar as they imply a human-
like shape, but not insofar as they imply a human-like personality.237
According to D257=B110.10, ‘all things have intelligence and a share in
thought’ (πάντα . . . φρόνησιν ἔχειν καὶ νώματος αἶσαν). The elements,
then, like all things in Empedocles’ universe, have mental faculties. Thus,
at D75=B35.6, they come together ‘willingly’ (θελημά, D75=B35.6).238
Aphrodite – who is in fact, the cosmic force of Love – is said to ‘fit things
together’ (D61=B71.4); she ‘fashioned eyes’ (D213=B86, ἔπηξεν), and ‘fitted
them together with bolts of affection’ (D214=B87, γόμφοις ἀσκήσασα
235
On the implicit division between divine and mortal language in Empedocles, see Inwood (2001)
37–8 and especially Willi (2008) 230–63.
236
Chapter 2 n. 19 above.
237
In addition to the anthropomorphic names for the four elements at D57=B6 and the references to
Love as ‘Aphrodite’ (D101=B22.5; D159=B66.1; D213=B86.1; D214=B87.1; D73.255=B17.24) or
‘Kypris’ (D25=B128.2), Strife seems to be referred to as ‘Ares’ at D25=B128.1.
238
For the four elements as personal deities, see e.g. Rangos (2012) 319 and Rowett (2016) esp. 84–8,
who argues that the elements are divine, personal agents. Note also Arist. Gen. corr. 2.6.333b20–2.
174 Empedocles
καταστόργοις). As Kypris, she ‘moistens earth in water’ and ‘gives it to fire
to rule over’ as she ‘busies herself over forms’ (D199=B73).239 There is such
a thing as Aphrodite, it is just that she is in fact the cosmic force of Love.
The divine names are not the only characteristically poetic feature that
Empedocles invites us to reinterpret in light of his cosmology. Often, it
seems to be specifically poetic expressions that are given a new significance:
the ‘mortal’ expressions in D53=B8 and D54=B9, ‘end of death’ (θανάτοιο
τελευτή) and ‘ill-starred fate’ (δυσδαίμονα πότμον), are rather poetic
locutions (cf. [Hes.] Sc. 357; Soph. OT 1302) that are replaced with the
more prosaic διαλλαξίς, so that we are encouraged to interpret the poetic
language in terms of the ‘scientific’ cosmology. Traditional hexameter
epithets are deployed in new contexts, illustrating the common origin of
humans, animals and plants.240 Trees are given the personifying epithet
‘lovely-haired’ (δένδρεσιν ἠυκόμοισιν, D36=B127.2), usually used of god-
desses and ladies (Il. 1.36; Hes. Theog. 241; Pind. Ol. 6.91 – they also ‘lay
eggs’, ὠιοτοκεῖ, D254=B79). As we noted above, pomegranates are referred
to as ‘late-born’ (ὀψίγονοι, D255=B80) because their fruit appears late in
the year, but the epithet is used by Homer of humans (Il. 3.353, Od. 1.302).
The narrator describes the mind of his addressee as ἔλλοψ in D258=B5, an
epithet of disputed meaning normally used of fish (e.g. Hes. [Sc.] 212,
either meaning ‘scaly’ or ‘dumb’, see LSJ s.v.). What at first may appear as
a figurative metaphor, then, highlights the common origins of tenor and
vehicle, for example, of humans and plants.241 Even characteristically epic
formal features acquire a new meaning. As is well known, Empedocles self-
consciously uses repetitions, a salient feature of orally derived epic, to
illustrate the iterations in the cosmic cycle (D73>B17, D45=B25), and his
extended similes, another distinctive epic technique, are in fact analogies,
demonstrating common underlying processes for both familiar and
unfamiliar phenomena (D60=B23; D215=B84; D201=B100).242
Empedocles defamiliarises the world of the audience’s experience by pre-
senting it in the forms and language characteristic of mythical epic. The
239
Note also D101=B22. 5, things are ‘made similar’ (ὁμοιωθέντα) by Aphrodite. See Sedley (2007) 52f.
for a strongly personal reading of Love. Contrast Curd (2013a).
240
As Gemelli Marciano (1990) 36–7, 52–4, 85, 96 and Rosenfeld-Löffler (2006) 18–21 demonstrate.
241
A point made by Rosenfeld-Löffler (2006) 18–21.
242
Hershbell (1968) catalogues the repetitions and traditional epithets, seeing them as evidence that
Empedocles was an ‘oral poet’; whilst it can scarcely be denied that Empedocles’ poetry was
designed for oral performance, this argument misses the point that these traditional features
illustrate particular features of the cosmology and so are not merely an expedient for oral compos-
ition. On Empedocles’ mimetic use of repetition, see Graham (1988) 303–7. On the similes, see esp.
O’Brien (1970) and Gheerbrant (2017) 271–386.
Conclusions: Empedocles in the History of Greek Poetry and Poetics 175
fact that the content of his poem is not the distant world of epic but,
among other things, the familiar world of the audience invites them to
apply an interpretation to the world that is analogous to their interpret-
ation of the poetry.
6 Conclusions: Empedocles in the History of Greek Poetry

and Poetics
In Section 5 above I have tried to show that Empedocles’ use of verse is not
a feature that can be explained by a single cause and then swiftly discarded.
Rather, his text works as poetry in a more essential sense. It is designed to
have a particular pragmatic effect on the audience, through the emotional
force of its narratives and imagery and through the penetrating, subtextual
interpretation it provokes. The ‘truths’ of Empedocles’ poetry are not
always to be identified with the obvious meaning of the text: some of the
propositional statements – that the ‘palace of Zeus’ receives the exile
(D12=B142) or even that things ‘are born’ (D54=B9) – are not to be
taken at face value. In self-consciously exploiting the distinction between
the literal sense of the text on the one hand and the underlying cosmology
on the other, Empedocles approaches an articulation of the notion that
poetry creates a self-contained image of the world, analogous to the visual
arts (cf. D60=B23) and valued, not for its historical accuracy, but for its
symbolic meaning and emotive force.
We saw a similar distinction at work in Parmenides’ division between
the world of the Doxa and the world of the Aletheia. Although
Empedocles grants more authority to the world of the senses than
Parmenides had done, he maintained the principle that the objects of
secure, divine knowledge are unchanging and small in number. Both
authors thus isolate the world of the senses as distinct from secure reality
but interpretable in such a manner as to lead to it. This epistemological
point foreshadows (and in Parmenides’ case directly influenced) Plato’s
theory of the forms, a theory which would provide the epistemological
basis for his attacks on poetry as a mimetic representation, ‘at third
remove from that which is’ (τριττὰ ἀπέχοντα τοῦ ὄντος, Resp. 10.599a).
The fact that Parmenides and Empedocles make this epistemological
point in verse and, more specifically, the fact they depict the world of
the senses as analogous to the world of poetry suggest that they also
regarded their theories of knowledge as having consequences for the
nature and function of poetry. In short, Parmenides and Empedocles
foreshadow Plato in treating an epistemological issue as an issue of
176 Empedocles
poetics. In Empedocles’ case, a further point of overlap with Plato is the
conceptualisation of poetry as an artisanal product, analogous to paint-
ing. Yet in contrast with both Parmenides’ Doxa and the Platonic views
expressed in the Republic, Empedocles stresses the positive dimensions of
both the world of the senses and of poetry as means to guide us to the
unchanging principles that underlie them. In stressing these positive
aspects of poetry and, in particular, exploiting its emotional force for
didactic ends, Empedocles arguably comes closer to the Aristotelian
conception of poetry as an emotionally beneficial source of philosophical
truths (cf. Poet. 9.1451a38–1451b4).
Most accounts of the rise of the Classical conception of poetry attribute
the development of the theory of literary mimesis to the rise of a literate
culture in the fifth and fourth centuries bce.243 More cautiously, we might
say that literacy facilitated such a shift.244 But a further factor that ought to
be taken into consideration is the epistemological achievement of the
Presocratics, especially Parmenides. In the Epilogue, we shall see that
certain key literary-critical developments in the fourth century bce were
inspired by the Presocratic poets. But first, let us draw together into
a conclusion some of the threads that have been running through these
three case studies.
243
Thus, the works listed at n. 83 above. In the most detailed study of the ancient concept of mimesis,
Halliwell (2002) stresses the deeper roots of this notion in earlier Greek literature (15–22) but points
out that ‘there is little surviving evidence for the theorising of mimesis before Plato himself’ (15).
Nagy (1989) 47–51 similarly stresses the continuity, regarding ‘re-enactment’ as the consistent core
sense.
244
Thus Ford (2002) 155.
Conclusion
My central aim has been to enrich our literary appreciation of these

sometimes-maligned poems. I see the main value of this project as residing
in the specific readings it has generated. The proof of this pudding will
have been in the eating. If I have convinced any readers that these texts are
more interesting literary artefacts than previously thought, then I am
sufficiently rewarded. But in the course of pursuing this overarching aim,
some general points have emerged.
It has been a running hypothesis that these authors were not restrained
by their medium but rather masters of it. Their use of verse cannot be fully
explained by recourse to their circumstances of production. It may well be
true, at least of the cultural context in which Xenophanes and Parmenides
operated, that prose was not yet a viable means of publication. But this
explanation can only tell us so much. The same thing could be said of
whoever composed the Homeric poems, but it would hardly provide an
adequate account of the significance or impact of the Iliad or Odyssey. Even
in a purely oral culture, verse is a distinctive speech-genre with particular
connotations, so that it does not only serve as a mnemonic or a means of
publication. At least one further reason for the Presocratic poets’ use of
verse is that it enabled them to manipulate, artfully and self-consciously,
the traditions of elegiac and hexameter poetry. One stimulating possibility,
in line with recent trends in literary studies that have drawn on the findings
of cognitive science, is that the use of particular narratives, images, and
literary genres helped these authors to forge new concepts.1 But my focus
1
For an example of literary criticism that explores this possibility, see Lyne (2011). The idea that
metaphors determine the arguments of Western philosophers (including the Presocratics) is argued
by Lakoff and Johnson (1999). Variants of the idea that literary genres determine the conceptual
content of philosophical texts have been argued by Marías (1971); Jordan (1981); Lang (1990) 24–44;
Ferrell (2002); and Lavery (2007). See also Mourelatos (2008a) on Parmenides, whose approach is
discussed in Chapter 2 Section 1 above.
177
178 Conclusion
has been not so much on the author’s mindset during composition as on
the other side to this coin, the effect of these features on their recipients.
Like other Archaic and Classical Greek verse, these texts are designed to
elicit emotional and cognitive responses from the audience that are contin-
gent on their particular form and phrasing. In this respect, they are not
mere verse, but count as poems in an essential sense. Xenophanes, by
presenting his god in the highly stylised Kunstsprache of hexameter verse,
imbues him with a grandeur and reverence befitting traditional anthropo-
morphic deities. Parmenides, in his vividly immersive proem, takes us on
a frightening yet enthralling trip beyond the boundaries of the familiar
universe. Empedocles confronts us with the terrifying image of a father
unwittingly sacrificing and eating his child, but also the sublime drama of
the cosmic cycle in all its workings. Through their use of suggestive
imagery and their comments on the limitations of mortal names,
Parmenides and Empedocles prompt a critical exploration of the true
underlying significance of their language. In doing so, they capitalise on
the characteristic tendency of hexameter poetry to invite a symbolic
interpretation.
These suggestions can help to add some nuance to the contested
question of whether or not verse plays an essential philosophical role in
these texts.2 One problem is that it is not always clear what is meant by
philosophy in this context since, as applied to these authors, the term
denotes an ‘etic’ concept, one used by the interpreter, but unfamiliar to the
culture under investigation.3 Certainly, we can extract philosophical claims
from the strict, semantic content of the language, translate them into
formal logic and evaluate their coherence, validity and veracity.4 If we
define ‘philosophy’ as this paraphrasable content, then the use of verse, by
definition, does not play an essential philosophical role. But, on my
account, this sort of approach misses at least two important functions of
the texts in question. First, if I am right in treating them as symbolic, they
imply claims that they do not make overtly. Xenophanes implies that his
god is worthy of veneration. Parmenides and Empedocles both imply that
2
Granger (2007) argues explicitly that verse plays no essential philosophical role in Xenophanes;
Barrett (2004) and Gheerbrant (2017) argue that it does, respectively, in Parmenides and
Empedocles, though on different grounds to my own. Barrett sees the difficulties of Parmenides’
verse as challenging us to reflect on the nature of linguistic reference, whilst Gheerbrant regards
Empedocles’ adaptation of traditional forms of Muse-inspiration as crucial to his message.
3
On the issue of the respects in which the Presocratics may or may not be considered philosophers, see
Laks (2018) 35–52. On the history of the term philosophia, see Burkert (1960).
4
This procedure is standard in work on the Presocratics but meets its apex in the work of Barnes
(1982).
Conclusion 179
the acquisition of knowledge is somehow analogous to mystic initiation.
Of course, such implied claims can never be identified with certainty, but
an interpretation that does not acknowledge their possibility is substan-
tially incomplete. Secondly, as theorists have often pointed out, the essen-
tial value of a work of literature is not reducible to any message it may
impart.5 In arguing that these texts are true poems, I am claiming that they
count as literature in this sense, so that if we read them purely for their
‘message’, we neglect the affective experiences they provide.
Both these functions may be regarded as playing an important philo-
sophical role, the first more obviously so, since it is uncontroversial that
philosophy is concerned with propositional claims. The second may be
regarded as philosophical in a more subtle manner. If the experience of
these texts as literature demands a critical reflection on the meaning of
certain details, this could be characterised as a form of dialectic.6 Most
saliently, the obscurity of Parmenides’ use of esti may be part of the point,
since it gets the audience thinking about the meaning of being. So, the use
of verse is a way of inducing an active engagement with philosophical
questions. But the affective experience of these texts may also be character-
ised as philosophical in a further respect. Some theorists have argued that
literature is essentially concerned with a kind of knowledge that is non-
propositional.7 Through literature, we learn what it is like to view the
world from a particular perspective, to hold certain values and to feel
certain emotional responses. I have argued that the Presocratic poems are
designed to have these effects: they grant an experience of what it is like to
view the world in a certain way, a way that may be more appropriate to true
reality than habitual mortal attitudes. Of course, this theory of literary
knowledge is not explicitly articulated as such until modern times, but the
general idea is already implicit in Archaic depictions of the engrossing,
emotional powers of song and in Plato’s concerns over the ways in which
poetry entrenches certain values. If we count this type of ‘literary knowing’
as philosophical – this knowing how it feels to view the world from
a certain, value-laden perspective – then the poetic quality or ‘poeticity’
of these texts plays an essential philosophical role. The use of verse does not
guarantee the quality of poeticity, but it does indicate that this is a quality
to which they aspire.
At the same time, the explicit propositional claims made can shed light
on the ultimate purposes of these affective experiences. Though scanty, the
5 6
E.g. Sontag (1966); Cave (2016) 149–51. Thus, on Parmenides, Mackenzie (1982).
7
See Introduction Section 2 above.
180 Conclusion
surviving Xenophanean fragments that describe his singular god, in light of
his criticisms of Homer and Hesiod, serve the end of entrenching more
beneficial theological attitudes as a counterpoint to the morally pernicious
verses of canonical hexameter poetry. Parmenides’ engrossing proem, in
light of the goddess’s stringent conditions on true being and her introduc-
tion to the Doxa, illustrates the deceptiveness of sensory experience.
Empedocles portrays a tragic sacrifice to get us to refrain from killing
and eating animals. More subtly, his emotive, first-person narrative of
the daimon in exile encourages feelings of kinship with non-human organ-
isms that can be explained in terms of the cosmology. Taken together, the
explicit semantic content, the symbolic connotations activated by formal
aspects as well as by intertextual and intratextual associations, and the
cognitive and emotional impact of a performance are all directed towards
the twin purposes of communicating that which would normally be
ineffable and of bringing the audience from one set of beliefs to another.
There is every reason to think that the Presocratics’ use of verse for these
purposes was not unreflective or purely instinctive but quite deliberate. In
Xenophanes’ command that ‘these things be thought like true things’
(D50=B35), in the instruction of Parmenides’ goddess that we ‘judge by
reason’ her ‘refutation’ (D8.5=B7.5) and in Empedocles’ suggestion that we
‘plant’ his teachings under our ‘organs of thought’ (D257=B110.1), the
Presocratic poems feature meta-didactic comments that bespeak
a rumination on the nature and function of verse. Their allusions to one
another and, especially, to previous Muse-inspiration episodes form part of
a discussion surrounding the relationship between verse and truth which
goes back to Homer and Hesiod.
Through their poetic practice, they presuppose certain claims about poetry.
Rather than being straightforward Archaic ‘masters of truth’, they manipulate
the capacity of poetry to transport its audience into a vividly imagined yet
potentially deceptive mimetic world. Parmenides and Empedocles both iso-
late a small number of objects that can be the object of true knowledge – be
that singular true being or the six permanent entities – in relation to which the
epistemic status of poetry can be assessed. By presenting the world of the Doxa
in terms appropriate to poetic artifice, Parmenides emphasises the distance
between poetry and true reality. Empedocles, by contrast, valorises poets. His
poetry is an artisanal product, but although it is, in this respect, artificial, it can
nevertheless act as a conduit to knowledge of universal truths. Parmenides and
Empedocles are thus transitional figures in the movement towards an articu-
lation of the Classical conception of poetry, whereby poems are mortally
crafted, self-contained artefacts that present true claims only symbolically. The
Conclusion 181
claim that the objects of secure knowledge should be fundamentally unchan-
ging facilitated later articulations of the notion that poetry symbolically
represents such objects, whether through ainigmata or mimesis. This point
will be substantiated in the Epilogue, which will trace some points of influence
by the Presocratic poets on certain important developments in fifth- and
fourth-century Greek poetics.
A more specific bestowal to European literature is in the influence of
these authors on the genre of didactic poetry. Empedocles’ importance for
later poetry is now widely acknowledged and is most pervasive in the De
rerum natura. But his work would not have been possible without the
precedents of Xenophanes and especially Parmenides. In addition to the
didactic form and the use of hexameters, at least four features that would
become characteristic of the didactic genre are already evident in one or
both of the two earlier poets and arise from the challenge of bringing an
audience to a radically different understanding of the universe. First, their
meta-didactic comments, serving to guide the audience past the limitations
of mortal language, are instances of what Katherina Volk has termed
‘poetic self-consciousness’, a hallmark of Latin didactic poetry.8 Second,
for the three Presocratic poets, the interpretation of parts of the text is
analogous to the interpretation of the world: we must penetrate beneath
the misleading habitual sense of the language, just as we must look beyond
our potentially specious sensory experience. This text–world analogy
would recur in several prominent didactic poems: Aratus’ acrostic at
Phaenomena 783–7 reflects the way that constellations are hidden in the
stars;9 Lucretius draws an analogy between letters of his poem and the
elements (e.g. 1.196–7, 823–7, 912–14);10 Manilius, like Aratus, employs
acrostics in such a manner as to suggest that his text is analogous to the
cosmos it describes (e.g. 1.813–18).11 Third, Empedocles follows Parmenides
in presenting the process of education as a sort of mystic initiation,
a ‘didactic plot’ that would recur in both the De rerum natura and
Virgil’s Georgics. Initiation acts as a familiar cultural model for distinguish-
ing those ‘who know’ an ineffable, transformative truth from the ignorant.
Fourth, my analysis has drawn upon, and corroborated, Conte’s conten-
tion that Lucretius’ sublime aesthetic is part of Empedocles’ influence.12
Yet such an aesthetic is already utilised by Parmenides, who presents the
process of education as an extraordinary journey that culminates in
8
Volk (2002) 9–10, although she denies that Parmenides displays this feature (50).
9 10
See Hunter (1995). See Friedländer (1941); Snyder (1980); Dalzell (1987).
11 12
Colborn (2013). See Chapter 3 Section 5b above.
182 Conclusion
a confrontation with an awesome divinity. In both authors, sublimity
stems from an attempt to inculcate a conception of the world according
to which generally accepted beliefs are erroneous: the elevating emotional
trajectory conveys what it feels like to transcend the limitations of habitual
mortal thoughts.
A more detailed exploration of the relationship between the Presocratic
poets and later didactic poetry would require a separate study, but even
from this brief overview it is apparent that, through using hexameter poetry
to promote a radically new conception of the world, they expanded the
resources of the medium in a manner that would have a lasting impact,
affecting some of the most influential poems of antiquity. In addition to
their important role in the history of philosophy, Xenophanes, Parmenides
and Empedocles deserve a prominent place in the histories of Greek poetics
and ancient literature.
Epilogue: The Legacy of Presocratic Poetics
1 Introduction
To add some substance to my claim that the Presocratic poets played an
important role in the development of the Classical conception of poetry,
this epilogue shall highlight certain important developments in fifth- and
fourth-century bce poetics that they directly influenced. To be clear, these
are developments in the explicitly articulated theory, rather than the
practice, of poetry. The concepts of ainigmata and literary mimesis can
fruitfully be applied to the earliest surviving Greek poetry, and there is
some evidence to suggest that early poets had a sophisticated awareness of
the features covered by these labels.1 Nevertheless, later on, we find more
systematic and explicit formulations of the idea that poetry conveys pro-
found truths symbolically, or that, along with the visual arts, it is a mimetic
form that offers a representation of the world. Xenophanes, Parmenides
and Empedocles greased the wheels of this process.
2 Anaxagoras, Allegory and Ainigmata

