Pruning of the People: Ostracism and the Transformation of the Political Space in Ancient Athens

Abstract

Athenian ostracism has long captured democratic imaginations because it seems to present clear evidence of a people (demos) routinely asserting collective power over tyrannical elites. In recent times, ostracism has been particularly alluring to militant democrats, who see the institution as an ancient precursor to modern militant democratic mechanisms such as social media bans, impeachment measures, and lustration procedures, which serve to protect democratic constitutions from anti-democratic threats. Such a way of conceptualizing ostracism ultimately stems from Aristotle’s “rule of proportion,” or the removal of “outstanding” individuals in a polity who threaten to disturb the achievement of communal eudaimonia (Aris. Pol. 1284^a). However, this way of interpreting the institution only presents a truncated view, one which is overly centered on the ultimate expulsion of an individual from the polity, rather than on its broader contextual telos—the transformation of the ostracized individual and of the community. To move past this simplified view, this paper considers all elements of ostracism with equal force, and argues that ostracism offered a shared opportunity and shared space for all members of the polis—citizens, non-citizens, and elite members alike—to reform the character of the subject individual and to instill and reaffirm democratic values in the community.

1. Introduction

Athenian ostracism has long captured democratic imaginations because it seems to present clear evidence of a people (demos) routinely asserting collective power over tyrannical elites by temporarily ejecting them from the polity. In recent times, ostracism has been particularly alluring to militant democrats, who see the institution as an ancient precursor to modern militant democratic mechanisms such as social media bans, impeachment measures, lustration procedures, etc.—all of which serve to protect democratic constitutions from anti-democratic threats. Such a way of conceptualizing ostracism ultimately stems from Aristotle’s “rule of proportion,” or the proactive removal of “outstanding” individuals in a polity (especially with regard to excessive wealth, social capital, and/or prestige) who threaten to disturb the long-term achievement of the final end of the community (Aris. Pol. 3.13, 1284^a). However, when we look at ostracism in depth, this way of interpreting the institution only presents a truncated view, one which is overly centered on the ultimate expulsion of an individual from the polity, rather than on its broader contextual telos—the transformation of the ostracized individual and that of the community.

To move past this simplified Aristotelian view, this paper will consider all elements of ostracism with equal force: from the extended and expressive process of preliminary public debate, to the semi-ritualized communal casting procedure (ostrakophoria), to the eventual expulsion of the ostracized individual, and finally to his often celebrated return to the polity.1 This article will therefore argue that ostracism, though undeniably an expulsive and defensive institution—as Aristotle and militant democrats rightly deem it to be—was not exclusively so. Rather, ostracism offered a shared opportunity and shared space for all members of the polis—citizens, non-citizens, and elite members alike—to reform the character of the subject individual and to instill and reaffirm democratic values in the community. In turn, this article suggests that the reason modern ostracism-adjacent measures (as proposed by militant democrats) have limited potential to protect democracy today is because they do not effectively drum up pro-democratic sentiments in either targeted individuals or in the broader population. To garner lessons from Athenian ostracism, then, the institution must be understood as more than just an isolated, expulsive procedure; it also had as its objective the transformation of character and of community.

The first part of this paper will briefly introduce the technical procedure of ostracism, as transmitted by ancient sources, before examining how Aristotle analyzed the institution, and how this analysis has since guided modern scholarship on ostracism, especially in militant democratic literature [1] (p. 319). The Aristotelian view, it will be argued, is “truncated” and ignores the larger reformative projects of ostracism and the ostrakophoria, which are processes of instantiating demokratia through political ritual. The next part of this paper will examine ostracism through a more holistic, phenomenological framework, with reference to the rehabilitative, reformative, and expressive elements of the institution. Contrary to both Aristotle and modern militant democrats, the process of Athenian ostracism was inherently connected to rebuilding the character of the ostracized individual, as well as to inculcating a sense of collective activity and pro-democratic civic identity. Having made this case, the paper will conclude with a suggestion for militant democrats on the lessons they should draw from Athenian ostracism: namely, to treat it as a site of civic education and transformation, a point which they have heretofore largely set aside in their analyses.2

2. The Process of Ostracism

Before looking at how Athenian ostracism has been interpreted by modern political theorists, it is first necessary to understand how the process itself was conducted. Philochorus, an Atthidographer of the 3rd century BCE, provides one such detailed technical description.3 He writes:

Ostracism takes place as follows. (1) Before the eighth prytany, (2) the people vote on whether it is necessary to hold an ostracism. If it is necessary, the agora is fenced in with boards, leaving ten entrances, through which the people enter in their tribes, and deposit their sherds [ostraka] with the writing facing downwards. (3) The nine archons and the council oversee the process. (4) When the sherds have been counted to determine who has the most votes (and not less than 6000), (5) then this person must, after settling his personal commitments, leave the city within ten days, for a period of ten years (this was later reduced to five years). He is allowed to receive income from his possessions, but he must not come nearer [to Athens] than Geraestum, the headland on [the coast of] Euboea. (Philoch. Frag. 79b)

[4].

While the procedural steps could be reconstructed from this account in a number of different ways [5] (p. 129). There are at least five distinct elements of ostracism that Philochorus mentions: (1) a two-stage process, (2) an open and inclusive procedure, (3) oversight by outside officials, (4) a participation quorum, and (5) regularized penalties. Ostracism, then, can be considered a highly regularized and regulated political procedure, despite the fact that no formal accusations or defense speeches (akin to courtroom debates) were made, and most debate happened “outside” official institutional spaces ([Andoc.] 4.3), [6] (p. 160). It was both a dynamic affair, involving a large number of motivated individuals, while also being a restricted one, limited to official times and conventions [7].4

To better understand the proceduralized dynamics of ostracism, let us consider the first two “limits” mentioned by Philochorus. The two-stage process had two effects: first, it meant that only if a majority agreed to hold an ostracism would one take place—ensuring that there needed to be at least initial shared grievances among the citizens about someone (or something) wrong with the polis. Secondly, the procedure ensured that there would be ample time between the first vote (‘Should we agree to hold an ostracism?’) and the second (‘Whom shall we ostracize?’) for informal deliberation about who should receive the penalty [8] (p. 162). Moreover, though not mentioned in Philochorus’ account, the opportunity to ostracize was only granted once a year, which further limited the institution. In turn, the rarity of the event made it into a sort of social spectacle about which all could—and would—debate.5

Like many Athenian institutions, the ability to ostracize was open to all male citizens, who were protected from potential intimidation by the public nature of the event, by the anonymous ostraka sherds that concealed individual votes, and by high officials observing the process in its entirety [7].6 Moreover, the high participation quorum ensured that it was not just a vocal minority who were able to target and expel their political rivals, but rather a broad swath of the Athenian voting population who all felt there was something wrong in the polity (though they often disagreed about its cause) [5] (p. 4). In fact, the number of people involved in ostracism was, by the quorum, necessarily larger than the number constituting the courts.

Lastly, Philochorus mentions that the subjected individual received a standard penalty: he was required to leave the city for the mandated ten (or fewer) years, but was otherwise allowed to retain his property and status. One could argue, then, that this was more a temporary (albeit harsh) sting for the transgressor than a lasting scar, given that his family was permitted to remain in the city, his property was left untouched, and he could continue to profit from his revenues [8] (p. 233). Moreover, after the stated allotment of time, he was reinstalled with all previous rights of citizenship. The penalty was, in Sara Forsdyke’s words, “symbolic” insofar as it asserted the demos’ power over the elite, without needing to unnecessarily provoke the subjected individual’s malice towards the people [8] (pp. 149–165). The ostracized party still had to relocate from the polity for a period, but he had a greater incentive to patiently wait for his reintegration into the civic body than risk fomenting public ire by plotting revenge. In sum, the process of ostracism was technical, regularized, and rare. Only nine ostracisms are historically confirmed, starting with that of Hipparchus in 488-7 and ending with Hyperbolus in 416-5, though the option to ostracize an individual likely continued to be set forward in an Assembly (Ekklesia) meeting until sometime in the late fourth century [9] (p. 163).

