WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

Sophistic Interpretations 
and Greek Treaties 

Everett L. Wheeler 

O ATHS played an important role in Greek private and public life, 
and particularly in international relations.1 No promise, con­
tract, agreement, truce, or treaty had binding force without its 

validation by a sworn oath.2 Oaths, however, not only validated 
agreements but also guaranteed them, for oaths were taken in the 
name of the gods who were thought to punish the perpetrators of 
perjury, bad faith, and oath breaking as well as their families and 
descendants.3 To break an oath meant waging war with the gods, and 
even gods could suffer from committing perjury.4 Condemnation of 
mala .fides in international agreements is seen already in Homer, and 
oaths continued to be an indispensable component of treaties until 
the eighteenth century.5 

So much is general knowledge. The books of Hirzel and Plescia, 
however, pay little attention to epigraphical evidence. Oaths pre­
served in inscribed treaties can include such words and phrases as 
a8oAw~, ou8E T€XYTI ou8E 1-'11Xavfj, and a8oAw~ Kat af3Aaf3€w~, whose 
functions have been only cursorily examined.6 Are such phrases 
merely routine formulas confirming bona .fides, or does their inclu-

1 The following will be cited by author's name alone: Elias J. BICKERMAN, "An Oath 
of Hannibal," TAPA 75 (944) 87-102; Rudolf HIRZEL, DerEid:einBeitragzuseinerGe­
schichte (Leipzig 1902); Coleman PHILLIPSON, The International Law and Custom oj An­
cient Greece and Rome (London 1911); Joseph PLESCIA, The Oath and Perjury in Ancient 
Greece (Tallahassee 1970); Peter SIEWART, Der Eid von Plataiai (Vestigia 16 [1972]). 

2 Thuc. 2.5.5-6; Oem. 7.33-37; Phillipson I 406-07; Plescia 59. In addition to Plescia 
and Hirzel for general studies of oaths, see E. Ziebarth, RE 5 (1905) 2076-83 s. v. 
"Eid"; A. Steinwenter, RE 10 (1918) 1253-60 s. v. "Iusiurandum." 

3 Hdt. 6.86; Phillipson I 389. 
4 Waging war with the gods: Xen. Anab. 2.5.7, Hell. 3.4.11; perjury of gods: Hes. 

Theog. 783-806; Plescia 3-4. 
5 II. 3.276-91, 4.234-37, 7.351-53. The last instance of the use of an oath to confirm 

an international treaty was in the French-Swiss alliance of 1777 sworn at the cathedral 
of Solothurn: L. Oppenheim, International Law F, ed. H. Lauterpacht (London 1948) 
835; Arthur Nussbaum, A Concise History oj the Law oj Nations (New York 1954) 126; 
J. H. W. Verzijl, International Law in Historical Perspective I (Leiden 1968) 389. 

6 Siewart 33-34, cf 38-40; Bickerman 96; Paul Gratzel, De pactionum inter Graecas 
civitates jactarum ad bellum pacemque pertinentium appellationibus jormulis ratione (Halle 
1885) 32 n.2. 

253 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

254 SOPHISTIC INTERPRETATIONS AND TREATIES 

sion indicate a genuine fear of treachery or circumvention? These 
phrases, which I shall term 'anti-deceit clauses', usually occur within 
the oath rather than among the stipulations of the agreement, as the 
oath validates and guarantees those stipulations~ if the oath is abused 
or circumvented, then the entire pact loses its binding force. The 
terms of these anti-deceit clauses are suggestive. 80AO~, 'TEXVTJ, J.LTJ­
xav';', although generic words for trickery and deception in many 
contexts, also belong to the more specialized vocabulary of strata­
gems and sophistry.7 The intent of anti-deceit clauses in Greek trea­
ties needs to be examined from a ruse perspective. 

Anti-deceit clauses imply a fear of deception. How do they relate to 
sophistic interpretations of oaths, or what the Greeks called a OPKO~ 
UO<pur'TtKO~?8 A sophistic interpretation involved neither perjury, i.e. 
swearing something false, nor breaking an oath, but rather an overly 
literal interpretation of the wording of the oath or agreement, or 
playing on some ambiguity of meaning to produce an interpretation 
contrary to that intended and obvious, whereby the sophistic inter­
preter can, in any event, still claim fidelity to what was actually 
sworn. Probably the most famous sophistic interpretation is in Shake­
speare's Merchant 0/ Venice (IV.i.305-12, 324-25): a pound of flesh 
but not one drop of blood. As a genre of anecdotes in classical litera­
ture, sophistic interpretations have never been systematically col­
lected or studied from the perspective of international relations.9 

Naturally the interpretation of international agreements and the de­
sire to avoid ambiguity were of the utmost importance in antiquity, as 
they are today, 10 and it should not surprise if Greek treaties reflect 
this concern. 

While a Roman interest in prevention of sophistic interpretation of 
treaties has been noted,l1 the Greek evidence has not been discussed 

7 The Greek vocabulary of stratagems has received no detailed discussion from the 
standpoint of military affairs or international relations. As an introduction see Luitgard 
Camerer, Praktische Klugheit bei Herodot: Untersuchungen zu den Begri/Jen J.LTJx.avT" 
TEX.VTJ, (ToctXTJ (Tiibingen 1965). For the Latin side a survey is found in Antonio Carca­
terra, Dolus Bonus/Dolus Malus: Esegesi di D. 4.3.1.2-3 (Naples 1970) 34-54. 

8 See 259 infra. Of course an interpretation and an oath are not the same thing; I 
have coined the phrase 'sophistic interpretation' as a blend of ancient and modern 
expressions for this phenomenon, which in the tradition of modern international law is 
called an 'interpretation'. Cf Oppenheim (n.l 0 infra). 

9 Hirzel's concern (41-52) is not international relations, and Plescia (86) is even 
more brief and cursory. Also see K. Latte, RE 15 (1931) 348 S.v. "Meineid." 

10 Oppenheim (supra n.5) 858-62 lists fifteen rules for how to avoid sophistic inter­
pretations. Cf U.S. Department of the Army, The Law of Land Warfare (Washington 
1956, rev. 1976) 170, where prevention of sophistic interpretation in the terms of 
capitulation is strongly implied. 

11 Friedrich Lammert, RE Suppl. 6 (1935) 1356 s.v. "Kriegsrecht." 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT L. WHEELER 255 

in any detail. Thus Siewart (33-34) has connected anti-deceit clauses 
with deterring circumvention of oaths, but without seeing the rela­
tionship to sophistic interpretation and its genre of anecdotes. It is 
this relationship which this paper seeks to establish more clearly: the 
kind of interpretation illustrated by these anecdotes is precisely what 
in most cases an anti-deceit clause seeks to prevent, and comparison 
of Greek anti-deceit clauses with Roman formulaic practice will dem­
onstrate that the Greek clauses are not simply formulaic. Sophistic in­
terpretations of agreements were perceived as a real and present 
danger, as we shall see demonstrated in the collected exempla, in 
Cicero's condemnation of such interpretations, and in their theoreti­
cal discussion in the Anonymous Byzantine's Rhetorica militaris. 

I 

The investigation of anti-deceit clauses in Greek treaties will be 
limited in this section to the period from the seventh through the 
third century B.C., for which the texts are conveniently collected in 
the Staatsvertriige des Altertums series. Although most are not pre­
served on stone (compare the first two columns in the table below), 
only epigraphically attested treaties involving a Greek party will be 
considered here, since those preserved in literary sources may reflect 
textual and stylistic contamination, and with few exceptions they lack 
anti-deceit clauses. The statistical evidence is best presented in tabu­
lar form. 

Treaties Anti-deceit clauses 
Century Treaties on stone in treaties on stone 
VII 6 0 0 
VI 11 3 1 
V 91 33 8 
IV 170 64 6 
III 110 66 8 

Total 388 166 23 

The results initially appear inconclusive: of the 166 treaties on 
stone only twenty-three (13.9%) contain anti-deceit clauses.12 The 

12 These are Staatsvertr. 120 (Sybaris and the Serdaei, before 510 B.C.), 154 (Athens 
and Eretria, 446/5), 155 (Athens and Chaicis, 446/5), 162 (Athens and Rhegium, 
433/2), 163 (Athens and Leontini, 433/2), 184 (Athens and Halieis, 424/3), 186 
(Athens, Perdiccas, and Arrhabaeus, 423/2), 187 (Athens and the Bottiaeans, 422), 
204 (Thasos and Neapolis, after 411), 260 (Mausolus and Phaselis, 377-352), 289 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

256 SOPHISTIC INTERPRETATIONS AND TREATIES 

meager evidence from the sixth century would imply a rate of 33.3% 
of the treaties, while the subsequent centuries show a decrease in the 
percentages as the volume of evidence increases (24.2, 9.4, and 
12.1 % for the fifth, fourth, and third centuries). Use of anti-deceit 
clauses is clearly the exception, not the rule. This is puzzling, espe­
cially when viewed against a background, from the fifth century on, 
of increasing awareness of stratagems in historians and military writ­
ers, and of their simultaneous rise to prominence in Greek warfare.13 

Do these figures show that the official instruments of Greek ius 
gentium have little concern for this problem? 