The term allegoria is first attested in Cicero (Orat. 27.92–4), and first
attested in Greek in Plutarch (Mor. 19E–F), but ‘allegory’ has come to be
used by modern scholars for what fifth-century bce authors refer to as
ainigmata, a term also translatable as ‘riddles’.2 As we observed in the
Introduction (Section 2), ainigmata convey hyponoiai, or ‘hidden mean-
ings’, but these need not be ‘allegorical’ in the later sense that they are the
tenors of extended metaphors. The meanings can be hidden by any means,
1
See Most (1993b) for a possible allegorical interpretation in Homer.
2
On the development from ainigmata to allegoria, see Reinhardt (1960) and, more recently, Struck
(2004).
183
184 Epilogue: The Legacy of Presocratic Poetics
for instance, through the use of misleadingly vague terms, or through
ambiguous syntax.3 In keeping with scholarly convention, I shall refer
here to the understanding of poems as ainigmata in this sense as ‘allegor-
ical’, even though the later term has certain inapplicable connotations.
As we noted in Chapter 1, an orthodox scholarly view holds that the
allegorical exegesis of poetry originated with the shadowy figure of
Theagenes of Rhegium, as a defence of the Homeric poems against the
criticisms of Xenophanes. If this view is correct, then we have an obvious
case in which a Presocratic poet inspired a literary-critical development.
But Theagenes is only lately attested, we lack any verbatim fragments, and
Porphyry, after explaining the allegorical interpretation of the theomachia
(whereby Apollo, Helios and Hephaestus are fire; Poseidon and Scamander
are water; Artemis is the moon; and Hera is the air) only says that this
‘manner’ (tropos) of interpretation is very ancient and goes back to
Theagenes (ap. Schol. B ad Il. 20.67). This is rather flimsy evidence on
which to posit a major development in the history of Greek criticism.
What is more, allegorical interpretative practices were also attributed to
Pherecydes of Syros, who may have been earlier than Xenophanes and, in
any case, the conception of songs as enigmatically harbouring profound
truths is already evident in the earliest poems themselves. ‘Riddles’ have
been seen as a feature of Proto-Indo-European poetry.4 A more plausible, if
weaker, hypothesis is that Xenophanes’ attacks provided an impetus for
Theagenes’ specific type of allegorical defence which, if the reports are
accurate, understood the Iliad in terms of a Presocratic cosmology. If
Homer’s gods are not really fighting each other, then he may be less
vulnerable to the charge of portraying them in an unbefitting and impious
manner.
With greater confidence, we can say that the Presocratic poets influ-
enced the manifestation of this sort of interpretation among the fifth-
century Sophists.5 Anaxagoras of Clazomenae seems to have been a key
figure in this development. He was credited with bringing Ionian philoso-
phy to Athens (Clem. Strom. 1.63=P13<A7) and was most famous for being
exiled for claiming that the sun was a fiery mass (Diog. Laert. 2.12=P23<A1).
Even if this story is a later fabrication, influenced by the trial of Socrates,6 it
suggests a reputation as a denier of traditional mythological explanations
for natural phenomena. When this reputation is taken in combination
3 4
See further Richardson (1975) 66–7 and Ford (2002) 72. West (2007) 363–72.
5
On whom, see Richardson (1975) and Morgan (2000) 89–131 esp. 98–101.
6
Thus, Dover (1975) 31–4.
Anaxagoras, Allegory and Ainigmata 185
with the testimony that he was the first to claim that the ‘poetry of Homer
is about virtue and justice’ (Favorinus fr. 61 Barigazzi ap. Diog. Laert.
2.11=D98<A1), he starts to look like an allegorist (in the broad sense),
someone who argued that earlier mythological poems were not simply
about their ostensible subject matter (say, the Trojan war or the return of
Odysseus), but about the immutable truths sought by philosophers.7
Furthermore, he is credited with inspiring later allegorical interpreters.
According to Diogenes, he taught the notorious allegorist Metrodorus of
Lampsacus (Diog. Laert. 2.11=D98<A1). He is also widely acknowledged as
the principal source of inspiration for the cosmology that the Derveni
commentator sees encoded in the Orphic theogony.8
The sort of cosmological allegoresis of Metrodorus and the Derveni
commentator was clearly in the tradition of Presocratic cosmological
speculation and the fact that Xenophanes, Parmenides and Empedocles
wrote in hexameters may in particular have been an influencing factor. As
we have seen, Empedocles uses traditional divine names in an allegorical
(in this sense) manner. More tentatively, in one case, reported in Plato’s
Timaeus 22c–d, the myth of Phaethon is allegorised as a deviation in the
heavenly bodies that caused huge fires on the earth, perhaps under the
inspiration of Parmenides’ use of the same myth (see Chapter 2 Section 5b
above). A more specific connection is Parmenides’ influence on
Anaxagoras. Like Empedocles’ elemental theory, Anaxagoras’ cosmology
is an attempt to defend the existence of the pluralistic cosmos whilst
accepting the Parmenidean principle that nothing can be created ex nihilo
or destroyed in nihil (Anaxagoras D15=B17): ‘there is a portion of every-
thing in everything’ (D25=B6) except ‘mind’ (nous, D26=B11; D27=B12),
and each thing is ‘most manifestly’ that thing of which it contains the most
(D27=B12). In other words, what we perceive as gold contains within it
portions of all stuffs, but it contains more gold than anything else, so the
manifest substance appears to us as gold. For Anaxagoras, then, we can
distinguish the manifest from the pure type of any given stuff: the manifest
stuff – say, ‘gold’ – is what we perceive, but it actually contains portions of
all pure stuffs; the portions of pure gold it contains are greater in mass than
the portions of pure silver, or hair, or flesh, or any other stuff within it.
Creation and destruction are, then, explicable as rearrangements of these
putative ‘pure’ stuffs, just as, for Empedocles, they are explicable as
7
In support of this view, see Sider (2005) 61–6.
8
First argued by Burkert (1968), and since refined by e.g. Betegh (2004) 278–305. See also, on this
point, Piano (2016) 292–7.
rearrangements of the four elements. This point of similarity is marked by
the remarkably similar claims made by Anaxagoras D15=B17 and
Empedocles D53=B8 and D54=B9.9 This similarity is probably best
explained on the grounds that both authors are responding to
Parmenides.10 Whatever the exact nature of the relationship, an intellectual
genealogy can confidently be drawn between Parmenides and Anaxagoras,
a key figure in the rise of Sophistic allegorical interpretation.
One can speculate further that this connection with Anaxagoras
extended to the realm of poetic interpretation in the ‘Anaxagorean
School’.11 If Metrodorus did indeed read Homer as explicating
Anaxagoras’ cosmology, one of the cosmological details he identified in
Homer was that the moon shines with light reflected from the sun (cf.
Anaxagoras D40=B18). Parmenides seems to have been the first to discover
this phenomenon (D27=B14; D28=B15).12 He thus memorably describes
the moon as ‘ἀλλότριον φῶς’ (D27=B14) or ‘foreign light’, an expression
that was taken up by Empedocles (D139=B45) and which itself was an
adaptation of the Homeric formula ἀλλότριος φώς, ‘foreign person’ (Il.
5.214; Od. 16.102, 18.219). Parmenides, then, in his hexameter poem, adapts
a line describing a person to describe the moon; Metrodorus sees an Iliadic
character as in fact representing to the moon; and both Parmenides and
Metrodorus see the underlying point as being that the moon shines with
light reflected from the sun. It is tempting to suggest that Metrodorus’
allegorical interpretation of the Iliad was not only influenced (directly or
indirectly) by Parmenides’ astronomical discovery, but also by his own
practice of appropriating Homeric language to describe physical phenom-
ena in a personifying manner. Whilst such a claim is of course highly
speculative, it is hard to deny that Parmenides was a major influence on the
wider intellectual context in which Metrodorus was active, and his poem
could have provided an inspiration for these sorts of interpretation.
9
Note e.g. the vocabulary of ‘mixture’ and ‘separation’ and the emphasis on the language used:
Anaxagoras D15=B17: ‘The Greeks do not consider creation and destruction correctly; for no thing
is created or destroyed, but, from the things that are, is mixed together (συμμίσγεταί) and separates
(διακρίνεται). And thus they would correctly call (καλοῖεν) creation mixture and destruction
separation’; Empedocles D53=B8: ‘And I will tell you another thing: there is no creation of any at
all / of mortal things, nor is there any end of accursed death / but there is only mixture (μίξις) and
exchange (διάλλαξις) of the mixed things / but creation is the name given (ὀνομάζεται) to them by
mortals.’
10
See e.g. Graham (1999). The relative chronology of Anaxagoras and Empedocles is controversial, due
to the ambiguity of Aristotle’s testimony at Metaph. 1.984a11–13. See O’Brien (2005) 319–21.
11
For the notion of an Anaxagorean ‘school’ associated with allegorical interpretation, note Syncellus
Chron. 140C=DK61A6 and see Richardson (1975) 70; Morgan (2000) 98.
12
On this point, see Mourelatos (2012).
Anaxagoras, Allegory and Ainigmata 187
In at least one case, the distinctively Parmenidean epistemological
distinction between aletheia and doxa seems to have been invoked in the
explication of Homer, by the late fifth-century associate of Socrates,
Antisthenes (fr. 194 Prince=Dio Chrys. Or. 53 On Homer §5):
ὁ δὲ Ζήνων οὐδὲν τῶν τοῦ Ὁμήρου ψέγει, ἅμα διηγούμενος καὶ διδάσκων
ὅτι τὰ μὲν κατὰ δόξαν τὰ δὲ κατὰ ἀλήθειαν γέγραφεν, ὅπως μὴ φαίνηται
αὐτὸς αὑτῷ μαχόμενος ἔν τισι δοκοῦσιν ἐναντίως εἰρῆσθαι. ὁ δὲ λόγος
οὗτος Ἀντισθένους ἐστὶ πρότερον, ὅτι τὰ μὲν δόξῃ, τὰ δὲ ἀληθείᾳ εἴρηται
τῷ ποιητῇ. ἀλλ’ ὁ μὲν οὐκ ἐξειργάσατο αὐτόν, ὁ δὲ καθ’ ἕκαστον τῶν ἐπὶ
μέρους ἐδήλωσεν. ἔτι δὲ καὶ Περσαῖος ὁ τοῦ Ζήνωνος κατὰ τὴν αὐτὴν
ὑπόθεσιν γέγραφε καὶ ἄλλοι πλείους.
Zeno blames none of the verses of Homer, but in the course of setting them
out in detail he teaches that Homer has written some according to opinion
and others according to truth, in order that he not appear to be contradict-
ing himself in certain verses that seem to be said in opposition to each other.
This principle is previously from Antisthenes, that some things have been
said by the poet in opinion and some in truth. But he did not work it out,
whereas he [Zeno] showed it according to each of the parts. And further also
Persaeus pupil of Zeno has written according to the same hypothesis, and
many others. (Trans. Prince [2015])
Though it does not mention him by name, this source distinctly recalls
Parmenides. Dio’s language is very similar to that used by the doxographers
to describe the different parts of his poem. The expression ‘according to
opinion . . . according to truth’ (κατὰ δόξαν . . . κατὰ ἀλήθειαν) is used by
ancient commentators in reference to the philosopher’s work.13 Here, accord-
ing to one interpretation, the distinction is applied to the allegorical tenor and
its vehicle: what ‘Homer has written according to opinion’ refers to his
ostensible sense, which has to be decoded to yield his true meaning.14 This
view is supported by Antisthenes’ own creative use of myth to demonstrate
ethical points. Thus, he used Heracles as a paradigm for the ethical principle
that toil is good (fr. 85 Prince=Diog. Laert. 6.2), an example that would later
be mentioned alongside Prodicus’ more famous use of the hero (fr. 44 C
Prince=Julian Or. 7.216d–217b). The creative use of myth to illustrate certain
ethical points is a natural extension from the identification of ethical points in
canonical mythical texts and may strengthen the possibility that Antisthenes’
comments on ‘truth and opinion’ in Homer refer respectively to the
13
Theophrastus ap. Alexander, in Metaph. 31.9–14=R13<A7: οὐχ ὁμοίως περὶ ἀμφοτέρων δοξάζων,
ἀλλὰ κατ’ ἀλήθειαν μὲν . . . κατὰ δόξαν δὲ . . . Cf. also Simplicius in Cael. 7.556.12–14=R7<A14.
14
See Prince (2015) ad loc., who does not endorse this view. It is, however, supported by the fact that he
is associated here with Zeno, who was best known for his allegorical interpretations.
unchanging truths he communicates and the opinions in which he dresses
them, just as Parmenides’ Aletheia is the reality that we are to identify beneath
the world of the Doxa.15 On the other hand, recent commentators have tended
to reject the allegorical reading on the grounds that ‘truth’ and ‘opinion’ here
do not correspond to interpretative practices, but to parts of Homer’s
composition.16 Prince suggests instead that Antisthenes’ use of these terms
refers to a distinction between narrator-speech and character-speech, given
that he uses this distinction to explain an apparent inconsistency in the
depiction of the Cyclopes (fr. 189 Prince=schol. ad Od. 9.106):17 they enjoy
a land of spontaneous plenty in spite of the fact that at Od. 9.275 they are
‘heedless of aegis-bearing Zeus’, but Antisthenes points out that that line is
spoken by Polyphemus, not the primary narrator. If we accept this explan-
ation, the connection with Parmenides is still salient: Homer the omniscient
narrator becomes analogous to Parmenides’ authoritative goddess, whilst his
characters are like the heedless mortals.
In either case, Antisthenes’ use of the Aletheia/Doxa binary in the
context of Homeric exegesis, when taken alongside Parmenides’ influence
on Anaxagoras and the reports of Anaxagoras’ associations with allegorical
interpretation, amount to a cumulative case for Parmenides’ importance
for fifth-century Homeric ‘allegorical’ criticism. Indeed, Anaxagoras’ own
epistemology involved a distinction between the unseen, putative, ‘pure’
portions which alone occupy the category Aletheia and the manifest visible
world (D5=B21). One can speculate that this epistemological distinction
may have led to a distinction between the ‘true’ and the ‘apparent’
meanings of Homer’s language.
3 Gorgias
The fragments of Gorgias constitute some of the most important documents
in our history of fifth-century Greek poetics. His reported dictum that tragedy
is a ‘deception in which the deceiver is more just than the one who does not
deceive and the deceived is wiser than the undeceived’ (D35=B23) has been
lauded as the earliest valorisation of poetic deception.18 His Encomium of
15
Prince (2015) 668.
16
The allegorical reading is forcefully rejected by Tate (1953); see also Richardson (1975) 77; Prince
(2015) ad loc.
17
Prince (2015) 668.
18
As opposed to earlier pejorative comments on the deceptive capacity of poetry. E.g. Finkelberg
(1998) 177; similarly, de Romilly (1973) 160–1; Russell (1981) 22–4; Verdenius (1983) 29–30. Ford
(2002) 172–3 is more cautious, suggesting that the Gorgias’ point is primarily to showcase an ethical
paradox. Halliwell (2011) 267–84 argues that for Gorgias the deceptiveness of poetry is compatible
Gorgias 189
Helen (D24=B11) contains the earliest attested definition of poetry as versified
speech (§9) and also gives articulation to the notion that poetry causes the soul
of the listener to be affected emotionally by the affairs of others. Famously, as
we have had reason to observe already, this passage stipulates the characteristic
responses to poetry as pity and fear, the emotions that Aristotle would
canonise as the appropriate response to tragedy. Gorgias’ epistemological,
physical and metaphysical claims show the clear influence of Parmenides and
Empedocles. My contention is that this influence also encompasses his claims
pertaining to poetic theory.
Significantly for our purposes, Gorgias was known as the student of
Empedocles (Diog. Laert. 8.58–9=P5=A3; Olympiodorus in Gorgiam
Prooem. 9). Such teacher–student relationships in ancient biographies
are often met with scepticism, since the biographers have a tendency to
draw connections on little evidence,19 but chronological and geographical
considerations permit this possibility as both came from Sicily and were
born in the early part of the fifth century bce.20 Even if we choose to
disbelieve the story, it indicates that ancient readers saw a connection
between the work of these two authors, a connection that is borne out by
certain internal details. As Diels showed long ago, the surviving fragments
and testimonia for Gorgias’ work show both the philosophical and stylistic
influence of Empedocles. Plato (Meno 76a–e=D45a=B4) and Theophrastus
(De igne 73=D45b=B5) report that Gorgias held an Empedoclean theory of
vision as arising from effluences, and the surviving speeches are full of the
florid antitheses, repetitions and use of homoioteleuton that also characterise
the poet’s work.21
We can reasonably infer that this influence extended further. In the
Encomium of Helen, Gorgias famously argues that if speech (λόγος) per-
suaded Helen to go to Troy, she was inculpable (§8–10, 13–15). Speech is
a ‘great potentate’ (δυνάστης μέγας) which ‘with the tiniest and least-clear
body (σμικροτάτωι σώματι καὶ ἀφανεστάτωι) accomplishes the most
divine deeds’ (§8). This physical conception of speech as a ‘tiny body’
recalls the theories of the atomists, who, like Empedocles, were defending
a pluralistic cosmos against Parmenidean principles that what-is must be
unchanging and indestructible. But, given the association with
with the idea that it harbours truth (perhaps in the sense that it illustrates ethical truths, rather than
detailing accurate facts).
19
See e.g. OCD4 s.v. ‘biography, Greek’.
20
Olympiodorus (In Gorg. Prooem. 9) dates Gorgias’ birth to 470/469 bce; Diogenes Laertius (8.52)
places Empedocles’ floruit in the period of 456/452 bce.
21
Diels (1884), still the key study in this issue. Note also Segal (1962) 101.
Empedoclean theories of perception, we might in particular recall the
Empedoclean doctrine that sound (probably including speech) is com-
posed of ‘air’ (Aët. 4.16.1=D227=A93) which itself could be regarded as
a ‘tiny body’. Certainly, later ancient commentators regarded Empedocles’
elements as constituting ‘bodies’ (σώματα, e.g. Philoponus de anima
344.34 ad Arist. 418b.20<A57; Alexander ad Arist. de sens. 2.437b23<B84;
perhaps already Plato Leg. 10.889b=A48).
Gorgias’ comments on speech recall Empedocles in other respects. To
illustrate the powers of speech, he mentions emotive poetry (§8) and
incantations (ἐπωιδαί) that can induce pleasures or relieve pains and are
examples of sorcery (γοητεία) and magic (μαγεία). Later on, persuasive
speech is said to have the same effect on the soul as drugs (φάρμακα) have
on the body (§14). Some speeches ‘drug and bewitch the soul’ (τὴν ψυχὴν
ἐφαρμάκευσαν καὶ ἐξεγοήτευσαν). This conception of song as ‘bewitch-
ment’ or ‘enchantment’ is at least as old as Homer.22 The most pertinent
model was the episode, mentioned in the Introduction, where Helen’s
drug causes forgetfulness of ills (Od. 4.221) and seems to symbolise the
function of her speech.23 But given the other connections with
Empedocles, we would be justified in identifying a further connection
with his poetry. The Empedoclean narrator’s persona is that of both
a healer (D4=B112.11–12) and a sorcerer (D43=B111). He has a reputation
for being able to speak a ‘healing utterance’ (D4=B112.11) to those suffering
ailments, in other words, a sort of incantation. This doubtless led to the
later anecdote involving Empedocles’ performance of Od. 4.221 to charm
the would-be murderer (Iamblichus VP 113=P17<A15). Empedocles’ poetry
aspires to persuasion (D61=B71.1) just as Gorgias’ comments concern the
persuasiveness of speech. It is as if Empedocles’ own poetry were designed
to conform to the linguistic principles set out in these sections of the
Encomium. His unification of speech, poetry, incantation and persuasion,
when taken in combination with the biographical and other philosophical
associations with Gorgias, permit the inference that his poetic practice
influenced the theory of language and poetry proposed in the speech.
If Empedocles’ practice provided a model of the sort of speech that
Gorgias theorises, the source of Parmenides’ influence seems largely to have
been his argumentative content. Gorgias’ notorious On What Is Not was
formulated in direct response to the Eleatics, of whom Parmenides was the
22
De Romilly (1973) 156–7 argues that in Homer this notion only usually occurs with divine singers
such as the Sirens, but as she notes, one (surely programmatic) counterexample occurs at Od. 1.337.
See further Halliwell (2011) 45–50.
23
See Introduction Section 1 above.
Gorgias 191
most prominent figure.24 In the treatise, which survives in summaries by
Sextus Empiricus (Math. 7.65–87=D26b=B3) and the pseudo-Aristotelian
On Melissus, Xenophanes, and Gorgias (§5–6, 979a–980b=D26a), Gorgias
notoriously argued that ‘there is not anything’, that ‘if there is, it is
unknowable’ and ‘if there is and it is knowable, then it is not communic-
able to others’. As Jaap Mansfeld argued, this probably is not to be taken at
face value to mean that nothing exists, but is rather directed specifically
against the speculative theoretical constructions of the Presocratics above
all, Parmenides.25 On What Is Not, as far as we can tell from the summaries,
redeploys the language and principles from Parmenides’ discussion of
being and not-being, while the MXG reports that the work began with
a summary of the ideas of earlier philosophers, including those who claim
that ‘what-is is one and not many’, that is, the Eleatics.
For Gorgias, it seems, Parmenides is mistaken in thinking that there is
a fundamental entity that transcends the world of opinion. Rather, the
world of subjective opinion is all that there is. This epistemological and
metaphysical principle has been seen to underlie the theory of language in
the Helen, for in that speech, as Calogero wrote, ‘logos does not have the
theoretical function of mirroring reality, or mirroring our knowledge of
reality, it has merely a certain practical and rhetorical function, viz., to put
the mind in a [certain] affective state through persuasion’.26 Such an
understanding of the function of language constitutes a response to
Parmenides: for Parmenides, knowledge of true reality is characterised
positively in contrast with the deceptive world of mortal names and
opinions; for Gorgias, there is no transcendent reality from which mortal
opinion is a corruption. Parmenides had used literary verisimilitude to
illustrate the gap between mortal opinions and reality; Gorgias vindicates
literary verisimilitude since he excludes the possibility that language has the
higher purpose of reflecting true reality. Thus, Gorgias defends literary
deception (D35=B23). Whilst an association between poets and deception
is at least as old as the Odyssey, Parmenides appears to be the earliest
surviving poet to describe part of his own work explicitly as deceptive
(D8.57=B8.52), and consequently Verdenius argued that Gorgias was
24
For this treatise as a response to Parmenides, see Wardy (1996) 6–24 and Mansfeld (1985).
25
Mansfeld (1985). See, similarly, Curd (2006).
26
Calogero (1977) 262, who also argued that Gorgias’ epistemology was largely indistinguishable from
that of Protagoras. This translation of Calogero is from Mourelatos (1987) 135–6, who expands on
Calogero’s point and argues that Gorgias attacks the notions that linguistic meaning is reference and
that it is a mental idea, assuming instead, in both the Encomium and On What Is Not,
a ‘behaviourist’ conception of meaning, whereby language is prompted by certain stimuli and has
certain effects on its listeners.
inspired by Parmenides on this point.27 The connection between
Parmenides and Gorgias here is strengthened by the fact that Gorgias
describes deception (as well as the charms of incantations) as working on
the doxa of the soul (D24=B11§10).
The idea that poetry is supposed to be a deception approaches the later
notion of fiction. Although this idea may well already be implicit in the
Odyssey and in the Homeric Hymn to Hermes, our sources do not present an
explicit articulation of it until Gorgias. An important development in
poetics, then, seems to have been formulated as a response to
Parmenides’ restrictions on what constitutes true reality: a specific and
narrow definition of what constitutes true reality was the catalyst for the
claim that poetry need not fulfil such a definition.
4 Democritus on Literature
The atomic theory, probably invented by the shadowy Leucippus and then
refined by the more prolific Democritus, was another defence of the
pluralistic cosmos that retained Eleatic principles against absolute creation
and destruction.28 In contrast with Empedocles and Anaxagoras, the
atomists denied the Parmenidean claim that there is no such thing as not-
being, positing the existence of ‘void’. The language they seem to have used
in reference to atoms and void – literally, ‘that which is’ and ‘that which is
not’ (τὸ μὲν ὄν, τὸ δὲ μὴ ὄν) – sounds distinctly Parmenidean.29 Along with
this Parmenides-influenced metaphysics, there is also a Parmenides-
influenced epistemology: for Democritus, atoms are the usually inaccess-
ible reality that underlies our sensory experiences; the latter are merely
variable ‘conventions’ (D14<B9; D23a=B125; D24=B117; cf. D15<B9;
D16=B10), comparable to Parmenidean ‘opinions’. Nevertheless, the
atomic theory provides an explanation for the causes of sensation: vision,
for instance, occurs because objects emit ‘effluences’ (ἀπόρροιαι) of atoms
in the shape of ‘images’ (εἴδωλα) that reach the eyes of the viewer
(D146=B123; Aët. 4.13.1=D145<A29; Theophr. Sens. 50=D147<A135).
Sound occurs when atoms of air pass quickly through void, a process
that takes place in the ears when we hear sounds (Theophr. Sens.
27
Verdenius (1981) 124–7.
28
On atomism as a response to Parmenides, note e.g. Furley (1967) 79–103; Wardy (1988); Sedley
(2008). This view goes back to antiquity (e.g. Arist. Gen. corr. 1.8.325a2; Simpl. In Phys.
28.4=Theophrastus Phys. Op. fr. 8).
29
E.g. in the report of Simplicius Physics 28.14–16 (<D32=A8, A38).
Democritus on Literature 193
55–6<A135).30 It is not clear which aspects of the theory were developed by
Leucippus and which were original to Democritus, but we have far more
evidence for the views and works of the latter.
Significantly for our purposes, among his copious writings, Democritus
was credited with having written eight works on poetry and music (Diog.
Laert. 9.48): On Rhythm and Harmony; On Poetry; On the Beauty of Words;
On Nice-sounding and Bad-sounding letters; On Homer or On Correct Words
and Unusual Words; On Song; On Words; Onomastic Matters. Although
only a few fragments survive, there is enough material to have led scholars
to credit him with what can broadly be distinguished as three important
conceptual developments in the field of poetics. First, Democritus pro-
posed an anthropological theory for the origins of poetry in human history,
as arising naturally through imitation of animals (D203=B154) rather than
as a divine gift, and as a leisure activity arising through surplus rather than
a necessary one (D204<B144).31 Second, although he still maintained that
poets could be divinely inspired (D217=B18), he seems to have offered
a physical explanation for this process, in atomic terms: divine inspiration
occurs when a ‘divine breath’ (D217=B18; cf. D221=B21), probably com-
posed of the same fiery atoms of which the soul is formed (D130<A101;
D132<A28; D133<A101, all from Aristotle’s summary at De an. 1.2.403b–
405a), enters the poet/singer.32 Third, as Andrew Ford writes, ‘his materi-
alist perspective could support a view of “making” poems as a process of
selecting elementary substances (atomic sounds) and organising them into
a structure or “system” that, by virtue of its constituents and their organ-
ization, produced a specific effect on the auditor’s psyche, itself having
a physiological substratum.’33 In support of this view, according to
Aristotle, the atomists illustrated their physical principles using the
example of the components of language: different sensible objects are
made up of different atoms, just as different words are made up of different
30
For a good overview of some of the exegetical problems raised by Democritus holding simultan-
eously that only atoms are real, but also that collections of atoms have undeniably real, sensible
properties not shared by the atoms themselves, see, concisely, Warren (2007) 166–71. For more
detailed studies, see Furley (1993) and O’Keefe (1997). For Democritus’ theories of vision and
hearing, see Sassi (1978) 96–109, 117–19.
31
Cole (1967) 43; Ford (2002) 145–6.
32
Thus Delatte (1934) 35–6; Ford (2002) 168–9. Heraclitus’ cosmos seems to have contained a similarly
fiery exhalation of which souls are composed. See Betegh (2013). This physical explanation runs
counter to the view, derived from Horace Ars poetica 295–7 and found in Dodds (1951) 82, that
Democritus was an irrationalist championing the inexplicable madness of poets, although it is
possible that Democritus explained ‘madness’ as an extraordinary arrangement of atoms that was
particularly receptive to the divine ‘breath’ (thus, Delatte).
33
Ford (2002) 164–5. See, similarly, Brancacci (2007).
letters and syllables (Arist. Metaph. 1.4.985b4=D31=67A6); the same atoms
can create different effects, just as the same letters make up both tragedies
and comedies (Arist. Gen. corr. 1.2.315b6=D56=67A9). For Ford, then,
Democritus’ atomic theory provided a scientific basis on which to analyse
poetry, as an entity put together out of distinguishable components, but
which is more than the sum of its parts.
In certain respects, these poetic developments are presaged by the
Presocratic poets and, in particular, can be seen as expansions of their
influence upon the physical doctrines. First, whilst Democritus’ anthropo-
logical explanations for human history are more sophisticated than any-
thing attested among the earlier Presocratics, Xenophanes offers a similar
perspective on primordial human discoveries. As we saw, Xenophanes
D53=B18 rules out the possibility of direct divine communication, instead
claiming that ‘in time, through seeking, men discover better’. On
a straightforward reading of this fragment, we would have to include
poetry and music within the discoveries that have been achieved by
humans without the aid of the gods. In this respect, Democritus’ theory
of the origins of poetry makes explicit an entailment of Xenophanes’ claim.
On a broader level, Parmenides may have furnished Democritus with an
epistemological framework within which to articulate this theory.
Parmenides’ goddess ruled out the validity of all previous claims to divine
inspiration (and indeed, of the factual content of all previous narrative
poetry): our beliefs are a result of ‘names’ that, at some prior point in time,
mortals ‘set down’ for what-is, ‘trusting them to be true’ (D8.44=B8.39).
For her, mortal beliefs do not arise from objective reality, or through divine
inspiration, but come as a result of fallible mortal agency. This sort of
explanation could have provided a precedent for Democritus’ naturalised
understanding of major discoveries in history: his ‘anthropology’ is
a continuation of the Eleatic tradition of rewriting human history in
a manner that debunks the traditional role played by divine-to-mortal
communication.
Secondly, I have argued that Empedocles presents a systematic cosmo-
logical explanation for divine inspiration: the Muse Kalliopeia is an agent
of cosmic Love. Democritus’ explanation similarly posits that poetic
inspiration is caused by the same basic processes that lie behind all other
sensible phenomena. Both provide a systematic explanation for this trad-
itional phenomenon. Within the wider Eleatic tradition, Democritus’
explanation for the divine inspiration of poets, like Empedocles’ use of
the Muse, can be read as a revalorisation of a traditional poetic feature
Democritus on Literature 195
against the criticism of Xenophanes and against Parmenides’ negative
implications for poetry.
More fundamentally, an atomic theory of the composition of song and
its effects on the auditor seems rooted in Parmenidean and Empedoclean
principles. As we noted in Chapter 2, Democritus adapts a Parmenidean
expression when describing Homer’s poetry:
Ὅμηρος φύσεως λαχὼν θεαζούσης ἐπέων κόσμον ἐτεκτήνατο παντοίων
Homer, having been allotted a divine nature, built an arrangement of all
kinds of words
(D221=B21)
Homer ‘builds’ a kosmos out of words, presumably in like manner to how
visible objects are made out of atoms.34 The verb translated here as ‘built’
(τεκταίνομαι) is used of carpenters or smiths.35 His ‘divine nature’ probably
means that he has an atomic constitution that is especially susceptible to
receiving the ‘divine breath’ or divine images (εἴδωλα) of atoms.36 Ford
argues that, ‘This is praise of Homer as word-constructor, not poem-
maker’, possibly a defence against Protagoras’ (not necessarily serious)
criticisms that Homer had not chosen the correct words.37 Certainly, this
view is supported by the fact that an alternative title for Democritus’ work
On Homer was On the Correct Words (i.e. orthoepeia) and Unusual Words.
Indeed, one example of this sort of practice survives: Democritus defended
the epithet for Athena, Tritogeneia, ‘thrice-born’ on the grounds that it
refers to her triple concerns with proper reason, speech, and action
(D293=B2). But it should be noted here that the praise of D221=B21 is
not only of word selection, but of the whole structure that is formed out of
those words. The Parmenidean intertext supports a reading of this frag-
ment as praise of the world Homer creates through all kinds of words.
Parmenides had used the expression ‘kosmos of words’ of the deceptive
account of the world of opinion (D8.57=B8.52). Given that the atomic
34
Thus Ford (2002) 169–70. 35 See Lanata (1963) 261–2; Russell (1981) 73.
36
See Guthrie (1965) 478 n. 1; Ford (2002) 170–1; Brancacci (2007) 201–2, who connects the use of the
verb λαγχάνω, ‘be allotted’, with D154=B166, where Democritus εὔχετο εὐλόγχων τυχεῖν εἰδώλων,
‘prays to meet with well-allotted images’; for the idea that it is the divine ‘breath’ that enters Homer
here, see Delatte (1934) 28–36.
37
Ford (2002) 170. Protagoras lectured on orthoepeia (D24=A28; D25=A29), and is reported to have
criticised Homer, for instance, for using a command instead of a request when invoking the Muse at
the opening of the Iliad (D25=A29=Arist. Poet. 19.1456b15–18). Fehling (1965) 212–7 argues that
Protagoras made these arguments as a display of Sophistic skill, rather than as an attempt at
a widespread linguistic reform. Democritus D222=B23 suggests that the atomist was concerned
with similar issues of linguistic propriety.
theory developed as a response to Parmenidean arguments against absolute
creation and destruction, we would be justified in reading this as
a reappropriation of Parmenidean language: a poetic cosmos of words
becomes no longer an inferior substitute for knowledge of the truth;
instead, its internal coherence is a virtue. Indeed, such a cosmos of
words, whilst clearly the product of mortal agency in Parmenides, becomes
divinely inspired in Democritus.
This response to Parmenides is paralleled by that of Empedocles. As we
saw, Empedocles adapted the same line (D73.257=B17.26). In a manner
comparable to Democritus’ comment on Homer, he presented his own
poetry as the product of a manual craft, formed of distinct components,
just as the physical world is composed of the four elements. If Democritus’
‘materialist perspective could support a view of “making”38 poems as
a process of selecting elementary substances’, the same could already be
said of Empedocles. Similarly, the idea that the effects of song depend
upon ‘the auditor’s psyche, itself having a physiological substratum’ may
already be presupposed by Empedocles D257=B110, where we must receive
the text in a certain manner which is presumably explicable in terms of
Empedocles’ elemental physiology.
5 Democritus and the Sophists on Language

