Janer 5_f7_138-156

1/31/06

5:36 PM

Page 139

CURSES AND BLESSINGS IN ANCIENT GREEK OATHS

CHRISTOPHER A. FARAONE
Like most ancient peoples, the Greeks used oaths to verify statements of fact and to give added weight and security to a wide variety of social, political and scal agreements. And they often enforced
the compliance and truthfulness of these oaths by sanctioning them
with a balanced pair of conditional self-blessings and self-curses,
such as we nd at the end of the well-known Hippocratic oath:1
If I fulll this oath without violating it, may it be granted that I enjoy a
happy life and profession, honored always among men. But if I violate it
and perjure myself, may the opposite befall me.

There are, in fact, quite a number of oaths that have this simple
kind of binary sanction, which rst describes blessings and then
briey mentions curses, like the proverbial carrot and stick, to
encourage good behavior with a promise of reward and to discourage bad behavior with a threat of punishment. In oaths of this
more balanced type the curse is usually shorter and very tame.
Indeed, it often consists, as it does here, of a simple negation of
the blessing.
There is also evidence, however, for the popularity of Greek
oaths that greatly emphasize their curses by making them longer
and much more vivid. See, for example, the self-curse described in
this account of a dramatic oath apparently sworn by Greeks living
on the island of Thera in the sixth century B.C.E.:2
On these conditions they made an agreement, those who stayed there (i.e.
on Thera) and those who sailed on the colonial expedition, and they put

J. L. Heiberg, Hippocratis opera 1.1 (Leipzig 1929) 27, in my translation.

SEG 9.4. The translation is that of A. J. Graham, Colony and Mother City in
Ancient Greece (Manchester 1964) 226. For the best Greek text see R. Meiggs and
D. Lewis, A Selection of Greek Historical Inscriptions (Oxford 1969) 5-9 no. 5; for discussion and additional bibliography, see Graham (loc. cit.) and C. A. Faraone,
Molten Wax, Spilt Wine and Mutilated Animals: Sympathetic Magic in Early
Greek and Near Eastern Oath Ceremonies Journal of Hellenic Studies 113 (1993)
60-62.
2

Koninklijke Brill NV, Leiden, 2006

Also available online www.brill.nl

JANER 5

Janer 5_f7_138-156

140

1/31/06

5:36 PM

Page 140

christopher a. faraone
curses on those who should transgress these conditions and not abide by
them. . . . They molded wax images and burnt them up while they uttered
the following imprecation, all of them, having come together, men and
women, boys and girls: May he, who does not abide by these oaths but
transgresses them, melt away and dissolve like the imageshimself, his seed
and his property. But to those who abide with these oaths let there be
many good things for themselves and their ospring.

In this case, a city in some kind of crisis (probably famine or overcrowding) is forced to make an unpopular decision: to send away
a percentage of its people to form a colony. In anticipation of
strong resistance and non-compliance, the city forces its own citizens to take a solemn oath to comply with the agreement and they
resort to a very memorable public ceremony to ensure that the
Theran people do not violate their oaths. In contrast to the balanced and the rather subdued sanction in the Hippocratic oath
quoted earlier (a blessing followed by a shorter curse) the Theran
people utter a very frightening self-curse rst and then a much
shorter blessing. The curse, moreover, is emphasized even more by
a vivid and horric ritual that involves the destruction of waxen
images, a device that clearly works according to the principles of
sympathetic or persuasively analogical magic:3 the Therans collectively wish that those who violate their oaths, along with their
families and their property, might melt away and dissolve like the
wax images that were melting before their eyes.
My goal in this essay is to describe and then explore the dierences
between oaths (like the Hippocratic) that end with a careful balance of conditional self-blessings and self-curses, and those more
fearful oaths (like that of the people of Thera) that have a very
lopsided sanction: a powerful and emphasized curse followed by a
shorter blessing or in some cases by no blessing at all. How does
one explain the dierence between the two forms? Is it a problem

3
I use the traditional terms sympathetic and sympathetically advisedly
throughout this article. S. J. Tambiah, Form and Meaning of Magical Acts: A
Point of View in R. Horton and R. Finnegan (eds.) Modes of Thought (London
1973) 199-229, dismisses the common view that sympathetic magic is based on
poor observation of empirical analogies. He distinguishes instead between the operation of empirical analogies (used in modern scientic discourse to predict future
actions) and persuasive analogies (used in rituals in traditional societies to encourage future action). Such rituals do not betray inferior observation skills, but rather
they reveal a profound belief in the extraordinary power of language. Cf. G. E. R.
Lloyd, Magic, Reason and Experience (Cambridge 1979) 2-3 and 7.

