Regarding the chronology and events of the Elean War, fought between Sparta and Elis ca 400 B.C., our sources are in notorious disagreement.[1]Cf. W. Judeich, Kleinasiatische Studien (Marburg 1892) 182f; G.E. Underhill, “The Chronology of the Elean War,” CR 7 (1893) 156-58; E. Meyer, Theopomps Hellenika (Halle 1909) 114-16; J. Hatzfeld, “Notes sur la chronologie des Helleniques,” REA 35 (1933) 395-409; and D.H. Kelly, Sources and Interpretations of Spartan History in the Reigns of Agesilaus II, Archidamus III and Agis III (diss. Cambridge 1975 [hereafter ‘Kelly’]) 21-41. The inveterate notoriety and difficulty of the issue is summed up in Underhill’s opening sentence: “A complete solution of all the difficulties involved in his problem, which has occupied the attention of commentators and historians for more than a century, is almost out of the question: we have to content ourselves with weighing one set of probabilities against the other.” Xenophon’s account (and that of Pausanias, which ultimately derives from it) appears to differ irreconcilably from that of Diodorus.[2]Xen. Hell. 3.2.21-3.3.1, Paus. 3.8.3-5; Diod. 14.17.4-12, 34.1. Though condensed, Pausanias’ account of the war is virtually identical with Xenophon’s both in material and in order of presentation, the only major difference being his statement that the war lasted three years, whereas many modern scholars interpret Xenophon’s phrase περιιόντι τῷ ἐνιαυτῷ to imply that the war lasted only two years (cf. n.3 infra). Pausanias’ other differences from Xenophon involve minor details, and are probably best explained as errors due to Pausanias’ faulty memory or his intermediate source, although Kelly (21) sees in them the traces of an additional tradition. A resolution of this crux is desirable for its own sake, but even more important are the substantive issues involved: reconstruction of the actual course of the war should help us gain a better understanding of Spartan foreign policy during the aftermath of the Peloponnesian War, allow us to learn the circumstances and date of the accession of King Agesilaus, and, perhaps most significantly, provide further insight into the reliability and methods of our two principal sources for the events of the period.

There is little dispute about the origins of the war. In 420 the Eleans barred Spartans from the Olympic Games because of an alleged truce violation and flogged Lichas, a prominent Spartiate who attempted to participate; some time later the Eleans refused King Agis permission to sacrifice at Olympia for victory in the war against Athens (Xen. Hell. 3.2.21f, Diod. 14.17.4, Paus. 3.8.3). Underlying these specific charges was Sparta’s general hostility towards Elis for the latter’s military adherence to the anti-Spartan coalition of Argos, Mantinea, and Athens formed during the course of the Peloponnesian War; and this, among the causes he offers for the outbreak of war, Xenophon chose to list first (Hell. 3.2.21).

Sparta opened the conflict by demanding that Elis grant independence to the outlying towns of the polis—a gesture aimed at fragmenting Elean strength, identical to Sparta’s later strategy for destroying Mantinean power (Xen. 5.2.1-7). When the Eleans refused to break up their state, the Spartans declared war and invaded Elis with an army led by one of their kings.

It is at this point that our accounts of the war diverge completely. According to Xenophon (3.2.23-31) King Agis led the first invasion, entering Elis from Achaea in the northeast, along the river Larissus; but an earthquake suddenly convinced him to break off his invasion, return to Sparta, and disband his army. We are told that Agis’ retreat emboldened the Eleans to send embassies to all Greek states hostile to Sparta, presumably in the hope of forming a general alliance. But in the following year[3]Xenophon’s actual phrase is “as the year was coming round” (περιιόντι τῷ ἐνιαυτῷ). Whether this should be interpreted to mean that Agis’ second campaign took place in the same season as the first had been an important point of contention: Judeich (supra n.1) argues that it should, and G.L. Cawkwell, ed ., Xenophon: A History of My Times (Harmondsworth/New York 1979) 155, appears to concur, letting stand Warner’s translation of the phrase “in the same year.” The Loeb version by G.L. Brown (London/Cambridge [Mass.] 1961) has “in the course of the year.” The most natural meaning of the Greek would seem to be “as the year was changing” or “at the start of the new year,” as argued by Underhill (supra n.1) 157 and Kelly 25 n.40; and this is clearly the sense in which Pausanias or his intermediate source took it (thus his figure of three years for the war; cf. supra n.2). In fact the main reason for confining Xenophon’s account of the war to two years is to reconcile it with Diodorus, who chose to distribute the war into two political years; but in view of Diodorus’ almost random chronological reckoning, this is a dubious advantage.

The pattern of events themselves also strongly supports giving Agis two summers of campaigning. It is possible to squeeze a Spartan mobilization, a short campaign in Elis, a Spartan demobilization, the dispatch of Elean embassies to various anti-Spartan states, a second Spartan mobilization (this time including nearly all of Sparta’s allies, notably the Athenians), and a second Elean campaign of some length into a single campaigning season, and still leave “the rest of the summer” for the ravages of Lysippus (Xen. 3.2.23-30); but such a reconstruction is forced and implausible. It is far more natural to attribute the two campaigns to separate summers, with diplomatic exchanges occurring during the intervening winter. the ephors again called out the ban; Agis led an army that included troops from all the allies except Corinth and Boeotia in a second invasion of Elis, now by way of Aulon in the south. Agis was more successful this time: after winning over a number of the outlying towns, he offered sacrifice at Olympia (thus fulfilling one of the Spartan objectives) and began to ravage Elean territory and the outskirts of the city. We are told that the Spartans believed that Agis could have captured the unwalled city[4]G.L. Cawkwell, “Agesilaos and Sparta,” CG N.S. 26 (1976) 75 n.48, argues that Xenophon actually means that only the acropolis was unwalled, rather than the city as a whole; this is primarily an unnatural attempt to reconcile Xenophon with Diodorus and Pausanias (see text infra), and its motive disappears if a simpler hypothesis will serve. Moreover, the logic of the argument seems flawed: Xenophon is using the absence of walls to prove that Elis was vulnerable to capture, and if the city walls had been as strong as Diodorus implies, an unwalled acropolis would not have been cited as proof of vulnerability. but chose not to; if this is true, he may have expected a pro-Spartan coup within Elis, for one did in fact occur soon after and narrowly missed succeeding. With the failure of the coup, Agis departed leaving behind a garrison under Lysippus at Epitalium near the Alpheus. These men ravaged Elis for the rest of the summer and the following winter, and at last forced the Eleans to sue for peace the next summer. At first the Eleans balked at accepting the Spartans’ demand that Epeum be freed, but were forced to comply with even this condition and peace was finally arranged. Such is Xenophon’s account, with which Pausanias (3.3.3-5) is in complete agreement.

