I Greek Tyranny and the Rule of Law

Modern parlance recognises a linguistic variance between ‘king’ and ‘tyrant’. The adjectives ‘kingly’ or ‘regal’ tend to imply a judgment of positive value about the ruler or leader to whom the description is being applied. Conversely, the label ‘tyrannical’ signals something negative, and even derogatory. The semantic distinction goes back at least as far as the sixteenth century and, in an important sense, back to Greek antiquity, which invented the idea of ‘king’ as good ruler and ‘tyrant’ as evil. Early modern Europe evolved the doctrine of the divine right of kings, according to which the king was a legitimate ruler who drew authority from God. Elsewhere, the Enlightenment challenged the moral distinction between kings and tyrants. Whereas Hobbes disregarded tyranny as conceptually meaningless, Locke saw tyranny and lawful rule as conceptual polarities. These discussions go back to Aristotle. Politics III (1279^b5–10) famously defines τυραννίς (‘tyranny’) as a deviation from βασιλɛία (‘kingship’) but sees both as subspecies of one-man rule or monarchy (μοναρχία). Aristotle in the Nicomachean Ethics likewise differentiates king from tyrant stating that, whereas the one possesses the means to devote attention to the interests of subjects, the other pursues his own good (ὁ μὲν γὰρ τύραννος τὸ αὑτῷ συμφέρον σκοπɛῖ, ὁ δὲ βασιλɛὺς τὸ τῶν ἀρχομένων) (Arist. Eth. Nic. 1160^a1–1160^b15). In consequence, ‘king’ and ‘tyrant’ have mostly been morally counterposed.

In the nineteenth century, the theoretical distinction between ‘king’ and ‘tyrant’ was explored in Drumann's pioneering treatise entitled De Tyrannis Graecorum, but in the English-speaking world, perhaps the most influential exponent in the twentieth century of the notion of a tyrant as usurper was Andrewes, who read the definitions of Aristotle back into earlier sources, where the distinction between ‘king’ and ‘tyrant’ is not readily felt.Footnote ¹ Andrewes argued that tyranny had three main explanations, the hoplite revolution, racial polarities, and a rise of a new moneyed class which began to challenge the economic dominance of the older and more established aristocracies of Greece.Footnote ² The idea that tyrants came to power on the back of a shift in military technology has won wide favour, especially in the works of Berve (Reference Berve1967), Pleket (Reference Pleket1969), Luraghi (Reference Luraghi1994) and Parker (Reference Parker1996). Others have dismissed that notion, arguing that the synchronisms are too difficult to establish to postulate with any confidence that tyranny and hoplite warfare were connected causally, even if tyrants may have relied upon hoplite warfare as military methods changed.Footnote ³ For most of the twentieth century, the belief persisted that there had been a distinction between conventional rulers and tyrants, seen to be usurpers. That notion was then challenged in a revisionist paper by Anderson, who pointed out that the evidence of the sixth century does not support a hard-and-fast distinction between tyrants and other types of ruler in Archaic Greece.Footnote ⁴ Anderson confronted the idea that tyranny should be understood as a normative construct, implying something negative about the ruler to whom the title was given. Studies of specific tyrannies, especially the Peisistratids at Athens, have undermined the claim that tyrants rose to power on the back of socio-economic changes, where the evidence seems to suggest that Peisistratus depended upon his own private bodyguard rather than upon a self-identifying social ‘class’, distinct from the aristocracy.Footnote ⁵ One trend has held that tyrants were little more than victors within a competitive world of rival aristocrats jockeying for power and dominance within the Archaic communities in which they held sway.Footnote ⁶ According to that view, there is no need to understand a tyrant as a challenger to a pre-existing aristocratic social class. Even more recently, it has been argued that the tyrants of early Greece continued in the tradition of the Homeric βασιλɛῖς, for convenience (though inadequately) translated as ‘kings’, emulating all the defining hallmarks of Homeric ‘kingship’, but having to establish themselves in a world which was beginning to recognise the rule of law, where the arbitrary rule of one ruler was increasingly being challenged.Footnote ⁷ Thus, tyranny was nothing new in the Archaic age, but what was new was the possibility for society to cohere without the ruling hand of a tyrant or monarch.

