PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

The Ancient Tradition on Antiphon 
Reconsidered 

Gerard]. Pendrick 

O NE OF THE STRONGEST ARGUMENTS for the existence in the 
late fifth century of a sophist Antiphon, author of I1Epi 
uA1l8dac; and other philosophical works, distinct from 

the politician and logographer Antiphon of Rhamnus, is a small 
but significant body of ancient 'separatist' evidence. In a recent 
discussion of the ancient tradition on 'Antiphon', Michael 
Gagarin seeks to explain away the separatist tendencies of the 
tradition and argues for the identity of the aforementioned 
Antiphons.1 I propose to re-examine Gagarin's treatment of the 
evidence2 -explicit and implicit-in order to demonstrate that 
the separatist case is stronger than Gagarin and others have 
supposed. 3 In addition, I shall argue that the 'unitarian' aspects 

1 "The Ancient Tradition on the Identity of Antiphon," GRBS 31 (1990: 
hereafter 'Gagarin') 27-44. Gagarin is in the main responding to G. Pendrick, 
"Once Again Antiphon the Sophist and Antiphon of Rhamnus," Hermes 115 
(1987: hereafter 'Pendrick') 47-60. 

2 Other recent discussions tending to unitarian conclusions include: f. 
Decleva Caizzi, "Le fragment 44 D.K. d' Antiphon et Ie probleme de son 
auteur: quelques reconsiderations," in 'H apxaia croqncrnJdj, The Sophistic 
Movement (Athens 1984) 97-107, «Ricerche su Antifonte. A proposito di 
POxy. 1364 fr. 1," in M. Capasso, F. De Martino, and P. Rosati, edd., Studi di 
filosofia presocratica (Naples 1985) 191-208, "II nuovo papiro di Antifonte: 
POxy. LII, 3647," in F. Adorno et al., Protagora, Antifonte, Posidonio, 
Aristotele. Saggi su frammenti inediti e nUO'lJe testimonianze da papiri 
(Florence 1986) 61-69, and "'Hysteron Proteron': La Nature et la loi selon 
Antiphon et Platon," Re'lJue de Metaphysique et de Morale 91 (1986) 291-310; 
M. Ostwald, "Nomos and Phusis in Antiphon's nEpl aA.T]8dm;." in M. 
Griffith and D. Mastronarde, edd., Cabinet of the Muses: Essays on Classical 
and Comparative Literature in Honor of Thomas G. Rosenmeyer (Atlanta 
1990) 293-306. These authors are not primarily concerned with the ancient 
testimonia and their arguments will not be considered here. 

3 Michel Narcy ("Antiphon [no. 209J," in R. Goulet, ed., Dictionnaire des 
philosophes antiques I [Paris 1989] 225-44, esp. 225-30) reviews the ancient 
testimonia on Antiphon. His discussion focuses on the evidence of the corpus 
Aristotelicum and of the pseudo-Plutarchean X orat., on the basis of which he 
confidently distinguishes Antiphon of Rhamnus from Antiphon the tragic 

215 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

216 ANTIPHON RECONSIDERED 

of the tradition do not constitute a decisive stumbling-block to 
the separatist position. 

The only surviving ancient discussion of the problem of the 
Antiphons is that of the rhetorician Hermogenes,4 who reports 
the opinions of Didymus of Alexandria and others as a preface 
to his own discussion of the style of the various works current 
under the name of Antiphon.s Hermogenes' report includes 
these points: (1) the claim that there have been many Anti­
phons, two of them crOqllcru:ucravu:~;(, (2) the identification of 
these latter two as Antiphon the rhetor (i.e., Antiphon of 
l\hamnus, cf 400.22 6 toivuv ·Paf.lVoucrlO~ 'Avnqlwv) and 

poet, but comes to no certain conclusions about the identity of the author of 
the sophistic works. The data furnished by the corpus Aristotelicum, however, 
are inconclusive (see n.27 infra) and the X orat. is a hopeless confusion. Here I 
shall concentrate rather on those strands of the ancient tradition that seem to 
permit a probable resolution of whether the author of the sophistic works is 
or is not identical with Antiphon of Rhamnus. 

4 On Hermogenes see Christ-Schmid-Stahlin I1.26 929-36; L. Radermacher, 
"Hermogenes (22)," RE 8.1 (1912) 865-77; and most recently M. Patill~n, La 
Theorie du Discours chez Hermogene Ie Rheteur (=Collections d'Etudes 
anciennes 117 [Paris 1988]), who distinguishes (8-19) the author of nept {Sewv 
("Hermogenes the rhetor") from Hermogenes the" sophist" who appears in 
Philostr. VS 577f. 

