TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

Plato's Alleged Epitaph 

Leonardo Taran 

I N HIS BIOGRAPHY of Plato, which occupies Book III of his lives of 
ancient philosophers, Diogenes Laertius has preserved three epi­
grams inscribed, he says, on the philosopher's tomb. To these he 

adds two which are his own work.! These last, however, are not 
relevant for the purpose of this paper, which is to determine whether 
or not any of the extant epitaphs is likely to be the original inscrip­
tion carved on Plato's tomb. For the same reason it is not necessary 
to discuss the third epigram; it is an obvious example of a purely 
literary epitaph, and even Diogenes himself says that it is later than 
the first two. It will be useful for the discussion that follows to tran­
scribe the first two epigrams and the context in which they occur. 

E1T'E"ypaCPTJ B' av'Tov 'TqJ 'TacfxtJ E'TT/:ypaJL/-W-'Ta 'TaBe 1T'PW'TOV 
O"wcppoa-UVf} 1T'pOCPEPWV lJVTJ'TWV .;)lJEt 'T€ BtKaicp 

EvlJaB€ B7] K€t'Tat lJ€to<; 'APtO"'TOKAE'Y1<;" 
., ~ , " ,.1.,." ~ ." Et u€ 'Ttl) €K 1T'av'TWV O"O'f'I.TJI) JLeyav €O"X€V €1T'CUVOV 

'TOV'TOV EXEt 1T'A€tO"'TOV Kat cP(JOVOI) OVX E1T'€'Tat . 
., ~, 

€'T€POV u€ 

yaw JLEV EV KOA1T'~ KPW'TEt 'ToB€ O"W/-W- TIAa'TWVOI), 
I/IVX7] B' cilJava'Tov 'TaeW EXEt /-W-KapWV 

viov 'Apln'TWVOI), 'TOV 'Ttl) Kat 'TTJAOlJl. vaLwV 
'TtM aV7]p ayalJOI) lJiiov iBov'Ta {3Wv. 

Kat aUo V€W'T€PO v 
aiHE, K'TA. 

7 yaw . .. IIX.aTwvo~ over erasure in Plan.a KOX.7rW Diog. Laert.: 
KOX.7rOL~ Anth.Pal. and Plan.a 

8 &8avaTov Diog. Laert. Band A nth. Pal. : &8avaTwv Diog. Laert. 
F P and Plan. a 

These two epigrams, and the third as well, have also been pre­
served in the Palatine Anthology2 in the same order and with minor 

1 Diog. Laert. 3.43-45. The edition cited is that of H. S. Long (Oxford 1964). The 
MSS. referred to below are: B = Neapolitanus Burbonicus iii B 29 (XII cent.); F = 
Laurentianus 69.13 (XIII cent.); P = Parisinus gr. 1759 (early XIII cent.). 

27.60-62. The MS., which once belonged to the Palatine Elector, is now divided 
between Heidelberg and Paris (Palatinus 23 + Parisinus Suppl. gr. 384). On the history 

63 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

64 PLATO'S ALLEGED EPITAPH 

variant readings.3 Moreover, the first epigram is there ascribed to 
Simias. The three appear also in Plan.a, where, however, the first 
epigram is anonymous (as it is also in Diogenes and in the Arabic life 
of Plato, about which see infra); and in Planudes it occurs earlier 
than the second and the third epigrams.4 The two latter ones appear 
together and in the same order as in Diogenes and in A nth Pal. On 
the other hand, Plan.b (3.26.7) has preserved a two-line epitaph, the 
same, but with some significant variants (discussed irifra), as lines 
1-2 of the second epigram in Diogenes Laertius, and with ascription 
to Speusippus. (Undoubtedly the Speusippus meant is Plato's nephew 
and his successor as head of the Academy, whose approximate dates 
are 410-339 B.C.)5 This two-line epitaph has also been preserved in E 
(= Sylloge Euphemiana) 56 but without any ascription at all. 

I.7TEVU{:tr7TOV 

uwp.a J-LEv EV KOA7To/'~ Ka'TEXEt 'T08E yata IIAa'Twvo~ 
.pVX7J 8' in08EOV 'Ta~,v EXEt p.aKapwv. 

1 I.1TEVUL1MTOV Plan. b: om. E 2 KOA1TOLt;' Plan. b: KOAmp E 
3 WOOEOV E: WOOEWV (from WOOEOV) Plan.b : WooEO'> 8ergk 

A majority of the scholars who have concerned themselves with 
this question have held that the distich preserved in Plan.b was writ­
ten by Speusippus and that it is in all probability the very epitaph 
inscribed on Plato's tomb. The second epigram in Diogenes Laertius 
would in that case be merely an expansion, with modification in the 

of this MS. and of the Anth.Pal. see A. S. F. Gow and D. L. PAGE, The Greek Anthol­
ogy, Hellenistic Epigrams I (Cambridge 1965) xxxiii-xxxviii, and A. D. E. CAMERON, 
The Greek Anthology: From Meleager to Planudes (forthcoming, the Clarendon Press); 
these works are cited hereafter by authors' names alone. I should like to thank Alan 
Cameron both for reading this paper and for permitting me to consult the typescript of 
his new book. I have referred a few times to this work, but because it is still in the 
press I have stated my points in full wherever I considered it appropriate. I cite the 
A nth. Pal. from H. Beckby, Anthologia Graeca 2 • 

a A glance at the cri.tical apparatus will suffice to show that the variants for the sec­
ond poem in Diog. Laert. (the only ones that are relevant to the purpose of this paper) 
are of minor importance and are such as often occur in MSS. of one and the same 
author. On the readings for the first and the third epigrams see Beckby II 46, 48. They 
may all be explained by the hypothesis that the epigrams in Cephalas (whose own 
anthology was excerpted by the scribes of the A nth. Pal. ) which ultimately come from 
Diog. Laert. were derived from a MS. different from the exemplar of the extant MSS. of 
Diogenes. Cf P. Waltz, Anthologie grecque 2 IV (Paris 1960) 15-16. 

4 The references are Plan.- 3.1.l, 28.2, 28.3. On the two editions of Planudes' anthol­
ogy see 73f infra. 

5 For the dates see my Speusippus of Athens, A Critical Study (Philosophia Antiqua 39 
[198I)) 7. There I briefly discussed and dismissed Planudes' ascription of the distich to 
Speusippus, but the whole question requires a detailed analysis of the sources, which I 
offer here. The Speusippean texts are cited from my edition. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO T ARAN 65 

first two lines, of Speusippus' original distich. Among those who 
subscribe to this interpretation one may mention F. Osann, G. Kai­
bel, T. Preger, T. Bergk, H. Stadtmiiller, P. Lang, 1. Geffcken, W. 
Peek, and M. Isnardi Parente.6 These critics, however, have accepted 
Planudes' version as the original Plato epitaph and his ascription of it 
to Speusippus without offering any detailed discussion of the issue. 
To the best of my knowledge the only scholar who has treated the 
whole question at length is 1. A. Notopoulos, in a paper published 
some forty years ago.7 His conclusion is that Planudes' two-line 
epitaph was written by Speusippus and was the actual epigram in­
scribed on Plato's tomb. D. L. Page, for his part, in his posthumously 
published Further Greek Epigrams,8 maintains that Planudes' ascription 
to Speusippus, though possible, is insufficient. Moreover, he believes 
that the related four-line epigram in Diogenes Laertius is a later 
remodelling and expansion of the distich preserved in E and in Plan. b 

However, as will be argued below, Page's discussion is indecisive; 
he does not take into account the evidence of the Arabic life of Plato 
by Ibn al-~if!i, nor does he seem to have been acquainted with 
Notopoulos' paper.9 Since my interpretation radically differs from that 
of Notopoulos, I shall here discuss at some length the arguments he 
has advanced. At the same time I shall offer probable proof that none 
of the extant epigrams is likely to be the original epitaph inscribed on 
Plato's tomb (if there was one) and that Planudes' ascription of the 
two-line epigram to Speusippus is probably merely a conjecture of his 
or of his source, but that in any case it cannot be traced back to 
classical antiquity, let alone to the fourth century B.C. Finally, I shall 
discuss the question whether the two-line epitaph or that in four lines 
is likely to be the earlier one, as well as Page's arguments in favor of 
the priority of the Planudean distich with its peculiar readings. 

