The Vow of Socrates

2015.04.17 | By Gregory Nagy

Lamentation over the Corpse of Socrates—Vincenzo Camuccini
Lamentation over the Corpse of Socrates—Vincenzo Camuccini

§1. In Plato’s Phaedo 118a, we read this description of the very last seconds before Socrates died from the poison that pervaded his body after he was forced to drink the potion of hemlock that the State had measured out for his execution:

Then he uncovered his face, for he had covered himself up, and said—this was the last thing he uttered—“Crito, I owe the sacrifice of a rooster to Asklepios; will you pay that debt and not neglect to do so?”

These last words ever spoken by Socrates, as quoted in Plato’s Phaedo, are referring to a ritual performed by worshippers of the cult hero Asklepios. It seems as if Socrates had made a vow to perform such a ritual.

§2. Right away, I need to say more about the ritual act to which Socrates is referring, which is the slaughtering of a rooster as a sacrifice to Asklepios. It is significant that, of all the followers of Socrates, Crito is being singled out here as the person who must perform the ritual act. This Crito—or, to spell his name in its non-latinized form, Kritōn—was a native of a dēmos or ‘deme’ by the name of Alopeke. All the territories controlled by the city of Athens were divided into districts known as dēmoi ‘demes’, and we know for a fact that Socrates as well as Crito was a native of the same dēmos, Alopeke. It makes good sense, then, that Socrates would have called out to his fellow demesman, Crito, to perform the sacrifice of a rooster, since all practices that we today would normally describe as “religious” were in those days grounded in local traditions. To say it another way, ancient Athenian practices associated with rituals and myths were strictly tied to wherever one was born and raised. That is why Socrates chose his demesman as the one who would stand in for him and sacrifice a rooster to Asklepios.

§3. There is external evidence that proves the existence of rituals involving such a sacrifice. In a minute, I will get into the details of this evidence, showing that the motive of worshippers in sacrificing a rooster to Asklepios was to seek a cure for an illness or to prevent an illness. But first, I need to review the essentials of the actual myth centering on Asklepios.

§4. In H24H 20§§29–33, I offered a brief sketch of these essentials. According to an early version of the myth, as I reported in the course of my sketch, Asklepios was a physician who had the power of bringing the dead back to life, and the divine order was so threatened by this power that Zeus himself incinerated the hero with his divine thunderbolt. Once dead, Asklepios became a cult hero, and, as such, he was believed to have the power to help those who worship him by stopping or preventing their illnesses.

§5. As I indicate in the analysis I just cited from H24H, Asklepios was still primarily considered to be a cult hero in the era when Socrates died, in 399 BCE, but, eventually, he evolved into a full-fledged theos or ‘god’ in the context of his sanctuaries, as at Epidaurus and at Pergamon: in the era of Pausanias, for example, who lived in the second century CE, Asklepios was by now primarily considered to be a god.

§6. One conventional way of worshipping Asklepios at his sanctuaries, as at Epidaurus and at Pergamon, was to experience a ritual incubation: you go to sleep after you pray to him and, in your sleep, you stay in mental contact with his sacred presence, hoping that Asklepios will give you some kind of response to your prayer. If he has heeded your prayer, you will be cured—but of course there is no guarantee that your prayer will be heeded.

§7. In this connection, I eagerly await the publication of an important book by Gil Renberg on rituals of incubation (see below for a bibliographical reference to his work). In this book, Renberg adduces a most relevant passage dating back to the second century CE. The source is Artemidorus, the author of a treatise on the interpretation of dreams, and here is what this source says about the ritual practice of sacrificing roosters to Asklepios:

ηὔξατό τις τῷ Ἀσκληπιῷ, εἰ διὰ τοῦ ἔτους ἄνοσος ἔλθοι, θύσειν αὐτῷ ἀλεκτρυόνα· ἔπειτα διαλιπὼν ἡμέραν ηὔξατο πάλιν τῷ Ἀσκληπιῷ, εἰ μὴ ὀφθαλμιάσειεν, ἕτερον ἀλεκτρυόνα θύσειν. καὶ δὴ εἰς νύκτα ἔδοξε λέγειν αὐτῷ τὸν Ἀσκληπιὸν ‘εἷς μοι ἀλεκτρυὼν ἀρκεῖ’. ἄνοσος μὲν οὖν ἔμεινεν, ὠφθαλμίασε δὲ ἰσχυρῶς· καὶ γὰρ μιᾷ εὐχῇ ὁ θεὸς ἀρκούμενος τὸ ἕτερον ἠρνεῖτο.

