Chapter 4
It was not only among the established states of the Balkan Peninsula that the Galatian invasions had a profound effect. The Galatians themselves were also affected. Three tribes moved east into Asia, and their fates will be examined in the next chapter – it is surely an extraordinary matter to find a new state of Kelts planted in Asia Minor, in the land of the Hittites and the Phrygians – but there were also established two Galatian states in the Balkans, one a long-standing fixture in the Balkan state system, the other a brief presence, if unpleasant. The effect of these two were as powerful on their neighbours as they had been on their Greek-speaking victims.
The Scordisci
The evacuation of Greece and Macedon took the beaten Galatians formerly commanded by Brennos and now (presumably) by Akichorios, in two directions. One group headed north through the Dardanian country, where the Dardani copied the Thessalians and turned on their former allies.1 The survivors returned to the Banat country. Some perhaps went onwards further north, if that had been their homes, perhaps as far as Germany, and in one case apparently to Gaul, the clue some of the loot at Tolosa. The second group, as already mentioned, turned east and moved along the Thracian coast into the lands bordering on the Hellespont and the Propontis.2
The result of their retirement (if that is not too gentle a word for their defeat) from Greece was eventually the formation in the Balkans of two new Galatian states by the survivors and others, the Scordisci and the Tylis. Neither of these two states is known at all well, the source material being as fragmentary and discontinuous as anything else about Thrace and the Galatians generally. The name of the Scordisci is first noted by Justin, writing in the second century AD, in relation to the return of one part of the Greek expedition, where he notes that the returnees formed the Scordiscian kingdom on their arrival.3
From that it is assumed that the state was founded in 278 or 277, but that is not necessarily to be inferred from Justin’s words, though he does say that they ‘chose the name of Scordisci for themselves’; this may be his assumption, knowing that they bore that name later, it is not at all certain that it was ever a kingdom, kingship being an institution only two of the Galatian (or Keltic) states had in the ancient world. Perhaps it had a king, though, at the very end, close to the time of Justin’s source, Pompeius Trogus, as a state with that name is not referred to in written sources until the beginning of the second century.4 I have, as will have been noted, used the name for the state which had come into existence in the Banat not long after 400 BC, and whose envoys met Alexander in 335.
Another group which appears to have left the Banat region at about the time that the defeated forces returned, and which moved into Thrace, became the founders of the Tylis kingdom – this is the other Galatian kingdom – while some of those who had defected from the main returning army also contributed to its formation. The kingdom of Tylis is discussed in the latter half of this chapter; the Scordisci may be taken before this, but only after consideration of the situation in Macedon and Thrace in the aftermath of the great retreat.
There was for some years a confused condition of affairs in both Macedon and Thrace after the invasions and the retreat, in the lands on either side of the Haemos mountain range. The invasions had had a decisive effect in Macedon, as noted in the last chapter, in eventually forcing the acceptance of a competent king; and it had hardly less effect on the tribes between the Macedonian frontier and the Danube. To the west the Illyrians had been badly damaged by the Galatian invasions, but had survived, though infiltrated and still often attacked by the invaders, almost as though the Scordisci preserved the Illyrians as prey so that they could practise their raiding techniques on them.
Their effect on Thrace was much greater. The tribal organizations of the Triballi and the Thracians in eastern Thrace had been all-but destroyed in the troubles since the invasion by Philip II. In combination with the invasions of Philip and Alexander between 350 and 335, and then of Alexander’s governors, and eventually of Lysimachos, all of whom fought the Thracians with ferocity, the Galatian invaders in 281-277 had probably performed the final social and political destruction to a tribal system which was already reeling towards collapse from the Macedonian attacks. There can have been only broken remnants of the old tribal order left by the time the Galatian movements subsided.
The initial Thracian revival after Alexander’s time, which was associated with Seuthes III and his city of Seuthopolis, had been stillborn; the city had been destroyed, either by Lysimachos or more likely by Kerithrios’ Galatian raid in 280; that Galatian raid also almost finished off the Triballi. The Macedonian colonies of Philippopolis and Kabyle certainly survived for a time, though it is possible that the former had suffered badly; the two border towns of Vetren and Pernik had been destroyed, though we cannot tell by whom.5 At Kabyle coins produced there later were in the name of a Thracian, who was either a magistrate or a ruler of some sort, his authority was probably restricted to the town and its immediate environs.6
However, a broken political system such as evidently had been the result in Thrace because of the several invasions and passages of armies, does not mean the extermination of the population. Certainly casualties were great – Alexander killed at least 5000, so Arrian claimed,7 and Philip and Lysimachos were probably even more lethal – but it is effectively impossible to exterminate a whole people without industrial methods, and not even then. Some will always survive. The Triballi, for example, were battered by Alexander, by Lysimachos (probably), and by Kerithrios, yet they were still identifiable as a much-weakened tribal group for some time after all that.8
What was destroyed in the fighting, of course, was the elite layer of the population, those whose conspicuous wealth gave them the authority of leading figures in the tribe, and made them the obvious targets for the invaders, both as possessors of portable valuables, which could be looted and stolen, and as the wielders of political and military power within the society. These were the men who had the authority to mobilize the people, either to resist attack, or to persuade them to submit to conquest. They were the glue which held such tribes together, and without such an elite social layer the tribes disintegrated into their component sections, which would have been clans, or even individual families.9 The elite were never numerous, and if they kept fighting every invader, they would soon be killed off – and if their wealth was stolen by those same invaders, they could not be replaced. In the place of the former tribal organization there remained the smaller population units, villages, families, clans, out of which eventually a new state might arise, but only after a period of peace and capital accumulation, financial and military; only then could a return be made to a renewed and larger political system.
