A renewed public interest in Harmodios and Aristogeiton at the end of the fifth century likewise suggests that the leaders of the opposition against the Thirty were associated by democrats with the original tyrannicides. Vase painting provides the best evidence for this. Although three or four Athenian vases with representations of Harmodios and Aristogeiton survive from the period circa 470 to 450, no such decoration appears on extant vases or fragments for the next half century.60 Intriguingly, however, the tyrannicides reappear on five vases that date to circa 400, shortly after the overthrow of the Thirty. Three of the five are Panathenaic prize amphoras, which likely date to the festival of 402.61 Following the tradition of Panathenaic prize amphoras, each of the vases carries a picture of Athena depicted in the act of advancing upon an enemy: her left leg leads her right leg, in her left hand she holds a spear in the throwing position, and in her right hand she holds a shield. On the shield is painted an image of the statue group of Harmodios and Aristogeiton by Kritios and Nesiotes (see Figure 1.2). One interpretation of the image is that democratic Athens (symbolized by Athena) is protected (symbolized by the goddess’s shield) by tyrant killers.62 Given that the dēmos commissioned the amphoras and that the Panathenaia in question was likely the first to occur after the overthrow of the Thirty and restoration of the democracy—not to mention the fact that two other vase fragments carrying an image of the tyrannicides also date to circa 400—it seems certain that the figures of Harmodios and Aristogeiton were intended to allude to the recent overthrow of the Thirty, and to characterize their fall as an act of tyrannicide.
Another sign of renewed democratic interest in the original tyrannicides is the fact that, after the overthrow of the Thirty, the Athenians may well have granted new honors for the descendants of Harmodios and Aristogeiton. The only such honor securely attested in a fifth-century source is sitēsis (IG I3 131, circa 440). The orator Isaios, however, in a speech delivered in 389 (5.47), mentions among their honors not only sitēsis, but also proedria and ateleia. We cannot be certain that the Athenians first granted the honors of proedria and ateleia years after the grant of sitēsis, and even if they did so, it does not necessarily follow that the additional honors were granted after the overthrow of the Thirty. Nevertheless, this remains a very reasonable reconstruction of events, both chronologically (because the honors are first attested soon after the fall of the Thirty) and contextually (because they would fit well with the heightened interest in the tyrannicides at the time). Assuming it is true, it would suggest an increased appreciation of the value of tyrant killing in preserving democracy, and further support the conclusion that Athenian democrats considered the Thirty to have fallen in an act of tyrannicide.
Figure 1.2. Panathenaic prize amphora, circa 402. Photo © The Trustees of the British Museum.
In order to facilitate their efforts to mobilize in defense of the democracy, Athenians swore an oath to kill tyrants and reward tyrant killers—in other words, to become tyrannicides, a point made explicit by their pledge to treat a fallen assassin “just like Harmodios and Aristogeiton.”63 After swearing the oath, many Athenians did in fact successfully mobilize against the Thirty and killed many of their supporters, in a movement characterized by democrats as a collective act of tyrannicide. The logical consequence is clear: Athenian democrats attributed the fall of the Thirty, at least in part, to the fact that all Athenians had sworn the oath of Demophantos.
A second indication that the oath of Demophantos was largely responsible for the success of the democratic response is that the Athenians, very soon after the mobilization against the Thirty, swore another mass public oath, the function of which was virtually identical to that of the oath of Demophantos. In this oath, known as the amnesty oath, they pledged “not to remember past wrongs” (μὴ μνησικακεῖν).64 As I attempt to demonstrate, this oath too affected the revolutionary thresholds of the participants and thus the ability of democrats to mobilize en masse, but whereas the oath of Demophantos lowered thresholds in order to facilitate stasis, the amnesty oath raised thresholds in order to make stasis much less likely.
