Chapter 10
The defeat of King Perseus of Macedon in 167 by the Romans to end the final Romano-Macedonian War also destroyed the Macedonian kingdom and its system of royal government. The Roman settlement abolished the monarchy, and replaced it with four republics. These proved to be weak and quarrelsome, and incapable in the event of defending the Macedonian territory, a matter which had always required the resources of the entire united country led by the king. In 148 the pretender Andriskos succeeded, with Thracian help, in overthrowing the republics and reinstating the monarchy, calling himself Philip VI. This regime lasted only a few months, but in that time he also defeated a Roman army and killed its commander, P. Iuventius Thalma, and then had to be himself defeated by a consular force under Q. Caecilius Metellus. The whole affair caused a great shock in both Greece and Italy.1
A continuous Roman military presence in Macedon, which had not been thought necessary under the four republics, now followed. It was probably not originally intended to be continuous, but the frontier remained disturbed for several years, with copycat pretenders – an Alexander, another Philip, a Perseus – all of whom gathered barbarian support in the Balkans and had to be fought and defeated.2 Andriskos’ invasion had at first seemed like a barbarian attack of the traditional sort, with his use of Thracian forces, and the later pretenders worked in the same way, though using different barbarian auxiliaries. None of these invasions and threats, of course, convinced the Macedonians to support them, and Andriskos/Philip was not seriously given any help in the end. As a result of the continuing threat, which lasted eventually for seven years, the Roman military presence became in effect permanent, and Roman commanders, usually of praetorian rank, were appointed annually.3
Rome had become involved in wars in Dalmatia and Illyria in the 150s, and in 156 a Roman army penetrated as far as Siscia in the valley of the Sava River, where it defeated a Scordiscian force in an obscure battle.4 This was at, or close to, the border of Scordiscian territory. It is tempting to link the two events, at Siscia and in Macedon, though they are really too far apart in time and space for any real connection. As far as the Scordisci were concerned they had no further repercussions for the moment. But in 141 a new raid against Macedon by a Scordiscian force was met by a Roman force and won the victory, though there is no further record of their actions on this occasion.5
Exactly what troops Rome maintained in Macedon is not quite clear. It is usually put at a legion, which is likely enough since this is the minimum Roman military formation which was usually employed, but this was not necessarily the whole force available to the Roman commander; and a legion would hardly be enough to defeat an invasion by the manpower of a whole Balkan tribe, who could usually muster perhaps 10,000 fighting men for an invasion. The Roman force, if it was kept concentrated, was in effect the reserve force backing up the local Macedonians, who would always turn out to combat a barbarian invasion. There were also forts, camps and posts at the frontier which were occupied by local forces, no doubt manned by rotating units, and it is probably with them that the first responsibility for the defence of the frontier lay. (This was, of course, the pattern of Vulso’s army in the attacks on Mounts Olympos and Magada – put the allies/auxiliaries in the front of the battle, saving the Roman citizen manpower.) The Macedonian forces had in fact failed to block the invasion of Andriskos’ forces in 148, undermined by a lack of organization in the rear and by the propaganda of a return to a monarchy put out by Andriskos. Now they were backed up by a Roman legion, so they were more likely to stand firm.
A second collision with the Scordisci came in 135, specifically in Thrace, and this time it was the Roman forces which scored a victory, under the command of C. Cosconius.6 Again there seems to have been no follow-up; a defensive victory was probably sufficient to ensure the safety of Macedon; the fight is said to have been in Thrace, so Cosconius had certainly already moved out of Macedon, his province, perhaps in order to intercept a raiding party, or just possibly in alliance with the local Thracians.
The Scordisci operating in Thrace were not necessarily aiming to attack Macedon. It is possible that they were attempting to enforce their suzerainty over a variety of neighbours, Illyrians, Dalmatians, Thracians, and any others within reach – or possibly establish that suzerainty – and had assumed that the Romans would confine themselves to Macedon. They would understand enough about Roman attitudes and practices to know that launching a war against a Roman-occupied and defended region would be a major undertaking, not to be begun lightly – though the 140s and 130s would be a good time to try something, with Roman forces heavily engaged in Africa, Spain, and Asia at various times and occasionally all at the same time.