5 nEpt {SEWV 399.18-400.21 Rabe: nEpt bi: 'Avtl<pwV'tos AEyoVtOS uvaYKTl 
1tponm:'iv, Ott, K08a1tEp UA.Aot tE <pOotv aUK oAiYOt KOt L\iSu~oS,o ypa~~ottK6S, 
1tpOS 8i:KOt U1tO tOtOpioS <poivEtm, 1tAdouS ~i:v YEy6vootv 'Avtt<pwvtES, 8uo 8i: 
oi OO<ptOtEUOOVtES, <1v KOt f...iJyov uvaYKT] 1totTIooo80v <1v cts ~EV (OttV 0 
PTItWP, otmEp Ot <pOVtKOt <pEpOVtm f...iJym KOt (Ot) 8Tl~TlYOPtKOt KOt oom toutms 
01;l0Wt, (tEPOS 8i: 0 KOt tEPOtOO)(01tOS KOt ovnpo)(pttTlS AEYO~EVOS YEVEcr8at, 
olmEp Ot tE nEpl tT\s UATl9doS dvm AEYOVtat f...iJym KOt b nEpl o~ovotos [)((Xl Ot 
8Tl~TlYOptKOll )(Ol 0 nOAtttKOS. (yw 8i: EVE)(O ~i:v tOU 8w<popou twv (V tOtS 
A6yotS toutOtS {8EWV 1td80~ot 8uo tOUS 'AvttcpwvtaS YEvEcr8m (1tOAU yap WS 
QVtWs to 1t0pOAAattOV tWV (1ttYPOCPO~EVWV tT\s 'AATl8dos f...iJywv 1tpOS tOUS 
AOt1tOUS), £VEKO 8i: tOU KOt 1tOpa nAutWVt KOt 1tOp' UAAOtS tOtOpOU~EVOU 
1taAtv ou 1tEi80~0t' 80UKU8i8Tlv yap 'AvttcpwvtOS dvat tOU 'Po~vouoiou ~08Tltljv 
U)(OUW 1tOAAWV AEyoVtWV, )(Ot tOY ~i:v 'Po~vouowv doroS (KEtVOV, O~1tEP dotv 
Ot cpovtKoi, tOY 80UKU8i8Tlv 8i: 1tOAAip KEXWPtO~EVOV KOt KEKOtvWVTlKOtU tip 
don twv tT\s 'AATl8cios A6ywv, 1taAtV OU 1tci80~at. OU ~ljv un' dtE rtS 0 
'Avtt<pwv £yEVEtO, Suo f...iJywv dbEm tOOOUtOV uUTIA.ooV OtEOtTlKOOt XPTloa~£vos, 
dn: )(Ot buo, XWptS i:KUtEpOS 0 ~i:v tOUtO 0 8i: (K£tVO ~EtEAewv, uvuYKT] XWptS 
1tEpt [KOtEPO'\) btEAeEtv' 1tAEtOtOV yap, wS ECPO~EV, to ~EtO~U. 

6 OOcptOtEuOOVtES probably means "teach rhetoric" (e.g. Strab. 13.1.66), or 
more generally "be a sophist" (e.g. D.L. 9.56); cf Pendrick 55 n.39; LSJ S.v. 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

GERARD J. PENDRICK 217 

another Antiphon, said to have been a dream-interpreter; and 
(3) a division of works. 7 

The evaluation of Didymus' views as reported by Her­
mogenes is of crucial importance to the problem at hand. Like 
others before him,8 Gagarin (36f, 43) contends that Didymus 
presented no "specific" or "significant" evidence-e.g. bio­
graphical details 9-in support of the distinction he drew 
between Antiphon of Rhamnus and the 'other' Antiphon, but 
relied merely on considerations of style. For if Did ymus had 
offered such evidence (so the argument goes), Hermogenes 
would surely have cited it in support of the distinction drawn 
by Didymus; instead, Hermogenes' exclusive reliance on 
stylistic arguments proves that he knew no other ones. This 
argumentum ex silentio dies hard, but it should be rejected for 
the following reasons. First, Hermogenes is a critic of style, and 
his emphasis on the stylistic criteria both for and against 
Didymus' distinction probably reflects his own preoccu­
pations, not those of Didymus. It is misleading to say, as 
Gargarin (36) docs, that although "Hermogenes' primary con­
cern is style," he "has evidently investigated carefully the 
matter of identity. He cites the evidence of Plato and others fa­
voring the unitarian position." In fact, Hermogenes' "evidence" 

7 When Gagarin writes (36) "J-Ie [Hermogenes] is inclined to assign the 
sophistic works to the latter because, he reasons, their style is so different that 
the same man cannot have written them," he appears to imply that the 
division of works is Hermogenes'. The wording of the text shows otherwise: 
oilm:p Ot n: ncp\. 'tll-; <lATj8da-; dvm AEyoV'tat MSYOl K'tA. 

8 Gagarin 36£, 43; for a list of others who have so argued see Pendrick 56 
n.42. Another recent exponent of this view is B. Cassin, "Histoire d'une 
identitc: les Antiphons," L 'ecrit du temps 10 (1985) 65-77, whose discussion 
(67-70) is flawed by her misunderstanding of Hermogenes' technical 
terminology (on which cf Patillon [supra n.4] 103-278; G. Lindberg, Studies 
in Hermogenes and Eustathios [Lund 1977] 8-128, 200-62). By way of the 
supposed ambiguity of £100-; ("style," "genre"), Cassin (71H, cf Gagarin 38) 
passes from stylistic differences to the notion of "competence" and intimates 
that separatists ancient and modern divide the Antiphons not only on the 
basis of stylistic considerations but also in the belief that the 'rhetorical' 
works should belong to an orator and the 'philosophical' works to a sophist. 
No evidence whatsoever suggests that this consideration influenced IIer­
mogenes, Didymus, or any other ancient critic. 