The internal grounds on which Notopoulos bases his ascription to 
Speusippus of the distich Planudes has preserved can' be quickly 

6 F. Osann, Beitrage zur griechischen und romischen Litferarurgeschichte (Darmstadt 
1835) 307ff; G. Kaibel on Epigrammata Graeca 56 (p.19); T. Preger, Inscriptianes Grae­
cae Metricae (Leipzig 189]) 9-]]; T. Bergk, Paetae Lyrici Graeci4 II (Leipzig 1882) 
329f; H. Stadtmiiller, Anthalagia Graeca Epigrammatum 11.1 (Leipzig 1899) 44; P. Lang, 
De Speusippi Academic; Scriptis (Bonn 191 J) 86; 1. Geffcken, Griechische Epigramme 
(Heidelberg 1916) 52 ad no. 141b; W. Peek, Gr. Vers-Inschr. 1756; M. Isnardi Parente, 
SpeusiPPD, Frammenti (Naples 1980) 123, 389-90. 

7 "Plato's Epitaph," AlP 63 (1942) 272-93 (hereafter 'Notopoulos'). 
8 Further Greek Epigrams, edd. R. D. Dawe and J. Diggle (Cambridge 1981: hereafter 

'Page') 305-07. 
9 This may be due to the author's death before he was able to complete his work. 

The editors (p. vii) transcribe a pencilled note by Page: "Ready for the Press, except 
that it would be better for a critical eye." 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

66 PLATO'S ALLEGED EPITAPH 

dismissed. Such arguments are seldom decisive in cases of positive 
ascription, since one must always take into account the possibility 
that a given poem may be the work of a skilful imitator. But in the 
present instance the question does not really arise, as only one epi­
gram, of a completely different nature, is known to have been written 
by Speusippus. lO One will readily concede that there is nothing in the 
Wording, motifs, and contents of the distich that is incompatible with 
a date of composition in the fourth century B. c. But neither does 
anything in it make it necessary or likely for that epigram to have 
been written then rather than later. Moreover, even if written in the 
fourth century, nothing in the poem requires us to suppose that it 
was written by a philosopher, let alone by Speusippus.ll 

The issue, then, must be decided on the basis of the external 
evidence alone. In this connection Notopoulos' main argument is as 
follows. In Plan.b 3.26.7 and in the Arabic life of Plato by al-~if!i 
(A.D. 1172-1248)12 one finds an epitaph in one distich only. (This 
Arabic life of Plato goes ultimately back, in part at least, to Por­
phyry's life of Plato and to another Greek source, or sources, different 
from Diogenes Laertius, who does not appear to have been known to 
the Arabs.)13 On the other hand, Diogenes, the Palatine Anthology, 

10 Notopoulos fails to refer to the only extant epigram almost certainly written by Speu­
sippus. This is found in the Aeademieorum Philosophorum Index Hereulanensis, where it is 
cited on the authority of Philochorus (Speusippus T2.3-II, F86): rcH(T&] O[mwd (JEa<; 
Xapml<; M[ov](Tat<; clVIH(J7)]KEV / :E1TEVCTL1r7T[O<;] Ao[yi]wv [Ei'v]EKa owpa TtAWV. 

II The contrast between the destiny of the body and that of the soul after death is a 
commonplace even in inscriptional epitaphs (ef. nn.46-47 infra), and the notion that 
the place of the soul is in the immortal rank of the blessed, though held by Plato, was 
not peculiar to him or to philosophers in general. In fact it is in lines 3-4 of the related 
four-line epitaph in Diog. Laert. that we find implied some knowledge of Plato's 
thought; but even apart from the fact that it is the Planudean distich, not the final two 
lines of the epigram in Diogenes, that is ascribed to Speusippus by Notopoulos and 
others, nothing in the lines in question requires us to ascribe them to Speusippus or 
even to a philosopher. 

12 Notopoulos did not himself use T. Roeper, Leefiones Abulpharagianae Alferae: De 
Honaini Vila PlalOnis (Gdansk 1866), but relied on Preger's summary of this work (supra 
n.6: 10-11). Both Preger and Notopoulos follow Roeper in ascribing the Arabic life of 
Plato to Hunain ibn Ishaq (IX cent.), and Notopoulos makes use of a translation of Bar 
Hebraeus' Chronicle, Bar Hebraeus himself having made use of the Arabic biography of 
Plato «(/ nn.44-45 infra). But Roeper's ascription of that life to Hunain is mistaken, ef. 
D. A. Khvol'son, Die Ssabier und der Ssabismus I (St Petersburg 1856) 787 n.2, and M. 
Steinschneider, AI-Farabi: Des arabisehen Philosophen Leben und Sehri/ien (Memoires St 
Petersburg VII.xiii.4 [I869]) 186 n.1. The biography in question is by al-I.<ifti, in whose 
Ta'rikh al-Hukama' it occurs. Roeper (9-22) gives an annotated translation of al-I.<ifF's 
life of Plato, both in the version given by M. Casiri, Bibilolheea Arabieo Hispana I 
(Madrid 1760) 301 f, and in a longer recension, as well as of Bar Hebraeus' statements 
on Plato. For the Arabic text see the edition of 1. Lippert (n.I7 infra). 

13 In his life of Plato al-I.<ifF cites Theon (probably Theon of Smyrna, II A.DJ. He is 
dependent also on another Greek source, probably Porphyry's life of Plato (ef. n.45 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO T ARAN 67 

and Plan.a 3.28.2 all give a four-line epitaph (the second in Diog­
enes). It is well known that for the classical epigrams the Palatine 
Anthology and Planudes ultimately depend on Cephalas' anthology,14 
and that Cephalas, who was 7TpwTo7Ta7Tas of the palace at Constan­
tinople in A.D. 917, and who composed his anthology ca 900, drew 
the epigrams now found in A nth. Pal. 7.60-62, 83-133, and others as 
well from a source which got them from Diogenes Laertius.15 Hence 
the ultimate source of the four-line epitaph with which we are here 
concerned, of its peculiar readings, and of its anonymous character is 
Diogenes Laertius himself.16 But in Plan.b 3.26.7 the two-line epitaph 
ascribed to Speusippus occurs independently of the group of epigrams 
on Plato's tomb found in Diogenes, in the A nth. Pal. , and in Plan.a 

Moreover the distich in Plan.b presents some readings different from 
those of the first two lines in our three sources for the four-line 
epitaph: a-{vJ.UX and yata have exchanged places; KaT€XEt appears 
instead of Kp tJ7TTEt, and inO(JEOV or icro(J€wv instead of a(JavaTov or 
a(JavaTwv. Notopoulos then concludes (274): 

Finally the epitaph [sc. in Plan.b] shows its independent source in 
the authorship. Plan. IIIb 26, 7 gives Speusippus as its author 
whereas all the other sources, including Plan. ilia 28, 2 give it as 
adespoton. The existence therefore of the second epitaph in the 
Planudean A nthology as adespoton, enlarged by a second distich, as 
part of the general group of Plato epitaphs also found in Diogenes 
Laertius and the Palatine Anthology, and the existence of this same 
epitaph under the name of Speusippus, independent and different 
in text, position, and size, show a source which ultimately con­
tained the Speusippusepitaph alone, before its inclusion with other 
Plato epitaphs. 

