Indicting the Athenians in the Melian dialogue

Abstract

It is widely agreed that Thucydides’ Melian dialogue presents the Athenian invasion of Melos, and the Athenian justification, in a negative light. Attention tends to focus on the immorality of ‘the rule of the stronger’ that the Athenians present in the dialogue. This essay argues that another feature of the dialogue triggering negative judgements of the Athenians is their criticism of the Melians’ resistance: it is voiced by the Athenians themselves and therefore provokes in readers a ‘speaker-relative’ normative judgement of the Athenians. Philosophers have explored how our normative judgements about statements often depend on the speaker. Because the Athenians have deliberately put the Melians into their perilous situation, and because part of Athenian self-mythology was heroic resistance against overwhelming numbers in the Persian Wars, Athenian criticism of the Melians is hypocritical and applies an asymmetrical ethics to the Athenians and the Melians. Reaction against these features of the dialogue exacerbates the moral abhorrence of the Athenians felt by many readers. Hence I disagree with Bosworth’s view of the dialogue as primarily critical of the Melians. Instead we see Thucydides here condemning not only the Athenian imperial project but also the rhetoric used to defend and sustain it.

I. Introduction

The Melian dialogue is a damning indictment by Thucydides. This is agreed by many. There has long been debate, however, over what Thucydides is indicting. A particularly harsh application of Athenian imperialism? Stark realism in foreign affairs? A ‘sophistic’ rule of justice that ‘might makes right’? An international system in which the Athenian attitude is regrettable but necessary? Immutable human aggression? Melian foolishness in the face of overwhelming odds?¹ What may be the most common general interpretation goes back at least to Dionysius of Halicarnassus, who criticizes Thucydides for depicting the Athenian envoys in Melos as ‘presenting the most disgraceful arguments and wrapping them in the most unpleasant language’, and as arguing that justice, under the law of nature, is ‘to rule whomever one is able to conquer’.² Although Dionysius blames Thucydides rather than the Athenians for the words that Thucydides reports, his criticism is founded partly on finding the Athenian arguments ‘not appropriate (προσήκοντα) to be spoken by Athenians to Greeks whom they had freed from the Persians’ (Thuc. 39). Francis Cornford’s observation that the Athenian argument suffers from a ‘horrible deformity’ is typical of a long line of judgements that the dialogue presents the Athenians as morally in the wrong.³ In a ‘deliberately provocative’ article published in these pages in 1993, however, A.B. Bosworth argues that the dialogue’s ethical judgement is aimed not primarily at the Athenians but at the Melians, and that the greater weight of moral indignation kindled by the dialogue is against the Melians and their ‘perverse determination to hold out against overwhelming force’.⁴

Thucydides provides material for the interpretation of Bosworth, who is largely persuaded by the arguments offered by the Athenians. And undoubtedly a number of readers over the centuries have had similar reactions, critical primarily of the Melians. But I will argue here that the dialogue, whose format is unique in Thucydides, gives us more reason to question than to endorse the Athenian arguments. Philosophers have recently introduced the concept of ‘speaker relativity’ as a label for the way in which certain facts about a person who offers ethical judgements or advice can affect our own judgement about both that person and their advice. Interpreting the Melian dialogue in the light of speaker relativity, I will argue that Athenian criticisms of the Melians are themselves morally objectionable. Three main speaker-relative facts compromise the Athenian arguments that the Melians, given their situation, should surrender peacefully to the Athenians. The first is that the Athenians themselves purposefully put the Melians into the situation. The second is that the Athenians treat the Melians as having agency but consider themselves to be following an unyielding law of nature. The third is that, when faced with a similar show of apparently overwhelming force, by Persia during the Persian Wars, Athenians had refused to submit without a fight. I will offer an explanation of how both the Athenians’ sanctimoniousness (for lack of a better word) and their hypocrisy weaken the aggressors’ case.

There are less and more ambitious versions of my argument. The less ambitious version concerns only the responses of some modern readers and seeks to deepen our understanding of their commonly harsh reactions to the Melian dialogue. The more ambitious version claims that ancient readers would have had similar reactions, for similar ethical reasons, and therefore my interpretation helps us understand Thucydides’ work in its historical context. Because the philosophical arguments I present are derived from modern philosophers, the question arises whether modern philosophy can help us understand ancient Greek ethical intuitions. This is a fair concern, and one that Bernard Williams has addressed.⁵ Like Williams I believe that on certain questions, above all the institution of slavery and the social roles and rights of women, Greek ethical views are so different from modern western views that modern philosophy does little to help us understand the Greek state of mind.⁶ But I also follow Williams in his belief that on a number of ethical questions Greek understanding and modern western understanding are closer than has often been acknowledged.⁷ On these questions, to adapt Peregrine Horden and Nicholas Purcell’s phrase, modern philosophy is ‘in’ but not exclusively ‘of’ modernity.⁸ Without arguing the point systematically, I will provide evidence that particular aspects of the ethical intuitions experienced by modern readers of the Melian dialogue were available to and alive for Thucydides and his early readers.