In Diodorus’ version (14.17.4-12) it is King Pausanias who led the first invasion with an army of 4,000 men, as well as many soldiers from nearly all the allies except, again, Corinth and Boeotia. His route was west from Arcadia across the mountainous Acrorea region; after capturing several small towns he reached Elis and began to invest the city. But the siege was carelessly conducted and the Eleans, backed by 1,000 recently hired elite Aetolian troops, managed a sortie to defeat the Spartan army in a decisive skirmish. This set-back convinced Pausanias that the city would be difficult to capture by siege; after further ravaging nearby territory and establishing fortified outposts in Elis, he took the bulk of his army to winter at Dyme. Diodorus’ narrative of the war breaks off at this point and resumes at 14.32.1 with a one-sentence account of Elis’ surrender.

The two accounts have almost nothing in common. Each is a highly detailed description of the war, containing the route of the Spartan invasion, the names of the Elean towns captured, and the specific course of the campaign; but in each case the details are utterly different. It is implausible to argue, as many have done,[5]Cf., among others, Judeich 182 and Meyer (both supra n.1); Cawkwell (supra n.4) 76 n.51. that Diodorus’ account is a ‘distorted’ version of Xenophon’s account: our two basic sources are in complete disagreement.

Ordinarily this might not greatly concern us. Xenophon is always well-informed about Spartan affairs,[6]It is not surprising that Xenophon’s Hellenica demonstrates a wealth of inside information about Spartan society and politics, for he had many Spartan friends, notably King Agesilaus, and spent the latter half of his life in the Spartan-controlled Peloponnese. Cf. Cawkwell (supra n.3) 12-15. while Diodorus’ errors are notorious. However, this passage of Diodorus concerns years covered by the Hellenica Oxyrhynchia; the precise details that Diodorus includes—the names of captured towns, troop strengths, the positions of opposing forces, the number of casualties, as well as the unrhetorical, even-handed tone—leave little doubt that the Oxyrhynchus historian is his ultimate source.[7]So Kelly 29; cf. I.A.F. Bruce, An Historical Commentary on the Hellenica Oxyrhynchia (Cambridge 1967) 20-22.

We are thus left with the problem of two good sources that provide detailed but totally different accounts of a single war. We can choose to accept one account as true and reject the other as fantasy, but this places us in the uncomfortable position of arguing that a reputable source, for no apparent reason, simply invented the vivid details of an imaginary campaign. Or we may begin with the working hypothesis that both accounts of the war are essentially correct but incomplete; on this simple assumption, the versions of Xenophon and Diodorus can be shown to mesh remarkably well.

If both Agis and Pausanias led invasions of Elis, the first issue to consider is the order in which these occurred. At the outset, it might seem more natural to place Pausanias’ invasion first:[8]Assumed by Kelly (30), without elaboration. Diodorus, after all, never hints that Pausanias’ campaign was responsible for ending the war, while Xenophon strongly implies that it was Agis who led Sparta to victory, even though he places the end of the war more than a year after Agis’ second campaign and death. Moreover, the long gap between Diodorus 14.17.12 (which describes Pausanias’ campaign) and 14.34.1 (which describes Elis’ surrender) might seem admirably suited for the insertion of an extra campaign or two. Finally, Epitalium, which Agis garrisoned, was captured by Pausanias (Xen. Hell. 3.2.29, Diod. 14.17.80).

While these arguments have merit, the weight of the evidence strongly supports the alternate view that Agis’ campaign preceded Pausanias’. The main arguments are these:

(a) The walls of Elis. Xenophon (3.2.27) informs us that during Agis’ invasion the Spartans believed that the city would be easy to capture because it had no walls;[9]Even if we leave aside the question whether Xenophon is referring to the city or to the acropolis (cf. supra n.4), the statement remains that the Spartans believed that the city would be easy to capture. This would seem absurd if a full Spartan army under Pausanias had battered its head against the city a year or two before and finally given up because Elis was too strong to besiege. according to Diodorus (14.17.10f) Pausanias had found Elis too strong to capture by siege, implying the existence of walls. Presumably these were erected by the Eleans during the year or so that separated the two campaigns. Although the Spartan forces remaining in Elis would have done their best to hinder this effort, they would have been no match for the full Elean army; and Greeks of this period could, under the pressure of necessity, construct fortifications with remarkable speed.[10]Cf., for example, the rapid Athenian fortification of Pylos (Thuc. 4.4f).

