¹ I am very grateful to Drs Amélie Kuhrt and Michael Whitby, and to Professor G. E. Rickman, for their careful scrutiny of an earlier version of this article. Their detailed comments have helped to improve the original substantially, though the responsibility for remaining imperfections is of course entirely my own.

² Walser, G., Hellas und Iran. Studien zu den griechisch–persischen Beziehungen vor Alexander (Darmstadt, 1984: Enträge der Forschung 209), pp. 27–35, at 29, 34Google Scholar, plays down the implications of Herodotus' account and virtually ignores tyranny as a factor in the revolt, which is put down to internal political and leadership struggles in the Greek cities. Graf, D. F., ‘Greek tyrants and Achaemenid polities’, in The Craft of the Ancient Historian. Essays in Honor of Chester G. Starr, ed. Eadie, J. W. and Ober, J. (1985), pp. 79–123, at 80, 84, 86Google Scholar, goes further in denying that there was resentment against tyranny as such. This does not seem convincing, and his argument that Herodotus' presentation of the Ionian revolt is anachronistic and coloured by his own experiences at Halicarnassus in the 450s (pp. 97–9) is rather forced. Resentment against tyranny did not have to wait till after the Persian Wars to develop. For a recent view of the causes of the revolt, see Murray, O. in CAH iv ². 473–80Google Scholar, though contrast C. Roebuck, ibid. 452f. on the question of trade.

³ Berve, H., Die Tyrannis bei den Griechen (Munich, 1967), i. 104, 118Google Scholar, notes the generally lenient treatment given by the Greeks in revolt to their tyrants; Coes of Mytilene was obviously a special case (Hdt. 5.38; see below). He remarks that individual tyrants may have enjoyed a strong personal following in their cities (this is clear at least in the case of Aristagoras at Miletus, a man of great wealth and local influence, cf. the implications of Hdt. 5.35–8, 99, 124–6). This may have been a further incentive to caution, and later in the revolt encouraged the Persians to believe that the Greeks under pressure might respond to conciliatory overtures from their former tyrants (Hdt. 6.9–10, cf. 13).

⁴ See especially Hdt. 5.36–8, 124–6; 6.9–10, 13, 22, 43; also 4.136–9 (the Greek tyrants at the Danube bridge in c. 514), which looks forward to the revolt, though probably anachronistically for the time. Aristagoras' attack on Naxos before the revolt can be read as implying that he felt his position in Miletus was already under threat, and needed strengthening vis-à-vis both Greeks and Persians by some conspicuous success (Hdt. 5.30–4).

⁶ See the writers mentioned by Graf (n. 2 above) p. 100 n. 1; see also Burn, A. R. in Cambridge History of Iran, ii, ed. Gershevitch, I. (1985), pp. 295fGoogle Scholar.

⁶ This is based on his Michigan, Ph.D. dissertation Medism: Greek Collaboration with Achaemenid Persia (1979)Google Scholar, which I have not seen.

⁷ T. Cuyler Young writes, referring to Graf (p. 68), ‘Since the conquest of Ionia under Cyrus it had been royal policy to rule the region through the local tyrants who were in power when the Persians arrived. Cyrus supported these tyrants not because he necessarily approved of tyranny, but because the Persians tried, where possible, to rule through existing forms of government. While in mainland Greece there had been a considerable swing away from such forms of government since Cyrus' time, in Ionia tyrants had continued to be supported in their rule by Persians.’

⁸ See CAH iv².42f. on Cyrus.

⁹ See T. Cuyler Young, pp. 68f., O. Murray, pp. 476f., 486, M. Ostwald, p. 341 (cited in nn. 7, 24, 57).

¹⁰ Hereafter referred to by the author's name only, and H followed by a number refers to an entry in Hofstetter's prosopographical catalogue.

¹¹ Exceptions are Gernet, L., Anthropologie de la Grèce Antique (1968), pp. 182fGoogle Scholar. and (citing him) Briant, P., Etat et Pasteurs au Moyen Orient Ancien (1982), pp. 90f.Google Scholar, both discussing the important case of Syloson and Darius (below); Herman, G., Ritualised Friendship and the Greek City (Cambridge, 1987), discussed in the next paragraphGoogle Scholar.

¹² See Edel, E., ‘Amasis und Nebukadrezar II’, Göttinger Miszellen 29 (1978), 13–20 (I owe this reference to Dr A. Kuhrt)Google Scholar.

¹³ See Wallinga, H. T., ‘The ancient Persian navy and its predecessors’, in Achaemenid History, i, ed. Sancisi-Weerdenburg, H. (Leiden, 1987), 947–77, at 59–66 for Amasis and PolycratesGoogle Scholar.

¹⁴ Mazzarino, S., Fra Oriente e Occidente (Florence, 1947), pp. 191–252Google Scholar.

¹⁵ Page, D. L., Sappho and Alcaeus (1955), pp. 226–34, at 231Google Scholar.

¹⁶ Mazzarino (n. 14), esp. pp. 191f. (Lydians and Greek tyranny), 244f. (contrast between Lydians and Persians).