By arguing that the Melian dialogue encourages a critical ethical view of the Athenians, I am making a case for Thucydides being to some extent an ethical writer even with regard to foreign affairs.⁹ The dialogue is often taken as a primary exhibit in the case for the common view of Thucydides as an ‘amoral’ observer, a strictly scientific historian, or an international relations realist avant la lettre.¹⁰ Bosworth appears to agree with me that the Melian dialogue is making an ethical point, although for him Thucydides’ position is that the Melians should have capitulated immediately and their decision to resist was ‘perverse’.¹¹ As I argue below, the central argument the Athenians make in the dialogue, namely that the Melians should capitulate because ‘might excludes right’ (in Simon Hornblower’s paraphrase), is itself an ethical position and is not amoral. If this is correct, then the Melian dialogue is not evidence for Thucydides’ exclusion of morality from foreign affairs but rather highlights Athenian attempts to cloak their ethics in the garb of natural necessity, a move that is in itself ethically objectionable. In this way my interpretation supports arguments made by others that the ‘realist’ reading of Thucydides over the centuries has been overly selective and fails to do justice to the complexity of his work.¹²

I will begin with a discussion of speaker-relative ethical judgements about normative arguments or statements. Building on insights from G.A. Cohen and R. Jay Wallace, I offer ethical interpretations of situations in which the circumstances and history of the speaker of a normative argument or statement affect a hearer’s judgement of that argument or statement. I will explain how these interpretations are relevant to our judgement of the Athenians’ criticisms of the Melian position. The Athenians in the dialogue make ethical (or moral) arguments, as opposed to mere prudential arguments based on bare facts about the world, and adopt a corporate persona under which they endorse the actions and opinions of Athenians generally. These factors cast the Athenians in an ethical light more negative than that shone upon the Melians, one that supports Dionysius’ judgement and reveals Bosworth perhaps to have fallen into one of Thucydides’ many traps.

II. Speaker relativity: kidnappers and hypocrites

i. Cohen’s kidnapper

G.A. Cohen, in noting a phenomenon rarely discussed by philosophers, explains how the ethical force of normative statements can be speaker-relative:

A normative argument will often wear a particular aspect because of who is offering it and/or to whom it is being addressed. When reasons are given for performing an action or endorsing a policy or adopting an attitude, the appropriate response by the person(s) asked so to act or approve or feel, and the reaction of variously placed observers of the interchange, may depend on who is speaking and who is listening.¹³

Cohen focuses on the case in which the speaker of an argument has made a choice that makes one or more premises of the argument true. He offers the example in which a kidnapper has stipulated that a child will be freed only if the parents pay him a ransom. As Cohen phrases it, the basic argument for paying the ransom might go as follows:

• A1. Children should be with their parents.

• A2. Unless they pay him, this kidnapper will not return this child to its parents.

• A3. So this child’s parents should pay this kidnapper.

In this form, the argument is put into the mouth of a third party (perhaps a police detective, or a friend of the parents), who simply takes as a fact the kidnapper’s intention not to release the child unless paid. In this form, the argument might have great force. But the force changes, Cohen argues, when we put it into the mouth of the kidnapper:

• B1. Children should be with their parents.

• B2. Unless you pay me, I shall not return your child.

• B3. So you should pay me.¹⁴

In this case the argument is ‘discreditable’ to the speaker:

[T]here is, of course, no mystery about why the argument’s presenter attracts discredit … He does so because the fact to which he appeals, which is that you will get your child back only if you pay, is one that he deliberately causes to obtain: he makes that true, and to make that true is morally vile.¹⁵

To rephrase Cohen’s point more explicitly: the fact that the speaker is an agent with free will who freely chose to bring about a ‘morally vile’ premise of his normative argument casts that argument in a negative ethical light.

But we must add a further observation to understand why the kidnapper’s expression of this argument creates an ethical problem beyond the mere fact of having committed a ‘morally vile’ act. As Cohen notes, in voicing this argument the kidnapper displays a form of ‘bad faith alienation from [his] own agency’.¹⁶ This alienation consists, I suggest, in the fact that from premise B1 the kidnapper moves to premise B2, which concerns action by the parents. But the kidnapper thereby ignores another means by which B1 might be fulfilled, namely by his own action of simply returning the child. By voicing B2 rather than an alternative focused on the kidnapper’s ability to act (‘I am holding your child, therefore I should return your child’), the kidnapper is alienated from his own agency by effectively denying it. This denial of the speaker’s own agency is strengthened by the contrast with his treatment of the listener as possessing agency. The implication is, ‘I can’t control my actions but you can control yours, and this is how you should act’. The resulting asymmetry of attributed agency in turn indicates an asymmetry of the application of ethical standards: the argument as spoken by the kidnapper implies that ethical standards should be applied to the parents’ behaviour but not to the kidnapper’s, since someone without agency cannot a fortiori have ethical agency.¹⁷ This is an identifiably Greek view, as Aristotle makes clear in his observation that praise and blame are accorded only to voluntary actions.¹⁸