At some point in the period after 141, probably in the 130s, the praetor Cn. Egnatius spent three years organizing a new imperial Roman road, called the via Egnatia after him.7 This was not so much a matter of physical construction, more a matter of occasional improvement, and mainly of marking the route by planting milestones with Latin inscriptions indicating distances; the aim of the road was to assist Roman military forces marching through on the way to somewhere else, hence the Latin. None of this would necessarily be of much local utility, for the locals already knew the way and the distances, and Latin would be of little use to them. The road already existed, of course as a traditional routeway, though perhaps only in local sections, but one can scarcely expect that the Macedonian kings would have ignored the need for roads leading to their dangerous frontiers, and especially this route, which facilitated travel behind the frontier line. Egnatius’ work, therefore, was in part the designation of an imperial Roman road, and the result was a very obvious and permanent indication of the permanent presence of Rome in Macedon. Militarily it would provide an easy and well-known road all along behind the frontier, a means by which any Roman force could march quickly to the source of any trouble. (A few years later M. Aquillius in Asia Minor did the same, so that by the 120s there was a designated Roman road all the way from the Adriatic to southern Asia Minor.)
The ability of the Scordisci to campaign as far south as the northern border of Macedon in 141 and 135 presupposes the subjugation of, or possibly alliance with, the peoples of the territory between that frontier and the Scordiscian homeland in the Belgrade area and the Banat. In support of this it is noticeable that the name of the Dardanians disappear from all events from the 180s, though they were not destroyed as a people, for they re-emerged in the 90s; that is, they appear to have become part of the Scordiscian polity, no doubt by conquest and reduction to vassal status.8 This would imply an extensive Scordiscian ‘empire’ stretching from the Macedonian frontier to the Drava-Danube junction, and from the Danube gorges to the site of Siscia, where they had clashed with another Roman force.
In 119 a new bout of Romano-Scordiscian warfare began, with another Roman attack on Siscia, which was apparently unsuccessful.9 The cause of the war is not known, but the two states were neighbours and no doubt there were plenty of problems and hostile incidents which could be magnified if one or other wanted to make war. Both Romans and Scordisci were expansionist; the Romans in particular had a horror of any powerful state on their borders, paranoiacally always imagining an imminent attack. The result was a lengthy period of warfare. The attack on Siscia appears to have triggered a Scordiscian response with another attack on Macedon, a response the Romans perhaps did not expect. In alliance with at least one Thracian group, the Maedi, the Scordisci invaded Macedon by way of the town of Argos in Orestis, an interesting way of approaching the province, coming by way of the routes through the mountains which had earlier been utilized by the Dardani.
The praetor Sex. Pompeius was killed in the fighting, and command was taken over by M. Annius, his quaestor. From the account given in a decree from the town of Lete in Macedon, honouring Annius, it appears that he found the Scordiscian forces scattered after their victory, recovered the dead at the battlefield, and then defeated a new Scordiscian attack, presumably by the re-concentrated invaders. This was then followed by a victory over the Maedi, who had chosen the moment of the original Roman defeat to join in. It may be assumed that some of the fighting took place near Lete, which in turn was near Thessaloniki, for it was there that the thanksgiving inscription was erected – so the invasion had penetrated deep into Macedon. No doubt the unusual approach route created surprise. Annius certainly deserved his award of an olive wreath and commemoration in the equestrian games at Lete.10
It is also clear that the ‘Gallic nation’, as the inscription identified the Scordisci, was a largely mounted force. Annius several times captured horses and horse gear, and he refused to call up the local forces, giving as his reason his not wishing to burden the Macedonian cities with the soldiers’ pay, but perhaps he also understood that a large infantry force would not be an appropriate counter to the Gallic cavalry; the Scordiscians scattering after the battle would be easier for cavalry than for infantry.
This fight and invasion in 119 was the beginning of a period of warfare lasting at least a decade. The Scordiscian territory and their area of suzerainty evidently covered a large part of the Balkans, from Siscia to the Macedonian border and into Thrace.11 They were thus, as Pompeius discovered, a formidable power, and were also allied with the Thracian Maedi, who clearly in 119 acted independently. The Maedi also joined in the fighting in the next years, so alliance is perhaps the best word for their relationship, in that they continued to act separately but in accordance with the Scordisci.