Now, to begin with, it is necessary to distinguish clearly three 
things which have become somewhat confused in Notopoulos' argu­
ment: (a) whether the two-line epitaph is likely to be earlier than the 
related four-line epigram; (b) whether the two-line epitaph is likely to 
be the epigram actually inscribed on Plato's tomb; (c) whether the 
two-line epitaph is likely to have been written by Speusippus. But 

infra). Porphyry's History of Philosophy, where Plato's biography occurred, was available 
to the Arabs in a Syriac translation (~'j: A. Muller, Die griechischen Phi/osophen in der 
arabischen Uberlieferung [Halle 1873] 5, 25, 30 n.l, 57 n.46), whereas they do not seem 
to have been acquainted with Diog. Laert. (MUller 42). Cf also n.35 infra. 

14 Cf Gow/Page xiv-xviii and Cameron's exhaustive study. See also n.21 infra. 
15 Cf R. Weisshaupl, Die Grabgedichte der griechischen Anrhologie (AbhArchEpigWien 

7 [1889]) 34-38; Stadt mUlier (supra n.6) LXIII-LXIV; Notopoulos 277-80. 
16 On the variants in these three sources see supra n.3. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

68 PLATO'S ALLEGED EPITAPH 

even if one could establish (a), that the two-line epitaph is earlier 
than the related one in four lines, this by itself would not be decisive 
regarding points (b) and (c): for how early need the two-line epitaph 
be? On the other hand, if one could establish (c), that the distich in 
question was written by Speusippus, then (a) must surely be an­
swered affirmatively and, with good probability, also (b). Finally, 
even if one could prove (a) and (b), that by itself would not suffice 
to establish that Speusippus was the author of the two-line epitaph. 
For that to be the case one would have to produce strong arguments 
in support of the· reliability of Planudes' ascription of the distich to 
Speusippus, since he is the only source for this ascription. As it is, 
however, it seems that Notopoulos has failed to prove, or to make a 
likely case for, any of these three points. 

It is therefore most important to discuss first of all the possible 
Speusippean authorship of the two-line epitaph. Notopoulos bases his 
attribution of this distich to Speusippus on three interrelated grounds: 
first, that in Plan.b the epigram is explicitly ascribed to Speusippus; 
second, that the transmission of the two-line epitaph with its ascrip­
tion to Speusippus and with its characteristic readings is independent 
of Diogenes Laertius (and of the sources which ultimately depend on 
him), and so it is also independent of the transmission of the related 
four-line epigram with its peculiar readings and its anonymous charac­
ter; third, that the two-line epitaph is independent of the transmis­
sion of the group of epigrams allegedly inscribed on Plato's tomb. 
The question arises, then, whether the ascription of the distich in 
Plan.b is likely to go back to Speusippus himself or, in any case, how 
early is it likely to be. However, it is regrettable that in his discussion 
Notopoulos kept a discreet silence concerning the evidence and the 
implications of two sources. For in the Arabic life of Plato by al-~ifF, 
and so presumably also in its Greek source (not later than the fourth 
century A.D.), which was not Diogenes Laertius, the two-line epitaph 
is transmitted as adespoton and is transmitted as part of the epigrams 
allegedly inscribed upon Plato's tomb. Since in the discussion that 
follows the evidence of the Arabic life will be of importance also in 
connection with several other points, I give a translation of the per­
tinent passage of Lippert's text of al-~if!i:17 

There is written on his [sc. Plato's] tomb in Roman [Rumi = 
Byzantine Greek] (this) of which the translation into Arabic is: 
"Here is the place of a man who was Aristocles the divine. He 
stood before men and above them in virtue and a character of 

171. Lippert, Ibn al-Qifti's Ta',ih al-Hukama' (Leipzig 1903) 24-25. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO T ARAN 

justice. Whoever praises wisdom (as) greater than all the other 
things praises this (man) very much because in him is the greatest 
[or mostl wisdom." 18 This is on one side of the tomb, and on the 
other side: "As for the earth, it covers the body of Plato; on the 
other hand, as for his soul,· it is in the rank of one who does not 
die."19 

69 

Moreover, in addition to the Arabic life of Plato, the two-line 
epitaph is also transmitted as adespoton in E. Now E is the Sylloge 
Euphemiana, a selection of epigrams from Cephalas' anthology made 
for a certain Euphemius "by a native of Hypata living in Constan­
tinople in the reign of Leo the Wise (886-911) and therefore a near 
contemporary of Cephalas. "20 It is a well established hypothesis that 
the Palatine Anthology, E, Plan.a, and Plan.b all go back ultimately to 
Cephalas' anthology,21 compiled ca A.D. 900. Hence the occurrence of 
the two-line epitaph in both E and Plan.b would probably mean that it 
was found in Cephalas' anthology (but see the caveat in the next 
paragraph). Since the distich is adespoton in E, a Sylloge that in gen­
eral is well equipped with ascriptions, it is surely illegitimate simply to 
infer, as Notopoulos implicitly does, that Planudes found the distich's 
ascription to Speusippus in Cephalas himself. And so, concerning this 
question, we would at best have had to be satisfied with a non liquet. 
Yet if we take into account all the facts it seems more probable than 
not that the distich's ascription to Speusippus is a conjecture of Pla­
nudes or of another author probably later than Cephalas, but that it 
was not found in the latter's anthology. In fact, as I shall argue, it is 
quite possible, in the light of the evidence, that the distich itself was 
not part of Cephalas' own anthology at all. 

To begin with, we must bear in mind three things. First, had the 
two-line epigram really been by Speusippus, this would mean that it 
was in all likelihood the very epitaph inscribed on Plato's tomb. 
Second, both in antiquity and in Byzantine times Plato was, together 
with Homer, the most famous of all classical authors. Third, Speu-

18 This first epigram is clearly the same as the first of the group of Plato epitaphs in 
Diog. Laert., even if some of the readings of al-~ifF's Greek source are different from 
those in Diogenes. 

19 The Arabic text was kindly translated for me by Professor David Pingree of Brown 
University (el also n.50 infra). 

20 Gow/Page xli. On the Sylloge Euphemiana ef. also Stadtmtiller (supra n.6) XXIX­
xxx. The preservation of the distich in Iriarte 105 = Matritensis XXIV is of no con­
sequence, since that Sylloge was made by Constantine Lascaris, who frequently took his 
epigrams from Planudes himself (el Stadtmtiller LXIV-LXVI). 