Anyone who hears the kidnapper’s argument is apt to get angry with him for this ethical asymmetry and to judge him in an even more negative light than for the kidnapping alone (although that deed, plus the demand for ransom, is certainly his most despicable act here). But we may press the case further and hold that the argument is also less persuasive coming from his lips. Its persuasiveness is reduced by an implied inconsistency between B1 and B2: the claim of B2 implies that the speaker himself does not in fact believe the truth of B1.¹⁹ If the kidnapper really believed that children should be with their parents, he would not have committed a kidnapping and imposed a ransom in the first place. This inconsistency in the argument undermines our ability to believe that its conclusion is true.²⁰

In short, the argument is apt to be both offensive and less persuasive coming from the kidnapper than from another speaker. Contrast the case of the parents discussing whether to pay the ransom. They would agree that the child should be with them; they would observe that the kidnapper seems adamant in his demands; and they would therefore conclude that they should pay the ransom. Here is no asymmetry of attributed agency because the parents concede that the kidnapper has agency. And there is no logical inconsistency between premises. Because the conclusion voiced by the parents is the same as that voiced by the kidnapper (‘You (= the child’s parents) should pay me (= the kidnapper)’), we see that the weakness of the kidnapper’s argument is speaker-relative.

ii. Wallace’s hypocrite

One common situation in which our ethical evaluation of a normative statement is speaker-relative, recognized already by ancient Greeks, is hypocrisy.²¹ As Wallace explains, when we judge someone to be a hypocrite we are claiming there is ‘an inconsistency in a person’s attitudes and behavior’:

They might have attitudes that cannot both be correct, as in the case where people publicly disapprove of practices that in private contexts they condone. People in this position seem to be saying both that the practices in question are impermissible and that they are not impermissible. Or the inconsistency might be between their official attitudes and their behaviour, for instance when they themselves actually engage in the very conduct that they censure on the part of others. Here we might think of the tension as involving a failure to live up to one’s own commitments, commitments that are undertaken through the public act of moral censure.²²

The specific kind of hypocrite I am concerned with here is the person who makes an argument or criticism based on a rule, or gives advice to act according to a rule, that the hypocrite himself violates or has violated.

Although criticism of hypocrites is ‘a nearly universal moral currency’, there is debate over what in hypocrisy is ethically objectionable.²³ One problem with hypocrisy is its deceptiveness about moral attitudes: hypocrites falsely represent ‘their own moral attitudes, with the intention of misleading others about what is the case’.²⁴ The more fundamental problem is an undermining of the hypocrite’s ‘standing to engage in moral criticism’; the hypocrite is not entitled to criticize someone else for things he himself has done.²⁵ The specific cause of this failure of ‘standing’, according to Wallace, is that ‘hypocrites attach greater importance to their own interests than to those of the victim’.²⁶ He explains at greater length:

Hypocrisy … involves an implicit ascription of differential significance to one’s own interests, as compared to the interests of others. In violating the same moral requirements, both the hypocrite and the agent the hypocrite criticizes lose their entitlement to protection from the opprobrium of others. The hypocrite seizes on this feature of the other agent’s situation, subjecting him or her to the kind of opprobrium that is constitutive of moral blame. At the same time, however, hypocrites arrogate to themselves a higher level of protection from opprobrium, insofar as they refuse to engage in the kind of critical self-scrutiny that would lead them to acknowledge that they have engaged in wrongdoing of just the same kind … This combination of attitudes effectively attaches different levels of importance to the hypocrite’s interest in protection from opprobrium and to that of the hypocrite’s target. This in turn opens hypocrites to a distinctively moral complaint, thereby depriving them of the standing to engage in the criticism that they have launched.²⁷

Wallace sees this ‘denial of equal moral standing’ as the most objectionable aspect of hypocrisy.²⁸

Although it is probably more common to object to hypocrisy when we agree with the ethical rule that the hypocrite invokes, and therefore object to the hypocrite’s own previous violation of that rule, hypocrisy is also objectionable in the inverse case. When we disagree with the ethical rule the hypocrite has cited, we can find it ethically objectionable that the hypocrite previously (and correctly) violated the rule but now invokes a false rule against someone else. The objectionability of the hypocrisy therefore depends not on the correctness of the ethical rule invoked but on the differential level of ‘protection from opprobrium’ the hypocrite applies. And thus, like the objection to Cohen’s kidnapper, the objection to the hypocrite is speaker-relative: it is the hypocrite’s own differential ethical attitude that offends, regardless of whether we agree or disagree with the criticism the hypocrite has levelled.

III. Speaker-relative criticism of the Athenians in the Melian dialogue

The Athenians in the Melian dialogue are subject to both the objections against Cohen’s kidnapper and those against hypocrites. My argument is based on a reading of Thucydides’ text rather than on conclusions drawn from that text about actual events. I am not claiming that the Athenian envoys in Melos were in fact like Cohen’s kidnapper, nor that they were actually hypocrites (although some of them may have been both). Instead I am offering a literary and philosophical interpretation of the text based in part on the context of historical facts. My reading is literary in proposing that the text has certain effects on readers. Because among those effects is the triggering of ethical reactions or intuitions, it is philosophical in offering explanations for the underlying grounds of those reactions and intuitions.²⁹

i. The Athenians’ ethical arguments

The Athenians ethically judge the Melians and make an ethical argument about what their goal should be in light of the Athenian invasion. They do, of course, make a prudential argument about what the Melians should do in order to maximize their own safety. But the Athenians also advise the Melians that safety should be their highest priority: they suggest to the Melians both an end and a means to that end. By justifying a prudential argument in moral terms, and thereby passing moral judgement on the Melians, Thucydides’ Athenian envoys invite moral judgement upon themselves.