Each of Democritus’ three achievements in poetics is, then, foreshadowed
in some part by the Presocratic poets, but there is a further relevant point of
contact with Empedocles in Democritus’ claims for language more gener-
ally. For Democritus, it seems, speech, like our sensory experience, is
a matter of convention (nomos) rather than nature. Proclus, in his com-
mentary on Plato’s Cratylus, presents Democritus as a champion of the
conventionalist theory of naming, that is to say that names have no natural
connection to their referents, but only a conventional one (Procl. in Crat.
16, pp.6.20–7.6=D205<B26).39 Names are adopted for things arbitrarily
(thesei). According to Proclus, Democritus developed four arguments in
favour of this view:
ἐκ τῆς ὁμωνυμίας· τὰ γὰρ διάφορα πράγματα τῷ αὐτῷ καλοῦνται
ὀνόματι, οὐκ ἄρα φύσει τὸ ὄνομα· καὶ ἐκ τῆς πολυωνυμίας· εἰ γὰρ τὰ
διάφορα ὀνόματα ἐπὶ τὸ αὐτὸ καὶ ἓν πρᾶγμα ἐφαρμόσουσιν, καὶ
38
Also notable in this regard is that Democritus is an early user of the word poietes (D217=B18, lit.
‘maker’) for ‘poet’. See Ford (2002) 136.
39
On this testimony, see Sluiter (1997) 172–3.
Democritus and the Sophists on Language 197
ἐπάλληλα, ὅπερ ἀδύνατον· τρίτον ἐκ τῆς τῶν ὀνομάτων μεταθέσεως· διὰ τί
γὰρ τὸν Ἀριστοκλέα μὲν Πλάτωνα, τὸν δὲ Τύρταμον Θεόφραστον
μετωνομάσαμεν, εἰ φύσει τὰ ὀνόματα; ἐκ δὲ τῆς τῶν ὁμοίων ἐλλείψεως·
διὰ τί ἀπὸ μὲν τῆς φρονήσεως λέγομεν φρονεῖν, ἀπὸ δὲ τῆς δικαιοσύνης
οὐκέτι παρονομάζομεν; τύχῃ ἄρα καὶ οὐ φύσει τὰ ὀνόματα
(1) from homonymy (for different things are called by the same name, hence
the name is not by nature); (2) from polyonymy (for if different names are
applied to one and the same thing, [sc. they will be applied] mutually as well,
which is impossible); (3) the third from the change of names (for why did we
change the name of Aristocles to Plato, and of Tyrtamus to Theophrastus, if
names are by nature?); and (4) from the lack of similar [sc. forms] (why do
we say phronein [‘to think’] derived from phronesis [‘thought’], but we have
no word derived from dikaiosune [‘justice’]?) So names are by chance and
not by nature. (trans. LM)
Some of the arguments are obscure: it is not clear what it means for names to
be applied ‘mutually’ or why that is ‘impossible’. It is also unclear how much
of this goes back to Democritus of Abdera: the example of Theophrastus must
post-date the fifth century. But these sorts of arguments, as Plato’s Cratylus
attests, are characteristic of fifth-century bce sophistic debates. In some
details, they resemble Empedocles’ comments on language. He anticipates
the polyonymy argument at D73.251–7=B17.20–6, where mortals have differ-
ent names for what he calls philotes. It is possible that he, too, understood
mortal language to be conventional on this basis. In D54=B9.9, according to
most editors,40 Empedocles refers to mortal names for birth and death as
a nomos, a ‘convention’, since in reality these processes are mixture and
disintegration. As we saw, Empedocles’ comments on language developed
from Parmenides’ comments on mortal names which reflect the deceptive
world of mortal opinion and both Parmenides and Empedocles draw upon
a traditional conception of the difference between divine and mortal language.
The Democritean comments on language can again be seen as part of this
tradition: for Parmenides, mortal language is deceptive and does not have
a secure connection to true reality; for Empedocles and Democritus, it is
conventional and similarly severed from the true objects of knowledge (in
Empedocles’ case, the six permanent entities; in Democritus’ case, atoms).
Even if the arguments reported by Proclus do not in fact go back to
Democritus, our sources amply demonstrate that the fifth-century Sophists
debated the two related questions of whether words originally arose some-
how naturally or by the arbitrary decision of a name-giver and whether
40
But not LM, who prefer ὁμῶς (giving the sense ‘I assent in the same way’, rather than ‘I assent to the
nomos’), a very slight emendation for ὅμως, a better attested manuscript reading than νόμῳ.
words today have a natural or merely a conventional connection to their
referents.41 In Plato’s Cratylus, the eponymous character is presented as
espousing a ‘naturalist’ position in answer to the latter question against
Hermogenes’ ‘conventionalist’ opinion (although, later in the dialogue,
Socrates argues that the question is in fact one of the origins of words,
425a–b). The Derveni commentator seems to treat Orpheus as an original,
semi-divine name-giver (Col. 22.1–3). Prodicus argued against the possibil-
ity of ‘polyonymy’ or synonyms, suggesting that all apparent synonyms in
fact have subtle distinctions in meaning (D6a<A19), perhaps a ‘naturalist’
response to Democritus’ conventionalism.42 Protagoras’ comments on the
‘correctness of words’ (orthoepeia) in Homer has sometimes been under-
stood as a plea for a ‘natural’ usage of language, so that, for instance,
‘masculine’ things (such as ‘wrath’) should have a masculine grammatical
gender (D24=A28=Arist. Soph. el. 14.173b17–22).43
Aside from the more specific connections with Democritus, this debate
more generally stems from Presocratic anxieties over the relationship
between language and the real world.44 Cratylus was a famous Heraclitean,
and some of Heraclitus’ own fragments explore the relationship between
words and their referents. The best-known instance is D53=B48, which plays
upon the fact that bios is a homonym for both ‘bow’ and ‘life’: ‘the name for
the bow (τόξωι) is life (βίος), but its function is death’.45 The name bios thus
reveals one of the two oppositions, for Heraclitus, that are united in the
bow.46 His own logos seems to be formulated in such a manner as to map
more closely onto his cosmos than ordinary speech, reflecting the notorious
doctrine of the ‘unity of opposites’ in its puns and antitheses.47 This aspect of
Heraclitus’ work may have been a factor in Plato’s decision to explore the
problem of words and their referents in a dialogue named after one of his
followers. Be that as it may, after Heraclitus, Parmenides and Empedocles, as
we have seen, continued to raise problems over the relationship between
mortal language and the true reality that mortals mistakenly believe their
language to represent.
Among the Sophists, the material for this debate was supplied, more
often than not, by canonical poems, especially the Homeric epics. It is,
41
For an overview of these debates and how they develop from Parmenidean and Empedoclean
principles, see de Jonge and van Ophuijsen (2010) 487–90.
42 43
Thus, Sluiter (1997) 176. Thus Kerferd (1981) 68–9; although see n. 37 above.
44
Kraus (1987) is a monograph devoted to this topic.
45
βίος: τῶι οὖν τόξωι ὄνομα βίος, ἔργον δὲ θάνατος.
46
Cf. also D55=B23; D45=B32. On Heraclitus on naming, see di Cesare (1980); Schmitter (1991);
Sluiter (1997) 169–70.
47
See the work cited at Introduction n. 19 above.
Plato and Mimesis 199
then, a further instance where the ontological and epistemological prob-
lems raised by the Presocratics provided the impetus for a development in
the criticism of poetry. With Protagoras and Democritus’ detailed discus-
sions over whether Homer used the ‘correct words’ (or the Derveni
commentator’s explication of his Orphic theogony) we see a fine-grained
analysis of a particular, fixed text, rather than, say, a spontaneous response
to a song.48 Both ‘naturalist’ and ‘conventionalist’ positions could be used
to defend the text: in the Cratylus, Socrates demonstrates the naturalist
position by arguing for the appropriateness of certain Homeric names
(Crat. 390d5–393b6), whilst Democritus seems to have defended Homer
as a successful manipulator of conventional language. Attacks on the poet
in the vein of Xenophanes, Heraclitus and eventually Plato, seem to have
been rare among the Sophists:49 Protagoras, in boldly criticising the
opening of the Iliad, appears as an exception. But in any case, this
particular discussion of Homer revolves around the relationship between
his vocabulary and its referents.
The treatment of this issue as a matter of poetic criticism may have been
inspired by Parmenides and Empedocles: they assume that poetry is suscep-
tible to this sort of close, fine-grained investigation concerning its relation-
ship to reality, by inviting us to apply such a type of interpretation to parts of
their own texts. Parmenides frames the Doxa as a deceptive kosmos of words,
demanding a different type of scrutiny from the rest of the goddess’ speech.
Empedocles invites us to reflect upon the relationship between certain
expressions he uses and the true, elemental cosmos to which they refer. In
this respect, then, Parmenides and Empedocles already gesture towards the
conception of poetry presupposed by Sophistic interpretative practices. The
more specific influence of Parmenides on the broader epistemological ques-
tions explored by Democritus and the Sophists supports the contention that
he was also a direct influence on their literary-critical work.
6 Plato and Mimesis

