• 1 Thuc. 2.70.3-4.

1War made many refugees in ancient Greece as in the modern world. Families might pack up and leave their cities rather than be caught by an invading enemy, or make their escape from a siege, or surrender and be allowed leave home with only the clothes on their backs. Once, when Athenian generals allowed the people of Potidaea to leave their besieged city with ‘an agreed amount of money for the road’ as well as one or two items of clothing, the Athenian assembly blamed them for agreeing such mild terms.1

• 2 Xen. Ages. 1.21-22.

2Often, however, a captured or surrendered community would not be driven into exile but seized and sold into slavery. Xenophon suggested that even this might create refugees, indeed give rise to a particularly acute refugee problem: the youngest, oldest and weakest captives might not find a buyer and simply be left by the roadside, without anywhere to live or anyone to look after them.2 Xenophon mentioned this because in his opinion the way in which Agesilaos II of Sparta (c. 400-360) dealt with the problem was a prime example of the king’s exceptional ‘benevolence’.

• 3 Xen. Ages. 1.16; Hell. 3.4.12.
• 4 Diod. Sic. 14.79.2.
• 5 E.g. Hermippus, fr. 63 K-A; Lewis, 2015, p. 318; id., 2011, p. 104-105, 109-110; Vlassopoulos, 2010
• 6 Pritchett, 1991, p. 404-416, collects the evidence for sale of spoils in Spartan armies.
• 7 Xen. Ages. 1.18.
• 8 Xen. Ages. 1.28; Hell. 3.4.19.

4The context of Xenophon’s remark about Agesilaos’ kindness towards abandoned captives was the king’s first ever military campaign, an invasion of Asia Minor in 396 BC, advertised as a war to liberate Greeks from the Persian Empire, but in effect a vast plundering expedition deep into Phrygia. Xenophon stressed that Agesilaos seized many unsuspecting cities and collected a great quantity of spoils.3 According to Diodoros the ‘market mob’ (ἀγοραῖος ὄχλος) of traders who followed Agesilaos’ army ‘also for the sake of plunder’ (καὶ τῆς ἁρπαγῆς χάριν) was as large as the army itself.4 Captives must have been a major part of these spoils: Phrygian slaves were in high demand at Athens, for example.5 The sheer volume of spoils produced a glut, and when the army’s official ‘booty-sellers’ (laphyropoloi)6 sold off the plunder to traders on the spot ‘everything was sold for next to nothing’. Agesilaos allowed his friends to buy at these minimal prices, on credit, so that they could sell at a large profit when the army returned to the Greek cities on the coast where the demand was much greater.7 Xenophon later adds that Agesilaos ordered male prisoners to be stripped naked when sold, so that his soldiers would see these ‘barbarian’ captives’ untanned skins, their ‘fat and inert’ bodies, and despise them as opponents no more dangerous than women.8

• 10 See e.g. Azoulay, 2004, p. 320 (= 2018, p. 192-193) on the importance of ‘benevolence’ in Xenophon’ (...)

6‘To guard’ (φυλάττειν) is an ambiguous verb in Greek as in English, meaning both ‘to guard against danger’, i.e. to protect, to preserve safely, and ‘to keep under guard’, i.e. to hold prisoner. Xenophon’s wording of Agesilaos’ instructions suggests that the king was telling his soldiers not to maltreat their prisoners but preserve their dignity as human beings, and this is no doubt the message Xenophon wanted to convey.10

• 11 Xen. Anab. 2.2.20.

7However, if we take ‘guard’ in its other sense, and accept that ‘revenge on wrongdoers’ would typically take the form of killing them, Agesilaos’ announcement in essence amounts to an order not to kill those whom they captured but to keep them alive as prisoners. This interpretation is all the more likely since Agesilaos did not offer his views as informal advice or exhortation, but ‘publicly ordered’ (προηγόρευε) this course of action. The same verb is elsewhere used explicitly for a public announcement made by ‘the commanders’ and formally proclaimed on their behalf by a herald.11 In that context, it seems likely that Agesilaos did not just tell his soldiers to be nice to prisoners of war but ordered them not to kill and instead take prisoners for sale into slavery. Since this campaign was a plundering raid, and since the revenue from spoils sold by the ‘booty-sellers’ flowed into the central treasury, it was very much in Agesilaos’ own interest to issue such orders.

• 13 A very young age is implied by the fact that they still needed to be carried.

9What exactly was the problem that Agesilaos noticed, and what kind of ‘care’ did he bestow on these abandoned babies and toddlers?13

• 14 As suggested by Krentz, 2007, p. 153; Hunt, 2018, p. 63.