The Athenians state at the outset that they want to talk to the Melians only if the Melians adopt the ethical stance that safety should be their highest priority. They threaten to ‘stop right now’ if the Melians have come for any purpose other than ‘to consult about the safety of your polis based on the present circumstances that you see before you’ (5.87). According to the Athenians, there is only one appropriate goal for the Melians to have in these circumstances: the Melians’ own safety.³⁰ As the conversation progresses and it becomes apparent that safety may not be the Melians’ highest priority, the Athenians criticize this attitude. In response to the Melian claim that it would be ‘awful and cowardly (πολλὴ κακότης καὶ δϵιλία) for us, while we are still free, not to do everything we can to avoid slavery’, the Athenians respond, ‘No’ (5.100–01). They deny that there is anything awful or cowardly in surrendering to them. Because the Melians are outmatched, this is not a contest in which they should seek to avoid shame (αἰσχύνη); rather they should deliberate prudently (σωφρόνως), that is, with regard to their safety (πϵρὶ δὲ σωτηρίας μᾶλλον, 5.101). Hornblower claims ‘there is no moral tinge’ in the use of σωφρόνως, but his judgement ignores that the Athenians are replacing the Melians’ ethical criteria of ‘awfulness’, ‘cowardice’ and ‘shame’ with their own criteria of ‘prudence’ and ‘safety’, which the Athenians claim should be guiding the Melians’ judgements.³¹ This ‘should’ implied in the Athenian response makes the Athenian criteria ethical. They claim the Melians have the wrong goals.³² The ethical (or moral) aspect of the Athenians’ argument becomes even clearer in their peroration. After again warning the Melians not to base their decision on a fear of shame (αἰσχύνη, 5.111.3), the Athenians turn around and phrase their own counsel of ‘prudence’ in the same terms: many people, they claim, drawn to a certain course of action by ‘the power of the seductive word “shameful”’, by their own lack of sense they take on ‘a shame even more shameful (αἰσχύνην αἰσχίω)’ than they would have by mere bad luck (5.111.3). This adoption of the Melian ethical vocabulary leaves no doubt that the Athenians are offering them ethical advice and an ethical critique. They drive the point home by insisting that there is one way for the Melians to ‘deliberate well’: by choosing the security of submission over the honour of fighting for independence (5.111.4).

ii. The Athenian envoys as both speakers and actors

One might agree the Athenians are making an ethical criticism of the Melians but argue that the dialogue as depicted by Thucydides still does not present a case like Cohen’s kidnapper or the hypocrite. After all, the speakers here are a few unnamed Athenians in the year 416. Because the Athenian decision to subjugate Melos was made by the democratic assembly, these envoys cannot themselves have been responsible for putting the Melians in this critical situation. And even less can they be called hypocrites based on actions taken by other Athenians six decades earlier. Thucydides’ text supports neither of these objections because the envoys actively adopt a transtemporal corporate ‘Athenian’ personality. By this expression I mean that, as Thucydides presents them, the Athenian speakers of the Melian dialogue willingly represent and adopt the viewpoint of the great bulk of Athenians both of their own day and of the previous two or three generations.

Bosworth argues that a proper interpretation of the dialogue takes into account that the decision to subjugate Melos had already been made by the dēmos in Athens, and that there was no way for the Athenian generals Cleomedes and Teisias, much less the envoys speaking in the dialogue, to revoke that decision.³³ On this view the envoys are less like Cohen’s kidnapper than friendly advisors to the Melians. ‘It’s unfortunate you are in this situation’, we might paraphrase the Athenian envoys as saying, under Bosworth’s interpretation, ‘but we all know there is no way to change what those imperial Athenians will do if you resist. So you should really just concede peacefully’. But Thucydides draws no distinction between the envoys on the one hand and the Athenian dēmos or military on the other, and gives the reader no reason to conclude that the envoys did not fully endorse the Athenian aggression. Unlike in the case of decisions by the assembly regarding Mytilene and Sicily, for example, Thucydides does not even mention, much less reproduce, an Athenian assembly debate regarding action against Melos. Nor does Thucydides put Bosworth’s point into the envoys’ mouths, along the lines of, ‘This was an irrevocable decision by the assembly, which we cannot modify and are legally bound to carry out’.