To their north the Scordisci were in contact with the Boii of north Pannonia and Bohemia, with the Taurisci of the north-west Balkan area, and beyond that with Noricum and the Helvetii. When the tribes of the Cimbri and Teutones on their wanderings throughout Western Europe bumped into the Boii and were repulsed, they then met, in succession, these three Keltic states of the middle Danube, probably without fighting them, but certainly not being welcomed, and then they headed west into Gaul, all the time searching for a new homeland, and no doubt being repeatedly discouraged; they seem to have avoided the Norican kingdom.12 The threat they posed was, however, as Rome discovered, powerful, and it may have helped to compel the Dacians to unite and to begin menacing their western neighbour, the Scordisci – whose own power may also have been one of the reasons forcing the Dacians to unite.
In the years immediately after the defeat of Pompeius and the expulsion of the Scordisci from Macedonia by Annius, there are records of several periods of fighting to the north of Macedon between the Scordisci and Roman forces. Details are absent in most cases, but it is known that there were battles, which presuppose both preliminary and succeeding campaigns, in 114, 113, 112 and 109.
These campaigns may well have been preceded, during the period between 119 and 114, by earlier campaigns (records are poor). In 114 the danger was sufficiently apparent in Rome to require the appointment of a consul, C. Porcius Cato, to the Macedonian command, possibly with increased forces, in place of the usual praetor. That the apprehension was correct is demonstrated by Cato’s defeat in the subsequent fighting. This permitted another Keltic invasion of Macedon and Greece which reached as far as Delphi,13 and next year the wandering Cimbri defeated another Roman consul, Papirius Carbo, in Illyria, also in the Balkan area.14 The Scordisci attacked again, or were attacked, in 112–111. The consul of 113, C. Caecilius Metellus Caprianus, and that of 112, M. Livius Drusus, both celebrated triumphs over the Scordisci, but only the latter is credited with a serious victory.15 The final fighting took place probably in 110 or 109, when the Roman forces under the consul (for 110), M. Minucius Rufus, defeated a Scordiscian invasion of Thrace – that is, Roman forces were again operating to the north of the Macedonian frontier. This took place at some point in his extended term of office as consul and proconsul, from 110 to 106.16
This was a major conflict, lasting at least a decade and probably more, but one which in the Roman sources is largely overshadowed by the contemporary wars in North Africa against Jugurtha, and against the Cimbri and Teutones, not to mention the ongoing internal political crisis within the city. But the Scordisci clearly mounted a major threat to the Macedonian frontier, and penetrated through the defences at least twice, in 119, and most spectacularly, in 114, when they reached Delphi. The reaction of the Greeks to this renewed Keltic spoliation of their major sanctuary was much less vehement than it was in 278, no doubt because of their consciousness of their military weakness, and their own inability to mount any sort of defence; Minucius Rufus, whose victory seems to have finally brought the war to an end, at least for the moment, was honoured with a statue at Delphi, or rather perhaps claimed the credit for the victory. The inscription lists his enemies as Gauls, Scordisci, Bessi, and ‘the remaining Thracians’, a catalogue amounting to the whole of the central and south-eastern Balkan peoples.17 The attack was clearly a major invasion.
The record of this war, thin and fragmentary as it is, does reveal that the Scordisci were the dominating power in the southern Balkans at the time. They had subdued the Dardanians of the Scopje basin, and this gave them access to the Macedonian frontier. They were also allied, it seems, with the Maedi in Thrace, again a neighbour of Macedon. To the north their territory extended along the Sava River valley at least as far as Siscia which they controlled in 156 and 119; along the Danube northwards they appear to have controlled, or dominated, the river and its banks as far as a boundary with the Boii, which was at some indeterminate distance north of the junction of the Drava and the Danube – if Siscia is a guide, the border was probably about a hundred kilometres beyond the junction. In this territory not only the Dardanians were under Scordiscian control, but also most of the Pannonians, who lived in the Sava valley, south of the later Roman province which was named after them; other Pannonians were part of the Tauriscian and Boiian states.