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 141

curses and blessings in ancient greek oaths

141

with our sources? Do we get exaggerated reports of these sympathetically magical curses, only because they are so weird and because
they are so memorable? In what follows, I argue that the oaths
with the longest and most emphatic curses, but few if any blessings are generally imposed by stronger powers on lesser ones. They
are also frequently sworn by a large group or its representative in
international or other public contextsfor example by an entire
city or by an entire army. These kinds of oaths, moreover, usually
have far-reaching and fatal consequences for the perjurer. Oaths
containing a more careful balance of blessings and curses, on the
other hand, seem to be designed for intramural situations or private oaths between individuals, and the sanctions attached to them
are rather vague, for example: in one popular formula, prosperity is the blessing and misfortune is the curse.
I shall argue, furthermore, that the lopsided oaths with emphatic
curses are generally attached to promises about future behavior (the
technical term is promissory oath), as in the case of the Theran
oath, in which the islanders promise that they will not return home
after the colony is sent out. The more balanced oaths, however,
appear in both promissory oaths (as in the Hippocratic oath quoted
earlier) and in testamentary oaths, that is: oaths in which a person
swears to the truthfulness of her or his testimony. We shall see,
however, that these lopsided, coercive oaths also appear in the
realm of private oaths by individuals in legal trials or other situations of high social tension in which perjury would have dire consequences for the entire city. I should stress, however, that this is
a preliminary study, in which I focus on a few test cases and try
to see how dierences in the length and intensity of the blessing
and the curse and other details (such as the order in which they
appear) might correlate to the social and historical circumstances
of the performance of the oath. In short, I shy away from any
monolithic, transhistorical model for these phenomena and prefer
a contextualized understanding of oaths, in which variations in the
length and impact of curses and blessings may provide important
clues to the social relations of the participants or the social tensions between them. In short, I am more interested in the dierences
than in the commonalities.

Janer 5_f7_138-156

142

1/31/06

5:36 PM

Page 142

christopher a. faraone
Section One: Coercive Oaths with Emphatic Curses

In a study published nearly a decade ago (see footnote no. 2), I

showed how the Greeks in the archaic period modeled their lopsided oaths, like that of the people of Thera quoted above, on two
kinds of Near Eastern oaths popular in the 8th and 7th centuries
B.C.E.: (i) the so-called vassal oaths which powerful Assyrian kings
like Esarhaddon forced weaker, dependent kings to swear; and (ii)
the so-called loyalty oaths which generals and kings forced upon
soldiers (often mercenaries) in large multinational armies. As far as
we can tell, these oaths were almost always used in the international sphere of activity, and were imposed by powerful states or
kings on weaker ones. They were, in short, designed to prevent
disobedience or rebellion. They are, moreover, usually sanctioned
by curses alone. I shall argue in this section that in the context of
this typically unbalanced political relationship, it perhaps makes
sense to see this kind of oath-curse as a vehicle for broadcasting
the threats that are implied in them and to understand that in this
sort of international context, at least, it was more expedient and
eective to control a conquered nation or a subservient army with
a stick rather than with a carrot.
Two other shared features of these Near Eastern curses are: (i)
their all-encompassing eect (the men of the subservient city or
army swear to them, but often their wives and children are embraced
by the conditions of the curse) and (ii) the very theatrical rituals of
a sympathetically or persuasively analogical nature that accompany
themsuch as the melting of wax gurines or the mutilation of
animals. I argued in the aforementioned article that such ceremonial acts were probably designed to communicate in a very immediate and graphic fashionand often to peoples who did not speak
a common languageprecisely what will result if they break their
oaths. A mid eighth-century B.C.E. Aramaic text found near Aleppo,
for example, preserves a treaty between two kings, Bargayah and
Mattiel, living on the southwestern periphery of the Assyrian empire.
The latter is forced to swear to the dire consequences which will
befall him and his state if he should violate the stipulations of the
treaty:4

4
The translation is by F. Rosenthal, in J. B. Pritchard (ed.), Ancient Near Eastern
Texts Relating to the Old Testament 3 (Princeton 1969) 659-60. For dating, detailed

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 143

curses and blessings in ancient greek oaths

143

As this wax is consumed by re, thus Ma[ttiel] shall be consumed b[y ]re.
As this bow and these arrows are broken, thus Inurta and Hadad (= names
of local deities) shall break [the bow of Mattiel] and the bows of his nobles.
As a man of wax is blinded, thus Mattiel shall be blinded.
[As] this calf is cut up, thus Mattiel and his nobles shall be cut up.

Since only Mattiel and his nobles swear the oath and since they
alone are threatened with the curses, it is clear that they are in a
politically or militarily weak position vis--vis Bargayah, who has
apparently forced this oath upon them. The deictic adjective this,
moreover, in three of the four curses encourages us to imagine that
wax images were disgured and burnt, weapons were ceremonially
broken, and domestic animals were cut up and otherwise mistreated
in plain sight of Mattiel and his people.
Although these oaths seem to have been designed and used primarily for international situations, we do nd them turning up occasionally in national or domestic settings, where a whole city forces
itself to swear a public oath. The coercive and self-reexive oath
of the Therans quoted earlier, for example, is imposed by the
Therans (or at least their civic leaders) on the entire city. This is
also true for an oath sworn by all the free inhabitants of Jerusalem
at a time roughly contemporaneous with the Theran oath ( Jeremiah
34:8-20):5
Everyone, ocials and people, who had entered into the covenant, agreed
to set their male and female slaves free and not to keep them enslaved any
longer; they complied and let them go. But afterward they turned about
and brought back the men and women they had set free, and forced them
into slavery again (vv. 8-11).