The naturalism/conventionalism debate concerned language in general
and was only incidentally concerned with the criticism of poetry. But the
48
I leave out of discussion here the vexed question of when the Homeric poems actually achieved
textual fixity (almost certainly by the end of the sixth century – for a good overview, see Cassio
[2002]). The question at hand does not concern when poetry actually did attain such a status, but
rather at what point authors started to talk about poetry in this manner.
49
As Morgan (2000) 97 puts it, with the Sophists, ‘Homer has become an authority instead of
a competitor.’
idea that it is a distinctive characteristic of poetry in particular to conjure
a world separate from reality is already implicit in Democritus’ comment
that Homer created a kosmos of words. The Parmenidean intertext supports
the notion that kosmos, in this instance, does not simply mean ‘arrange-
ment’, but could also have the connotation of ‘universe’.50 However, the
idea that it is an essential feature of poetry to create a world distinct from
reality would most explicitly, most systematically and most influentially be
articulated by Plato. As Halliwell puts it, Plato ‘took two momentous steps
toward turning mimesis into the backdrop for an entire philosophy of art.
The first was to pose certain fundamental challenges to the status and value
of artistic mimesis – challenges that have remained unsettling and less than
completely resolved to this day. The second was to orientate questions of
mimetic art around larger philosophical concerns with the relationship
between mind and reality.’51 It is far beyond my scope to offer anything
approaching a substantial treatment of the enormous topic of Plato’s
thoughts on mimesis. More modestly, I wish to close this study by pointing
out some respects in which the second of the steps identified by Halliwell –
Plato’s orientation of questions of mimetic art around larger epistemo-
logical questions – seems to have been substantially influenced by
Xenophanes and Parmenides.
By the time of Plato, the verb mimeisthai and its cognates had long been
used of poetic activity.52 Most famously, the Delian maidens in the
(probably late sixth-century bce) Homeric Hymn to Apollo53 are able to
‘mimic’ the voices of all people, to such a degree that ‘anyone would think
that he himself were speaking’ (hAp. 162–4). Already, here, such
a supernatural skill might seem ominous, a potential form of deception
(as Helen’s similar ability is at Od. 4.279). In the fifth century,
Aristophanes has Agathon defend his effeminate dress on the grounds
that he is composing plays about women. When men compose about
men, they have what they need, but when they compose about women,
he says, ‘what we don’t possess, mimesis captures’ (Ar. Thesm. 155–6). Later,
in the same play, the character of the Inlaw uses the verb mimeisthai of
acting the part of Euripides’ Helen (850).54 Democritus is the earliest
50 51
See Chapter 2 Section 3a above. Halliwell (2002) 37.
52
The most judicious overview is Halliwell (2002) 15–22, who rightly criticises the often-repeated view
of Koller (1954) and Nagy (1989) 47–51 and (1994) that the term has its origins in ritual, a claim for
which evidence is lacking.
53
See Janko (1982) 112–5; West (2003a) 9–12; Richardson (2010) 13–15.
54
Austin and Olson (2004) ad loc. argue that the Inlaw’s use of the word simply means ‘I will imitate’
and not ‘I will perform’ (citing the parallel Aristophanic uses at Nub. 559; Vesp. 1019; Eccl. 545), but
attested philosopher or sophist to use the term mimesis, in claiming that
humans learned song through mimesis of birds (D203=B154).
In Plato’s Republic, when Socrates first introduces the concept of mimesis
(3.392d5) in the course of discussing the education of the guardians of the
ideal city, it is fully in accordance with these earlier usages: Homer engages
in mimesis when relating direct speech, in other words, when, as primary
narrator, he impersonates particular characters (392d5–394b1). Homer’s
non-mimetic speech, on the other hand, is diegesis, usually translated as
‘narration’ or ‘narrative’. Socrates then categorises three genres on the basis
of this distinction (394b9–c5): the ‘mixed’ genre, which includes epic and
involves both mimesis and diegesis; the purely mimetic genre, which
includes drama since it comprises entirely character-speech (at least, in its
Greek versions); and the purely diegetic genre, which consists exclusively of
narration, lacking direct speech, and according to Socrates includes
Dithyramb.55 The point is then made that an individual is generally only
good at imitating one thing, rather than many things (394e8–9). Even
actors seem to excel only at either comedy or tragedy (395a4–5). Moreover,
what people ‘imitate’ becomes ingrained in their natures and affects their
habits even when they are not imitating. For these reasons, the guardians
must imitate only what is appropriate for them rather than attempting to
imitate a variety of things and must especially avoid imitating that which is
wicked, slavish, vulgar or mad (395b9–396b8). There is a practical conse-
quence for the censorship of poetry within the city: only the pure poet-
imitator of the decent sort of person will be admitted (397d1–4); the man
who, like the Delian maidens, or like the Odyssean Helen, can imitate all
sorts of people, though he may be admirably talented, will be excluded
(398a1–b4). This is in spite of the fact that the latter’s work will be more
pleasant to all (397d5–8). Although Plato is often described as ‘banishing
the poets’, he does not banish all poets, but retains the composer of the
duller, less varied type of poetry (398a9–b4). It might be objected that
Socrates has illicitly blurred the distinction between a performer’s depic-
tion of their literary content and an audience’s reception of the perform-
ance, but this is to overlook the pervasive influence of Greek musical
culture (of which poetry was a part) on society at large. The audience is
not at risk of spontaneously copying the actions they see represented
the distinction is a fine one. Cf. Ar. Ran. 109, where mimesis refers to Dionysus’ dressing up as
Herakles.
55
Despite the fact that our surviving Dithyrambs do sometimes contain direct speech (e.g. Bacchyl.
15.50–63; 17.20–46 etc.). On this issue, and the difficulty later ancient authors have in categorising
Dithyramb, see Peponi (2013).
onstage; rather, the risk is that the performances will ingrain detrimental
societal norms.56 Mimesis, at this point in the dialogue, denotes a similar
activity to the earlier attested usages: the ‘imitators’’ imitations are com-
parable to the Delian maidens’ singing in various different voices, or to
Agathon’s impersonation of women, or to the Inlaw’s impersonation of
Helen.
Notoriously, however, the discussion surrounding poetry is resumed in
the tenth book, where a more expansive sense of mimesis is explored. At the
start of the book, Socrates recalls that he and his interlocutors did not
permit any poetry that was ‘mimetic’ into the city (Τὸ μηδαμῇ
παραδέχεσθαι αὐτῆς ὅση μιμητική, lit. ‘the not-allowing of it [sc.
Poetry], as much as is mimetic’, 595a1). Contrary to some commentators,
there is no contradiction here with the previous discussion, since not all
poetry was ‘mimetic’, and the type to be permitted involved the least
amount of mimesis in the narrow sense of ‘impersonation’ or direct speech
(cf. 3.396e3–7).57 Conversely, it is drama especially that appears to be
excluded.58 Later, we are told more specifically that the only permissible
poetry is to be ‘hymns to the gods and encomia to good people’ (607a3–5).
Again, this conclusion coheres with the earlier discussion, since such poems
would be of the duller, more austere type allowed into the ideal city.
What emerges from book 10, however, is a more detailed explication of
why mimetic poetry has an especially pernicious effect on the psyche, in
appealing to the irrational part of the soul. To this end, Socrates expands
the notion of mimesis to encompass the visual arts, in order to illustrate its
ontological and epistemological aspects (cf. 597b1–3). A painter who paints
a couch imitates the sensible couch, which was built by a carpenter, but the
carpenter himself imitates the form of the couch, which could only have
been built by a divine master-craftsman (597b1–e9). The painting of the
couch is thus at ‘third remove from truth’: an imitation of an imitation of
the form of the couch. The painter may be able to paint a lifelike picture of
a couch, but you could not sit on it, nor would the painter need to know
what makes a good couch in order to paint one – the only thing he would
need to know about couches would be how one appears from a particular
56
This feature is well discussed by Burnyeat (1997) 255–86. See also Halliwell (2002) 52.
57
As Belfiore (1984) 126–7; Ferrari (1989) 124–5; and Burnyeat (1997) 289–92 argue, contra (e.g.) Annas
(1982) 27 n. 37; Nehamas (1982) 51; Murray (1996) 184–7. The mistake derives from the application
of a sense of mimesis as a quasi-pictorial representation that Socrates only develops later in the book
and would become the basis for the Aristotelian and later senses of the term.
58
Note how tragic lamentations and comic jokes are the examples of how mimesis corrupts even decent
people (605c5–606c10), whilst the examples of the ‘ancient quarrel’ between poetry and philosophy
come exclusively from comedy (607b2–d2), on which see Most (2011).
angle (598a1–c5). This example helps debunk the claim that poets are
experts in all things: poets need not have knowledge of all the things
they describe, just as a painter need not have knowledge of the couch he
paints (598c5–601b1). Poets need not have knowledge or right opinion
concerning generalship or justice or any other topics they depict; imitation
is merely a kind of game (601a4–602b10). Since rational calculation would
establish that trompe l’œil paintings are paintings, or that a stick in water
only appears bent, this sort of imitation must appeal to the irrational part
of the soul (602c4–603b5). Poetry can be an especially pernicious form of
imitation, since it imitates actions in a manner that implies an evaluation.
Poems may fool us into thinking that indulgent lamentation is an accept-
able response to the loss of a loved one (603c5–606c1). Nevertheless, poets
in search of praise will typically indulge the irrational part of the soul,
portraying such sentimental scenes. It is for this reason that there is an
‘ancient quarrel’ between poetry and philosophy and almost all poets will
be banned from the ideal city.
In expounding the problems of the relationship between mimesis and
reality, Socrates uses the term in a sense that might seem applicable, not
only to the more mimetic poetry, such as drama, but to poetry and
language in general.59 However, it should be noted that the emphasis of
Book 10’s discussion of poetic mimesis is still very much upon the depiction
of particular actions (603c5–9): poetry in which the performer speaks in
propria persona, without using direct speech, in praise of gods or of great
men still need not be included within this category.60 But, in any case,
from book 3, the discussion has shifted from the general societal effects of
mimesis to its ontological and epistemological basis, in which respect
mimetic poetry is analogous to painting. What is significant, for our
present purposes, is that the epistemological and ontological grounds on
which Socrates raises these problems for poetry are in large part derived
from Parmenides.
Book 10’s claim that poetry is at ‘third remove from the truth’ relies
upon the theory of the forms, elaborated in books 5–7. As has been
acknowledged, this theory was clearly strongly influenced by
Parmenides.61 In the Republic, it is introduced towards the end of book 5
59
Note how the discussion of poetic mimesis at 598d8–599a9 focuses on poets’ knowledge of their
subject matter more generally, rather than on specific episodes of direct speech.
60
Later on, Aristotle removes the emphasis on direct speech, but maintains that mimesis involves the
representation of actions, so that Empedocles’ physical verse is non-mimetic (Poet. 1.1447a–b).
61
On Plato’s reception of Parmenides, see especially Palmer (1999), who discusses the connection with
the sight-lovers in Republic 5 in detail (esp. pp. 31–55).
to distinguish true philosophers or ‘lovers of wisdom’ from the mere ‘lovers
of sight’ (475d1–e4). The language and arguments that Socrates proposes to
convince the sight-lover that he loves only a ‘likeness’ rather than a real
thing (i.e. the form) bear an unmistakable resemblance to Parmenides’
poem: the person who knows has knowledge of what-is (ὄν) rather than
what-is-not (οὐκ ὄν) (476e6–477a1). As Glaucon asks, ‘how could what-is-
not be known?’ (477a1, πῶς γὰρ ἂν μὴ ὄν γέ τι γνωσθείη; cf. Parm.
D6.7=B2.7, ‘for you would not know what-is-not’, οὔτε γὰρ ἂν γνοίης
τό γε μὴ ἐὸν). Socrates then argues that ‘what is completely is completely
knowable, and what is in no way is in every way unknowable’ (477a2–5; cf.
Parm. D6.5–6=B2.5–6, where the way of not-being is παναπευθέα, ‘com-
pletely undiscernible’, and D8.16=B8.11, where what-is ‘is completely’,
πάμπαν πελέναι). Knowledge is of what-is and ignorance is of what-is-
not (477a10–11), but what is between knowledge and ignorance is doxa,
opinion (477b4–5). Socrates then reasons that since knowledge and opin-
ion are different powers, they are set over different things (477c1–478b2).
Knowledge is set over what-is, ignorance over what-is-not, but opinion is
set over that which is and is not at the same time (478d5–9, note especially
the resemblance between d5–6, ἅμα ὄν τε καὶ μὴ ὄν, ‘at the same time being
and not being’, and Parmenides D7=B6.7–9, where mortals believe that
‘being and not-being are the same and not the same’). Moreover, there is
only one of each form, notably, only one form of the beautiful that is the
object of knowledge, but many apparently beautiful things that are the
object of opinion (478e7–479d9). The sight-lovers are lovers of opinions
rather than lovers of knowledge (479d10–480a13). Just as Parmenides’
goddess argues, the object of knowledge is singular, whilst the objects of
opinion are plural and various. Socrates’ sight-lovers are like Parmenides’
mortals who are associated with ‘that which-is and is-not’ and who have
opinions that they mistake for knowledge.
This section already foreshadows book 10’s association of the theory of
the forms with poetics, since the sight-lovers are in particular lovers of
theatrical performances who attend all the Dionysiac festivals and listen to
every chorus (475d1–e1). Book 10 then recalls the earlier, Parmenides-
influenced discussion of the forms when Socrates invokes the claim that
the carpenter does not make what he and his interlocutors have called the
‘being’ (that is, the form) of the bed (597a1–2). Indeed, one reason why the
discussion of literary mimesis is postponed until book 10 is that it relies
upon the epistemological and ontological claims made in books 5–6 and so
would have made little sense in book 3. In short, then, the theory of literary
mimesis expressed in book 10 of the Republic is conspicuously based upon
Parmenides’ division between what-is and the world of mortal opinion.
That Plato associated Eleatic arguments more generally with mimesis is
further suggested by the fact that another theory of mimesis is put into the
mouth of the Eleatic stranger in the Sophist (234b4ff.), an avowed follower
of Parmenides.62 With respect to the more specific issue of literary
mimesis, I suggest that Parmenides’ own use of verse was an influence
on Plato’s presentation of these ontological/epistemological theories as
a problem for poetics. As I have argued, Parmenides’ goddess, in intro-
ducing the world of opinion as a ‘deceptive kosmos of words’
(D8.57=B8.52) casts it as a world of poetic verisimilitude. There is
already, then, the idea that the world of poetry is especially characteristic
of the world of opinion. Plato combines this idea with the concept of
mimesis, a term that had previously tended to refer more specifically to
impersonation. This notion of mimesis qua representation (as opposed to
mimicry or impersonation) would be taken up by Aristotle and would
become an enduring topic in poetics and aesthetics.63
Furthermore, the aesthetic problem of the fact that the more varied
‘mimetic’ poetry will be more appealing to the masses than the homogen-
ous, austere poetry seems an expansion of issues already evident in the
Presocratic poets. The two genres that Socrates permits into the ideal city
(10.607a3–5) are precisely the two genres of which Xenophanes approves in
the symposium (D59=B1.13–14, 19–21). There is a good chance that Plato
was influenced by Xenophanes in particular at this point, since
Xenophanes is elsewhere presented by Plato as the founder of the ‘Eleatic
tribe’ (Soph. 242d4–6) and, here, Socrates has relied upon Eleatic-inspired
arguments. A division between a more appealing, more varied poetry and
a less attractive, more homogeneous (but ultimately more justified) type is
exhibited in Parmenides’ poem. As we saw, this text has often met with
scathing evaluations of its literary merits, but some praise has been reserved
for the Doxa.64 A more positive literary assessment of this section may be
due to the same considerations Socrates adduces for the popularity of the
62
On the relationship between the theories of mimesis in the Sophist and the Republic, see Halliwell
(2002) 62–7.
63
On which see, above all, Halliwell (2002). Cf. Ferrari (1989) 110 ‘This [sc. Book 10’s argument that
imitativeness has, by its very nature an ethical effect] is the truly path-breaking aspect of Plato’s
critique.’
64
Most emphatically, Beaufret (1952) 8 declared that D27=B14, ‘night-shining foreign light, wander-
ing around the earth’ was ‘one of the most beautiful lines in the Greek language’, a judgement
qualifiedly endorsed by Mourelatos (2008a) 224, but described by Tarán (1977) 653 n. 5 as ‘an
attitude that reveals either little acquaintance with Greek poetry or poor taste or both’. De
gustibus . . .
more mimetic kind of poetry: the little evidence we possess for the Doxa
suggests that it was more varied and featured lush descriptions of the
sensible world that contrast starkly with the singularity of the Aletheia’s
account of what-is. Yet the more attractive Doxa is still a world of mortal
deception. By devaluing what is a world of beautiful poetry on epistemo-
logical and ontological grounds, Parmenides anticipates Plato’s ‘quarrel
between poetry and philosophy’, although he does not yet put it in those
terms.
Bibliography
Ahl, F., and Roisman, H. (1996). The Odyssey Re-Formed. Ithaca, NY.
Allan, R. J., de Jong, I. J. F., and de Jonge, C. C. (2017). ‘From Enargeia to
Immersion: The Ancient Roots of a Modern Concept’. Style 51: 34–51.
Allan, W. (2006). ‘Divine Justice and Cosmic Order in Early Greek Epic’. JHS
126: 1–35.
Aloni, A. (2009). ‘Elegy: Forms, Functions and Communication’. In Budelmann
(ed.), 168–88.
Aloni, A., and Iannucci A. (2007). L’elegia greca e l’epigramma dalle origini al
v secolo. Florence.
Annas, J. (1982). ‘Plato on the Triviality of Literature’. In Plato on Beauty, Wisdom,
and the Arts, ed. J. M. E. Moravcsik and P. Temko. Totowa, NJ: 1–28.
Arrighetti, G. (1983). ‘L’eredita dell’ epica in Parmenide’. In Festschrift für Robert
Muth zum 65. Geburtstag, ed. P. Haendel and W. Meid. Innsbruck: 9–16.
Asheri, D., Lloyd, A., and Corcella, A. (2007). Commentary on Herodotus Books i–
iv. Oxford.
Atherton, C. (ed.) (1998). Form and Content in Didactic Poetry. Bari.
Atkins, J. W. H. (1934). Literary Criticism in Antiquity: Volume 1, Greek: A Sketch of
Its Development. Cambridge.
Attridge, D. (2004). The Singularity of Literature. London.
(2015). The Work of Literature. Oxford.
(2019). The Experience of Literature. Oxford.
Austin, C., and Olson, S. D. (2004). Aristophanes: Thesmophoriazusae. Oxford.
Austin, N. (1972). ‘Name Magic in the Odyssey’. California Studies in Classical
Antiquity 5: 1–19.
Babut, D. (1974a). ‘Xénophane critique des poètes’. AC 43: 83–117.
(1974b). ‘Sur la “théologie” de Xénophane’. Revue philosophique de la France et
de l’étranger 164: 401–40.
Bachvarova, M. R. (2009). ‘Hittite and Greek Perspectives on Travelling Poets,
Texts and Festivals’. In Hunter and Rutherford (eds.), 23–45.
Bakker, E. J. (2005). Pointing at the Past: From Formula to Performance in Homeric
Poetics. Washington, DC and Cambridge, MA.
(2013). The Meaning of Meat and the Structure of the Odyssey. Cambridge.
(2017a). ‘Trust and Fame: The Seal of Theognis’. In Bakker (ed.), 99–121.
207
208 Bibliography
(ed.) (2017b). Authorship and Greek Song: Authority, Authenticity, and
Performance. Leiden.
Barfield, R. (2011). The Ancient Quarrel between Philosophy and Poetry. Cambridge.
Barker, E. T. E. (2015). ‘New Religious Movements’. In International
Encyclopedia of the Social & Behavioral Sciences, ed. J. D. Wright, 2nd ed.
Oxford: 805–8.
(2009). Entering the Agon: Dissent and Authority in Homer, Historiography, and
Tragedy. Oxford.
Barnes, J. (1982). The Presocratic Philosophers. Rev. ed. London.
(2011a). ‘The History of Philosophy’. In Method and Metaphysics: Essays in
Ancient Philosophy I. Oxford: 17–22.
(2011b). ‘Philosophy within Quotation Marks?’ In Method and Metaphysics:
Essays in Ancient Philosophy I. Oxford: 23–42.
Barrett, J. (2004). ‘Struggling with Parmenides’. Ancient Philosophy 24: 267–91.
Barrett, W. S. (1964). Euripides: Hippolytus. Oxford.
Barthes, R. (1974). S/Z. Translated by R. Miller. Oxford. Originally published as
S/Z, Paris, 1973.
Beaufret, J. (1952). Le Poème de Parménide. Paris.
Becker, O. (1937). Das Bild des Weges und Verwandte Vorstellungen im
frühgriechischen Denken. Berlin.
Beekes, R. (2010). Etymological Dictionary of Greek. Leiden.
Belfiore, E. (1984). ‘A Theory of Imitation in Plato’s Republic’. TAPhA 114:
121–46.
Bernabé, A. (1979). ‘Los filósofos presocráticos como autores literarios’. Emerita
47: 357–94.
(1987–2007). Poetarum epicorum Graecorum testimonia et fragmenta. 4 vols.
Berlin.
(2014). ‘On the Rites Described and Commented Upon in the Derveni
Papyrus, Columns i–vi’. In Papadopoulou and Muellner (eds.), 19–52.
Bernabé, A., and San Cristóbal, A. I. J. (2008). Instructions for the Netherworld: The
Orphic Gold Tablets. Leiden.
Bernays, J. (1850). ‘Heraklitische Studien’. RhM 7: 90–116.
Betegh, G. (2004). The Derveni Papyrus: Cosmology, Theology and Interpretation.
Cambridge.
(2013). ‘On the Physical Aspect of Heraclitus’ Psychology, with New
Appendices’. In Doctrine and Doxography Studies on Heraclitus and
Pythagoras, ed. D. Sider and D. Obbink. Berlin: 225–61.
(2016). ‘Colocation’. In ΣΩΜΑ. Körperkonzepte und körperliche Existenz in der
antiken Philosophie und Literatur, ed. T. Buchheim, D. Meissner and
N. Wachsmann. Hamburg: 393–421.
Bidez, J. (1894). La Biographie d’Empédocle. Brussels.
Bierl, A. (2014). ‘“Riddles over Riddles”: “Mysterious” and “Symbolic” (Inter)
Textual Strategies: The Problem of Language in the Derveni Papyrus’. In
Papadopoulou and Muellner (eds.), 187–210.
Bignone, E. (1916). Empedocle: studio critico. Turin.
Bibliography 209
Bing, P. (1981). ‘The Voice of Those Who Live in the Sea: Empedocles and
Callimachus’. ZPE 41: 33–6.
Blank, D. L. (1982). ‘Faith and Persuasion in Parmenides’. ClAnt 1: 167–77.
Blumenberg, H. (1966). ‘Sprachsituation und immanente Poetik’. In Immanente
Ästhetik, ästhetische Reflexion: Lyrik als Paradigma der Moderne, ed. W. Iser.
Munich: 145–55.
Boedeker, D., and Sider, D. (eds.) (2001). The New Simonides: Contexts of Praise
and Desire. Oxford.
Böhme, R. (1986). Die verkannte Muse: Dichtersprache und geistige Tradition des
Parmenides. Bern.
Bollack, J. (1965–9). Empédocle. 4 vols. Paris.
(2001). ‘Voir la haine’. Methodos 1. Available online at https://journals
.openedition.org/methodos/169 (accessed 23 August 2019).
(2003). Purifications. Paris.
(2005). ‘Empedocles: Two Theologies, Two Projects’. In Pierris (ed.), 45–72.
Bowie, E. L. (1986). ‘Early Greek Elegy, Symposium and Public Festival’. JHS 106:
13–35.
(1993). ‘Lies, Fiction and Slander in Early Greek Poetry’. In Lies and Fiction in
the Ancient World, ed. C. Gill and T. P. Wiseman. Exeter: 1–37.
(2016). ‘Cultic Contexts for Elegiac Performance’. In Iambus and Elegy: New
Approaches, ed. C. Carey and L. A. Swift. Oxford: 15–32.
Bowra, C. M. (1937). ‘The Proem of Parmenides’. CPh 32: 97–112.
(1938). ‘Xenophanes, Fragment 1’. CPh 33: 353–67.
(1964). Pindar. Oxford.
Boys-Stones, G. R. (2003). ‘The Stoics’ Two Types of Allegory’. In Boys-Stones
(ed.), 189–216.
(ed.) (2003). Metaphor, Allegory, and the Classical Tradition: Ancient Thought
and Modern Revisions. Oxford.
Brancacci, A. (2007). ‘Democritus’ Mousika’. In Democritus: Science, the Arts, and
the Care of the Soul, ed. A. Brancacci and P.-M. Morel. Leiden: 181–206.
Bredlow, L. A. (2011). ‘Aristotle, Theophrastus, and Parmenides’ Theory of
Cognition (B 16)’. Apeiron 44: 219–63.
Bremer, D. (1980). ‘Aristoteles, Empedokles und die Erkenntnisleistung der
Metapher’. Poetica 12: 350–76.
Bremer, J.-M. (1998). ‘Giving and Thanksgiving in Greek Religion’. In Reciprocity
in Ancient Greece, ed. C. Gill, N. Postlethwaite and R. Seaford. Oxford:
127–38.
Bremmer, J. (1983). The Early Greek Concept of the Soul. Princeton, NJ.
(1987). ‘What Is a Greek Myth?’ In Interpretations of Greek Mythology, ed.
J. Bremmer. London: 1–9.
(2001). The Rise and Fall of the Afterlife. London.
(2014). Initiation into the Mysteries of the Ancient World. Berlin.
Broadie, S. (1999). ‘Rational Theology’. In Long (ed.), 205–24.
Brooks, C. (1947). The Well-Wrought Urn: Studies in the Structure of Poetry.
Cambridge, MA.
210 Bibliography
Bryan, J. (2012). Likeness and Likelihood in the Presocratics and Plato. Cambridge.
Budelmann, F. (ed.) (2009). The Cambridge Companion to Greek Lyric.
Cambridge.
Bugno, M. (ed.) (2005). Senofane: interpretazioni antiche e contesto culturale.
Naples.
Bundy, E. L. (1962). Studia Pindarica. Berkeley, CA.
Burkert, W. (1960). ‘Platon oder Pythagoras? Zum Ursprung des Wortes
“Philosophie.”’ Hermes 88: 159–77.
(1968). ‘Orpheus und die Vorsokratiker’. A&A 14: 93–114.
(1969). ‘Das Proömium des Parmenides und die Katabasis des Pythagoras’.
Phronesis 14: 1–29.
(1972). Lore and Science in Ancient Pythagoreanism. Cambridge, MA.
(1979). Structure and History in Greek Mythology and Ritual. Berkeley and Los
Angeles, CA.
(1982). ‘Craft versus Sect: The Problem of Orphics and Pythagoreans’. In Jewish
and Christian Self-Definition, Vol. 3: Self-Definition in the Greco-Roman
World, ed. B. Meyer and E. P. Sanders. Philadelphia, PA: 1–22.
(1983). Homo Necans: The Anthropology of Ancient Greek Sacrificial Ritual and
Myth. Translated by P. Bing. Berkeley and Los Angeles, CA. Originally
published as Homo Necans: Interpretationen altgriechischer Opferriten und
Mythen, Berlin, 1972.
(1985). Greek Religion. Translated by J. Raffan. Oxford. Originally published as
Griechische Religion der archaischen und klassischen Epoche, Stuttgart, 1977.
(1987a). Ancient Mystery Cults. Cambridge, MA.
(1987b). ‘The Making of Homer in the Sixth Century bc: Rhapsodes versus
Stesichorus’. In Papers on the Amasis Painter and His World, ed. D. von
Bothmer. Malibu, CA: 43–62.
(1992). The Orientalizing Revolution: Near Eastern Influence on Greek Culture in
the Early Archaic Age. Cambridge, MA.
(1996). Creation of the Sacred: Tracks of Biology in Early Religions.
Cambridge, MA.
(2008). ‘Prehistory of Presocratic Philosophy in an Orientalizing Context’. In
Curd and Graham (eds.), 55–87.
(2014). ‘How to Learn about Souls: The Derveni Papyrus and Democritus’. In
Papadopoulou and Muellner (eds.), 107–14.
Burnyeat, M. (1999). ‘Culture and Society in Plato’s Republic’. Available online at
http://tannerlectures.utah.edu/_documents/a-to-z/b/Burnyeat99.pdf
(accessed 23 August 2019).
Burton, D. (2011). ‘Response and Composition in Archaic Greek Poetry’.
Antichthon 45: 58–76.
Bury, R. G. (1936). Sextus Empiricus. 3 vols. Cambridge, MA.
Buxton, R. (1980). ‘Blindness and Limits: Sophokles and the Logic of Myth’. JHS
100: 22–37.
(1994). Imaginary Greece: The Contexts of Mythology. Cambridge.
(ed.) (1999). From Myth to Reason? Oxford.
Bibliography 211
Cairns, D. L. (1993). Aidōs: The Psychology and Ethics of Honour and Shame in
Ancient Greek Literature. Oxford.
(2002). ‘The Meaning of the Veil in Ancient Greek Culture’. In Women’s Dress
in the Ancient Greek World, ed. L. Llewellyn-Jones. Swansea: 73–93.
(2014). ‘Exemplarity and Narrative in the Greek Tradition’. In Defining Greek
Narrative, ed. D. L. Cairns and R. Scodel. Edinburgh: 103–36.
(2015). ‘The Horror and the Pity: Phrikē as a Tragic Emotion’. Psychoanalytic
Inquiry 35: 75–94.
Cairns, D. L., and Howie, J. G. (2010). Bacchylides. Five Epinician Odes (3, 5, 9, 11,
13). Cambridge.
Cairns, F. (1972). Generic Composition in Greek and Roman Poetry.
Edinburgh.
Calame, C. (1974). ‘Réflexions sur les genres littéraires en Grèce archaïque’.
QUCC 17: 113–28.
(2005). Masks of Authority: Fiction and Pragmatics in Ancient Greek Poetics.
Translated by P. M. Burk. Ithaca, NY.
Calogero, G. (1977). Studi sull’Eleatismo. 2nd ed. Florence.
Campbell, D. A. (1982). Greek Lyric Poetry: A Selection of Early Greek Lyric, Elegiac
and Iambic Poetry. 2nd ed. Bristol.
Canevaro, L. G. (2015). Hesiod’s Works and Days: How to Teach Self-Sufficiency.
Oxford.
Cassin, B. (1987). ‘Le Chant des Sirènes dans le poème de Parménide’. In Études
sur Parménide tome ii: Problèmes d’interprétation, ed. P. Aubenque. Paris:
163–9.
Cassio, A. C. (2002). ‘Early Editions of the Greek Epics and Homeric Textual
Criticism in the Sixth and Fifth Centuries bc’. In Omero tremila anni dopo,
ed. F. Montanari. Rome: 105–36.
Caston, V., and Graham, D. W. (eds.) (2002). Presocratic Philosophy: Essays in
Honour of Alexander Mourelatos. Aldershot.
Cave, T. (2016). Thinking with Literature: Towards a Cognitive Criticism. Oxford.
Cerri, G. (1996). Platone sociologo della comunicazione. 2nd ed. Lecce.
(2001). ‘Physikà e Katharmoì di Empedocle’. Aevum(ant) 1 (n.s.): 181–96.
Chantraine, P. (1968). Dictionnaire étymologique de la langue grecque. Paris.
Cherniss, H. (1935). Aristotle’s Criticism of Presocratic Philosophy. New York, NY.
Chitwood, A. (1986). ‘The Death of Empedocles’. AJPh 107: 175–91.
Classen, C. J. (1989). ‘Xenophanes and the Tradition of Epic Poetry’. In Ionian
Philosophy, ed. K. Boudouris, 91–103. Athens.
Clay, D. (2012). ‘Empedocles at Panoplis and Delphi’. In Presocratics and Plato:
Festschrift in Honor of Charles Kahn, ed. R. Patterson, V. Karasmanis and
A. Hermann. Las Vegas, NV: 59–77.
Clay, J. S. (1993). ‘The Education of Perses: From “Mega Nepios” to “Dion