21 Cf supra n.14. Aubreton's attempt to deny this ("La tradition manuscrite des 
epigrammes de I' Anthologie grecque," REA 70 [I968] 32-80 is unconvincing and has 
been refuted by Cameron in his forthcoming book. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

70 PLATO'S ALLEGED EPITAPH 

sippus' relation to Plato was well known even in Byzantine times 
through Diogenes Laertius (c! nn.30-34 irifra) , Hesychius of Mile­
tus, and others. Is it therefore likely that, had the ascription of this 
distich to Speusippus been found in Cephalas, the scribes of the 
Palatine Anthology (composed ca A.D. 930-950)22 would have failed 
to preserve this epitaph altogether, and that E, which has preserved 
it, would have failed to transcribe its ascription to Speusippus? And 
there is also evidence-negative, but nevertheless as significant as 
the previous item - from- an additional source not hitherto cited in 
connection with the point at issue here. 

A number of marginal notes in the manuscript of the Palatine 
Anthology show that for books VI and VII the corrector C, a careful 
scholar, collated the Palatine MS. against the MS. of an anthology 
written by one Michael Chartophylax (a title which at this time des­
ignated the head of the Patriarchal Chancellery), 23 who was a contem­
porary of Cephalas. And yet C did not transcribe the distich Planudes 
ascribes to Speusippus. This probably means that the distich's ascrip­
tion to Speusippus was not found in Michael either. The importance 
of this is seen from a note C wrote on A nth. Pal. 7.432: EW<; ~Se 'Tn 
'TOV KVPOV MtXa1]A 'TOV J.W.Kapwv 7Tepte'ixoll E'Trt'YpaJ..LJ.W.'Ta a'Ttlla iSw­
Xeipw" mho" E'YpaljJell EK 'Tij" f3if3AOV 'TOV KecpaAa K'TA. For the impli­
cations to be drawn from this text are: (a) C had not a copy of, but 
Michael's own autograph; (b) Michael took his epigrams from the 
book of Cephalas; (c) the words E'YpaljJell EK 'Tij" f3if3AOV 'TOV KecpaAci 
most probably mean that Michael took his epigrams from Cephalas' 
own copy or from Cephalas' own autograph.24 For Cephalas was not 

22 This date for the composition of the MS. containing the Anth.Pal. is earlier by 50 
to 30 years than that generally accepted until a few years ago. But there are strong rea­
sons, both palaeographical and from the contents of the marginal notes, for thinking 
that the MS. is not much later than Cephalas. Cf A. Diller, "The Age of Some Early 
Greek Classical Manuscripts," Serta Turyniana (Urbana 1974) 514-24, esp. 520-21; J. 
Irigoin, Annuaire de I'licole pratique des hawes etudes 1975/6, 281-95, esp. 281-89; and 
ch. 5 ("The Palatine MS. and its Scribes") of Cameron's forthcoming book. 

23 Therefore an important person: most scholars take <5 XapTocpvAat to mean 'the 
archivist', but see Cameron. 

24 From several remarks by scribe C it is to be inferred that he had good reason to 
think that Michael had access to a complete copy of Cephalas' anthology; and, given 
C's note on 7.432, it follows that he thought Michael to have had access to Cephalas' 
own copy or to Cephalas' own autograph. Thus in his note on 7.428, C, after making 
corrections in the text of the Palatinus from Michael's copy, states that "even Mi­
chael's text contained errors." On 6.269 C states Ei~ TO aIJTt/30AtlJ OV KEtmt TOV KVPOV 
MtXa-r,AOV' 7rOOEIJ O~IJ E'Ypacfn/ OVK olOO: that is, assuming Michael to be a better wit­
ness to Cephalas than the A nth. Pal. , C infers that 6.269 was interpolated in the Palatine 
MS., since it is absent from Michael's autograph. Finally, in the Palatine MS. there are 
gaps after both 6.125 and 6.143, and C says, respectively, ov AEi7rEt, We; o'4rot, and ov 
AEi7rEt, We; oif.LUt, ovsE ElJmv(Ja. From this we may infer that there was no lemma after 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO TARA.N 71 

the 'author' of the Anthology except in the sense that he himself 
compiled the definitive Byzantine anthology of classical epigrams. If 
this is so, then it is possible that even the distich itself preserved by 
E and by Planudes was not to be found either in Michael or in Ceph­
alas himself. There is no reason to think that Planudes had access to 
a complete copy of Cephalas' anthology. Rather it is clear that neither 
of the two MSS. of the anthology he used was a complete copy of 
Cephalas, since Plan.a lacks most of the epigrams found in Plan.b And 
both Plan.a and Plan.b together lack epigrams found in the Palatine 
Anthology and in other sources, all of which go ultimately back to 
Cephalas' anthology.25 

The compiler of E, for his part, given the fact that the Sylloge 
Euphemiana must have been produced only a few years after Ceph­
alas' own compilation, may very well have had a complete copy of the 
latter's anthology. But apart from the fact that E has not preserved the 
distich's ascription to Speusippus, there are a few instances in which E 
has included epigrams which in all likelihood were not to be found in 
Cephalas.26 On the other hand, the scribes of the Palatine Anthology, 
given the date of the Palatine MS. and the marginal remarks of the 
corrector C, in all probability did have a complete copy of Cephalas, 
and the same thing is true also of Michael Chartophylax.27 Hence, it is 
possible that the distich itself, which only Planudes ascribes to Speu­
sippus, was written in the margin or in the body itself of a MS. of the 
anthology, but that it was not found in Cephalas himself. Moreover, 
even if the distich was included in Cephalas, then E must have got it 
from him, whereas Planudes did not have a complete copy of Cepha­
las. In that case, the fact that in E the distich is adespoton, as it was 
also in the Greek source of the Arabic life of Plato by al-J.(if!i, makes 
it appear likely that the ascription to Speusippus was added by Planu­
des himself or in any case by another author later than Cephalas. It is 
well known that Planudes was not averse to making conjectures of all 
sorts, including ascriptions.28 And it is not at all difficult to see why 

either of these poems in Michael's autograph. For Michael's copy as C's only exemplar 
for collation cf Gow/Page xxxv f and Cameron ch. 5. 

25 The fact that Planudes in Plan.b transcribed once more some of the epigrams 
already included in Plan.a shows that he was trying to make his copy of the anthology 
as complete as possible. (In the case of the repeated epigrams Plan.b contains additional 
information.) On the other hand, Planudes did not have access to the A nth. Pal. , cf 
Gow/Page xxxviii with n.3. 

26 Cf Cameron ch. II. 
27 Cf C's note on A nth. Pal. 7.432 (cited above), and supra n.24. 
28 0: Gow/Page xxxi-xxxii and xxxix. Concerning Planudes' ascription of the distich 

in question to Speusippus, Page says (306): "To say that the evidence for the ascription 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

72 PLATO'S ALLEGED EPITAPH 

he-or another Byzantine scholar-may have come to ascribe the 
distich in question to Speusippus: he knew the longer, four-line epi­
gram in comparison with which the distich appears to be the original 
'inscription' .29 Then mere acquaintance with Diogenes Laertius would 
suggest Speusippus as the most likely candidate for its authorship, if 
one was looking for one. For in Diogenes we find the following infor­
mation about Speusippus: he was Plato's nephew30 and his student,31 
he succeeded Plato as head of the AcademY,32 he was one of the 
executors of Plato's last Will,33 and-most important for the point at 
issue here-he wrote a "Funeral Banquet" on the death of the philos­
opher or an "Encomium of Plato" or both.34 

However that may be, even if Planudes found the distich's ascrip­
tion to Speusippus in Cephalas himself, it is highly improbable that 
this ascription goes back either to classical antiquity or even to Hel­
lenistic times. And it is also unlikely to go back to a good and reliable 
source. For it is hardly credible that such a source for the ascription 
was still available to Cephalas but was not known to Diogenes Laer­
tius and to the Greek source of al-J.GfWs life of Plato, both of which 
probably go back to an earlier biography of Plato.35 (In connection 

to Speusippus is insufficient is certainly not an overstatement of the case against it." To 
my mind the extant evidence justifies the conclusion that that ascription is with good 
probability a conjecture of Planudes or of a source later than Cephalas. Sometimes we 
find conjectural ascriptions also in the A nth. Pal. For example in 7.60 the first Plato 
epitaph is ascribed to "Simias." This ascription has rightly been rejected by most schol­
ars because that epitaph is anonymous not only in Diog. Laert. and in Plan.a but also in 
al-Jpf!i's life of Plato. The evidence of the first two sources indicates that the ascription 
to "Simias" was probably not found in Cephalas either. All in all Cephalas seems to 
have compiled his anthology with more mechanical methods and seems not to have 
been prone to making conjectures, including ascriptions. 