Yahweh complains to Jeremiah about the perjury and promises

punishment:
I will make the men who violated my covenant, who did not fulll the
terms of the covenant which they made before me, [like] the calf which
they cut in two so as to pass between the halves. The ocers of Judah and

discussion and bibliography, see: J. A. Fitzmyer, The Aramaic Inscriptions of Sere,

Biblica et Orientalia 19 (Rome 1967) 52-58 and A. Lemaire and J.-M. Durand,
Les inscriptions aramennes de Sr et lAssyrie de Shamshi-ilu (Geneva 1984).
5
The translations used here and immediately below are from The Jewish
Publication Society, Tanakh: A New Translation of the Holy Scriptures (New York 1985).
For discussion see H. Tadmor, Treaty and Oath in the Ancient Near East: An
Historians Approach in G. M. Tucker and D. A. Knight (eds.) Humanizing Americas
Iconic book: Society of Biblical Literature Centennial Addresses 1980 (Chico, CA 1982) 136.

Janer 5_f7_138-156

144

1/31/06

5:36 PM

Page 144

christopher a. faraone
Jerusalem, the ocials, the priests and all the people of the land who passed
between the halves of the calf shall be handed over to their enemies who
seek to kill them. Their carcases shall become food for the birds of the sky
and the beasts of the earth (vv. 18-20).

This incident dates to the end of the seventh or the beginning of

the sixth century and concerns an oath sworn by the people of
Jerusalem, probably when Nebuchadrezzar laid siege to the city in
586 B.C.E. As was true in the circumstances surrounding the Theran
oath, an emergency forces the political leadership to take a very
unpopular step: they liberate all the slaves to help in the military
defense of the city and they promise that these freed slaves will
never be re-enslaved. In order to ensure that this unpopular decision is universally upheld, all of the free people of the city participated in the oath-ceremony, which apparently included a walk
between the severed halves of a calf and the conditional selfimprecation that they would suer like the bisected calf if they perjured themselves.6 These self-reexive oath-curses, then, may be
peculiar to city-states like Jerusalem or Thera, where the collective
will of the people is seen to work in an almost tyrannical manner
against itself, especially in times of impending disaster.
As it turns out, most of the Near Eastern and Greek examples
of this kind of oath date to the same period:7
Hittite Military Oaths (1450-1350 B.C.E.)
Aramaic Sere Inscription (mid-8th Century B.C.E.)
Treaty between Assurnerari V and Matiilu (mid-8th Century B.C.E.)
Vassal Treaties of Esarhaddon (672 B.C.E.)
?Oath of Pittacus and Alcaeus (late 7th Century B.C.E.)
Oath of the Founders of Cyrene (late 7th Century B.C.E.)
Oath of Zedekiah and the People of Jerusalem (late 7th/early 6th Century
B.C.E.)
Oath of the Phocaeans (mid 6th Century B.C.E.)

6
This passage is usually discussed in tandem with Genesis 15:15-19, where
Abraham bisects a three-year-old heifer, a three-year-old she-goat, a three-yearold ram and a turtle dove, and lays the halves opposite one another. When it
gets dark, a re pot and torch (emblems of divinity) pass between the pieces (verses
17-18), indicating Yahwehs agreement to the covenant. See S. E. Loewenstamm,
Zur Traditionsgeschichte des Bundes zwischen den Stcken VT 18 (1968) 500506, esp. 503 and C. Westerman, Genesis 12-36: A Commentary, trans. J. J. Scullion
(Minneapolis 1985) 225-28, for detailed discussion and bibliography.
7
This list of the historical or quasi-historical oaths has been adapted from
Faraone (n. 2) 79-80. The quasi-historical oaths are preceded by a question mark.
For the inclusion here of the Phocaean and Delian League oaths, see H. Jacobson,
The Oath of the Delian League Philologus 119 (1975) 256-58.

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 145

curses and blessings in ancient greek oaths

145

?Oath before the Battle of Plataea (479 B.C.E.)

Oath of the Delian League (478/77 B.C.E.)
Oath of the Greek and Persian Mercenaries in Xenophon (401 B.C.E.)