By swearing the amnesty oath, the Athenians generated common knowledge of a credible commitment “not to remember past wrongs.” Andokides (Myst. 90) makes it clear that all Athenians participated, describing it as “the oath in which the whole city joined, the oath which you swore one and all after the reconciliation: ‘ … and I will not remember the past wrongs (οὐ μνησικακήσω) of any citizen except for the Thirty, the Ten, and the Eleven.’ …” We do not know how the Athenians actually swore the oath, but it is certainly conceivable that they did so “by tribe and by deme,” as in the case of the oath of Demophantos, which provided the only real Athenian precedent for such a large-scale oath ritual. But even if the amnesty oath was sworn in a different manner, the important point remains that everybody knew that everybody knew that everybody knew that all Athenians had solemnly pledged “not to remember past wrongs.”
This pledge was tantamount to a pledge not to engage in violent acts of vengeance against members of the oligarchic faction. Traditional interpretations of the phrase μὴ μνησικακεῖν stress its relevance to “partisan” activities directly involving the law courts: according to this interpretation, a citizen pledged not to bring anyone (with a few exceptions) to trial for crimes that he may have committed during the reign of the Thirty, and a juror pledged not to convict individuals for such crimes.65 Three points, however, strongly suggest that the phrase “not to remember past wrongs” had a broader relevance, referring to violent acts of vengeance generally speaking, not just to those involving the law courts. First, if the Athenians had intended to refer only to actions involving the courts, they easily could have crafted a more specific law that explicitly forbade the indictment of individuals for actions they may have committed during the time of the Thirty. Second, the author of the Platonic Seventh Letter (336e) directly associated the verb μνησικακεῖν (“to remember past wrongs”) with σφαγή (slaughter, butchery). The term σφαγή specifically refers to intentional killing and appears to have been associated with political violence.66 In the decree of Patrokleides (Andok. Myst. 79), for example, σφαγή is paired with tyranny. And Xenophon (Hell. 4.4.2) called the bloody massacre in Korinth in 392 a σφαγή. Third, according to the author of the Ath. Pol. (40.2), Archinos, a leading democratic figure in the period immediately following the civil war, convinced the members of the boulē to execute an unnamed individual because he “began to remember past wrongs” (ἤρξατο μνησικακεῖν). The passage in no way suggests that the unnamed individual did so in the law courts.67
Thus, by swearing the amnesty oath, the Athenians generated common knowledge of (at least apparently) credible commitment not to engage in violent acts of vengeance against members of the oligarchic faction. In order to understand what that has to do with the potential effectiveness of the oath of Demophantos, it is necessary to determine how an oath containing the pledge “not to remember past wrongs” might have helped prevent stasis and thus helped to defend the recently reestablished democracy. Let us posit that every Athenian has a “stasis threshold” that represents the number of individuals who must participate in stasis activity before he does. An individual with a stasis threshold of 2, for example, will participate only if two other individuals (out of a population of ten) have already done so. This individual wants neither to initiate stasis nor to join in its earliest stages, for he is afraid that an insufficient number of other individuals will follow him and thus he will be punished as a revolutionary. If 20 percent of the population has already joined in the stasis, however, he too will join, in order to protect himself or those with whom he is in sympathy, and because he now believes that others will follow.68 An extreme radical on the other hand, an individual who wants to reignite the civil war and eliminate members of the rival faction, will have a stasis threshold of 0: he believes that, if he commits some conspicuous act of violence, ot
hers will follow and thus the polis will be engulfed in stasis.
The most important consequence of the fact that all Athenians swore the amnesty oath is that most individuals would have raised their individual stasis thresholds and thus participated in a stasis later than they otherwise would have. Most individuals, that is, would have waited until more people joined in the stasis before they themselves decided to join. The oath, by generating common knowledge of an apparently credible commitment not to engage in acts of stasis, would lead most individuals to think that, if they were to engage in such acts, others might not follow them, because those others would think that others might not follow them, and if an insufficient number of individuals should take part, those who had done so would be more easily punished as revolutionaries or agitators. Since credibility is never perfect, however, an individual Athenian might not raise his stasis threshold by much, but he would still probably raise it, say, by 1 (e.g., from 2 to 3).