Two articles published very recently have maintained that monarchical rulership, far from being an historical exception, was the general rule in Greece down to the late Classical period and even beyond. Rhodes (Reference Rhodes2019) has argued that the fifth century was an exceptional period in Greek history, when political stability could be maintained without resort to one-man rule. This might explain why Thucydides at the end of the fifth century could hark back to a so-called ‘age of tyrants’, as if a past phenomenon (Thuc. 1.13), when tyranny was commonplace and re-entered the limelight when larger power structures, like the Delian and Peloponnesian Leagues, were not available anymore to uphold democracies or oligarchies in areas of the Greek world which had once fallen within the respective orbits of Athens and Sparta. In a related way, Mitchell (Reference Mitchell2019) has argued that negative attitudes to tyranny began to develop no earlier than in the aftermath of the Persian Wars, and that a morally affirmative notion of kingship evolved at Athens in the century that followed, at a time when democracy was increasingly held up to criticism among an intellectual elite. In support, Mitchell contends that the theory that ‘the best man must rule’ finds its most likely origin in the great (though probably fictitious) constitutional debate at Persia (Hdt. 3.82.2–4), and gained momentum in intellectual circles, when democracy was being held up to scrutiny.Footnote ⁸ By the fourth century, those theories had grown current in ‘monocratological’ pamphlets of Xenophon (Xen. Cyr. 1.4.2–4; 1.6.7–8; Oec. 7.3–10.13) and Isocrates (2.9, 15–16, 21–3; 3.32, 41). Mitchell goes on to claim that the Funeral Oration entails a critique of democracy by rotation (Thuc. 2.37.1), even though ideas of geometric equality were not developed in Greek political theory until a century later (see Isoc. 7.21–2; Pl. Resp. 8.558c; Leg. 6.757bc; Arist. Pol. 5.1301^b29–32).Footnote ⁹

Though in line with the prevalent trend to erode nuanced semantic distinctions between ‘kings’ and ‘tyrants’ until the age of the political philosophers, both revisionist approaches raise as many questions as they answer. The second speaks of a ‘positive theorising of kingship’ without defining what is meant by ‘kingship’, and moreover goes on to claim, in the absence of decisive evidence, that the moral schism between democracy and one-man-rule originated in the years following Marathon, when the Persians attempted to reinstate Hippias as tyrant of Athens.Footnote ¹⁰ No doubt, the victory over Persia and the repulse of Hippias will have reinforced aversion to monarchical systems, but aversion to tyranny as a political concept at Athens was long-standing. In the seventh century, a notable called Cylon led an unsuccessful bid to establish tyranny at Athens (Hdt. 5.71; Thuc. 1.116; Plut. Sol. 12.1–2). Whether this coup resulted in the passage of the first anti-tyranny law is disputable, but there can be little doubt that by the early sixth century, the first laws against tyranny were in place.Footnote ¹¹ Over a century later, when the last of the Peisistratids were expelled, the dēmos united around Cleisthenes to resist Sparta's attempt to install Isagoras as tyrant (Hdt. 5.70–74.1; Ath. Pol. 21.3). Though Herodotus uses morally ambivalent language to describe Cleisthenes’ motives when courting popular support, the point that democracy had an intrinsic advantage over tyranny is clear. The first claims that ‘in the fifth century … there were still kings or tyrants, without as yet a clear distinction between the two terms: in some cases at least they were not autocrats but there were also civic institutions with which they had to engage’.Footnote ¹² The problem here is that English, rather than Greek, is applied in analysis. What is perhaps meant in context is that βασιλɛύς and τύραννος could overlap, as is clear in the case of Alexander I of Macedon, described by Herodotus (9.44.1) as στρατηγός τɛ ἐὼν καὶ βασιλɛὺς Μακɛδόνων, but elsewhere the system of rulership in Macedonia is described as a tyranny (8.137.1).Footnote ¹³ Even if there was a conceptual overlap, it is misleading to claim that the terms were not differently nuanced.

Revisionist efforts to undermine the old distinction between βασιλɛύς as good and τύραννος as bad ruler are justified insofar as moral differentiation between the variant species of monarchy is not firmly set until the second half of the fourth century, with Aristotle. But in another sense, discussion to date has been flawed. Deconstructions of Aristotle have tended to focus on a moral distinction but have failed to explain why Aristotle produced the distinction. Political philosophers, both ancient and modern, often favoured clean conceptual categories without the messy nuances of everyday application and usage. But modern interpreters know no better than Aristotle, the greatest political theorist of the ancient world, who possessed a vast range of documentary material lost to us. Aristotle may have used ‘king’ and ‘tyrant’ in a special way to drive home the point that some rulers, whom he called ‘kings’, ruled in accordance with the law and others, whom he called ‘tyrants’, did not. Even if that distinction is less obvious earlier, rulers ruled either within legal parameters or without reference to law. Once Greeks developed the rule of law, arbitrary government fell into disrepute. This is not to deny that tyrants could reign long after the first laws were written. For example, Pindar eulogised the king of Aetna who ruled in accordance with the established laws (Pyth. 1.61–9). Yet, even those rulers who continued to rule after the emergence of political systems that respected the rule of law would no doubt have needed to eschew the negative connotations which tyranny had acquired.Footnote ¹⁴