9 Narcy (supra n.3: 228) remarks that Hermogenes' information on the 
'other' Antiphon" est dcpourvue d'elcments biographiques," but fails to note 
that the same holds for his information on Antiphon of Rhamnus. 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

218 ANTIPHON RECONSIDERED 

for the unitarian position is stylistic as well. 10 Secondly, 
Hermogenes may well have encountered Didymus' views on 
Antiphon in an abbreviated form (since Didymus' works were 
often excerpted) 11 or at second hand, which could account for 
his failure to cite the sort of specific evidence that Gagarin de­
siderates. Thirdly, neither Gagarin nor any other unitarian cites 
any evidence for the supposition that Didymus would base his 
solution to a literary-historical problem such as the identity of 
Antiphon exclusively or primarily on stylistic evidenceY 
Finally, Hermogenes does in fact allude to evidence other than 
that of style in support of Didymus' distinction, when he says 
npo~ DE Kat uno l(rtOpla~ <palVE'tat. Whatever Hermogenes 
means by tatopia, 13 his words imply acquaintance with factual 
information of some sort in support of the separatist posi tion, 
and it is not unreasonable to assume that Didymus also was 
acquainted with such information. (It may even be that 
Hermogenes is indebted to Didymus for his knowledge on this 
point.) In sum, nothing in Hermogenes' report warrants the in­
ference that Didymus based his distinction of the Antiphons 
exclusively or primarily on stylistic considerations; the ref­
erence to lO"topia implies rather the reverse. 

10 Hermogenes interprets Plato's remark (Menex. 236A) about someone 
taught rhetoric by Antiphon who could win repute praising Athenians among 
Athenians as an allusion to Thucydides (cf ]. S. Morrison, .. Antiphon," PCPS 
N.S. 7 [1961] 56) and evidence that Antiphon of Rhamnus was Thucydides' 
teacher. But since Thucydides' style resembles that of ITEpt uAT)edac;, the 
author of that work, Hermogenes reasons, should be identical with the 
Antiphon mentioned by Plato (i.e., the Rhamnusian) who was Thucydides' 
teacher. 

II Cf L. Cohn, "Didymos (8)," RE 5 (1903) 446. 
12 An additional argument against the notion that Didymus divided the 

Antiphons on grounds of style is the division of works that Hermogenes 
reports (probably from Didymus; cf supra n.7). For the substantial stylistic 
differences between the extant fragments of ITEpt uAT)edac; on the one hand 
and ITEpt ollovoiac; on the other (which led W. Nestle, Vom Mythos zum 
Logos 2 [Stuttgart 1942] 385-88, 39M with nn.114ff, 399 with n.130, to attribute 
these works to different Antiphons) suggest that something other than 
stylistic criteria lay behind the ancient attribution of these works to one and 
the same author. 

13 Gagarin (36 n.31) criticizes C. Wooten (Hermogenes' On Types of Style 
[Chapel Hill 1987]) for translating iotopiac; as "history," a sense that is by no 
means impossible (cf LS] S.v. II). He thinks it refers to Hermogenes' own 
"inquiry" or "investigation." 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

GERARD J. PENDRICK 219 

It ought furthermore to be emphasized that on Hermogenes' 
testimony (which there is no reason to doubt) Didymus was far 
from alone in distinguishing the Antiphons, as "many others" 
(aAAol ... OUK OAtYOl, litotes for "quite a few") shared his view. 
Therefore, when Gagarin characterizes the unitarian tradition in 
antiquity as "nearly unanimous" (43) and the separatist tradition 
as "isolated" (43; cf "widely scattered exceptions," 44), he un­
justifiably ignores Hermogenes' explicit testimony to the 
contrary. 

Whatever evidence Didymus may have had for his views on 
Antiphon, for us Xen. Mem. 1.6.1-15 provides the strongest 
support for the separatist thesis. Xenophon depicts the at­
tempts of "Antiphon the sophist" ('AvncproV'ta 'TOV aocptcHT)v) to 
win over some of Socrates' associates (6 yap 'Avncprov non: 
~ouA6~EVOs 'TOUs aUVOUataa'Tas au'Tou naPEAEaeat npoaEA­
emv 'Tip LWKpa'TEt K'TA.).14 Two interrelated features of Xeno­
phon's Antiphon that bear on the question of his identity are his 
status as a professional, paid teacher (implied by the entire 
episode) and the designation "sophist," which Xenophon 
attaches to him. Like other unitarians, Gagarin (30-33) argues 
that these features are compatible with the belief that Xeno­
phon's Antiphon is the Rhamnusian. 

In the first place, however, the evidence adduced by Gagarin 
and others to prove that Antiphon of Rhamnus was a teacher is 
very weak. 1S It consists of the following: (1) Socrates' jest at PI. 
M enex. 236A that even someone more poorly educated than 
himself, someone who had been taught music by Lamprus 16 

and rhetoric by Antiphon of Rhamnus, would be able to win 
repute by praising Athenians before Athenians; (2) the late 
tradition that Antiphon of Rhamnus was the teacher of 
Thucydides; (3) the Tetralogies attributed to the Rhamnusian. 
But the M enex. passage is perfectly intelligible on the assump­
tion that Antiphon is mentioned because he was one of the 
outstanding practitioners of dicanic rhetoric in the later fifth 

14 The encounters of Antiphon the sophist and Socrates in M em. are 
probably fictitious, whether invented by Xenophon or borrowed from 
another Socratic author (cf n.28 infra). For a full discussion of this passage 
with references to earlier literature, see Pendrick 47-52. 

15 Cf. Andrewes, in A. W. Gomme, K. ]. Dover, and A. Andrcwcs, A 
Historical Commentary on Thucydides V (Oxford 1981) 173f. 

16 On Lamprus see H. Abert, "Lamprokles (1)," RE 12.1 (1924) 586£ and A. 
von Blumenthal, "Sophokles (1)," RE IlIA.1 (1927) 1042f. 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

220 ANTIPHON RECONSIDERED 

century; his functioning as a teacher need be-and probably 
is-no more than an accomodation to the context. 17 The late 
tradition that Antiphon was Thucydides' teacher rests in all 
likelihood on mere inference from Thucydides' high praise of 
the Rhamnusian (8.68) and is a good example of Hellenistic 
scholars' fondness for constructing pupil-teacher relation­
ships.18 Finally, there is good reason to doubt that the 
Tetralogies were written as models for the use of students; 
other motives for their composition and 'publication' are 
conceivable and do not by themselves prove that their author 
taught rhetoric. 19 Gagarin, to be sure, argues (30) that" Anyone 
with the intellectual interests and wide influence of the 
Rhamnusian could surely be considered a teacher in some 
sense.» But Xenophon's depiction of Antiphon as a professional 
educator and rival to Socrates seems to require that this Anti­
phon have been rather more than "a teacher in some sense.» 