3. The Aristotelian Concept of Ostracism

Now that the technical procedure of ostracism has been explained, we are better able to review the dominant way in which the institution has been interpreted, as largely determined by Aristotle [10] (pp. 92–94). To Aristotle and his followers, ostracism in Athens seemed to serve two related, but distinct, defensive purposes. On the one hand, allegedly, it was introduced by Cleisthenes to protect the fledging Athenian democracy from tyrannical individuals. Yet, later Aristotle describes ostracism in broader terms, claiming that it would be employed against anyone whose greatness is so preeminent that “they can be no longer regarded as part of the state” (Aris. Pol. 3.13, 1284^a) [11]. Aristotle thus leaves his interpreters with a (perhaps intentional) ambiguity: how sensitive was ostracism to democratic “threats”? Was it meant to target and expel only clearly anti-democratic figures, or did it also aim at anyone who the people thought could pose a threat to democracy? Or, was there even a distinction to be made between “potential” tyrants and proven “tyrannical individuals?” Regardless of technicality, this understanding of the institution as an anti-tyrannical expulsive measure maps on to the way Athenian ostracism is invoked in political science literature today. The remainder of this section will further detail the Aristotelian view, and then connect this view with later scholars’ understandings of ostracism.

In the Athenian Constitution, written (most probably) by a follower of Aristotle, it is claimed that ostracism in Athens had a blunt objective: it was introduced to prevent tyrannic overthrow. Ostracism, the author of the Constitution writes, “had been passed from a suspicion of those in power, because Peisistratus had started as leader of the people and strategos, and become tyrant” (Aris. [Ath. Pol.] 22). Accordingly, the first person “legally” ostracized was Hipparchus, and in the following three years the demos continued to ostracize the “friends of tyrants” [12]. Yet, even within the first generation of ostracism, the institution came to be directed not only against tyrants and their associates, but also against “any one who seemed to be more powerful than was expedient” (Aris. [Ath. Pol.] 22). Ostracism may have thus emerged, under the Aristotelian view, as an anti-tyrannic mechanism, in tandem with other more explicitly anti-tyrannic measures (e.g., capital punishment, permanent confiscation of property, refusal of burial rites), but it quickly came to be used against anyone who might pose a threat to the public order [8] (pp. 153–154).

Hence, in his Politics, Aristotle claims that:

If, however, there be some one person, or more than one…whose virtue is so pre-eminent that the virtues or the political capacity of all the rest admit of no comparison with his or theirs, he or they can be no longer regarded as part of a state; for justice will not be done to the superior, if he is reckoned only as the equal of those who are so far inferior to him in excellence and in political capacity. Such a man may truly be deemed a God among men. Hence we see that legislation is necessarily concerned only with those who are equal in birth and in capacity; and that for men or pre-eminent excellence there is no law—they are themselves a law. Anyone would be ridiculous who attempted to make laws for them […] And for this reason democratic states have instituted ostracism; equality is above all things their aim, and therefore they ostracized and banished from the city for a time those who seemed to predominate too much through their wealth, or the number of their friends, or through any other political influence. (3.13, 1284^a)

[11]7

Whether it be the case of a single excellent man, or of an otherwise excessively wealthy or popular individual, democrats’ apparent concern in subjecting an individual to ostracism remains the same: to prevent him from being able to subsume or circumvent the laws of the state, however biased those laws may be. In short, as Aristotle sees it, where laws were ineffective for preserving the ruling order, ostracism was. Hence, later Aristotle will assert that ostracism is based upon a “kind of political justice,”, e.g., ensuring “equality” in the democratic polis, but acknowledges that it is only a second-best solution when a regime starts to go awry, when “one or more persons have a power which is too much for the state” (1284^b; 5.3, 1302^b). In short, Aristotle sees ostracism as a sort of natural and necessary solution—a “safety valve”—in a regime whenever the rulers (be they singular, few, or many) believe there is some individual who could in theory, or has in practice, used his power against the state.8 “Such a measure is just and expedient,” Aristotle writes, “but it is also clear that it is not absolutely just” (Aris. Pol. 3.13, 1284^b). And indeed, despite the reigning ideals of equality in Athens, economic, political, and rhetorically skilled elites still had looming power over the demos. According to Aristotelian thinking, then, ostracism was introduced in Athens with tyrants in mind, as those were the ones “not in proportion” to the state (1284^b10), but soon came to be used against potential tyrants as well. Whether Aristotle was correct in his judgment has been subject to ongoing historical debate, as tyranny is “never explicitly linked to ostracism in fifth-century sources,” but Aristotle’s understanding of ostracism has nonetheless largely guided the way the institution has been interpreted through the millennia [15] (p. 502), refs. [9,16,17].

In sum, Aristotle sees two related “angles” to ostracism: it was introduced to remove undeniably tyrannical individuals from the polity, but once introduced, it also ended up granting the demos the opportunity to determine and expel anyone it considered threatening to the democratic regime. Within this analysis, however, little thought is devoted to how the community comes to its decision or how the ostracized individual is treated post-expulsion. Most, if not all, pre-modern and contemporary discussions of ostracism have likewise overlooked these questions and instead taken up the Aristotelian viewpoint, by singling out the legal expulsion of an elite citizen as constituting the ultimate telos of ostracism. This single-minded emphasis on the expulsive side of the institution is exemplified by Benjamin Constant’s “Liberty of the Ancients Compared with that of the Moderns” (1819) [18], in which Constant, a leading theorist of republican liberty in post-revolutionary France, lambasts his contemporaries for trying to bring back ostracism à la ancient Athens. A similar view of ostracism is developed in classicist Sara Forsdyke’s seminal Exile, Ostracism, and Democracy (2005) [8] and has been more recently taken up by militant democrats, who see the potential in the Athenian model of democratic self-protection against (aspiring) anti-democrats. In the following section, I articulate these views in more detail, before moving on to a broader conceptual critique.

4. The Expulsive View: Benjamin Constant to Militant Democrats

Though nearly all societies have used political banishment or social exile—in one way or another—to discipline those who pose a threat to the social order, maintain their political systems, and/or craft a civic identity, nonetheless the rigorously codified procedure of ostracism stands out amongst other measures of social and political exclusion due to its sheer entrenchment in the Athenian political system [19]. For this reason, the expulsive-centered view of ostracism has captivated political theorists for centuries—either as source of denigration or inspiration. Benjamin Constant, for example, was deeply concerned about his contemporaries invoking ideals of “ancient liberty” to legitimize their seemingly authoritarian demands, which put the “empowerment of the state over the individual” [20] (p. 223). In particular, he was worried about the attempted introduction of ostracism into France in 1802, whose proponents, Constant noted, were struck by the “liberty of Athens and all the sacrifices that individuals must make to preserve this liberty” [18]. To Constant, ostracism “worked” in Athens because it was based “on the theory that society had complete authority over its members”—a value that the moderns had forgone as nations grew and the commercial world became the primary zone of individual development [18]. Unsurprisingly, therefore, Constant abhors the attempt to banish citizens under the pretense of safeguarding the public welfare. He writes:

No-one has the right to exile a citizen unless he is legally convicted by a regular court, following a law which explicitly assigns the penalty of exile for the action of which he is guilty. No-one has the right to tear the citizen from his county, the owner from his possessions, the merchant from his trade, the husband from his wife, the father from his children, the writer from his studious meditations, the old man from his accustomed way of life. All political exile is a political abuse. Any exile pronounced by an assembly for alleged reasons of public good is a crime that this assembly commits against the public good, which resides only in respect for the laws, in the observance of forms, and in the maintenance of safeguards

[18].