29 Note the absence of lines 3-4 in Diogenes' version and the occurrence of the 
antithesis uW/Ul ~II ••• !jJVXT, 8E, which is common in true sepulchral epitaphs from 
the fourth century B.C. onwards. Cj. n.47 irifra. 

30 Speusippus T l.2-4 (Diog. Laert. 4.0 and T4 (Diog. Laert. 3.4). 
31 Speusippus T5 (Diog. Laert. 3.46). 
32 Speusippus T l.4-5 (Diog. Laert. 4.0. 
33 Speusippus T37 (Diog. Laert. 3.43). 
34 Speusippus T l.58 (Diog. Laert. 4.5) and F la (Diog. Laert. 3.2); cj. Tanin (supra 

n.5) 228-35. . 
36 Since Diog. Laert. was not known to Arabic authors (cj. supra n.13) and since the 

ultimate Greek source of al-Jpftcs biography of Plato did not use Diog. Laert., the fact 
that both he and the Arabic life have preserved the first two Plato epitaphs (that the 
second epigram in the Arabic life is two lines long, whereas in Diogenes it has four 
lines, is not relevant to the point at issue here) shows that the two lives ultimately go 
back to an earlier biography of Plato which already contained the two epitaphs in the 
same order as they appear in the extant sources. (A comparison of the Arabic life with 
Diogenes' biography of Plato shows that the Greek source of the former was not Diog­
enes at all.) 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO T ARAN 73 

with this, it is noteworthy that al-~ifti:'s Greek source for the Plato 
epitaphs was either Porphyry or a source earlier than Porphyry.) 
Moreover, Notopoulos says nothing about this difficulty: if the distich 
is the text Speusippus had inscribed on Plato's tomb, how was the 
author's name transmitted? It was almost certainly not written on the 
stele, and the epigram itself does not contain its author's name as 
does the only epigram attested for Speusippus (see supra n.lO). Page 
(306) says that "it might nevertheless have been remembered that 
the author was Speusippus." But he can scarcely mean to say that the 
ascription to Speusippus had an oral transmission from 34817, the 
year of Plato's death, to ca A.D. 900, the approximate date of Ceph­
alas' anthology. And the evidence seems to show that Speusippus' 
works were not available even to Porphyry or Diogenes Laertius.36 

Hence Cephalas' likely source for the distich as preserved in E and in 
Plan.b could hardly have been anything but a biography of Plato, for 
example that by Hesychius of Miletus.37 Is it at all likely, however, 
that the distich's ascription to Speusippus in such a late biography of 
Plato could come from a good source which was not known to Diog­
enes Laertius, to the Greek source of the Arabic life of Plato, and 
even to the earlier biography of Plato which was the source of both? 
To me the obvious answer is in the negative. Moreover, as I have 
argued above, it is more likely than not that even if the distich was in 
Cephalas, the ascription to Speusippus was not. 

Nor is Notopoulos justified in his inference that the distich preserved 
by E and Plan.b is attested independently of the group of epigrams on 
Plato's tomb in Diogenes Laertius, in the A nth. Pal. , and in Plan. a It is 
well known that Planudes' autograph manuscript of the anthology is 
extant; it is Marcianus gr. 481 (colI. 863). He dated the end of the 
composition of Plan.a to September 1299, and he arranged the epi­
grams he preserves in seven books, now customarily referred to as la, 
lJa, etc. After writing out these seven books Plan'udes obtained another 
MS. in which he found epigrams absent from his source for Plan.a He 
wrote out these epigrams-as well as some already included in Plan.a -

as supplements to each of the first four books of Plan.a It is now cus­
tomary to refer to these supplementary books as Ib, lib, etc.38 In addi-

;J6 Cf Taran (supra noS) 233, 407-08. 
:J7 On Hesychius and his work, which was available to Suidas/the Souda (roughly 

contemporary with Cephalas), c:l Suidas s. v. 'HmJXLO~ MLA-r1aLo~ (II 594.15-25 Adler); 
H. Schulz, RE 8 (1913) l322-27 S.v. "Hesychios 10"; A. Adler, RE IVA (1931) 
706-09 s. v. "Suidas." 

.18 Cf Gow/Page xxxviii-xxxix, Cameron's forthcoming study, and Irigoin (supra 
n.22) 289-95. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

74 PLATO'S ALLEGED EPITAPH 

tion, merely from Planudes' separation of the first Plato epitaph On 
Diog. Laert. and in the A nth. Pal. ) from the second and the third it is 
clear that he introduced changes into the arrangement of epigrams in 
Cephalas. Hence it is impossible to know in what context Planudes 
found many of the epigrams that are in Plan.b, and so it is unwar­
ranted to infer, as Notopoulos does, that Plan.b 3.26.7 was trans­
mitted independently of the group of epigrams allegedly inscribed on 
Plato's tomb.39 Similarly, in the case of the Sylloge Euphemiana noth­
ing can be inferred as to the context in which its compiler found the 
two-line epitaph. On the other hand, it is noteworthy that in al-~if!i's 
life of Plato and hence also in its Greek source, which was not Diog­
enes Laertius, the two-line epitaph is transmitted together with, not 
independently of, another Plato epitaph, which is clearly the same as 
the first epitaph in Diogenes.4o In short, from the extant evidence 
there is no reason to believe that the transmission of the one-distich 
epitaph is independent of the group of epigrams allegedly inscribed 
on Plato's tomb. Moreover, it appears that the preservation of the 
two-line epitaph is not necessarily connected with its ascription to 
Speusippus found in Plan.b , since in E the epitaph is anonymous, as it 
is also in al-~if~l's life of Plato. 

It is highly improbable that either the two-line epigram or the 
related one in four lines is the original epigram inscribed on Plato's 
tomb. Nor is it likely that the ascription of either version to Speusip­
pus fell off from the source ultimately common to Diogenes Laertius 
and the Greek source of al-~ifF. Diogenes explicitly tells us that the 
epigram in 3.43 came first, i.e. before the four-line epitaph in 3.44, 
the epigram related to the distich that Plan.b ascribes to Speusippus. 
And Diogenes also states that the third epigram is "more recent" 
(lIEWrepoll). The same position relative to each other of the first two 
epigrams in Diogenes may be inferred for the Greek source of al­
KiftI. For the latter states that on one side of the tomb (i.e. of the 
~teie) there was one epigram (the same as the first epigram in Diog. 
3.43) and on the other side another epigram. This is the Arabic 
translation of a Greek distich roughly identical to the first two lines of 
the second epigram in Diogenes (more about this infra). And this 
means that in the Greek source of al-~ifti the two epigrams were 
given in the same order as in Diogenes. 