(b) The Aetolian mercenaries. According to Diodorus (14.17.9f), Pausanias’ attack on the city was defeated by 1,000 elite Aetolian troops whom the Eleans had recently hired. It is doubtful that the Eleans would have taken such an expensive step[11]The pay for 1,000 mercenaries at 3 obols per day (perhaps more for elite troops) would have come to at least 2.5T of silver per month, a considerable expense for a largely agrarian Peloponnesian state. merely as a precaution against a potential Spartan threat, before their lands had been ravaged for some time. Xenophon’s account makes no mention of this significant military force, and we have no reason to suppose that the Eleans would have sent them away after the war began.

(c) Agis’ abortive first invasion. That Agis discontinued his first campaign after an earthquake suggests that the invasion was intended more as a warning or show of force than as a serious attempt to bring Elis to its knees. Such action would make little sense coming on the heels of a major invasion by Pausanias during the previous year, which had seen the Spartans establish fortified posts for a sort of Decelean War strategy after having been beaten in an initial encounter. On the other hand, a show of force is plausible for a first campaign, especially if the Spartans were still too exhausted by the long Peloponnesian War to relish a full scale struggle with a first-class hoplite power such as Elis, and hoped that a threat would be sufficient to achieve their ends.[12]We should note that Archidamus had adopted a similar strategy at the beginning of the Peloponnesian War (Thuc. 2.18). War in earnest seems more likely after Elis’ attempt to organize an anti-Spartan coalition.

(d) The participation of the Spartan allies. Xenophon’s specific emphasis on the participation of all Sparta’s allies (except Corinth and Boeotia) in Agis’ second campaign (3.2.35) implies their absence in the first campaign. If, however, Pausanias had brought in allies and had nonetheless been defeated the previous year (Diod. 14.17), Agis would have been unlikely to leave them behind. But if Agis’ first campaign opened the war and constituted more a show of force than an actual attack, the inclusion of large allied contingents would not have been vital.

(e) Elean diplomacy. According to Xenophon (3.2.24) the Eleans attempted to organize an anti-Spartan coalition after Agis’ abortive first invasion. This would be very strange timing if Pausanias had already invaded the previous year, ravaged the country thoroughly, been defeated outside the city walls, and then established fortified bases for a protracted war: why would the Eleans have waited another year and a half to send out their ambassadors? As we have pointed out, it is much more plausible that Agis’ ‘warning campaign’ began the war; when Elis called Sparta’s bluff by raising the diplomatic stakes, Sparta was forced to go to war in earnest.

(f) The choice of Spartan commander. The Eleans’ insult of Agis had provided one of the main pretexts for the war (Xen. 3.2.22, Diod. 14.17.4), and he avenged this insult during his first full invasion (Xen. 3.2.26). Agis, rather than Pausanias, would have been the natural choice to command the first invasion of Elis, and it was probably only his death soon after the second invasion that caused Pausanias to lead the third.

(g) The attempted pro-Spartan coup. If the Eleans had, more or less, the better of the first year of the war (against Pausanias) and had ‘frightened off the Spartans’ during the second year (as any Elean would have viewed Agis’ abortive first invasion), it seems implausible that they would have nearly fallen victim to a poorly-organized pro-Spartan coup at the beginning of the third year (Xen. 3.2.27-29), especially if the 1,000 Aetolian mercenaries were still at their disposal.

Even if we accept these arguments and assume that Pausanias’ campaign followed that of Agis, we still need to place Pausanias’ invasion within the context of Xenophon’s narrative. A likely possibility is not difficult to find. According to Xenophon, Elis sought peace in the summer following Agis’ second invasion, but Sparta found the Elean terms unacceptable and “compelled” (ἠνάγκασεν, 3.2.31). Elis to make peace on Sparta’s own terms instead. We might ordinarily assume that such compulsion referred merely to further threats, but we are required to place a campaign by Pausanias around this time; and the notion that Xenophon chose to mask with euphemism a full invasion of Elis led by a Spartan king under the vague term “compelled” is not at all at odds with what we know of Xenophon’s often obscure style.[13]Cf. Cawkwell (supra n.3) 15ff, esp. 33f. Xenophon is guilty either of a severe distortion of the facts or of the graver offense of omitting all reference to the third Spartan invasion of Elis.

The likely reason for Xenophon’s apparent reticence is not difficult to guess: a full account of this year of the war might have shown Xenophon’s beloved Sparta to be both oppressive and incompetent. According to Xenophon, Sparta’s ‘compulsion’ of Elis followed rejection of an Elean offer of surrender that met virtually all Sparta’s demands. But Sparta exacted the last ounce of flesh by requiring that Elis surrender Epeum, a strip of territory that she had bought honestly for a price of thirty talents (Xen. 3.2.30f). When Pausanias invaded Elis to enforce Sparta’s demands, his army of 4,000 Lacedaemonians and many allies was surprised and defeated outside the city walls by the Eleans and their 1,000 newly-hired Aetolian mercenaries (Diod. 14.17.6-10). Xenophon may have feared that a complete account might leave the reader to conclude that Agis’ two invasions and Sparta’s harsh negotiating tactics had achieved little beyond giving the Eleans time to build up their walls and hire mercenaries; the eventual (and inevitable) Elean surrender would not have offset this impression. Xenophon’s other omissions and distortions in service of his pro-Spartan sentiments are similar in nature.[14]Cf. Cawkwell (supra n.3) 17, 20, 35-38, 40-43. There is no doubt that Xenophon’s version of the Elean War is much kinder to Sparta than Diodorus’ or any melding of the two.