Though Constant did not fully reject Athens—and in fact approbated its commercial energy, regular jury trials, and zones of free speech—to him, Athens’s “use of ostracism made clear that she conformed in essentials to the ancient construct for the state” [20] (p. 225). To him, ostracism was prima facie illegitimate because it cut off a member from the community, without regard to the individual’s rights, safety, or overall standing within a political system. He saw ostracism, then, as a primarily expulsive institution, with little to be gained from its use (even in ideal conditions, such as Athens).

Though Constant disdains ostracism as a legitimate method of political control, more recent theorists have viewed the institution in a positive light, as it both gave Athenians the opportunity to step into elite-dominated politics and empowered them to do so.9 Forsdyke, for example, characterizes ostracism as a form of demotic conflict resolution which served to end the “aristocratic politics of exile,” whereby sets of elites would extralegally—and often violently—expel political rivals from the polity [21]. In turn, this would set off a series of destabilizing contests between those in and out of power to wrestle over political control. Once the power to ostracize was granted to the demos, however, the citizens were able to use this potential threat as a “symbolic reminder” to the aristocracy of popular political supremacy, as the people were the ones who ultimately decided the “winners” of elite contests [22].10 However, Forsdyke’s mode of analysis, similar to that of Aristotle and Constant, overlooks the fact that the ostracized party was expected (and often invited) to return to Athens at the appropriate moment––be it after ten years, or whenever the demos authorized. Return and reintegration, as will be argued in the following sections, were considered just as essential to ostracism as was the expulsion. Moreover, though Forsdyke does recognize the “ritualistic” (i.e., the symbolic or expressive) side of ostracism, she still treats the “ritual” as being focused on proving the demos’ negative power to expel elites, rather than on the positive, community-building aspect of the ostrakophoria. Despite the more expansive view of ostracism than that of her predecessors, Forsdyke still presents a “truncated” view.

Contemporary militant democrats––who cite Aristotle and Forsdyke’s accounts of ostracism almost exclusively––have unsurprisingly latched on to the expulsion-centered view, seeing it as a source of inspiration for modern protective democratic measures [24,25,26,27]. In his Theory of Militant Democracy, for example, Alexander Kirshner briefly touches on ostracism, and understands it as the Athenians’ “popular response” to threats by antidemocratic elites. Yet, he worries about any protective measure that would ultimately “silence or ostracize” political enemies, as if silence and permanent expulsion were integral parts of Athenian ostracism [28] (p. 25). Similarly, in an article from 2016, Jan-Werner Müller categorizes Athenian ostracism as a form of citizen disenfranchisement, thus belonging to a category which he deems too blunt and permanent to be implemented into democratic regimes today [29].11 More recently, Müller has found promise in conceptualizing ostracism as a measure of “individual militant democracy,” whereby individuals who exhibit a pattern of anti-democratic behavior might be barred from entering political competitions, or be subject to more intense scrutiny as their terms in office conclude. Yet he seems to disregard the fact that Athenian ostracism was a community-building mechanism that helped to both promote and reaffirm collective pro-democratic sentiments [26]. Both Kirshner and Müller, then, worry about the inflexibility of ostracism as a measure of social or political stigma, as well as its perceived permanence, without regard to the institution’s more positive, transformative aspects in the Athenian context.

In contrast to the views outlined above, Anthoula Malkopoulou emphasizes the more participatory and popular element involved in Athenian ostracism, which she has variously called a “negative election procedure” and a system of “negative representation.” In her article, “De-presentation Rights as a Response to Extremism,” Malkopoulou highlights how ostracism can provide citizens a way (or right?) to “cast a negative judgement and express their objection to specific representatives” [30]. Although a militant democrat herself, Malkopoulou distinguishes herself from her colleagues insofar as she emphasizes that members of the demos, not political courts or high officials, ought to be the ones to decide who needs to be expelled or debarred from engaging with politics [5] (p. 11). Yet, Malkopoulou’s analysis is still framed in terms of the expulsion of an individual, with little thought given to how the ostracized individual would be rendered “more fit” for democracy through the process of ostracism.

Drawing from militant democratic theory, American legal scholars have also latched onto ostracism. Like militant democrats, legal theorists tend to understand ostracism as providing a helpful historical example of a form of electoral disqualification for those who pose a clear threat to “relatively minimalist, electorally-focused conception[s] of democracy” [31] (p. 1). For example, Tom Ginsburg, Aziz Z. Huq, and David Landau, in their blueprint for the “optimal” disqualification scheme for anti-democratic politicians, briefly touch on ostracism, if only to suggest that “disqualification is as old as democracy itself” [31] (p. 18). Likewise, Alex Zhang, in his 2021 article “Ostracism and Democracy,” argues that ostracism-adjacent measures, such as legal electoral disqualification or restrictions on public appearances, should be introduced to protect weakened democracies [24] (246–247, 258). Although Zhang and Ginsburg, Huq, and Landau alike argue that ostracism presents a compelling framework for dealing with anti-democratic actors, none address exactly how the ostracized individual would be rendered less threatening in the long term by such measures, nor how citizen control of ostracism would actually render democracy more stable.

All such militant democratic views—be they individualist, popular, or legalist—offer strong reasons to draw from Athenian ostracism. Nonetheless, they focus too narrowly on the expulsive side of the institution, namely, the removing, barring, or disqualifying of individuals, however temporarily, from the political fray. In doing so, these theorists ignore the broader reformative projects and effects inherent to the Athenian practice of ostracism. In short, expulsive theorists are overly focused on a single result of the procedure, to the exclusion of the equally important contextual “before” and “after” of both the community and the ostracized individual [32] (p. 124). In turn, this narrow focus on the supposed immediate benefit of removing harmful individuals from the community might call into question the overall sustainability of other militant mechanisms they seek to introduce. If they overlook the “softer” elements of ostracism—which the Athenians considered necessary features of the institution—do militant democrats do the same with other defensive democratic procedures, such as social media bans, impeachment measures, lustration methods, or restrictions on certain political rights?

5. Doing Democracy: Individual Reform and Ritual Ostrakophoria

Fifth and fourth century Athenians understood ostracism to be a “live” institution of their political reality, always available for use if the need should arise [33] (p. 94). But, perhaps frustratingly, actual instances of banishment do not seem to track reliably onto how much power or prestige particular politicians seem to have held in the fifth and fourth centuries [34]. It has been considered somewhat of an unresolved mystery, for example, why an actual ostrakophoria did not occur in the first few decades after it was supposedly introduced by Cleisthenes, though circumstances might have seemed to reasonably condition one [34] (p. 72). Even if we assume, taking up Marek Węcowski’s game theoretic analysis, that it was the threat of an ostrakophoria alone that served to regulate transgressive elites, there still would have needed be at least of few iterations of the “game being played” after it was introduced in order for elites to learn the optimal strategy of cooperation [34].