The envoys do warn the Melians not to believe that they can persuade (πϵίσϵιν) the envoys not to attack (5.89). But the envoys’ claimed inability to be persuaded is not premised on their hands being tied (à la Bosworth). On the contrary. The envoys use the first person plural in describing ‘our’ actions not only in attacking the Melians now (νῦν ἐπϵξϵρχόμϵθα), but also in holding power justly because ‘we’ justly defeated the Persians six decades earlier (δικαίως τὸν Μῆδον καταλύσαντϵς ἄρχομϵν), as well as in describing what ‘we really think’ (ἀληθῶς φρονοῦμϵν): that those in power do what is possible and the weak yield (5.89). This posture of the envoys as fully representative of the Athenian dēmos and endorsing the entire imperial project as well as the present action is upheld throughout the dialogue.³⁴ By leaving no ‘gap’ between the words and actions of the envoys and those of the dēmos, Thucydides creates the effect that the envoys speaking to the Melians are the very people who have deliberately launched a military expedition against them. They adopt the position of a kidnapper blaming the parents for hesitating to pay the ransom.

The envoys’ language thus opens them to a charge of hypocrisy. They endorse, and even take credit for, the actions of Athens generally: here and now at Melos, as well as during the Persian Wars, and in the work of empire during the intervening years (5.89–91). This form of identification with one’s polis or state as it has existed and acted across time, the adoption of what I am calling a transtemporal corporate personality, is consistent with the corresponding attribution of such corporate personality to others, and both phenomena can be seen elsewhere in Thucydides. We see it already in the first speeches reported in the work, by the Corcyreans and Corinthians regarding Athenian intervention in their dispute of 433 (1.32–43). The Corcyreans admit they are being ‘inconsistent’ (ἄλογος) because, ‘although never before in our history willing to be allies with anyone, we have now come to seek this from others’ (1.32.4). That the Corcyreans feel the need to address an ‘inconsistency’ shows that they see themselves, and expect the Athenians to see them, as a transtemporal corporate personality, ‘the Corcyreans throughout their history’. That this personality is transtemporal is indicated by the phrase ‘never before in our history’ (πω ἐν τῷ πρὸ τοῦ χρόνῳ, more literally ‘never in previous time’). The Corcyrean ambassadors realize they, as ambassadors of the current decision-making body of Corcyra, might be blamed for adopting a policy contrary to that adopted by previous Corcyrean decision-making bodies. What I am calling transtemporal corporate personality, and ethical concerns deriving from it, permeate Thucydides’ work and seem to reflect a common Greek attitude.³⁵

Thucydides depicts Greeks not only adopting a single corporate personality but also, at times, drawing distinctions between themselves and the political decision-makers of a polis, present or past. At the Spartan ‘trial’ of the Plataeans in 431 the Thebans speaking for the prosecution carefully distinguish themselves from earlier Thebans who decided to side with the Persians against the allied Greek force in 480. They argue that, because Thebes during the Persian Wars was ruled by a small clique (δυναστϵία ὀλίγων ἀνδρῶν), the polis was not in charge of its actions at that time (ἡ ξύμπασα πόλις οὐκ αὐτοκράτωρ οὖσα), and therefore the entire polis should not be blamed for the decisions made when it ‘lacked a constitution’ (3.62.3–4). A fortiori, they imply, the current generation of Thebans should not be blamed for previous Theban medizing.³⁶ Yet this comes across as special pleading, and therefore the exception proving the rule: among Greeks as depicted by Thucydides, there is a default reflex to assume an identity of transtemporal corporate personality in the context of inter-polis affairs, and to grant individuals (at least when they are representing their polis in some respect) credit or blame for the official corporate actions of their fellow citizens, present or past. At the same time, it is always possible for individuals to argue in any particular case, as the Thebans do at Plataea, that no such identity exists and to claim they should not be blamed for the decisions or actions of their countrymen.³⁷

Thucydides depicts the Athenian envoys speaking at Melos as fully identifying with the official Athenian decision-makers, and with Athenians generally, both past and present. As the Thebans at Plataea show us, the Athenian envoys might not have done so; they might have attributed to the assembly the Athenian decision to attack Melos. That they fail to do so renders Bosworth’s appeal to such a possibility unpersuasive as an interpretation of Thucydides’ text. As other passages indicate, it was usual to accord praise or blame on an assumption of transtemporal corporate personality such as that adopted by the Athenian envoys. Thucydides’ text therefore encourages the reader to treat the envoys both as speakers representing the views of all Athenians and as actors who themselves both jointly led the Greek fight against the Persians and decided to attack Melos. Thucydides thus exposes the envoys to criticism based in part on actions or decisions taken by any official or semi-official group of Athenians.

iii. The Athenians as hypocrites

The Athenian envoys criticize the Melians for considering resisting an invasion force significantly larger and apparently stronger than their own defensive forces.³⁸ The criticism is hypocritical because the Athenians and their Greek allies had during the Persian Wars resisted an invading army and navy that they believed to be considerably larger and stronger than their own forces. Although hindsight may tell us that the allied Greek forces of 490 and of 480/79 were not in fact weaker than the Persian forces in the specific circumstances under which they fought the major battles, the Persian Wars were celebrated by Greeks, and by Athenians in particular, as a victory by an underdog over a powerful, almost numberless invader.³⁹ Because Thucydides does not present the Melian hope to survive the Athenian invasion of 416 as obviously more far-fetched than the Athenian hope to survive the Persian invasions of 490 and 480, the Athenian criticism of the Melians is hypocritical.⁴⁰