This was an extensive territory. To the west the Romans were in control of most of the Adriatic coast, and were allied with the Keltic kingdom of Noricum; there were probably a series of small independent groups between the Roman and the Scordiscian territories, which were useful buffers between them, and likewise were their occasional victims. The Romans also controlled Macedon to the south, and so, apart from the small ‘buffer’ tribes, the Scordisci had Roman territory as their neighbour all the way around from Thrace to the northern Balkans, and the two were in direct contact in the south and in the Save Valley, and perhaps elsewhere. The geography may explain the Scordiscian ability to exert pressure on the Romans at several points, using their control of internal routes – notably at Siscia and on the Macedonian frontier almost simultaneously in 119. They had other Keltic states to the north, the Boii, and the north-west, the Taurisci, with which they were at least friendly and possibly even allied – there is never any suggestion of conflict with these states. To the east were the Dacians, who by the end of the first century BC were becoming increasingly organized and powerful.
The domination which the Scordisci exercised over their subject tribes was somewhat tenuous. The Dardani, the Pannoni, and the other tribes under them – the Moesi and probably the Andizetes of the lower Sava and Drava valleys, for example – were not disturbed in their homelands, but were subjected by Scordiscian control in some way, and were no doubt compelled to pay tribute and to supply troops on demand. It was therefore not too difficult for them to detach themselves from the Scordisci when they got the chance – and it is by their later appearance in the records that we know of them. The Romano-Scordiscian conflict which continued from 119 (or before) to about 106 quite possibly resulted in a decisive weakening of the Scordiscian state and the dissolution of its empire. It would certainly be a sensible political move by the Roman commanders to encourage independence among the Scordiscian subjects. This Scordiscian ‘empire’ essentially consisted of the domination of a relatively small Keltic-Illyrian group which had become the Scordisci over its tribal neighbours, and the Scordisci may well have suffered severe casualties in the Roman wars. The account of the fighting preserved in the inscription from Lete describes the Scordisci as a cavalry army, and such a force is always expensive to maintain, and therefore somewhat deficient in numbers – the infantry will have been largely supplied by the subject peoples, and will have probably been untrained and unenthusiastic, just as were the Galatian armies facing Vulso in 189. In the 119-106 war the Scordisci probably suffered considerable losses, and their empire and dominating authority was no doubt weakened by their ultimate defeat.
The peace with the Scordisci after 106 or so was firm enough that the praetor T. Didius in 101 or 100 directed his military strength to the conquest of a Thracian tribe called the Caeni in the Chersonese area,18 and the next praetor had the task of sorting out and establishing the government for the conquered territory, according to a decree of the Senate which regulated the conquests.19 The peace on the Macedonian frontier did not actually last long, but when fighting took place again, in 97 or thereabouts, the Romans’ enemies were the ‘Dardani and the Maedi’.20 That is, on this occasion the Dardani were fighting on their own behalf, and not as Scordiscian clients, and they had apparently inherited the alliance with the Maedi. However, this is in the note by an excerptor, who was essentially not much interested in Rome’s wars, and it is quite possible that the name of the Scordisci was either omitted or otherwise dropped out of the record. But naming the Dardani as a participating tribe implies its independence, when for the previous century it had been under Scordiscian power. If the notice is in error or perhaps incomplete in not naming the Scordisci (for there is evidence that the three tribes were later allied), then it could be that as late as 97 the Scordisci were still active in the South Balkans, and in good relations with the other two peoples named. Nevertheless, the separate mention of the Dardani does indicate that they were now independent, just as the Maedi had been all along. The invasion of 97 was apparently defeated, but the praetor C. Sentius was himself defeated by the Maedi alone at some point in his extended governorship between 93 and 87. This was probably in 92, but he later recovered and won a victory.21
The collapse of the Roman state into civil warfare in the 80s inevitably tempted the Balkan tribes and Roman victims into a new invasion. The Scordisci, Dardani, and Maedi were allied in an invasion of Macedonia and Greece in 85 or thereabouts, during which, once again, a (part-) Keltic army reached and looted Delphi. The temple and shrine would seem to have become a fixation among the barbarian tribes.22
L. Cornelius Sulla in 85 or 84 attacked ‘the Thracians’, or maybe just the Maedi,23 which notice is sufficiently vague to imply only a Roman raid, but it may well have stirred them up sufficiently to reply. While in Macedonia, Sulla also made war on the Dardani and two more tribes, the Emeti and the Sinti, presumably Thracians. The reason, according to Appian, is that he wanted to ‘exercise his troops and enrich them’.24 There is no sign of any Scordiscian involvement in this conflict. Either these tribes were now wholly independent of the Scordisci (as the Maedi always were) or the Scordisci deliberately stayed out of the fighting – either way it surely meant the end of any Scordiscian authority over those peoples.