Tyrants are often depicted as guarantors of the law. In the fourth century, Dionysius of Syracuse is represented as having put executive decisions about war and peace to the vote (Diod. Sic. 14.45). Pittacus of Mytilene was elected αἰσυμνήτης to combat στάσις (Arist. Pol. 3.1285a). Pheidon of Argos introduced a system of weights and measures (Ath. Pol. 10.2). Cleisthenes of Sicyon re-organised the tribal structure of the city (Hdt. 5.67) and was commemorated as a law-abiding autocrat (Arist. Pol. 5.1315b). Similar connections to justice and the rule of law are made for Mycerinus of Egypt (Hdt. 2.129), Cadmus of Cos (Hdt. 7.164), Gelo of Syracuse (Hdt. 7.163–4), and Cypselus of Corinth (Hdt. 5.92b). Superficially, it might be tempting to infer from these examples that the polarity between tyranny and law should be eroded. It may be that many of the tyrants of Classical Greece ruled fairly and justly, perhaps to the point that these were the rule and not the exception. These tyrants modelled their rule on the precedent of the Homeric βασιλɛῖς, the most famous among them Achilles, who resolved arguments in the presence of the assembled host and, at the funeral games of Patroclus, arbitrated between the contestants (Il. 23.488–98, 536–41, 551–2, 576–8, 822–5). In the Homeric paradigm, the βασιλɛῖς were expected to uphold right and justice (Il. 9.295–8; Od. 19.107–15; Hes. Op. 37–8). The tyrants themselves followed that paradigm as far as they could. Peisistratus kept good order in exchange for regular tribute (Hdt. 1.64; Thuc. 6.54; Ath. Pol. 16.4). At face value, it could be argued that law-abiding rule did not exclude tyranny.Footnote ¹⁵

The crucial point, however, is not whether a particular rule was just, but whether justice in a city was guaranteed by the rule of a tyrant, as was the case with Peisistratus and other enlightened rulers of his ilk, or by the rule of law, as in the case of democratic Athens. To call governance just is not the same as to claim that such governance operated according to the same criteria which underpinned the rule of law. As an analytical category, the rule of law has several key defining characteristics. The first is that the law should apply to all citizens equally, except where objective differences justify differentiation, such as mental incapacity or other extenuating circumstances when a crime has been committed. Another is that all public officials should be held accountable for their actions. A third requirement is that the law should be openly and universally accessible, and a fourth that there should be no punishment outside the law for offences committed against private individuals or the state.Footnote ¹⁶ Clearly, the rule of law implies justice, but the second is a necessary, not sufficient, condition of the first. Legal theorists, in addition, distinguish between ‘thick’ and ‘thin’ definitions, the former incorporating universal human rights, the latter omitting that requirement. To assert that Greeks respected the rule of law could only amount to a ‘thin’ definition insofar as they lacked a Bill of Rights comparable to the American Declaration of Independence or the Déclaration des droits de l'homme of 1789, which precipitated the French Revolution. Some have argued that the Greeks never got close to perfecting rule of law as understood in modern times.Footnote ¹⁷ Yet while there are some obvious differences between ancient and modern conceptions, the most important being the absence in antiquity of any known objection to the institution of slavery (see Pl. Resp. 5.468a–b; Arist. Pol. 1.6.1255a 6–7; Xen. Cyr. 7.5.73) and the use of torture in the courts,Footnote ¹⁸ the evidence presented by the best known of the Greek cities, Athens, illustrates that the rule of law was conscientiously applied, and thereby democracy could function.

Written law and the rule of law are distinct but related concepts. The rule of law implies the existence of written laws in most cases, though Sparta could here be cited as an exception. But written law by itself does not guarantee the rule of law. Though early Greece began to have written laws which can be connected to the rise and evolution of democracy, this should not mean that it understood the rule of law, at least not as understood at Athens in the fifth and fourth centuries. Close inspection of some very early laws shows that the idea of the rule of law was early in the making, which was incompatible with tyrannical rule. An inscription from Dreros, dating from the Archaic period, specifies that no man should hold the office of κόσμος more than once within a ten-year cycle (ML 2). A sixth-century law from Gortyn forbids a man to be κόσμος within three years or κόσμος with foreigners within five (IC IV.14 G-P). In neither community could a tyrant have been possible, because of the principle of rotation. Magistrates in Archaic cities which had rejected tyrants were held accountable. A law from Lyttos in Crete, dating c. 500, specifies that a κόσμος could be put on trial (SEG 35.991). Two sixth-century laws from Eretria imposed fines on officials who failed to perform their duties (Nomima 1.91). Similar fines on delinquent officials are in evidence in other parts of Greece dating as far back as the seventh century.Footnote ¹⁹ Citizen rights at Gortyn were protected by a system of fines on any κόσμος who seized a citizen illegally (IC IV.72: 1.51–5). An inscription from Locris dating from the sixth century prescribes penalties for the δημιουργοί if they profited from the sacred offerings to Apollo (ML 13B). Such a system of accountability and control over magistrates was present at Athens a century later (Ath. Pol. 45.2), and a similar constraint was imposed on the Spartan ‘kings’, who could be tried and condemned (Hdt. 6.65–6, 72, 82, 85). The laws at Gortyn forbade seizure without fair trial (IC IV 72 l.2–2.2).Footnote ²⁰