Genos” and Back’. MD 31: 23–33.
(2003). Hesiod’s Cosmos. Cambridge.
(2011a). ‘The Homeric Hymns as Genre’. In The Homeric Hymns: Interpretative
Essays, ed. A. Faulkner. Oxford: 233–53.
212 Bibliography
(2011b). Homer’s Trojan Theater: Space, Vision, and Memory in the IIiad.
Cambridge.
(2015). ‘Commencing Cosmogony and the Rhetoric of Poetic Authority’. In
Cosmologies et cosmogonies dans la littérature antique = Entretiens sur
l’Antiquité classique 61. Vandœuvres: 105–47.
Clayman, D. L. (2009). Timon of Phlius: Pyrrhonism into Poetry. Berlin.
Clinton, K. (2003). ‘Stages of Initiation into the Eleusinian and Samothracian
Mysteries’. In Greek Mysteries: The Archaeology and Ritual of Ancient Greek
Secret Cults, ed. M. B. Cosmopoulos. London: 50–78.
Cobley, P. (2014). Narrative. 2nd ed. Abingdon.
Colborn, R. (2013). ‘Solving Problems with Acrostics: Manilius Dates
Germanicus’. CQ 63: 450–2.
Cole, T. (1967). Democritus and the Sources of Greek Anthropology. Ann Arbor, MI.
(1983). ‘Archaic Truth’. QUCC 13 (n.s.): 7–28.
Collins, D. (2004). Master of the Game: Competition and Performance in Greek
Poetry. Washington, DC.
Collobert, C. (2004). ‘L’odyssée ou la naissance de la fiction’. RPhilos 129: 15–26.
Conche, M. (1996). Parménide: Le poème. Paris.
Conte, G. B. (1986). The Rhetoric of Imitation: Genre and Poetic Memory in Virgil
and Other Latin Poets. Ithaca, NY.
(1994). Genres and Readers: Lucretius, Love Elegy, Pliny’s Encyclopedia.
Baltimore, MD.
Cordero, N.-L. (1984). Les Deux chemins de Parménide. Paris.
(ed.) (2012). Parmenides, Venerable and Awesome. Las Vegas, NV.
Cornford, F. M. (1912). From Religion to Philosophy: A Study in the Origins of
Western Speculation. London.
Cosgrove, M. R. (2011). ‘The Unknown “Knowing Man”: Parmenides, B1. 3’. CQ
61: 28–47.
Coxon, A. H. (1968). ‘The Text of Parmenides Fr. 1. 3’. CQ 18: 69.
(2009). The Fragments of Parmenides: A Critical Text with Introduction and
Translation, the Ancient Testimonia and a Commentary. Rev. ed.
M. Schofield, with translation by R. D. McKirahan. Las Vegas, NV.
Csapo, E. (2005). Theories of Mythology. Oxford.
Culler, J. (1975). Structuralist Poetics. London.
(1981). The Pursuit of Signs. London.
(2015). The Theory of the Lyric. Cambridge, MA.
Curd, P. (1998). The Legacy of Parmenides: Eleatic Monism and Later Presocratic
Thought. Princeton, NJ.
(2002). ‘The Presocratics as Philosophers’. In Laks and Louguet (eds.), 115–38.
(2005). ‘On the Question of Religion and Natural Philosophy in Empedocles’.
In Pierris (ed.), 137–62.
(2006). ‘Gorgias and the Eleatics’. In Sassi (ed.), 183–200.
(2013a). ‘Where Are Love and Strife? Incorporeality in Empedocles’. In Early
Greek Philosophy: The Presocratics and the Emergence of Reason, ed. J. McCoy.
Washington, DC: 113–38.
Bibliography 213
(2013b). ‘The Divine and the Thinkable: Toward an Account of the Intelligible
Cosmos’. Rhizomata 1: 217–47.
Curd, P., and Graham, D. W. (2008). The Oxford Handbook of Presocratic
Philosophy. Oxford.
Currie, B. (2012). ‘Hesiod on Human History’. In Greek Notions of the Past, eds.
J. Marincola, L. Llewellyn-Jones and C. Maciver. Edinburgh: 37–64.
(2016). Homer’s Allusive Art. Oxford.
Currie, G. (1986). ‘Fictional Truth’. Philosophical Studies 50: 195–212.
(1990). The Nature of Fiction. Cambridge.
(2010). Narratives and Narrators: A Philosophy of Stories. Oxford.
d’Alessio, G. B. (1995). ‘Una via lontana dal cammino degli uomini (Parm. Frr. 1+6
D.-K.; Pind. Ol. vi 22–27; Pae. viib 10–20)’. SIFC 13: 143–81.
(2009). ‘Defining Local Identities’. In Hunter and Rutherford (eds.), 137–67.
Dalzell, A. (1987). ‘Language and Atomic Theory in Lucretius’. Hermathena 143:
19–28.
de Jong, I. J. F. (1988). ‘Homeric Words and Speakers: An Addendum’. JHS 108:
188–9.
(2014). Narratology and Classics: A Practical Guide. Oxford.
de Jonge, C. C., and van Ophuijsen, J. M. (2010). ‘Greek Philosophers on
Language’. In A Companion to the Ancient Greek Language, ed. E. Bakker.
Malden, MA, Oxford, and Chichester: 483–98.
de Rijk, L. M. (1983). ‘Did Parmenides Reject the Sensible World?’ In Graceful
Reason: Essays in Ancient and Medieval Philosophy Presented to Joseph Owens,
ed. L. P. Gerson. Toronto: 29–53.
de Romilly, J. (1973). ‘Gorgias et le pouvoir de la poésie’. JHS 93: 155–62.
Deichgräber, K. (1933). ‘Hymnische Elemente in der Philosophischen Prosa der
Vorsokratiker’. Philologus 88: 347–61.
(1958). Parmenides’ Auffahrt zur Göttin des Rechts: Untersuchungen zum
Prooimion seines Lehrgedichts. Wiesbaden.
Delatte, A. (1934). Les Conceptions de l’enthousiasme chez les philosophes
présocratiques. Paris.
Delattre, D. (2007). Philodème de Gadara sur la musique, livre iv. 2 vols. Paris.
de Man, P. (1983). Blindness and Insight: Essays in the Rhetoric of Contemporary
Criticism. 2nd ed. London.
Detienne, M. (1986). The Creation of Mythology. Translated by M. Cook. Chicago,
IL. Originally published as L’Invention de la mythologie, Paris, 1981.
(1989). ‘Culinary Practices and the Spirit of Sacrifice’. In The Cuisine of Sacrifice
among the Greeks, ed. M. Detienne and J.-P. Vernant, translated by
P. Wissing. Chicago, IL: 1–20. Originally published as La Cuisine du sacrifice
en pays grec, Paris, 1979.
(1996). The Masters of Truth in Archaic Greece. Translated by J. Lloyd.
Cambridge, MA. Originally published as Les Maîtres de vérité dans la Grèce
archaïque, Paris, 1967.
di Cesare, D. (1980). ‘Heraklit und die Sprache’. In Sprachphilosophie in Antike
und Mittelalter, ed. B. Mojsisch. Amsterdam: 1–16.
214 Bibliography
di Marco, M. (1989). Timone di Fliunte, Silli. Rome.
Diels, H. (1884). ‘Gorgias und Empedokles’. SBBerl 49: 343–68.
(1897). Parmenides Lehrgedicht. Berlin.
Diggle, J. (2004). Theophrastus: Characters. Cambridge.
Dodds, E. R. (1951). The Greeks and the Irrational. Berkeley, CA.
Doležel, L. (1980). ‘Truth and Authenticity in Narrative’. Poetics Today 1: 7–25.
(1998). Heterocosmica: Fiction and Possible Worlds. Baltimore, MD.
(2010). Possible Worlds of Fiction and History. Baltimore, MD.
Dolin, E. F. (1962). ‘Parmenides and Hesiod’. HSPh 66: 93–8.
Dougherty, C. (1993). The Poetics of Colonization: From City to Text in Archaic
Greece. Oxford.
Douglas, M. (1966). Purity and Danger: An Analysis of the Concepts of Pollution and
Taboo. London.
Dover, K. J. (1964). ‘The Poetry of Archilochus’. Entretiens sur l’Antiquité classique
10: 183–223.
(1975). ‘The Freedom of the Intellectual in Greek Society’. Talanta 7: 24–54.
Dowden, K. (1992). The Uses of Greek Mythology. London.
Durkheim, E. (1912). Les Formes élémentaires de la vie religieuse. Paris.
Eagleton, T. (2008). Literary Theory: An Introduction. 2nd, rev. ed. Malden, MA.
Edmonds III, R. G. (2004). Myths of the Underworld Journey: Plato, Aristophanes,
and the ‘Orphic’ Gold Tablets. Cambridge.
(2013). Redefining Ancient Orphism: A Study in Greek Religion. Cambridge.
Edmunds, L. (1985). ‘The Genre of Theognidean Poetry’. In Figueira and Nagy
(eds.), 96–111.
Edmundson, M. (1995). Literature against Philosophy, Plato to Derrida: A Defence of
Poetry. Cambridge.
Effe, B. (1977). Dichtung und Lehre. Untersuchungen zur Typologie des antiken
Lehrgedichts. Munich.
Eliot, T. S. (1920). ‘Hamlet and His Problems’. In The Sacred Wood, London: 87–94.
Else, G. F. (1957). Aristotle’s Poetics: The Argument. Cambridge, MA.
Elsner, J. (2002). ‘Introduction: The Genres of Ekphrasis’. Ramus 31: 1–18.
Faraone, C. A. (2008). The Stanzaic Architecture of Early Greek Elegy. Oxford.
Faulkner, A. (2008). The Homeric Hymn to Aphrodite: Introduction, Text, and
Commentary. Oxford.
Federico, E. (2001). ‘La Katharsis di Epimenide ad Atene: La vicenda, gli usi e gli abusi
ateniesi’. In Epimenide Cretese, ed. E. Federico and A. Visconti. Naples: 77–128.
Feeney, D. C. (1991). The Gods in Epic: Poets and Critics of the Classical Tradition.
Oxford.
(1995). ‘Criticism Ancient and Modern’. In Ethics and Rhetoric, ed. D. C. Innes,
H. Hine and C. B. R. Pelling. Oxford: 301–10.
Fehling, D. (1965). ‘Zwei Untersuchungen zur griechischen Sprachphilosophie’.
RhM 108: 212–29.
Ferrari, G. R. F. (1989). ‘Plato and Poetry’. In The Cambridge History of Literary
Criticism: Volume 1: Classical Criticism, ed. G. A. Kennedy. Cambridge:
92–148.
Bibliography 215
Ferrell, R. (2002). Genres of Philosophy. Aldershot.
Feyerabend, B. (1984). ‘Zur Wegmetaphorik beim Goldblättchen aus Hipponion
und dem Proömium des Parmenides’. RhM 127: 1–22.
Figueira, T. J. and Nagy, G. (eds.) (1985). Theognis of Megara: Poetry and the Polis.
Baltimore, MD.
Finglass, P. J. (2007). Sophocles: Electra. Cambridge.
(2018). Sophocles: Oedipus the King. Cambridge.
Finkelberg, M. (1998). The Birth of Literary Fiction in Ancient Greece. Oxford.
Finnegan, R. (1977). Oral Poetry: Its Nature, Significance and Social Context.
Cambridge.
Fischer-Hansen, T., Nielsen, T. H., and Ampolo, C. (2004). ‘Sikelia’. In An
Inventory of Archaic and Classical Poleis, ed. M. H. Hansen and
T. H. Nielsen. Oxford: 172–248.
Flower, M. (2008). The Seer in Ancient Greece. Berkeley, CA.
Floyd, E. D. (1992). ‘Why Parmenides Wrote in Verse’. AncPhil 12: 251–65.
Ford, A. (1985). ‘The Seal of Theognis: The Politics of Authorship in Archaic
Greece’. In Figueira and Nagy (eds.), 82–95.
(1988). ‘The Classical Definition of ΡΑΨΩΙΔΙΑ’. CPh 83: 300–7.
(1992). Homer: The Poetry of the Past. Ithaca, NY.
(2002). The Origins of Criticism: Literary Culture and Poetic Theory in Classical
Greece. Princeton, NJ.
(2003). ‘From Letters to Literature: Reading the “Song Culture” of Classical
Greece’. In Written Texts and the Rise of Literate Culture in Ancient Greece, ed.
H. Yunis. Cambridge: 15–37.
(2015). ‘The Purpose of Aristotle’s Poetics’. CPh 110: 1–21.
(2016). ‘Catharsis, Music, and the Mysteries in Aristotle’. Skenè 2: 24–41.
Fowler, A. (1982). Kinds of Literature: An Introduction to the Theory of Genres and
Modes. Oxford.
Fowler, D. (1991). ‘Narrate and Describe: The Problem of Ekphrasis’. JRS 81:
25–35.
(1997). ‘On the Shoulders of Giants: Intertextuality and Classical Studies’. MD
39: 13–34.
(2000). ‘The Didactic Plot’. In Matrices of Genre: Authors, Canons and Society,
ed. M. Depew and D. Obbink. Cambridge, MA: 205–19.
Fowler, R. L. (1987). Nature of Early Greek Lyric: Three Preliminary Studies.
Toronto.
(2011). ‘Mythos and Logos’. JHS 131: 45–66.
(2013). Early Greek Mythography: Volume 2: Commentary. Oxford.
Fränkel, H. (1955). Wege und Formen frühgriechischen Denkens: Literarische und
philosophiegeschichtliche Studien. Munich.
(1993). ‘Xenophanes’ Empiricism and His Critique of Knowledge (B34)’. In The
Pre-Socratics: A Collection of Critical Essays, ed. A. P. D. Mourelatos, 118–31.
Princeton, NJ. Originally published in 1925.
(1975a). ‘Studies in Parmenides’. In Studies in Presocratic Philosophy Volume 2:
The Eleatics and Pluralists, ed. D. J. Furley and R. E. Allen. London: 1–47.
216 Bibliography
(1975b). Early Greek Poetry and Philosophy. Translated by M. Hadas and
J. Willis. Oxford. Originally published as Dichtung und Philosophie des
frühen Griechentums, New York, NY, 1951.
Frede, M. (2008). ‘Aristotle’s Account of the Origins of Philosophy’. In Curd and
Graham (eds.), 501–29.
Frege, G. (1960). ‘On Sense and Reference’. In Translations from the Philosophical
Writings of Gottlob Frege, ed. P. Geach and M. Black. Oxford: 56–78. Essay
originally published in 1892 as ‘Über Sinn und Bedeutung’, Zeitschrift für
Philosophie und philosophische Kritik, 100: 25–50.
Friedländer, P. (1941). ‘Pattern of Sound and Atomistic Theory in Lucretius’. AJPh
62: 16–34.
Furley, D. (1967). Two Studies in the Greek Atomists. Princeton, NJ.
(1993). ‘Democritus and Epicurus on Sensible Qualities’. In Passions and
Perceptions: Studies in Hellenistic Philosophy of Mind, ed. J. Brunschwig and
M. C. Nussbaum. Cambridge: 72–94.
Furley, W. D. (1995). ‘Praise and Persuasion in Greek Hymns’. JHS 115:
29–46.
Furley, W. D., and Bremer, J. M. (2001). Greek Hymns: Selected Cult Songs from the
Archaic to the Hellenistic Period. 2 vols. Tübingen.
Gadamer, H.-G. (1986). The Relevance of the Beautiful. Translated by N. Walker.
Cambridge. Originally published as Die Aktualität des Schönen, Stuttgart,
1977.
(2004). Truth and Method. 2nd, rev. ed. Translated by J. Weinsheimer and
D. J. Marshall. London. Originally published as Wahrheit und Methode,
Tübingen, 1960.
Gagné, R. (2006). ‘L’Esthétique de la peur chez Empédocle’. RPhA 24: 83–110.
(2013). Ancestral Fault in Ancient Greece. Cambridge.
Gale, M. (1994). Myth and Poetry in Lucretius. Cambridge.
Gallavotti, C. (1975). Empedocle: Poema fisico e lustrale. Verona.
Gallop, D. (1984). Parmenides of Elea. Toronto.
Gantz, T. (1993). Early Greek Myth: A Guide to Literary and Artistic Sources.
Baltimore, MD.
Garani, M. (2007). Empedocles Redivivus: Poetry and Analogy in Lucretius. London.
Garner, R. (1990). From Homer to Tragedy: The Art of Allusion in Greek Poetry.
London and New York, NY.
Garvie, A. F. (1968). Aeschylus’ Supplices: Play and Trilogy. Cambridge.
Gazis, G. (2018). Homer and the Poetics of Hades. Oxford.
Gemelli Marciano, M. L. (1990). Le metamorfosi della tradizione: mutamenti di
significato e neologismi nel Peri physeos di Empedocle. Bari.
(2001). ‘Le “demonologie” Empedoclee: problemi di metodo e altro’.
Aevum(ant) 1 (n.s.): 205–35.
(2005a). ‘Senofane: interpretazioni antiche e contesto culturale. La critica ai
poeti e il cosiddetto “Monismo”’. In Senofane ed Elea tra Ionia e Magna
Grecia, ed. M. Bugno, 63–76. Naples.
(2005b). ‘Empedocles’ Zoogony and Embryology’. In Pierris (ed.), 373–404.
Bibliography 217
(2008). ‘Images and Experience: At the Roots of Parmenides’ Aletheia’. AncPhil
28: 21–48.
(2013). Parmenide: suoni, immagini, esperienza: A cura di Livio Rossetti e Massimo
Pulpito, ed. L. Rossetti and M. Pulpito. Sankt Augustin.
Gentili, B. (1988). Poetry and Its Public in Ancient Greece. Translated by A. T. Cole.
Baltimore, MD. Originally published as Poesia e pubblico nella Grecia antica,
Milan, 1984.
George, A. (1999). The Epic of Gilgamesh. Rev. ed. London.
Gera, D. L. (2003). Ancient Greek Ideas on Speech, Language, and Civilization.
Oxford.
Gerber, D. E. (1999). Greek Elegiac Poetry. Cambridge, MA.
Gernet, L. (1981). The Anthropology of Ancient Greece. Translated by J. B. Hamilton
and B. Nagy. Baltimore, MD. Originally published as Anthropologie de la
Grèce antique, Paris, 1968.
Gheerbrant, X. (2017). Empédocle, une poétique philosophique. Paris.
Gibson, J. (2009). ‘Literature and Knowledge’. In Oxford Handbook of Philosophy
and Literature, ed. R. Eldridge. Oxford: 467–85.
(ed.) (2015). The Philosophy of Poetry. Oxford.
Goldhill, S. (1991). The Poet’s Voice: Essays on Poetics and Greek Literature.
Cambridge.
(1999). ‘Literary History without Literature: Reading Practices in the Ancient
World’. SubStance 28: 57–89.
González, J. (2013). The Epic Rhapsode and His Craft. Washington, DC.
Goody, J. (1977). The Domestication of the Savage Mind. Cambridge.
Goody, J., and Watt, I. (1963). ‘The Consequences of Literacy’. Comparative
Studies in Society and History 5: 304–45.
Gould, J. (1973). ‘HIKETEIA’. JHS 93: 74–103.
(2001). ‘On Making Sense of Greek Religion’. In Myth, Ritual, Memory, and
Exchange: Essays in Greek Literature and Culture. Oxford: 203–34.
Graeser, A. (1977). ‘On Language, Thought, and Reality in Ancient Greek
Philosophy’. Dialectica 31: 359–88.
Graf, F., and Johnston, S. I. (2013). Ritual Texts for the Afterlife: Orpheus and the
Bacchic Gold Tablets. 2nd ed. London.
Graham, D. W. (1988). ‘Symmetry in the Empedoclean Cycle’. CQ 38: 297–312.
(1999). ‘Empedocles and Anaxagoras’. In Long (ed.), 159–80.
(2002). ‘Heraclitus and Parmenides’. In Caston and Graham (eds.), 27–44.
(2006). Explaining the Cosmos: The Ionian Tradition of Scientific Philosophy.
Princeton, NJ.
(2010). The Texts of Early Greek Philosophy: The Complete Fragments and Selected
Testimonies of the Major Presocratics. 2 vols. Cambridge.
Granger, H. (2007). ‘Poetry and Prose: Xenophanes of Colophon’. TAPhA 137:
403–33.
(2008). ‘The Proem of Parmenides’ Poem’. AncPhil 28: 1–20.
(2010). ‘Parmenides of Elea: Rationalist or Dogmatist?’ AncPhil 30: 15–38.
Graziosi, B. (2002). Inventing Homer: The Early Reception of Epic. Cambridge.
218 Bibliography
(2006). ‘Il rapporto tra autore ed opera nella tradizione biografica greca’. In
L’autore e l’opera nella Grecia antica, ed. F. Roscalla. Pavia: 155–74.
(2013). ‘The Poet in the Iliad’. In The Author’s Voice in Classical and Late
Antiquity, ed. A. Marmodoro and J. Hill. Oxford: 9–38.
Grethlein, J. (2008). ‘Memory and Material Objects in the Iliad and the Odyssey’.
JHS 128: 27–51.
(2010). The Greeks and Their Past: Poetry, Oratory and History in the Fifth
Century bce. Cambridge.
Grethlein, J., and Huitink, L. (2017). ‘Homer’s Vividness: An Enactive Approach’.
JHS 137: 67–91.
Griffin, J. (1986). ‘Homeric Words and Speakers’. JHS 106: 36–57.
Griffith, M. (1983). ‘Personality in Hesiod’. ClAnt 2: 37–65.
(1990). ‘Contest and Contradiction in Early Greek Poetry’. In Cabinet of the
Muses: Essays on Classical and Comparative Literature in Honor of Thomas
G. Rosenmeyer, ed. M. Griffith and D. J. Mastronarde. Atlanta, GA:
185–207.
Grube, G. M. A. (1965). The Greek and Roman Critics. London.
Gudeman, A. (1934). Aristoteles: Peri Poietikes. Berlin.
Guthrie, W. K. C. (1962). A History of Greek Philosophy: Volume 1, The Earlier
Presocratics and the Pythagoreans. Cambridge.
(1965). A History of Greek Philosophy: Volume 2, The Presocratic Tradition from
Parmenides to Democritus. Cambridge.
Gutzwiller, K. (2010). ‘Literary Criticism’. In A Companion to Hellenistic
Literature, ed. J. J. Clauss and M. Cuypers. Malden, MA: 337–65.
Hahn, R. (2001). Anaximander and the Architects: The Contributions of Egyptian
and Greek Architectural Technologies to the Origins of Greek Philosophy.
Albany, NY.
Halliwell, S. (1986). ‘Where Three Roads Meet: A Neglected Detail in the
Oedipus Tyrannus’. JHS 106: 187–90.
(1997). ‘Review of Nightingale (1995)’. AncPhil 17: 452–7.
(1998). Aristotle’s Poetics. 2nd ed. London.
(2002). The Aesthetics of Mimesis: Ancient Texts and Modern Problems.
Princeton, NJ.
(2011). Between Ecstasy and Truth: Interpretations of Greek Poetics from Homer to
Longinus. Oxford.
Harder, A. (2012). ‘Callimachus’ Aetia’. In A Companion to Hellenistic Literature,
ed. J. J. Clauss and M. Cuypers. Oxford: 92–105.
Hardie, A. (2013). ‘Empedocles and the Muse of the Agathos Logos’. AJPh 134:
209–46.
Hardie, P. (1985). ‘Imago mundi: Cosmological and Ideological Aspects of the
Shield of Achilles’. JHS 105: 11–31.
(1995). ‘The Speech of Pythagoras in Ovid Metamorphoses 15: Empedoclean
Epos’. CQ 45: 204–14.
Harriott, R. M. (1969). Poetry and Criticism before Plato. London.
Harvey, A. E. (1955). ‘The Classification of Greek Lyric Poetry’. CQ 5: 157–75.
Bibliography 219
Haubold, J. H. (2010). ‘Shepherd, Farmer, Poet, Sophist: Hesiod on His Own
Reception’. In Plato and Hesiod, ed. G. R. Boys-Stones and J. H. Haubold.
Oxford: 11–30.
Havelock, E. A. (1958). ‘Parmenides and Odysseus’. HSPh 63: 133–43.
(1963). Preface to Plato. Cambridge, MA.
(1983). ‘The Linguistic Task of the Presocratics’. In Language and Thought in
Early Greek Philosophy, ed. K. Robb. La Salle, IL: 7–82.
Heath, M. (1985). ‘Hesiod’s Didactic Poetry’. CQ 35: 245–63.
(1987). The Poetics of Greek Tragedy. London.
(1991). ‘The Universality of Poetry in Aristotle’s Poetics’. CQ 41: 389–402.
(1999). ‘Longinus On Sublimity’. PCPhS 45: 43–74.
(2002). Interpreting Classical Texts. London.
(2009). ‘Cognition in Aristotle’s Poetics’. Mnemosyne 62: 51–75.
(2013). Ancient Philosophical Poetics. Cambridge.
Heidegger, M. (1962). Being and Time. Translated by J. Macquarrie and
E. Robinson. Oxford. Originally published as Sein und Zeit, Tübingen,
1927.
(1992). Parmenides. Translated by A. Schuwer and R. Rojcewicz. Bloomington,
IN. Originally published as Parmenides, Frankfurt, 1982.
(2002). ‘The Origin of the Work of Art’. In Off the Beaten Track. Translated by
J. Young and K. Haynes. Cambridge. Originally published as Holzwege,
Frankfurt, 1950.
Heiden, B. (2007). ‘The Muses’ Uncanny Lies: Hesiod, “Theogony” 27 and Its
Translators’. AJPh 128: 153–75.
Heitsch, E. (1966). ‘Das Wissen des Xenophanes’. RhM 109: 193–235.
(1983). Xenophanes: Die Fragmente. Munich.
Helms, M. W. (1988). Ulysses’ Sail: An Ethnographic Odyssey of Power, Knowledge,
and Geographical Distance. Princeton, NJ.
Henry, M. (1986). ‘The Derveni Commentator as Literary Critic’. TAPhA 116:
149–64.
Herington, J. (1985). Poetry into Drama: Early Tragedy and the Greek Poetic
Tradition. Berkeley, CA.
Hershbell, J. P. (1968). ‘Empedocles’ Oral Style’. CJ 63: 351–7.
Hinds, S. (1998). Allusion and Intertext: Dynamics of Appropriation in Roman
Poetry. Cambridge.
Hobden, F. (2013). The Symposion in Ancient Greek Society and Thought.
Cambridge.
Hölscher, U. (1965). ‘Weltzeiten und Lebenszyklus: Eine Nachprüfung der
Empedokles-Doxographie’. Hermes 93: 7–33.
Hunter, R. L. (1995). ‘Written in the Stars: Poetry and Philosophy in the
Phaenomena of Aratus’. Arachnion 2: 1–34.
(1996). Theocritus and the Archaeology of Greek Poetry. Cambridge.
(2009). Critical Moments in Classical Literature. Cambridge.
(2014). Hesiodic Voices: Studies in the Ancient Reception of Hesiod’s Works and
Days. Cambridge.
220 Bibliography
Hunter, R. L., and Russell, D. A. (eds.) (2011). Plutarch: How to Study Poetry.
Cambridge.
Hunter, R. L., and Rutherford, I. (eds.) (2009). Wandering Poets in Ancient Greek
Culture: Travel, Locality and Pan-Hellenism. Cambridge.
Hussey, E. (1972). The Presocratics. New York, NY.
(1982). ‘Epistemology and Meaning in Heraclitus’. In Language and Logos, ed.
M. Schofield and M. C. Nussbaum. Cambridge: 33–59.
(1990). ‘The Beginnings of Epistemology: From Homer to Philolaus’. In
Epistemology, ed. S. Everson. Cambridge: 11–38.
(2006). ‘Parmenides on Thinking’. In Common to Body and Soul: Philosophical
Approaches to Explaining Living Behaviour in Greco-Roman Antiquity, ed.
R. A. H. King. Berlin: 11–38.
Inwood, B. (2001). The Poem of Empedocles: A Text and Translation with
a Commentary. 2nd ed. Toronto.
Ioli, R. (2003). ‘Senofane B34DK e il conoscere’. GIF 55: 199–219.
Irwin, E. (2005). Solon and Early Greek Poetry: The Politics of Exhortation.
Cambridge.
Jaeger, W. (1932). ‘Tyrtaios über die wahre Άρετή’. SBBerlin 23: 537–68.
(1939). Paideia: The Ideal of Greek Culture: In Search of the Divine Centre. 3 vols.
Translated by G. Highet. Oxford.
(1947). The Theology of the Early Greek Philosophers. Oxford.
Jakobson, R. (1960). ‘Linguistics and Poetics’. In Style in Language, ed.
T. A. Sebeok. Cambridge, MA: 350–77.
Janko, R. (1981). ‘The Structure of the Homeric Hymns: A Study in Genre’.
Hermes 109: 9–24.
(1982). Homer, Hesiod and the Hymns: Diachronic Development in Epic Diction.
Cambridge.
(1992). The Iliad: A Commentary: Volume 4, Books 13–16. Cambridge.
(2000). Philodemus on Poems, Book 1. Oxford.
(2004). ‘Empedocles, “On Nature” i 233–364: A New Reconstruction of “P.
Strasb. Gr.” Inv. 1665-6’. ZPE 150: 1–26.
(2010). Philodemus on Poems, Books 3–4: With the Fragments of Aristotle on Poets.
Oxford.
(2016). ‘Parmenides in the Derveni Papyrus: New Images for a New Edition’.
ZPE 200: 3–23.
Jauss, H. R. (1970). ‘Literary History as a Challenge to Literary Theory’.
Translated by E. Benzinger. New Literary History 2: 7–37. Originally pub-
lished as sections x–xii of Literaturgeschichte als Provokation der
Literaturwissenschaft, Konstanz, 1967.
Johansen, T. (2016). ‘Parmenides’ Likely Story’. OSAPh 50: 1–29.
Jordan, M. D. (1981). ‘A Preface to the Study of Philosophic Genres’. Philosophy &
Rhetoric 14: 199–211.
Kahn, C. H. (1960). ‘Religion and Natural Philosophy in Empedocles’ Doctrine of
the Soul’. Archiv für Geschichte der Philosophie 42: 3–35.
Bibliography 221
(1979). The Art and Thought of Heraclitus: An Edition of the Fragments with
Translation and Commentary. Cambridge.
(2001). Pythagoras and the Pythagoreans: A Brief History. Indianapolis, IN.
(2002). ‘Parmenides and Plato’. In Caston and Graham (eds.), 82–93.
(2003). ‘Writing Philosophy: Prose and Poetry from Thales to Plato’. In Yunis
(ed.), 139–61.
(2005). ‘Parmenides and Being’. In Frühgriechisches Denken, ed. G. Rechenauer.
Göttingen: 216–26.
(2009). Essays on Being. Oxford.
Kamtekar, R. (2009). ‘Knowing by Likeness in Empedocles’. Phronesis 54: 215–38.
Kannicht, R. (1988). The Ancient Quarrel between Philosophy and Poetry: Aspects of
the Greek Conception of Literature. Christchurch.
Kennedy, G. A. (1989). ‘Language and Meaning in Archaic and Classical Greece’.
In The Cambridge History of Literary Criticism: Volume 1: Classical Criticism,
ed. G. A. Kennedy. Cambridge: 78–91.
Kerferd, G. B. (1981). The Sophistic Movement. Cambridge.
Kern, O. (1888). ‘Empedokles und die Orphiker’. Archiv für Geschichte der
Philosophie 1: 498–508.
Kingsley, P. (1995). Ancient Philosophy, Mystery, and Magic: Empedocles and
Pythagorean Tradition. Oxford.
(1999). In the Dark Places of Wisdom. Inverness, CA.
(2002). ‘Empedocles for the New Millennium’. AncPhil 22: 333–413.
(2003). Reality. Inverness, CA.
Kirk, G. S. (1990). The Iliad: A Commentary, Volume 2: Books 5–8. Cambridge.
Kirk, G. S., and Raven, J. E. (1957). The Presocratic Philosophers: A Critical History.
Cambridge.
Kivilo, M. (2010). Early Greek Poets’ Lives: The Shaping of the Tradition. Leiden.
Koethe, J. (2009). ‘Poetry and Truth’. Midwest Studies in Philosophy 33: 53–60.
Koller, H. (1954). Die Mimesis in der Antike: Nachahmung, Darstellung, Ausdruck.
Bern.
Koning, H. (2010). Hesiod: The Other Poet. Leiden.
Kouremenos, T., Parássoglou, G. M., and Tsantsanoglou, K. (2006). The Derveni
Papyrus. Florence.
Kövecses, Z. (2010). Metaphor: A Practical Introduction. 2nd ed. Oxford.
Kowalzig, B. (2006). ‘The Aetiology of Empire? Hero-Cult and Athenian
Tragedy’. BICS 87: 79–98.
(2007). Singing for the Gods: Performances of Myth and Ritual in Archaic and
Classical Greece. Oxford.
Kranz, W. (1916). ‘Über Aufbau und Bedeutung des Parmenideischen Gedichtes’.
Sitzungsberichte der königlich preussischen Akademie der Wissenschaften 47:
1158–76.
(1944). ‘Lukrez und Empedokles’. Philologus 96: 68.
Kraus, M. (1987). Name und Sache: ein Problem im frühgriechischen Denken.
Amsterdam.
222 Bibliography
(2013). ‘Parmenides’. In Die Philosophie Der Antike. Vol. 1: Frühgriechische
Philosophie, ed. H. Flashar, D. Bremer, and G. Rechenauer. Basel: 441–530.
Krischer, T. (1965). ‘ΕΤΥΜΟΣ und ΑΛΗΘΗΣ’. Philologus 109: 161–74.
Kristeva, J. (1980). Desire in Language: A Semiotic Approach to Literature and Art.
New York, NY.
Kullmann, W. (1985). ‘Gods and Men in the Iliad and the Odyssey’. HSPh 89: 1–23.
Kurke, L. (1990). ‘Pindar’s Sixth Pythian and the Tradition of Advice Poetry’.
TAPhA 120: 85–107.
(2010). Aesopic Conversations: Popular Tradition, Cultural Dialogue, and the
Invention of Greek Prose. Princeton, NJ.
Laird, A. (2003). ‘Figures of Allegory from Homer to Latin Epic’. In Boys-Stones
(ed.), 151–75.
(2006). ‘The Value of Ancient Literary Criticism’. In Ancient Literary Criticism,
ed. A. Laird. Oxford: 1–36.
Lakoff, G., and Johnson, M. (1980). Metaphors We Live By. Chicago, IL.
(1999). Philosophy in the Flesh: The Embodied Mind and Its Challenge to Western
Thought. New York, NY.
Laks, A. (1999). ‘Soul, Sensation and Thought’. In Long (ed.), 250–70.
(2001). ‘Ecriture, prose, et les débuts de la philosophie grecque’. Methodos 1:
131–51.
(2002). ‘Reading the Readings: On the First Person Plurals in the Strasburg
Empedocles’. In Caston and Graham (eds.), 127–37.
(2004). Le Vide et la haine: éléments pour une histoire archaïque de la négativité.
Paris.
(2005). ‘Some Thoughts about the Empedoclean Cosmic and Demonic Cycles’.
In Pierris (ed.), 265–82.
(2013). ‘Phenomenon and Reference: Revisiting Parmenides, Empedocles, and
the Problem of Rationalization’. In Modernity’s Classics, ed. S. C. Humphreys
and R. Wagner. New York, NY: 165–86.
(2018). The Concept of Presocratic Philosophy. Translated by G. W. Most.
Princeton, NJ. Originally published as Introduction à la ‘philosophie
présocratique’, Paris, 2006.
Laks, A., and Louguet, C. (2002). Qu’est-ce que la philosophie présocratique?/ What
Is Presocratic Philosophy? Paris.
Laks, A., and Most, G. W. (eds.) (1997). The Derveni Papyrus. Oxford
Lamarque, P. (2009a). ‘Poetry and Abstract Thought’. Midwest Studies in
Philosophy 33: 37–52.
(2009b). The Philosophy of Literature. Oxford.
Lanata, G. (1963). Poetica preplatonica, testimonianze e frammenti. Florence.
Lang, B. (1990). The Anatomy of Philosophical Style: Literary Philosophy and the
Philosophy of Literature. Oxford.
Lapini, W. (2013). Testi frammentari e critica del testo: problemi di filologia filosofia
greca. Rome.
Latona, M. J. (2008). ‘Reining in the Passions: The Allegorical Interpretation of
Parmenides B Fragment 1’. AJPh 129: 199–230.
Bibliography 223
Lavery, J. (2007). ‘Philosophical Genres and Literary Forms: A Mildly Polemical
Introduction’. Poetics Today 28: 171–89.
Lecznar, A. (2020). Dionysus after Nietzsche. Cambridge.
Ledbetter, G. M. (2003). Poetics before Plato: Interpretation and Authority in Early
Greek Theories of Poetry. Princeton, NJ.
Lefkowitz, M. (2013). The Lives of the Greek Poets. 2nd ed. London.
Lesher, J. H. (1981). ‘Perceiving and Knowing in the Iliad and Odyssey’. Phronesis
26: 2–24.
(1983). ‘Xenophanes’ Scepticism’. In Essays in Ancient Greek Philosophy, Vol. 2,
ed. J. Anton and T. Preus. New York, NY: 20–40.
(1984). ‘Parmenides’ Critique of Thinking. The Poluderis Elenchos of Fragment
7’. OSAPh 2: 1–30.
(1992). Xenophanes of Colophon: A Text and Translation with a Commentary.
Toronto.
(1994). ‘The Emergence of Philosophical Interest in Cognition’. OSAPh
12: 1–34.
(1999). ‘Early Interest in Knowledge’. In Long (ed.), 225–49.
(2008). ‘The Humanizing of Knowledge in Presocratic Thought’. In Curd and
Lewis, D. (1978). ‘Truth in Fiction’. American Philosophical Quarterly 15: 37–46.
Lilja, S. (1968). On the Style of the Earliest Greek Prose. Helsinki.
Lightfoot, J. L. (2009). Hellenistic Collection: Philitas, Alexander of Aetolia,
Hermesianax, Euphorion, Parthenius. Cambridge, MA.
Lincoln, B. (1999). Theorizing Myth: Narrative, Ideology, and Scholarship. London
and Chicago, IL.
Llewellyn-Jones, L. (2003). Aphrodite’s Tortoise: The Veiled Woman of Ancient
Greece. Swansea.
Lloyd, G. E. R. (1987). The Revolutions of Wisdom: Studies in the Claims and
Practice of Ancient Greek Science. Berkeley, CA.
(1990). Demystifying Mentalities. Oxford.
(2006). ‘The Pluralism of Intellectual Life before Plato’. In Principles and
Practices in Ancient Greek and Chinese Science. Aldershot: 1–13. Originally
published as ‘Le Pluralisme de la vie intellectuelle avant Platon’, translated by
A. Laks, in Laks and Louguet (eds.), 1–13.
Lloyd-Jones, H. (1985). ‘Pindar and the Afterlife’. Entretiens Hardt 17: 245–83.
Long, A. A. (1963). ‘The Principles of Parmenides’ Cosmogony’. Phronesis 8:
90–107.