As regards the epithet 'sophist', it must be conceded to 
Gagarin (31 f) and others that this term could be applied to a 
logographer like Antiphon of Rhamnus, especially in a hostile 
context. 20 The question, however, is not whether Xenophon 
could have called the Rhamnusian a 'sophist' but whether he 
would have done so, and why. The Rhamnusian is usually 
identified in ancient texts by name and demotic, or by name 
along with the epithet PT)'twp (="politician"), or by bare name; 
'sophist' would be an unusual designation for him,21 and we 

17 Cf R. Clavaud, Le Minexene de PLalon et La rhetorique de son temps 
(Paris 1982) 76, 263-77; Pendrick 49 n.ll. Note that in his speech of defense 
on the charge of treason Antiphon of Rhamnus refers (fr. 1, p.109.2f 
Thalheim) to the prosecutors' mention of his paid logographical activity, but 
says nothing about teaching oratory for pay. 

18 Cf [Plut.] X orat. 832E: KandAtO~ 0' EV 10 lH:pl a\J'tou (1)V1aYflan 
801)1(1)oiOo1) 'tOU (1)yypa<pE(o~ 1(a8TlYll'tl)v [Wyttenbach: fla8Tl'tl)v M ss.] 'tE 1(­
,wipE'tat YLYOVEVat £1; (;)v btatVEt'tUl nap' aU1w 6 'AV'tHPWV Cf further 
supra n.10 and Pend rick 49 n.11. 

19 Cf K. J. Dover, "The Chronology of Antiphon's Speeches," CQ 44 (1950) 
44-60, esp. 59. Gagarin's claim (30 n.13) that one purpose of the Tetralogies 
"must have been to provide others with a model for techniques of forensic 
oratory" takes no account of the difficulties involved in this supposition to 
which Dover calls attention. 

20 Cf Pendrick 51 f with nn.19-23. 

21 "Antiphon the sophist" appears (apart from Hermogenes' ao<ptm£\'mav­
't£~) elsewhere only in Simp!' in Phys. 273.36 Diels and the confused entries in 
the Suda s.v. 'Av'tt<pwv (A2744-46 Adler: from Hesychius, cf Adler ad lac.; H. 
Schultz, "Hesychios (10)," RE 8 [1913] 1322ff). 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

GERARD J. PENDRICK 221 

should have to account for Xenophon's use of it. Gagarin 
senses this and suggests that Xenophon may be trying to 
distinguish his Antiphon, the Rhamnusian, from Antiphon the 
trierarch (son of Lysonides, cf [Plut.] X orat. 832F-33B) who is 
mentioned (in a speech of Theramenes) at Hell. 2.3.40. But this 
fails to account for the specific designation 'sophist' and 
especially for the "perjorative touch" that Gagarin (33) admits 
inheres in the designation. On the other hand, if Xenophon's 
Antiphon was a professional teacher ('sophist') distinct from the 
Rhamnusian, this would account not only for the epithet 
'sophist' and the hostility of Xenophon's portrait as a whole, but 
also more generally for Xenophon's use of this Antiphon as a 
type of the sophistic opponent of Socrates. 22 It is noteworthy 
that Gagarin, like other unitarians, does not even attempt to 
explain why Xenophon would choose Antiphon of Rhamnus 
(of all people) as a typical sophistic opponent of Socrates, nor 
why he would harbor towards the Rhamnusian (with whose 
politics at least one would suppose he sympathized) the 
animosity evident in his portrait of Antiphon. Gagarin, how­
ever, objects (31 f) that Xenophon could not have intended to 
identify a sophist Antiphon distinct from the Rhamnusian 
unless such a sophist had been well-known in the fourth 
century, and even then would "surely" have distinguished him 
by means of a demotic, patronymic, or ethnic. That the sophist 
Antiphon was well-known in the fourth century Gagarin 
regards as "very unlikely" because the next reference to such a 
figure occurs nearly a millenium later (Simp!. in Phys. 273.36 
Diels). But the scarcity of later references to "Antiphon the 
Sophist" does not rule out the possibility of his having been a 
familiar figure in the fourth century B.C. And in fact, 
Xenophon's very wording CAvncpwv'ta 'tOY crOCPlcr'tl)V) seems to 
imply that he was sufficiently well-known to be identified 
briefly thus as "the sophist." 23 As to the objection that 
Xenophon would have identified his Antiphon by means of an 

22 On Xenophon's Antiphon as a type cf. U. von Wilamowitz-Moellen­
dorff, Der Glaube der !lellenen II (Berlin 1932-33) 217 n.1; O. Gigon, Kom­
mentar zum ersten Buch von Xenophons Memorabilien (=Schweizerische 
Beitrdge zur illtertumswissenschaJt 5 [Basel 1953]) 152, 165. 