Our epigraphical evidence is admittedly fragmentary, and it is im­
possible to know the extent to which chance of preservation has 
affected these statistics. Moreover, from the standpoint of power 
politics not all extant treaties have equal value: some involve major 
powers of the Greek world and others small or insignificant po/eis. 
Yet given the volume of evidence from the fifth century on, we 
should probably trust the percentages revealed by the stones. Inas­
much as anti-deceit clauses appear so infrequently, their inclusion 
could be significant evidence of mistrust and suspicion between the 
contracting parties. But infrequency is not the only reason for think­
ing that such clauses are not rote formulas. 

The evolving language of the anti-deceit clauses is instructive. 
Anti-deceit clauses are found neither in Homeric conventional oaths 
nor in the alleged treaties of the seventh century known from later 
literary sources.14 The earliest clause, a8oA.o~ ('without trick'), first 
appears in the alliance of Sybaris and the Serdaei concluded before 
510 B.C., and a8oAo~ or one of its variants, usually the adverb a&)­
Aw<;, continues into the third century B.C. as the most frequent form 
of the clause, occurring in thirteen of the twenty-three epigraphical 
examples.15 The second most frequent form, Oli8E Texvn ov8E I-''TIXa­
vii ('and not with craft nor with stratagem'), first occurs in the mid 

(Athens and Ceos, 362), 297 (Arcadian Orchomenos and Euaemon, 360-350), 308 
(Philip and the Chalcidians, 357/6), 309 (Athens and Cetriporis, Lyppeius, and Gra­
bus, 356), 403 (Corinthian League, 338/7), 463 (Aetolians and Boeotians, 292 [?)), 
481 (Eumenes I and his troops, 263-241), 492 (Smyrna and Magnesia by Sipylos, after 
243), 551 (Rhodes and Hierapytna, ca 200), 552 (Rhodes and Otus, ca 200), 578 
(Itanos and Hierapytna [?1, third century), 581 (Praesos and Lyttos), 584 (Cnossos and 
Dreros, ca 200). 

13 I shall treat this topic in detail elsewhere. 
14 Phillipson I 385-87; Staatsvertr. 101-06. On the concept of perjury in Homer and 

Hesiod see Plescia 83-85. 
15 Staatsverlr. 120, 162, 163, 184, 186, 187, 260, 308, 309, 463, 551, 578, 581. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETI L. WHEELER 257 

fifth century, and is found in nine of the twenty-three examples.I6 

Although some variation of these two phrases is found by the late 
fifth centuryp the introduction of new phrases and much greater 
variations begin ca 360 B.C.: at/JEv8.,Jwv ('may I be true', i.e. not 
lie or deceive), a8oAw~ Kat aTEXv~ ('without trickery and without 
craft'), ov8e KaKOTEXV1}U"W 1Tf.pt TOV OPKOV TOVTOV ov8iv OVTE T€Xvn 

oihE 1TapEVpEUEt, OV8EJ.L~ ('and I shall devise nothing fraudulent 
about this oath, neither by craft nor by any false pretext'), and a80-
Aw~ Kat a1Tpocfxx.uia-Tw~ ('without trick and without evasion').I8 

Two trends can thus be discerned in the statistics: decreased inci­
dence of anti-deceit clauses over time, as the number of extant trea­
ties and the concern for trickery in international affairs increase, and 
the appearance ca 360 B.C. and later of new phrases or greater varia­
tions in the language. The first trend becomes even more obvious if 
the four centuries are grouped in pairs: anti-deceit clauses occur in 
28.75% of the treaties of the sixth and fifth centuries combined (or 
24.2% for the fifth century alone) by comparison with only 10.8% for 
the fourth and third centuries combined. 

In the light of these figures (especially if we exclude the one item 
from the sixth century), we may examine the origins of the treaties. 
All the fifth-century evidence for anti-deceit clauses is Athenian and 
concerns either the Athenian empire or one of its former members, 
whereas only two of the fourteen examples from the fourth and third 
centuries occur in Athenian treaties. Variations in the language after 
360 could thus reflect the greater diversity of the contracting parties 
(i.e. most are not Athenian) and also perhaps the natural evolution 
of the Greek language in later centuries. On the other hand, it seems 

16 Staatsvertr. 154, 155, 204, 289, 308, 403, 481, 492, 584. The oath of the heliasts 
preserved in Oem. 24.149-50 and allegedly of Solonian date contains the phrase om-f: 

Texvn oun /-LT)Xavji, which as we have seen is not epigraphically attested in treaties 
before the fifth century. From this point of view the Solonian date of the oath, at least 
in the form given in Demosthenes, seems in doubt. For literature on attempts to 
reconstruct the oath of the heliasts see Plescia 26 n.30. 

17 Thuc. 5.18.3 (Peace of Nicias in 420, nt~ (T1TOVoa~ ... a&lAov~ Kat a{3h.a{k/8 Kat 
Kant y7JV Kat Kant 8aAa(T(TaV, cf. 5.18.9; 5.47.8 (alliance of Athens, Argos, Mantinea, 
and Elis in 420), E/-L/-Lf:VW rn {v/-L,."ax0 Kant nt MKf:i/-Lf:va 8LKaLw<; Kat a{3h.a/3W<; Kat 
a&lA~, Kat ou 1Tapa{3~(To,."aL Texvn ov8e /-LT)Xavji OV/-Lf:8u!, cf. 5.23.2 and Staatsvertr. 
260. The phrase aVEv TE &lAOV Kat a1TaTT)~ in proposals of alliance occurs three times 
in Herodotus 0.69.2, 8.140a.4, 9.7a}; Bickerman 96 n.48 claims Herodotus' phrase is a 
standard formula of oriental diplomacy, but cites no evidence except Herodotus. Anti­
deceit clauses are not found in oriental treaties: see Dennis J. McCarthy, Treaty and 
Covenant: A Study in Form in the Ancient Oriental Documents and in the Old Testament 
(AnaiBibl21 [l963}). 

18 Staatsvertr. 297, 308,481, 551. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

258 SOPHISTIC INTERPRETATIONS AND TREATIES 

fair to add that the burdens of empire and also the interests of the 
sophists have produced in fifth-century Athenian treaties (of which 
all containing anti-deceit clauses date after 450) a clear concern about 
stratagems, circumvention, and trickery. 

Still another factor that could affect the frequency of anti-deceit 
clauses is the degree of faith in the religious bond of the oath. It has 
been thought that from the fifth century on the confidence that the 
gods punish perjury and oath-breaking, and belief in the very exist­
ence of the gods, waned under the influence of the individualism and 
rationalism of the sophists; oaths in general, and not just in treaties, 
lost their previous authority.19 Pritchett and Lonis, however, argue 
that the sophistic movement did not abolish faith in religious prac­
tices, which continued to exert strong influence on military and 
international affairs.20 Lonis nonetheless concedes that by the fourth 
century oaths lost their value as guarantees of international pacts and 
that the use of hostages did not prove a better means to insure fi­
delity to agreements.21 The depreciated religious force of the oath is 
further seen in the Hellenistic philosophical schools: both the Epicu­
reans and the Stoics sought to replace the religious sanction with an 
ethical regard for justice and bona fides itself. 22 

Thus it could be argued that if the religious value of the oath had 
become nil, insertion of an anti-deceit clause would be futile verbiage 
in a legal/religious mechanism that was already a dead letter: hence 
the infrequency of Greek anti-deceit clauses. Oaths, however, wheth­
er as a true religious sanction or as a conventional element, con­
tinued to be an essential component in Greek treaties; and whatever 
the cause of their increasing diversity of formulation, anti-deceit 
clauses likewise continued to be used. These clauses are framed 
not to prohibit perjury (which was thought punishable by the gods) 
but rather the strictly legitimate if ethically dubious circumvention 
through interpretation; so the explanation of waning religious faith in 

19 Plescia 86-87, cf Hirzel 79-104. 
20 W. Kendrick Pritchett, The Greek State at War (Berkeley 1971-1979) III esp. 2-10; 

Raoul Lonis, Les usages de la guerre entre grecs et barbares (Paris 1969) 130 and Guerre 
et religion en Grece a I'epoque c1assique (Paris 1979) 50, 55-57. 