39 The distich in question may well have appeared in Cephalas or in a later copy of 
Cephalas as a marginal addition or in the text together with the other Plato epitaphs. 

40 0: supra nn.18 and 35. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO TARAN 75 

Now Notopoulos himself rightly thinks that the first epigram in 
Diogenes cannot be earlier than the Hellenistic age, for it presup­
poses the notion that Plato's original name was Aristocles. And 
Notopoulos himself has shown that Plato was the original name of 
the philosopher and that the legend about his name being Aristocles 
originated in the Hellenistic ageY Therefore there is no reason to 
think that the second epigram's ascription to Speusippus fell off from 
the Greek source ultimately common to Diogenes and al-I~jf!i. Nor is 
it likely that the second epitaph in Diogenes, whether in the two- or 
the four-line version, was the original epitaph inscribed on Plato's 
tomb. For had it been, it would be hard to explain how this epitaph 
came to occupy the second place-after an obviously Hellenistic epi­
gram-in the group of epitaphs allegedly inscribed on Plato's tomb. 
Surely it is impossible to take at face value the statement in Diogenes 
and in al-~ifti that the epigrams were inscribed on Plato's tomb, 
especially when it is seen that the first epitaph is a literary epigram 
written not earlier than the Hellenistic age. It is clear that only if 
Planudes' ascription of the distich to Speusippus were reliable would 
there be a probability that the epitaph on Plato's tomb (if there was 
one) had survived. But I trust it has been shown that Planudes' 
ascription, whether his own or not, is probably nothing but a late 
conjecture. To summarize: so far as our evidence goes, the conclu­
sion seems unavoidable that both the two-line epitaph and the related 
one in four lines were known in antiquity as anonymous epigrams 
and that there are no good grounds for thinking that either version 
was the original epitaph inscribed on Plato's tomb. 

There is an additional point against Notopoulos' interpretation which 
is also telling. The most important variant reading in the distich 
preserved in E and in Plan. b is that in line 1 fIWI-W and yaw are 
found in places different from those in which they appear in Diog. 
Laert. 3.44 and in the sources dependent on him. Thus one finds 
ya'La j.L€lJ ... fIWI-W in Diogenes but UWI-W j.L€lJ ... yma in E and in 
Plan.b But the Arabic life by al-~if!i, and hence also its Greek source 
for this epigram, has yaw j.L€lJ ... uWI-W, i.e. the same reading as in 
Diogenes. (Nothing can be inferred with certainty from the Arabic 
text concerning the other variant readings, though in line 2 the 
Arabic, as will be argued later, favors Diogenes' &8alJaTOlJ over the 

41 J. A. Notopoulos, "The Name of Plato," CP 34 (I 939) 135-45, and A. Riginos, 
Platonica. The Anecdotes Concerning the Lite and Writings of' Plato (Columbia Studies in 
the Classical Tradition 3 [Leiden 1976]) 35-38. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

76 PLATO'S ALLEGED EPITAPH 

in08EOV in E and in Plan.b) This shows, pace Notopoulos, that not 
even the transmission of the variant readings in E and in Plan.b need 
be connected with the alleged independent transmission of the distich 
as it appears in these two anthologies, or even in Cephalas. The likely 
inference is that the variant uWf.UX JLEv . .. yata is merely the con­
jecture of a late or, more probably, of a Byzantine scholar prompted 
either by the desire to emphasize the contrast between the destiny of 
Plato's body and that of his soul after death and/or by acquaintance 
with some genuinely sepulchral epitaphs where one finds UWf.UX JLEV 
... / tjJVXT, 8' (see in/ra). 

In view of the preceding discussion, the fact that the first epigram 
in Diogenes and in the Arabic life is Hellenistic at the earliest sug­
gests that so too is the second. What with good probability can be 
ascribed to the Greek source ultimately common to these two au­
thors is the preservation of two of the epigrams allegedly inscribed on 
Plato's tomb, written during Hellenistic times. It is well known that it 
had then become a topos to write literary epigrams on the tombs of 
famous men.42 The second epigram in either its four-line or its two­
line version is neither a variation nor an imitation of the first epigram 
found in both Diogenes and al-~ifti. Whichever of the two versions 
is the original or earlier, one must infer that it is a purely literary 
exercise in which its author tried to reproduce the simplicity and 
almost formulaic character of the true sepulchral epitaph. The main 
idea of the distich and of the first two lines of the related epigram in 
Diogenes is the contrast between the fate of Plato's body and that of 
his soul after death, a motif well attested in true sepulchral epitaphs 
from the fourth century B.C. on. 

In what precedes I have discussed the issue on the assumption that 
the Arabic life of Plato by al-~ifti gives a translation of the second 
epitaph as it appeared in its Greek source. There are good reasons for 
doing so even despite the fact that we do not have the original ver­
sion of al-~ifti's Ta'rikh al-Hukama', the work in which Plato's biog­
raphy appears, but an epitome of it by al-Zawzani written in 1249,43 
al-~if!i himself having died in 1248. For one thing, as Professor 
Franz Rosenthal of Yale University tells me in a private communica­
tion, even if al-Zawzani really abridged al-~if!i's original, he probably 
did not shorten, nor modify, Plato's biography. One cannot assume 
that he did abbreviate, unless good grounds are adduced in favor of 

42 Book VII of the Greek Anthology (i.e. the A nth. Pal. supplemented by Plan.a, 
Plan.b, and the Syffogae Minores) provides sufficient evidence of this. 

43 c.r. Lippert (supra n.17) 11-13, 16-17; A. Dietrich, Encyclopedia of [slam 2 3 
(971) 840 s.v. "Ibn al-~if!i." 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO T ARAN 77 

such a hypothesis. Secondly, as 1. Lippert, the editor of al-~ifti, 
says,44 Bar Hebraeus (died A.D. 1286) seems to have made use of a 
complete text of al-~ifF's work, and it is noteworthy that the two 
Plato epitaphs as they appear in Bar Hebraeus' Chronicle are identical 
to the version found in al-~ifti.45 Nor is there any reason to think 
that al-J.<if!i himself would have abbreviated the second epitaph to 
two lines if his Greek source had had the four lines found in Diog­
enes Laertius. Nevertheless, since there is a remote possibility that 
he may have done so, it is well to consider the consequences this 
would have for both Notopoulos' arguments and my own. 