Diodorus’ omission of the first two years of the war is even easier to understand, for his history generally retains only a portion of the historical material presumably contained in his sources. He includes the downfall of the Thirty at Athens (14.33.5f) but not the final reannexation of Eleusis (Xen. 2.4.43). He fails to mention the trial and exile of King Pausanias after the death of Lysander at Haliartus (Diod. 14.81.3; cf. Xen. 3.5.25). He completely omits any discussion of the disputed succession of Agesilaus or the conspiracy of Cinadon, though it is possible that his source (not being privy to Xenophon’s Spartan information) was equally ignorant of these last events. It therefore seems not at all surprising that Diodorus would have left out the first two Spartan campaigns of the Elean War, which were marked by no battles and failed to accomplish much of note.

In fact the only improbability in our hypothesis is that Xenophon and Diodorus merely chanced to produce partial accounts of the war that contain virtually no overlap, lending a peculiar patchwork quality to our reconstruction. Normally this might be a significant argument against the case presented, but an exactly parallel situation exists in the way Xenophon and Diodorus describe Sparta’s campaigns in Asia during this same period. Xenophon fails to mention either Tissaphernes’ attack on Cyme (which precipitated Spartan intervention in Asia) or Thibron’s first campaign, both of which are described in detail by Diodorus (14.35.6-36.3).[15]Xen. 3.1.5 does contain a vague sentence that could be taken to summarize these events; if so, this would parallel the manner in which his vague reference to Spartan ‘compulsion’ may summarize Pausanias’ invasion of Elis (supra 34f). Diodorus then proceeds to omit Thibron’s second campaign, which Xenophon discusses at 3.1.6f. Both include Dercyllidas’ campaigns, although Diodorus omits his eight-month siege of Atarneus (Xen. 3.1.8-3.2.20; Diod. 14.38.2f, 6f, 39.5f).

The Asian campaigns of Agesilaus are even more poorly recorded: here Xenophon’s omissions are underlined by the survival of large sections of the Hellenica Oxyrhynchia for these years. While Diodorus passes over Agesilaus’ initial negotiations with Tissaphernes and the three-month truce he arranged (Xen. 3.4.5f), Xenophon matches this by leaving out all mention of Agesilaus’ post-Sardian campaign, as well as his campaign in Mysia (Hell. Oxy. 12f, 21f; Diod. 14.80). Diodorus, for his part, proceeds to omit Agesilaus’ campaigns in Phrygia and Paphlagonia and his negotiations with Pharnabazus (Xen. 4.1). On Agesilaus’ return march to Sparta, Diodorus describes his battle against the Thracians but not his skirmishes against the Thessalians, while Xenophon does exactly the reverse (Xen. 4.3.1-9, Diod. 14.83.3; Plut. Ages. 16 lists both).

As these examples indicate, the selective method of both Xenophon and Diodorus entails the omission of important material, and often, as it happens, results in accounts that complement each other exactly. This is very likely the cause of our disparate versions of the Elean War.

According to our reconstruction, therefore, the Elean War probably lasted three campaigning seasons, with Agis’ invasions coming in the first two summers and Pausanias’ in the third. Elis’ surrender probably came during the third winter: although she had held her own military, the Spartans seemed determined to ravage her territory indefinitely; exhaustion and the costs of the war steadily mounted, and the complete hopelessness of her long-term strategic position became apparent. By the time of the Elean surrender, Agis had probably been dead for over a year and had been succeeded by Agesilaus. Xenophon tells us that Agis died shortly after dedicating a share of his booty at Delphi; although he mentions this after describing the end of the war, he is probably referring to the booty that Agis himself had captured during the second invasion, and is therefore violating strict chronological sequence in order to maintain the continuity of his topical narrative.[16]Xenophon often digresses in just this way to describe the eventual outcome of an event before returning to chronological sequence: cf. the death of the exiled Pausanias (3.5.25). On our hypothesis Xenophon here digresses to sum up the end of the Elean War before returning to Agis’ activities following the end of the second Elean campaign. Otherwise, we would expect Agis to have dedicated his booty from the invasion immediately, rather than wait a year or ore until the end of the war. As I argue elsewhere, similar digressions on the part of Thucydides have been responsible for much confusion in the chronology of the Pentecontaetia: cf. R.K. Unz, “The Chronology of the Pentekontaetia,” CQ (forthcoming). As suggested above, Agis’ death was presumably the reason that King Pausanias was chosen to lead the third invasion.

Assuming that our reconstruction of the events of the Elean War is correct, we must still attempt to place the war within the overall chronology of the period. Diodorus lists the war under the years 402/1, but this is undoubtedly due to his notoriously muddled attempt to fit the topical histories of his sources into the Procrustean bed of his own chronological scheme, and counts for little.[17]For example, Diod. 14.32 places the overthrow of the Thirty at Athens (403) just after the return of the Ten Thousand (399), and both under the political year 401/0. His numerous other chronological blunders are almost as severe. Fortunately there are more reliable indications of the date of the war. Xenophon (3.2.1) tells us that the Elean War took place at the same time as the campaigns of Dercyllidas in Asia. Although the value of this synchronism has been sharply attacked,[18]For example, by Cawkwell (supra n.3) 154 and Kelly 24-26, whose efforts at chronological reconstruction are hindered by their refusal to consider the possibility that Agis may have died before the end of the Elean War, and that Xenophon’s account of the end of the war may represent a chronological digression (supra n.16). and obviously ought not to be taken to mean that the two wars began and ended simultaneously, we should nonetheless understand Xenophon to mean that there was a sizable amount of overlap between the two; and we should assume, without contrary evidence, that he is correct. Xenophon was himself serving as a commander in Dercyllidas’ Spartan-led expeditionary force and cannot have remained wholly ignorant of Sparta’s military doings at home, either during this time or later, when he composed his history while residing in the Peloponnese. Since Dercyllidas arrived in Asia during summer 399 and campaigned there until ca summer 397,[19]The year of Dercyllidas’ arrival in Asia is solidly based and is accepted by nearly all the scholarly literature. For discussion cf. K.J. Beloch, Griechische Geschichte III.2 (Berlin 1922) 211f. the synchronism narrows the possibilities considerably: the three-year Elean War probably began in summer 401, 400, 399, or 398.