10Evidently, the slave traders who bought captives acquired some small children, whom they tried to sell when the army moved camp, but left behind when they could not find a buyer. By implication, these traders did not do the same to adult captives: normal practice was not to sell slaves on immediately, but to keep them under guard and bring them along, buy more as the march continued, and finally sell them at the end of the campaign in a market where demand was high, i.e. in a Greek coastal city. Agesilaos encouraged his friends to do the same. But the traders evidently did not feel this was a viable procedure for the small children, so the question arises why they bought child captives at all. The answer must be that the booty-sellers sold families together,14 or at least sold women together with their children.

• 15 IG I³ 422 ll. 195-197, as convincingly explained by Schmitz, 2011.
• 16 Dem. 19.305-306.
• 17 Tchernetska et al., 2007.

11This practice is implied by an entry in the Attic Stelae, where three Thracian slaves are auctioned off for a single large sum, evidently a mother with her son and daughter.15 The same custom is hinted at in a description of a group of 30 ‘women and little children’, taken captive in the sack of Olynthos in 348 BC, following their new master on foot.16 It is now also explicitly attested in a new fragment of Hypereides’ Against Timandros17, which denounces as evil the forced separation of siblings from one another:

• 18 Diod. Sic. 14.79.3.
• 19 Xen. Hell. 3.4.13, 26; cf. 4.1.1, 15.
• 20 See e.g. Lewis, 2018, p. 276-277.

13The idea that slave dealers kept children and their mothers together ‘as much as possible’ implies that in some circumstances they might nevertheless break up families, and there were indeed special circumstances which account for the scale of abandonment during Agesilaos’ campaign of 396 BC. This expedition was unusually long, lasting ‘most of the summer’ until ‘the start of autumn’, and marched deep inland, into ‘Phrygia and adjoining regions’;18 Xenophon specifies that the army got as far as Dascyleion in north of Phrygia before returning to Ephesos,19 a trek of nearly 200 miles each way as the crow flies and in practice of course much longer still. Transporting large numbers of slaves over such distances for such a long time posed a particular challenge.20

• 21 Hunt, 2018, p. 63. Cf. Lewis, 2015, p. 331-333, for overland transport of slaves in ‘coffles’, stri (...)
• 22 Xen. Ages. 1.18.

14One reason for abandoning the little children may have been simply in order to keep up with the army, which would be harder with infants and toddlers in tow to ‘slow and distract their parents’.21 But Xenophon’s reference to the difficulty of ‘feeding’ the children suggests an additional and perhaps even more compelling reason. On a very long campaign the cost of feeding slaves until they could be sold would be unusually high, while the sheer numbers of slaves for sale meant that profit margins would be unusually low, so that the less valuable slaves were simply not worth keeping. A merchant who wanted to buy an adult female slave had no choice but to buy her children as well, but would abandon them if their maintenance threatened to cost more than what they could be expected to fetch eventually when sold. The same cost-benefit analysis underlies Agesilaos’ scheme for enabling his friends to make large profits on the sale of slaves and other spoils: he not only let them buy on credit, but also ‘told them that he would be going down to the coast soon with the army’,22 i.e. he provided them with insider information that so that they could carefully time their purchases and make sure that they did not have to feed their newly bought slaves for any longer than strictly necessary.

• 23 One of the referees for Pallas notes the evidence for children being sold as slaves (e.g. Xen. Anab(...)

15On shorter campaigns which brought fewer slaves to the market, profit margins might have been good enough for traders not to abandon slave children, and the problem need not generally have been as widespread as it was under Agesilaos’ command in 396.23 However, abandonment was evidently at the discretion of the slave trader, and economic considerations might very well override any moral qualms about such a course of action.

• 24 Funari, 2013: portati con loro da qualche parte, ‘carried with them somewhere’, which presumably en (...)

16What, then, was Agesilaos’ solution to the problem? ‘He took care of them, ensuring that they would be gathered together somewhere.’ Several modern translations try to make sense of this elliptic statement by adding something that does not feature in the Greek. ‘He would show concern in behalf of these poor waifs and have them conveyed to some place of safety’ (Dakyns, 1890-7), ‘some place of refuge’ (Marchant, 1925; Azoulay, 2018, p. 208), ‘quelque endroit sûr’ (Azoulay, 2004, p. 341). The assumption is that the children were somehow ‘rescued’ (Krentz, 2007, p. 153) or ‘taken into care’ (Gray, 2011, p. 32). Yet the verb used by Xenophon, συγκομίζω, does not elsewhere have any connotation of ‘conveying to safety’: it simply means bringing together in one place things that had previously been dispersed, and it is regularly used of ‘gathering in the harvest’, for example (see LSJ s.v.). Nor does the verb mean ‘to bring along’, as seems to be assumed in a recent Italian translation,24 and in an unusual interpretation of the passage which envisages Agesilaos’ soldiers picking up abandoned children encountered during the march (Pritchett, 1991, p. 411).