There was a man who prayed [eukhesthai] to Asklepios, [saying] that if he [= the man] gets through a whole year without any illness, he [= the man] would sacrifice [thuein] to him [= Asklepios] a rooster. A day later, he prayed again to Asklepios, [saying] that if he [= the man] did not develop any eye-ailment, he would sacrifice [thuein] a second rooster. Then, during the night, Asklepios appeared to him in a dream and said: “one rooster is enough for me.” What happened after that is that he [= the man] stayed free of any illness, except that he developed a very serious eye-ailment. So the god [Asklepios] was content with one prayer [eukhē] but said no in the second case.

Artemidorus Interpretation of Dreams 5.9

In this passage, the person who prays to Asklepios is vowing to sacrifice a rooster—or two roosters—‘to the god’ only after one year has passed. So the sacrifice here would be an act of thanksgiving that takes place only after a wish has been fulfilled—and only after the incubation, accompanied by a dream, had already taken place. There is a parallel that we find in Herodas Mimiamboi 4.11–20, where a rooster is sacrificed as an act of thanksgiving for a cure that Asklepios has already accomplished for the person who is saying her prayer.

§8. But there are other situations where the sacrifice of an animal to Asklepios, with or without the additional offerings of cakes to him, takes place before the incubation—and so before it is known whether the wish of the person who is praying will or will not be heeded. A shining example is the text of Inschriften von Pergamon 161AB, dated to the first half of the second century CE.

§9. In the case of Plato’s Phaedo 118a, I argue, the wording of Socrates indicates this kind of preliminary sacrifice to Asklepios—a sacrifice that takes place before an incubation.

§10. In my posting on 2015.03.27, I had in mind such a sacrifice when I offered my own interpretation of the last words of Socrates as quoted in Plato’s Phaedo. I interpreted these words as a mystical sign that was intended not only for Crito but also for all followers of Socrates: if they want to follow Socrates, they must be like his first followers and sacrifice a rooster to Asklepios. By sacrificing a rooster, you will be like the worshippers of Asklepios: you too, like those worshippers, will experience some kind of a ritual incubation. And then you will get a “wake-up call,” as it were, when you hear the crowing of roosters in the morning. Here is the way I said it in that posting:

After sacrificing a rooster at day’s end, sacrificers will sleep the sleep of incubation and then, the morning after the sacrifice, they will wake up to hear other roosters crowing. So, the words of Socrates here are referring to rituals of overnight incubation in the hero cults of Asklepios.

The relevant wording that I used in my posting, as I just quoted it, is a repetition of what I had said in H24H 24§46, where I went on to say:

So, Asklepios is the model for keeping the voice of the rooster alive. And, for Socrates, Asklepios can become the model for keeping the word alive.

My interpretation of the symbolism built into the last words of Socrates in Plato’s Phaedo 118a resembles, at least in part, an earlier interpretation that is mentioned in passing in a monograph by Mabel Lang (1977, pp. 28–29; see the bibliography below) concerning the material remains of the sanctuary of Asklepios in Corinth. In this monograph, she interprets the crowing of roosters in the morning as a symbol of a mystical awakening that follows an overnight incubation in the worship of Asklepios: “After the night of dreaming in the shrine it was the morning awakening that brought the cure, signaled and symbolized, as it were, by the cock’s crow” (p. 28). For a tracking of this and other such interpretations concerning the sacrifice of a rooster for Asklepios, I refer to the work of Emma Stafford (2008 in the bibliography, especially pp. 210–12).