¹⁷ Thus Berve i.89f.; Roebuck, C., CP 50 (1955), 38 n. 29, citing MazzarinoGoogle Scholar; Hornblower, S., Mausolus (1982), p. 18, citing RoebuckGoogle Scholar.

¹⁸ See Hornblower, S., Mausolus (1982), pp. 17f., 142f., 158Google Scholar.

¹⁹ See Salmon, J. B., Wealthy Corinth (Oxford, 1984), pp. 224–6Google Scholar.

²⁰ cf. Herman, pp. 88f.; a young Athenian noble (possibly an Alcmaeonid) was called after Croesus, cf. Jeffery, L. H., BSA 57 (1962), 143f. no. 57Google Scholar, and Davies, , APF, p. 374Google Scholar.

²¹ Note also the activity of Thales of Miletus as an engineer in the service of Croesus (Hdt. 1.75), reminiscent of Greeks in Persian service later (below).

²² He does not mention (for example at 1.77, 79, 81, 83), the (clearly well known and long-remembered) story of Eurybatus of Ephesus (H 110), who was given money by Croesus to recruit mercenaries, but went over to Cyrus with the money.

²³ Thus Berve i.90, 100.

²⁴ Thus Huxley, G. L., The Early Ionians (1966), p. 121Google Scholar: ‘Already under Cyrus the Persian policy of supporting tyrants in the subject Greek cities had begun’ and he goes on to mention the case of Pytharchus of Cyzicus, discussed below; 144: ‘After the conquests of Harpagus the Persians imposed tyrants on the Greek cities of Asia and ruled through them’. Also Ostwald, M., CAH iv ².341Google Scholar: ‘the appointment of Persian henchmen as tyrants, as had happened in Ionia after the Persian conquest of Lydia’, and O. Murray, ibid. p. 486 writes of ‘fifty years of Persian-backed tyranny’ (sc. in Ionia by 499).

²⁵ See also Young, T. Cuyler in CAH iv ².68 (following Graf)Google Scholar, who presents Cyrus as supporting already existing tyrants, not imposing them.

²⁶ 1.100 and 2.577.

²⁷ On this institution see for example de Ste Croix, G. E. M., The Origins of the Peloponnesian War (1972), pp. 37–40Google Scholar; Briant, P., ‘Dons de terres et de villes: l'Asie Mineure dans le contexte achemenide’, REA 87 (1985), 53–72CrossRefGoogle Scholar; Herman, pp. 106–15.

²⁸ 1.85f.; Herman, p. 111, seems to conflate the different cases, but see Briant's discussion (previous note).

²⁹ Hegesistratus, a son of Peisistratus, was set up by him at an unknown date as tyrant in Sigeum after its capture from Mytilene (Hdt. 5.94; H 133). Berve i.88 conjectured that he must have submitted to the Persian king soon after the fall of the Lydian monarchy and have become a vassal of the Persians, but there is no specific evidence for this, though Hofstetter repeats the suggestion. Hdt. 1.169 mentions that the Ionians in the islands submitted to Cyrus out of fear after Harpagus had reduced the mainlanders, but no further details are known, and in practice this was probably no more than a token surrender without effects on the internal affairs of the Greek cities.

³⁰ Cambyses had sent back unharmed to Cyrene Ladike (H 188), the Cyrenaean wife of Amasis (Hdt. 2.181; see above), and this may have influenced the Cyrenaean submission.

³¹ The Cypriots submitted to Cambyses before his expedition to Egypt and participated in it (Hdt. 3.19), but no further information is available about the role of trie individual rulers in the various cities.

³² See Shipley, G., A History of Samos, 800–188 B.C. (Oxford, 1987), pp. 80, 96fGoogle Scholar.

³³ Cook, J. M., The Persian Empire (1982), p. 71Google Scholar notes Darius', inclination ‘to work through responsible individuals or bodies’ and also notes (p. 75)Google Scholar Darius' clemency towards such individuals.

³⁴ Herman, pp. 28f., 128–30, notes the wide range of services that friends might expect from each other.

³⁵ For a particularly sceptical view, see Griffiths, A., ‘Democedes of Croton: a Greek doctor at Darius' court’, in Achaemenid History, ii: The Greek Sources, ed. Sancisi-Weerdenburg, H. and Kuhrt, A. (Leiden, 1987), pp. 37–51Google Scholar.

³⁶ See CAH iv ². 201–3Google Scholar.

³⁷ See Momigliano, A., The Development of Greek Biography (1971), pp. 29f., 33Google Scholar. See also on him Hornblower, S., Mausolus (1982), pp. 20fGoogle Scholar.

³⁸ The story is dismissed by How, and Wells, , A Commentary on Herodotus, i (1912), p. 298Google Scholar; Herman, , Friendship, p. 41Google Scholar describes it as ‘picturesque and apparently fictitious’. See Shipley, G., A History of Samos, 800–188 B.C. (1987), pp. 103–7 for a narrative of eventsGoogle Scholar.

³⁹ The provision of military support was one of the services expected from a friend (Herman, pp. 97–105).