21 Lonis, Les usages (supra n.20) 108-11, 150, 133, 139; "La valeur du serment dans 
les accords internationaux en Grece c1assique," DHA 6 (I980) 267-86; "Les otages 
dans les relations internationales en Grece c1assique," Melanges offerts a Leopold SMar 
Senghor (Dakar 1977) 215-34. 

22 Cic. Fam. 7.12, cf Plut. Mor. l1l2e; Cic. Off. 3.102, 104. Cf Curt. 7.8.29: iuran­
do gratiam Scythas sancire ne credideris; colendo fidem iurant. Graecorum ista cautio est, qui 
pacta consignant et deos invocant; nos religionem in ipsa fide ponimus. qui non reverentur 
homines, fallunt deos (a Stoic sentiment placed in the mouths of Scythians). 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT L. WHEELER 259 

oaths cannot be wholly satisfactory. The infrequency of anti-deceit 
clauses thus remains problematic. 

II 

It is widely agreed that the function of anti-deceit clauses is to 
deter circumvention of the oath,23 and we have hypothesized that 
circumvention should be identified with sophistic interpretations. Be­
fore examining the epigraphical evidence linking anti-deceit clauses to 
this type of circumvention, we should consider ancient definitions 
and examples of sophistic interpretations. 

Eustathius (ad Od. 19.396 [1870-71]) offers the most detailed 
definition of sophistic interpretation, or m Greek terms a OPKO~ 
UOcPtUTLKO~ or UOcPtUI.ux:rw8..,,~:24 

OPKO~ 8E VVV OVXt <> KaT' i.'1TWPKLaV· TOVTO yap oihE Eu9Aov av-
8po~, OVTE KOUJ.'EL av9pW7f'ov, Kat ov8E 9EO~ 8i&Uutv' aua <> EV 

~ ,1.,.1" ., ~,,~, ", ,~,', Katpcp (TOopt."oJ,J.,EVOC;, OTE 0Tll\.aoTl aVTOC; J,J.,EV TtC; EVOPKWV E(TTL, TOUt; 
8E E'1TtKA.07T~ aVTc!J EYKaAoVvTaC; 7TapaAoyi{ETat ola JLT, EXOVTaC; 
VOEtV TO UOCPWp,a. OVTW 7T~ Kat '08VU(TE~ avo'1Ttv E(TocpiuaTO TT,V 
llTlVEA.07T"f1V OPKCP EV Tc!J ... TOWVTOC; 8E Tte; OPKOC; 7Tpoe; aAAote; 
JLvpWte;, T1YOVV UOCPtUTtKO~, Kat <> T(;JV <>p,ouaVTWV Toua~ T,p,Epae; 
ayEtV (T7Tov8ae;, EV 8E Tc!J p,E(TCP VVKTWP E'1Tt9Ep,EVWV TOLe; Ex9poLe;, 
w.; 8ij9Ev Tij~ uvlLf3a(TE~ TWV (T7TOVSWV T,J,J.,EpWV J.'ElLvTlILEVTle;, ov 

\ , ,.., ' '''' ~ A.. '~ f' JLTlV KaL VVKTWV ... KaL OVTOL J,J.,EV OPKOL UOopL(Tp,aTWoEt~ 07TOWLe; 
,I.,.~' , ,~, ~ '!"-'-tVETaL XaLpEtV Kat LoLUV'fA'C;. 

In Eustathius' view the swearer of a sophistic oath does not commit 
perjury or break the oath but remains EVOPKO~, as other sources also 
attest.25 The outrage of the deceived party is quite apparent in Eu­
stathius and the scholiast, but Polybius, who identifies Antiochus 
IV's sophistic interpretation of a truce in the capture of Pelusium as a 

23 Lammert (supra n.l1) (Roman evidence) and Siewart 33-34. 
24 The Romans never developed a suitable Latin expression for sophistic interpreta­

tion and remained content with periphrastic phrases: Cic. Off 1.33, calumnia quadam et 
nimis callida. sed malitiosa iuris interpretationes; Val. Max. 7.3.4, improbo ... praestigiarum 
genere; Tac. Hist. 4.41.2, trepidis et verba iuris iurandi per varias arIes mUlandibus; Gell. 
6.18.6, 10, commenticio consilio . .. Jraudulenta callidas. Cf Hirzel 47 n.1. 

25 Schol. Od. 19.396, OVK E1TWpK{;JV, aMa aocfx(,oJ,LElJo<; 70v., OPKOV<;, aV70<; ~IJ EV­
OPKO<; SLEJLELIIE, TOV<; SE E'YKaAOvIlTa<; StU TIJ<; E7TLIIOI.a<; €7TEL(JEII; Ae\. VH 12.8, Af:YWV 
I-L7J 7Tapaa7TOIISEI,II I-L7/SE E1TWPKEI,II; Polyaen. 6.53, <> SE ovSEII EcfrrI aStKEI,II. In Polybius 
and Pausanias sophistic interpretation is expressed succinctly by aocPi~Ea(Jat and ao</xa­
/-LU: Polyb. 6.58.12, TOil SE ao</xaaJLEllolJ 7TPO<; TO Avaat TOil 0PKOII; Paus. 10.30.2, 70V 
E7T!, 741 0PKq> J,LETaaxolJTa aOrPW/-LU70<;; cf Hirzel 44 n.2. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

260 SOPHISTIC INTERPRETATIONS AND TREATIES 

stratagem, stresses that, like sycophants in the law courts, Antiochus 
observed the letter of the law although without regard for justice.26 

aequitas and iustitia aside, none of the sources defining sophistic in­
terpretation question its legitimacy from a strictly legal point of view. 
The sophisma of a sophistic interpretation lies in the ambiguity of 
what was sworn, the al-'4xfJo)..la or Of.UJJVV~, as the Anonymous 
Byzantine puts it.27 Indeed Camerer's definition of a sophistic trick 
states the basic principle of sophistic interpretation: "Es handelt sich 
urn die Fahigkeit, den Partner dadurch zu tibertOlpeln und einzu­
schtichtern, dass man eine Vereinbarung trifft, bei der irgendein un­
auffalliger, unbedeutender Umstand vom Partner unbeachtet bleibt, 
aber spater gegen ihm verwendet wird. "28 

Since the oath lends a religious sanction to an agreement, we 
should note Greek religious lore about the chicanery of sophistic in­
terpretations. Homer's gods often practise deception, but subsequent 
generations honored more moral divinities. Yet even Zeus, the god 
of justice, had the epithet E'7T(,K)..07TELO~, probably derived from his 
deceptions of Hera and his extramarital affairs.29 Thus it is no sur­
prise that his son Hermes should have a great reputation for trickery 
and deceiPo In fact Hermes might be said to have sworn the first so­
phistic oath, when as an infant he stole Apollo's cattle, hid them in a 
cave, and craftily swore to Zeus that he had not driven them home. 

26 Polyb. 28.18, with F. W. Walbank, A Historical Commentary on Polybius (Oxford 
1957-1969) III 352. Diod. 30.18.1-2: 1TPO<TTjKOII'TW'; av TL~ a1Tocfr{,vaLTo TOV 'AVTWXOV, 
Ka(Ja1TEp TO~ am> TWV &Kau1TiPl.wv UVKocfxlVTa~, TO JUV P1JTOV TOV VOJ.£l>V TE1TiP1J­
KEvaL p.ETa T£l~ aVOKa~ r7JV KaTa),:rJlPLV 1TE1TOt1J"uVOV, TO "uVTOL yE BiKawv Kat KaAOV, 
6t' WV 0 1TaVTWV UVVEXETat {3io~, JLT, TET1Jp1JKEvat. &a 'Yap r7JV crvyyEVELaV &/>EiAwV 
tPEi&u9aL TOV p.EIfXXKWV, Ka9a1TEp aVTo~ ~cfnlUE, TovvavTwv E~a1TarTwa~ E1TEJ30:AETO 
n),8 OAOt~ uc/YiiAm TOV 1TE1TLUTEVKOTa. Cf Val. Max. 7.3.4: ceterum etsi circum venti No­
lani ac Neapolitani queri nihil potuerunt secundum ipsorum demonstrationem dicta sententia. 
See n.65 infra for Antiochus' stratagem as a sophistic interpretation. 