Under a political system that respected the rule of law, citizens were protected under the law and officials held to account. In the second half of the fourth century, Aeschines was clear that the key difference between a city under the law and another under a tyrant is that the first breeds respect for the laws, whereas the second is subject to the whim of one man (Aeschin. 1.4–6). Early laws were frequently protected by entrenchment clauses which impose severe penalties on those who tried to change or undermine them. An inscription from early sixth-century Argos imposes a curse on any who ignores the laws (IG IV.506). A law from Locris dating from about the same time imposed a curse, confiscation of property, and destruction of the house of any who proposed further division of property (ML 13A. 7–15). By putting measures in place whereby laws could not be changed or disregarded either by citizens or by officials, the early lawgivers ensured that tyrants were unable to rise and impose their unaccountable will on communities which were governed by the rule of law. This crucial innovation meant that tyranny, by definition, was becoming an anathema in the Archaic period.

The emergence of the rule of law in Greece meant that democracy became possible. In turn, this meant that tyrants needed to accommodate their rule to lawful principles if they were to survive politically. The emergence of written law itself did not render tyranny impossible, as can be demonstrated by the fact that many tyrannies continued to spring up in the sixth century and beyond. Yet political and legal equality was impossible to maintain if a city was under the governance of a sole ruler, as Aristotle (Pol. 5.1313^a3–5) clarifies. Once the rule of law became established, it became possible for an early community to govern itself without reference to the will or authority of a tyrant. The recent attempts to describe Archaic tyranny in an evaluatively neutral way, or indeed to claim that monarchy was the rule in Greece rather than the exception right down until the Classical period, underestimate that once a community could govern itself by laws, tyrants and despots were no longer needed, even though, to be sure, they continued in many parts of the Greek world. Athens from the second half of the fifth century began to develop a positive theory of monarchy, but the ‘kings’ of Attica were mythological constructs which plugged into a democratic doctrine of virtue and self-sacrifice, qualities which the mythical kings represented. None of this, however, should imply that in a practical sense, monarchy was ever coveted or seriously proposed as an institution during the Classical period. The ‘good’ monarchs of popular mythology, such as Theseus, were democratic heroes through and through. Real monarchy, which meant tyranny, was an unpalatable alternative to democratic government, and the hatred harboured for tyrants in the fourth century is of a piece with a tradition of anti-tyrannical thought reaching back to Archaic times.

II Tyranny in Popular Perception in Greece

It has been argued thus far that though tyrants did exist after written laws came into existence in Greece, tyrants nevertheless had to rule in accordance with legal principle if their rule was to survive. Those who failed to do so, like Hippias in Athens, quickly discovered that their rule was untenable. This is important to bear in mind when we turn to consider how Greeks of the Classical age came to understand tyranny. Tyranny was typically regarded as a negative and destructive force once the Greek cities began to apply the rule of law. From an early stage, tyrants were viewed as arbitrary, lawless, and almost apolitical. There is not the evidence to assert that the objection to tyranny which surfaces in fifth-century literary texts was somehow exceptional or unprecedented. The antiquity of the Athenian anti-tyranny laws shows that from the seventh century in Athens, tyranny was held in suspicion. To be sure, Athens possessed a tyranny in the sixth century, but the tradition is clear that Peisistratus ruled in accordance with the laws (Ath. Pol. 16.10). The case of Athens shows that written laws, which date from the seventh century in their earliest formulations, offered no fool-proof guarantee against the emergence of a tyrant, though it might be said that where there were written laws, tyrannical government needed to be tempered by observance of those laws. More importantly, as argued above, the rule of law was not guaranteed by written law. Many communities had written laws, but never quite developed democracy or the rule of law. Once the latter became established, however, democracy in Greece became possible and tyranny was thus anathematised.

It has become fashionable to claim that negative constructions of tyranny in Greece were late coming. Much of that assertion bases itself on the observable fact that in the Archaic period not all tyrants were necessarily bad. Lavelle (Reference Lavelle2005, 17–29), for example, has studied the Peisistatid tyranny, arguing that the later anathematisation of Peisistratus and his sons was ex post facto. In his own day, Peisistratus did much in public perception to push his claims to heroic ancestry back to the reign of the Neleids of Pylos and, during the sixth-century Athenian war with Megara, modelled a genealogy on Codrus and Melanthus, who according to the myth are said to have defeated Xanthus of Thebes in a war with Athens.Footnote ²¹ Ancient tradition presented Peisistratus as having purified the island of Delos as a symbolic gesture to Ionian heritage (Hdt. 1.64.1–2; Thuc. 3.104.1–2; cf. Hymn to Delian Apollo, 146–64). The extensive building work and growth of prosperity under Peisistratus indicates that whatever his government amounted to, it was far from unpopular or resented at Athens. Even so, the sources are clear that Peisistratus rose to power by factional means and maintained power through force. The tradition that he ruled in accordance with the law needs to be read with a degree of fluidity, in that the anti-tyranny law which predated him was directly violated by his rule. He was ousted twice and was tolerated, perhaps because his rule was mild and was the most reliable way for the time being to prevent Attica sinking back into political disintegration. Once the rule passed to his sons, patience was short-lived, and with Spartan help the Athenians expelled the tyrants.Footnote ²²