This new, post-invasion condition was one, therefore, of small, even minuscule, political units. This was what existed in the territory north of Macedon in the period before the Galatian invasions of Macedonia and Greece, and after they had returned, defeated, from Greece and Macedon. It was out of these fragments that new tribes, cities, and kingdoms had to be constructed, two of these being the Scordisci and Tylis states. In these the new elite were the Galatians, some of them became rich from their predatory raids; the rural population whom they subjugated were a mixture of peoples, Illyrians, Thracians, other Galatians, even Greeks and Macedonians. Beyond them, the Dardanians had survived the invasions well, thanks to the devious and successful diplomacy of their king, and they now emerged strengthened; to the west, the Illyrians similarly had survived, and some of them developed a new state which troubled the Romans and the Macedonians during the next century.
Macedon itself slowly recovered during the 270s. Antigonos imposed peace on the kingdom which permitted it to recover. The work took several years, and the detailed recovery very much longer. For the next thirty years he operated to encourage this recovery, taking care that his wars were defensive and limited, even reducing his involvement in the affairs of Greece to a minimum, relying mainly on alliances with local groups (as in Athens) or on supporting local tyrants in other cities, such as Argos.10
One of the results of this necessary policy of recovery was that Antigonos had to relinquish any conquests to the north of Macedon proper, and give up any hope of recovering the lands which had been gained by the kings from Philip II to Lysimachos. These extensive conquests had always been difficult to hold, and their original conquest had never been more than superficial, and had had to be repeated several times. The people there were still rebellious even after fifty years of Macedonian conquest and rule; with the drastically-reduced resources of Macedon after the conquests in Asia and the Galatian wars, holding such an extensive territory was impossible; Antigonos apparently made no attempt to exert his authority beyond the old Macedonian boundaries.
The northern frontier of Macedon therefore retreated to where it had been at the beginning of Philip II’s reign, along an approximate line from the mouth of the Nestos River to the Pindos Mountains, utilizing the convenient mountain ranges as the frontier but holding in some strength the passages through them. This was, as earlier Macedonian experience had shown, a defensible line, one which had very largely held at first against the Galatians. The line was close enough to the centres of Macedonian manpower in the plain to be guarded and to allow Macedonian forces, if an attack was threatened, to arrive in time close to the frontier to drive out any invaders, and in most cases to prevent their penetrating the defensive line in the first place.
This left the land north of that frontier, as far as the Danube, the land still called Thrace, outside any of the organized kingdoms, at least for the moment. It was in this area and out of the fragmented populations which had survived the invasions and conquests that the Galatians crafted their two new states.
There were two kingdoms north of Macedon which had survived the invasions, like Macedon, battered but unbroken. Due north of Macedon were the Paionians, who had been part of the Macedonian family of peoples, with an autonomous dynasty in office from the time of Philip II (and perhaps before) for well over a century. Their fight against Brennos’ forces in 279 had no doubt left them weakened, but in the new circumstances it was in Antigonos’ interest that the kingdom should survive and continue; from Macedon’s viewpoint, Paionia was a most useful buffer state, which could absorb blows from the north – just as it had done at the beginning of the Galatian attacks – and which, moreover, might give good notice of any northern troubles.
The other kingdom which had survived was that of the Dardanians. The king had played a clever game amid the Galatian invasions, first allying with them, then attacking them, and no doubt fleecing them of as much of their Greek and Macedonian loot as he could seize. As a result, the Dardanian kingdom emerged from the invasions both strengthened and powerful, and no doubt it absorbed some of the damaged and smaller neighbours, and so increased its population. It occupied a fairly extensive territory north and west of the Paionians, between them and the Galatian Scordisci, who dominated the Banat and the surrounding lands. So between them these two surviving states covered much of Macedonians’ northern frontier; but while the Paionians were friendly, the Dardanians were consistently hostile; it was Dardanian enmity which kept the northern frontier active throughout the third century. To the east of the Paionians was Thrace, which did not recover for some time, partly because of the battering it had received, but also due to the absence of politically-effective leadership.
The Galatian returnees, not surprisingly since their expedition had been composed of voluntary groups who came together for the single occasion, broke up once the dangerous journey seemed over. There emerged at least three, perhaps four, separate groups. The largest became the Scordisci and settled back into the Banat region, many of the participants probably having set out from that area in the first place. A second group, perhaps under a chief called Kommontorios, who actually may not have taken part in the Greek expedition, moved out of the Banat into Thrace, following the route of Kambaules and Kerithrios, and became the founding group of the Tylis kingdom. The third group was that under Leonnorios and Loutarios, which had broken away from the main force while it was still in Macedon, or the Dardanian country, and eventually became the founders of the Asian states of Galatia. It is also possible that a fourth group abandoned the Balkans altogether. These are said to have gone ‘home’ to Gaul, to the Tektosages of southern Gaul, and to have deposited their stolen Greek treasures in the sacred lake of the temple of Tolosa (Toulouse), whence some of it was relooted by a Roman expedition in 106 BC. This story is often doubted, but we know from further examples that contact continued to be made between the several branches of the Tektosages, a tribe much given both to dividing and to each section retaining its Tektosagian identity. There were continuing Tektosagian groups in Gaul and Asia and Germany, and the idea of some individuals from the Greek expedition heading to the Gallic homeland to boast of their deeds and display their loot is hardly unbelievable; boasting seems to have been a male Galatian character trait. Justin also notes that this treasure display inspired more of the Tektosages of Gaul to migrate eastwards when they settled in Pannonia. The story of a ‘return’ to Tolosa may be doubted, but one may not necessarily doubt the mass movement of a group from Gaul to the Hungarian Plain, and the second to a degree depends on accepting the first.11 A Tectosagian group was certainly present in the Greek invasion force, and followed the other two into Asia.