27 Anon. Byz. Rhet.mil. 2.2. On this source see n.37 infra. 
28 Camerer (supra n.7) 67. Since a sophistic interpretation is not perjury, we need not 

enter into the Peripatetic and Stoic controversies on the definition of perjury. On these 
debates see Plescia 84-85, Hirzel 75-79, and the general discussion of perjury in Latte 
(supra n.9) 346-57. Likewise the verbalvoluntas problem, seen in Euripides' famous 
"My tongue swore, not my mind" (Hipp. 612), has little bearing on our discussion. 
Even in Roman law this issue is post-classical: Erich Hans-Kaden, "Die Lehre von 
Vertragsschluss im klassischen romischen Recht und die Rechtsregel: Non videntur qui 
errant consentire," in Festschrift Paul Koschaker I (Weimar 1939) 334-57; Fritz Schultz, 
Geschichte der romischen Rechtswissenschajt (Weimar 1961) 155-57. 

29 See Karl Deichgraber, Der Iistensinnende Trug des Gottes (Gottingen 1952); Hesych. 
s. v. E1TLKA01TEW~; Hirzel 42 n.5. 

30 Deichgraber (supra n.29) 109-13; Chr. Sherer, Roscher Lex. I 2371 s. v. "Her­
mes"; cf Marcel Detienne and Jean-Pierre Vernant, Cunning Intelligence in Greek 
Culture and SOCiety, tr. J. Lloyd (Atlantic Heights 1978) 23 n.3; S. Eitrem, RE 8 (I922) 
755, 780-81 s. v. "Hermes 1." 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT 1. WHEELER 261 

It is significant that Zeus felt no anger but only amusement at his 
son's cunning.31 

Hermes passed on his sophistic skill at oaths to Autolycus, Odys­
seus' maternal grandfather, who according to Homer (Od. 19.395f) 
Cx.1I8pdmolJf; EKaKaCT'TO KAE7TTOuVV[I 8' OPKCP TE and whom later sources 
call a son of Hermes. His fame as a trickster eventually passed into the 
stratagematic tradition, although in our evidence it is only Eustathius 
who first clearly associates his use of oaths with sophistic interpreta­
tion.32 Much the same can be said about Sisyphus, whose trickery has 
Homeric roots and a rich development in later authors, but whom 
Eustathius again first directly connects with sophistic oaths, even as­
serting that he could not be outwitted by Autolycus.33 

Eustathius' favorable view of sophistic interpretations thus belongs 
to a strong Greek tradition. They were legitimate (i.e. neither perjury 
nor oath-breaking)~ a good man (ECT8AOS') could use them~ and it was 
a skill not only given by a god but approved of with humor by Zeus. 
Taking advantage of another's oversight or gullability in this way did 
not provoke divine wrath.34 

bona fides, however, which does not have its roots exclusively in 
legality or religion, forms the basis of any viable system of interna­
tional relations. Dissenting voices to the view seen in Eustathius had 
to shape their reproaches on moral and ethical grounds.35 The Stoic 
condemnation of sophistic interpretations is summarized in Cicero: 

exsistunt etiam saepe iniuriae calumnia quadam et nimis callida, sed 
malitiosa iuris interpretatione. ex quo illud summum ius summa iniuria 
factum est iam trilUm sermone pro verbum. quo in genere etiam in re 

31 Hymn. Hom. Mere. 62-495: the sophistic phrase at 379 and Zeus' amusement at 
389-90. 

32 PI. Resp. 334A-B~ Polyaen. 1 praef. 6; Eust. 1870-71. 
33 I/. 6.l53~ AT. Aeh. 391 and schol.; Polyaen. 1 praef. 5; Phrynichus Arabicus in 

A need. Bekk. 164; Eust. 1870-71; Wilsch, Roscher Lex. IV 958-72 S.v. "Sisyphos." For 
Autolycus' attempted deception of Sisyphus see Suda s. v. I.iuvct>o<;. 

34 Hirzel 43, 45. 
35 PI. Resp. 334A-B; schol. Ar. Ach. 391; Diod. 30.l8.l-2 (Polybius); Phrynichus 

(supra n.33). Plato's approval of Autolycus is ironic, but he emphasizes Autolycus' 
thievery. The scholiast and Phrynichus characterize Sisyphus with the pejorative rrO:II­
ovp-yo<;, although it cannot be argued that this is aimed specifically at Sisyphus' use of 
sophistic oaths. Polybius' objection based on justice is essentially the same as that 
raised by the Stoics (see infra n.36). Does Eustathius express his own view, or reflect 
earlier sources? His approval of sophistic interpretations could reflect the rather liberal 
Byzantine view of stratagems seen in the Anon. Byz. De re strategiea, Maurice, Leo, 
and other military texts: ef. ad I/. 15.541 (1030), &lAo<; ov IjIEK'TO<; ow (J"Tpa'TLW'TtKo<; 
(his defense of Teucer killing Cleitor with a shot in the back). On stratagems in Byzan­
tine military thought see most recently Walter E. Kaegi, Jr, Some Thoughts on Byzantine 
Military Strategy (Brookline 1983). 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

262 SOPHISTIC INTERPRETATIONS AND TREATIES 

publica multa peccantur . . . quocirca in omni est re jugienda ta/is solJer­
tia. 36 

Stoics did not have the last word on the subject in antiquity, and 
the pendulum eventually swung in the opposite direction. An un­
known Greek, probably in the age of Justinian, composed a treatise 
on statesmanship, of which the practical military side is discussed in 
the De re strategica. In a companion treatise on the verbal aspects of 
statesmanship, the less well known Rhetorica militaris, the author 
encourages the use of sophistic interpretation to gain advantage over 
foreign nations (2.2-3.1): both written and unwritten public speeches 
in domestic and military affairs demand clarity, as do letters to sub­
jects. But ambiguity and equivocation present the opportunity for 
profit when writing to the rulers of foreign nations, and the Anony­
mous Byzantine supports his argument with two examples, one de­
fensive and one offensive.37 Eustathius' favorable view of sophistic 
oaths had a theoretical as well as many practical precedents. 

The catalogue of forty-five sophistic interpretations given in the 
Appendix is intended to illustrate the extent of this practice in an­
tiquity, with no claim of comprehensiveness. Most examples are 
derived from historical and other prose sources, as these rather than 
the deceptions of tragedy, comedy, and romance seem more relevant 
to an assessment of sophistic oaths in treaties. The historicity of all 
the examples is not guaranteed; certainly some are the inventions of 
propaganda or dubious traditions, while others may be folktales and 
proverbs. For our purposes the significance of these forty-five exam­
ples lies in their distribution in time from Homer to Procopius and in 
the pervasiveness and awareness of sophistic interpretation in the 
ancient psyche that they reflect. The contrasting views of the Stoics 
and the Anonymous Byzantine represent a real debate on a practical 
issue. 

In these examples, who perpetrates the sophistic oaths, and how 
many concern chicanery in military and international affairs? Bar­
barians are seen to swear sophistically in nine of the forty-five cases, 

;16 Off. 1.33, cf 3.113-15 and Caecin. 51-52. Whether Off. 1.33 and 3.113-15 are 
Cicero's own opinion or derive from Panaetius or Poseidonius directly or from Panae­
tius through Poseidonius is immaterial here. For a recent assessment of the problem 
see Horst-Theodor Johann, Gerechtigkeit und Nutzen: Studien zur ciceronischen und hel­
lenistischen Naturrechts- und Staatslehre (Heidelberg 1981). 

37 See Appendix nos. 9 and 21. On the Anonymous Byzantine see H. Kochly and W. 
Rlistow, Griechische Kriegsschriftsteller 11.2 (Leipzig 1855) 2-40, who give the only 
printed edition of the Rhetorica militaris. Their grouping of the Rhetorica with the De re 
strategica goes back to Lucas Holsten and is accepted in Alphonse Dain, "Les strate­
gistes byzantins," Travaux et memo ires 2 (1967) 343-44. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT L. WHEELER 263 

Romans also in nine, and Greeks in twenty-seven.38 Only eleven of 
the forty-five examples do not concern military or international affairs 
in some respect.39 This sampling of the evidence thus yields that 
Greeks are three times as likely as Romans or barbarians (60% ver­
sus 20%) to perpetrate sophistic interpretations; and that three of 
every four instances of sophistic interpretation involve military or 
international affairs. 

