Appendix

The Spartan king-lists

The purposes of this Appendix are twofold: to discuss the source and significance of the lists of Agiads and Eurypontids preserved in Herodotus (Table 1) and to illustrate briefly the cardinal role that the upper reaches of these and comparable lists appear to have played in the elaboration of a chronology for early Greek history.

TABLE 1

Agiads (7.204)

Eurypontids (8.131.2)

Herakies Hylios Kieodaios Ari stomach os Aristodamos

Did not reign

We must first decide what these lists represent. Are they, as the ancients believed (although, as we shall see, they differed in detail), king-lists? Or are they, as Henige (1974, 207-13) has now argued in a fundamental study of oral tradition, merely the pedigrees of Leonidas I (reigned c.490-480) and Latychidas II (c.491-469)? The way in which Herodotus introduces the lists suggests the latter, and this view is apparently supported both by the omission of otherwise recognized kings (from the Agiads Kleomenes I; from the Eurypontids Agasikles, Ariston and Damaratos) and by his cross-reference at 9.64.2, where he points out that regent Pausanias had the same ascendants (progonoi) as Leonidas. On the other hand, Herodotus states explicitly

(8.131.3) that ‘all except the two named immediately after Latychidas (viz his father and grandfather) became kings of Sparta’, and this may indeed imply that he believed all Leonidas’ ascendants had done so too. At any rate, this was how he was understood by all later writers.

Neither interpretation is entirely cogent, but on balance I think Herodotus did indeed mean the lists for king-lists. The major obstacle in the way of this interpretation is the omission of recognized kings from both lists, but this has been adroitly circumvented by Prakken (1940) with the suggestion that Herodotus was adapting lists compiled, perhaps by Hekataios (below), in the joint reign of Kleomenes I and Damaratos, neither of whom was succeeded by a son. Whether or not Hekataios (or whoever) was the first to produce and publish written king-lists we cannot of course say.

A minor objection, that not even the Spartans believed Leonidas’ and Latychidas’ ascendants before Aristodamos to have been kings of Sparta, has been proven groundless by Huxley (1975b), who rightly distinguishes between kings ‘of’ Sparta and kings ‘in’ Sparta; we may add that in 371 the Dioskouroi, ‘the model and divine guarantee of the Spartan dyarchy’ (Carlier 1977, 76 n. 42), could be referred to as ‘fellow citizens’ of the Spartans (Xen. Hell. 6.3.6). Moreover Herodotus’ confusion over the name of Latychidas II’s grandfather—Agesilaos in Table 1, but Agis at 6.65.1—seems most easily explicable if, as Herodotus himself states, Agesilaos/Agis did not in fact reign. Thus, since Latychidas II and Damaratos were coevals (they fell out over the girl they both wanted to marry: 6.65.2), Latychidas’ grandfather could have been a younger brother of Damaratos’ grandfather, Agasikles, the co-king of Leon. This would make Latychidas and Damaratos second cousins, closely enough related for Kleomenes I to use the former as an acceptable replacement for the latter.

To these negative arguments in favour of the interpretation of the lists as king-lists we may add the evidence of two papyri from Oxyrhynchos published after Prakken’s important paper. One of these (2390, with Harvey 1967) proves that Latychidas I did indeed reign—probably c.600, since Alkman sang of him. The other (2623.1, a choral lyric fragment attributable to Simonides or Bacchylides) mentions a Zeuxidamos (see Table 2, below) and perhaps a Hippokratidas (cf. Table 1), apparently in a royal context. We know that Stesichoros sang before a Spartan prince in Sparta c. 550 and that he lent his voice to Heraklid propaganda (West 1969, 148). Perhaps the poet of this fragment was doing the same for a Eurypontid prince in the early fifth century.

Granted then that we have access to Spartan king-lists in Herodotus, two further and related questions arise. How far may we accept them as true records of the dyarchy? Second, what role or roles might the lists have played in forming the Spartans’ view of their past and in determining the way that past was presented to or used against outsiders?

We may start with the connection of the eponyms Agis (I) and Euryp(h)on through the twin sons of Aristodamos to Herakles. The Heraklid connection is first explicitly attested in Tyrtaios (fr. 2.12-15W), but it should go back to the dissemination of the Homeric or similar poems in Sparta and the foundation of the Menelaion sanctuary in the late eighth century, if not to the incorporation of ‘Achaean’ Amyklai c. 750. Indeed, it is likely enough to have been forged at the same time as the dyarchy itself, which perhaps began with Archelaos and Charillos (Chapter 8). From Tyrtaios the assertion of the connection can be traced in an almost unbroken chain of poetical references through the Lakonians Kinaithon and Alkman in the seventh century, Stesichoros and the Delphic Oracle in the sixth to Pindar, the contemporary of Leonidas I and Latychidas II.

As far as the outside world was concerned, the function of the Heraklid connection was to legitimate Spartan supremacy in Lakonia and indirectly, the Peloponnese. Within Sparta itself, however, it had other functions. All ‘Heraklids’ were Spartans, but not all Spartans were ‘Heraklids’ (Hdt.

8.114.2). Moreover, within the Spartan aristocracy there were other ‘Heraklid’ families besides the Agiads and Eurypontids, and families like the Aigeidai (Hdt. 4.149.1) who were not Heraklid at all. The king-lists therefore were a very special kind of genealogical charter (Malinowski’s expression) or ‘mnemonic of social relationships’ (Goody and Watt 1963, 309), serving to affirm the superior blue-bloodedness of the Agiads and Eurypontids against the claims of other aristocratic families and to distinguish the aristocracy from the commons.