Despite this disclaimer, there is little doubt that the Scordiscian state was organized and established in the distinctive geographical situation in the Banat area well before the return of the defeated refugees from Greece. The wealth acquired by the participants became the social basis for the aristocrats who were the rulers. By the time they were returning northwards, having got through the Dardanians’ lands, they were under the leadership of a man called Bathanattos.12 He was a survivor of the Greek expedition: Akichorios was no longer the leader; either he died in one of the various fights during the return, or he was deposed as leader. Bathanattos appears to have been the leader of those who were determined to return to the Danube area and to settle down there, apparently with unpleasant memories of their experiences. The people are said to have refused to use gold in reaction to the ill fortune of the Delphian attack; they were perfectly willing to use silver, however.13 (This, if it is true, might imply their failure to capture the temple, or simply be a reaction to their defeat and the large number of casualties they had suffered.) No doubt the returnees’ arrival caused a good deal of political unrest (they were, after all planning to settle elsewhere), but Bathanattos established some authority, and can be regarded as the re-founder of the state. But since Galatian settlement in the area was by this time a century old, what occurred was the reinforcement of the earlier settlement; Bathanattos emerged as a major figure partly because he had already emerged as the returnees’ leader, and partly no doubt because he had held on to a substantial quantity of Greek loot.
In fact, the returnees had little choice as to their destination. Unless they were to go off on their travels again, perhaps following in the tracks of the invaders of Thrace or Asia, the Banat area was as far as they were likely to get. To the north was the Pannonian area, the Hungarian Plain, on both sides of the Danube, which was by this time fully organized and controlled by Keltic groups – from whom the Kelts of the Banat had themselves probably come as part of the continuous migration and conquests during the previous centuries. To the east, the Dacians had received some Keltic immigrants and goods, but were now probably less than welcoming to more, and later became actively hostile.14 And it must be presumed that, after their homeward retreat over a distance of almost 1000 kilometres, fighting repeatedly, and subdividing more than once, it is unlikely that the survivors were either numerous or were willing to go further. Their repeated battles will have cost them many casualties, and more were suffered by the harassment they received on their return march.
Most of the men probably were from the Banat region in the first place, and those that were not would be from other Keltic areas, especially Pannonia, Italy and Germany. With a determined leader such as Bathanattos, whose name was remembered as much as that of Brennos, it was possible to form a new and defensible – and aggressive – polity. They must also have been conscious that they themselves had been seriously weakened by their casualties, and may have feared a revenge attack from Macedon – but hardly from Greece, with its divisions. All around they could see the weakness of the fragmented and damaged tribes of the Balkans, and as foreigners in the land, the Galatians no doubt fully realized that only in unity would they be likely to survive. This, of course, was hardly new; they had dominated the region before the Greek expeditions, which in fact were clearly anomalous.
The Scordisci are usually referred to as a tribe, or as a tribal state, but they actually formed a typically Keltic state. Bathanattos is not described as a king, but he was clearly a charismatic leader, giving his name to the road they had travelled along in their retreat from Greece, and fathering a noted family of descendants- either that or the term Bathanatti referred to the group of survivors he had led and their descendants.15
From the names of Scordiscians recorded in written sources and on inscriptions it is clear that the underlying population of the state south of the Danube, the Galatians’ subjects, remained Illyrian or Thracian, with the two being separated by the Morava River, though both were part of the state and the separation was not total.16 There were perhaps other groups in the territory, and, given that the Keltic element in the population was fairly small, it may well be that it was the heterogeneous nature of the population which ensured the evolution of a more integrated state. The Keltic element will have formed the ruling group, no doubt, though it seems that the integration of the various groups was successful, since Illyrian and Thracian names occur in the later ruling group.17
The political status of the Scordisci as a kingdom is not clearly attested until fairly late in its history, and then not certainly. But when the Scordisci were under severe pressure they may have accepted the leadership of one man. This they did in the expeditionary forces of 281–277 under Brennos, then Akichorios, and as did the survivors of that disaster under Bathanattos – but he was not a king. Later, in Asian Galatia, kings emerged, but only under Roman pressure; the Romans wanted a single man that they could hold responsible for order, hence a king. An intermittent kingship is possible, though that may only be the effect of the intermittent nature of the sources. It does not appear, from Athenaios’ notice of him, that Bathanattos founded a royal dynasty; later leaders all seem to be mentioned as coping with an emergency, such as, later, the Dacian invasions; that is, they were leaders who emerged for the emergency only. In the place of a fully-established kingship, one must assume an oligarchic system, probably hereditary (and therefore, more correctly, aristocratic), and in the hands of the Keltic elite – though it seems that the several ethnic groups in the state merged so that a good deal of integration took place over the next two centuries.18 It may be that the aristocracy was internally very competitive, though no internal disputes among the Scordisci are recorded.
The territory they inhabited has been referred to here as the Banat, which is the provincial name of the area around Belgrade in Ottoman times, but which is useful as denominating a clear geographical area. This was centred on the fortress city of Singidunum (a Keltic name), on the site of modern Belgrade, which was at a most advantageous geographical situation. It stands at the junction of the Sava and the Danube rivers, thereby dominating three major river routes, to north, east, and northwest; a little downstream to the east there was the Morava River, along which there was a clear route to the south. Nearly 150 kilometres further north-west, the Drava River joins the Danube, providing still another route, this time into Pannonia. The Sava developed into a major land route towards northern Italy, where for a time in the third and second centuries the Kelts of the Scordisci, and those of Pannonia, had continuing contacts with the Gauls of Cisalpine Gaul. To the north of Italy there also developed the Keltic kingdom of Noricum, which was for a long time a state firmly allied to the Roman Republic.19 In addition to the Sava and Drava routes to the north-west, the Danube provided routes to the north into Pannonia, the Hungarian Plain, and on towards Germany, and also to the north-east, avoiding Dacia by going north of the Carpathian Mountains, and into the valley of the Borysthenes (Bug) River and so on to the Black Sea, the Dniepr Valley, and the Greco-Skythian Bosporan kingdom in the Crimea. To the north was the Tisza River route, parallel for much of the way to the Danube, but closer to Dacia, and to the north-east were the mountains of Dacia, like Noricum a major source of metals. With such wide access to these lands, resources and trade routes, the fortress town at Singidunum was an obvious centre for trade, and so a source of wealth, and it rapidly developed into the Scordiscian capital. A second urban foundation, Taurunum, lay across the Sava from Singidunum, so that, together these two towns controlled the entrance to that river valley from the Danube, and the Danube route itself.