Occasion has already arisen to refer to the variant ways in which Herodotus refers to the Macedonian leadership. Describing Alexander I as a βασιλɛύς, Herodotus adopts a specific context, because as general of his army he takes on the role of leadership and appears to occupy a constitutional position. By contrast, when Herodotus describes Alexander earlier in Book 8 as a tyrant, he does so to emphasise the monarchical nature of his rule. Ferrill (Reference Ferrill1978) argued for a broadly negative view of monarchy in Herodotus, but in a fresh analysis of the problem, Fitzsimons (Reference Fitzsimons2017) has recently put forward the case that the matter was not whether monarchy on the Persian model is categorically evil, but the variant styles of government which each of the four successive kings of Persia, Cyrus, Cambyses, Darius and Xerxes, exhibit, with the first being described as πατήρ, the second δɛσπότης, the third κάπηλος, and the last as a consultative leader. If so, the view of absolute rule in Herodotus is not fixed, but this is not to suggest that it was either ambivalent or insensitive to a range of different nuances of rulership. If Herodotus used the concepts of kingship and tyranny interchangeably to describe Macedonian rule, this need not imply that the two were interchangeable without variance of connotation. When describing Alexander I as βασιλɛύς, he does so in a special context, because as general of an army he assumes the role of leadership and appears to occupy a constitutional position.Footnote ²³

Thucydides (1.13) understands the distinction between βασιλɛύς and τύραννος in terms of the proportion of authority that each wielded, with βασιλɛίς enjoying more limited and harnessed powers, whereas tyrants came to power on the back of money and wealth.Footnote ²⁴ Rhodes (Reference Rhodes2019) argues that Thucydides consigned tyranny to the past because the Delian and Peloponnesian Leagues had, in his own day, ruled out the need for tyrants. This is simply to beg the question. The great power blocs of the Hellenistic and Roman periods regularly installed tyrants to prop up their power bases, and if the two main alliances of the fifth century had been concerned only to maintain power at the expense of all else, there is no obvious reason why they could not have resorted to local tyrannies to accomplish that end. The reason they did not is because, as a matter of political principle, tyranny was contrary to the ideologies of both alliances. Athens presented herself as the champion of the freedom of the Greeks and as the protector of victims of injustice, whilst Sparta too saw her traditional role as defender of Greek liberty.Footnote ²⁵ Xenophon attests that the Peloponnesians followed Sparta only when it professed the defence of freedom (Xen. Hell. 2.4.30). Earlier, Sosicles of Corinth had reminded the Spartans, when they contemplated restoring Hippias from exile, of what his own city had suffered under the tyranny of Cypselus and Periander (Plut. Mor. 858). It is otiose to claim, without further argument, that tyranny vanished from Greece in the fifth century for the accidental reason that the two predominant Hellenic alliances outlawed it. The point at issue is that the two great leagues in the fifth century outlawed tyranny because it was no longer tenable.

The underlying question is why those alliances felt the need to outlaw tyranny. Mitchell (Reference Mitchell2019) claims that tyranny as a political concept became anathematised only once the Persians invaded, and the notion of one-man rule had fallen into disrepute. Greeks united twice to resist Persia because of a common loathing of tyranny. Had tyranny been habilitated in Greek ideology at the time of the Persian invasion, it is unclear why the Greeks should have risked everything to resist a new Persian suzerainty if, as in Asia Minor previously, it was possible for tyrants to rule with the blessing of the satrap. Mitchell builds her case on the Funeral Oration of Pericles, which, following Rhodes, she understands to be a critique of fifth-century democracy. The moot reference is at Thuc. 2.37.1, ἀπὸ μέρους, which Rhodes (Reference Rhodes1988) and Mitchell take to be a critique of democracy by rotation. This is to mistranslate ἀπὸ μέρους, as if Pericles were claiming that in an ideal democracy officials are selected not by lot, but on the basis of merit. Pericles’ point, however, is not that democracy should not practise rotation, but that in democratic Athens the tenure of office does not confer honour if unaccompanied by honourable and capable character. Pericles speaks not of an abstract political ideal but of Athens itself.Footnote ²⁶ Whether led by Athens or Sparta, the Greeks of the fifth century anathematised tyranny not because it was politically convenient to do so, but in a deeper sense because tyranny conflicted with established doctrines of freedom, which the Peloponnesian and Delian Leagues, at least by self-proclamation, aimed to uphold. Their professed ideologies are explicable only if tyranny was already intolerable.