23 I n any case, the rarity of the designation" Antiphon the sophist" is at 
least as strong an argument against the notion that Xenophon would have 
referred in this way to Antiphon of Rhamnus as it is against the possibility 
that a sophist Antiphon was well-known in the fourth century B.C. 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

222 ANTIPHON RECONSIDERED 

ethnic, demotic, or patronymic, Gagarin seems to forget that 
Xenophon seldom employs additional means of identification 
beyond the bare name, and especially eschews demotics in the 
identification of Athenians. 24 The likeliest conclusion remains 
that Xenophon designates his Antiphon as "the sophist" in 
order to distinguish him from the (better-known) Rhamnu­
sian.25 

Considerable evidence suggests that the identity of Xeno­
phon's Antiphon was controversial already in antiquity. The 
author of the pseudo-Plutarchean X orat. or his source identi­
fied Xenophon's Antiphon with the Rhamnusian. 26 But in 
Diogenes Laertius' citation of a list of disputes involving famous 
poets and philosophers from the third book of Aristotle's On 
Poets,27 Antiphon 6 UPCX-rOO'K61tos is said to have contended 

2. Cf D. Whitehead, "Athenians in Xenophon's H ellenica," LivClM 13 
(1988) 145ff, who concludes (146) that "Xenophon took no systematic care, 
either with Athenians or in general, to distinguish between homonyms in his 
narrative by providing them with patronymics or other marks of additional 
identification." On the other hand, Xenophon's use of such identificatory tags 
as Callistratus 0 OllJlllYOPO~ (Hell. 6.2.39, 3.3) and Nicostratus 0 lWA.O~ 
btllCaAOUJlEVO~ (Hell. 2.4.6; all three passages cited by Whitehead 146) shows 
that the conditions Gagarin (.32) lays down for the use of supplementary 
designations are much too rigid. That the sophist Antiphon was an Athenian 
is likely on other grounds and is suggested as well by Xenophon's failure to 
identify him by means of an ethnic (cf Mem. 4.4.5, Symp. 4.62; An. 2.6.17, but 
contrast Symp. 1.5,4.62 and Mem. 2.1.21). 

25 The evidence of Xenophon fully justifies the modern custom of dis­
tinguishing Antiphon "the sophist" from Antiphon of Rhamnus, despite the 
objections of Morrison (supra n.10: 54), Narcy (supra n.3: 230), and others. 

26 832c: OWtPl~fJV or ouvrOtllOE [Antiphon of Rhamnus] lCat LCOlCpatEl tip 
qnAooo<pql olE<prpEtO tfJv \mrp tOOV A.6ycov ow<popav OU <plAovdlCCO~ aAA' 
EAE'{lCtllCOO~, w~ 3EVO<pOOV iOtOPlllCEV EV to'i~ 'A1t0JlVllJlOVEUJlamv. (On 
qnAovdlCco~ cf below.) Likewise Photo Bibl. 486 (=VII 42 Henry) and the 
anonymous Vito Antiphontis 7, both of which seem (pace Gagarin 39 with 
n.43) to depend on [Plut.]. 

27 2.46=fr. 75 Rose. In the remainder of the corpus Aristotelicum the sophist 
Antiphon is referred to three times without distinguishing epithet (Ph. 185a17, 
193a12ff; Soph. E I. 172a7); the Rhamnusian is referred to three times by name 
alone (Ath. Pol. 32.2, Eth. Eud. 1232b6-9, fr. 624 Rose), and Antiphon the 
tragic poet is referred to twice with the epithet o1tOlllnldRh. 1385a10, [Mech.] 
847a20) and three times without it (Eth. Eud. 1239a37, Rh. 1379b15, 1399b27). 
Like others before him (e.g. Morrison [supra n.l0] 51-55; K. Joel, Geschichte 
der antiken Philosophie I [Tiibingen 1921] 663 n.3), Gagarin (33) infers from 
Aristotle's manner of citation that he identified the author of the sophistic 
works with the Rhamnusian. But Aristotle probably felt it unnecessary in 
most cases to identify explicitly which Antiphon he had in mind, especially in 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

GERARD J. PENDRICK 223 

eristically (£<jnAOVElKH) with Socrates. Because Aristotle in all 
probability refers to the disputes in the Memorabilia, 28 the epi­
thet't£pa'!oO"Korcoc; applies to Xenophon's Antiphon. Whether 
it goes back to Aristotle (as is usually assumed) or was added by 
Diogenes or by an intermediate source, the epithet "diviner" 
implies that its author considered Xenophon's Antiphon to be 
someone other than the Rhamnusian (who would hardly be 
referred to in th is way). 29 Even clearer evidence is provided by 
a passage of Athenaeus (63 7E -P), in which Democri tus charges 
that after Adrastus had written a monograph in six books on 
historical and Ii terary problems in Theophrastus' ITEP!. f)9wv and 
Aristotle's Nicomachean Ethics,3o in which he discussed at 
length Antiphon the tragic poet and the character Plexippus 
from his M eleager, a certain Hephaestion plagiarized 31 Adrastus 

works such as the Ph. and the Soph. EI., which were addressed to the 
specialized and knowledgeable audience of his school. If Aristotle could take 
for granted his audience's knowledge of the details of Antiphon's quadrature 
(as he evidently does in Ph. 185a14-17 and Soph. E I. 172a7), surely he could 
also assume their familiarity with the identity of the Antiphon in question. 
Narcy (supra n.3: 225ff) ignores this consideration in his discussion of the 
Aristotelian evidence. 

28 Some scholars (e.g. Gigon [supra n.22J 152; K. Joel, Der echte und der 
xenophontischc Sokrates II.2 [Berlin 1901J 638; G. Altwegg, De Antiphontc 
qui dicitur Sophista Quaestionum particula I. De libro ITEPI OMONOIAI 
scripta [Basel 19081 7f) have argued that Aristotle depends not on Xenophon 
but on an older Socratic work that was Xcnophon's source as well. Although 
this is not improbable, their chief argument (Aristotle's designation of 
Antiphon as n:pm:oO'Korro<; against Xenophon's 'tOY O'O<plO''tljV) is very weak. 