The state founded in the fourth century BC was initially small, perhaps including no more than the Banat area itself. It is known to have expanded along the Sava River Valley later, and the area later called Pannonia by the Romans to the north was already occupied and dominated by European Kelts, who were separate politically from the Scordisci, but were probably friendly, for it was from this population that the original Galatian population of the Scordisci derived, and to which some of the survivors of the Delphi expedition returned; family relationships are to be expected. Expansion east along the Danube valley would be difficult, for the river flows through difficult gorges for some distance, but equally that would provide a helpful barrier against enemies moving upriver. Similarly, expansion southwards along the Morava River into the region called by the Romans Upper Moesia, proved to be difficult also, but for a different reason. For to the south was the Dardanian kingdom, controlling the Scopje basin and dominating the surrounding lands; the hostility between the Scordiscian Galatians and the Dardanians which had begun during the former’s return march northwards, appears to have continued after they were both settled in their new states with new boundaries. The Scordiscian state extended into Illyria, and many Scordiscian names in the first centuries BC and AD were ‘Illyrianised’.20
One of the reasons for the consolidation of the Galatians in the Banat into an organized state, which was distinct from other Keltic states in the area of central Europe, was that they faced the developing Dardanian hostility, which burst into open warfare between the two in 197, in a war in which Philip V of Macedon and the Scordisci were, if not active allies, both hostile to the Dardani (though these had no doubt had earlier confrontations back to the Greek expedition’s return).21 They were compelled to develop a state system in order to resist the expansionary aims of their surrounding rivals and enemies, and to organize their own expansion. (In the south, it was Dardani who were now the inveterate raiders across the Macedonian frontier.)
This was, of course, a condition of persistent hostility, and hostilities are always one of the main factors in studying the formation of a state out of a group of separate peoples or clans. The kingdom of Noricum was formed from several Keltic and non-Keltic elements in part as a result of the development of local wealth derived from mining and the manufacturing of high quality metal goods. Republican Rome was a prime market for such products, and this was one of the main reasons the two states were politically friendly – one produced the goods, the other bought them.22 Such wealth stimulated the envy of enemies, which in turn stimulated the development of a stronger governmental system for defence, including the integration of the several elements of the population, who were all equally threatened by enemies.
The position of the Scordisci at a major trading crossroads had the same effect as the metals of the Noricans, and to hold and develop and defend that trading source of wealth a properly organized state was certainly required. The initial expansion of the Scordisci would seem to have been along the Sava River valley to the north-west of the Banat, either exploiting or contriving the trade route along that valley. The major Balkan crisis which we call the Second Romano-Macedonian War (200–197) involved not merely Macedon and Rome, but the Dardanians as well, and the Scordiscian state also. It was a situation in which a series of enmities coalesced – Macedon’s enemies were Rome and the Dardani, who were therefore allied; the Dardani and the Scordisci were enemies, and so the latter allied with the Macedonians. It was at this crisis that the Scordisci turned decisively southwards to engage in active warfare against the Dardanians, while Philip V campaigned against them and defeated them in battle at Stobi in Paionia. Once such active fighting began, mutual hostility continued as the Scordisci expanded southwards, partly at Dardanian expense.23
Archaeologically, the presence of elements of the ‘La Tène’ Iron Age culture are taken as indicative of Galatian settlement, and there is a strong concentration of such material in the Banat, especially along the valleys of the Danube and the Sava and Drava rivers, and around their confluences. Beyond, to the east, there is another scatter of such finds, much less concentrated, in what became the Roman province of Lower Moesia, south of the Danube and north of the Haemos Mountains.24 There is no sign of such evidence in Thrace (which is odd, given the existence there of the Tylis kingdom), or in the Dardanian territories, which may be explained by their mutual hostility. Dacia, on the other hand, received a considerable migration of Kelts, probably as individuals – assuming that the remains indicate the arrival of people, not just goods acquired in trade. They were concentrated into the metal-rich area of the north of Dacia, the territory which later formed the heartland of the Dacian kingdom, which suggests that the Keltic immigrants might have been professional miners and metalworkers – a link between developments in Dacia and Noricum is obviously likely.
The Scordisci, therefore, by the middle of the third century BC, after a century and a half of occupying the Banat and its contributory river valleys, was a serious power in the region, probably the most vigorous and expansionist state in the Balkans, and so one at enmity with all its neighbours, particularly those to the south and the east. This was an expanding Galatian state, willing to accept other ethnic groups into its population as well as any Kelts. It was clearly Keltic in origin, but developed as a multi-ethnic state, and as such settled confidently into the existence as one more of the ‘tribal states’ of the region. It was, like every contemporary state, willing to seize on the weakness of any neighbour, and to ally with any other state if it brought advantage.