So much for the roles of the lists. Now for their accuracy. We must at once admit the depth of our ignorance. We do not know when, if ever, after the introduction of writing the lists were committed to script at Sparta; nor, if and when the transmission was purely oral, how that transmission was effected; nor how much circumstantial detail was passed on in association with any particular name. However, we do know that by the time of Pausanias (3.2.17; 3.1-8; 7.1-10) Herodotus’ lists were regarded as king-lists and were taken to imply both that each Agiad king had had his one Eurypontid counterpart and that succession had been hereditary from father to son over fifteen generations within each house. Why then do the Eurypontid lists of the two authors differ (Table 2)?

TABLE 2 Eurypontids

Did not reign

The introduction of Soos is easy to explain: the Eurypontid list in Herodotus was one shorter than the Agiad, and he was probably inserted in the fourth century (Kiechle 1959, 90-101). The discrepancies after Theopompos are more difficult. In effect Pausanias has Zeuxidamos and Anaxidamos for Herodotus’ Anaxandridas (I) and Anaxilaos, and he has omitted Herodotus’ Latychidas (I) and Hippokratidas. Of these Latychidas I was certainly a king and Hippokratidas may have been referred to as such in the choral lyric fragment of the early fifth century (above). On the other hand, Pausanias’ Zeuxidamos may be the man who appears in the same fragment and so may also have some claim to have ruled; alternatively, he could be the son of Latychidas II (Hdt. 6.71). Either way, it is possible that Pausanias had independent access to genuine Eurypontid tradition, perhaps ultimately through Charon of Lampsakos (262T1 Jacoby). If therefore we accept that both Herodotus and Pausanias may preserve the truth about the ruling members of the Eurypontid house, the most economical hypothesis to explain the discrepancies between their lists is that both lists are selective king-lists, the one confined to Latychidas II’s direct ascendants, the other recording those of Damaratos.

This hypothesis has many merits, of which two may be singled out here. First, as we shall see in the second part of this Appendix, it helps to solve a puzzle in early Greek chronography. Second, it does away with the glaring contradiction between the allegedly unbroken father/son succession down to

Kleomenes I and Damaratos and the situation thereafter. For between c.491 and 219 lineal succession broke down in no fewer than twelve out of the twenty-six instances; of the remaining fourteen successions the largest number, five, were consobrinal—brother succeeding brother. We must make some allowance for changed political conditions after c.491, which saw the first attested deposition of a Spartan king (Ste. Croix 1972, 350-3). But as Henige (1974, 210) rightly says, it ‘beggars the imagination’ to postulate two series of unbroken father/son succession in a single state over the same sixteen generations from Eurysthenes and Prokles. In other words, even if Herodotus’ lists are adaptations of king-lists drawn up in the joint reign of Kleomenes I and Damaratos, we should make allowance for an unknowable number of collateral successions.

By a still more opaque process than those of their creation and transmission the Spartan king-lists transcended their local political significance to occupy a unique niche in the chronography of early Greek history. The first exponent of ‘scientific’ chronography, a byproduct of the shift in emphasis of Ionian historia from nature to man (Chapter 5), was probably Hekataios. It could then have been he who drew up the king-lists which Herodotus adapted (Jacoby 1949, 306 n. 25, 323 n. 28, 357 n. 26). He too it may have been who interpreted the fifteen kings in each line from Agis and Eurypon to Kleomenes I and Damaratos as fifteen generations and, making allowance for the Heraklid connection, gave to each generation the notional value of forty years (Meyer 1892, 153-88, esp. 169ff., 179-82). However, since ‘it is impossible to accept a generation average as high as forty years over a period of fifteen generations, no matter what contingencies are postulated’ (Henige 1974, 208), the hypothesis about collateral, and especially perhaps consobrinal, succession may again be invoked to account for it, if indeed it is felt that the forty-year generation has any basis in fact.

However that may be, exact lengths were subsequently attached to the reigns at least down to those of Alkamenes and Theopompos. Various candidates for the role of first calculator have been proposed, of whom the third-century Lakonian Sosibios (595 Jacoby) has possibly the strongest claim. Eratosthenes of Cyrene, also in the third century, brought the lists into an acceptable relationship with the First Olympiad, which was for him the dividing line between ‘mythical’ and ‘historical’ Greece (Fraser 1970, esp. 190, 196f.). From Eratosthenes descends the ‘vulgate’ chronology of early Greek history through Apollodoros (c.100) and Diodorus to Eusebius (AD 263-339).

It goes without saying that the absolute dates arrived at by these erudite men have no truly scientific foundation, and that differences between their dating and ours are to be expected. On the other hand, to tamper with their relative chronology is hazardous. In general, their absolute dates are too high, a natural consequence of the Heraklid distortion. If we substitute the more plausible allowance of thirty years per generation for the ‘Hekataian’ forty, we achieve a satisfying congruence between potsherds and pedigrees, at least for the Agiads: Agis I could have been on the throne around the last third of the tenth century (Forrest 1968, 21). At the same time, however, we cannot pretend that in the present state of our knowledge this is much more than a happy coincidence.

Note

I was greatly helped in the preparation of the original version of this Appendix by my late friend Richard Ball, although the responsibility for any remaining errors is of course entirely mine.