At about the same time, but in a direct reply to the raid which reached Delphi, the praetor L. Cornelius Scipio Asiagenus (who was consul in 83, and so probably praetor in 85), conducted a major campaign which brought a decisive defeat upon the Scordisci.25 Assuming that Asiagenus was operating from Greece with the Roman army stationed there, and assuming also that the Scordisci were defeated in their home territory, this must mean that the Roman forces were able to march through the territories of the southern Balkan tribes in order to reach the Scordisci; no fighting in Thrace or against the Dardani implies that they permitted the Roman forces to march through their territory to fight the Scoedisci. These two military crises, coming in the same year, but from different Roman commanders (who were no doubt on different sides in the Roman civil wars) would be the moment for those tribes to reach finally for their independence. The result would seem to have been that the Scordiscians’ tribal subjects all finally broke free from their domination. The relationship between these campaigns is not at all clear, but from the Scordiscian point of view the result is perfectly clear. After their defeat by Asiagenus, their empire collapsed and they are scarcely heard of again for the next seventy years. And just in case the southern tribes facing the Macedonian province felt called on to celebrate their freedom from domination by a raid into Greece, Sulla’s cynical campaign to exercise his troops as their expense will have stopped them.
The Scordisci did not actually disappear, of course, any more than their earlier tribal subjects had disappeared. They were, however, stripped of those subordinates, reduced, says Appian, to living on some islands in the Danube.26 This is clearly a curious exaggeration, or diminution, but they were certainly confined to their core territory in the Banat from then on. Their decline was illustrated in 74 when C. Scribonius Curio, commanding against the Dardani in a war which lasted from 76 to 73, marched his army all the way to the Danube, the first time a Roman force had reached that river, or so it was claimed. He arrived in the land of the Moesi, recently released from Scordiscian overlordship, and so becoming known to Rome for the first time also.27
In about 60 BC (or between 56 and 50, perhaps) the Scordisci were defeated by the Dacians, who were now powerfully united under their King Burebista. His power, which reached south by the 50s to threaten the Macedonian frontier, was clearly replicating that of the former Scordiscian empire, but with a larger, richer base – like Noricum, Dacia was wealthy in metals. The Roman war against the Dardani no doubt sufficiently weakened that people that they were, like the Scordisci, relatively easily dominated.28 In their easy submission, the Scordisci were perhaps fortunate. The Dacians also fought the Boii, who had combined in some way with the Taurisci under a single king called Kritasirs; but they were so damaged and frightened that they fled from their lands, and headed west, where they helped push the Helvetii westwards; it was stopping this movement which in turn gave Julius Caesar his excuse to begin his conquest of Gaul. The former Boiian lands were left as ‘desert’, though in fact, as ever, the removal of the overlords merely revealed the existence of their subjugated tribal subjects, who were Pannonians. The Boii who fled were no doubt the aristocracy, their overlordship having collapsed at their defeat, just as had that of the Scordisci a generation earlier.29
Strabo gives a clue to the fate of the Scordisci in commenting that though Burebista ‘destroyed the Boii and the Taurisci, he used the Scordisci as allies’.30 This must mean that the Scordisci submitted quickly when attacked, like their own former subjects the Dardani. They therefore survived and kept their homeland, but it looks as though both of these tribes were demoralized – it was unlike either of them to submit easily.