As we can see, this type of coercive public oath, with its vividly
enacted curses but with few if any blessings, was popular in the
eastern Mediterranean basin in the 8th-5th centuries, but seems to
have fallen into disuse by the late classical period at least in the
context of international or corporate agreements.
Although there are some clear indications that by the early
Hellenistic period the rationale behind these rituals was not so
clearly understood,8 this emphasis on and forward placement of the
curse continues to survive in some parts of the Greek world, for
example: in this oath from Naupactus (IG 9.22.3 no. 609 A 15-16):
that destruction may befall the one who transgresses this precept,
himself his progeny and all [his property?], but for him who acts
piously, the gods must be propitious. This pattern also shows up
in Hellenistic Crete:9
If I should swear falsely to the things that I have sworn and agreed, may
the gods by whom I have sworn be angry and may I be destroyed in the
most miserable way, and may earth nor trees bear fruit, nor the women
bear according to the laws of nature, and in war may I be defeated. But
if we keep our oath, may the gods be propitious and may everything go
well.

This Cretan oath is a bit odd, in the way that it switches from the
rst person singular in the curse to the rst person plural in the
blessing, but the anomaly may in fact be explained by the political
context: the Cretan city of Hieraptna and several of her outlying
settlements have agreed upon a military and political alliance, but
only the people living in the settlements seem to utter the text
quoted above, not the people of Hieraptna. In short: the I in
this text (i.e. those liable to be cursed if they fail to keep their
oaths) refers only to the individual oath swearers living in the

8
The Hellenistic translators of the Septuagint, for example, completely botch
their Greek version of the description of Zedekiahs oath ( Jer. 34:18-20; quoted
extensively above), primarily because they do not understand why the calves were
cut in two. H. S. German, Some Types of Errors of Transmission in the LXX
VT 3 (1953) 397-400.
9
A. Chaniotis, Die Vertrge zwischen kretischen Poleis in der hellenistischen Zeit,
Heidelberger Althistorischer Beitrge und Epigraphische Studien 24 (Stuttgart 1996)
432-39 no. 74.

Janer 5_f7_138-156

146

1/31/06

5:36 PM

Page 146

christopher a. faraone

outlying settlements, whereas the we (those liable to the blessing)

perhaps includes the people of Hieraptna as well. Thus although
the frightening ceremonies are gone, the oath (like its more vivid
ancestors) seems to be a coercive one that a stronger statein this
case Hieraptnaforces on some weaker ones.
This implied relationship can be seen more clearly when foregrounded by other oaths from Hellenistic Crete, which are particularly well represented in the epigraphic corpus:10
1) Three instances of a curse followed by a blessing of equal length (nos.
26, 31 and 42).
2) Two instances of a blessing followed by a curse of equal length (nos. 59
and 64).
3) One instance of a curse followed by a longer blessing (no. 16).
4) One instance of a blessing followed by a longer curse (no. 27).
5) Two instances of a long curse followed by a short blessing (nos. 10 and
74).

The longest and most violent curse in the collection (no. 74) was
the one quoted above between Hieraptna and her satellites. On the
other hand, the predominance of balanced sanctions (or those with
more emphatic blessings) in six of the nine examples suggests that
the political playing eld was generally a level one for the smaller
city-states of Hellenistic Crete.
These kinds of lopsided, coercive oaths also persist in promissory
oaths sworn by individuals in connection with two very old cultural institutions in Greece: the Olympic games and the archonship in Athens. Pausanias gives us a detailed description of the
special ritual performed by the athletes, their families and their
entourage prior to competing in the quadrennial games at Olympia
(5.24.9-11):
But the Zeus in the council chamber (i.e. of the city of Elis) is of all images
of Zeus the one that has been designed to strike fear in men who do wrong.
His epithet is Horkios (Of the Oath) and in each hand he holds a thunderbolt. It is required that beside this image the athletes, their fathers and
their brothers and even their trainers swear an oath upon the cut-up bits
(t tmia) of a boar that there will be no misdeed on their part in the competition at Olympia. . . . Before the feet of (i.e. this statue of Zeus) Horkios,
there is a bronze plaque with elegiac verses inscribed upon it designed to
instill fear in those who forswear themselves.

10
Chaniotis (n. 9). The numbers in parentheses refer to the numbered oaths
in his collection.

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 147

curses and blessings in ancient greek oaths

147

Here it is our misfortune that Pausanias does not quote the actual
words of the oath or the verses on the statue base. He does, nonetheless, stress that the especially fearful and solemn nature of both
the gods image and the oath is designed to dissuade men who
would presumably act unjustly if they were not forced to swear the
oath. The cut up bits of boar (t tmia) serve the same function
here as the mutilated animals do in the Near Eastern oathsin
this case the aspiring Olympic athlete probably touched the tmia
while swearing that his own esh should be so treated if he broke
the rules of the game.11 This is, of course, the oath of a private
individual, but if we fore-ground it with my previous discussion of
Near Eastern promissory oaths, its coercive and lopsided form perhaps does make sense, for the games were held in an international
setting dominated by the host city Elis, in whose council chamber
the contestants were forced to swear their oaths.
This type of coercive oath was also sworn by some public ocials
at Athens. The Aristotelian Constitution of Athens (55.5), for example,
speaks of a special oath that the nine archons (executive ocers)
swore at the beginning of their term of oce, while standing upon
a special stone in the Agora:12
. . . they go to the stone upon which are the cut-up bits (t tmia) . . . and
mounting on this stone they (i.e. the archons) swear they will govern justly
and according to the laws, and that they will not take bribes.