This range of possibility can be limited further by applying our information on the regnal dates of King Agesilaus. If Agis died soon after this campaign in the second summer of the war, this will determine the official date of Agesilaus’ accession. According to Plutarch (Ages. 40.2) Agesilaus reigned for 41 years. Now, from Plutarch and from Xenophon’s Agesilaus we learn that Agesilaus went out to Egypt as a mercenary commander in 361 and died there either during winter 359/8 or, more likely, 360/359, depending on how much time we assign to the various Egyptian campaigns and how seriously we regard the various analyses of the Babylonian records.[20]Cf. Kelly 33-35. Plutarch further informs us that Agesilaus reigned for more than thirty years before the battle of Leuctra (summer 371). Since Plutarch’s information is probably derived from an annalistic source, we may assume that all his figures refer to political years. Still, we have no means of determining whether Agesilaus’ official accession occurred before or after the turn of the Spartan political year, nor do we know whether Leuctra itself occurred in 372/1 or 371/0. Given this uncertainty, the best we can conclude (if we credit Plutarch’s figures) is that Agis may have died as late as early summer 400. This narrows the possible starting date of the war to summer 401, with summer 400 remaining a possibility if we regard Plutarch’s phrase “more than thirty years” as rhetorical exaggeration.[21]Note that Plutarch’s full claim is that Agesilaus “had been regarded as the king and leader of almost the whole of Hellas” down to Leuctra, which is clearly false. Kelly (32 n.62) speculates that Plutarch’s error may follow from his own uncertainty as to the exact date of Leuctra.

In fact the choice of 401-399/8 for the Elean War accords well with the known date of the Athenian annexation of Eleusis, which fell in the Athenian political year 401/0 (Arist. Ath. Pol. 40.4). Sparta’s strength was committed to the Peloponnese in summer 400, during Agis’ second Elean campaign, and Athens might have viewed this as an opportune time to move against Eleusis. Sparta’s lack of response to the overthrow of her Athenian settlement of 403 is puzzling unless we posit some such Spartan military preoccupation. Further, the absence from Athens of a sizable number of the conservative members of her hoplite and cavalry classes (serving with Agis’ army in the Peloponnese) might have tipped the political balance in the ekklesia, just as Cimon’s expedition to the Peloponnese seems to have made the reforms of Ephialtes possible a half-century earlier.[22]Cf. J.R. Cole, “Cimon’s Dismissal, Ephialtes’ Revolution and the Peloponnesian War,” GRBS 15 (1974) 369-85.

By contrast, the alternative of 400-398/7 for the war would accord poorly with the events of Eleusis: Sparta would not have been fully committed to conflict with Elis by summer 400, nor would a large number of Athenian hoplites have been absent from the city then. Hence we are left with 401-399/8 as the only plausible date for the war.[23]Admittedly these dates would require that the Elean War was interrupted by the Olympic games of 400, and that this interruption escaped the interest of our sources, despite the fact that Sparta’s disqualification from the games was a principle cause for the conflict. Underhill (supra n.1) 156 and Hatzfeld (supra n.1) 397 both doubt the likelihood of this. But arguments from silence are extremely weak when Diodorus or Xenophon are in question, and the interruption of wars or campaigns by regularly scheduled truces for athletic competitions are almost enver mentioned by even our most thorough sources. Indeed, in view of the reconstruction offered above, one might speculate that it was the approaching Olympic truce of summer 400 that persuaded Agis to depart Elis despite the military successes that had placed victory within his grasp.

While the resolution of such bothersome and venerable puzzles is of some interest for its own sake, our ultimate aim is historical analysis; it is time to ask what of substance we can learn from the reconstruction offered above.

First, we should note that, according to Xenophon, Spartan hostility and anger towards Elis was long-standing and severe; given Elis’ behavior, we have no reason to doubt this.[24]Kelly (22f) argues that Elis had reverted to the Spartan alliance “at the eleventh hour before the final defeat of Athens” in the Peloponnesian War. A priori this does not seem implausible, but the evidence Kelly cites does not counter the implication in Xenophon of unbroken Spartan-Elean hostility going back to 420. Furthermore, if Elis had returned to the Spartan fold in hope of assuaging Spartan anger, she would presumably have readmitted Sparta to the Olympic games at the same time for the same reason; yet all our sources agree that this was not the case at the start of the war. Yet it seems that Sparta waited until summer 401 to translate her feelings into military action. And even as late as the outbreak of the war, King Agis’ first campaign probably amounted only to a show of force, perhaps aimed at securing Spartan readmission to the approaching Olympic games of 400, and was not intended to escalate into full-scale war. This is a strong indication of Sparta’s own war-weariness following decades of brutal struggle with Athens, and of her reluctance to commit herself to a war with a first-class hoplite power. Sparta may have had few qualms about sending a Spartiate commander or two to lead a few thousand freed helots and Arcadian mercenaries to victory or death in distant Ionia,[25]Xen. 3.1.4, 4.2; the “allied troops” are presumably de facto mercenaries. Spartans obviously would have considered helots with military training as expendable an expeditionary force as could be imagined. It should be noted that Lysander persuaded Agesilaus to lead an expedition to Asia Minor in summer 396, a date firmly established. Lysander had long been hoping to find a way to regain his old command in Asia, and Agesilaus was heavily indebted to him politically for the succession (Xen. 3.31-3, 4.3); allowing Agesilaus several years to establish his personal power and influence, we would expect Lysander to have been able to obtain an Asian command for himself within a very few years of Agesilaus’ accession. An expedition prepared in winter 397/6 would somewhat strengthen the case for an accession in summer 400 (proposed above) as compared with an earlier year. but a war against a tough Peloponnesian state, requiring the lengthy deployment of significant Lacedaemonian forces, was another matter entirely. The pattern of the Elean War and its starting date are a testament to Spartan caution.