Herodotus emphasizes the almost inconceivable size of the Persian force that invaded Greece under Xerxes in 480, not only by calling it ‘by far’ the largest expedition known to him (7.20.2) but also by taking 15 pages of Greek text (7.59–100) to describe an army that he estimates at 1.7 million men (7.60), 1,207 triremes (7.89) and 3,000 support vessels (7.184). He reports sensational feats performed by Xerxes’ army, such as digging a canal across the isthmus northwest of Mount Athos (7.22–24), bridging the Hellespont (7.33–37), drinking entire rivers dry (7.43, 7.58, 7.108) and eating meals so large they required months of preparation (7.119). Herodotus describes the loss of some of the Persian fleet (400 ships lost in a storm off the coast of Magnesia, 7.190; 15 captured near Artemision, 7.194), but gives the impression that the Persians still greatly outnumbered the Greeks at the battle of Salamis, at least two to one.⁴¹ This ratio is conservative compared to the earliest Athenian description we have of the battle, from Aeschylus’ Persians. In that play, produced in 472, the Persian messenger reports to Queen Atossa that the battle involved either 1,000 or 1,207 Persian ships compared to 300 (or 310) Greek (Pers. 339–43).⁴² Modern scholars generally consider these accounts to exaggerate the Persian forces.⁴³ But Thucydides does not suggest that he or anyone else thought the reports inflated. While he does claim the Peloponnesian War was greater than the Persian Wars by various standards, such as the duration of the later contest and the total extent of the strife involved, both on and off the battlefield (1.23), Thucydides does not call into doubt the size of the Persian invading force that his readers would have learned not only from sources such as Herodotus and Aeschylus but also from local and oral traditions.

The glory of the Greek victory over Persia, and especially of the Athenian role in that victory, became a topos of Athenian rhetoric in particular. Surviving sources repeatedly claim or imply that Athenians demonstrated daring or bravery in facing steep odds by opposing the far greater numbers of Persian forces. Already by 432, according to Thucydides, the Athenians in Sparta found it ‘annoying’ (δι’ ὄχλου, 1.73.2) to be constantly mentioning their role in the Persian Wars. They claim that they first ‘braved danger alone’ at Marathon (μόνοι προκινδυνϵῦσαι, 1.73.4), that they displayed ‘eagerness and much greater daring’ at Salamis (προθυμίαν δὲ καὶ πολὺ τολμηροτάτην, 1.74.2) and that they ‘took a risk for a city existing on slender hopes’ (ὑπὲρ τῆς ἐν βραχϵίᾳ ἐλπίσι οὔσης [sc. πόλϵως] κινδυνϵύοντϵς, 1.74.3). Rather than give in to fear, the Athenians took the desperate measure of abandoning Attica for the sea. Overall these Athenian boasts are fairly measured, at least when compared with the examples I discuss below, but their reserve is explained by the Spartan audience here: any specific Athenian claim to have fought against immense numerical odds would be easily trumped, in the minds of the Spartans, by the Spartan bravery at Thermopylae.⁴⁴ Most importantly, the Athenians do not suggest that the decision to fight the Persian invasion was based on a calm calculation that Greek forces were roughly equal to Persian forces or that victory over the Persians was reasonably likely.

The steep odds, and Athenian daring and bravery in face of them, are emphasized even more in Athenian sources from the first half of the fourth century. Lysias, in his funeral oration from around 390 (Lys. 2), claims that the Athenians at Marathon ‘ran a great risk’ (διϵκινδύνϵυσαν, 2.20) in facing 500,000 Persians (πϵντήκοντα μυριάδας, 2.21).⁴⁵ They ‘did not calculate the dangers of war, but believed that a glorious death leaves behind the undying fame of excellent men’ (2.23). They showed that it is better to fight for freedom with a few men than to be enslaved as subjects of a king (2.41). Stephen Todd argues that Lysias’ funeral speech is probably the most representative surviving example of the genre, giving ‘a better sense of the norms of the genre than we gain from the more spectacular but more maverick speeches of Thucydides or Hypereides’.⁴⁶ This exemplary status may be reflected in how the funeral speech offered by Socrates in Plato’s Menexenus provides a similar report of the Athenians in the Persian Wars.

In the Menexenus Socrates, delivering a speech he claims was taught him by Aspasia, emphasizes Athenians overcoming great odds in the Persian Wars. At Marathon the Athenians showed ‘that the Persian power was not invincible, and that any great number of men and any amount of wealth yield to virtue’ (240d).⁴⁷ The Greeks who fought against the Persians warded off the ‘greatest and cruellest’ danger (241c). This speech is likely a satire of actual funeral speeches, but ‘satire only works if it is based on something real’.⁴⁸ Plato’s emphasis on the trope of Athens as an underdog against Persia shows he found it typical of late fifth-century funeral orations.⁴⁹ Further examples abound, and are not limited to funeral speeches.⁵⁰ What they have in common is the claim that much of what was praiseworthy in the Athenians who fought against the Persians, what should be imitated by later Athenians, was their willingness to fight against an apparently much greater enemy.