⁴⁰ On the functions an d usefulness from the Persian point of view of the Greek (and other) tyrants they backed, see Berve i.85 with ii.569; Herman, pp. 39f., 74f., 102. To call them ‘tyrants’ is of course to see them from the perspective of the Greek cities. From Darius' point of view they were dependable friends; indeed, the Persians could be said never to have understood the spirit of the Greek polis (Herman, p. 80).

⁴¹ On the function of this institution in the Persian empire see Wiesehöfer, J., ‘Die “Freunde” und “Wohltäter” des Grosskönigs’, Studia Iranica 9 (1980), 7–21, esp. 16f.Google Scholar; Briant, P., Etat et Pasteurs au Moyen Orient Ancien (1982), pp. 88–94Google Scholar and in Achaemenid History, i, ed. Sancisi-Weerdenburg, H. (1987), pp. 23fGoogle Scholar. and ii, ed. A. Kuhrt and H. Sancisi-Weerdenburg, pp. 5f. See also H, , Sancisi-Weerdenburg, , Historia 37 (1988), 372–4Google Scholar.

⁴² See for example Finley, M. I., The World of Odysseus (2nd ed. 1977), pp. 64–6Google Scholar; Herman, index s.v. ‘reciprocity’.

⁴³ Herman repeatedly emphasizes the international character of relationships between members of the élite, pp. 8, 12, 31–4, 44, 72, 74f., 106, 128, 130, 162.

⁴⁴ Darius Behistan i.20–2 and iv.65–7 respectively, cited from Kent, R. G., Old Persian: Grammar, Texts, Lexicon (New Haven, 2nd ed. 1953), pp. 119, 132Google Scholar.

⁴⁵ Thus Chamoux, F., Cyrène sous la monarchie des Battiades (1953), pp. 147–51Google Scholar.

⁴⁶ Mitchell, B. M., ‘Cyrene and Persia’, JHS 86 (1966), 99–113, at 99–103CrossRefGoogle Scholar with suggested chronological table p. 103, further defended in JHS 94 (1974), 174–7CrossRefGoogle Scholar.

⁴⁷ It is hard to believe that Arcesilas could have done this without at least the tacit acquiescence of whoever was in control of the island then, but Herodotus (4.163) is silent on who that may have been.

⁴⁸ Davies, , APF, p. 452Google Scholar.

⁴⁹ The many chronological and political problems associated with the career of Miltiades cannot be discussed here. See most recently Shimron, B., ‘Miltiades an der Donaubrücke und in der Chersonesos’, Wiener Studien 100 (1987), 23–34Google Scholar, though his reconstruction is not followed here. He dates the sending of Miltiades by the Peisistratids to the Chersonese to c. 508, i.e. after the Peisistratids were expelled from Athens and moved to Sigeum (pp. 29–32), and is even inclined to doubt whether Miltiades was tyrant in the Chersonese at the time of the Scythian expedition (pp. 32–4). If this is correct, the distortions in Herodotus' account are so deep as to be beyond viable analysis.

⁵⁰ Graf, , op. cit. (n. 2), p. 84Google Scholar.

⁵¹ cf. Seibert, J., Die politischen Flüchtlinge und Verbannten in der griechischen Geschichte, 2 vols. (Darmstadt, 1979), i.392–4Google Scholar; Herman, pp. 43f.

⁵² Athenian tradition may have retrospectively magnified the role of Hippias in promoting the first Persian invasion, and obscured the significance of the Athenian disavowal of their envoys' offer of earth and water to Artaphemes in 507. In Persian eyes, the Athenians then compounded their offence by giving support to the Ionians in revolt. See Kuhrt, A., ‘Earth and water’, in Achaemenid History, iii: Method and Theory, ed. Kuhrt, A. and Sancisi-Weerdenburg, H. (Leiden, 1988), pp. 87–99, esp. 91–3, 98fGoogle Scholar. on Athens and the Persians.

⁶³ See Hofstetter, pp. 192–4 for a complete list.

⁵⁴ On both see Briant, P., REA 87 (1985), 62–4CrossRefGoogle Scholar.

⁵⁵ See already How, and Wells, , Commentary on Herodotus ii.80Google Scholar.

⁵⁶ See for example How and Wells (previous note); Berve i.105f. and ii.581; Burn, A. R., Persia and the Greeks (2nd ed., 1984), p. 222Google Scholar; Hornblower, S., Mausolus (1982), p. 22 and n. 128Google Scholar.

⁵⁷ Especially Young, T. Cuyler, CAH iv ².68f.Google Scholar: ‘[after the Ionian revolt] there was a marked change in the imperial attitude towards the tyrants […] Henceforward we witness the rather ironical fact that the despotic Persians became the staunch supporters of democratic government for the Asia Greeks’. Also O. Murray, ibid. p. 490, who accepts without comment or qualification Herodotus' account of Mardonius' action; Hornblower, S., Mausolus (1982), p. 18Google Scholar: ‘a partial return to Lydian conditions’.

⁶⁸ Cook, J. M., The Persian Empire (1982), p. 96Google Scholar.

⁵⁹ On the hereditary nature of personal friendships, cf. Herman, pp. 16f., 69f., 152.