The Kingdom of Tylis
The other new Galatian state in the Balkans, the daughter, in a way, to the Scordisci, was the kingdom of Tylis, established in Thrace. It was founded in the immediate aftermath of the retreat from Greece, beginning with the settlement of a Galatian group under a leader called Kommontorios, referred to as their king,25 a most unusual distinction amongst the Galatians, though perhaps it meant no more than their commander and leader – and yet it was always referred to as a kingdom in our sources. Part of the group was composed of survivors of the Greek expedition who had broken away from the main group while it was heading for home, and part was another group which had left their Danubian homeland during the return of the defeated army, perhaps aiming originally to take part in the great raid. They had instead moved into Thrace, clearly searching for a new base, and intent on creating a new kingdom, possibly part of the expedition led by Kerithrios, men who had already prospected the possibilities in Thrace.26 The conditions in Thrace were evidently very confused, and it seems there were other autonomous Galatian warbands at large apart from those of Kommontorios, Leonnorios and Loutarios. One of these bands was defeated and massacred by Antigonos Gonatas in his ambush of a Galatian force at Lysimacheia in 277, an event which no doubt compelled a greater concentration by the rest.27
Antigonos recruited as mercenaries other Galatian bands who were present in Macedon; they proved to be more loyal to him than the Macedonians, but then Galatian mercenaries in Macedon must have led precarious lives, while the Macedonians were considering him as the latest candidate as their king, not just their employer. Being employed by the king in establishing a renewed kingdom was much the same work for the mercenaries as the other bands were doing in the Banat with Bathanattos, or Kommontorios’ men in Thrace. And, of course, they ceased to be so much of a public danger when under some degree of discipline as part of the royal army, not to mention that Antigonos used them up in his battles, thereby getting rid of them, protecting his own Macedonians, and saving on their pay, all at the same time.
The majority of Kommontorios’ men, perhaps under some pressure from their by-now numerous enemies in Thrace, consolidated into a single group with the latecomers from the Danubian area and established themselves as a kingdom which became called Tylis. Antigonos and Kommontorios absorbed or killed off the stray wandering Galatian bands, while others moved off into Asia. And all this happened within at the most a couple of years of the return of the great raid from Greece.
The kingdom of Tylis has been perhaps the most mysterious of the Galatian states founded in this period of the invasions of the East. We know something of its origin, as noted in the last paragraph, and something of its destruction, but little or nothing of its history between these events; nor is its location altogether clear, though it is known to have been in Thrace. As a political entity it lasted only a little over sixty years.28
The condition of Thrace in 277 was no doubt both confused and confusing. The native Thracian population had been battered by half a century of warfare against Macedonian rulers, and now had to cope with the wandering Galatian bands. Of the Galatian forces in the region, two groups had separated from the greater raiding force and moved eastwards under the command of Leonnorios and Loutarios. This group was evidently composed of men dissatisfied with the results of the Greek/Macedonian expeditions, and, being accompanied by their families, had always intended to continue raiding and invading new lands until they were able to settle as a community somewhere. The second group, commanded by Kommontorios, seems, at least by their actions, to have intended from the start to secure territory for themselves in Thrace, and perhaps raiding for loot had always been a secondary activity. They arrived and planted themselves without delay, and did not move. This suggests that at least some of the leaders were already familiar with Thrace, and had arrived with a clear plan of action – this, of course, would fit well with the obvious planning by Brennos and his colleagues and by Bathanattos, evidently a Galatian characteristic.
These two groups remained quite separate and evidently had quite different aims. It would, however, be naive to pretend that they did not meet and mingle. It may be assumed that some broke away from each of the groups to join the other, or even completely away from both. They both spent time in eastern Thrace before reaching decisions on their final intentions. Quite probably the firm and early decision of Kommontorios’ group to settle in Thrace was made clear to the others, and that the Leonnorios/Loutarios group was therefore not welcome. There, after a time, it is likely that both suffered defections, and their numbers were reduced.
Several influences and actions contributed to this reduction. There were the inevitable casualties. There may not have been many wounded surviving from the Greek expedition – the Greek sources claim that before leaving central Greece the wounded from the fighting at Thermopylai and Delphi had been killed29 – but one would suppose that the march along the Thracian coast had not been unopposed, and that further deaths had occurred; there were a series of Greek cities along that coast, but although there is no indication that any fighting took place, that is no guarantee that all was peaceful. The number of people in the Leonnorios/Loutarios group is put at 20,000,30 and while this is as unacceptable as any number from the ancient world, it may be in the neighbourhood of probability. But whatever the number when the group split from Akichorios’ main body, it will have been considerably lower by the time they reached Thrace; 20,000 (which included the families) is probably their approximate number when they crossed into Asia.
In Thrace, further problems chipped away at the size of the forces. The ambush of Antigonos at Lysimacheia is said to have caused the deaths of ‘18000’ Galatians,31 while Antigonos’ subsequent recruiting of Galatian soldiers as mercenaries was probably at the expense, at least partly, of these groups. And finally, the two commanders were recruited, with their whole warband, by the king of Bithynia and transported across the Bosporos into Asia. Altogether it is likely that the whole of the Leonnorios/Loutarios warband can be roughly accounted for by these various actions, moves, treacheries, and calamities – though it is surely probable that some also remained in Thrace, while others will have joined with them. The force which Antigonos massacred was another, it seems, which had broken away from the breakaway forces, and these directionless groups may have been gradually disintegrating ever since they arrived in Thrace. At least one other substantial group was the Tektosages, who followed the Asian groups across the Bosporos; they must have been in Thrace before they moved east, but nothing is known of them at that time
The force under Kommontorios may or may not have joined with that under Leonnorios and Loutarios at some point, but he did not join in the Asian expedition, nor does he seem to have been involved in the Macedonian expedition earlier, nor in the fighting against Antigonos. Instead all the indications are that he immediately established his force in Thrace and organized them as the kingdom of Tylis.
In the immediate aftermath of the failure of the Greek campaign, therefore, apart from the main group retreating north with Bathanattos, there were at least five substantial Keltic forces in Thrace: those led by Leonnorios and Loukarios; the Tektosages; that under Kommontorios, which may have been in two parts; and that in the Chersonese which was caught and destroyed by Antigonos. Within a short time they all either moved away or consolidated into a kingdom – or, of course, died.
We may see Kommontorios, then, as one of those men who saw an opportunity and worked determinedly to realize it. Leonnorios and Loutarios by contrast seem to have been more adventurous, seeking new opportunities to gather loot, and enjoying the excitement of raiding and fighting before searching for their new homes; they may also have been under pressure from their people to continue moving and raiding. Certainly Kommontorios succeeded in establishing his little kingdom, and probably, given the dire condition of Thrace at the time, this took place fairly soon after his force’s arrived there.