If Thucydides’ text encourages us to attribute to the envoys a transtemporal corporate Athenian identity, we may also attribute to them this well-attested Athenian attitude to the Persian Wars, since they do not disclaim it. That attitude allows us to see the envoys’ hypocrisy clearly. Just before the Melian dialogue proper, Thucydides states the size of the invading force: 38 ships, 3,100 hoplites, 300 archers. The Athenians dwell on the imbalance of strength between themselves and the Melians. They point out to the latter that they are ‘weaker than οthers’ yet facing the ‘masters of seafaring’ (5.97); the fight is not fair because the Athenians are much stronger (5.101). They claim the Melian ‘forces are too small to endure those already arrayed against them’ (5.111.2) and that they are facing ‘the greatest polis’ (111.4). The considerations the Athenians offer the Melians are precisely the considerations Athenians bragged about ignoring for over a century after the Persian Wars: a great imbalance of forces and impending disaster. Their arguments that the Melians should surrender peacefully, when the Athenians had refused to surrender to Persia, are blatantly hypocritical. In the conclusion I will elaborate on the ethical problem of this hypocrisy.

iv. The Athenians as Cohen’s kidnappers

Thucydides portrays the Athenians as like Cohen’s kidnappers in having deliberately put the Melians in the desperate situation in which they find themselves and criticizing their preference for resistance. The invasion of Melos appears as an unprovoked aggression with no legitimate immediate justification. Thucydides tells us merely:

The Melians are colonists of the Spartans, and they were not willing to submit to the Athenians like the other islanders, but at first they stayed quiet, on neither side; later, when the Athenians tried to compel them by ravaging their land, they came into a position of open warfare (ἐς πόλϵμον φανϵρὸν κατέστησαν). (5.84.2)⁵¹

The Greek does not make clear the subject of κατέστησαν, but it implies the Athenians were unprovoked aggressors.

The Athenians do hint, near the beginning of the dialogue, that they could produce a pretext for their invasion if they wanted to. They use a rhetorical praeteritio to say they will not make a long speech ‘with fine phrases, such as that we rule justly because we defeated the Mede or that we are attacking now because we were wronged’ (5.89). But they fail to specify a Melian offence, and even the Melians, desperate for any argument, do not disavow or justify any specific injury they might have inflicted on Athens. These passages foster the conclusion that there is no reasonable justification for the Athenian attack: the Athenians are in the wrong. Recall that Cohen’s kidnapper does not justify his apparently immoral act of kidnapping, the act whereby he puts the parents in the position of deciding whether to ransom their child. Aside from their specious arguments invoking compulsion, which I discuss below, the Athenians similarly offer no justification for their apparently immoral act of putting the Melians in the position of deciding to submit or resist.

IV. Conclusion: the Athenians’ ethical offences

We can now draw together the threads of the argument to see where precisely the ethical offences of the Athenian envoys lay. These two distinct offences are both speaker-relative and derive from their criticisms of the Melians. Like Cohen’s kidnapper, the Athenians offend by denying their own human agency. They treat their invasion of Melos as a simple fact of nature rather than a voluntary choice. The envoys make this view explicit in one of the most quoted phrases of the dialogue:

From what we hear of the gods and clearly know of men, we believe that an eternal compulsion of nature (ὑπὸ φύσϵως ἀναγκαίας) causes them to rule whenever they have power to do so. We did not make this law (νόμος) and we were not the first to use it after it was established. (Thuc. 5.105.2)

The Athenians claim to be acting under compulsion (ἀναγκαία) and law (νόμος), not willingly. In contrast, they repeatedly address the Melians as being confronted with a ‘choice’ (αἱρέσϵως, 5.111.4) and ‘deliberating’ about how to act. The Athenians use terms for deliberation (forms of βουλϵύϵσθαι or βουλή) seven times to describe what their Melian counterparts are doing.⁵² These words are never used in the dialogue to describe the Athenians.⁵³ Aristotle’s discussion of voluntary actions (Eth. Nic. 3.1–5) is crowded with similar vocabulary: virtue and vice are ‘up to us’ (ἐφ’ ἡμῖν), the result of deliberation and voluntary choice. I argued above that the Athenians make an ethical argument that the Melians ought to value their own safety above any concern about ‘shame’ or ‘baseness’. By attributing deliberation and choice to the Melians, describing their actions in words evoking an ethical viewpoint, while denying their own agency in having brought about this state of affairs, the Athenians asymmetrically apply an ethical standard to the Melians but not to themselves.⁵⁴ Therefore, not only was their attack on Melos ‘morally vile’, to use Cohen’s expression, their denial of its voluntary nature invites further ethical opprobrium from the reader.⁵⁵

The Athenians also offend with their hypocrisy. I argued above that Thucydides’ text encourages us to attribute to the Athenian envoys the well-known Athenian pride in having faced, against heavy odds, a large and dangerous invading force during the Persian Wars. I also showed that the Athenians criticize the Melians for considering taking a similar risk. By criticizing the Melians without acknowledging (much less justifying) their own previous ‘wrongdoing’ on this count, the hypocritical Athenians lack moral standing and become moral offenders.