(1966). ‘Thinking and Sense-Perception in Empedocles: Mysticism or
Materialism’. CQ 16: 256–76.
(1985). ‘Early Greek Philosophy’. In The Cambridge History of Classical
Literature Volume 1: Greek Literature, ed. B. M. W. Knox and
P. E. Easterling. Cambridge: 245–57.
(1992). ‘Stoic Readings of Homer’. In Homer’s Ancient Readers: The
Hermeneutics of Greek Epic’s Earliest Exegetes, ed. R. Lamberton and
J. J. Keaney. Princeton, NJ: 41–66.
224 Bibliography
(1993). ‘Empedocles’ Cosmic Cycle in the ’Sixties’. In The Pre-Socratics:
A Collection of Critical Essays, ed. A. P. D. Mourelatos. Princeton, NJ:
397–425.
(1996). ‘Parmenides on Thinking Being’. Proceedings of the Boston Area
Colloquium of Ancient Philosophy 12: 125–51.
(1999). ‘The Scope of Early Greek Philosophy’. In Long (ed.), 1–21.
(ed.) (1999). The Cambridge Companion to Early Greek Philosophy. Cambridge.
Long, A. G. (2017). ‘Immortality in Empedocles’. Apeiron 50: 1–20.
Long, H. S. (1949). ‘The Unity of Empedocles’ Thought’. AJPh 70: 142–58.
Louden, B. (2011). Homer’s Odyssey and the Near East. Cambridge.
Lowe, N. J. (2000). The Classical Plot and the Invention of Western Narrative.
Cambridge.
Lyne, R. (2011). Shakespeare, Rhetoric and Cognition. Cambridge.
Mackenzie, M. M. (later publishing as McCabe, M. M.) (1982). ‘Parmenides’
Dilemma’. Phronesis 27: 1–12.
Mackenzie, T. (2016). ‘The Contents of Empedocles’ Poem: A New Argument for
the Single-Poem Hypothesis’. ZPE 200: 25–32.
(2017). ‘Parmenides and Early Greek Allegory’. MD 79: 31–59.
(forthcoming). ‘Empedocles, Personal Identity, and the Narrative of the Fallen
Daimōn’. Phoenix.
Macleod, C. W. (1982). Homer: Iliad Book xxiv. Cambridge.
Maehler, H. (1963). Die Auffassung des Dichterberufs im frühen Griechentum bis zur
Zeit Pindars. Göttingen.
Malkin, I. (1987). Religion and Colonization in Ancient Greece. Leiden.
Manetti, G. (1993). Theories of the Sign in Classical Antiquity. Bloomington, IN.
Mansfeld, J. (1964). Die Offenbarung des Parmenides und die menschliche Welt.
Assen.
(1985). ‘Historical and Philosophical Aspects of Gorgias’ On What Is Not’.
SicGymn 36: 243–71.
(1992). Heresiography in Context: Hippolytus’ Elenchos as a Source for Greek
Philosophy. Leiden.
(1995). ‘Insight by Hindsight: Intentional Unclarity in Presocratic Proems’.
BICS 40: 225–232.
(2005). ‘Minima Parmenidea’. Mnemosyne 58: 554–60.
Mansfeld, J., and Primavesi, O. (2012). Die Vorsokratiker: Griechisch/Deutsch.
Stuttgart.
Marcovich, M. (1978). ‘Xenophanes on Drinking-Parties and Olympic Games’.
ICS 3: 1–26.
Marías, J. (1971). ‘Literary Genres in Philosophy’. In Philosophy as Dramatic
Theory. Translated by James Parsons. University Park, PA: 1–35.
Marmodoro, A., and Hill, J. (eds.) (2013). The Author’s Voice in Classical and Late
Antiquity. Oxford.
Martin, A., and Primavesi, O. (1999). L’Empédocle de Strasbourg. Berlin.
Martin, R. P. (1989). The Language of Heroes: Speech and Performance in the Iliad.
Ithaca, NY.
Bibliography 225
(1992). ‘Hesiod’s Metanastic Poetics’. Ramus 21: 11–33.
(1993). ‘The Seven Sages as Performers of Wisdom’. In Cultural Poetics in
Archaic Greece: Cult, Performance, Politics, ed. C. Dougherty and L. Kurke.
Cambridge: 108–30.
(2009). ‘Read on Arrival’. In Hunter and Rutherford (eds.), 80–104.
Maslov, B. (2015). Pindar and the Emergence of Literature. Cambridge.
Mason, R. (1988). ‘Parmenides and Language’. AncPhil 8: 149–66.
McCabe, M. M. (previously publishing as Mackenzie, M. M.) (2000). Plato and
His Predecessors. Cambridge.
McNelis, C., and Sens, A. (2016). The Alexandra of Lycophron: A Literary Study.
Oxford.
Meijering, R. (1987). Literary and Rhetorical Theories in Greek Scholia. Groningen.
Metcalf, C. (2015). The Gods Rich in Praise. Oxford.
Meuli, K. (1935). ‘Scythica’. Hermes 70: 121–76.
Miller, M. (2006). ‘Ambiguity and Transport: Reflections on the Proem to
Parmenides’ Poem’. OSAPh 30: 1–47.
Mitsis, P. (1993). ‘Committing Philosophy on the Reader: Didactic Coercion and
Reader Autonomy in De Rerum Natura’. MD 31: 111–28.
Mogyoródi, E. (2006). ‘Xenophanes’ Epistemology and Parmenides’ Quest for
Knowledge’. In Sassi (ed.), 123–60.
Montiglio, S. (2000). ‘Wandering Philosophers in Classical Greece’. JHS 120:
86–105.
(2005). Wandering in Ancient Greek Culture. Chicago, IL.
Morgan, K. A. (2000). Myth and Philosophy from the Presocratics to Plato.
Cambridge.
Morrison, A. D. (2007). The Narrator in Archaic Greek and Hellenistic Poetry. 1st
ed. Cambridge.
Morrison, J. S. (1955). ‘Parmenides and Er’. JHS 75: 59–68.
Moss, J. (2007). ‘What Is Imitative Poetry and Why Is It Bad?’ In The
Cambridge Companion to Plato’s Republic, ed. G. R. F. Ferrari.
Cambridge: 415–44.
Most, G. W. (1993a). ‘Hesiod and the Textualization of Personal Temporality’. In
La componente autobiografica nella poesia greca e latina fra realtà e artificio
letterario, ed. G. Arrighetti and F. Montanari. Pisa: 73–91.
(1993b). ‘Die Früheste erhaltene griechische Dichterallegorese’. RhM 136:
209–12.
(1999). ‘The Poetics of Early Greek Philosophy’. In Long (ed.), 332–62.
(2002). ‘Heidegger’s Greeks’. Arion: A Journal of Humanities and the Classics 10:
83–98.
(2007). ‘ἄλλος δ’ ἐξ ἄλλου δέχεται. Presocratic Philosophy and Traditional
Greek Epic’. In Literatur und Religion, 1: Wege zu einer mythisch-rituellen
Poetik bei den Griechen, ed. A. Bierl, R. Lämmle and K. Wesselmann. Berlin:
271–301.
(2011). ‘“What Ancient Quarrel between Philosophy and Poetry?”’ In Plato and
the Poets, ed. P. Destrée and F.-G. Herrmann. Leiden: 1–20.
226 Bibliography
Mourelatos, A. P. D. (1987). ‘Gorgias on the Function of Language’. Philosophical
Topics 15: 135–70.
(2002). ‘La Terre et les étoiles dans la cosmologie de Xénophane’. In Laks and
Louguet (eds.), 331–50.
(2008a). The Route of Parmenides: A Study of Word, Image, and Argument in the
Fragments. Expanded ed. Las Vegas, NV.
(2008b). ‘Heraclitus, Parmenides, and the Naïve Metaphysics of Things’. In
The Route of Parmenides: A Study of Word, Image, and Argument in the
Fragments. Rev., expanded ed. Las Vegas, NV: 299–332.
(2012). ‘Parmenides, Early Greek Astronomy, and Modern Scientific Realism’.
In Parmenides, Venerable and Awesome. Plato, Theaetetus 183e: Proceedings of
the International Symposium, ed. N.-L. Cordero. Las Vegas, NV: 167–90.
Munteanu, D. L. (2012). Tragic Pathos: Pity and Fear in Greek Philosophy and
Tragedy. Cambridge.
Murdoch, I. (1970). The Sovereignty of Good. London.
Murray, O. (1993). Early Greece. Cambridge, MA.
Murray, P. (1981). ‘Poetic Inspiration in Early Greece’. JHS 101: 87–100.
(1996). Plato on Poetry: Ion; Republic 376e–398b9; Republic 595–608b10.
Cambridge.
Nagy, G. (1979). The Best of the Achaeans: Concepts of the Hero in Archaic Greek
Poetry. Baltimore, MD.
(1989). ‘Early Greek Views of Poets and Poetry’. In The Cambridge History of
Literary Criticism Vol. 1: Classical Criticism, ed. G. A. Kennedy.
Cambridge: 1–77.
(1990). Greek Mythology and Poetics. Ithaca, NY.
(1994). ‘Genre and Occasion’. Mètis: Anthropologie des mondes grecs anciens 9–10:
11–25.
Nehamas, A. (1982). ‘Plato on Imitation and Poetry in Republic 10’. In Plato on
Beauty, Wisdom, and the Arts, ed. J. M. E. Moravcsik and P. Temko. Totowa,
NJ: 47–78.
(1988). ‘Plato and the Mass Media’. The Monist 71: 214–34.
(2002). ‘Parmenidean Being/Heraclitean Fire’. In Caston and Graham (eds.),
45–64.
Nelis, D. P. (2004). ‘Georgics 2.458–542: Virgil, Aratus and Empedocles’.
Dictynna 1. Available online at https://journals.openedition.org/dictynna/16
1 (accessed 18 September 2019).
Nestle, W. (1942). Vom Mythos zum Logos. Stuttgart.
Nightingale, A. W. (1995). Genres in Dialogue: Plato and the Construct of
Philosophy. Cambridge.
(2007). ‘The Philosophers in Archaic Greek Culture’. In The Cambridge
Companion to Archaic Greece, ed. H. A. Shapiro. Cambridge: 169–98.
Norden, E. (1913). Agnostos Theos: Untersuchungen zur Formengeschichte religiöser
Rede. Leipzig.
Noussia-Fantuzzi, M. (2010). Solon the Athenian the Poetic Fragments. Leiden.
Novitz, D. (1987). Knowledge, Fiction, and Imagination. Philadelphia, PA.
Bibliography 227
Nünlist, R. (1998). Poetologische Bildersprache in der frühgriechischen Dichtung.
Berlin.
(2005). ‘Poetological Imagery in Empedocles’. In Pierris (ed.), 73–92.
(2009). The Ancient Critic at Work: Terms and Concepts of Literary Criticism in
Greek Scholia. Cambridge.
Obbink, D. (1993). ‘The Addressees of Empedocles’. MD 31: 51–98.
(2006). ‘A New Archilochus Poem’. ZPE 156: 1–9.
O’Brien, D. (1969). Empedocles’ Cosmic Cycle. Cambridge.
(1970). ‘The Effect of a Simile: Empedocles’ Theories of Seeing and Breathing’.
JHS 90: 140–79.
(1995). ‘Empedocles Revisited’. AncPhil 15: 403–70.
(2001). ‘Empedocles: The Wandering Daimon and the Two Poems’.
Aevum(ant) 1 (n.s.): 79–179.
(2005). ‘Empedocles: A Synopsis’. In Frühgriechisches Denken, ed.
G. Rechenauer. Göttingen: 316–42.
(2006). ‘Life beyond the Stars: Aristotle, Plato and Empedocles’. In Common to
Body and Soul: Philosophical Approaches to Explaining Living Behaviour in
Greco-Roman Antiquity, ed. R. A. H. King. Berlin: 49–102.
O’Keefe, T. (1997). ‘The Ontological Status of Sensible Qualities for Democritus
and Epicurus’. AncPhil 17: 119–34.
Osborne, C. (later publishing as Rowett, C.). (1987). ‘Empedocles Recycled’. CQ
37: 24–50.
(1998). ‘Was Verse the Default Form for Presocratic Philosophy?’ In Form and
Content in Didactic Poetry, ed. C. Atherton. Bari: 23–35.
Owen, G. E. L. (1960). ‘Eleatic Questions’. CQ 10: 84–102.
Page, D. L. (1955). Sappho and Alcaeus: An Introduction to the Study of Ancient
Lesbian Poetry. Oxford.
Palmer, J. (1999). Plato’s Reception of Parmenides. Oxford.
(2008). ‘Classical Representations and Uses of the Presocratics’. In Curd and
(2009). Parmenides and Presocratic Philosophy. Oxford.
(2013). ‘Revelation and Reasoning in Kalliopeia’s Address to Empedocles’.
Rhizomata 1: 308–29.
(2016). ‘Parmenides’. In The Stanford Encyclopedia of Philosophy, ed.
E. N. Zalta. https://plato.stanford.edu/archives/fall2016/entries/parmeni
des/ (accessed 18 September 2019).
Palumbo, L. (2007). ‘Empedocle e il linguaggio poetico’. In Empedocle tra poesia,
medicina, filosofia e politica, ed. G. Casertano. Naples: 83–107.
Papadopoulou, I., and Muellner, L. (eds.) (2014). Poetry as Initiation: The Center
for Hellenic Studies Symposium on the Derveni Papyrus. Cambridge, MA.
Parker, R. (1983). Miasma: Pollution and Purification in Early Greek Religion.
Oxford.
(1995). ‘Early Orphism’. In The Greek World, ed. A. Powell. London: 483–510.
(1996). Athenian Religion: A History. Oxford.
228 Bibliography
(1999). ‘Through a Glass Darkly: Sophocles and the Divine’. In Sophocles
Revisited: Essays Presented to Sir Hugh Lloyd-Jones, ed. J. Griffin. Oxford: 11–30.
(2011). On Greek Religion. Ithaca, NY.
(2017). Greek Gods Abroad: Names, Natures, and Transformations. Oakland, CA.
Patzer, A. (1986). Der Sophist Hippias als Philosophiehistoriker. Munich.
(2006). Wort und Ort: Oralität und Literarizität im sozialen Kontext der
frühgriechischen Philosophie. Munich.
Pavel, T. (1986). Fictional Worlds. Cambridge, MA.
Pellikaan-Engel, M. E. (1974). Hesiod and Parmenides: A New View on Their
Cosmologies and on Parmenides’ Proem. Amsterdam.
Peponi, A.-E. (2013). ‘Dithyramb in Greek Thought’. In Dithyramb in Context, ed.
B. Kowalzig and P. Wilson, 353–67. Oxford.
Petrovic, A., and Petrovic, I. (2016). Inner Purity and Pollution in Greek Religion:
Volume i: Early Greek Religion. Oxford.
Pfeiffer, R. (1968). History of Classical Scholarship: From the Beginnings to the End of
the Hellenistic Age. Oxford.
Piano, V. (2016). Il papiro di Derveni tra religione e filosofia. Florence.
Picot, J.-C. (2007). ‘Empedocles, Fragment 115.3: Can One of the Blessed Pollute
His Limbs with Blood?’ In Reading Ancient Texts. Volume i: Presocratics and
Plato, ed. S. Stern-Gillet and K. Corrigan. Leiden: 41–56.
(2008a). ‘Empédocle pouvait-il faire de la lune le séjour des bienheureux?’
Organon 37: 9–38.
(2008b). ‘La Brilliance de Nestis (Empédocle, Fr. 96)’. RPhA 26: 75–100.
Picot, J.-C., and Berg, W. (2015). ‘Lions and promoi: Final Phase of Exile for
Empedocles’ daimones’. Phronesis 60: 380–409.
(2018). ‘Apollo, Eros, and Epic Allusions in Empedocles, Frr. 134 and 29 DK’.
AJPh 139: 365–96.
Pierris, A. L. (ed.) (2005). The Empedoclean Κόσμος: Structure, Process and the
Question of Cyclicity Part 1: Papers. Patras.
Popa, T. (2000). ‘The Reception of Parmenides’ Poetry in Antiquity’. StudClas
34–6: 5–27.
Porter, J. I. (2010). The Origins of Aesthetic Thought in Ancient Greece: Matter,
Sensation, and Experience. Cambridge.
(2016). The Sublime in Antiquity. Cambridge.
Pozdnev, M. (2016). ‘Homerstudien zur Zeit des Xenophanes’. WS 129: 7–24.
Pratt, L. H. (1993). Lying and Poetry from Homer to Pindar: Falsehood and
Deception in Archaic Greek Poetics. Ann Arbor, MI.
Price, S. R. F. (1984). ‘Gods and Emperors: The Greek Language of the Roman
Imperial Cult’. JHS 104: 79–95.
Primavesi, O. (2001). ‘La daimonologia della fisica empedoclea’. Aevum(ant) 1 (n.
s.): 3–68.
(2005). ‘Theologische Allegorie: Zur philosophischen Funktion einer poe-
tischen Form bei Parmenides und Empedokles’. In Wissensvermittlung in
Dichterische Gestalt = Palingenesia 85, ed. M. Horster and C. Reitz. Stuttgart:
69–93.
Bibliography 229
(2007). ‘Zur Überlieferung und Bedeutung des Empedokleischen Titels
“Καθαρμοί”’. In Katharsiskonzeptionen vor Aristoteles zum kulturellen
Hintergrund des Tragödiensatzes, ed. M. Vöhler and B. Seidensticker.
Berlin: 183–226.
(2008a). ‘Empedocles: Physical and Mythical Divinity’. In Curd and Graham
(eds.), 250–83.
(2008b). Empedokles ‘Physika’ I, Eine Rekonstruktion des zentralen
Gedankengangs. Berlin.
(2013a). ‘Le Chemin vers la révélation: lumière et nuit dans le proème de
Parménide’. PhilAnt 13: 37–82.
(2013b). ‘Empedokles’. In Die Philosophie der Antike. Vol. 1: Frühgriechische
Philosophie, ed. H. Flashar, D. Bremer and G. Rechenauer. Basel:
667–739.
(2016). ‘Empedocles’ Cosmic Cycle and the Pythagorean Tetractys’. Rhizomata
4: 5–29.
Prince, S. (2015). Antisthenes of Athens: Texts, Translations, and Commentary. Ann
Arbor, MI.
Pucci, P. (1987). Odysseus Polutropos: Intertextual Readings in the Odyssey and the
Iliad. Ithaca, NY.
Puelma, M. (1989). ‘Der Dichter und die Wahrheit in der griechischen Poetik von
Homer bis Aristoteles’. MH 46: 65–100.
Pugliese Cratelli, G. (1988). ‘La Θεά di Parmenide’. PP 43: 337–46.
Purves, A. C. (2010). Space and Time in Ancient Greek Narrative. Cambridge.
Putnam, H. (1978). Meaning and the Moral Sciences. London.
Rabel, R. J. (1997). Plot and Point of View in the Iliad. Ann Arbor, MI.
Race, W. H. (1982). The Classical Priamel from Homer to Boethius. Leiden.
(1997). Pindar: Nemean Odes, Isthmian Odes, Fragments. Cambridge, MA.
Rangos, S. (2012). ‘Empedocles on Divine Nature’. Revue de Métaphysique et de
Morale 3: 315–38.
Ranzato, S. (2015). Il kouros e la verità: polivalenza delle immagini nel poema di
Parmenide. Pisa.
Rashed, M. (2008). ‘Le Proème des “Catharmes” d’Empédocle. Reconstitution et
commentaire’. Elenchos 29: 7–38.
(2018). La Jeune fille et la sphère: études sur Empédocle. Paris.
Reibaud, L. (2012). Xénophane, Œuvre Poétique. Paris.
Reinhardt, K. (1916). Parmenides und die Geschichte der griechischen Philosophie.
Bonn.
(1960). ‘Personifikation und Allegorie’. In Vermächtnis der Antike: Gesammelte
Essays zur Philosophie und Geschichtsschreibung, ed. C. Becker.
Göttingen: 7–40.
(1993). ‘The Relation between the Two Parts of Parmenides’ Poem’. In The Pre-
Socratics: A Collection of Critical Essays, ed. A. P. D. Mourelatos. Princeton,
NJ: 293–311.
Renehan, R. (1980). ‘On the Greek Origins of the Concepts Incorporeality and
Immateriality’. GRBS 21: 105–38.
230 Bibliography
Richards, I. A. (1970). Poetries and Sciences. Rev. ed. London.
Richardson, N. J. (1974). The Homeric Hymn to Demeter. Oxford.
(1975). ‘Homeric Professors in the Age of the Sophists’. PCPhS 21: 65–81.
(1983). ‘Recognition Scenes in the Odyssey and Ancient Literary Criticism’.
PLLS 4: 219–35.
(1992). ‘Panhellenic Cults and Panhellenic Poets’. In The Cambridge Ancient
History: Volume 5: The Fifth Century bc, ed. D. M. Lewis, J. K. Davies,
J. Boardman and M. Ostwald. 2nd ed. Cambridge: 223–44.
(2010). Three Homeric Hymns: To Apollo, Hermes, and Aphrodite. Cambridge.
Riedweg, C. (1987). Mysterienterminologie bei Platon, Philon und Klemens von
Alexandrien. Berlin.
(1995). ‘Orphisches bei Empedokles’. A&A 41: 34–59.
(2008). Pythagoras: His Life, Teaching, and Influence. Translated by Steven
Rendall. Ithaca, NY.
Ritoók, Z. (1989). ‘The Views of Early Greek Epic on Poetry and Art’. Mnemosyne
42: 331–48.
Robbiano, C. (2006). Becoming Being: On Parmenides’ Transformative Philosophy.
Sankt Augustin.
Rodriguez, C. M. (2005). Orfeo y el orfismo en la poesia de Empedocles. Madrid.
Romano, A. (2012). ‘Euripidean Explainers’. In Greek Notions of the Past in the
Archaic and Classical Eras, ed. J. Marincola, L. Llewellyn-Jones and
C. Maciver. Edinburgh: 127–43.
Romm, J. (1992). The Edges of the Earth in Ancient Thought. Princeton, NJ.
Rosenfeld-Löffler, A. (2006). La Poétique d’Empédocle: cosmologie et métaphore. Bern.
Rösler, W. (1980a). ‘Die Entdeckung der Fiktionalität in der Antike’. Poetica 12:
283–319.
(1980b). Dichter und Gruppe. Munich.
(1983). ‘Der Anfang der “Katharmoi” des Empedokles’. Hermes 111: 170–9.
Rossi, L. E. (1971). ‘I generi letterari e le loro leggi scritte e non scritte nelle
letterature classiche’. BICS 18: 69–94.
Rowett, C. (previously publishing as Osborne, C.) (2016). ‘Love, Sex and the
Gods: Why Things Have Divine Names in Empedocles’ Poem, and Why
They Come in Pairs’. Rhizomata 4: 80–110.
Rumpf, L. (2005). ‘Lukrez und Parmenides’. Philologus 149: 78–95.
Russell, D. A. (1981). Criticism in Antiquity. London.
Russell, D. A., and Winterbottom, W. (1972). Ancient Literary Criticism: The
Principal Texts in New Translations. Oxford.
Rutherford, I. (2011). ‘Singing Myth: Pindar’. In A Companion to Greek Mythology,
ed. K. Dowden and N. Livingstone. Malden, MA: 109–23.
Rutherford, R. B. (1982). ‘Tragic Form and Feeling in the Iliad’. JHS 102: 145–60.
(1992). Homer: Odyssey Books xix and xx. Cambridge.
Samons, L. J. (1999). ‘Aeschylus, the Alkmeonids and the Reform of the
Areopagos’. CJ 94: 221–33.
Santamaría, M. A. (2018). The Derveni Papyrus: Unearthing Ancient Mysteries.
Leiden and Boston, MA.
Bibliography 231
Sassi, M. M. (1978). Le teorie della percezione in Democrito. Florence.
(1988). ‘Parmenide al bivio. Per un’interpretazione del proemio’. PP 43:
383–96.
(ed.) (2006). La costruzione del discorso filosofico nell’età dei presocratici. Pisa.
(2018). The Beginnings of Philosophy in Greece. Translated by M. Asuni.
Princeton, NJ. Originally published as Gli inizi della filosofia in Grecia,
Turin, 2009.
Schäfer, C. (1996). Xenophanes von Kolophon: ein Vorsokratiker zwischen Mythos
und Philosophie. Wiesbaden.
Schein, S. L. (1984). The Mortal Hero: An Introduction to Homer’s Iliad.
Berkeley, CA.
Schibli, H. S. (1990). Pherekydes of Syros. Oxford.
Schmalzriedt, E. (1970). Peri physeos. Zur Frühgeschichte der Buchtitel. Munich.
Schmitt Pantel, P. (1992). La Cité au banquet: histoire des repas publics dans les cités
grecques. Rome.
Schmitter, P. (1991). ‘Vom Mythos zum Logos: Erkenntniskritik und
Sprachreflexion bei den Vorsokratikern’. In Geschichte der Sprachtheorie 2:
Sprachtheorien der abendländischen Antike, ed. P. Schmitter. Tübingen:
57–86.
Schwabl, H. (1956). ‘Empedokles Fr. B110’. WS 69: 49–56.
(1963). ‘Hesiod und Parmenides: Zur Formung des Parmenideischen
Prooimions (28 B 1)’. RhM 106: 134–42.
Scruton, R. (2015). ‘Poetry and Truth’. In Gibson (ed.), 169–61.
Scully, S. (2003). ‘Reading the Shield of Achilles: Terror, Anger, Delight’. HSPh
101: 29–47.
Seaford, R. (2004) Money and the Early Greek Mind. Cambridge.
Sedley, D. (1982). ‘Two Conceptions of Vacuum’. Phronesis 27: 175–93.
(1998). Lucretius and the Transformation of Greek Wisdom. Cambridge.
(1999). ‘Parmenides and Melissus’. In Long (ed.), 113–33.
(2007). Creationism and Its Critics in Antiquity. Berkeley, CA.
(2008). ‘Atomism’s Eleatic Roots’. In Curd and Graham (eds.), 305–32.
Segal, C. P. (1962). ‘Gorgias and the Psychology of the Logos’. HSPh 66: 99–155.
(1974). ‘Eros and Incantation: Sappho and Oral Poetry’. Arethusa 7: 139–60.
(1983). ‘Kleos and Its Ironies in the Odyssey’. AC 52: 22–47.
Shankman, S. (1983). ‘Led by the Light of the Maeonian Star: Aristotle on Tragedy
and “Odyssey” 17.415–444’. ClAnt 2: 108–16.
Sharrock, A., and Morales, H. (eds.) (2000). Intratextuality: Greek and Roman
Textual Relations. Oxford.
Shaw, M. M. (2014). ‘Aither and the Four Roots in Empedocles’. Research in
Phenomenology 44: 170–93.
Shields, C. J. (2013). Aristotle. 2nd ed. London.
Shklovsky, V. (1990). ‘Art as Technique’. In Theory of Prose, translated by B. Sher.
Elmwood Park, IL: 1–14.
Sider, D. (1984). ‘Empedocles B 96 (462 Bollack) and the Poetry of Adhesion’.
Mnemosyne 37: 14–24.
232 Bibliography
(2001). ‘Fragments 1–22 W2: Text, Apparatus Criticus, and Translation’. In
Boedeker and Sider (eds.), 13–29.
(2005). The Fragments of Anaxagoras. 2nd ed. Sankt Augustin.
(2006). ‘The New Simonides and the Question of Historical Elegy’. AJPh 127:
327–46.
Silk, M. S. (1974). Interaction in Poetic Imagery: With Special Reference to Early
Greek Poetry. Cambridge.
Slaveva-Griffin, S. (2003). ‘Of Gods, Philosophers, and Charioteers: Content and
Form in Parmenides’ Proem and Plato’s “Phaedrus”’. TAPhA 133: 227–53.
Slings, S. R. (1990). ‘The I in Personal Archaic Lyric: An Introduction’. In The
Poet’s ‘I’ in Archaic Greek Lyric, ed. S. R. Slings. Amsterdam: 1–30.
Sluiter, I. (1997). ‘The Greek Tradition’. In The Emergence of Semantics in Four
Linguistic Traditions: Hebrew, Sanskrit, Greek, Arabic, ed. W. J. van Bekkum,
J. Houben, K. Versteegh and I. Sluiter. Amsterdam: 148–224.
Snell, B. (1924). Die Ausdrücke für den Begriff des Wissens in der vorplatonischen
Philosophie. Frankfurt.
(1944). ‘Die Nachrichten über die Lehren des Thales und die Anfänge der
griechischen Philosophie- und Literaturgeschichte’. Philologus 96:
170–82.
(1953). The Discovery of the Mind. Oxford.
(1973). ‘Wie die Griechen lernten, was geistige Tätigkeit ist’. JHS 93: 172–84.
Snyder, J. (1980). Puns and Poetry in Lucretius’ De Rerum Natura. Amsterdam.
Sontag, S. (1966). Against Interpretation. New York, NY.
Solmsen, F. (1965). ‘Love and Strife in Empedocles’ Cosmology’. Phronesis 10:
109–48.
Sommerstein, A. (1989). Aeschylus: Eumenides. Cambridge.
Stanford, W. B. (1983). Greek Tragedy and the Emotions: An Introductory Study.
London.
Stehle, E. (1997). Performance and Gender in Ancient Greece: Princeton, NJ.
(2001). ‘A Bard of the Iron Age and His Auxiliary Muse’. In Boedeker and Sider
(eds.), 106–19.
(2005). ‘The Addressees of Empedokles, Katharmoi Fr. B112’. AncPhil 25:
247–72.
Stein, H. (1852). Empedoclis Agrigentini fragmenta disposuit recensuit adnotavit
H. Stein. Præmissa est de Empedoclis scriptis disputatio. Bonn.
Steiner, D. (1993). ‘Pindar’s Oggetti Parlanti’. HSPh 95: 159–80.
Stevens, W. (1951). The Necessary Angel: Essays on Reality and the Imagination.
New York, NY.
Stolnitz, J. (1992). ‘On the Cognitive Triviality of Art’. British Journal of Aesthetics
32: 191–200.
Stokes, M. C. (1971). One and Many in Presocratic Philosophy. Washington, DC.
Stroh, W. (1976). ‘Hesiods lügende Musen’. In Studien zum antiken Epos, ed.
H. Görgemanns and E. A. Schmidt, 85–112. Meisenheim am Glan.
Struck, P. (2004). Birth of the Symbol: Ancient Readers at the Limits of Their Texts.
Princeton, NJ.
Bibliography 233
Svenbro, J. (1984). La parola e il marmo: alle origini della poetica greca. 2nd ed.
Turin.
Swift, L. A. (2010). The Hidden Chorus: Echoes of Genre in Tragic Lyric. Oxford.
(2012). ‘Archilochus the “Anti-hero”? Heroism, Flight and Values in
Homer and the New Archilochus Fragment (P.Oxy lxix 4708)’. JHS
132: 139–55.
Taplin, O. (1992). Homeric Soundings: The Shaping of the Iliad. Oxford.
Tarán, L. (1965). Parmenides: A Text with Translation, Commentary, and Critical
Essays. Princeton, NJ.
(1977). ‘Review Article of Mourelatos (1970)’. Gnomon 49: 651–66.
Tate, J. (1927). ‘The Beginnings of Greek Allegory’. CR 41: 214–15.
(1953). ‘Antisthenes Was Not an Allegorist’. Eranos 51: 14–22.
Thomas, R. (1992). Literacy and Orality in Ancient Greece. Cambridge.
Too, Y. L. (1998). The Idea of Ancient Literary Criticism. Oxford.
Tor, S. (2013). ‘Mortal and Divine in Xenophanes’ Epistemology’. Rhizomata 1:
248–82.
(2015). ‘Parmenides’ Epistemology and the Two Parts of His Poem’. Phronesis
60: 3–39.
(2016). ‘Heraclitus on Apollo’s Signs and His Own: Contemplating Oracles and
Philosophical Enquiry’. In Theologies of Ancient Greek Religion, ed.
E. Eidinow, J. Kindt and R. Osborne. Cambridge: 89–116.
(2017). Mortal and Divine in Early Greek Epistemology: A Study of Hesiod,
Xenophanes and Parmenides. Cambridge.
Traglia, A. (1952). Studi sulla lingua di Empedocle. Bari.
Trépanier, S. (2003). ‘Empedocles on the Ultimate Symmetry of the World’.
OSAPh 24: 1–57.
(2004). Empedocles: An Interpretation. London.
(2010). ‘Early Greek Theology: The Gods as Nature and Natural Gods’. In The
Gods of Ancient Greece: Identities and Transformations, ed. A. Erskine and
J. Bremmer. Edinburgh: 273–317.
(2014). ‘From Wandering Limbs to Limbless Gods: Δαίμων as Substance in
Empedocles’. Apeiron 47: 172–210.
(2017). ‘From Hades to the Stars: Empedocles on the Cosmic Habitats of Soul’.
ClAnt 36: 130–82.
Tsagalis, C. (2008). The Oral Palimpsest: Exploring Intertextuality in the Homeric
Epics. Washington, DC.
Ustinova, Y. (2009). Caves and the Ancient Greek Mind: Descending Underground
in the Search for Ultimate Truth. Oxford.
van der Ben, N. (1975). The Proem of Empedocles’ Peri Physios: Towards a New
Edition of All the Fragments. Amsterdam.
van Groningen, B. A. (1958). La Composition littéraire archaïque grecque: procédés et
réalisations. Amsterdam.
(1971). ‘Empédocle, poète’. Mnemosyne 24: 169–88.
van Noorden, H. (2014). Playing Hesiod: The ‘Myth of the Races’ in Classical
Antiquity. Cambridge.
234 Bibliography
Verdenius, W. J. (1942). Parmenides: Some Comments on His Poem. Groningen.
(1970). Homer, the Educator of the Greeks. Amsterdam.
(1981). ‘Gorgias’ Doctrine of Deception’. In The Sophists and Their Legacy, ed.
G. B. Kerferd. Wiesbaden: 116–28.
(1983). ‘The Principles of Greek Literary Criticism’. Mnemosyne 36: 14–59.
Vernant, J.-P. (1982). The Origins of Greek Thought. Translated from the French.
Ithaca, NY. Originally published as Les Origines de la pensée grecque, Paris,
1962.
(1983). Myth and Thought among the Greeks. Translated from the French.
London. Originally published as Mythe et pensée chez les Grecs, Paris, 1965.
Versnel, H. S. (2002). ‘The Poetics of the Magical Charm: An Essay in the Power
of Words’. In Magic and Ritual in the Ancient World, ed. P. Mirecki and
M. Meyer. Leiden: 105–58.
(2011). Coping with the Gods. Leiden.
Vieira, C. (2013). ‘Heraclitus’ Bow Composition’. CQ 63: 473–90.
Vlastos, G. (1946). ‘Parmenides’ Theory of Knowledge’. TAPhA 77: 66–77.
Volk, K. (2002). The Poetics of Latin Didactic: Lucretius, Vergil, Ovid, Manilius.
Oxford.
von Fritz, K. (1943). ‘NOOΣ and ΝΟΕΙΝ in the Homeric Poems’. CPh 38: 79–93
(1945). ‘ΝΟΥΣ, ΝΟΕΙΝ, and Their Derivatives in Pre-Socratic Philosophy
(Excluding Anaxagoras): Part I. From the Beginnings to Parmenides’. CPh
40: 223–42.
(1946). ‘ΝΟΥΣ, ΝΟΕΙΝ, and Their Derivatives in Pre-Socratic Philosophy
(Excluding Anaxagoras): Part II. The Post-Parmenidean Period’. CPh 41:
12–34.
Walsh, G. B. (1984). The Varieties of Enchantment: Early Greek Views of the Nature
and Function of Poetry. Chapel Hill, NC.
Wardy, R. B. (1988). ‘Eleatic Pluralism’. Archiv für Geschichte der Philosophie 70:
125–46.
(1996). The Birth of Rhetoric: Gorgias, Plato and Their Successors. London.
Warren, J. (2007). The Presocratics. London.
(2013). ‘Gods and Men in Xenophanes’. In Politeia in Greek and Roman
Philosophy, ed. V. Harte and M. Lane. Cambridge: 294–312.
Webb, R. (2009). Ekphrasis, Imagination and Persuasion in Ancient Rhetorical
Theory and Practice. Farnham.
Weber, M. (1947). The Theory of Social and Economic Organization. Translated by
A. M. Henderson and T. Parsons. Oxford. Originally published in Wirtschaft
und Gesellschaft, Tübingen, 1922: 1–180.
Weil, H. (1862). ‘Über Spuren strophischer Composition bei den alten grie-
chischen Elegikern’. RhM 17: 1–13.
West, D. (1969). The Imagery and Poetry of Lucretius. Edinburgh.
West, M. L. (1966). Hesiod: Theogony. Oxford.
(1971). Early Greek Philosophy and the Orient. Oxford.
(1974). Studies in Greek Elegy and Iambus. Berlin.
(1978). Hesiod: Works and Days. Oxford.
Bibliography 235
(1981). ‘The Singing of Homer and the Modes of Early Greek Music’. JHS 101:
113–29.
(1983). The Orphic Poems. Oxford.
(1992). Iambi et elegi graeci ante Alexandrum cantati, Vol. ii. 2nd ed. Oxford.
(1997). The East Face of Helicon. Oxford.
(1999). ‘The Invention of Homer’. CQ 49: 364–82.
(2003a). Homeric Hymns: with Homeric Apocrypha and Lives of Homer.
Cambridge, MA.
(2003b). Greek Epic Fragments: From the Seventh to the Fifth Centuries bc.
Cambridge, MA.
(2007). Indo-European Poetry and Myth. Oxford.
(2010). ‘Rhapsodes at Festivals’. ZPE 173: 1–13.
(2011). ‘Rhapsodes’. In The Homer Encyclopedia, ed. M. Finkelberg. Oxford.
Wilamowitz-Möllendorff, U. v. (1902). ‘Lesefruechte’. Hermes 37: 321–488.
(1962). Kleine Schriften, Vol. 4. ed. K. Latte. Berlin.
Willi, A. (2003). The Languages of Aristophanes: Aspects of Linguistic Variation in
Classical Attic Greek. Oxford.
(2008). Sikelismos: Sprache, Literatur und Gesellschaft im griechischen Sizilien (8.–
5. Jh. v. Chr.). Basel.
Williams, B. (2002). Truth and Truthfulness. Princeton, NJ.
(2006). The Sense of the Past: Essays in the History of Philosophy. ed. M. Burnyeat.
Princeton, NJ.
Wöhrle, G. (1993). ‘War Parmenides ein schlechter Dichter? Oder: Zur Form der
Wissensvermittlung in der frühgriechischen Philosophie’. In Vermittlung
und Tradierung von Wissen in der griechischen Kultur, ed. W. Kullmann
and J. Althoff, Tübingen: 167–80.
Wolfsdorf, D. (2009). ‘Empedocles and His Ancient Readers on Desire and
Pleasure’. OSAPh 36: 1–71.
Wordsworth, W., and Coleridge, S. T. (2005). Lyrical Ballads. ed. R. L. Brett and
A. R. Jones with an introduction by N. Roe. Abingdon. Originally published
London, 1798.
Wright, M. R. (1985). The Presocratics. Bristol.
(1995). Empedocles: The Extant Fragments. 2nd ed. London.
(1998). ‘Philosopher Poets: Parmenides and Empedocles’. In Form and Content
in Didactic Poetry, ed. C. Atherton. Bari: 1–22.
Yunis, H. (ed.) (2003). Written Texts and the Rise of Literate Culture in Ancient
Greece. Cambridge.
(2011). Plato: Phaedrus. Cambridge.
Zanker, G. (1981). ‘Enargeia in the Ancient Criticism of Poetry’. RhM 124: 297–311.
Zuntz, G. (1971). Persephone: Three Studies in Religion and Thought in Magna
Graecia. Oxford.
Index
Aeschylus, 136 emotion and literature, 9, 178, 179