For these reasons, ostracism cannot be understood as merely a “protective” or militant institution against tyrannical actors, as ancient and contemporary theorists have often made it out to be. It may simply be beyond the realm of modern scholars to understand exactly what contemporary circumstances or collective states of mind precipitated an actual banishment once the decision to hold one had been decided. And indeed, by trying to analyze the supposed immediate utility of such a procedure, scholars overlook other, more social aspects of the institution. The two-stage procedure of ostracism took time and deliberative energy—a fact in which Athenians seemed to relish. Indeed, contemporary testimonia describe the praxis of ostracism as generating the sort of energy of a “game” (καὶ βλέψειας ὀστρακίνδα, Ar. Eq. 855), complete with its own euphemistic language (Adespota, fr. 363 K–A, cf. Plutarch Arist. 7.2 above), and arguments over who was deserving of whichever outcome (Plato Comicus, fr. 203).

To the Athenians, then, ostracism was more akin to other collective political actions—such as elections, mass protests, or public trials—which were just as much “goal-oriented” as they were instructive, informative, and expressive of the community’s values and problems. In the following section, I articulate the ways in which Athenian ostracism is best understood as having been equally a defensive institution as well as collective political ritual—and the lessons we can, and should, draw from ostracism, understood in this way. In doing so, this paper does not mean to disagree with the traditional interpretations of ostracism (e.g., that it served to protect democracy, diffuse intra-elite conflict, make demotic power manifest, etc.) but merely to suggest that any purely defensive reading of ostracism only captures a part of the Athenian understanding of the institution [23] (p. 138). Through the ostracism, and especially the ostrakophoria, the members of the Athenian demos came to reform and shape the character of the political elites, feel their personal and political grievances collectively expressed, and instantiate democracy “with their feet” in the agora.

This section will first consider the role of character reformation of the individual subjected to an ostracism. The second part will examine how the community was transformed through the collective activity of the ostrakophoria, which served as both a moment of shared remonstration and of democratic socialization in a shared public space [33] (p. 351).12

5.1. Reform, Return, and Recall of Leaders

Though ostracism occasioned a physical removal from the city, it was not a form of permanent social exclusion. The ostracized individual did temporarily “[lose] his public status” and was “deprived of his right to participate in political decision-making and litigation during his absence,” but he was not ever fully rejected from the Athenian community [35] (p. 72). It is in this temporariness and, indeed, leniency of decree that the more rehabilitative and reformative elements of ostracism might first be gleaned, especially with respect to modern “correctional” measures [36]. To the Athenians, expulsion was merely one step in a complex, character-reforming process.

To begin, many modern scholars—drawing on ancient sources—have recognized that ostracism was never considered as, or intended to be, a punitive institution [37,38], (cf. Herodotus 8.79.1; Thucydides 1.135.3, 8.73.3; Diodorus 11.55.1-3; Didymus, Commentary on Demosthenes 23.205.). Plutarch, for example, writes that:

For the banishment by ostracism was not a chastisement of evil character. Whereas euphemistically called a lessening and curtailing of overbearing pride and power, it was in fact a humane consolation of resentment (φθόνου), which thus vented its malignant desire to injure not in some irreparable evil, but in a mere removal for ten years. (Plut. Arist. 7.2)

[39].

Though modern scholars are generally willing to take Plutarch at face value regarding the non-punitive nature of ostracism, they are less willing to engage with the way ostracism both shaped the individual ostracized (to be discussed below), and the community doing the ostracizing [9,12,40]. In Plutarch’s account, ostracism was rehabilitative in two senses: in a literal sense, it relocated the individual in question outside the community, with the intention of reintegrating him once hostilities towards him (or his own hostilities) had died down; in a metaphorical sense, ostracism restored the health of the community by giving the people an outlet to express their mistrust of current elites, or their current political grievances [41] (pp. 224–237).

Relatively little attention has been paid to the transformative effect the Athenian demos hoped ostracism would produce in the expelled individuals. Indeed, Athenians maintained both the expectation and hope that ostracized individuals would return from their sentence for the better [31] (p. 124). This “rehabilitative” idea of temporary exile can be observed in a variety of Greek sources. In the Theogony, for example, Hesiod rehearses a pre-Cleisthenic version of temporary exile, where an immortal who has sworn false upon the “binding oath” of the gods, is “cut off” from Olympus for ten years, and compelled to reflect upon his actions before he may “rejoin the company of the immortals” (Hes. Theog. 775–806 [42]).13 Importantly, for Hesiod, it is through this period of reflection that the immortal learns to uphold and respect the cosmic social order, divine themis (the embodiment of righteousness), and his sacred oaths, as well as to recognize the limitations of his supposedly “omnipotent power” [34] (pp. 11–12, 25). A similar theme also occurs in Plato’s Laws, where the Athenian Stranger prescribes temporary exile as a form of reformative punishment for those who commit homicide in fits of passion, with more “severe penalties on those who slay with intent and in anger” (Pl. Leg. 867c4–d3.) [44].

Though the cases cited above are connected to a rule having been violated, it is important to note that the reasoning behind the punishment in both cases was explicitly connected to the overall betterment of the individual, rather than other potential justifications, such as for retribution, deterrence, or permanent incapacitation [45]. Benjamin Gray too argues that the “underlying reasoning” for this punishment was “probably [that] temporary exile would have the transformative effect of checking the offender’s propensity for anger, perhaps because the harsh conditions of exile would demand more prudent and restrained habits for mere survival” [37] (p. 127).14 Earlier in the text, Plato also comments on the possible reformative effects of being “cast out,” though he does less to explain how such a transformation would occur (Pl. Leg. 854d–e).15 In this way, it is the rehabilitative intention of temporarily sidelining an individual for a perceived transgression that connects the punitive cases about with ostracism.

Although the notions of reformation have thus far been explicitly connected to punishment, there is also evidence of the reformative potential of ostracism itself. In On the Mysteries, Andocides pleads with the demos to not expel him again from Athens, claiming that he has learned the lesson to respect the Athenian way of life from his previous temporary exile. “I know what it is to be an alien sojourning in the lands of neighbouring peoples,” asserts Andocides, continuing that “I have learnt the meaning of self-control and good sense; I have learnt what it is to suffer for one’s mistakes.” Moreover, Andocides says he is willing to share what he has learned as a reformed individual (Ando. 1.144-5). Based on historical evidence, Andocides’ pleas seemed to have worked; the demos were convinced by the “formative hardships” of his exile [37] (p. 127). Nor was the concept of civic “reintegration”—even of supposedly permanent exiles or attested war criminals—foreign to the Athenians [37] (pp. 80–98), ref. [48]. All this would actually seem to lend support to the people’s euphemistic gloss of ostracism, as referenced by Plutarch—namely, that ostracism could serve as a “lessening and curtailing of overbearing pride and power” that served to change the ostracized individual to be more in line with democratic values. Though Plutarch suggests that the relief of phthonos was the actual reason for the people’s decision to ostracize, this does not imply that the institution did not also, if unintentionally, have the people’s claimed effect of humbling the ostracized individual and curtailing his seemingly oppressive behavior.

Indeed, even in Plutarch’s handling of Aristides, Aristides’ public behavior did seem to have changed in response to his ostracism. If before his ostracism the people had misinformed anti-democratic opinions about Aristides, as Plutarch asserts, then afterwards Aristides began to more dutifully publicize his overtly democratic and cooperative behavior in publicly exalting, “for the sake of common deliverance,” his greatest foe (Plut. Arist. 8.1). This shows that it was indeed Aristides’ ostracism alone, not his actual intentions (which, arguably, had held constant), which prompted such behavioral self-reflection. Aristides learned that, when maneuvering in democratic politics, it was not just his substantive actions that he needed to take into account, but also the perception of those actions. In essence, rumors, euphemism, and “renown” (Plut. Arist. 7.1), however faulty, were part and parcel of operating in Athens, a lesson that a powerful man like Aristides was able to learn through a temporary removal from the community.