The aftermath of the war may also provide us a glimpse of internal Spartan policy: in particular, the policies of Sparta’s kings. A certain school of modern scholarship sees the Sparta of this era as dominated by the struggles of two (or three) contending Spartan political ‘parties’ centering on the kings and divided along ‘imperialist’ and ‘anti-imperialist’ lines.[26]Cf. C.D. Hamilton, “Spartan Politics and Policy, 405-401 B.C.,” AJP 91 (1970) 294-314, and Sparta‘s Bitter Victories (Ithaca 1979). Pausanias is viewed as the leader of the ‘anti-imperialist’ faction, while the Spartan war with Elis (as well as the Spartan expulsion of the Messenians settled at Naupactus immediately after the war) are seen as clear moves by the ‘imperialist’ faction.[27]Cf. Hamilton (supra n.26) 120f. Yet, as we have seen, it was actually Pausanias who led the final and decisive campaign of the Elean War—the only campaign that involved direct fighting. Furthermore, there is every reason to believe that it was again King Pausanias who actually captured Naupactus and expelled the Messenians.[28]In Diodorus’ account (14.17.4-12) we are told that Pausanias wintered at Dyme, just across the gulf from Naupactus, after his Elean campaign. When the narrative resumes (14.34.1-3) we are told of the Elean surrender and the Spartan attack on Naupactus that directly followed. The conclusion seems inescapable that it was Pausanias and his army who led that attack. Such points should cause us to question an over-simplified imperialist/anti-imperialist paradigm of Spartan policy-making.

Our reconstruction may also clear up one issue in the disputed succession following the death of King Agis, whose son Leotychides was declared illegitimate and displaced in his claim by Agis’ younger brother Agesilaus (Xen. 3.3.1-4; Plut. Ages. 3.1-4, Alc. 23.7f, Lys. 22.3-6). The story is an interesting one and naturally attracts the attention of our sources, who emphasize the important role played by Lysander in ensuring Agesilaus’ succession. But there is no sign at all that King Pausanias was involved in the dispute. Such a lack of involvement would be surprising, for Pausanias clearly had a strong interest in influencing the choice of his co-regent, and as a king of a long and generally successful reign, he would have possessed considerable auctoritas and might have had a dominant influence on the decision. Obviously such an argument from silence is far from conclusive—especially when Xenophon is our key source—but it is tempting to suggest that Pausanias had already departed with his army before the official period of mourning for Agis ended and the issue of the disputed succession arose. Perhaps this may explain how Lysander managed to obtain the kingship for his protégé Agesilaus over the express wishes of the dying Agis, and may further explain why Cinadon chose to organize his conspiracy to destroy the homoioi when he did: a large portion of the Spartiate manpower was in Elis with Pausanias, separated from Sparta by almost a hundred miles of rugged terrain.

There is, finally, a certain negative significance in our reconstruction of a plausible chronology that incorporates the accounts of both Xenophon and Diodorus: failing this result, we should have been forced to admit that one (or both) of our leading sources for this period was guilty of fabricating a convincing and detailed account of an important crisis for no obvious reason whatever; all credibility in that source would be greatly compromised.[29]Oddly, scholars who accept the truth of the antecedent fail to recognize the gravity of the consequent. Yet if (for example) Diodorus’ account of the Elean War were complete fiction, despite its precise detail, lack of rhetorical inflation, and even tone (all of which strongly point to the Oxyrhynchus historian as the ultimate source), then Diodorus’ factual statements must be assigned as little weight as his chronological pretensions. The implications of this conclusion would affect much of our picture of this entire era. But in our analysis we have merely found further examples of the sins of omission characteristic of both Xenophon and Diodorus; our view of the reliability of each is scarcely altered. This is a negative conclusion, but an important one.

Appendix: Suggested Chronology for the Elean War

Summer 401: Spartans under Agis launch ‘warning invasion’ of Elis, aimed at coercing her into admitting Spartans to the Olympic games of 400 and fulfilling other Spartan demands (Xen. 3.2.23f)

Winter 401/0: Elis reacts by calling Sparta’s bluff and attempting to create a united front of anti-Spartan states (Xen. 3.2.24)

Early summer 400: Agis leads a second, much more determined, invasion of Elis, this time including contingents from Athens and all the other Spartan allies except Corinth and Boeotia; Elean territory is ravaged, including the outskirts of the city, and a pro-Spartan coup narrowly misses succeeding; Agis withdraws his army, leaving a garrison under Lysippus (Xen. 3.2.25-29).

Early summer 400 (in Athenian archon year 401/0): Democratic forces at Athens use the opportunity of Sparta’s preoccupation in the Peloponnese and the absence of large numbers of the conservative members of the Athenian hoplite and cavalry classes (serving in Elis with Agis) to overturn Pausanias’ Athenian settlement of 403 and reincorporate Eleusis (Arist. Ath. Pol. 40.4, Xen. 2.4.43).

Summer 400: Olympic games held. After returning to Sparta and disbanding his army, Agis travels to Delphi to donate the spoils of his second campaign, then falls ill and dies soon after (Xen. 3.3.1).

Late summer 400 to early summer 399: Spartans under Lysippus continue ravaging Elean territory (Xen. 3.2.30); meanwhile the Eleans fortify their city against a future Spartan assault.