It is from this period that the evidence begins to appear which suggests that the Kelts of the Scordisci had become a population better described as Keltic-Illyrian or Keltic-Pannonian. Studies of the surviving names of the people in the area are generally from this period those of the underlying populations; the Kelts had thus been absorbed, or at least had amalgamated with their subjects.31 In fact, it seems likely that the frequent wars they had fought had helped reduce the numbers of the aristocracy drastically. They certainly lost their empire, and then their independence. The Dacians’ empire broke up after Burebista’s death in 44, and so once more the Scordisci, the Dardani, and other tribes or states, re-emerged from the ruin of a brief empire, some in Dacia itself, while others emerged from the wreck of the Boiian state, at least a dozen peoples in all, not counting the Thracians.
The political situation in the Balkans, therefore, for the first time in well over a century, was now one where no overarching inland Balkan power exercised control over many of the tribes, and so was powerful enough to confront Rome. The Scordisci had exercised some control over the southern Balkans for a time, and their successor, Burebista’s empire, lost its peripheral conquests, in the north and in Thrace. This included the Scordisci themselves, while Burebista’s own homeland territory was divided among several separate kingdoms. No new greater power emerged to take control of the situation, though the Romans, with their Civil War winding down and ending in 31 BC, were beginning to scent the possibilities of decisive conquests in the Balkan area beyond the Adriatic coast and the northern Macedonian frontier.
Julius Caesar as dictator between 48 and 44 spread rumours that he intended to attack Dacia,32 in part because Burebista had offered to join Pompey in his dispute with Caesar, back in 48,33 but then both Caesar and Burebista fell to assassins in 44, and Caesar’s carefully circulated rumour was not put to the test. Burebista’s power had been essentially personal, however, while Caesar’s, also in many ways personal, was based on seizing control of the well-organized Roman Republic, and the state even survived yet another and worse civil war. Caesar’s grandnephew, who eventually changed himself from the teenage terrorist Octavian to the mature statesman Augustus, inherited Caesar’s personal legacy, his wealth, his clients, his position, and much of the political authority he had wielded. Octavian had, of course, to prove himself as Caesar’s heir, and he did so by a political ruthlessness, by gathering competent men around him, by playing on his adoptive father’s legacy, and by using his clever political sense; rumours of a Dacian war lasted well into his personal rule, before fading away. Yet Caesar’s ambitions and plan were part of his legacy.
Octavian/Augustus was not himself a very competent military commander, but he and his circle could spot strategic opportunities and possibilities with clarity. One of these was the necessity for Italy (his essential power base) to be properly defended if his political support was to be maintained. In the Balkans this meant controlling the passageway from the northern Balkans into northern Italy and its approaches, through the Julian Alps. One of the keys to this was the former Scordiscian town at Siscia, which controlled the Sava River route from the Banat westwards. This had become Pannonian territory with the collapse of the Scordiscian domination, and in 35 BC the town was taken by an army under Octavian’s ultimate command. This, of course, was also, if anyone in Rome remembered, a place where a Roman army had earlier suffered defeat, which Octavian could, once he had taken the place, make claim to be avenging; the main result of the conquest, however, was to set up an advanced defence for Italy, while at the same time acquiring for Rome an advanced base from which further Balkan conquests could be prepared and begun. He left a major force of 25 cohorts – over 10,000 men if the cohorts were fully manned – in garrison in the place. This force became useful as soon as the main Roman force withdrew, when it had to put down a rebellion in the winter of 35-34.34
As it happened, it was not from Siscia that the decisive Balkan campaign by Roman forces was launched, though the place did function as a major base in later stages of the war of conquest. Instead the campaign came, appropriately, out of Macedonia. After defeating his opponents, Octavian gave the governorship of Macedonia to M. Licinius Crassus, who had switched sides in the Civil War at a crucial moment. Whether it was realized just how ambitious Crassus was is not known – though he was a grandson of the Crassus who had been a triumvirate colleague with Caesar and Pompey, and in those years, any Roman governor was inevitably ambitious – but it was perhaps not expected that he would launch himself into a major war of conquest.