III 

We have seen that anti-deceit clauses use the language of strata­
gems and that the two most frequent forms are aSoAo~ and ovSe 
TeXVn ovBe I-''TIXavfi. If these clauses often appeared as isolated ele­
ments, i.e. seemed to be inserted routinely in the oaths, this would 
support the proposition that they are rote formulas. In fact the op­
posite is the case. aSoAo~ and its variants occur in only thirteen of the 
166 treaties on stone and in isolation only four times. It is most 
frequently coupled with a reinforcing element, such as a form of 
?TL(TTO~ (four examples) or with af3Aaf3ew~ (three examples). Sim­
ilarly, ovSe Texvn ovSe I-''TIXavfj is found in only nine of the 166 trea­
ties and in isolation only once.40 

The aim of an anti-deceit clause is of course to close all avenues to 
circumvention, and we have seen that these clauses exhibit a ten­
dency toward greater elaboration and variety. Such supplements, often 
addressing diverse contingencies and even specific acts, are first at­
tested in Athenian treaties of the fifth century and reflect Athenian 
fear of revolt among its allies. It was no longer sufficient to swear ovBe 
Texvn ovBe l-''TIxavfj; one also had to renounce hostile action by word 
or deed, ovB' E?TEt ovBe EpYWt.41 During the Peloponnesian War we 
find for the first time the specification that oaths cover operations 
both on land and on sea, KaTlX yijv Kat KaTa 8a'Aaa-a-av, as in the 
Peace of Nicias and first epigraphically recorded in the League of 
Corinth. By the third century a stipulation of both night and day (W'.,TE 
EV VVKTt I-'T,TE ?TES' al-'epav) first appears, which matches precisely a 
species of sophistic interpretation found four or five times in the cata-

38 Barbarians: nos. 4, 7, 9, 23, 27, 33, 36, 38-39; Romans: 1, 3, 6, 17, 26, 29, 30, 40, 
44; Greeks: 2,5,8,10-16, 18-22,24-25,28,31-32,34-35,37,41-43,45. 

39 Nos. 1,3, 12, 19-20,22,35,38,40-41,45. 
40 a8oAo~ in isolation: Staatsvertr. 184, 309, 463, 578; with 1TLO'Tck 120, 162, 184 

(restored), 187; with a{3Aa{3EWC;: 163, 186, 260; ov8e TEX"7I KTA. in isolation: 289. 
41 Staatsvertr. 154, 155, cf 204. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

264 SOPHISTIC INTERPRETATIONS AND TREATIES 

logue.42 We should note that Philip II, who achieved so much by chica­
nery, had it sworn in the treaty of the League of Corinth (if the various 
restorations are correct) that no member of the League would capture 
a city, fort, or harbor of another member by craft or stratagem.43 

Certainly a80Ao~, TEXlIT1, and J.LT1xall..q are general terms, but the 
vocabulary of some treaties of the fourth and third centuries points 
more explicitly to sophistic interpretation. Words such as TrapEVpE­
UtS ('pretext') , aTrpocpaUlnTW<; ('without evasion'), and KaKoTEXlIEw 
('deal fraudulently'), although by no means new in the Greek lan­
guage, are new variations in anti-deceit clauses; in this application 
they cannot be traditional formulas, mere regional preferences in 
language, or even hollow routine verbiage without a more immediate 
and relevant purpose.44 

If we may cross briefly the lower chronological terminus set for this 
study and consider the second century, we find in Crete a good 
example of the close connection between anti-deceit clauses in trea­
ties and stratagems in international relations. Cretan treaties are rela­
tively numerous in the late third and second centuries. Indeed four of 
our last five epigraphical examples of anti-deceit clauses in the third 
century come from Crete, and this tendency continues in the second: 
without claiming an exhaustive search, I find that of six treaties 
employing anti-deceit clauses four are from Crete.45 A new variant 
form of anti-deceit clause appears in all four Cretan texts, ciTrAOW<; 
Ka8oAw~ ('openly and without trickery'). This is reminiscent of an 
Athenian treaty of the fifth century, unfortunately mostly restored 
and thus not secure evidence: 1TL]UTll Kat. a80Aa Kat. h[a1TMl (Staats­
vertr. 162.11f). Furthermore, Hierapytna's treaty of isopoliteia with an 
unknown city contains one of the most extensive anti-deceit clauses 
attested (I.Cret. III iii 5): EJ.LJ.LElIW Ell TOtS avyKELJ.LElIOL~ Kai ov KaKO­
TEXlIT1UW OV(JElI TWlI Ell TlIL8E TelL W01TOALTEUU yeypaJ.LJ.LElIwlI OVTE 

, "., • ~ , "\ \ I.>' ,r,"" , , AOYWL OVTE EpyWL, OVuE al\.l\.WL E1TLTpao/w EKWlI KaL YLlIWUKWlI Trap-
EvpeuEL OV8Ep,WL OV8E TP0TrWL OV(JElIl.. 

This flourishing of anti-deceit clauses in Cretan documents of the 
late third and second centuries is illuminated by the historical back-

42 By land and by sea: Thuc. 5.l8.3, Staatsvertr. 403, I. Cret. IV 186B (Gortyn and 
Lappa, II B.C,). Night and day: Staatsvertr. 584, cf Appendix nos. 7-10 and 31. 

43 Staatsvertr. 403; cf 309.40-43, where the subsequent clause (ov npOKaTaAVaWUXL 
'TOV nUAEIJ.OV av[w KE'TPL1TOpW~ K'TAJ can be seen as a partial explanation of a&)~. 

44 Staatsvertr. 481,551, 552, cf I.Cret. IV 186B. 
45 Third century: Staatsvertr. 551, 578, 581, 584. Second: I.Cret. IV 174 (Gortyn, 

Hierapytna, and Priansos), IV 186B (Gortyn and Lappa), III iii 3B (Hierapytna and 
Lyttos), III iii 5 (Hierapytna and an unknown city), Milet III 149 (Miletus and Pidasa); 
Syll.3 633 (Miletus and Heraclea). 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT L. WHEELER 265 

ground known from literary sources. Cretans, notorious as pirates in 
this period, were seen not only as treacherous liars, but also as fre­
quent users of stratagems and sophistic interpretations.46 Polybius' 
boyhood hero Philopoemen campaigned in Crete during 200-192, the 
very period from which at least some of these Cretan anti-deceit 
clauses derive, and Plutarch's description of Philopoemen's activities 
probably has its source in Polybius' three-book encomium of the 
Achaean genera1:47 

EV 8E 'rfi Kpr1711 tTlJVE7TOAEj.UL 'TOIS ropTVvioL~, ovx ~ nEA01TOV­
V7}(TLoS' CWTIP Kat 'ApKas a1TAOVV nva Kat 'YEJlJlaWJI 1TOAEJ.WJI, aAAa 
TO KP'YITLKOV i,80') ev8v') Kat TOL') EKElvwv UocpWJUX.(rL Kat MAOL') 
KAw7rEiaL~ TE Kat AOXLU/-LOIS xpWj.UVO~ E1T' aVTOV~ TaXV 1TaL8a~ 
a1T€8Et4EV aVO'YITa Kat KEva 'TTpOf) ep,'TTELpiav aA'YI8LV~V 'TTaVOVp-

~ 

,),OVVTa,). 

The Cretan anti-deceit clauses and the contemporary style of warfare 
on the island fit together like hand in glove. 

A document of the third century, although preserved in a literary 
source, also supports our thesis about anti-deceit clauses. Hannibal's 
oath confirming his alliance with Philip V of Macedon in 215 B.C. is 
found in Polybius 7.9. As Bickerman demonstrated, Polybius' text is 
a Greek translation of a Punic document produced in Hannibal's 
chancellery, and the problems of the Carthaginian translator in ren­
dering the Punic into suitable legal Greek are revealing.48 The anti­
deceit clause reads (7.9.8): 

, , n' ,~\ ~, , 8 " ,~~,~ \ OVK E'TTLpOVAEVUOp,EV OVoE 1\.0x~ XP'YIuop,E a E'TT al\.l\.'YII\.OL')· p,ETa 
1T&:ua~ 8E 1TpO(JVp,ia~ Kat EVJlOia~ aJlEV MAOV Kat E1TLf30VAij~ EUO­
j.U8a 'TTOA€p,LOL TOL~ 'TTpOf) Kapx'YIoovlov.:; 'TTOAE/-LOVUL ... 