This oath is mentioned in passing in a few other places (e.g. ibid.

7.1 and Plut. Solon 25.2), and undoubtedly involved a self-curse in
which the person swearing the oath stepped upon or otherwise
came into contact with the mutilated animal esh. Here this special curse seems designed to encourage politicians not to take bribes
or otherwise break the laws of Athens, but at the same time signals to us that such behavior, then as now, was an endemic problem. Since the bad behavior of national leaders, like the archons,
could potentially have disastrous eects on the whole city, these
leaders are forced to swear an especially dire oath and to perform
publicly a bloody self-cursing ceremony (but no self-blessing). On

11
For a more detailed discussion of the signicance of the tmia in this passage and those discussed below and also of the term sfgia used in a similar
way to describe a similar ritual, see Faraone (n. 2) 65-72.
12
See also Aeschines Tim. 114-115 for a similar oath against bribery.

Janer 5_f7_138-156

148

1/31/06

5:36 PM

Page 148

christopher a. faraone

the other hand, we sometimes nd in classical Athenian texts (like

Aristophanes comedy The Knights) the popular idea that the people (i.e. the dmow of the Athenians) could rule in a tyrannical
manner over its individual leaders. Thus it seems quite appropriate that the Athenians use the same kind of oath originally deployed
by Near Eastern despots to keep its subordinates in line: here the
archons are truly being treated as the servants of the people.
More Balanced Oaths with a Slight Emphasis on Blessings
I now turn to oaths that strike a more equal balance between curses
and blessings. In the Greek tradition this kind of sanction often
appears in promissory oaths, like those discussed in the previous
section, and in testamentary oaths concerned about the veracity of
statements. The Hippocratic oath quoted at the start of this study
serves as a good example of a promissory oath of this type.13
If we can use the Athenian evidence as a reliable indicator of
the wider Greek experience, testamentary oaths were equally popular.14 Thus, for example, when a young Athenian child was enrolled
in his or her phratry (= extended family group) during the Apaturia
festival, the father and two or three other witnesses had to make
a sacrice to Zeus and swear an oath asserting the childs true
paternity. This oath usually ended with a simple pairing of blessing and imprecation: If my oath is true, may I prosper, if false
may the opposite befall me.15 Like the Hippocratic oath, this one

13
For other examples, see the oath sworn by the Eretrians about the future
use of land near the city of Chalkis, in which we get the typically short version
of this kind of balanced sanction: If I am faithful to my oath, may I prosper; if
I am a perjurer, may misfortune befall me. We nd a somewhat fuller example
in the oath of representatives to the Delphic Amphictyonya group of Greek
states linked together by treaties, who swore: Thus do I promise by Pythian
Apollo and Leto and Artemis: If I swear a good oath I ask that they grant me
many good things, but if I swear a false oath I ask that they give me evils (SIG 3
no. 145.13-15). In both cases the promise of prosperity or blessings, is neatly
balanced by the threat of misfortune or evils, but there is no explicit mention of
death or destruction.
14
The discussion of legal oaths that follows depends heavily on C. A. Faraone,
Curses and Social Control in the Law Courts of Classical Athens Dike: Revista
di storia del diritto greco ed ellenistico (1999) 99-121, reprinted with minor changes in
D. Cohen (ed.) Demokratie, Recht und soziale Kontrolle in klassischen Athen, Schriften des
Historischen Kollegs Kolloquien 49 (Munich 2002) 77-92.
15
SIG 3 921. 14-15. The oath is, however, sworn over a sacrice to Zeus Patrios,

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 149

curses and blessings in ancient greek oaths

149

has the simplest and most benign form of sanction: perjury will
result in the punishment of only one person and compliance will
bring the opposite. Two historical anecdotes suggest, however, that
in emotionally charged situations, questions of paternity might result
in some ad hoc variations to this standard form of the oath.
Andocides, for instance (1.126), tells the story of the prominent
Athenian politician Callias, who was unexpectedly confronted at
this same Apaturia festival by the relatives of a former mistress,
who demanded as he was making the sacrice to Zeus that he recognize her son as his own. But instead of yielding to this request
and swearing the usual oath, Callias, instead took hold of the altar
and swore that the only son he had or had ever had, was
Hipponicus . . . if that was not the truth, he prayed that he and
his house might perish completely. Here, Callias seems to improvise on the spot a much more powerful and lopsided form of oath
in order to signal publicly his complete resistance to their appeal;
he gives a rhetorically exaggerated form of the oath by saying that
Hipponicus (the son born to Callias lawful wife) was the only son
he had or had ever had, and he drops the self-blessing altogether
and performs a dramatic and intensied version of the self-curse,
by gripping the altar of Zeus Phratrios and calling for the complete destruction of himself and his household if he is lying. Note
that in his performance of this improvised oath, Callias, by taking
hold of the altar and thereby laying his hands on or near the cut
up sacricial victim, comes quite close to performing the ritual that
often accompanies the coercive curses found in the promissory oaths
discussed in the rst section of this essaycurses which involve
standing upon or otherwise touching the chopped up bits (t tmia)
of a mutilated animal.
Herodotus (6.67-68) reports an even stranger oath sworn with
regard to the paternity of Demaratus, the Spartan king, who was
deposed as a bastard. According to the historian, Demaratus, while
sacricing a bull to Zeus, unexpectedly forced his mother to hold
its bloody innards (t splgxna), while she swore to the true identity
of his father. Here, too, it seems probable that Demaratusbecause