It is easy to see how even a small increase in an individual’s stasis threshold could prevent an initial act of stasis from initiating a “stasis bandwagon.” Imagine the following stasis threshold sequence: {0, 1, 2, 3, 3, 3, 3, 4, 5, 7}. The person in the first position in the sequence, with a threshold of 0, is a radical who wants to provoke civil strife; perhaps he hopes to eliminate those citizens who sided with the oligarchs. He thus commits some conspicuous act of violence in the hope that other individuals will follow. And they will, for after he acts the person in the second position in the sequence, with a stasis threshold of 1, will join in, and his action will draw in the next person, with a threshold of 2, and so on. Very soon the entire population could find itself taking part in the stasis, even if a majority might not wish to do so.69 If, however, after swearing the amnesty oath, all but the most radical individuals raised their stasis thresholds by 1, the threshold sequence would look very different: {0, 2, 3, 4, 4, 4, 4, 5, 6, 8}. In this scenario, the radical represented in the first position in the sequence would still act in the hope that others would follow him. Since there is no citizen with a stasis threshold of 1, however, nobody would follow him and his efforts to initiate a stasis bandwagon would fail.70
The amnesty oath was thus functionally similar to the oath of Demophantos. Both represent an attempt to control sparks and bandwagons in order to defend the democracy. The Athenians swore the oath of Demophantos in order to increase the likelihood that someone would go first and initiate a revolutionary bandwagon. They swore the amnesty oath in order to prevent someone from going first and initiating a stasis bandwagon. They are two sides of the same coin, both working to defend and preserve the democracy.
The similarity of both means and ends between the amnesty oath and the oath of Demophantos suggests that—in the minds of the Athenians, at any rate—the successful response against the Thirty should be attributed in part to the oath of Demophantos. After the fall of the regime in 403, the Athenians must have concluded that the oath of Demophantos had been shown to be an effective tool, and they decided to use the tool once again in the amnesty oath to achieve a very similar end. If the goal of the oath of Demophantos was to facilitate democratic efforts to mobilize in response to organized attempts to overthrow the democracy, the conclusion that the oath had actually worked must have been based on the conviction that it did, if fact, facilitate mobilization. Since the only such mobilization that postdated the oath of Demophantos and antedated the amnesty oath was that conducted against the Thirty, the conclusion is clear: Athenian democrats saw a causal connection between the swearing of the oath of Demophantos and the mobilized response to the Thirty.
Table 1.1. The Law of Eukrates and the Decree of Demophantos
A final point in support of this view of the oath is that the next time the Athenians feared an attempt to overthrow the democracy they passed a law that virtually quoted the language concerning tyranny contained in the oath of Demophantos. This was the law of Eukrates, dating to the spring of 336, when democrats feared a pro-Macedonian domestic coup. I will provide a full analysis of that law in chapter 3. So all I demonstrate here is that it does, in fact, incorporate the important language found in the oath and decree of Demophantos.71 Table 1.1, based on that crafted by Martin Ostwald in his classic article on Athenian anti-tyranny legislation, makes the case most efficiently.72 In the right-hand column, (a) refers to language found in the rule proper, (b) refers to the oath.
The verbal echoes make it clear that the provisions concerning tyranny in the law of Eukrates should be interpreted as a reaffirmation or a reminder to potential revolutionaries of the decree and the oath of Demophantos. Such a reaffirmation was almost certainly predicated on the widespread consensus that the earlier oath had worked, and indeed there is other evidence (Dem. 20.159–162; Lykourg., Leok. 124–27) that fourth-century Athenians thought highly of its effectiveness in generating mass action in defense of the democracy. Since, as noted above, the only such action that postdated the decree of Demophantos was that against the Thirty, one should again conclude that Athenian democrats causally connected the swearing of the oath with the success of the response to the Thirty.73
The natural conclusion of the arguments presented above is that the successful democratic response against the Thirty should be attributed, at least in part, to the fact that all Athenians swore the oath of Demophantos. I have shown, first, that the Athenian democrats attributed the fall of the Thirty to an act of tyrannicide; second, that very soon after the successful mobilization against the Thirty, they swore another public oath (the amnesty oath), the function of which was complementary to that of the oath of Demophantos; and third, that the next time the Athenians feared an oligarchic attempt to overthrow the democracy, they promulgated a law that virtually quoted the anti-tyranny language contained in the oath of Demophantos. These three points might not be conclusive individually, but their cumulative weight seems to validate the thesis.