Where exactly the kingdom was, however, is unclear. It used to be thought to be located in the Haemos Mountains, then it was suggested that it was in the valley of the Thundza River.32 Both of these are in fact in much the same area as Seuthes III’s defunct kingdom (and where Kambaules’ raid in 298 came to grief), and would have included the wrecked site of Seuthopolis, where, to be sure, there is some archaeological evidence for a Galatian presence.33 A more detailed consideration of the distribution of Keltic artefacts in Thrace, combined with some details in the written sources, now suggests that the kingdom was closer to the coast, in the area inland of the Gulf of Burgas, and close to the Greek cities of Apollonia Pontica, Mesambria, and Odessos.34 It is thought in one theory that the kingdom controlled the former Thracian/Macedonian city of Kabyle at one point, but the evidence is not strong, consisting, as at Seuthopolis, of the finding of a few Keltic-type artefacts, which do not necessarily imply more than trade.35
The clue to the kingdom’s location lies in appreciating that these Galatians kept themselves aloof from the Thracian population they dominated.36 In order to do so these Galatians retained to a degree their former predatory habits, modified into a blackmail regime. This was certainly their relationship with the city of Byzantion. Perhaps, however, the other coastal cities did not suffer from the same attentions. Byzantion was some way away from the centre of the Tylis kingdom, whereas the other cities were geographically contiguous to its territory and appear to have had settled into a more or less comfortable relationship based on protection – for which they would pay, of course – and trade; indeed, under the circumstances they had no choice. Byzantion, on the other hand, had to be periodically menaced by expeditions which ravaged its mainland territory as an instance of what the kings could do if denied their reward. Polybios notes that a series of tributary payments were made by the city, beginning with one to Kommontorios in the 270s, amounting to 3000 gold pieces; later demands were for 5000, then 10,000, and finally a demand for an annual payment of eighty talents. The successive rises in the demands, were, of course, the direct result of paying the blackmail in the first place. How many times the demand came in is not known, but the first was in the 270s and the last in the 220s. The final demand, the one for eighty talents annually, was the one which finally brought Byzantion to resist, and appears to have been made in 221 or 220 BC.37
The cities closer to, or enveloped by, the kingdom – Apollonia Pontica, Mesambria and Odessos – probably did not need to be blackmailed; they no doubt could simply be taxed regularly – as indeed was Byzantion, in its particular way. The same happened no doubt to the Thracian peasantry in the country around Tylis, which is thought to be the Galatians’ own city, or perhaps camp, but has not been located, other than by assuming that it was at one of several places whose modern names might somehow resemble it. The fact that deliberate expeditions had to be made to levy blackmail on Byzantion implies that, unlike Apollonia and the other cities, Byzantion was not close to the Tylis kingdom, and was increasingly reluctant to pay. These considerations, together with some archaeological evidence and some place-name evidence, suggest that the territory of the kingdom was close to the smaller Greek cities, and that it stretched inland from the coast at the Gulf of Burgas for perhaps forty or fifty kilometres, the territory being an approximate semi-circle with its centre perhaps at the head of the bay. The city of Kabyle, which is thought to have formed part of the kingdom for a time, is about fifty kilometres from the coast, and securing control of it may well have been an attempt by one of the kings to extend his power further inland; Byzantion was rather further away, and was a stronger city than any of the others.
This state was therefore different from all other Galatian states in the east. First, it appears from the origin to be ruled by a king. This is, as already noted, highly unusual for a Galatian state, but the evidence seems conclusive, in that when it became involved with a major crisis near the end of its life, the ruler was always described as a king. On the other hand, this may be because the Greek sources we have could not think of any other way to describe the ruler. But the state was also run on different lines from others: it does not seem to have integrated its Galatian and Thracian populations, but to have kept them separate, the Thracians, of course, being subordinate. It kept its predatory habit, to the extent that it prayed regularly on the local Greek cities. All this – king, racial separation, predation – makes it out as a singular and distinctly different state and society.
These characteristics made Tylis an insufferable neighbour and ruler, and its methods will have built up considerable resentment both in the Thracians who were in effect it serfs (and probably subject to enslavement and sale) and among the Greek cities. If Byzantion’s experience is typical, the taxation regime applied to the other cities also became increasingly unendurable. The end was a crisis which first broke the predatory system and finally eliminated the kingdom.
When the crisis of the payments by Byzantion came in 221 or 220, Tylis was ruled by a king called Kavaros. He was perhaps king from the 230s, though the exact date of his accession is not accurately known. We may assume he had ruled for some time before 220; his eventual intervention in the dispute suggests a canny politician and a king fully familiar with Greek diplomatic methods. The institution of kingship may well have been part of the original Galatian settlement of Tylis, which would make Kommontorios the first of the kings. Kingship, as opposed to chieftainship, did not go well with a community living by predatory raids, in which individual chieftains used their power to distribute wealth to their clients; these were a more suitable means of authority. And if Tylis was a state in which the ruling group, the Galatians, remained aloof from their Thracian subjects, it would be more likely that a king existed, if only to centralize the menacing and taxation powers which would maintain Galatian control; the demands of the king’s clients and the other chiefs would not remain steady, but would necessarily increase – hence the steadily greater demands on Byzantion, and probably on the other cities.
We can therefore infer something of the kingdom’s origin, of its social structure, of its political organization, and of its territorial extent. We know the king regularly levied tribute-blackmail on at least one rich Greek city, and no doubt enforced tribute payments, collected equally regularly and greedily, on the Thracians within his reach, and on the neighbouring Greek cities on the coast. On the other hand, it was not expansionist, unlike the Scordiscian state for example. It also appears that Kavaros’ tribute will have been one of the major factors which led to the collapse of the state. It was, therefore, from the start a predator state. This would make it vulnerable to internal disputes over the division of the tax-take, and to any stronger power whom it alienated; it was, in short, inherently unstable.