For this reason, Kosmin suggests that ostracism served as a “rite de passage” that “transform[ed] a dangerous or treacherous politician into a safe member of the Athenian community” [32] (p. 124). In this way, the ostracized subject can be contrasted with another expelled individual—the pharmakos, who, unlike the ostrakismenos, was not permitted to rejoin the community. When a pharmakos ritual was carried out, one or two individuals were permanently expelled from the polis, and sometimes ritually killed, because it was believed that this would apotropaically cleanse the city in times of social, religious, or political upheaval, or natural disaster. The pharmakos was most often a criminal or enslaved individual, though they could be any member of an abject caste of society [49].

Though some have likened ostracism to the pharmakos ritual as serving an opposite but complementary role, there are more contrasts than similarities between the two procedures [50]. Three in particular are worth bearing in mind: First, the pharmakos ritual was explicitly connected to the “purification” of the polis after a communal disaster, whereas the language of pollution and purification is not associated with ostracism in ancient sources [51]. Secondly, the pharmakos was chosen from the lower social classes, therefore not from the echelon generally faced with ostracism. Indeed, various ancient authors suggested that once Hyperbolus, a non-chrestos (non-elite), was subject to an ostracism, the institution suffered an abuse and thereby lost its purpose—as characters like Hyperbolus were not fit to be ostracized (cf. Cf. Thuc. 8.73.3; Plato Com. fr. 187; Plut. Nic. 11.6-7.), ref. [52]. Finally, even if the pharmakos was not necessarily killed, by banishing him from the polis, the demos “kill[ed] him ideologically” [17] (p. 57). Thus, the pharmakos became a simultaneous scapegoat of the city and sacrifice to its gods. On the other hand, the ostracized person was kept alive precisely so as to be able to return to and reintegrate with the community [49] (p. 104). Indeed, those ostracized rarely had to carry out the full course of their term and usually remained close enough to the city to still affect Athenian politics, perhaps indicating that the location or “duration of the exile was not as important as the fact that the ostracized man was banished from the city in the first place” [35] (p. 88), ref. [53].

To add to this, an elite person who was ostracized from Athens could also be recalled for the very defense of that city. Around 480, for example, a general amnesty recalled all ostracized elites to help defend Athens against the second Persian invasion and, accordingly, Xanthippus (father of Pericles), Aristides ‘the Just,’ and Megacles returned to the polis to serve as either strategos or other military leaders in the war (Hdt. 8.79.1, [Dem.] 26.6; Andoc. 1.107; Arist. Pol. 22.7-8; Themistocles Decree/Troezen Inscription). But it was not just emergencies that occasioned a recall; at the behest of the demos, Pericles recalled Cimon once the latter had proven, once and for all, that he did not bear pro-Spartan sympathies.16 The fact that these men not only were recalled early from their exiles, but also came to serve prominent civic roles, demonstrates that ostracism did not necessarily result in lasting stigma or loss of influence. Finally, it should be remembered that Cleisthenes, who codified the law of ostracism, was himself once recalled in an early democratic act. Forsdyke reads this as a powerful moment: “at the same time as the Athenian people took control over political power in the polis (as evidenced by the democratic reforms that followed their uprising), they also took control over decisions of exile.” According to her analysis, in Cleisthenes’ recall, the Athenian demos not only took over the power to ostracize, but equally to bring back members of the community [8] (p. 136).

Though Forsdyke highlights the ability to recall as an important feature of the demos’ power, she does not treat the more general “moment of return” (be it recall, or the end of the term) as a significant moment in itself. Indeed, Forsdyke, along with other scholars of ancient tyranny and democratic theorists, tend to telescope ostracism into other mechanisms of the demos’ power to impede or prevent potential unlawful power seekers. But these views conflict with the fact that those suspected of aiming at tyranny would be subject to much harsher penalties than ostracism, including disenfranchisement (atimia) and death (Arist. [Ath. Pol.] 16.10), ref. [17] (p. 57).

Yet, despite the fact that ostracism was not as harsh as these anti-tyrannic measures, it still was not a pleasant experience. This, naturally, begs the question: why would elites tolerate the existence of ostracism—either as a threat, or as an actual decree—at all? Two reasons are apparent: first, elites believed that being ostracized would actually contribute to their reputation. Often, heroic or legendary status was attributed to individuals who returned from temporary exile, especially in a moment of need. Jérôme Carcopino traces this back to the myth of the Athenian hero and founder-king, Theseus, who, after having been unjustly expelled from Athens, was celebrated “with great pomp” when allowed to return to Athens at a time of great factional strife (Ar. Plut. 627, Plut. Thes. 32–35), refs. [54,55]. Aristotle similarly likens the hero Heracles to having been ostracized by the Argonauts (Aris. Pol. 1284a). Cleisthenes’ own exile and return certainly also contributes to the semi-heroic ethos involved in the rite du passage.

For most, then, the experience of being ostracized was paradoxically more “a compliment rather than otherwise,” as it gave the ostracized citizen the possibility of being recalled into the city prematurely for his talents or skills, only further increasing his reputation [7] (p. 369), ref. [56] (p. 39). Hence, in looking back on the role of ostracism in 5th century Athens, Aelius Aristides writes:

Indeed, in reckoning up a man’s virtue it is a far greater and nobler thing to be sent into exile and then recalled than not to be sent into exile at all… It is jealousy [that] would seem to have been responsible for Cimon’s being expelled and becoming an exile, but one cannot attribute being recalled and summoned to help to anything other than the virtue that made the Athenians feel ashamed before him even in his absence and convinced that they would benefit from his presence. (Aristid. Or. 148)

[57].17

The reputation of Aristides the ‘Just’ was also increased when he willingly inscribed his own name on an ostrakon for an illiterate peasant and when, as he was leaving Athens, he prayed that Athens would have no need to recall him prematurely from exile (Plut. Arist. 7.7–8.).

Thus, individuals who complied with the term of their ostracism had two probable “end states”: either they returned to the polity after the allotted time, with their civic privileges restored and with their popularity increased as a result of their ostracism; or they were recalled early from their ostracism on account of their excellence, and thus gained greater opportunity to prove their worth to the community. This accords with the story that Hyperbolus, a non-elite Athenian, orchestrated an ostracism against his aristocratic rivals, Alcibiades and Nicias, recognizing the situation as a “win-win.” Either one of his rivals would be removed from the polis, clearing the way for his own political rise, or Hyperbolus himself would be ostracized, thus allowing him to attain a “cultural legitimacy for his leadership and that of his class” [33] (p. 341) (cf. Plut. Nic. 11.5). Even the comic poet Plato recognizes that ostracism tended to brand on its subjects an undeserved honor, albeit democratically chosen [8] (p. 153) (Kassel-Austin, PCG fr. 203).

Secondly, elites likely tolerated ostracism as an institution, rather than working against it, due to the difficulty of manipulating ostracism in their favor, and to the relative leniency of the decree.18 Given that ostracism branded no lasting stigma on elites and that the “boundaries” which exiles could not cross were, by all accounts, only a spear’s throw from Athens proper, elites, as a class, had more to gain from abiding by an occasional ostracism, rather than trying to rig the system in their favor, or eliminate the institution altogether [10] (pp. 94, 99). An argument akin to this is made in Against Meidias, in which Demosthenes argues that the laws protect “our common possession” of security, which the wealthy have reason to tolerate, given that nobody questions their advantages (Dem. Meid. 21.210), ref. [37] (p. 167). In other words, the elites recognized that they had more to gain than lose from enduring, or being subject to, ostracism as an institution.