29 Gagarin (41 n.50) rejects this line of reasoning because we know neither 
who added the epithet 't£pa'toO'Korro<; nor why it was added. But our 
ignorance on the former point does not, in my view, seriously weaken the 
conclusion drawn in the text above. If Aristotle really called Xenophon's 
Antiphon b 't£pa'toO'Korro<;, this would confirm Hermogenes' identification of 
his 'other' Antiphon, i.e., the sophist, with Antiphon the diviner. 

30 The Ad rastus (Casaubon's correction for the meaningless aopavwv of the 
Mss) in question is almost certainly the well-known Aristotelian commentator 
of the second century, on whom see P. Moraux, Der Aristotelismus bei den 
Griechen II (Berli n 1984) 294-332, esp. 323-30 on his monograph on Aristotle 
and Theophrastus; on Theophrastus' IT£Pl Tj8wv cf O. Regenbogen, "Theo­
phrastos (3)," RI:' Suppl. 7 (1940) 1479f. 

31 This Hcphaestioll is perhaps the well-known metric ian, although other 
identifications are conceivable: cf. Moraux (supra n.30) 295 with nn.7f with 
further literature. On accusations of plagiarism in ancient literture cf E. 
Stemplinger, J)cr Plagiat in der griechischen Literatur (Leipzig 1912); K. 
Ziegler, "Plagiat," R E 20.2 (1950) 1956-97. It is impossible to te II whether 
Athenaeus' (Dcmocritus') allegation is true, false, or exaggerated. 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

224 ANTIPHON RECONSIDERED 

and published a monograph entitled32 On the Antiphon in 
Xenophon's Memorabilia. Unfortunately, nothing further is 
known of this work. But it is difficult to believe that the identity 
of Xenophon's Antiphon was not one of the topics discussed 
by Hephaestion. Gagarin rejects this inference and argues (41£) 
that Adrastus probably discussed the characters of Antiphon 
the poet and Plexippus and that Hephaestion in turn discussed 
Antiphon's manner of arguing with Socrates (c{. the passages 
from [Plut.], Phot., and the Vito Antiph. cited above).33 But 
Gagarin has failed to take into account the available evidence 
that suggests that Adrastus' monograph devoted extensive 
attention to literary-historical questions of all kinds. 34 In light of 
our evidence for his broad interests and learning, and in view of 
Athenaeus' remarks (Kat 1tAElcr'ta 5cra Kat 1tEpt au'tou 'tou 
'Avncp&vwc; £l1tOVWC;), it is likely that Adrastus' discussion of 
Antiphon the poet was more wide-ranging than Gagarin 
suggests and that it touched on the question of the identity of 
this Antiphon vis-a-vis other people of the same name, in­
cluding Xenophon's Antiphon.35 

Many other ancient and medieval sources, from Aristotle to 
the Suda, quote from or refer to the works attributed in 
Hermogenes to the sophist Antiphon without distinguishing 
their author from Antiphon of Rhamnus. 36 Among these, 

32 Gagarin (41 with n.48) follows Morrison in interpreting Athenaeus' 
words bt€ypa\jf£v 'tt ~t~Aiov as" added a book." But the same verb is used a 
few lines earlier in the sense of "entitle" (cf LS] S.7.I. IL2), and there is no 
reason to think it means anything different here. 

33 Gagarin argues further that Hephaestion's title "does not suggest that the 
work concerned the question of identity." Perhaps; but still less does it suggest 
that Hephaestion discussed" Antiphon's manner of argument with Socrates." 

H Cf Moraux (supra n.30) 323-30; on Adrastus' philological and literary­
historical interests and knowledge cf Moraux passim, esp. 314-17, 323ff, 330ff. 

35 It is not necessary to suppose, as some have (e.g. W. Aly, Formprobleme 
der fruhen griechischen Prosa (=Philologus Suppl. 21.3 [Leipzig 1929]) 112 and 
Untersteiner, in M. Untersteiner and A. Battegazzore, Sofisti. Testimonianze e 
Frammenti IV [Florence 1962] 26), that Adrastus or even Hephaestion 
identified Xenophon's Antiphon with the tragic poet. 

36 Ancient writers from Thucydides (8.68) and Lysias (12.67) onwards who 
cite or refer to Antiphon of Rhamnus without distinguishing him from the 
sophist Antiphon represent a different case. The Rhamnusian was a famous 
politician and orator familiar to anyone in antiquity who had read 
Thucydides or studied the history of rhetoric or fifth-century Athens. It is 
quite unlikely that ancient authors who wished to refer to him felt it necessary 
to distinguish him from the much less familiar sophist, even if they were 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

GERARD J. PENDRICK 225 

Gagarin sees particularly strong support for the unitarian 
position in Valerius Harpocration, whose Lexicon of the Ten 
Attic Orators37 cites speeches of the Rhamnusian and the works 
attributed to the sophist Antiphon in precisely the same way 
without any hint that they may belong to different authors. 38 

Thus Gagarin infers that Harpocration must have known of 
Didymus' separatist arguments and dismissed them, and further­
more could not have been aware of "any significant scholarly 
opinion favoring a separatist view." These inferences are 
unjustified. In the first place, although Harpocration cites 
several commentaries of Didymus on the Attic orators and 
utilizes some of his lexicographical works, we do not know 
where Didymus expressed his views on Antiphon, and it is 
completely uncertain (pace Gagarin) whether Harpocration 
knew of them.39 In the second place, although Harpocration 
questions the authenticity of two speeches ascribed to Anti­
phon, it is unclear on what source (or sources) he depends for 
ascriptions in the corpus Antiphonteum (and elsewhere).40 

aware of the distinction. Critics such as Gagarin, Morrison (supra n.10) 
passim, Aly (supra n.35) 109-13, Altwegg (supra n.28) 6-12, and Narcy (supra 
n.3) 225-30, who have collected and analyzed the ancient evidence on 
"Antiphon" with a view to the problem of identity, have unjustly neglected 
this consideration. 