a god who regularly sanctions oaths in Greece, so we are probably right to imagine that he will in fact cause the misfortune to occur if the father forswears his
oath.

Janer 5_f7_138-156

150

1/31/06

5:36 PM

Page 150

christopher a. faraone

his father Ariston is deadis improvising a more powerful version

of the Spartan paternity oath, just as Callias does when he grips
the altar and exaggerates the curse in his oath. Although the exact
wording of the mothers oath is not quoted, it seems quite likely
that at the end of the oath she pronounced a conditional curse on
herself as she touched the entrails of the dead bull. The oath
Demaratus forces upon her seems, in short, to have been as lopsided and as coercive as those described earlier, for it is clearly
designed to force the unwilling mother to speak the truth. Callias,
on the other hand, is not forced to take this more powerful form
of oath; he, in fact, uses it to trump the brothers of his mistress,
for he willingly improvises an exaggerated version of the standard
sanction in order to emphasize the fact that he was speaking truthfully and they (by implication) were not. To sum up, then: the traditional Greek paternity oath apparently ended with a balanced
pair of blessing and cursewith the blessing coming rstbut it
apparently could be transformed into a lopsided oath with an
emphatic self-curse in cases where there was some doubt about the
paternity of the child or some extraordinary social tension surrounding the performance of the ritual.
Athenian men took many other testamentary oaths in connection with legal matters and lawsuitsoaths, which like these paternity oaths can be arranged along a scale of increasing severity and
fearfulness. It is, moreover, abundantly clear that the Athenians
themselves were aware that most of their testamentary oaths were
unremarkable, because they repeatedly point out the fact that the
form of oath sworn by participants in murder trials was extraordinary. Antiphon, for instance, describes the oath sworn at a
homicide trial as the greatest and most powerful oath (5.11).
Demosthenes gives us our most detailed description of this oath:16
On the Areopagus, where the law allows and orders trials for homicide to
be held, rst the man who accuses someone of such a deed will swear an
oath invoking destruction on himself and his family and his household, and
no ordinary oath either, but one which no one swears on any other subject, standing upon the cut up pieces (t tmia) of a boar, a ram, and a

16
Demosthenes 23.67-68. I give the translation of D. M. MacDowell, Athenian
Homicide Law in the Age of the Orators (Manchester 1963) 90-91, with one change:
upon the cut pieces instead of over the cut pieces.

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 151

curses and blessings in ancient greek oaths

151

bull, which have been slaughtered by the right persons on the proper days,
so that every religious requirement has been fullled as regards the time
and as regards the executants.

What is it, then, that makes these oaths so special? Like the oath
forced upon Demaratus mother, we must imagine that this very
elaborate curse ceremony (unlike any ordinary oath, as Demosthenes
says) was designed to be more fearful, and thereby presumably
more dicult for the potential perjurer to swear falsely, especially
in a public ceremony that may have been witnessed by members
of his own family and household, that is: people who would suer
under such a curse if he should foreswear it. A few lines later in
the same passage Demosthenes says that those who were accused
in a murder trial also swore this same oath, a point that is corroborated by Antiphon and Lysias.17 Other evidence suggests, moreover, that witnesses at murder trials also had to perform the same
type of self-curse. Thus in one legal speech the defendant complains that the prosecutor has cleverly bypassed the correct procedure for prosecuting a homicide in order to avoid having his witnesses
swear the proper oath: The witnesses are giving evidence against
me unsworn, although they ought to swear the same oath as you
(i.e. the man who is prosecuting him) and touch the slaughtered
animals (t sfgia) before giving evidence against me (Antiphon
5.12). It would appear, then, that this very strong form of testamentary oath (with its vivid and bloody self-curse) was made by
the most important participants at a murder trialthe prosecutor,
the defendant and the witnesseswho all assert by their oaths that
what they are about to say at the trial is true.
In fact, of all the traditional oaths made at an Athenian murder trial, only one contains both a curse and a blessing, but it is
quite extraordinary. Aeschines tells us that the man who was victorious at a murder trial (either the prosecutor or the defendant)
had to make an additional oath at the end of trial:18
In homicide trials at the Palladion our ancestors very properly introduced
the rule (and you have maintained this tradition up to the present day) that
those who are victorious in the voting cut the cut pieces (t tmia) and
swear that those of the jurors who voted for him were making the true and

17
18

Demosthenes 23.67. Cf. idem 59.10, Antiphon 6.16 and Lysias 11.
Aeschines 2.87. This is the translation of MacDowell (n. 16) 91-92.