The payment demanded by Kavaros from Byzantion in about 220 was for eighty talents. It is not altogether clear how long this level of payment had been so high, but for a time Byzantion clearly paid at something like this level. In 220/221, however, the city finally baulked at the demand. The eighty talents has been calculated to be the equivalent of 24,000 gold staters (eight times the original first payment)38 and the city decided this could not be met out of Byzantion’s own treasury. The alternatives were to refuse, thereby incurring a war in which the city’s territory, much of which will have been owned by the city’s oligarchs, was seriously damaged, or to shift the burden elsewhere; the city chose the second. The city government resolved to impose a transit tax, or toll, on ships passing through the Bosporos, thereby attempting to shift the weight of the tribute onto the Greek mercantile world as a whole. Complaints ensued. There is no real suggestion that the tax was exorbitant or unaffordable, but it would certainly increase shipping costs, and raise the prices in the cities to which the ships were taking the goods. In effect, as has been pointed out, Kavaros was compelling Byzantion to become his tax collector, and the merchants and shippers were expected to pay Kavaros’ tribute, and therefore their customers; Kavaros was laying the whole Greek world under tribute.39
The complaints no doubt began with the merchants and ships’ captains, but they were quickly taken up by their home cities. The cities appealed to Rhodes, currently the greatest naval and mercantile city in the Aegean region, to intervene – no doubt Rhodes will have been one of the original complainers as well. The Byzantines refused the demand from Rhodes to cancel the tax, and war followed, in which King Prusias of Bithynia joined, aiming to take advantage of Byzantion’s plight to filch some of the city’s territory on the Asian side of the Bosporos. The Rhodian tactics were, however, less violent, though perhaps more effective; they instituted a blockade of Byzantion, which no doubt meant that the passing ships would be protected from the Byzantine exactions, but also that they were prevented from calling at Byzantion. So Byzantion was both unable to collect the tax, and was deprived of the trade (and the customs duties) which was brought by the ships under normal conditions; further, the city was losing control of some of its territory. It could find no sympathizers. It was helped by Herakleia Pontike with cash, but by no one else – Herakleia was constantly at odds with the Bithynian kings; Byzantion’s allies, King Attalos of Pergamon and the Achaian League, gave it no real support.
Byzantion was quickly forced to ask for terms. Ironically, or perhaps deliberately, it was King Kavaros who arranged the terms. He was probably feeling the pinch as much as Byzantion, since Byzantion was clearly not paying up, and would not be able to pay while the blockade continued. By the treaty Kavaros arranged that the city would give up collecting the tax, and would become the ally of both Prusias and Rhodes; Prusias returned the territory he had taken.40 Nothing appears to have been said about Kavaros’ original demand for the tribute, and it may have been abandoned, or more likely reduced. Alternatively, Kavaros may have simply agreed to collect it over a longer period, which in fact will have put Byzantion in a similar tributary position to the other Greek cities already overseen by Tylis. In any case it was clearly impossible to extract the full sum from Byzantion at once, and the result of the dispute was a modified victory for the city.
These characteristics made Tylis an insufferable neighbour and ruler, and its methods will have built up considerable resentment both in the Thracians who were in effect it serfs (and probably subject to enslavement and sale) and among the Greek cities. If Byzantion’s experience is typical, the taxation regime applied to the other cities also became increasingly unendurable. The end was a crisis which first broke the predatory system and finally eliminated the kingdom.
Kavaros was, therefore, one way or another, deprived of a handsome sum of money. He was clearly greedy for it, for whatever reason. Possibly he had aimed to use it in wars in the Thracian interior, possibly he simply wanted to accumulate a large sum of money; the very large sums he had already collected had perhaps been distributed among the several chiefs of his kingdom, or used to pay mercenaries. If Tylis kept a fairly rigid separation between the elite and their Thracian subjects, its armed forces would necessarily be the followers of the elite plus any hired mercenaries Kavaros could recruit; the Thracians were no doubt not trusted and were kept disarmed. (One thinks of Spartan helots as an example, or serfs in many other parts of the Greek world, including Bithynia.) This would also mean that power within the kingdom was distributed among the Galatian chiefs, and Kavaros as king would be only primus inter pares, though no doubt richer than the others. Attempting to increase the tribute might have been a way for Kavaros to increase his internal power. The alternative means of getting money would be for him to impose greater demands on his own subjects, that is, the Thracians. The Galatians of Tylis, if they really were an oligarchic community whose contribution to the state was their military abilities and their armed followers, would not be taxed, but would expect to receive shares of the king’s exactions; in other words, they were living on the tribute of the Thracians, and Kavaros’ demands on Byzantion were probably fuelled by the expectations of handouts to his Galatians.
A new demand may have been the blow which finally brought the Thracians to rise against Galatian domination. The failure of Kavaros’ scheme to mulct Byzantion will perhaps have made him impose these new demands. The details of the overthrow, as ever, escape us, but it does appear that it was a Thracian attack or revolt which brought about the kingdom’s destruction.41 Exactly which Thracians were involved, however, is not known. They could have been the Galatians’ subjects, driven to desperation, but it could also be that Thracians from outside the kingdom felt increasingly threatened by the wealth being gathered by Kavaros; the move against Kabyle might have suggested that the kingdom was looking to expand. There was also a Thracian kingdom, ruled by a scion of the old Odrysian royal family, which may have inspired the Galatians’ subjects to rebel, either passively, by merely existing, or actively, by intrigues. There was certainly a powerful Thracian kingdom in Thrace in the years before about 200, which was able to capture and sack at least one Greek city, Lysimacheia.42 One might also suggest the possibility that the Byzantines, frightened by their experience of the demands of Kavaros and the war they provoked, might well have instituted intrigues themselves to secure his overthrow, though it would be surprising if they had not resorted to such tactics earlier.