Eumenides, 160 in Empedocles, 162–6
Aesop, 51 in Parmenides, 84–90
ainigmata. See allegory Empedocles, 43, 102–76, 183, 185, 192
Alcaeus, 58 ancient biography of, 1–3, 130–1
allegory, 12–13, 94, 97–100, 157, 161, 166, 171, and Democritus, 194–5, 196–7
183–8 and fifth-century theories of language, 198, 199
Anacreon, 58 and Gorgias, 189–90
anagnorisis, 164, See also Aristotle, Poetics as a god, 131–2
Anaxagoras, 184–6, 188, 192 one or two poems, 104–7
Antisthenes, 187–8 enargeia. See vividness, literary
Apollo, 28, 131, 137, 171, 184, See also Homer, Epimenides, 136–8, 139
Hymn to Apollo ethics, 116
Aratus, 181 Euripides, 88, 200
Archilochus, 108 exile, 5
Aristeas of Proconnesus, 50 and Empedocles, 132–5, 152, 158, 159
Aristophanes, 125, 200–1
Aristotle, 51, 106, 176, 189 Frege, Gottlob, 16
Poetics, 126–7, 162–4
treatment of Presocratics, 3 Geistesgeschichte, 3, 23
genre, 27, See also elegy; hexameter; prose
charismatic, sociological type of the, 129 hymnic, 40–1, 49
Cicero, 183 of Didactic poetry, 102–76
Cleanthes, 19 of theogony, 35, 117, See also Hesiod, Theogony
colonisation, 52, 135–8 gods. See also language, of the gods
Cratylus, 198, See also Plato, Cratylus hard for mortals to understand, 14
hard to look upon, 86
defamiliarisation, 99–100, 174 in Empedocles, 131–2, 170–4
Democritus, 71, 77, 192–8, 200 in Xenophanes, 34–41
Derveni papyrus, 12–14, 152, 185, 198, 199 Gorgias, 3, 6, 72, 188–92
Dionysius of Halicarnassus, 79
Dionysus Zagreus, myth of, 152 Heidegger, Martin, 15–16
Hekate, 85
ecphrasis, 22, 86–7, 122 Helms, Mary, 49
Elea, 56 henotheism, 38
and Xenophanes, 51–2 Hephaestus, 86, 184
elegy, 46–63 Hera, 12, 42, 86, 95, 170–1, 184
historical, 26, 46 Heraclitus, 7, 198
narrator in, 47, 55, 61–3 and Empedocles, 108–9
sympotic, 46, 56–61 Herodian, 43
Eleusis. See Initiation, mystic Herodotus, 43, 77
236
Index 237
Hesiod, 23, 36, 41, 106, 108 myth, 55, 60, 63
Catalogue of Women, 137, 158 and reason, 24–5, 45–6
criticised by Xenophanes, 20 in Empedocles, 157–62
Precepts of Cheiron, 128 parenetic, 11
Theogony, 14–15, 19, 30, 32, 35, 36, 37, 60, 72,
119, 133, 158, 160, See also Muses in Hesiod narrative, 40, 69, 80–2, See also plot, didactic
Works and Days, 10, 16, 32, 60, 83, 133, 135, 139, and early Greek philosophy, 126
143, 144, 161 in Empedocles, 126–56
hexameter, 27–46, 64 narrator
perspective of narrator in, 29, 47, 61–3 fictional in didactic texts, 128–9
Hippias of Elis, 3, 67 Near Eastern literature, 5, 80–2
Homer, 36, 41, 108, 125, 198, 199
criticised by Xenophanes. See Hesiod, orality, 9, 177
criticised by Xenophanes Orpheus, 13, 50, 152, 198
Hymn to Apollo, 39, 200 Orphism, 84, 104, 107, 142, 147
Hymn to Hermes, 42, 72
Iliad, 17, 38, 86, 99, 135, 177, 184 paraphrase, heresy of, 6
Odyssey, 1–2, 17, 32, 33, 39, 62, 66, 72, 77, 94–7, paretymology, 119
135, 177, See also Sirens Parmenides, 43, 101, 102–76, 183
and Democritus, 192, 194, 195–6
initiation, mystic, 90–1, 142–7, 151, 152–6, and fifth-century allegorical interpretation,
166–7, 179 185–8
intertextuality, 4–5, 6, 7, 180 and fifth-century theories of language, 198,
in Empedocles, 110–13 199
in Parmenides, 94–7 and Gorgias, 190
in Xenophanes, 32–3, 35–6 and Platonic mimesis, 200, 203
intratextuality, 27, 93–4 reception by Empedocles, 109, 124–5, 167
use of verse criticised by scholars, 65
katabasis, 76–7, 80–2, 85, 94–5, 98 performance, context of, 8, 27, 63, See also
katharsis, 163, See also Aristotle, Poetics rhapsode
for Empedocles, 106–7
language for Parmenides, 67
and atomism, 193–4 Persephone, 85, 152–3
fifth-century theories of, 196–9 persona, narratorial, 47–9, 51, 106
mortal, 36, 173–4, 197 Pherecydes of Syros, 184
of the gods, 70, 173–4, 197 philosophy. See also poetry, philosophical
Parmenides on, 20–1, 69–70 functions of
Leucippus, 192, 193 methodology in history of, 4, 5–6
Lobon, 26 Pindar, 43, 97, 152–3
Lucretius, 7, 87, 106, 149–50, 181 Plato, 4, 51, 106, 125, 179, 189
and mimesis, 199–206
Manilius, 181 Cratylus, 196, 198, 199
metaphor eschatological myths of, 84
conceptual, 17, 66, 177 Ion, 18
in Xenophanes, 36–7 Parmenides, 67
Metrodorus of Lampsacus, 12, 185, 186 Phaedrus, 89
mimesis, 76, 78, 101, 126, 175, 181, 183, 199–206 Republic, 44, 82, 152, 175, 201–6
Mimnermus, 49, 50 Theaetetus, 92
Mourelatos, Alexander, 66, 93, 95, 96 Timaeus, 185
Muses, 11, 29, 30, 103, 113, 180 treatment of Presocratics, 3
different functions in epic and elegy, 47 plot, didactic, 127, 142, 152
in Empedocles, 115, 118–19, 154–6, 158 Plutarch, 2, 33, 117, 130, 183
in Hesiod, 14, 74, 75, 99 reception of Parmenides and Empedocles
mysticism in interpretations of the Presocratics, by, 92
8, 66–7, 106–7 poeticity, 7, 179
238 Index
poetics. See also Aristotle, Poetics Solon, 25, 47, 50, 52, 58, 71, 72, 130, 131
history of Greek, 20–2, 24–5, 101, 125–6, 182, song, 8
183–206 Sophocles, 77, 136, 140, 164
immanent, 21 Stesichorus, 43
poetry sublime, the, 18–19, 87–90, 166, 178, 181–2
applicability of the term to early Greek texts, symbolism, 9–15, 39, 41, 93–100, 170–5, 178
8–9, 23 symposium, 25, 46, 56–61, See also elegy,
Classical conception of, 22, 125–6, 176, 180, sympotic
183–206
distinct from verse, 6–7 Thales, 130
philosophical functions of, 15, 63–4, 178–9 Theagenes of Rhegium, 12, 24, 184
Poseidon, 37, 184 Theognis, 25, 58, 62
priamel, 53–4 Theophrastus, 189, 197
Proclus, 196 Timon of Phlius, 26
Prodicus, 187, 198 titles for early Greek texts, 27, 43
prophecy, 14, 98–9, 109 tragedy, 72, 140–1, 162–4
prose, 6–7, 41, 82, 177 truth
Protagoras, 198, 199 in early Greek poetry, 9–20, 64, 72–3, 114–16
purification, 2, 50, 141–2 perspectival theory of literary, 16–18, 82–3, 179
in Empedocles, 134–8, 152–6, 159 Tyrtaeus, 49, 50, 54–6
Pythagoras, 129, 142
Pythagoreanism, 105, 107, 142, 147 Virgil, 106, 181
vividness, literary, 9, 17, 76–7, 78–80, 166–70
quarrel between philosophy and poetry, 3–4, 24
wandering, as a basis for authority, 49–52, 61, See
reception, literary, 4–5 also exile and Empedocles
rhapsode, 27–8, 67
Xenophanes, 24–64, 102–76, 183, 184, 194
sacrifice, 51, 103, 137, 138–9, 140–2, 150–1, 152, 154, and Empedocles, 108–9
160, 164, 180 and Platonic mimesis, 200, 205–6
Salmoxis, 76–7 as rhapsode, 27–8
Sappho, 54 epistemology of, 30–2
Semonides, 49, 50 Peri Phuseōs, 27
Sextus Empiricus, 35, 42, 97, 191 Silloi, 26, 28
Simonides, 47
Sirens, 73–6 Zeus, 12, 13, 37–40, 42, 170–1