§11. So, by now we have seen that worshippers of Asklepios could sacrifice a rooster to him, and the sacrifice would have to be correlated with a wish for health. Further, that wish would have to be formulated verbally in a prayer. As we saw in the passage I quoted from Artemidorus, the Greek word for ‘pray’ is eukhesthai. And such an act of praying would be what is normally called in English a vow. The idea is, I vow the sacrifice of a rooster in the context of my saying a prayer.

§12. Since Socrates is quoted as saying, literally, that ‘we owe the sacrifice of a rooster to Asklepios’, it is implied that Socrates had made a vow in the context of having said a prayer. He was praying to get something. He had wished for something. In terms of my argument, as I indicated in H24H 24§46, he wished for the Word to come back to life again, just as Asklepios himself had once upon a time come back to life, ready to heal those who are suffering from illness. And the Word, I argued, was Socratic dialogue. The followers of Socrates—and their followers in turn—must keep the conversation going, as it were. The conversations that Socrates started must be continued to keep the Word alive. I return once again to my formulation in H24H 24§46:

So, Asklepios is the model for keeping the voice of the rooster alive. And, for Socrates, Asklepios can become the model for keeping the word alive.

For me there still remains a big question to be asked about the vow of Socrates to sacrifice a rooster to Asklepios. Evidently, the sacrifice that he vowed must take place in the future. But when will the sacrifice take place? I argue that, in this case, the sacrifice would have taken place before an incubation. And the sacrificer owes such a preliminary sacrifice because he has already vowed it in prayer.

§13. To back up this point, I build on the research that is summarized in a chapter entitled ‘The Vow’ in a book by Emile Benveniste, Indo-European Language and Society, where the analysis focuses on Greek eukhesthai and its Latin cognate, vovēre.[1] The Latin verb vovēre can be translated as ‘vow’ in contexts where someone is praying to a divinity and asking for a favor to be granted, in return for which favor a vow is made to do something that is meant to gratify the divinity. Such a translation also applies in comparable contexts of the Greek verb eukhesthai. So, when you make a vow in a prayer, as expressed by way of the word eukhesthai, you are saying to a divinity that you will do or are doing or have done something in the hope that the divinity to whom you are praying will grant what you are wishing for. For a most pertinent example in the Iliad, I cite a situation where the hero Pandaros is being urged (misleadingly, by Athena in disguise) to make a vow-in-prayer as expressed by the verb eukhesthai (4.101: εὔχεο): this hero, by way of making a vow-in-prayer to Apollo, would be vowing that he would perform an animal sacrifice (4.102) in the hope that the god would grant him what he is wishing for, which is a safe homecoming (4.103). Pandaros then goes ahead and makes a vow-in-prayer (4.119: εὔχετο), vowing that he will in fact perform an animal sacrifice (4.120) in hopes of a safe homecoming (4.121). So, as Benveniste (1973 part 6 sub-part 4) says about the meaning of eukhesthai, ‘the “prayer” is not distinguished from the “vow”: it is one and the same operation’. Or, as I would prefer to say it, the wish-in-prayer is not distinguished from the vow-in-prayer. I can paraphrase in terms of the Latin noun vōtum, translated as ‘vow’, which is a derivative of the Latin verb vovēre, translated as ‘vow’. When you pray to a divinity, the word for what you vow to do is vōtum, but the word for what you wish for is likewise vōtum. In the case of the hero Pandaros in the Homeric Iliad, his wish—and therefore his prayer—is a failure, since he will soon be killed on the battlefield (5.290–296). I have more to say about this and other Homeric examples in an essay that I am preparing on the subject of Song 17 of Sappho.

