{1-42.} 2. THE DACIAN KINGDOM (András Mócsy)

The date and credibility of the earliest reports concerning a Dacian people are contested. The ancient assumption, that slaves, who figured in New Attican (4th century B.C.) plays under the name of Daos, were in fact Dacians, is less than plausible, for the first confirmed report of the Dacians' existence refers to a time at least two centuries later, whilst the Romans first reached the Danube even later, between 76-73 B.C. The utility of our data is limited both by the fact that only fragments of the detailed chronicles survive, and by the fact that the reflections on individual peoples found in these chronicles often compress events stretching over several historical periods. Thus historical research has not been able to establish clearly whether King Oroles of Dacia made war against his eastern neighbours, the Celtic Bastarnae, in the 2nd century B.C., or in some much later, equally indeterminate period. There is a similar lack of consensus over such an essential question as the identity of King Rubobostes, who is claimed by one source to have built up the power of Dacia; was he the first significant Dacian ruler, some time during the 2nd century B.C., or was his name merely a misspelling of Burebista, the king who is generally credited with founding a powerful Dacia? A further difficulty derives from the fact that more than one name has been attributed to the Dacians. The tribes that spoke Thracian and lived in the eastern half of the Balkan peninsula, the lower Danube valley, and Transylvania were called by a variety of names in Greek and Roman literature. The Thracians proper, who had very early contact with Greek culture, inhabited a region bounded in the north by the Balkan Mountains and in the west by Macedonia, while the Getae lived in a region north of the Balkan Mountains, along the lower reaches of the Danube. The Dacians of Transylvania, who were the {1-43.} last Thracian-speaking people to come to the notice of Greco-Roman world, are also called Getae in Greek sources; and Roman historians, who drew upon Greek sources, often — and arbitrarily — translated the appellation 'Getae' as 'Dacian', even when, as it happened, they were referring to authentic Getae. Thus caution must be exercised when dealing with the fragmentary sources that mention Dacians in the context of the wars, waged by the Romans in the 2nd and 1st centuries B.C., on the northern borders of Macedonia, against various Thracian, Getian, and Celtic tribes.

The Getae came into early contact with the Greeks. The development of their trade, crafts, and art was stimulated by contact with the Greek cities on the Black Sea, both directly and through the mediation of Thracian tribes. Within Transylvania, there is no significant trace of such Greek influence in the late Iron Age. It is all the more surprising, then, that Greek coins from the 2nd century B.C. (Macedonian and Thasos tetradrachma) have been found in larger number in southern Transylvania than in the lowlands south of the Carpathians. It is not known when these coins reached Transylvania. What is clear is that the Getae started producing silver copies of Greek coins well before such minting began — around 50 B.C., according to a recent study — in Transylvania. Thus the beginning of coinage in Transylvania coincides with that of Burebista's rule, and finds abound precisely at the center of the nascent Dacian state, in an area bounded by the Ruszka, Szörényi and Kudzsir mountains, and, in the north, by the Maros valley.

That region's mountainous borders to the south impeded access, but the Maros, Zsil, and Olt valleys allowed access to the Danube. The center of the nascent state was in the valley of the Városvíz, a tributary of the Maros. Numerous fortified hill settlements were established in the 1st century B.C. Recent research points to a significant growth in the number of non-fortified settlements in the 2nd and 1st centuries B.C., which implies a progressive transformation of Dacian society, but the nature of this change {1-44.} is not clear. The strong Celtic influence northern Transylvania presumably contributed to technological development. Besides the iron tools, potter's wheels, and other technical innovations, the hill fortresses can also be attributed to the influence of late Celtic, Oppidum culture. Coins and the advent of local minting indicate that silver was becoming a measure of the value of certain goods and a means of accumulating wealth. However, as in the case of the Danubian Celts, it cannot be concluded that the existence of coinage signified general acceptance of this measure of value or the advent of market-oriented production. By all indications, the Dacians lived in a predominantly subsistence economy; even domestic pottery was produced mainly in homes, and not for the market but for personal or local consumption. In this respect, even at the time of the Dacian kingdom, the economy of the Dacians was less developed than that of the Danubian Celts.

The building of fortified settlements, the accumulation of silver and other valuables, and, above all, the spectacular and unexpected consolidation of Dacian power all point to a high degree of social stratification. What internal factors might have spurred this development? The little light shed by historical sources on Dacian society projects the image of two sharply-divided, caste-like strata: the 'tarabostes', or 'cap-wearing people', and the 'long-haired people'. Indeed, subsequent portrayals of Dacian notables show them wearing headgear made of felt, sometimes of fur, and resembling Phrygian bonnets. The polarization between a small caste of notables and a vast, exploited mass of people explains the duality of Dacian archaeological finds, a feature that is absent in other late Iron Age finds in the Danube region. The typical finds in Dacian fortresses consist of Greek pottery and mirrors, and elaborate, locally-produced silver objects; the finds in unfortified settlements, on the other hand, consist of hand-formed pottery that is very simple, even coarse, and bears the mark of the early Iron Age, as well as of primitive metal goods.