Early summer 399: Elis seeks peace, but refuse to accept Sparta’s harsh terms (Xen. 3.2.30f); fearful of a new Spartan invasion, Elis hires 1,000 elite Aetolian mercenaries (Diod. 14.17.9)

Summer 399: Pausanias leads a third invasion to compel Elis to accede to Sparta’s harsh demands; after ravaging Elean territory, he is defeated in a skirmish outside the city walls, but maintains Spartan pressure and winters with his army at Dyme, on Elean soil (Diod. 14.17.4-12).

Summer 399: After the official Spartan period of mourning for Agis, Agesilaus comes to the throne in a disputed succession; soon afterward he discovers and crushes the conspiracy of Cinadon (Xen. 3.3.1-11).

Early 398: Recognizing that Sparta will not compromise and abandon the war, Elis surrenders (Xen. 3.2.31, Diod. 14.34.1f); Pausanias leads his Spartan forces from Dyme to Naupactus and expels the Messenians settled there (Diod. 14.34.1-3).

References

[1] Cf. W. Judeich, Kleinasiatische Studien (Marburg 1892) 182f; G.E. Underhill, “The Chronology of the Elean War,” CR 7 (1893) 156-58; E. Meyer, Theopomps Hellenika (Halle 1909) 114-16; J. Hatzfeld, “Notes sur la chronologie des Helleniques,” REA 35 (1933) 395-409; and D.H. Kelly, Sources and Interpretations of Spartan History in the Reigns of Agesilaus II, Archidamus III and Agis III (diss. Cambridge 1975 [hereafter ‘Kelly’]) 21-41. The inveterate notoriety and difficulty of the issue is summed up in Underhill’s opening sentence: “A complete solution of all the difficulties involved in his problem, which has occupied the attention of commentators and historians for more than a century, is almost out of the question: we have to content ourselves with weighing one set of probabilities against the other.”

[2] Xen. Hell. 3.2.21-3.3.1, Paus. 3.8.3-5; Diod. 14.17.4-12, 34.1. Though condensed, Pausanias’ account of the war is virtually identical with Xenophon’s both in material and in order of presentation, the only major difference being his statement that the war lasted three years, whereas many modern scholars interpret Xenophon’s phrase περιιόντι τῷ ἐνιαυτῷ to imply that the war lasted only two years (cf. n.3 infra). Pausanias’ other differences from Xenophon involve minor details, and are probably best explained as errors due to Pausanias’ faulty memory or his intermediate source, although Kelly (21) sees in them the traces of an additional tradition.

[3] Xenophon’s actual phrase is “as the year was coming round” (περιιόντι τῷ ἐνιαυτῷ). Whether this should be interpreted to mean that Agis’ second campaign took place in the same season as the first had been an important point of contention: Judeich (supra n.1) argues that it should, and G.L. Cawkwell, ed ., Xenophon: A History of My Times (Harmondsworth/New York 1979) 155, appears to concur, letting stand Warner’s translation of the phrase “in the same year.” The Loeb version by G.L. Brown (London/Cambridge [Mass.] 1961) has “in the course of the year.” The most natural meaning of the Greek would seem to be “as the year was changing” or “at the start of the new year,” as argued by Underhill (supra n.1) 157 and Kelly 25 n.40; and this is clearly the sense in which Pausanias or his intermediate source took it (thus his figure of three years for the war; cf. supra n.2). In fact the main reason for confining Xenophon’s account of the war to two years is to reconcile it with Diodorus, who chose to distribute the war into two political years; but in view of Diodorus’ almost random chronological reckoning, this is a dubious advantage.

The pattern of events themselves also strongly supports giving Agis two summers of campaigning. It is possible to squeeze a Spartan mobilization, a short campaign in Elis, a Spartan demobilization, the dispatch of Elean embassies to various anti-Spartan states, a second Spartan mobilization (this time including nearly all of Sparta’s allies, notably the Athenians), and a second Elean campaign of some length into a single campaigning season, and still leave “the rest of the summer” for the ravages of Lysippus (Xen. 3.2.23-30); but such a reconstruction is forced and implausible. It is far more natural to attribute the two campaigns to separate summers, with diplomatic exchanges occurring during the intervening winter.

[4] G.L. Cawkwell, “Agesilaos and Sparta,” CG N.S. 26 (1976) 75 n.48, argues that Xenophon actually means that only the acropolis was unwalled, rather than the city as a whole; this is primarily an unnatural attempt to reconcile Xenophon with Diodorus and Pausanias (see text infra), and its motive disappears if a simpler hypothesis will serve. Moreover, the logic of the argument seems flawed: Xenophon is using the absence of walls to prove that Elis was vulnerable to capture, and if the city walls had been as strong as Diodorus implies, an unwalled acropolis would not have been cited as proof of vulnerability.

[5] Cf., among others, Judeich 182 and Meyer (both supra n.1); Cawkwell (supra n.4) 76 n.51.

[6] It is not surprising that Xenophon’s Hellenica demonstrates a wealth of inside information about Spartan society and politics, for he had many Spartan friends, notably King Agesilaus, and spent the latter half of his life in the Spartan-controlled Peloponnese. Cf. Cawkwell (supra n.3) 12-15.

[7] So Kelly 29; cf. I.A.F. Bruce, An Historical Commentary on the Hellenica Oxyrhynchia (Cambridge 1967) 20-22.

[8] Assumed by Kelly (30), without elaboration.

[9] Even if we leave aside the question whether Xenophon is referring to the city or to the acropolis (cf. supra n.4), the statement remains that the Spartans believed that the city would be easy to capture. This would seem absurd if a full Spartan army under Pausanias had battered its head against the city a year or two before and finally given up because Elis was too strong to besiege.