• 25 Similarly, Waterfield, 1997 (‘rounded up and taken somewhere’); Bartlett, 2018 (‘conveyed somewhere (...)
• 26 Van Wees, 2004, p. 149.

17Instead, the children were merely ‘gathered together somewhere’.25 When the army left camp, and the merchants packed up their own tents and booths, babies and toddlers were left behind here and there, and when Agesilaos noticed this – one imagines that he heard their mothers screaming and crying – he had them collected in a single place. Note that he did not order the merchants to take these children along with their parents and older siblings until they could eventually be sold – as I assumed in a brief previous discussion of the passage26– which would have been an obvious and relatively humane thing to do. Instead, he apparently merely arranged for their collection.

• 28 The main verb προστάσσω + dative and infinitive means ‘to give an order to someone to do something’ (...)

19The translation I have offered is grammatically straightforward28 and its meaning is clear: Agesilaos gave an order to elderly captives, who were also abandoned by the slave traders as insufficiently valuable wares, to look after the abandoned children and protect them from dogs and wolves. In order words, Agesilaos’ intervention was limited to collecting together all the abandoned prisoners and telling them to look after one another: the old and infirm were ordered to protect the young and defenceless from predators.

• 30 Pritchett, 1991, p. 411.

22Much nearer the correct interpretation is the oldest English translation I have found, which says that children were ‘conveyed to some place of safety; or he would entrust them to the care of fellow-prisoners also left behind on account of old age; in no case must they be left to ravening dogs and wolves’ (Dakyns 1890-7). This introduces a contrast between two ways of helping the children when the Greek has only one, but is otherwise right. Also nearly correct is a paraphrase in W.K. Pritchett’s The Greek State at War, which has the army on the march picking up children abandoned by the roadside who are then ‘given over to the charge of those elderly natives whom age and feebleness had spared from being sold’.30 The suggestion seems to be that the children were left with the few remaining inhabitants of local villages, whom the soldiers had left alone. That cannot be right, because the elderly were also ‘captives’ who were ‘left behind’, i.e. by traders who had previously bought them. That aside, Pritchett did see the main point: Agesilaos provided no care for the elderly, but on the contrary added to their cares by making them look after the children.

• 31 This campaign thus supports the argument of Lewis, 2018, p. 284-286, that Greek demand for slaves c (...)

23In sum, Agesilaos’ campaign in Phrygia caused huge displacement across the region as he made slaves everywhere,31 ordering his soldiers to make prisoners rather than kill, while giving his friends the opportunity to make a killing on the slave market. A side-effect of his extensive raiding was that slave traders, seeing their profit margins shrink, overcame whatever qualms they might have had about separating mothers and children, and left infants and toddlers behind, along with elderly slaves too old and weak to keep up. Agesilaos made no attempt to stop traders from doing this, but did go so far as to have soldiers search his abandoned camps and round up all those left behind, then leaving each newly formed group on the spot to try and survive as refugees, with nothing but the exhortation to look after one another and not to let the babies be eaten by wild animals.

• 32 So most recently, with specific regard to Agesilaos, Bartlett, 2018, p. 79-106.
• 33 This paper has benefited from the helpful comments of Laura Loddo and two referees for Pallas; any (...)

24If food, clothes or shelter had been provided, Xenophon would surely have made a point of mentioning this, but he does not. Incredible as it may seem, the mere gesture of bringing refugees together before leaving them in the lurch is what constitutes Agesilaos’ ‘humanity’, which, Xenophon insists, was so exceptional that ‘not only people who heard about this but even the captives themselves took a liking to him’ (καὶ αὐτοὶ οἱ ἁλισκόμενοι εὐμενεῖς αὐτῷ ἐγίγνοντο, Ages. 1.22). This jaw-dropping claim may be taken by some as evidence that Xenophon’s seemingly straightforward work is to be understood as deeply ironic throughout, and his apparent eulogy of Agesilaos as a sarcastic indictment.32 But most of those in antiquity who escaped slavery only to end up living as refugees would not even have enjoyed the relative safety of being part of a group, and I would argue that Xenophon’s words were genuinely intended as flattery. Such was the brutality of Greek warfare that even a token gesture of concern by a general for those whom he had made slaves or refugees could legitimately be praised as notable proof of humane ‘benevolence’.33