§14. In that same essay on Sappho, I offer further interpretations concerning the concept of a “vow,” adjusting the formulation of Benveniste as I have outlined it here. These interpretations are indebted to the work of Leonard Muellner, who has shown that the English translation ‘vow’ for such words as eukhesthai works only in situations where the human who prays to a divinity is announcing an act that will happen in the future (see Muellner 1976:55–56). But the fact is, the act of gratifying a divinity can happen in the present or even in the past. What you announce in prayer does not have to be a promise about the future: it can also be an announcement about the present or even about the past (for a survey of Homeric examples, see Muellner 1976:36–37, 55–56). So, the translation ‘vow-in-prayer’ for eukhesthai does not cover the full range of meanings for these verbs. From here on, accordingly, I will translate eukhesthai simply as ‘announce-in-prayer’, not ‘vow-in-prayer’. And I must emphasize that, in each case of an announcement-in-prayer, the other side of the coin is a wish-in-prayer.

§15. But the divinity to whom a human prays is not obligated to heed a prayer. So, the divinity is not obligated to return the favor of, say, a sacrifice that is announced-in-prayer. The making of a sacrifice that you announce in prayer—whether that sacrifice takes place in the past, present, or future—does not guarantee that you will get your wish from the divinity to whom you are praying.

§16. In the mythical world of heroes, a wish expressed by a hero who makes an announcement-in-prayer to a divinity is often not heeded by the divinity. For example, at Iliad 2.402–429, when Agamemnon sacrifices an ox to Zeus (402–403, 422), he makes a wish-in-prayer, as expressed by the verb eukhesthai (411), that he will conquer the city of Troy (414–415) and kill Hector together with as many other enemies as possible (416–418)—all within the space of one single day (413). But Zeus refuses to bring this prayer to fulfillment (419)—even though the god accepts the offering of the sacrifice (420) and even though Agamemnon and his guests go ahead and cook the meat after killing the ox, dividing the beef among themselves and then feasting on it together (421–429). Although the narrative leaves it open whether, one fine day, Agamemnon will still succeed in his wish to conquer the city (419), it is made clear that the present wish-in-prayer, as performed by the hero on the occasion of this particular sacrifice, is a failure (again, 419). To paraphrase in Latin terms: the vōtum as a ‘wish-in-prayer’ is not granted here. And we have already seen another relevant example in the Iliad: when the hero Pandaros makes his announcement-in-prayer, as expressed by the verb eukhesthai (4.119), he says that he will perform an animal sacrifice (4.120) in hopes that Apollo, the god to whom he is praying, will grant him what he is wishing for, which is a safe homecoming (4.121). But the wish—and therefore the prayer—is a failure, since Pandaros will soon be killed on the battlefield (5.290–296). To paraphrase again in cognate Latin terms: the vōtum as an ‘announcement-in-prayer’ is a failure here because the same vōtum as a ‘wish-in-prayer’ is not fulfilled: the hero Pandaros will never return home safe and sound.

§17. Returning to the vow of Socrates, I ask: was his ‘wish-in-prayer’ fulfilled? My answer is yes: every time we engage in Socratic dialogue, Asklepios is once again granting for Socrates his fondest wish—that the Word will once again come alive.



Benveniste, E. 1969. Le vocabulaire des institutions indo-européennes. Paris.

Benveniste, E. 1973. Indo-European Language and Society. Translation by E. Palmer of Benveniste 1969. Coral Gables FL.

Lang, M. 1977. Cure and Cult in Ancient Corinth: A Guide to the Asklepieion. Princeton: American School of Classical Studies in Athens.

Muellner, L. 1976. The Meaning of Homeric ΕΥΧΟΜΑΙ through its Formulas. Innsbrucker Beiträge zur Sprachwissenschaft 13. Innsbruck. [link will soon be active]

Renberg, G. 2017. Where Dreams May Come: Incubation Sanctuaries in the Greco-Roman World. Leiden.

Stafford, E. 2008. “Cocks to Asklepios: sacrificial practice and healing cult.” Le sacrifice antique: Vestiges, procédures et stratégies (ed. V. Mehl and P. Brulé) 205–21. Rennes.


[1] Benveniste 1973 part 6 sub-part 4. For the online version of this chapter, see here.

Image by Tomisti via Wikimedia Commons in the public domain