Another understanding of the elite toleration of ostracism was recently put forth by Marek Węcowski. According to his analysis, elites in Athens, always being potentially subject to the demos’ decision to hold an ostracism, were incentivized according to game theoretic logic to, in effect, “play by the rules” of the democratic game and cooperate to avoid the instability, stasis, and aggressive political mobilization that would be generated by a looming ostrakophoria. Węcowski writes that “[a]s long as this ‘weapon’ [the law on ostracism] worked properly, by its mere prospect forcing the Athenian élites to act together and cooperate, there was no opportunity for its violent use on the day of ostrakophoria” [59] (p. 233). In turn, this would generate a system of elite self-regulation which would be not only tolerated, but perhaps even appreciated by the elites themselves. The theory goes that elites would, over time, come to recognize that it would be better for each of them to subsist on less power than optimal and compromise with others than to waste the resources necessary to survive an ostrakophoria [59] (pp. 230–231). Though Węcowski’s analysis largely focuses on the intended pragmatic effects of the Cleisthenic law on ostracism, we should not discount the civilizing effect the law would have also had on elites. Elite cooperation and compromise were, in effect, baked into the institution from its very establishment. Only if such values among elites broke down were ostracized individuals compelled to “learn the hard way”—namely, through temporary expulsion. Only through the ostrakophoria alone did the banished elite learn that he, above others, was most out of touch with the people’s opinions.

The fact that reform, return, and/or recall were all central aspects of the institution demonstrates that Athenians bore deep cultural and affective relations with those whom they ostracized, and they, indeed, hoped that the elites would come back to the polis at the appropriate moment. Individual reformation and civic reconciliation were thus integral aspects of Athenian ostracism. In contrast, today, the “fate” of those subject to militant democratic measures is rarely, if ever, discussed in the literature, except for how pushing them “out” of the political space might have adverse consequences [60] (p. 86).19 Moreover, any active measures of individual reform—such as political deprogramming, or de-radicalization—are rarely considered tools in the militant democrat’s arsenal [62].20

Though it might not be unreasonable to suggest that ostracism-adjacent acts, like social media bans or political debarment, might be effective ad hoc measures to hinder anti-democrats, it is short-sighted not to consider the treatment of anti-democratic actors once they have been identified as threatening to the political order, as did the Athenians with their treatment of ostracized individuals in the 5th century BCE. If not given a path towards reintegration, for example, socially/politically ostracized, “cancelled,” or otherwise expelled individuals might return to the political scene with a vengeance, and perhaps even a greater following. Instead, then, to be effective in the long term at reducing anti-democratic threats, militant democratic measures must involve intervention at more than just the pragmatic “expulsive” level; they must also commit to ideological and affective change in the problematic individual [62] (p. 98). In particular, anti-democratic individuals must somehow be reconciled with the “democratic rules of the game” and be granted a support system outside of their anti-democratic network.

5.2. Democracy in Action: Ostrakophoria as a Moment of Collective Expression

Yet the reformative leader-centric view of ostracism still only presents a partial view of the Athenian institution. Ostracism, although it affected political elites, was effected by the demos, who participated in it as an engaging and lengthy process of collective grief, action, and communication. To lift an idea from modern behavioral theory, ostracism had just as much in common with expressive political engagements as with instrumental ones [63]. Although Węcowski has recently suggested a move away from scholars’ focus on the ostrakophoria and towards an emphasis on (especially elite) discourse surrounding the preliminary voting procedure (epicheirotonia), this does not imply that the final collective process of physically voting someone out, when it occurred, did not independently hold democratic worth. It seems that, Węcowski, similar to those scholars with whom he engages, collapses the ostrakophoria with the final result of banishing someone from the polity, without sufficient conceptualization of the actual process of coming together in the agora and going through meaningful political actions [59] (p. 252). If democratic politics had become so frayed as to occasion an ostrakophoria, then the centrality of reevoking, restaging, and rehearsing collective democratic action becomes all the more important [59] (p. 206). Moreover, the political architecture of the Athenian agora endowed the ostrakophoria with powerful democratic and symbolic meaning, as individual citizens came to see themselves as physically carrying out the task of protecting their democracy from transgressive elites.

During an ostracism, participants incised on the ostraka any name they felt fit, for whatever reason. Though there is evidence of certain expert writers pre-inscribing names for illiterate citizens, there is no evidence that any ostrakophoria was systematically rigged in anyone’s (dis)favor.21 Despite attempts by politicians at mobilizing their supporters, “the people’s sentiments were entirely unpredictable and could shift even at the last moment” [59] (pp. 224, 218). Hence, among the surviving ostraka are a great variety of insults, names, and curses of expulsion for individuals who are not otherwise referenced in ancient sources [17] (pp. 60–63). For this reason, it is historically inaccurate to view ostracism as a predetermined choice between rival politicians, even if there were often a few main contenders [1] (p. 319). The decision to hold an ostrakophoria might have been the “public recognition that something was rotten,” but it was by no means a singular “diagnosis of that evil” [32] (p. 142). In this way, ostracism can be described as an expressive political procedure—like an “inverted popularity contest”—that said just as much about who was doing the casting as whom was being cast out [64] (p. 73).

I understand “expressive” political procedures (e.g., voting) to be situations where the rational decision-maker is “aware that his vote cannot [likely] decide the electoral outcome [but] votes for the utility gained by carrying out what may be considered a civic duty and by expressing support for a policy or a person rather than his self-interest” [65] (pp. 445–446). Carrying on this analysis, I will argue that ostracism, like voting, was both an inherently expressive political procedure (focused on both building a democratic community and airing out collective grievances) as well as a rational one (which did, as previous scholars have noted, display the demos’ “symbolic power” to expel and reform a threatening elite) [8] (p. 233, in passim). Ostracism’s expressiveness came about in two ways. First, citizens often cast ostraka for causes that had little “functional” or “practical” purpose. Secondly, the ostrakophoria was a location-sensitive public ritual, and the location (the agora) endowed the procedure with a positive, democracy-building valence, and which previous analyses of ostracism have largely overlooked.22 Ostracism was both rational and affective, with each aspect supporting the other.

Let us begin first with the expressive use of ostraka. As mentioned above, found in archeological troves are a large number of stray or “scatter votes” for individuals who, presumably, would not have had a realistic chance of being ostracized, indicating that there were “many more issues” than the historiographical records suggest that would have “determined the way an individual would vote” [38] (p. 337). Stray votes, Kurt Raaflaub hypothesizes, could have been cast against a second or third “candidate” who was associated with the more likely successful target of an ostrakophoria. He points to the example of Damon, who received a large number of scatter votes, as a sign of the demos’ distaste for his arrogance and closeness to Pericles, but who was not ultimately ostracized because he was simply not threatening enough to warrant collective expulsion. In casting ostraka against Damon, then, the people were not “expecting” any result except for Damon himself to get the message that his actions were being perceived as anti-democratic [1] (p. 322).