Instead of the two most common forms of anti-deceit clauses, ov8i 
T€XVV ov8£ /-L'Y/xavfj and a80Aol), the Carthaginian translator pre­
sents a quite literal rendering of the idea expressed by the former, 
specifically referring to plotting against each other and the use of 
ambush, and approximates the latter with avev 80AOV, a phrase with­
out parallel in Greek treaties.49 The poetic AOX0f) had been replaced 
by Ev€8pa in Thucydides, which became the standard prose word for 
'ambush' thereafter, although here the translator probably took AO-

46 See Appendix no. 5; Callim. Jov. 8; Polyb. 4.8.10-11, 6.47.5, 8.19.5, 21.5; Plut. 
Lys. 20.2, Aem. 23.6. For Cretan piracy see most recently P. Brule, La piraterie cretoise 
hellenistique (Paris 1978). 

47 Plut. Phil. 13.6, cf Polyb. 10.21.5-6. 
48 See Bickerman, and his "Hannibal's Covenant," AlP 73 (952) 1-23. Cf. Wal­

bank (supra n.26) II 42-56. 
49 Bickerman 96 with n.48. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

266 SOPHISTIC INTERPRETATIONS AND TREATIES 

xo~ to mean trickery in general as well as ambush, just as insidiae in 
Latin.50 

Explicit references to plotting against the other party do not occur 
in Greek treaties on stone except as ex post facto promises not to 
repeat such behavior when it has already occurred, and this is really 
the language of civic loyalty oaths rather than international conven­
tions.51 The point is best illustrated in the agreement between Eume­
nes I and his mercenaries after their mutiny, which includes a prom­
ise not to plot against the king (Staatsvertr. 481.31), but an anti-deceit 
clause for the oath separately (45f). Of course deterrence of decep­
tion, i.e. secret plotting against the other party, is the purpose of an 
anti-deceit clause, and the overly literal renderings of the Carthagin­
ian translator with their explicit mention of plotting and trickery lay 
bare the essential nature of anti-deceit clauses, which the more subtle 
Greek usage leaves somewhat vague. 

Lacking Punic evidence, we cannot know whether the Carthaginian 
translator was attempting to render a Punic legal formula by a Greek 
anti-deceit clause-which, I have argued, is not a routine feature of 
Greek treaties and reflects a more immediate concern about trickery. 
There is no lack of evidence, however, if we compare Greek and 
Roman usage; and the linguistic connection of Greek 8OA.o~ with 
Latin dolus offers an excellent instance for study. 

Hirzel argued that in antiquity the border between adherence to an 
oath and perjury was more fluid than modern standards permit and 
that the Greeks had greater freedom to interpret oaths than the 
Romans, as Greek oaths lacked formulas such as ex animi sententia 
iurare and si sciens fallo. 52 I would add to this list of Roman formulas 
absent from Greek dolus malus and sine dolo malo, for which 8oA.o~ 
and a8o~ offer linguistic but not, we shall see, legal equivalents. 
That dolus malus is a legal formula is general knowledge and requires 
no demonstration; its frequency is such as even to merit an epi-

50 See Pritchett (supra n.20) II 178. 
51 Bickerman 96 with n.47. Loyalty oaths: Syll.3 360 <rauric Chersonesus, ca 300-

280), I. Cret. III iv 7-8 (Itanos, ca 300), cJ Hesperia 5 (936) 422 (honorific decree 
for Kephisodorus, 196/5, who thwarted plots against Athens). Ex post facto promise: 
IosPE 12 402 (oath of Pharnaces I of Pontus, 179; he had attempted to conquer 
Chersonesus and was checked by Bithynia, Pergamum, and Cappadocia: M. Ros­
tovtzeff, CAH IX (1932) 218; Edouard Will, Histoire politique du monde hellenistique II 
[Nancy 1967) 242-44; a redating of the inscription to 155 B.C. is proposed by S. M. 
Burstein, "The Aftermath of the Peace of Apamea: Rome and the Pontic War," AJAH 
5 (1980) 1-12). 

52 Hirzel 22, 41, 48; Cic. Off. 3.108, Luc. 146; Liv. 22.53.11; additional references in 
Latte (supra n.9) 354. For modern standards of interpretation see Oppenheim (supra 
n.lO). 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT L. WHEELER 267 

graphical abbreviation, d.m.53 dolus belongs to the Roman vocabulary 
of stratagems, just as BOAO~ does to that of the Greeks, but dolo malo 
or sine dolo malo in a Roman document is unquestionably a rote 
formula about malicious intent rather than a pointed and immediate 
guard against stratagem or circumvention. Clear proof is found in the 
Lex Iulia maiestatis (Dig. 48.4.4): 

cuiusque dolo malo iureiurando quis ad actus est, quo adversus rem 
publicam faciat: cuiusve dolo malo exercitus populi Romani in insidias 
deductus hostibusve proditus erit: factumve dolo malo cuius dicitur, quo 
minus hostes in potestatem populi Romani veniant: cuiusve opera dolo 
malo hostes populi Romani commeatu armis tellis equis pecunia aliave 
qua re adiuti erunt: utve ex amicis hostes populi Romani ./iant: cuiusve 
dolo malo factum erit, quo rex exterae nationis populo Romano minus 
obtemperet. 

Certainly these treasonous acts could be seen as stratagems from one 
point of view, but dolo malo characterizes the performance of these 
acts with evil intent and not with regard for the deception of a strata­
gem. 

The Roman clarity of dolo malo as a formula does not exist in 
aBoAo~ or aBoAw~. Romans were of course just as aware of sophistic 
interpretation as the Greeks and desired just as much to avoid it. The 
fetial oath for treaties demonstrates this: ut ilia pa/am prima postrema 
ex illis tabulis cerave recitata sunt sine dolo malo, utique hic hodie rectis­
sime intel/ecta sunt, illis legibus populus Romanus prior non deficiet.54 The 
real anti-deceit clause, however, is not sine dolo malo but utique hie 
hodie reetissime intel/eeta sunt. 

If aBoAo~ or aB6Aw~ were a legal formula in Greek treaties, we 
should expect them in translations of the Latin formula as the most 
obvious legal as well as linguistic equivalent. The epigraphical and 
some literary evidence suggests otherwise. The Romans translate 
their dolus malus and sine dolo malo as B6AO~ Trov'T1p6~ and aVEv 

53 The literature is voluminous. See e.g. Carcaterra (supra n.7); Alfred Pernice, 
Labeo: romisches Privatrecht im ersten lahrhundert der Kaiserzeit 11.1.2 (Halle 1895); 
Kleinfeller, RE 5 (1903) 1292-94 S.v. "Dolus 4"; Adolf Berger, "Dolo malo," En­
cyclopedic Dictionary of Roman Law (T APhS 43.2 [1953]) 440; Andreas Wacke, "Zum 
dolus-Begriff der actio de dolo," RIDA 27 (1980) 349-86. The abbreviation: e.g. ILS 
6085. 

54 Liv. 1.24.7. The archaic date of the oath has been questioned by R. M. Ogilvie, A 
Commentary on Livy, Books 1-5 (Oxford 1965) 109-10. On the significance of formulas 
in early Roman law see Schultz (supra n.28) 28-30. From Liv. 38.11.2 and Hdt. 1.69.2 
(c! n.56 infra and supra n.17), Erich S. Gruen has concluded that Roman use of sine 
dolo malo in treaties is an imitation of Greek anti-deceit clauses: The Hellenistic World 
and the Coming of Rome (Berkeley 1984) I 29 n.85. The point is not argued in detail, 
and the evidence studied here indicates that it is invalid. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

268 SOPHISTIC INTERPRET A TIONS AND TREATIES 

&)AOV 1TOVT/pOV.55 Bickerman (02) holds that ancient chancelleries 
emphasized strict literalness in translation of legal texts and cites 
MAO~ 1TOVT/PO~ as an example, but this explanation does not seem 
convincing when such a close verbal equivalent as aBoAo~ was avail­
able, unless aBoAo~ was not felt as a legal formula. 