Aristotle in the fourth century distinguished kings from tyrants on moral grounds, as if not all monarchical rule was fundamentally bad. The reference at Thuc. 1.13 to kings and tyrants implies a conceptual distinction predating Aristotle, even if Thucydides engages in statements which historically might be difficult to defend. Recent scholarship has been disposed to doubt Thucydides’ definitions on the grounds that many of the so-called ‘kings’ in the Archaic period, and even before, became very wealthy and were not shy to use wealth as a means to bolster their authority, and at the same time many so-called ‘tyrants’ exercised ancestral rule and did so in the interests of the people they ruled.Footnote ²⁷ In post-Mycenaean society, ruling houses in the small Early Iron Age communities could have seated more than one family and in one case may have seated up to two hundred members. The case of Nichoria in Messenia makes the idea that Greek communities postdating the collapse of the Mycenaean world had solitary rulers who claimed uninterrupted lineage over generations prima facie unlikely.Footnote ²⁸ Yet archaeology also shows that these Early Iron Age societies were artistically and economically vibrant and that rulers sought to justify their power by connecting themselves with the warriors of the mythical past.Footnote ²⁹ Mythical scenes on vases dating from the period show that the elites actively sought to link themselves to an heroic narrative dating back to the late Mycenaean age.Footnote ³⁰ The Toumba Building at Lefkandi has often been likened to an heroön, which scholars have argued was used by the local rulers in the tenth and ninth centuries to legitimise their rule by reference to a notion of shared ancestry.Footnote ³¹ A similar picture emerges from Eretria.Footnote ³² Morgan (Reference Morgan2009) concludes that while ancestry in most cases was fictional, it continued to function in an important symbolic way for the rulers of the Early Iron Age civilisations. Thus, even if genealogical claims were fictitious, they were important ideologically and symbolically.

The vanishing of these large ruling houses from the eighth century is often linked to the disappearance of institutional kingship in Greece, but more recently the case has been made that what shifted between the Early Iron Age and the early Archaic period was not how much wealth rulers had at their disposal, but how they chose to display it, with an emerging emphasis being placed on the public display of wealth rather than on its use for private ostentation.Footnote ³³ This would certainly explain the fashion for austerity among tyrants such as Periander of Corinth, Gelo of Syracuse, and Pittacus of Mytilene.Footnote ³⁴ The link between tyranny and wealth had been established as a trope by the time Thucydides was writing. The first use of the word ‘tyrant’ in Greek literature appears in a fragment of the lyric poet Archilochus, who claims to despise the wealth of Gyges and repudiate tyranny (fr. 19 West). Modern scholarship has seen in this evidence for the connection in the popular imagination between tyranny and wealth.Footnote ³⁵ Archilochus was speaking proverbially, but the fragment shows that from earliest attestation tyranny is associated with something lavish, despotic, and foreign. It cannot be claimed that the link between tyranny and wealth was merely dreamt up by Thucydides. From a philological angle, it is crucial to recognise that wealth and tyranny were bedfellows from the point that the latter entered Greek vocabulary. Ostentation, luxury, decadence, and despotism were at the heart of what Greeks had always understood by tyranny. The fact that more enlightened tyrants, like Peisistratus, took to using wealth for public, rather than private, display should not lessen this stereotype. No doubt he, Cypselus of Corinth, and Gelo of Syracuse were each familiar with the advice of Bacchylides to rulers to use their wealth responsibly (fr. 3).Footnote ³⁶ But tyrants and wealth were, it seems, intimately associated.

Mitchell (Reference Mitchell2013, 57–151) has pointed to the commonality of the claim to heroic ancestry which unites tyrants with other types of ruler in early Greece. Of course, there will have been benevolent tyrants, just as there are examples of benevolent dictators in the modern world, but to recognise this is not to deny that autocratic rule was regarded as distasteful at best, uncivilised at worst. Occasion has already arisen to note the early advent of anti-tyranny laws at Athens. The lawgiver Solon expressly renounced tyranny, as did Archilochus, though some have argued that this was an effort to rid himself of the smear which associated itself to the massacre of the Cylonians.Footnote ³⁷ Whether, however, Solon can be argued to have been closer to a tyrant than not, a contention that could only ever rest on speculation, none of these observations subvert the fact that by the start of the sixth century, if not from before, tyranny and lawful government had a tenuous relationship. It might be objected that Solon anyway was offered tyranny, which he refused, and a generation later the Athenians installed a tyrant in the person of Peisistratus. How seriously, then, were the anti-tyranny laws taken? This would be to miss the larger point, that the anti-tyranny laws came about with the rejection of Cylon and the first attempt to write down the laws under Draco. Legislation against tyranny at Athens coincided with the writing down of the first laws. That does not mean, of course, that tyranny was an impossibility thereafter, as events would later disprove, but what it does mean is that tyrants needed to be aware of civic and legal institutions, as was the case with Peisistratus. As the later tyranny of Hippias shows, those tyrants who refused to do this were politically short-lived. As soon as the Peisistratids were out, the cult of the tyrant-slayers became entrenched in the political culture of democratic Athens, though some have argued for a later origin.Footnote ³⁸ Yet hatred of tyranny was ancient, and even if it took a while for Athenians to establish an odium against Peisistratus, this does not alter the fact that tyranny was viewed as distasteful.Footnote ³⁹ When Pericles described the Athenian Empire as a tyranny, the point was precisely that the Empire had to maintain itself by force, not by the goodwill of its subjects. When Thucydides famously characterises Periclean Athens as the ‘rule of one man’ (2.65), an intended irony is at play, namely that though Athens was a democracy, and therefore the opposite of a tyranny, even under the most accomplished and developed of democracies it was possible for tyrant-like figures to exert their influence. But the fact that Thucydides uses this language should not be taken to mean that it was value neutral; rather, its pungency lies in the conceptual contradiction which it conveys.