Moreover, citizens often cast ostraka against prominent individuals for somewhat context-specific reasons that do not fit comfortably within purely “rational” tyranny-preventative models of ostracism [66] (p. 660). Peter Siewert, who categorized extant ostraka by their stated charge, unsurprisingly found that the largest number of sherds involve political accusations, but not necessarily accusations of tyranny. Given the context of fifth century Athens, a good number of ostraka reference candidates’ Median (pro-Persian) stances, treason, or those who had “betrayed the interest of Athenians in favor of foreign states” [16]. Other political offenses included general mismanagement and sacrilege. Like Medism and treason, these could be indications of tyrannic leanings, but also indicate a more practical concern with day-to-day politics: la politique, not le politique [67].23 As much as ostracism was a defensive institution, then, it might have also been a way for citizens to let out their more routine anxieties.

There are also accusations against the imbalanced “preeminence” of certain politicians—their ambition, wealth, hubris, etc.—upon which the majority of ostracism scholarship has focused. However, Siewert explains that, within the context of the Persian wars, Athenians were hypersensitive to those who wanted to attribute polis-wide success to themselves alone: overt pride was seen as antithetical to Athenians’ collective war efforts [16] (pp. 9–10). Finally, many ostraka also amount to moral accusations of politicians: greed, malevolence, inner corruption, sexual deviancy, and personal conflict are frequent targets [16] (pp. 11–13). What is interesting here is not just the reasons for ostracism, but also who was ostracized for what reason, as it points to a difference between how Athenians conceived of ostracism vs. how ancient authors have done so. For example, in the historiographical tradition, Megacles was ostracized for his love of wealth (Pind. Pyth. 7), Themistocles for his arrogance and power (Plut. Them. 22.3; Dio. Sic. XI.55), and Cimon for his Laconophilia (Plut. Cim. 17), whereas ostraka cast out Megacles for his adultery, land disputes, and personal animosities; Themistocles for his sexual passivity; and Cimon for incest with his sister [17] (pp. 64–65). In short, ostracism was more than just an anti-tyrannic institution; it also had as its object moral offenses, accusations of corruption, and personal conflicts. Matteo Barbato has likewise analyzed ostracism as a collective response to “perceived dishonourable behaviour” [15] (p. 513).

Athenians, in this way, seemed to “exploit the institution to vent their own feelings,” and not necessarily to threaten proto-tyrants or transgressive individuals [52] (p. 271). Similarly, Athenians would also use ostraka to cast out more general anxieties against hunger (limos), poverty, and the “nobly born” (eupatrides) [33] (p. 337), ref. [32] (p. 133). “Far from being banal writing-surfaces for a rational, political process, ostraka seem to have been treated by at least some Athenians as ritual symbols,” writes Kosmin, “[t]hat is to say, the coherence of object, function, and symbol made ostraka as appropriate to expressive, affective satisfactions as to narrowly instrumental objectives” [32] (p. 134). Through the ostrakophoria, then, people came to transfer their personal anxieties—moral, personal, or economic—to the political arena [17].24

The link between personal and political expression was not, however, completely harmonious. There was consensus neither in the target of an ostracism, nor the reasons for targeting a particular individual, nor even the reasons for holding an ostracism [32] (p. 134). Collective democratic expression was therefore only made manifest in the actual ostrakophoria—the physical coming together and casting of ostraka in the agora. Ostracism was, then, a community-building mechanism that did not so much depend on shared sentiments, as on shared action and communication. It did this in several ways. First, a great number of citizens had to take part in an ostracism for it to be felicitous, and the “timing” of the event ensured that such a requisite number would be easily obtainable. Philippe Gauthier explains that the quorum number (6000) held symbolic value to the Athenians, as an idealized representation of the whole demos [68].25 In order to attain such a number, the preliminary and determining vote had to be taken during city-wide festivals (the Lenaia and Dionysia, respectively), which coincided with times of the year when agricultural demands were less, allowing rural Athenians to congregate in the city and cast ostraka (Aris. [Ath. Pol.] 43.5), ref. [10] (p. 99), ref. [23] (p. 140).

The interval between the first and second vote facilitated even greater community building activity, as people could—and did—openly debate about whom they should ostracize. Ostracism was not only an “inverted popularity contest,” but also a thoroughly popular, even festive, procedure. Just as Athenians keenly followed the assembly debates, by cheering and booing speakers, one could easily imagine this behavior spilling into the open forum and keeping citizens (and non-citizens alike) entertained for the two-month interval. Raaflaub adds that “[w]e should certainly not underestimate the impact of polemics, propaganda, organized campaigning, and attempts at influencing the vote [i.e., the ostracism]; the schedule offered plenty of time for such activities…” [1] (p. 323) [69]. The interval also meant that there was a “cooling off” period that would seem to hinder rash decision-making and encourage open campaigning and informal discussion.

Ostracism also instantiated expressive democracy in its very location: the agora. “[T]he Agora setting was deliberate,” writes Kosmin, chosen for its central location, “association with quotidian anti-elite behavior,” and as it was considered the “foundational site of democratic freedom” [32] (pp. 145, 150). As the setting for public administration, private exchange, and religious festivals, the agora represented both the economic and social center of society, and the place “for the negotiation of [contentious] political ideology” [70] (p. 76), ref. [33] (p. 348), ref. [71]. Though literal barriers went up to “separate institutional from non-i nstitutional space and time,” such as during an ostrakophoria, this nonetheless did not prevent the typical, less-than-couth loidoria (personal invective or slander against prominent individuals) from affecting the procedure, as evidenced by the myriad of insult-laden ostraka [72] (pp. 15–16, 81). Indeed, the agora was recognized as one of the few open spaces where loidoria was tolerated, though certainly not encouraged.26 Even the adjective agoraios, when applied to those who hung around the agora, came to have the meaning “boisterous,” “vulgar,” and “ne’er-do-well.” This concentration of expressive (not necessarily truthful) gossip, campaigning, and ridicule gave the agora an (often pejorative) association with democratic attitudes [70] (p. 76) (cf. Aristoph. Lys. 457–460; Plut. Mor. 521e; Ar. Nub. 991; Isoc. 7.48.).

The agora also held symbolic value as the site of democratic freedom, as it was in the shadow of Tyrannoktonoi (lit. “Tyrannicides”), a famous statue group of the tyrant-killers, Harmodius and Aristogeiton (Paus. 1.8.5; Arr. Anab. 3.16.8.). To Tonio Hölscher, the statues’ poses were intentionally “paraenetic,”, i.e., meant to encourage imitation, not only of the anti-tyrannical action itself, but also of the broader demotic attitude [73] (pp. 158–160). Or, as Ober puts it: “the killers, acting as a cooperative team, boldly advancing upon their foe, are caught by the sculptors at the moment just before the death blow was struck; the viewer is drawn into the action and invited to complete the narrative for himself” [74] (p. 219). Moreover, in Athens at the time was a prohibition on erecting any other honorific statues near the pair, endowing them with an uncontested prominence, visibility, and isolation [75] (pp. 113–114).

Democratic ideology and imagery were thus fused into ostrakophoriai, such that they became symbolic reperformances of democracy’s founding moment itself. Members of the demos did not just see themselves as banishing a nefarious individual, but also as participating in an inherited political experience, meant to rid their own anxieties and preserve and create their democratic regime anew. Ostracism was therefore just as much a generative, expressive act as a pragmatic one. In contrast, modern measures of militant “neo-ostracism” are more divisive than generative. Indeed, it is not just that people disagree on who should be subject to militant democratic procedures, but also that the people object to the necessity of the procedures themselves [60] (p. 86). Moreover, as Zhang and Malkopoulou have written, measures of modern “ostracism” are outside of democratic control—either relegated to constitutional courts or internal regulatory boards of private organizations—and hence, do not have the democratic ethos associated with the ancient institution [5,24].