Bickerman's point, however, carries no weight for an author, espe­
cially one who knew both Roman practice and Hellenistic Greek very 
well. When Naravas, a Numidian chieftain, enters Hamilcar Barca's 
camp during the mercenaries' revolt to offer his alliance, Polybius re­
cords that the crafty barbarian's proposal was given aBoAw~, a usage 
perfectly consistent with anti-deceit clauses in inscriptions. On the oth­
er hand, when Polybius gives the terms of the Roman-Aetolian treaty 
of 189, he translates sine dolo malo as XWPIS BOAOV rather than aMAw~. 
He thus uses, to indicate a Roman legal formula, an expression that 
has no precedent in Greek treaties, which suggests once again that 
a80~, the most obvious verbal counterpart, was not a legal formu­
la.56 Furthermore, the author of 1 Maccabees, in recounting the alleged 
treaty of 161 between Rome and the Jews (8.28), renders sine dolo 
malo as ov J..'ETU BOAOV, a phrase paralleled in the Rome-Methymna 
treaty (JG XII.2 510) but in no treaty between Greeks.57 

55 Some examples: IG XII.2 SlO with Suppl. p.116 (Rome and Methymna, ca 129, or 
as late as the 90's [Gruen (supra n.S4) II 741]); Sherk, Roman Documents 16 (Rome 
and Astypalaea, 105); SEG 3.378c (piracy law, ca 100); I.Delos 1511 (piracy law, 67); 
Sherk 58 (Rome and Rhosus, 30's); 73 (Mytilene, date unknown); OGIS 629 (Pal­
myra, A.D. 137). The preposition may vary: alJEV i.Detos IS11, SEG 3.378c; I-UTiX IG 
XII.2 SIO; xwpis OGIS 629. But none of these phrases occurs in inscribed treaties 
before the second century or outside a Roman context. I do not understand Bicker­
man's assertion (90 n.13) that the word BOAOS- disappears after composition of the 
Septuagint, is lacking in Hellenistic papyri and inscriptions, and reappears in legal 
Greek as a translation of dolus; d:80AWS" certainly occurs in Cretan inscriptions of the 
second century (supra n.45). We should also note Nep. Ages. 2.4, where Nepos trans­
lates the d:80AWS" of Xen. Hell. 3.4.S-6 as sine dolo. As neither passage is a legal text, 
my argument is not affected. It cannot be determined if Nepos (erroneously in my 
view) saw d:80AwS- as a legal formula and thus rendered it sine dolo, or simply offered 
the most obvious Latin translation. The uncertainties of such cases in literary sources 
explain and justify the emphasis on epigraphical evidence in this study. 

56 Polyb. 1.78.7,21.32.7, cf Liv. 38.11.2, 5. sine dolo malo occurs twice in Livy's ver­
sion of the treaty; a lacuna is found in Polyb. 21.32.2 where the first instance of a 
Greek equivalent of sine dolo malo should occur. Various restorations have been pro­
posed, but it seems clear from 21.32.7 that only XWptS- BOAOV can be correct. Some 
scholars assume that Livy's version is a Latin translation of Polybius, which is ques­
tionable. In any event, Polybius apparently translated from a Latin text. For the prob­
lem see Walbank (supra n.26) III 131-32. We do find d:80AWS" equated with sine dolo 
malo in a Roman-Vandal treaty of 478 (Malchus fr.13 [FHG IV 120)), but Late Roman 
practice has little validity for the period under consideration here. 

57 For bibliography and an argument for the treaty's historicity see Gruen (supra 
n.54) I 43-44. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT L. WHEELER 269 

We can conclude that aooil.o~ and Greek anti-deceit clauses in gen­
eral are not routine formulas to confirm good faith. Greek practice in 
treaties differs from the explicit formulas and the consistency of Ro­
man usage. Any rote formula should be found on the stones far 
more often than it is 03.9% over four centuries), particularly when 
the issue involved is trickery in international relations. The anti­
deceit clauses we find in Greek treaties were occasioned by real and 
immediate concern about deception and especially sophistic inter­
pretation. 

ApPENDIX: 

A Catalogue of Sophistic Interpretations 

The forty-five examples gathered here, though doubtless not exhaustive, 
will illustrate the range of sophistic interpretations to be found in our literary 
evidence. They are classified according to theme or similarity of the inci­
dents; the seventh and largest group contains all the examples that defy more 
specific categorization. 

I. Religion 
1. Ov. Fast. 3.285-354 
2. Paus. 10.18.5 
3. Macrob. Sat. 1.7.35 

1-3 involve sophistic interpretations of a vowed or demanded 
sacrifice, so as to permit a substitution. 

4. Procop. Bel. 1.3.17-22 
The Sassanid king Peroz outwits the Ephthalite king by sealing his 
oath with obeisance to the rising sun according to the rite of the 
Magi, while the Ephthalite thinks that the obeisance is to him.58 

II. Arbitration 
5. Zenob. 4.62; cf Diog. 5.58, 5.92, Apostol. 10.7, Suda s. v. npo~ KpTlm 

KPT/Ti~EtIJ . 
6. Liv. 3.71-72, Dion. Hal. Ant.Rom. 11.52; cf Cic. Off. 1.33, Val. Max. 

7.3.459 

In both cases the arbiter of a dispute, having sworn to choose 
the deserving party, adds himself to the parties and chooses 
himself. 

58 In the Persians' version as given in Tabari no sophistic interpretation is found and 
Peroz's renewal of the war and his subsequent defeat and death are seen simply as 
breaking faith and its just reward. See Th. Noldeke, tr., Tabari, Geschichte der Perser 
und Araber zur Zeit der Sasaniden (Leiden 1879) 124-25. 

59 The contesting parties are Ardea and Aricia in Livy and Dionysius but Nola and 
Neapolis in Cicero and Valerius. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

270 SOPHISTIC INTERPRET A TIONS AND TREATIES 

III. Days and Nights 
7. Strab. 9.2.4 (Ephorus FGrHist 70F119), Polyaen. 7.43, Suda s.v. 0P~Kia 

1TUPfVPWV;, Zenob. 4.37; (:f Suda s. v. epi!Kfc; OPKW OVK E1TL(rTUIITat, 

Zenob. 4.32, Diog. 5.25, Apostol. 8.91 
8. Cic. Off. 1.33; Plut. Mar. 223A-B 
9. Photius and Suda s. v. cfJOtllLKWII alJllfJijKUt (Demon FGrHist 327F13); 

Anon. Byz. Rhet.mil. 2.4; Diog. 8.67; Macar. 8.74, cf Strab. 6.1.15; 
Dion. Hal. Ant.Rom. 19.360 

10. Polyaen. 6.53, cf Thuc. 4.102.3 
All four play on the ambiguity of the terms 'day' or 'day and 
night'. 

IV. Hiding Gold 
11. Ath. 8.360E-61c (Ergias of Rhodes FGrHist 513F1, Polyzelus 521F6)61 
12. Stob. Anth. 3.28.21 Hense 

Both cases are attempts to avoid surrender of gold: to the victor 
of a siege (11) or to a foreigner who deposited it with a friend 
(12). 

V. Eliciting an Enemy Commander under Siege 
13. Thuc. 3.34.3; Polyaen. 3.2 
14. Polyaen. 2.662 

15. Polyaen. 2.19 
In 13-15 a besieged commander agrees to a parley in the besieg­
ers' camp. His town is then taken by assault; the commander is 
returned safe to the town in accordance with the agreement, but 
then killed. In 14 the commander's life is threatened at the 
parley and he is returned after agreeing to open the gates to the 
enemy. 

16. Dio. 17.84.1-5; Polyaen. 4.3.20, cf Plut. Alex. 59.6-7; Arr. Anab. 
4.27.3-4 

Alexander massacres an Indian garrison after its surrender: he 
had sworn that they could leave the fort, but not that they were 
released or were his friends. 

60 Strabo and Dionysius attribute the incident to the foundation of Metapontum and 
Callipolis respectively, while the other sources give Carthage. 

61 This case is unique in containing two sophistic interpretations, one to counter the 
other. The surrendering party circumvents one stipulation of the terms of capitulation, 
while the victors reply in kind with chicanery on another point. 

fit This sophistic interpretation of Dercyllidas is especially interesting because of his 
nickname Sisyphus, whose trickery in oaths Eustathius directly connects with sophistic 
interpretations: Xen. Hell. 3.1.8; Ephorus FGrHist 70F71; supra n.33. How he acquired 
the nickname is unknown; his other two recorded stratagems do not concern oaths: 
Xen. Hell. 3.1.17-19; Jul. Afric. Cest. 7.1.11 (attributed to Clearchus in Polyaen. 
2.2.9). 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT L. WHEELER 

VI. Returning Half 
17. Val. Max. 7.3.4 
18. Polyaen. 6.15 

VII. 

19. 

A stipulation to return half of something to the enemy results in 
ships (17) and arms (18) being cut in half.63 

Unclassified Examples 

ad. 19.303-07; Eust. ad ad. 19.396 (1870-71) 
Odysseus tricks Penelope. 