The condition of the Macedonian frontier had not essentially changed since the time of the death of Alexander the Great. The line of control of the Roman garrison was much the same as that from which Philip II had headed north in his temporary conquest of the lands as far as the Danube in the 330s. Since then, on at least three major occasions, the northern barbarians – Thracians, Scordisci, Dardani, and others – had penetrated that defence line, in some cases to reach and loot the sanctuary at Delphi, and on several other occasions had raided into Macedon. Previous Roman governors had defended the frontier either by fighting inside Macedon when an invasion came, or marching beyond it to intercept a raiding party somewhere in Thrace and Dardania, or to pre-empt a planned invasion. There had been no attempt to move the frontier forward, in part because there was no obvious line of defence before one reached the Danube, except perhaps the line of the Haemos Mountains (Philip’s forward defence line in Thrace), and the conquest of the Balkan Peninsula was unlikely to be welcomed in Rome – successful competitors with Octavian/Augustus were not needed; and no governor would usually have the time in office to achieve much more than the defence of the existing line. Now, however, for the first time, a Roman commander was aiming to shift the Roman-Macedonian frontier well forward. One might expect that the Greeks and Macedonians who had recently suffered from the campaigns of Roman armies fighting a Roman civil war – the various battles concluding each section of that war had been fought at Pharsalus in Thessaly, Philippi in Macedonia, and at Actium, with the accompanying ravaging and military damage – would be only too happy to see Crassus moving that frontier well away from their settled lands, and Crassus himself and his soldiers as well.
The object of Crassus’ conquests were some of the Thracians and the Getai, but also the Bastarnae of the Danube mouth and delta, and the groups and tribes between the Macedonian frontier and the Danube, a whole area of minor states, tribes, and the occasional city. These groups could not resist a Roman army consisting of considerable numbers of war-hardened soldiers. The conquests of the eastern Balkan area was completed in two years, by 28 BC,35 although risings and rebellions did follow. As a result of these campaigns in the north and south-east, there was left outside Roman control just the central Balkans, the valley of the Morava River, the area which later became the province of Upper Moesia, sandwiched between the Romans along the Adriatic and Crassus’ conquests in Thrace and the lower Danube area. This was the territory of the Scordisci and the Dardani and some others, not recorded as being active in any way since Burebista’s time; it was in fact roughly the area which had been the Scordiscian ‘empire’ two generations before. Crassus had come close to the Scordisci in his campaigns, for he operated in the land of their neighbours the Moesi, further along the Danube36 – and Augustus’ conquest of Siscia was of a place of proud memory for Scordiscian warriors. The Scordisci no doubt felt threatened from both sides, but were given a reprieve, and Crassus returned to Rome to celebrate a triumph.
It may be assumed that these various conquests in the Balkans had led to the ‘pacification’ of the conquered peoples – or their ‘subjection’ – but this did not necessarily mean that there had been any serious campaigns against them.37 When he had acquired Siscia, Octavian accepted the submission of the local Pannonians, but did nothing more.38 This may have been what happened to the other groups who were not campaigned against; the presence of a victorious Roman army in the region might well have persuaded them simply to render a formal submission and so to avoid a serious ravage by an army which was all too competent at such activity.
This situation held for over a decade. Then in 16 BC an attack by the Scordisci and the Denthelitai against Macedonia is recorded; Dio’s words imply that they broke through the frontier defences and ravaged widely, though this may be either his guesswork or an exaggeration (or both).39 These were two of the Balkan peoples who had not been noted as either conquered or subdued during Crassus’ campaigns, though the Denthelitai had been Roman allies and had been defended by Crassus against attacks by, for example, the Bastarnae.40 The Scordisci, rather surprisingly, are not mentioned at all in either Octavian’s campaign in 35, or Crassus’ in 29–28. But it seems evident from events in 16–15 that the apprehensions and the tensions amongst both groups had been growing. In 16 the two launched a joint attack on Macedonia; and in the same year the Pannonians, only theoretically subdued by Augustus in 35, attacked the Roman position in Istria, at the gateway into Italy. These Pannonians were apparently allied in this enterprise with the Noricans, whose kingdom has been a Roman ally for a century or so.41
This was evidently a coordinated attack by these several peoples, who must all have felt under an urgent and immediate threat from Rome. That one of them was a Roman subject tribe – the Pannonians – and two were Roman allies – the Denthelitai and Noricum – only emphasizes that they feared something, which could only be a Roman campaign of conquest. So the three who were in treaty relationship with Rome were pre-empting an apprehended attack, which they must have expected, by their ally, and their object was presumably to drive the Romans back to the borders of Italy. By attacking first, of course, they could be accused of bad faith, whereas in fact it was their perception of Roman bad faith which had provoked them. Noricum had been a Roman ally for a century, and the Denthelitai had been saved from their enemies by Crassus, so both clearly owed Rome a major political debt. The Pannonians had the well-garrisoned Siscia, a town taken from them, in their midst. All three of these peoples in their rebellion could be accused – and no doubt were guilty – of faithlessness, but none of them would be likely to undertake a rebellion against their powerful protector without a very good reason. The only possible reason is that they feared that that protector, Rome, intended them harm.