The dialogic nature of the Melian dialogue heightens the reader’s perception of the Athenian arguments as ethically offensive by creating the impression that we are witnessing both a private conversation and a philosophical discussion. The discussion is literally private, held among a small group behind closed doors (Thuc. 5.84.3). But also, as both Maria Fragoulaki and H.L. Hudson-Williams argue, back-and-forth dialogue of this sort was considered a ‘private’ genre in Athens in the fifth and fourth centuries, as opposed to the ‘public’ genre of longer alternating speeches frequently represented in Thucydides.⁵⁶ The Athenians explain that if they were giving a speech to ‘the many’ they would use attractive but deceptive arguments (5.85), making explicit that a dialogue instead permits the Athenians to be both blunt and offensive.⁵⁷ Thucydides’ early readers would have known that including such blunt arguments in a public diplomatic speech was unrealistic. By setting them in a private dialogue, where it is easier to believe the Athenians might actually have spoken them, Thucydides makes the arguments more tangible, with a sharper edge.

The dialogue format also heightens sensitivity to the ethical implications of the Athenians’ comments because dialogue was, by the time of the Peloponnesian War, a notably philosophical genre used to air intellectual debates, including ethical debates.⁵⁸ The dialogues of Socrates, as recreated by Xenophon and Plato, are the best-known representations of such conversations, but Plato’s dialogues in particular depict Socrates participating in an extant intellectual culture.⁵⁹ Often in such arguments one party tried to maintain a consistent position with no internal contradictions while the other sought to show that a particular statement led to contradictory conclusions.⁶⁰ What may have been a Socratic or Platonic contribution was to insist that the views defended must also be sincerely held.⁶¹ The Melian dialogue is recognizable as a form of philosophical conversation not only because of its dialogic and private nature but also because of the distinctly ethical questions of justice that it raises.⁶² I suggest that in the minds of Thucydides’ early readers, familiar with this genre of competitive dialogue, these features would have heightened awareness of the internal consistency of the disputants’ arguments, and perhaps also, in a Socratic vein, of their sincerity. In this way the dialogue form makes more apparent the Athenians’ ethical offences, rooted in inconsistency and insincerity. The Athenians, like the hypocrite and Cohen’s kidnapper, are inconsistent in unjustifiably applying different ethical standards to the Melians and to themselves and past Athenians; and this inconsistency indicates the insincerity of at least some of the Athenians’ ethical claims.⁶³

As I noted earlier, the more ambitious version of my argument claims Greek ethical views were largely in line with those underlying the analyses above. I have presented evidence that Greeks condemned what we would recognize as hypocrisy and that it was a contemporary view that ethical judgement was most appropriately applied to voluntary acts. Because these two attitudes are central to my ethical reading of the Melian dialogue, we have reason to believe that my interpretation was available to Thucydides and his early readers, even if their diagnosis of the flaws in the Athenian arguments might not have been identical to mine.⁶⁴ Thucydides’ readers likely would have responded to the dialogue in different ways, depending on their views of the Athenian empire. Those who thought the empire justifiable could probably have convinced themselves that the Athenian arguments and subsequent slaughter of the Melians was an aberrant excess, and some might, like Bosworth, have blamed the Melians more than the Athenians. Those who came to regret the empire, or never endorsed it in the first place, may have seen the hypocrisy and rejection of agency on display in the dialogue as symptomatic of a structurally flawed political system, one that (not coincidentally) underwent significant reforms at the end of the fourth century.⁶⁵ As a difficult literary work aimed at ‘serious’ intellectuals (1.21–22), many of whom held anti-democratic political sentiments, Thucydides’ History likely had many imperial critics among its readers.⁶⁶

At the beginning of this article, I asked what Thucydides is indicting with the Melian dialogue. The Athenians’ greatest crime at Melos was their unprovoked invasion and mass slaughter. They also deserve blame for the cynical and logically incoherent argument for ‘realism’ in international affairs that they express to the Melians. It is largely because of these elements that the Melian dialogue has been infamous for so long. My interpretation of it, unlike Bosworth’s, removes none of the Athenian blame for those acts. And, again unlike Bosworth’s, it keeps the focus more on the Athenians than on the Melians, while never denying the validity of the anger and regret that readers might feel at the stubbornness of the Melian leadership. But I have focused on features of the dialogue that I believe have been perceived by many but not previously singled out for analysis, features reflecting negatively on the Athenians. If correct, my interpretation of the dialogue matters for our broader view of Thucydides: we see here not a Thucydides excessively biased in favour of Athens, as Ernst Badian argues; not a Thucydides insensitive to the moral aspects of Athenian imperialism, as presented by Jacqueline de Romilly; not a Thucydides who views Greek international relations as a realm where the only considerations are power and fear, with no place for morality, as portrayed by Arthur Eckstein.⁶⁷ Instead we find a Thucydides critical of the Athenians, ethically judgemental of their imperialism and insistent that even self-styled ‘realists’ in fact base their advice on ethical considerations. The Melian dialogue is Thucydides’ indictment not only of Athenian imperial aggression but also, as Dionysius of Halicarnassus recognized long ago, of their ‘most disgraceful arguments’, arguments that come off in a worse light by virtue of being spoken by Athenians.