20. Ath. 8.338c; Eust. 1870-71 
Hiding a stolen fish with another person: the one who took it 
did not have it and vice versa. 

21. Anon. Byz. Rhet.mil. 2.3 
Besiegers receive the golden reflection of sunlight on the city's 
image rather than the gold from the city's cult statue-a play on 
the word eiKtfJJl. 

22. Paus. 10.30.2 
Cretan Pandareos' sophistic oath after stealing the golden dog of 
Zeus.64 

23. Hdt. 4.201; Polyaen. 7.34 
An oath is taken over a concealed trench with stipulation of the 
oath's validity as long as the earth stands. 

24. Ael. VH 12.8 
Cleomenes I gains power with the aid of Archonides, swearing 
to "do nothing without his head" (without consulting him); he 
then decapitates Archonides and keeps the embalmed head with 
him during state business. 

25. Polyb. 12.6.3-5; Polyaen. 6.22 
The Locrians swear an oath that is to last as long as they tread 
on earth and have heads on their shoulders; but they have put 
earth in their shoes and heads of garlic on their shoulders under 
their clothing. 

26. HA Aur. 22.5-23.3 
Aurelian, besieging Tyana, vows not to leave even a dog alive if 
he takes the town; successful, he repents and kills the dogs but 
not the people. 

27. Tac. Ann. 12.47 
Murder is committed by smothering, since poison or a sword is 
prohibited in an oath. 

271 

63 A similar trick occurs in no. 11, where ships without rudders, sails, and oars 
are given, since the agreement stated only ships. Cf Macrob. Sat. 2.6.2; Dig. 50. 
16.242. 

64 See M. van der Kolf, RE 18 (1949) 501. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

272 SOPHISTIC INTERPRETATIONS AND TREATIES 

28. Frontin. Str. 4.7.17 
An army surrenders with an oath to lay down their swords (fer­
rum)~ they are slaughtered because they retained iron clasps on 
their cloaks. 

29. Liv. 22.58.4-61.1O~ Cic. Off. 3.113-15~ Polyb. 6.58.2-12~ Gell. NA 6.18 
Ten Roman prisoners from the battle of Cannae are sent by 
Hannibal to negotiate ransom~ one circumvents the oath to 
return by going back to camp immediately on the pretext of 
having forgotten something. 

30. Dion. Hal. Ant. Rom. 2.40.2~ Fab. Pict. HRR J2 fr.8~ Liv. 1.11.6-8~ Val. 
Max. 6.9.1~ Plut. Rom. 17, Mor. 309B-C~ Servo ad Aen. 8.348~ Zonar 
7.3 

Tarpeia betrays Rome upon the Sabines' oath to give her what is 
on their left arms~ rather than give her their armbands, they 
crush her with their shields. 

31. Diod.30.18.1-2 
Antiochus IV's capture of Pelusium apparently involved the so­
phistic interpretation of a truce.65 

32. Thuc. 3.52.2-53.2, 68.1 
The Spartans lure the Plataeans into voluntary surrender by stip­
ulating that only the guilty will be punished and no one contrary 
to justice; at the trial justice is defined as service to Sparta dur­
ing the war. 

33. Polyaen. 6.14 
An agreement of the Samnites with their enemies to remove 
one row of stones from the Samnite walls is executed by remov­
al of the bottom row, thus demolishing the walls. 

34. Polyaen. 5.12.2 
Timoleon gains the surrender of the Syracusan tyrant by promis­
ing not to prosecute him; on taking the city he demands of the 
assembly (rather than in court) that the tyrant be killed, saying 
that this is not a prosecution. 

35. Hdt. 4.154 
Themison throws the daughter of Etearchus into the sea as 
sworn, but she is attached to a rope and is retrieved. 

65 C! supra n.26. No source explicitly states the nature of Antiochus' stratagem, but 
Polybius' reference (preserved in Diodorus) to observance of the letter of the law 
without regard for justice fits the pattern of a sophistic interpretation. The truce seems 
to have lasted some months before some Egyptian officers, impressed by Antiochus' 
generosity after the battle near Mt Casius, assisted in the capture of Pelusium. If the 
terms stated that the truce would last so many days, without reference to nights, then 
Antiochus' stratagem may have involved a night attack in conjunction with the treach­
ery of the Egyptian officers. This type of sophistic interpretation is seen in the cata­
logue (nos. 7-10), in Eustathius' discussion of sophistic oaths (1870-71), and also in 
treaties (Staatsvertr. 584). On the capture of Pelusium see Diod. 30.14; Walbank (supra 
n.26) III 352; Otto Merkholm, Antiochus IV of Syria (Copenhagen 1966) 64-77. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

EVERETT 1. WHEELER 

36. Amm. Marc. 27.5.9 
The Goth Athanaricus, under oath not to set foot on Roman ter­
ritory, makes peace with the Romans from boats in the middle of 
the Danube. 

37. Plut. Dem. 4, cf Servo ad Aen. 3.402 
Bound by an oath of silence, Demetrius writes a message on the 
ground. 

38. Judges 21; Jos. AJ 5.155, 169-74; Ambros. De off. 3.19 
The Hebrews, having sworn not to give their daughters to the 
Benjaminites, arrange for the Benjaminites to steal them. 

39. App. Lib. 1.1 
The Phoenicians swear only to take as much African territory as 
an ox-hide encompasses; they then cut the hide into a thin rope 
and surround what is later the citadel of Carthage. 

40. Tac. Hist. 4.41.1-2 
Senators devise circumventions to the oath that they neither 
harmed a citizen nor profited during the civil war of 69-70. 

41. Clem. AI. Strom. 3.6.50.4-51.1; Ael. VH 10.2 
An athlete in a foreign city promises to take a woman home 
with him and later departs with her picture.66 

42. Polyaen. 5.5.1 
When the Megarians wish to settle at the Chalcidian colony of 
Leontini, held jointly with the Sicels, Theocles says he cannot 
drive out the Sicels because of his oath to them. But he opens 
the gates by night to the Megarians, who expel the Sicels. 

43. Polyaen. 8.66; Plut. Mor. 244F-45B 
In a war with Erythrae the Chians agree to surrender Leuconia 
and depart from the city with only a cloak and tunic. The Chian 
women, however, recall the ancient custom to call a spear a cloak 
and a shield a tunic. Thus the Chians depart armed from the city. 

44. Polyb. 20.9-10; Livy 36.27-28 
The famous confrontation of 7TI.U'TLC; and fides. In 191 B.C. Aci­
lius Glabrio tries to trick the Aetolians into a quick surrender by 
urging them to enter into fides (7TtUTLC;) , but without making 
clear that fides and deditio are the same.67 

66 The name of the athlete is Aristotle in Clement and Eubatas in Aelian. 

273 

67 Gruen has argued that Polybius and modern historians have misinterpreted this 
incident, which in his view reveals no sharp differences between Greek 7Tun". and 
Roman .fides or between Greek and Roman values: "Greek 7TiuTL<; and Roman Fides," 
Athenaeum 60 (I982) 50-68, c): Gruen (supra n.54) 28 n.80, where he seems more 
cautious about deditio, .fides, and 7TWTL<;. Despite his arguments in 1982, not all of 
which convince equally, the incident as told by Polybius has a sophistic interpretation 
of the term .fides as its basis. If, as Gruen argues, the Aetolians understood Roman 
.fides, if 7TiuTL<; and fides were indistinguishable, and if fides and deditio were not syn­
onyms, then Glabrio's 'trick' was no trick at all and Polybius' clear equation of fides 
and deditio (20.9.12) is blatantly wrong. 



WHEELER, EVERETT L., Sophistic Interpretations and Greek Treaties , Greek, Roman and 
Byzantine Studies, 25:3 (1984) p.253 

274 SOPHISTIC INTERPRETATIONS AND TREATIES 

45. Hdt. 6.62-63.1 
A man in love with another's wife confirms an agreement with 
him by oath that each will give whatever the other asks. The 
second must give up his wife, as he did not suspect that his mar­
ried friend would demand his Spouse.68 

THE INSTITUTE FOR ADVANCED STUDY, PRINCETON 
October. 1984 

68 The sophistic interpretation is misunderstood by W. W. How and J. Wells, Com­
mentary on Herodotus I (Oxford 1912) 89. 

I wish to acknowledge that my research has been supported in part by a fellowship 
from the Alexander von Humboldt Foundation.