The strongest argument that scholars from Osann and Preger to 
Notopoulos have advanced in support of the independent existence in 
antiquity of a two-line epitaph related to the first two lines of the 
second epitaph in Diogenes is the occurrence of that distich in the 
Arabic life of Plato. But if in the Greek source of al-~ifti's biography 
the second epigram also was in four lines, then the epitaph preserved 
by E and by Plan.b would at best find its earliest attestation in Ceph­
alas' anthology composed ca 900. Hence, there would be no evidence 
of the distich's independent existence before that date. Moreover, its 
ascription to Speusippus would still rest entirely on the authority of 
Planudes, since in E the epitaph is adespoton. And the fact that the 
epitaph was not included in the Palatine Anthology either by the 
scribes of the Palatine MS. or by the corrector C would still create a 
strong presumption against the possibility that the ascription to Speu­
sippus was to be found in Michael Chartophylax and hence also in 
Cephalas. However, even if the epitaph and its ascription to Speusip­
pus had been found in Cephalas, that epitaph could scarcely be re­
garded as anything but the work of a late or Byzantine scholar who 
shortened and modified the first two lines of the epigram in Diogenes 
or in a similar source both in order to differentiate it from the second 
epitaph in the group of the Plato epitaphs and in order to make it 
appear more genuinely inscriptional. And its ascription to Speusippus 
would have been prompted by the same reasons as those given 
above. On the other hand, suppose that the epitaph in al-~if!i's 

44 Lippert (supra n.17) 17. 
45 For an English translation of the epitaphs in Bar Hebraeus cf E. A. Wallis Budge, 

The Chronography of Gregory AbCI'! Fara) I (Oxford/London 1932) 36. An English 
translation (due to J. Obermann) of the epigrams in Bar Hebraeus from the edition of 
Bedjan (Paris 1890) is found in Notopoulos, "Porphyry's Life of Plato," CP 35 (I 940) 
284-93, esp. 286. (In this paper Notopoulos, following Roeper [supra n.12], tries to 
ascribe to Porphyry's life of Plato the passage in Bar Hebraeus' Chronicle which was 
itself based on al-~if~U For a Latin translation of the epitaphs as they appear in al-~fF 
and Bar Hebraeus see Roeper 13. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

78 PLATO'S ALLEGED EPITAPH 

Greek source was originally of the same length as the second epitaph 
in Diogenes. Since its readings insofar as they can be inferred from 
the Arabic are basically the same as those in Diogenes, one would 
have to infer that it was the only version known in antiquity, that it 
was anonymous, and that it had four lines. In other words, neither 
the ascription to Speusippus of the distich in Plan.b nor the indepen­
dent existence of this distich (as preserved in E and in Plan.b) in 
antiquity would be more probable than it was on the more reasonable 
hypothesis that in the Greek source of al-~ifti the second Plato epi­
taph was only two lines long. 

Now D. L. Page (305-07), who does not take into account the 
Arabic life of Plato, argues on purely internal grounds that the distich 
preserved by E and by Plan. b is the original version of this epigram 
and that the related four-line epigram in Diogenes Laertius is later. 
The latter would be a variation with subsequent expansion of the 
former epitaph. Page's arguments are: (a) The second distich of the 
four-line epitaph is obscure and ill-phrased, and looks like something 
tacked on. (b) In line 1 the antithesis demands a-wJ.Ul JJkv ... I/Jv)(T, 
Be, which we find in E and in Plan.b46 (c) In the second line, the 
author of the two-line epitaph put a very bad epithet before T(igLV 
(i.e. Ur08EOV or Uro8ewv). And so one must infer that the &8avaTov 
(or &8avaTwv) in the four-line epigram was a deliberate change, 
which fact points to the priority of the distich in E and in Plan.b 

Page's second and third arguments are both based on a tacit as­
sumption, similar to the principle of the tectio difficilior in textual 
criticism. It is therefore necessary to keep in mind that tectio difficilior 
praeferenda is not of universal application, or any mistake or absurd­
ity would be a tectio difficilior. And similarly here. Take for example 
the problem of line 1. a-WJ.Ul JJkV ... I/Jv)(T, Be is apparently the more 
obvious antithesis, and it is a common antithesis found in sepulchral 
epitaphs of this type from the fourth century B.C. on.47 But in none of 
the attested examples do we find anything like the difficulty Page 

46 For the conventional beginning Page cites Peek, Gr. Vers-Inschr. 1781, [(Twl~ ~v 
fV KOA1TOL<; [xlaSE 'Ya'L' 'Ia[TploKAEia<; / TT,V s[El ci:PErT,V KTA. (Athens, IV B.C.), and 
1782, (TW~ (TOV fV KOA1TOL", KaAAwTm, 'Yaw KaAlnrrEL, / (TTi .. SE <ci:>PETTi .. KTA. 
(Peiraeus, mid IV B.C.). However, the epigraphical evidence is richer and more varied 
than Page supposed, cf. infra. 

47 Cf. Peek, Gr. Vers-Inschr. 1889.5-6, (TW~ ~v fv(}aSE (TOV, aWVV(TLE, yaw KaAV7r­
TEL / .pvXT,V SE ci:(}avaTov KOLVO<; EXEL Ta~<; (Athens, second half IV B.C.); 1758.1-2, 
(TW~ ~v TiSE X(}wv K[ahEXEL, NLKa<; SE KEKAT/~L, / .pVXT, s[El f'Y ILEAEWV KTA.. (Ery­
threae, III B.c.L 1766.1-2; 1773.1; 1774.1:3; 1776.1-3. In this connection, it is note­
worthy that Diogenes Laertius himself composed an epitaph on Solon which contains 
the common epigraphic formula (TW~ ~v ... .pVXT, SE (l.63 = A nth. Pal. 7.87). 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO TARAN 79 

himself recognizes in the distich E and Plan.b have preserved, viz. the 
word order uW/UX ,",Ev ... ToBe ydia TIAaTWVOI). Nevertheless, Page, 
taking for granted that the antithesis demands uW/UX ,",EV ... tjJVXT, BE 
(and saying nothing of the fact that the beginning yaw ,",Ev is also 
conventional),48 believes that the author of the original epitaph was 
therefore willing here to accept the juxtaposition of the words ToBe 
yata TIAaTWVOI). I submit, however, that it is at least as likely that the 
author of the original epitaph, knowing the standard formula O"w/UX 
J.LEV ... t/JVXT, BE, purposely wrote yaw J.LEV ... T()B€ O"wJ.La I1Achwvo<;, 
/ t/JVXT, Be, not only in order to avoid the awkward O"wJ.La J.LEV ... 
ToBe yaw llAaTwvo<;, but also to avoid reproducing a trite formula. 
Thereby, moreover, the antithesis does not at all lose force, as Page 
appears to think, because the J.LEV affects the whole of the first sen­
tence and subordinates it, as is often the case, to the BE clause as a 
whole: "Though the earth conceals in its bosom Plato's body, his 
soul is in the immortal ranks of the blessed" (or "in the rank of the 
blessed, which is immortal"). And it is perfectly intelligible that a 
Byzantine pedant-acquainted or not with the inscriptional formula 
UWJ.La J.LEV ... tjJVXT, Be-wished to modify the line in order to bring it 
into agreement with the common antithesis 'body/soul'. Similarly, in 
the case of the second line, it is just as probable that aOavaTov was 
changed to to"oOeov as vice versa. Moreover, it appears that to"oO€Ov is 
not "a very bad epithet," as Page calls it. For, as he himself says, it 
modifies J.LaKapWV TagtV~ but J.LaKapWv here need not mean, as Page 
believes, "of the gods." It probably means "of the blessed" and 
refers to all men who have attained perfect happiness in the after-life. 
And there is therefore no reason, pace Page, why iO"oOeov cannot 
refer to Plato's attainment of that rank. Hence the second line of the 
distich in E and Plan.b means, "but his soul is in the rank of the 
blessed, which is equal to that of the gods." (1 agree with Page that 
Bergk's iO"oO€o<; for iO"oOeo v is unlikely to be right and is scarcely 
anything but a mere lectio facilior.) No doubt aOavaTov here is more 
appropriate than iO"oO€OV~ but a Byzantine looking for a variant may 
have modified aOavaTov in order further to differentiate his epitaph 
either from the first two lines in the related four-line epigram or from 
a distich such as the Greek source of al-~if!I must have had. 