37 Gagarin's attribution to this Harpocration of a nCpl ,mv 'AV1HpmV1o~ 

()Xl1~6:,wv must be a slip; the Suda attributes it to a Gaius Harpocration: cf L 
Radermacher, "Harpokration (4)," R E 7.2 (1912) 2412. My citations of 
Harpocration are by page and line of Dindorf's edition. 

38 Some of Gagarin's figures (39) for Harpocration's citations of Antiphon 
are inaccurate: Harpocration cites nCpl O:ATJOEiac;, nEpl61l0voia~, and nOAnt1(6~ 
by title 29 times, not 26; and ncp't O:A110£ia~ is cited once (pace Gagarin) 
without specification of book (s:v. urOt, 4.8). 

39 On Harpocration's use of Didymus see Cohn (supra n.11) 458f. Harpo­
cration cites Didymus 39 times (not 36, as Gagarin states), but not one of 
these has to do with a problem of ascription. There is no evidence that 
Didymus wrote a commentary on Antiphon (cf Cohn 460)-although of 
course he may have. 

40 The speeches in question are npo~ (I>iAt1t1tov O:1toAoyia (90.7£) and Ka'ta. 
1tp'\)'t6:vcw~ (269.14f). The only authorities that Harpocration cites by name on 
a problem of ascription are Dionysius of Halicarnassus and Callimachus on 
the authenticity of n po<; Kpniav 1tEPl 'tau £VC1ttO"xTJIlIl0,1O<;, attributed to 
Demosthenes (133.4-7). There is reason to believe (see K. J. Dover, Lysias and 
the Corpus Lysiacum [Berkeley 1968J 15-19) that Harpocration made use of 
Dionysius' opinions on the ascription of speeches in the Lysianic corpus, but 
no evidence that Dionysius devoted attention to such problems in the corpus 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

226 ANTIPHON RECONSIDERED 

Caecilius of Caleacte is known to have devoted critical attention 
to problems of ascription in the corpus Antiphonteum, but 
Harpocration cites him only once (118.9), and not on a matter 
of ascription. 41 Further, it is clear from several cases where the 
evidence of papyri has made it possible to compare Harpo­
cration's work with the sources available to him that he offers 
only a small selection from the learning at his disposal. 42 In view 
of these considerations, it is quite unwarranted to draw the 
inferences Gagarin does concerning Harpocration's knowledge 
of discussions on the identity of Antiphon, and in particular to 
infer that he knew and rejected the separatist views of Didymus 
and others. 

What Harpocration's citations from DEpt eXATl8dac;, DEpt 
O/lOVOlac; and DOAt'ttKOC; do demonstrate is that these works 
formed part of the corpus Antiphonteum in antiquity.43 How 

Antiphonteum. It. should be noted that Harpocration is not always consistent 
in the matter of ascription. For example, he cites Ka'tCt Nwipa<;, attributed to 
Demosthenes, five times with the qualification d yviJOtO<; (79.15, 89.19, 96.11, 
161.8, 188.18), but seven times without it (e.g. 24.14 etc., 49.12, 65.7, 76.7, 79.4). 
Similarly, KmCt epao'U~ouAo'U, attributed to Lysias, is cited four times with, 
and four times without, the qualification d yviJOtOC;; and Ka'tCt N tKiDo'U 
(apyia<;), attributed to the same author, is cited three times with, but six times 
without, the d yviJcno<;. 

41 On Caecilius, who according to [Plut.] X orat. 833c=fr. 100 Of en loch 
rejected twenty-five of the sixty speeches current under the name of Antiphon, 
cf Blass, Au. Ber. 12 (1887) 102ff; J. Brzoska, "Caecilius (2)," RE 3.1 (1897) 
1181 f. Blass claimed (102) that Harpocration reflects Caecilius' views on the 
ascription of Antiphontean speeches; this may be right but is not supported 
by evidence. Although Aly (supra n.35: 113) and others have maintained that 
Caecilius did not distinguish the sophist Antiphon from the Rhamnusian, 
Blass (1 02f) thought that he did and that accordingly he classified the 
sophist's writings among the works falsely ascribed to the Rhamnusian. The 
reason why Harpocration does not question their ascription, according to 
Blass (103), is because "sie doch von einem Antiphon stammten." Again, this 
may be right but there is no evidence. Note that Harpocration seems not to 
have utilized, at least directly, Caecilius' lexicographical works on the orators: 
see Brzoska 1186. 

42 Cf H. Schultz, "Harpokration (5)," RE 7.2 (1912) 2413f with references. 
We possess, however, only an abridged version of Harpocration's Lexicon. In 
his attempt to use the evidence of Harpocration to discredit that of Didymus, 
Gagarin (39f) presents an exaggerated picture of the former's scholarship. 