Poetryandpoeticsinthepresocratic Philosophers

Uploaded by

Copyright:

Available Formats

Poetryandpoeticsinthepresocratic Philosophers

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Poetryandpoeticsinthepresocratic Philosophers

Uploaded by

Copyright:

Available Formats

POETRY AND P O E T I C S I N T HE P RE S O C RA T I C

Of the Presocratic thinkers traditionally credited with the foundation

tom mackenzie is a Research Fellow in the Department of Greek

Cambridge University Press is part of the University of Cambridge.

Abbreviations of ancient authors and works follow those of OCD4 or, in

BNJ I. Worthington (ed.) (2007–). Brill’s New Jacoby. Leiden.

I generally refer to fragments and testimonia of the Presocratics and

1 Aims, Methods and Assumptions

Iamblichus’ wonderful story (probably deriving from the c. 100 ce

2 Truth, Symbolism and Sublime Perspectives in Early Greek Song

2 Xenophanes and the Hexameter Genre

a Epistemology and the Immersive World of the Text

b Xenophanes’ Theology and the Language of Hexameter Poetry

οὖλος ὁρᾶι, οὖλος δὲ νοεῖ, οὖλος δέ τ’ ἀκούει.

ἀλλ’ ἀπάνευθε πόνοιο νόου φρενὶ πάντα κραδαίνει

αἰεὶ δ’ ἐν ταὐτῶι μίμνει κινούμενος οὐδέν

D18=B25 reworks the earlier conception of Zeus’s ability to deter-

c The Attacks on Homer and Hesiod

πάντα θεοῖσ’ ἀνέθηκαν Ὅμηρός θ’ Ἡσίοδός τε,

ὡς πλεῖστ(α) ἐφθέγξαντο θεῶν ἀθεμίστια ἔργα,

a Xenophanes as a ‘Wandering Poet’

ἀλλ’ εἰ μὲν ταχυτῆτι ποδῶν νίκην τις ἄροιτο

4 Xenophanean Elegies and Xenophanean Hexameters

2 Poetry and Parmenides’ Philosophy

3 Poetry and Deception

a Deception and the Doxa

c Deception and the Proem

4 Profound Visions and Sublime Emotions

a Seeing the Unseeable

c Perspective and Character

e Parmenides and the Sublime

g The Reception of Parmenides’ Poem

5 Poetry and Symbolism

c Allegory and Subtext

οὐδέ ποτ’ ἦν οὐδ’ ἔσται, ἐπεὶ νῦν ἔστιν ὁμοῦ πᾶν

Parmenides’ adaptation of the formula illustrates the contrast between the

2 Religion and Philosophy, One Poem or Two

a Empedocles and His Predecessors on Poetry

As commentators have noted, the allusion can be read as a ‘correction’ of

b Love and Poetry in Empedocles’ Cosmology

c Love’s Craft and Empedocles’ Poetry

a The Narrative of the Daimon

b Initiation, D257=B110 and the Education of the Singular Addressee

But, gods, turn away their madness from my tongue

5 The Effects of Empedocles’ Narratives

a Empedocles and Myth

b Emotions and Aesthetics

These lines have sometimes been taken as an appeal to empirical evidence:

d Interpreting the Text, Understanding the World

The four names conventionally refer to anthropomorphic deities and the

οἱ δ’ ὅτε μὲν κατὰ φῶτα μιγέντ’ εἰς αἰθέρ’ ἵ<κωνται>

In D53=B8 mortals typically use φύσις, ‘birth’ or ‘creation’ for what is in

6 Conclusions: Empedocles in the History of Greek Poetry

My central aim has been to enrich our literary appreciation of these

2 Anaxagoras, Allegory and Ainigmata

5 Democritus and the Sophists on Language

6 Plato and Mimesis

Aeschylus, 136 emotion and literature, 9, 178, 179

You might also like