[10] Cf., for example, the rapid Athenian fortification of Pylos (Thuc. 4.4f).

[11] The pay for 1,000 mercenaries at 3 obols per day (perhaps more for elite troops) would have come to at least 2.5T of silver per month, a considerable expense for a largely agrarian Peloponnesian state.

[12] We should note that Archidamus had adopted a similar strategy at the beginning of the Peloponnesian War (Thuc. 2.18).

[13] Cf. Cawkwell (supra n.3) 15ff, esp. 33f.

[14] Cf. Cawkwell (supra n.3) 17, 20, 35-38, 40-43.

[15] Xen. 3.1.5 does contain a vague sentence that could be taken to summarize these events; if so, this would parallel the manner in which his vague reference to Spartan ‘compulsion’ may summarize Pausanias’ invasion of Elis (supra 34f).

[16] Xenophon often digresses in just this way to describe the eventual outcome of an event before returning to chronological sequence: cf. the death of the exiled Pausanias (3.5.25). On our hypothesis Xenophon here digresses to sum up the end of the Elean War before returning to Agis’ activities following the end of the second Elean campaign. Otherwise, we would expect Agis to have dedicated his booty from the invasion immediately, rather than wait a year or ore until the end of the war. As I argue elsewhere, similar digressions on the part of Thucydides have been responsible for much confusion in the chronology of the Pentecontaetia: cf. R.K. Unz, “The Chronology of the Pentekontaetia,” CQ (forthcoming).

[17] For example, Diod. 14.32 places the overthrow of the Thirty at Athens (403) just after the return of the Ten Thousand (399), and both under the political year 401/0. His numerous other chronological blunders are almost as severe.

[18] For example, by Cawkwell (supra n.3) 154 and Kelly 24-26, whose efforts at chronological reconstruction are hindered by their refusal to consider the possibility that Agis may have died before the end of the Elean War, and that Xenophon’s account of the end of the war may represent a chronological digression (supra n.16).

[19] The year of Dercyllidas’ arrival in Asia is solidly based and is accepted by nearly all the scholarly literature. For discussion cf. K.J. Beloch, Griechische Geschichte III.2 (Berlin 1922) 211f.

[20] Cf. Kelly 33-35.

[21] Note that Plutarch’s full claim is that Agesilaus “had been regarded as the king and leader of almost the whole of Hellas” down to Leuctra, which is clearly false. Kelly (32 n.62) speculates that Plutarch’s error may follow from his own uncertainty as to the exact date of Leuctra.

[22] Cf. J.R. Cole, “Cimon’s Dismissal, Ephialtes’ Revolution and the Peloponnesian War,” GRBS 15 (1974) 369-85.

[23] Admittedly these dates would require that the Elean War was interrupted by the Olympic games of 400, and that this interruption escaped the interest of our sources, despite the fact that Sparta’s disqualification from the games was a principle cause for the conflict. Underhill (supra n.1) 156 and Hatzfeld (supra n.1) 397 both doubt the likelihood of this. But arguments from silence are extremely weak when Diodorus or Xenophon are in question, and the interruption of wars or campaigns by regularly scheduled truces for athletic competitions are almost enver mentioned by even our most thorough sources. Indeed, in view of the reconstruction offered above, one might speculate that it was the approaching Olympic truce of summer 400 that persuaded Agis to depart Elis despite the military successes that had placed victory within his grasp.

[24] Kelly (22f) argues that Elis had reverted to the Spartan alliance “at the eleventh hour before the final defeat of Athens” in the Peloponnesian War. A priori this does not seem implausible, but the evidence Kelly cites does not counter the implication in Xenophon of unbroken Spartan-Elean hostility going back to 420. Furthermore, if Elis had returned to the Spartan fold in hope of assuaging Spartan anger, she would presumably have readmitted Sparta to the Olympic games at the same time for the same reason; yet all our sources agree that this was not the case at the start of the war.

[25] Xen. 3.1.4, 4.2; the “allied troops” are presumably de facto mercenaries. Spartans obviously would have considered helots with military training as expendable an expeditionary force as could be imagined. It should be noted that Lysander persuaded Agesilaus to lead an expedition to Asia Minor in summer 396, a date firmly established. Lysander had long been hoping to find a way to regain his old command in Asia, and Agesilaus was heavily indebted to him politically for the succession (Xen. 3.31-3, 4.3); allowing Agesilaus several years to establish his personal power and influence, we would expect Lysander to have been able to obtain an Asian command for himself within a very few years of Agesilaus’ accession. An expedition prepared in winter 397/6 would somewhat strengthen the case for an accession in summer 400 (proposed above) as compared with an earlier year.

[26] Cf. C.D. Hamilton, “Spartan Politics and Policy, 405-401 B.C.,” AJP 91 (1970) 294-314, and Sparta‘s Bitter Victories (Ithaca 1979).

[27] Cf. Hamilton (supra n.26) 120f.

[28] In Diodorus’ account (14.17.4-12) we are told that Pausanias wintered at Dyme, just across the gulf from Naupactus, after his Elean campaign. When the narrative resumes (14.34.1-3) we are told of the Elean surrender and the Spartan attack on Naupactus that directly followed. The conclusion seems inescapable that it was Pausanias and his army who led that attack.

[29] Oddly, scholars who accept the truth of the antecedent fail to recognize the gravity of the consequent. Yet if (for example) Diodorus’ account of the Elean War were complete fiction, despite its precise detail, lack of rhetorical inflation, and even tone (all of which strongly point to the Oxyrhynchus historian as the ultimate source), then Diodorus’ factual statements must be assigned as little weight as his chronological pretensions. The implications of this conclusion would affect much of our picture of this entire era.