The status of the Scordisci at the time is, however, unknown. The absence of notices of the Scordisci in the events in the Balkans since Burebista’s assassination is uncharacteristic. For two centuries and more they had been a warrior people, and for a time they had controlled an unsteady local empire for rather longer than had Burebista. They were long familiar with Rome and its wars and methods. Their subjection to, and then alliance with, Burebista suggests that they had been so weakened by rebellions and defeats by his time that they were not capable easily of continuing in independence. Yet they held the most important strategic position in the Balkan region, and they had bold relations with all the peoples around them. Possibly their protection against trouble was that very strategic position, which none of their potential enemies would wish to see fall into other hands. It is evident in this connection that they were about the only Balkan state which was still without any political relationship with Rome other than old conflicts. This may well be the clue to the events of 16-15.
The pattern of the attacks launched in 16 by the Pannonians to the north-west and by the Scordisci and the Denthelitai to the south is very reminiscent of those which had been sent by the Scordisci alone in 119 and after, when they were at the height of their power. Then they had launched simultaneous attacks against the two areas of Roman power they could reach, out of Siscia and on Macedonia, no doubt in the hope of unbalancing the whole Roman defensive system. Further, the Pannonians were former Scordiscian subjects whose city had been taken from them, the Noricans were fellow Kelts, and the Denthelitai were their immediate neighbours. It seems obvious that the outbreak of attacks on Roman positions all around the Balkans in 16 was coordinated and pre-planned, and it may be plausibly suggested that the central element in the alliance would seem to have been the Scordisci.
It is almost needless to say that the allies’ attacks failed, assuming that their purpose was in some way to disrupt the Roman positions and intentions, or to drive the advanced positions of the Romans back. Instead it provided a stimulus for what they had probably feared – widespread and permanent Roman annexation of their lands. The participation of the Noricans might suggest a hope that other peoples, perhaps in Gaul, might join in, or perhaps in the Alpine regions; the participation of the Denthelitai might suggest that a successful raid into Macedonia might persuade some of Crassus’ Thracian victims to join in the war. To some degree the allies did stimulate others – two Alpine tribes, the Cammuni and the Vennii, did join in42 – but in the Balkans the recent Roman campaigns had been all too successful, and Crassus’ old victims remained at home.
It was almost as if the Romans were waiting for this war to begin. Augustus sent his militarily-capable stepson Tiberius to command, and he, with characteristic military precision, located the heart of the problem as the Scordisci country. P. Silius Nerva defeated the Pannonian attack, and then went on next year (15) to annex Noricum.43 Tiberius then attacked the Pannonians and the Scordisci.44 And while the Pannonians rebelled again soon afterwards, and gave their name to the succeeding war, the Scordisci are not mentioned again. It has been suggested that, having been crushed by Tiberius, they then became faithful Roman allies.45 It is perhaps more likely that, given the important strategic position they occupied at Singidunum/Belgrade and in the Banat, their country would now be heavily garrisoned and their young men forcibly recruited into the Roman forces, the traditional, and effective, Roman methods of holding down conquered territory. The land had certainly been badly damaged in the course of the war.
Whatever the fate of the people or their country, this was the end of the Scordisci as a political entity. The population was already less than purely Keltic, according to the recorded names, and it would not be difficult to classify the remaining people as Illyrian or as Thracians. The surviving Scordiscian aristocrats could be subverted into Roman citizenship. In the event, when the Roman province system was delimited, they put the Scordiscian territory into Pannonia, named after their former subjects; it was the final extinction.