Though no direct democratic procedure of neo-ostracism is likely to emerge in the near future, there are still lessons to be drawn from the Athenian model. During an ostrakophoria, citizens not only came to exercise their political freedoms, but also to imbibe, embody, and reproduce democratic values. They became symbolic tyrant-slayers, even as they openly disagreed with their fellows about whom to ostracize. In effect, the ostrakophoria and the hubbub surrounding it both publicized what might constitute “tyrannic behavior” and generated a pro-democratic spirit through the very performance of that procedure. Similar modern militant democratic measures, then, might include community coaching projects that would aim at preventing individuals from being drawn to anti-democratic ideologies, and that would, at the same time, promote pro-democratic sentiments at the local level [62].27 Such pro-active, even “preventive” measures, are not new to militant democratic theory, but have generally been sidelined as too repressive, historically contingent, or necessarily elitist [76]. But, as Karl Loewenstein, the founding theorist of militant democracy, avowed in his “Militant Democracy and Fundamental Rights,” democratic self-protection cannot be based on formal provisions alone—a democracy needs active democrats [77].

6. Conclusions

During an ostrakophoria, thousands of citizens, brandishing coarsely incised potsherds, would converge in the agora, after much hubbub and energetic politicking, to ostracize a prominent individual [59] (pp. 24–30). Afterwards, the “loser” (or “winner”?) of this contest would be compelled to leave Athens for a set number of years, although, in practice, he was often recalled early, having acquired a greater appreciation of, and respect for, Athenian political culture. The account adumbrated above highlights two often overlooked facets of Athenian ostracism: first, that it was just as much reformative as it was expulsive, and, second, that it was just as much expressive of the demos’ positive taste for democracy as it was protective of that very regime. Such observations were noted early on by George Grote, one of the first to show a nominal appreciation of the institution:

“[Cleisthenes realized that] it was necessary to create in the multitude, and through them to force upon the leading ambitious men, that rare and difficult sentiment which we may term a constitutional morality… Next, through such tranquil working of the democratical forms [i.e., ostracism], a constitutional morality quite sufficiently complete was produced among the leading Athenians, to enable the people after a certain time to dispense with that exceptional security which the ostracism offered. To the nascent democracy, it was absolutely indispensable; to the growing yet militant democracy, it was salutary; but the full-grown democracy both could and did stand without it”

[77] (pp. 155, 160).

Grote thus saw constitutional moral-building as a much more central aspect of ostracism than is now usually attributed to it, and this lesson should be recalled when suggesting modern forms of political ostracism.

A more fundamental tension has also recently been raised as to whether ostracism necessary was as inherently tied to democracy as modern interpreters, including Grote, Forsdyke, Siewert, Rosenbloom, and many others have made it out to be. After all, ostracism-adjacent measures seem to have been a feature throughout the Greek world. Aristotle, for example, although writing decades after the last ostracism in Athens took place, nonetheless suggests that some institution like ostracism naturally crops up in all political systems to address the “universal” problem of disportionality (1284^b3).

I do not think this fact, however, discounts the democratic nature of ostracism in Athens, nor in other Greek states, more generally. In Syracuse, for example, “petalism” seemed to involve a two-stage voting procedure, similar to the Athenian model (Diod. Sic. XI 86.4–87.6). Likewise, a small trove of ostraka found in Chersonesos also attests to the presence of a certain amount of internal politicking consistent with Athenian practice [59] (pp. 34–36). These systems all had a measure of popular involvement and clearly had an effect on public consciousness, even if they took place in decidedly oligarchic regimes. Ostracism and similar measures could be more or less democratic, but insofar as they tended to involve a greater number of people in the political system, they could be said to be rather inclusive, deliberative institutions in the main. In Athens, in particular, ostracism was a preeminently democratic feature, if but for the forms of collective life it generated.

When discussing measures of “neo-ostracism” today, however, it is human rights concerns (e.g., Hannah Arendt’s “right to have rights” [78] (p. 296)), territorial and boundary issues, and citizenship considerations that are the most cited factors that reasonably prevent anything like the institution’s reintroduction [5] (p. 11), ref. [79] (pp. 575–578), ref. [80] (pp. 291–295). But even those who, having sidelined such issues, still advocate for the pro-democratic benefits of “neo-ostracism” do so under the assumption that ostracism would in fact be an effective hinderance against anti-democrats. But this is not necessarily the case. Martin Ostwald, for example, observes that, in Athens, not only were “there simpler and more permanent ways of eliminating political opponents [than ostracism],” but there could also “be no guarantee that the same procedure would not backfire or be used in a subsequent year against the very person who first instituted it” [38] (p. 336). Ostwald and others bring up a few other “paradoxes” that make ostracism largely inadequate as a prophylactic against coming tyrants. For one, given that coups were (and still are), more often than not, spontaneous or momentary events, it is not clear how the sudden arrival of an anti-democratic individual would be immediately hindered by a regular and recurrent ostracism. To add, ostracism would also be an ineffective measure against demagogues—the prototypical proto-tyrants—as they would be revered, not held in contempt, by the people doing the ostracizing. On the obverse, according to Lindsay G.H. Hall, “a dangerous individual or group prepared to resort to extra-legal means in pursuance of their goals would be rendered not a whit less dangerous by the law of ostracism” [10] (p. 93). Why did ostracism, then, function so well in Athens? W. Robert Connor explains:

[The Athenian constitutional framers] devised an instrument that endured…and one nicely adapted to the political conditions of democratic Greek polis. Behind it was a recognition that what was most to be feared was not ideas or policies but men. The institution worked well because it was adapted to a system of politics in which political ties were in the first instance personal ties. As long as this was the case the removal of the leader would dissolve or at least temporarily incapacitate the group. Ostracism would tend, as Plutarch once phrased it, to ‘undo the hetaireia’ by dissolving the ties which held it together.

In contrast, today, eliminating or expelling the leader of a party or interest group would simply allow for a new member to take their place, at most making any form of neo-ostracism into a mere “exercise of factional vindictiveness” and likely making a martyr for a certain political cause [64] (p. 75).

This ineffectiveness as a measure against anti-democrats poses a great obstacle to modern theorists’ one-sided view of ostracism. However, there remains an even more fundamental challenge to militant democrats’ conceptual deployment of the institution: when they reference ostracism as a “forerunner” of militant democratic measures, they ignore all the “processual, expressive, emotional, and embodied strangeness” that the ancient procedure carried with it—the democratic imagery, collective catharsis, reform-minded outlook, and ritualistic atmosphere which permeated the whole process [32] (p. 125). The protection of the political community through the ejection of a member was merely one aspect of the institution, but equally important to Athenians was the process of reforming the character of accused individuals and cultivating democratic sentiments in the political community. Unless militant democrats are willing to somehow replicate the broader cultural context of ostracism—such the institution’s expressive links to democratic values and the individual leaders’ readiness to be reformed—then even updated measures of “neo-ostracism,” such as social media bans or lustration procedures, will not work in the long term. Democracies cannot rely on mere expulsion to reduce the threat of anti-democratic individuals, nor to hamper anti-democratic movements.

What Athenian ostracism demonstrates to us, then, is that any militant democratic procedure, or any institution to “protect” democracy more broadly, must engage with the “softer elements” of democratic self-defense, such as civic (re-)education and community reformation [26] (p. 13), refs. [3,81]. The question about whether and how to apply protective institutions should not merely be concerned with how to eliminate threats to democracy, but also how to make democracy more resilient against those threats and how to convince anti-democrats of the value of democracy, and democratic processes, themselves. If democracy needs protecting, then the first step is to ensure that democrats themselves recognize the importance, worth, and need in doing so.