Janer 5_f7_138-156

1/31/06

152

5:36 PM

Page 152

christopher a. faraone
right decision, and that he had spoken no lie, and that otherwise he invokes
destruction on himself and his house, but prays that the jurors who voted
for him have many blessings.

Here, Antiphon makes it clear that the special ritual of swearing

upon the cut up bits (t tmia) of a mutilated animal was an old
custom introduced by their ancestors. (In fact, in this ceremony the
oath-swearer was even more involved in the ritual, for it appears
that he himself cut up the bloody bits as he was invoking destruction upon himself.) According to Aeschines, then, the winner of the
trial asserted two things in his oath. First, he swears that he himself has spoken no lie which is, of course, the standard assertion
of a testamentary oath. But then he also asserts that the jurors who
voted for him made the true and right decision, which is a very
odd andas far as I knowa unique stipulation in an ancient
Greek legal oath. The winner of the trial then sanctions his twopart oath, by pronouncing a binary pair of curse and blessing. The
curse will fall upon himself and his household, if he in the role of
prosecutor has lied and (in the process) successfully prosecuted an
innocent man for murder or if he in the role of defendant successfully evaded prosecution even though he was guilty. But in every
case a blessing will fall upon the jury, even if they inadvertently
voted wrongly for conviction or acquittal because the winner of the
lawsuit perjured himself. Here the strange and otherwise unparalleled pairing of a curse for the oath-taker, if he perjures himself,
but a blessing for the jury if they believe his perjured testimony,
seems to be an awkward transformation of the traditionally binary
sanction that we nd in ordinary testamentary oaths.
The question arises, of course: why do homicide trials alone
among various legal actions have such special, powerful oaths? If
such lopsided oaths prevent perjury more eciently that the more
balanced ones, why do they not spread over time to all court cases?
For surely the Athenians did not think that mendacity was greater
in homicide trials than in other legal cases. I have argued elsewhere, in fact, that the principle criterion for the use of a lopsided
oath is a concern for the health and well being of the whole city,
that is: such emphatic oaths are sworn by individuals primarily in
cases where the behavior of these individuals threatens the welfare
of the entire state.19 This principle is, I think, most obviously at

19

Faraone (n. 14) 83-85.

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 153

curses and blessings in ancient greek oaths

153

work in the strange winners oath sworn by the man who successfully prosecutes or fends o a conviction of murder, for as
Aeschines suggests, there was a real fear that if the jurors falsely
acquit a murderer or falsely condemn an innocent man to death,
some danger might arise for them and the whole city.20 Therefore,
by forcing litigants and witnesses in homicide trials to swear these
especially fearful oaths, the city probably could expect that some
proportion would in fact be dissuaded from perjury by the oath.
But the more important feature seems to have been that the wording of the oathand especially the wording of its unique blessingprotected the jurors and the city in those cases where murderers
successfully eluded a capital charge or where others successfully
prosecuted an innocent man for murder.
Let me summarize, then, what we have learned about these balanced oaths. They are deployed, like the coercive oaths discussed
in the previous section, in ratifying promises about future behavior (as in the Hippocratic Oath) but they are equally popular in
testamentary oaths. In their normal or ordinary Athenian form,
these oaths were apparently sanctioned by a short conditional blessing followed by a complementary self-curse, which is usually circumspect about the precise punishment that awaits the perjurer.
The Athenians apparently used this polite and balanced set of sanctions regularly in testamentary oaths required by courts of law and
by family groups intent on ascertaining the precise paternity of its
members. In both of these situations, however, the more frightening coercive oath could be improvised or substituted if circumstances required it. In the case of the paternity oaths, this could
be accomplished in an ad hoc manner by the person performing
the oath, who was responding to abnormally high tensions, such
as the debate over a bastard king at Sparta or the alleged child of
a very wealthy Athenian aristocrat. In the case of the testamentary
oath in the Athenian courts, however, there is a much more clearly
dened division between the normal kind of balanced oath apparently used in non-homicide cases (e.g. of theft or in squabbles over
a deceased relatives will), and the extraordinary lopsided forms that

20
J. Mikalson, Athenian Popular Religion (Chapel Hill 1983) 31-38, has shown in
fact that other speeches make frequent references to the fear that pollution (masma)
or perhaps angry ghosts would attack the jurors themselves in the case of a
wrongful conviction or acquittal.