Conclusion
The viability of democratic governance in a Greek polis during the Classical period depended on it supporters’ ability to solve a revolutionary coordination problem. If the supporters were able to solve that problem, they could capitalize on their numerical superiority and respond to a successful coup or, better still, prevent a coup from succeeding in the first place. Should the democrats be unable to coordinate in defense of their regime, however, their superior numbers would be meaningless and the anti-democrats would quickly dominate the polis. A fundamental question for historians of ancient Greek democracy should thus be, how did the citizens of the various democratically governed poleis solve, or attempt to solve, the revolutionary coordination problem?
In this chapter, I have attempted to demonstrate that Athenian democrats in the late fifth century were able to solve the problem in large part because of the fact that all Athenians swore the oath of Demophantos. The public performance of that oath accomplished two ends. First, it generated common knowledge of credible commitment to defend the democracy. Individual democrats were thus willing to oppose an anti-democratic coup earlier than otherwise would have been the case, for they believed that other democrats would follow them. Second, the oath gave an incentive to particularly brave individuals to take the all-important first step in defense of the democracy—to “kill a tyrant” and thus initiate a revolutionary bandwagon. The oath directly contributed to the successful mobilization against the Thirty: Thrasyboulos and his men were confident enough to go first, and other individuals were subsequently confident enough to follow them. As a result, the democrats were able to capitalize on their numerical superiority and reclaim control of the polis.
This analysis suggests that the remarkable stability of the fourth-century Athenian democracy should ultimately be attributed to the fact that all Athenians swore the oath of Demophantos, and to the successful mobilization against the Thirty Tyrants that followed. After those events, anti-democrats would have known that, should they stage a coup, democrats would in fact be able to mobilize en masse in defense of their poli
teia; and since the number who favored democracy was greater than the number who opposed it, the former would almost certainly prevail. Thus anti-democrats were deterred from staging further coups.74 This strategic dynamic goes a long way toward accounting for the persistence of democracy in fourth-century Athens.75
* * *
1 The oligarchs’ conspiracy: Lys. 12.43–44. The oligarchs’ attempt to make Athens’s politeia similar to Sparta’s: Krentz (1982: 57–68).
2 The assembly was compelled to vote to change the politeia: Ath. Pol. 34.3; Lys. 12.75. Specifically, the assembly ratified the decree of Drakontides (Xen. Hell. 2.3.2, 2.3.11; Lys. 12.73; Diod. Sic. 14.3.5). There is some controversy over the content of that decree. But it almost certainly called for the selection of thirty men both to craft new laws and to act as a provisional government until a new, permanent constitution was established. On the decree of Drakontides, see Rhodes (1993: 434–35); Ostwald (1986: 476–78); Krentz (1982: 50).
3 The ancient sources do not agree on the number of men who accompanied Thrasyboulos on his march from Thebes. Numbers range from “not more than 50” (Nep. Thr. 2.1) to 70 (Xen. Hell. 2.4.2). The other ancient sources (conveniently collected in Krentz [1982: 70–72n4]) are [Arist.] Rhet. ad Al. 8 (50 men); Aristid. Or. 1.254 Lenz-Behr (just over 50 men) and 43.556 Dindorf (70 men); Paus. 1.29.3 (60 men); Plut. Mor. 345d (70 men). Krentz believes that Thrasyboulos left Thebes with 30 men. For a detailed account of the life of Thrasyboulos and his role in Athenian politics, see Buck (1998).
4 The rebels’ numbers increasing to 700 men: Xen. Hell. 2.4.5; to 1,000 men: Xen. Hell. 2.4.10; to well over 1,200 men: Diod. Sic. 14.33.1. The political status of the men at Phyle is a vexed question. For a cogent and persuasive discussion, see Taylor (2002b). Taylor argues that about 60 percent of the men besieged at Phyle were citizens and that about 90 percent of the men at Phyle at the time of the march to the Piraeus were citizens. She accepts important conclusions reached by Raubitschek (1941) and refutes Krentz’s conclusion (1982: 83–84) that the vast majority of the men first at Phyle were not Athenian citizens.
Death to Tyrants! Page 7