43 The same conclusion follows from Hermogenes and Philostr. VS 500 
(ooqno'ttKOt DE Kat En:pot /lEV, ooqno'ttKwn:pO<; DE 0 'Y1tEP 'tT]<; o/lovola<;). In 
contrast, the complete absence in the lexicographical tradition of citations 
from the dream-book current in antiquity under the name of Antiphon 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

GERARD ]. PENDRICK 227 

and when they came to be included therein is not known; a 
likely guess is that Callimachus catalogued them among the 
works of Antiphon of Rhamnus in his 1tlVa~ of orators. 44 Now, 
the bare fact of their inclusion no more proves that their author 
is identical with Antiphon of Rhamnus than (say) the inclusion 
of speeches of Apollodorus son of Pasion in the corpus Demos­
thenicum proves that Apollodorus was the same person as 
Demosthenes. 45 Yet there can be little doubt that this inclusion 
exercised great influence in later antiquity in helping to 
perpetuate (if not to create) the confusion of the sophist 
Antiphon with Antiphon of Rhamnus. 46 For once these works 
were included in the corpus Antiphonteum, it was possible for 
anyone to cite them without reservation as belonging to 
'Antiphon' or 'Antiphon of Rhamnus', whether or not he may 
have known or suspected that they belonged to a sophist 
Antiphon distinct from the Rhamnusian. The corpus 
Platonicum provides an instructive parallel: clearly, later writers 

implies that this work was not included in the standard corpus Antiphon­
teum. 

44 On Callimachus' 1ttvmw; cf R. Pfeiffer, History of Classical Scholarship 
I (Oxford 1968) 127ff; for the Pll'toptlca cf frr. 430ff, 443-48 Pfeiffer. 

45 Cf A. Lesky, Geschichte der griechischen Literatur 2 (Bern 1971) 671. 
Dover's discussion (supra nAO: 23-26) of the uncertainties involved in the 
ascription of works of the Attic orators and the value of Callimachus' 
ascriptions is important in this context. 

46 A passage of Philodemus (On Poetry 187.3=87B93 D.-K.) has been 
variously interpreted as the earliest extant instance of confusion between the 
Antiphons (so Altwegg [supra n.28] 8) and Untersteiner [supra n.35] 167) or as 
evidence, almost contemporary with that of Didymus, for the unitarian view 
(so Gagarin 33 with n.23; F. Decleva Caizzi, "Le fragment" [supra n.2J 97 
with n.13). In the course of a discussion about the effects of the sounds of 
letters (cf F. Sbordone, "Filodemo e la teoria dell' eufonia," RendNap N.S. 30 
[Naples 1955J 25-51), Philodemus refers to 'ttvO~ ,oov apxat(()v 'A V't\c:poov,o~, 
£1,' o-ov Pll'OptlCO~ £11:£ Kal c:plA6croc:po~ iJpoUAl"t' dvat. T. Gomperz ("Philodem 
und die asthetischen Schriften der Herculanischen Bibliothek," SBVienna 
123.6 [1891] 49 n.3) saw here a reference to Glaucus of Rhegium's On the 
Ancient Poets and Musicians, which some in antiquity attributed to 
"Antiphon" ([Plut.J X orat. 8330). But D.-K. ad loco more plausibly suggest 
that Philodemus is referring to the 'Pll'toPlKo.l'tEXVat current in antiquity 
under the name of Antiphon (cf especially fr. 76 Blass-ThalheirIl, as well as 
frr. 73, 75). In any case, it is doubtful whether this passage has anything to do 
with the identity of Antiphon. It appears rather that Philodemus is simply 
expressing a sarcastic judgement on what he sees as the philosophical 
pretensions (Kat c:ptA6croc:po~ iJpoUAE't' dvat) of the author of the work in 
question. 



PENDRICK, GERARD J., The Ancient Tradition on Antiphon Reconsidered , Greek, Roman and 
Byzantine Studies, 34:3 (1993:Fall) p.215 

228 ANTIPHON RECONSIDERED 

felt free to cite as Plato's even works that they knew were 
spurious, on the principle that any work included in available 
editions of Plato could be quoted as Plato's.47 In these cir­
cumstances, the 'unitarian' tendencies of the ancient tradition 
on Antiphon, to which Gagarin 48 and others attach great weight, 
lose much of their force as evidence. For Galen, Harpocration, 
and the other authors who cite the sophistic works without dis­
tinguishing their author from Antiphon of Rhamnus simply 
reproduce the mistaken ascriptions handed on in the corpus 
Antiphonteum. 

In conclusion, a significant body of ancient evidence points to 
the probable 49 existence in the late fifth century of a sophist 
Antiphon distinct from Antiphon of Rhamnus. This emerges 
most clearly from Xenophon's portrait of "Antiphon the 
sophist" in the Memorabilia, and is confirmed by the opinion of 
Didymus (and many others) as reported by Hermogenes. The 
uncertainty in later antiquity regarding the identity of Xeno­
phon's Antiphon lends futher support to this conclusion. On 
the other hand, if the majority of (extant) ancient authors who 
refer to the works of the sophist Antiphon ignores this dis­
tinction, this does not prove that it is mistaken. The failure to 
distinguish Antiphon the sophist may result from the inclusion 
of his works in the corpus Antiphonteum. 

GEORGIA STATE U~IVERSITY 

May, 1994 

47 Cf L. Taran, Academica. Plato, Philip of Opus and the Pseudo-Platonic 
Epinomis (Philadelphia 1975) 7 with n.23. 

48 Passim, esp. 27f, 37, 39f, and 42f. 
49 It must be stressed that the arguments presented here do not amount to 

conclusive proof of the separatist case. To admit this, however, is not to 
detract from the conclusions reached in this paper; for on the problem of the 
identity of Antiphon, probability is the best that can be achieved without new 
and decisive evidence.