Janer 5_f7_138-156

1/31/06

5:36 PM

154

Page 154

christopher a. faraone

were deployed in homicide trials. Since two of our fourth-century

sources claim that these extra powerful oaths were of great antiquity, it seems best to consider them as a traditional and important
part of the Athenian legal system and not simply an ad hoc variation.
Conclusions
There is, then, quite a dierence between the binary pair of blessings and curses used to sanction most ancient Greek oaths and the
lopsided and often very violent oaths that focus exclusively or primarily on cursing the potential perjurer. And it is not simply a
matter of quantity, for as we have seen in addition to being longer
and more emphatic, there is sometimes a marked intensity to the
curses, which often call for complete destruction, and not simply
misfortune. They are also often accompanied by sympathetic or
persuasive rituals, such as the melting wax gurines used in the
Theran oath ceremony. These dierences can be laid out in an
exaggerated and schematic form as follows:

usually sworn by:

usually sworn by:
usual context:
usual sanction:
rituals:
type:

More Lopsided
Coercive Oaths

More Balanced
Non-Coercive Oaths

nations, cities, armies or

individuals in murder trials
an inferior to a superior
extramural or international or
domestic murder trials
curse only or
longer curse, then blessing
fearful; sympathetically magical
usually promissory
(except testamentary
oaths at murder trials)

individuals
an equal among equals
intramural or domestic
blessing, then curse
(of equal length)
none
promissory or
testamentary

This division is, of course, admittedly somewhat crude. Indeed we

have seen evidence for a wide spectrum of dierent types and that
this dichotomy represents the poles of this spectrum, not two rigid
non-permeable categories. In my discussion of testamentary oaths,
for example, I have argued that there are, in fact, two forms, the
standard or ordinary form, which ts the description of the right
hand column, and the extraordinary form used in murder trials,
which we see in the left hand column. Thus in the end I have followed the lead of Demosthenes and Aeschines in claiming that the

Janer 5_f7_138-156

1/31/06

5:36 PM

Page 155

curses and blessings in ancient greek oaths

155

balanced oath (with an initial blessing) is the typical or ordinary

type of testamentary oath, whereas the the greatest and most powerful oaths (as Demosthenes describes them) are sworn only by
the participants in murder trials.
There is, moreover, a problem with the sources that I have used
in this paper: the second half of my paper discusses Greek testamentary oaths, but nearly all of the quoted examples are from
Athenian legal speeches. This fact makes me somewhat uncomfortable, because I have spent a good number of years battling the
traditional Athenocentric bias in the study of Greek religionthat
is: the often unexamined assumption that what goes on in Athens
is true for the rest of the Greek world. In this case, however, the
type of evidence that we get from Athens is unique for the classical period, because it reveals to us how oath curses and blessings
worked in oaths sworn by individuals, especially in instances where
they swear to the truthfulness of their words. The great majority
of our other examples of oaths and their sanctions come from
inscriptions that preserve public oaths sworn by cities or individuals in ocial roles representing their cities. Thus I would argue
that in this case, at least, the density and richness of the evidence
from the Athenian speeches give us important and unique insights
into the testamentary oaths sworn by private individuals.
In the end it would seem that, although the use of extended and
frightening curses in highly dramatic settings (e.g. while melting
wax egies or standing in the agora atop a stone covered with the
body parts of a mutilated animal) or in highly emotional and improvised oaths (e.g. the paternity oath Demaratus forces upon his
mother) may capture the imagination of modern scholars, it is
nonetheless ill-advised to suggest that such lopsided sanctions were
typical. Indeed, it seems more plausible to follow Demosthenes
and Aeschines and recognize that there was an ordinary or typical form of sanction, with a balanced pair of curse and blessing,
and extraordinary forms that were used more rarely and only in
special circumstances. In the case of curses used in promissory oaths,
these situations included special international or corporate agreements, such as those forced upon Assyrian vassals or those which
the people of Thera or Jerusalem forced upon themselves, as well
as oaths sworn publicly by important individuals, such as Athenian
archons or Olympic athletes, who were exposed to much greater
temptations to defraud the government or cheat at the games.
Likewise in the case of testamentary oaths, where the ordinary

Janer 5_f7_138-156

156

1/31/06

5:36 PM

Page 156

christopher a. faraone

paternity oath could occasionally be altered on the spot in response

to an extraordinary circumstance, or where the balanced oath-curse
typically used in Athenian trials was transformed in the case of
homicide to the lopsided and more frightening version, which had
been used (as the Athenians tell us quite plainly) for as long as
they could recall.21

21
An earlier version of this study entitled More Curses than Blessings: A
Curious Imbalance in Ancient Greek Oath Formulas? was presented at a conference Benedictio/Maledictio: What Have Curses to do with Blessings? at the
American Academy of Rome in April 2001. I would like to extend many thanks
to Lester Little, the organizer of the conference, for his hospitality and industry
in organizing the conference, and to the other participants, who made it an especially memorable event. Special thanks are owed to Henk Versnel for his comments on the paper both before and during the conference.