The result was not only the removal of Kavaros, but the complete destruction of the Tylis kingdom, apparently sometime after 220. This speedy end was assisted by the intrigues of a man called Sostratos of Kalchedon, who became a favourite of the king. No doubt his influence was resented, and contributed to Kavalos’ end.43 The failure of his extortion campaign against Byzantion was also presumably an element in the kingdom’s collapse.
The kingdom vanished without leaving a trace of its existence. The site of Tylis is unknown, the names of most of its kings are unknown, and the extent of the kingdom’s territory can only be estimated. The fate of the individual Galatians was probably to be killed – it is likely that there were not many of them; otherwise, they might escape to refuge with the Scordisci or into the Galatian communities in Asia. That is another thing we do not know.
Settlers Eastwards
There is one further region into which Galatians moved in this period of confusion, migration, raids, and conquest. North of the Black Sea, in the steppe lands of the Ukraine, east of the Carpathian Mountains, there is some archaeological evidence of a scattered Galatian population. This is spread fairly widely, from Moldova to the lands north of the Sea of Azov. Little information of this is recorded in any of the written sources, though one inscription is helpful and telling, so it is from archaeology that most of the information derives.44 This has been noted already, but here more detail is appropriate.
The migrants came from the Pannonian settlers, and perhaps from Dacia, Bohemia, and Moravia. The route to the east led them to the great rivers flowing through the Ukraine south to the Black Sea, the Dniestr, the Bug, the Dniepr, and the lands along the north of the sea; at the estuaries of these cities there were a series of old Greek cities, founded 200 or 300 years before; Tyras at the mouth of the Dniestr, Olbia at the joint estuary of the Bug and the Dniepr, Pantikapaion at the Strait of Kerch between the Crimea and the mainland, and others. The Galatian migrants arrived at a time when the steppe was relatively quiet, with no great nomad conquerors rampaging. Not that there were no great movements at the time – their arrival is an indication of some restlessness, however. The evidence of their presence is scattered but extensive, though dating is difficult – since it was only by about 400 BC that a large Galatian population was firmly located in Pannonia, it must have been some years, or decades, after that date that the Ukraine Galatians arrived.
The archaeological evidence consists in part of constructed earthworks which are thought or presumed to be of Keltic construction, of burials identified as specifically Keltic, and of isolated finds of La Tène-type metalwork, which might imply the presence of Keltic workers, who may have been itinerant.45 There is a concentration of Keltic materials in the upper Dniepr region, and a string of forts-cum-earthworks spread at approximately even distances from the upper Dniestr to the Dniepr. There is another concentration, this time mainly of burials and La Tène objects, in the middle of the Dniepr valley, in the area where Kyiv was later founded. The earthworks are largely in the open steppe, arguing a defensive purpose, but also close to or on the rivers;46 the heavier concentrations of apparent occupation are in the zone of mixed steppe and forest north of the steppelands.47
There are also scattered Keltic place names and tribal names in this region, at least in the western part. The Costobogi (a similar name to the Tolistobogii in Asia) are recorded in the central area of modern Moldova, and the Britolagi have been located just north of the Danube estuary; on opposite sides of that river, above the estuary, are the Keltic place names Allobrix and Noviodunum. Such place names and the fortifications do suggest a permanent Keltic presence, whereas finds of jewellery, or even single burials, might be no more than the evidence of occasional visits or passing groups.
Exactly what all this amounts to is therefore less than clear. Many of the metalwork finds are in graves which are clearly not of Keltic people, but were probably Skythians – but the Skythians had been displaced by the Sarmatians during this time, though neither of these was more than an elite of warriors ruling over the same indigenous population. An earthwork, whether amounting to a fort or not, is not necessarily a permanent or long-lasting structure indicating the presence of inhabitants, but could be just a temporary camp. The concentration of a mixture of items at the upper Dniestr and in the area of Kyiv, however, are sufficient to imply the arrival of settlers in those places, living among the native populations, where their skills – in warfare, in metalworking – were welcome. Tribal names imply a ruling group, place names imply permanent settlers, and these imply the existence of organized Keltic states on the usual ‘tribal’ pattern. It is also possible that these groups arrived not from Pannonia, though this was the area of the heaviest nearby Galatian population, but derived from the Gauls who had moved into Poland; the place names by the Danube estuary could signal Kelts coming along the river.
The one worthwhile written source in this first period is an inscription from Olbia, at the mouth of the Bug.48 The city had come under attack by an allied force of Kelts and Scirians. This looks very like the same sort of attacks made in Greece and Thrace and Asia Minor, if perhaps with a considerably smaller group of Kelts, if they were allied with others. The attackers encouraged the serfs working in the fields for the citizens to abscond into some sort of freedom, or kidnapped them. The attackers could have been based in some of the earthworks in the interior north of Olbia, or from one or more of the Keltic states in the east. On the whole, however, neither the settlements nor the raids seem to have had much effect on the local political or social situation in the lands north of the Black Sea. The presence of any Kelts was no doubt swallowed up by the indigenous inhabitants.
The eastward move of the Kelts produced a curious variety of results. The Scordisci state was an important local power in the Balkans for three centuries; the Tylis kingdom, a predatory state, went down to ruin after only sixty years; those Kelts who penetrated into the Ukraine were, like most steppe invaders or migrants, absorbed into the local population with scarcely any record. But the three were all different. The Scordisci was an expanding state typical of the Hellenistic period, only the Keltic origin of its rulers distinguishing it from other states all around the Mediterranean. The Tylis kingdom emphasized a different element in the heritage of the Kelts, the practice of raiding, which became institutionalized as predation on its neighbours; it was ended by those neighbours when the predations became too greedy or violent or both. In the Ukraine the migration was almost stateless, possibly predacious, possibly peaceful, perhaps both, but was far too weak in numbers to make much of a mark.