Concerning Page's first argument, it may be said at once that, even 
if valid, it would affect only the question of the original length of the 
epitaph (one or two distichs), not that of the readings of the first two 

48 For 'Yaw ~/l at the beginning of the hexameter cf Peek, Gr. Vers-Inschr. 1759 
(Athens, first half III B.c.L 1750 (Ancyra, I A.D.). 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

80 PLATO'S ALLEGED EPITAPH 

lines; for the distich in the Arabic life presupposes, as we have seen, 
a Greek text basically identical to that of lines 1-2 in the related four­
line epigram in Diogenes. In the latter Page objects to the expression 
8lioll WOll'Ta piall in line 4. He believes, rightly, that it means "be­
cause he [sc. Plato] saw that (man's) life has some god-like quality." 
But he contends that if this is what the author did mean, "he should 
have taken the trouble to choose a more suitable expression" (307). 
However, even if the objection were well taken, it would not consti­
tute a cogent ground for arguing that lines 3-4 are later than lines 
1-2. Surely it is at least as likely that the alleged inappropriateness of 
8EWlI wollTa piall and/or other problems in lines 3-4 could have 
caused someone to eliminate the second distich altogether. As it is, if 
one takes into account the context, the phrase 8EWlI i80llTa piall is 
neither vague nor inappropriate. The author is saying that a good 
man, even if he dwells afar, honors Plato because he has seen that 
man's life has a god-like quality. One must bear in mind that Plato 
himself, whose thought may be presumed to have been known to the 
author of the epigram, considered that the truly virtuous and hence 
wise man is happy and divine.49 Plato, then, having seen and taught 
what is divine in human life (i.e. virtue and wisdom), is honored by 
the good (i.e. the virtuous) man, for he too, following Plato's teach­
ing, will become a divine being. We must assume that the author of 
the epitaph was acquainted with Plato's doctrine and gives its full 
value to allr,p a'Ya8o~. Now in line 2 it has been said that after his 
death Plato achieved the immortal rank of the blessed, and it is 
noteworthy that Plato himself thought that happiness in this life 
(which is attained only through virtue) is the necessary condition for, 
and becomes perfect happiness after death. To my mind, it is the 
evidence of the Arabic life of Plato, which preserves the first two 
lines of the epitaph only, that suffices to leave open the possibility 
that lines 3-4 are a later addition. If they are, I should think that the 
only thing objectionable in lines 3-4, which is not mentioned by 
Page, is the emphatic position of vwv 'ApiuTWIlO~ at the beginning of 
3, for it comes immediately after what one feels should be a full stop: 
1/IvXr, 8' a8allaTOll Ta~tll EXEt IUlKapwlI. But it is not impossible that 
for this very reason the four-line epitaph was shortened to one dis-

49 For the relation between virtue, wisdom, and happiness in Plato cf. Leg. 631 B-
6320, 660E-664B, 7420-743c, 960B-968A, with L. Tanin, Academica: Plato. Philip of 
Opus. and the Pseudo-Platonic Epinomis (Philadelphia 1975) 54 with nn.235-37. And 
for Plato any perfectly good soul, whether embodied or not, is divine (cf. Tanin 35 
with n.159). Cf. also, in its context, Leg. 95184-5 and the well-known OJ-LOiwuL<; (JEeP 
KaTll TO 8vvaT() v of the Theaetetus (1750 -177 A). 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

LEONARDO T ARAN 81 

tich. Hence, I should pronounce a non liquet on the question of the 
original length of the epigram. 

To come back to the question of the original readings of lines 1-2, 
I submit that in this case the matter cannot be decided merely on 
internal grounds. Once again it is the question of attestation that 
takes precedence. Now if the epigram preserved by al-I,(ifp is a non­
shortened version of what his ultimate Greek source had-as there is 
good reason to believe - then the readings of lines 1-2 in Diogenes 
are probably the original ones. For the Arabic text implies that its 
Greek source had yata l .. d.v.50 This by itself creates a presumption in 
favor of a8avaTOv as the original reading in line 2. Moreover, the 
Arabic version of line 2, "his soul is in the rank of one who does not 
die," seems to presuppose that the Greek source had a8avaTwv 
rather than W-08EWV. And the coincidence in the key readings be­
tween Diogenes and al-~ifti's Greek source is decisively in favor of 
thinking that the text of the epitaph as it appears in E and Plan.b is a 
later, modified version. But even if the epigram in al-~ifti's Greek 
source also had four lines, it nevertheless follows that yata J.LEV is the 
only attested ancient reading, whereas fTwJ.W. J.LEV is not attested 
before ca A. D. 900 and need not be much earlier. 

In concluding, I should like to stress the following points. (a) The 
ascription to Speusippus in Plan.b is probably a conjecture of Planudes 
himself or of his source. But even if, as is possible though unlikely, 
he or his source found this ascription in Cephalas, the ascription is 
hardly likely to go back to early antiquity or to a good source. It is 
intelligible why a late or a Byzantine scholar made such a conjecture, 
and similar conjectural ascriptions occur often in our sources for the 
Greek Anthology.51 (b) It is highly improbable that the four-line epi­
gram or the related distich either with the readings in Diogenes 
Laertius or with those in E and in Plan.b is the original epitaph in­
scribed on Plato's tomb. (c) In the light of the extant evidence the 
epitaph in question-whether in two or in four lines-was transmitted 
as the second of a group of epigrams allegedly found on Plato's tomb. 
These were included in a biography of Plato that is ultimately the 
source both of Diogenes and of the Greek source of the Arabic life 
of Plato. This and the fact that the first epitaph is Hellenistic suggest 
that also the second epitaph is Hellenistic at the earliest. (d) The 
evidence does not suffice to ascertain whether originally the epitaph 

50 Professor Pingree tells me that there can be no question that the Arabic translator 
had yaw: ~lJ ..• / I/Jvxr, B' in his Greek source. 

51 Cf e.g. supra n.28 and Gow/Page xxviii ff with references. 



TARAN, LEONARDO, Plato's Alleged Epitaph , Greek, Roman and Byzantine Studies, 25:1 
(1984) p.63 

82 PLATO'S ALLEGED EPITAPH 

with which we are concerned was in two or in four lines. (e) The 
readings 'Yata /-LEV ... / ljJvXTJ B' £i8avarov are probably the original 
ones. The very nature of the evidence has made it necessary to 
discuss in this paper several different possibilities in the transmission 
of the two versions of this epitaph. But in none of them does the 
common opinion gain likelihood that the distich which E and Plan.b 

have preserved was written by Speusippus and was the epitaph ac­
tually carved on Plato's tomb. 

Notopoulos cites Waltz's statement that of several epitaphs devoted 
to the same person only one can be the original inscription on his 
tomb.52 But it is also possible that none of them is the original epi­
taph. Such is the case with the epitaphs allegedly inscribed on Plato's 
tomb. And so the dogmatism of Wilamowitz, who refused even to 
comment on the Plato epitaphs,53 turns out to be correct as an opin­
ion, though not as an attitude towards the evidence. On the contrary, 
the analysis of the transmission of the Plato epitaphs shows that, 
apart from its interest to the Platonist, it is also of importance for 
students of the Greek Anthology itself. 

COLUMBIA UNIVERSITY 

January, 1984 

52 Cf Waltz (supra n.3) 38-39. 
53 U. von Wilamowitz-Moellendorff, Platon 1 (Berlin 1919) 709, who, relying on 

Paus. 1.30.3, says: "Auch von seinem [sc. Plato's] Grabe wissen wir nichts weiter, als 
dass es nicht weit von der Akademie war."