The suspicions against tyranny were harboured not only at Athens but in other parts of the Greek world during the Archaic period. In the lyric age, there was a growing awareness of the dangers that monarchical rule could bring, and even if the word τύραννος is not always used, the poets were au fait with a common political thread, the nature of tyrannical rule and the deleterious effects which despotic government wrought if not tempered by the rule of law. The moralising poem of Pindar warning Hiero of Syracuse not to commit the proverbial errors of Phaleris of Acragas, who burned his political opponents alive in a hollow brazen bull, shows that the reality of one-man rule was omnipresent towards the end of the Archaic age and that its hazards were well known (Pind. Pyth. 1.96–7). The image that tyrants knew nothing of justice but ruled in accordance with their own whims was well established in the Archaic period. Theognis of Megara warned Cyrnus to observe religious piety and slay the tyrant:

This instruction to kill a tyrant with moral impunity is difficult to wed with the theory that Greeks in the Archaic age felt morally neutral about tyrants. As Raaflaub argued, the democratic ideals of fifth-century Athens originated not in the fifth century but much earlier, when the Greek elites began to confront tyranny and developed norms which, later, fed into the anti-tyrannical doctrines of the democratic and oligarchic states of the fifth century.Footnote ⁴¹ It is therefore difficult to maintain that the hatred of tyranny was a concoction of the fifth century, whereas previously tyranny had been normalised. It was precisely because tyranny, well before the advent of the Classical age, had fallen foul of moral norms that democracy could develop.

In the fifth century, the tragedians were preoccupied with monarchical government. It is far from true that rulers were always wicked, but even by the age of the tragedians, tyranny had acquired a negative connotation. Sophocles’ greatest tragedy, titled Oedipus Tyrannus, whilst eliciting great sympathy for the king of Thebes, is a tale of a usurper who acquires power after an unseemly outrage is committed against both parents; once power has been acquired, Oedipus acts like a tyrant towards Creon and Teiresias. Yet even if a ruler did not receive the designation of τύραννος, it was widely understood that rulers who ruled selfishly, despotically, and without reference to the will of their subjects were behaving in ways which, according to an emerging, if unwritten, political code, had become anathematised as being contrary to the principles of well-ordered society which understood the rule of law. Tragedians refer to monarchical government often, but not always, in a pejorative light. An example is the dream of Queen Atossa, who envisaged two sisters in different garb, one Persian, the other Doric. In the dream the king yoked each of the two women to a chariot, whereupon one rode along obediently, the other struggled to break free and eventually dashed the chariot in two (Aesch. Pers. 176–99). The story shows how uneasily Greeks sat with the idea of monarchy: Persian garb, luxuriant and decadent, becomes a symbol of how Greeks perceived the East, tyrannical, puffed up, and subservient to an autocratic master. To be a monarch was to be associated with something which was un-Greek, luxuriant, servile, and lacking vigour.Footnote ⁴²

But tragedians did not always regard monarchy as unequivocally evil, provided it tempered itself by the rule of law. Rulers had the choice to govern in accordance with the laws and in the interests of those they ruled, or else to behave in a tyrannical and autocratic fashion. In her reproach to King Creon, Antigone accuses him of becoming a tyrant because he does and says whatever he likes and even starts to personify the city:

The Antigone is a complex play, and it is far from clear that Antigone is always in the right. But the point she makes here would have been widely recognised as true, that autocracy which did not defer to a principle higher than itself was a moral failure. The hateful nature of tyranny is also visible throughout the plays of Euripides, especially in the Phoenician Women and the Suppliants. In the latter, Theseus is made to say that there is nothing more hateful to a city than a tyrant (Eur. Suppl. 429), but earlier claims that he has given sovereignty to the dēmos. The word is μοναρχία, seemingly an oxymoron, but which on closer reading comports the notion of constitutional, lawful government as distinct from lawless, arbitrary tyranny: καὶ γὰρ κατέστησ᾽ αὐτὸν ἐς μοναρχίαν/ ἐλɛυθɛρώσας τήνδ᾽ ἰσόψηφον πόλιν (‘for I established it [viz. the dēmos] in a position of sole-rule/ having liberated this city equal in vote’, 352–3). Of course, not all rulers in tragedy are wicked. King Pelasgus, in Aeschylus’ Suppliants, regularly addresses the assembly.Footnote ⁴⁴ He is a good ruler because he submits the decision to the Assembly, which the Danaids from Egypt do not understand. Antigone is right about the burial.Footnote ⁴⁵

Tragedy is not the only evidence for how Greeks thought about absolute monarchy. The comedies of Aristophanes are rich in allusions to monarchy and its excesses, especially in the way the King of Persia is viewed. The Great King is a despot, oozes in riches, dines lavishly, and anoints himself with fragrant ointments (Ar. Ach. 62, 68–76, 85–6, 88–9, 102; Plut. 170; Av. 486–7; Eq. 1330–2). The image of the oriental despot was typeset by the time of Aristophanes in the fifth century, who drew on stereotypes that went back to Archilochus two centuries earlier. To the democratic mindset, monarchical rule was foreign, un-Greek, unlawful, extravagant, and contrary to civilised society. Neither the comic poets nor the tragedians made distinctions between ‘kings’ on the one hand and ‘tyrants’ on the other. Much more important for them was the contrast between a free democratic society, where no king or ruler existed, and an unfree society ruled over by an arbitrary monarch.Footnote ⁴⁶ Though Zeus is referenced in Birds by Aristophanes as a βασιλɛύς, the comic point is just that Zeus himself had in legend been a usurper, having overthrown the Titans who preceded him to establish rule on Mount Olympus. Zeus is no better than the vile Pisthetaerus who seeks to upend him, and who in various places is described specifically as a τύραννος (Ar. Av. 467–506, 1708). Similarly, Pericles is likened in comedy to a tyrant, which cannot have been flattering (Crat. fr. 258 KA; Adesp. fr. 703). The link drawn in the narrative of Thucydides between the activities of Alcibiades during the religious scandals of 415 and the Peisistratid tyranny, though not immediately obvious, implies an antithesis between democracy and tyranny (Thuc. 6.27.3; 23.2; 53–61; cf. Isoc. 16.38).Footnote ⁴⁷ It has long since been pointed out that Thucydides’ account of Hippias is a good deal more flattering than the account of the assassins, which points to the connection made between the excursus in Book VI and the attempt to establish oligarchy in 415.Footnote ⁴⁸ But in telling the story as he does, Thucydides expressly subverts a popular perception of tyranny as a moral evil. If the excursus indicates anything about how tyranny was perceived in Greece in the fifth century, it was as a form of government for which it was difficult to find a credible moral approbation.

Even if tyranny was regarded positively in some very special and limited circumstances, such as the tyranny of the dēmos in acting as patron of the arts, in the mouths of democracy's critics tyranny was also weaponised as a term of opprobrium (Pl. Grg. 481b–527e).Footnote ⁴⁹ But the crucial point is that unlike βασιλɛύς, which need not have referred to a monarch and seems to have been applied to monarchs only in special contexts, τύραννος could only ever refer to an autocrat. For this reason alone, a case could be made to translate τύραννος, and not βασιλɛύς, as ‘king’. In the Hymn to Ares, it refers to the absolute power of a god (Mart. 5). Its reference to the gods, whose power is unchecked, is evident especially in tragedy (Aesch. PV 753; Soph. Trach. 217; Eur. fr. 136). References to historical figures as tyrants, though not automatically pejorative, always nevertheless comported monarchical rule (Hdt. 1.6; 5.67; 7.10.3; Thuc. 1.14; Xen. Hell. 6.5.33; Ath. Pol. 13.5; Arist. Pol. 1275^b36; Pl. Symp. 182c). Though, in the lyric age especially, use of the term may have been neutral descriptive in some contexts, it could also be accompanied by a negative evaluative implication (Thgn. 1181; Hdt. 3.80; Pl. Grg. 510b; Plt. 301c; Resp. 569b; Soph. OT 873). As an adjective, τύραννος when used in tragedy means ‘kingly’ or ‘regal’ and nothing much more (Aesch. PV 751; Soph. Ant. 1169; OT 588; Eur. Med. 1125; Hipp. 843; Andr. 3; Hel. 478; Tro. 474), but in prose uses it can mean something closer to our modern word ‘tyrannical’ (Thuc. 1.122, 124; Luc. Ner. 2). To behave tyrannically is in one place portrayed in contrast to behaving according to the values of a civilised man (Ath. Pol. 16.2), and by the fourth century tyrannical behaviour is pitted against the behaviour of a βασιλɛύς (Pl. Resp. 575a; Isoc. 5.154). When used as an abstraction to refer to the rule which a tyrant possesses (τυραννίς), in many attested instances the term comports something very negative (Simon. 71; Hdt. 3.53, 81; Ar. Vesp. 417; Dem. 2.30; Pl. Leg. 863e), but it can also refer neutrally to monarchy (Archil. 25; Pind. Pyth. 2.87; 11.53; Soph. OT 535; Eur. Bacch. 43; Aesch. PV 10) without entailing an implied value judgment.