{1-45.} To be sure, the history of the Balkan and Danubian regions in the last centuries B.C. reveals other instances of a tribe, or group of tribes, investing vast territories in short order, and of linguistically different tribes being recruited as allies or subjugated and exploited. Even before the Dacians had come to power in the lower and middle reaches of the Danube, Roman Macedonia had come under attack, first by the Celtic Scordisci, and then, after 100 B.C., by the Dardani. From the middle of the 3rd century B.C., the Scordisci based themselves in the region of what is now Belgrade; the Dardani's ancient homeland lay in the southern part of today's Serbia and in Macedonia. Early sources identify the Scordisci as Rome's sole or principal enemy in the Balkans; the Dardani were seldom mentioned, and then only as one of the Thracian tribes allied with the Scordisci against Rome. When the Scordisci's power waned around 100 B.C., they came to be listed less and less frequently among Rome's Balkan foes, while references to Dardani and other Thracian 'allies' of the Scordisci become more frequent.

Similarly, the political history of the Carpathian basin in the last few centuries B.C. is marked by transitional supremacy of a tribe or group of tribes. At the end of the 2nd century B.C., the northwestern part of the Carpathian basin was ruled by an alliance of Celtic tribes led by the Boii. The Scordisci were predominant in the Sava valley, and they kept the Pannonic tribes, who lived between the Sava and the Drava rivers, in subjection until ca. 100 B.C.; their local hegemony collapsed probably in consequence of the successive defeats they suffered at the hands of Romans attacking from Macedonia. When King Mithridates of Pontus made plans to attack the Romans by way of the Balkan peninsula and the Apennines, he referred to the Pannonic tribes, and not to the Scordisci, as masters of the region on his path; it appears, therefore, that around 70–60 B.C., the Pannonic tribes were no longer subjugated.

Around 50 B.C., these tribes of the Carpathian basin and Balkan peninsula were confronted by the Dacian ruler Burebista, {1-46.} who began suddenly to expand his domain. As noted, the sources do not indicate clearly whether Burebista was the original unifier of the Dacian tribes, or whether his efforts at unification built upon the work of his predecessors, such as the mysterious Rubobostes. The sources date the period of his rule by reference to important events in Roman history: the high priest Decaineus, Burebista's chief assistant and adviser, reached Dacia in the year of Sulla's accession to power (82 B.C.), and the assassination of Burebista coincided with that of Caesar (44 B.C.). These correlations indicate that Burebista reigned in the middle of the 1st century B.C. By all accounts, he ruled for a long period, perhaps as much as forty years. According to our most important source, Strabo, Burebista completed his vast conquest in a short period of time. An epigraph at Dionysiopolis (Balczik, Bulgaria) indicates that by 48 B.C. he was regarded as the 'first and greatest king to rule over Thracia'.[9]9. G. Michailov, Inscriptiones Graecae in Bulgaria repartae I (2nd ed., Sofia, 1970), no.13 = Dobó, Inscriptiones... 837. The same inscription mentions that Dionysiopolis sent a delegation to the father of a Getae ruler, and that the latter met the envoys at Argedava. Due to the fragmented state of the text, it cannot be ascertained whether this ruler in Argedava was Burebista's father. Nor is it evident that Argedava is the same place as Arcidava (Varadia), situated on the eastern fringe of the Banat. If that is the case, then the ruler who received the Greek delegation in Argedava may well have been Burebista's father. On the other hand, there is no indication that Burebista exerted any influence over Greek cities on the Black Sea during the first half of his reign. The Dacians do not feature in the bold plans of King Mithridates in 70–60 B.C., either as enemies or as allies, which suggests that their sphere of influence did not yet extend to the lower Danube or the Black Sea coast. In all likelihood, then, Burebista completed his conquests in a brief space of time between 60–50 B.C. In the earlier, and longer phase of his rule, he must have concentrated on unifying the Dacian tribes, and on establishing and consolidating a centralized realm. In this task, which no doubt involved bloodshed, he had the aid and {1-47.} counsel of the high priest Decaineus, who was said to enjoy 'virtually regal' powers.

With the considerable accomplishment of unifying the Dacian tribes behind him, Burebista proceeded to conquer in short order vast territories. The sequence of his conquests is not clear, for sources provide only a few bare facts. Burebista expanded the Dacian domains in three major directions. To the southeast, he reached as far as the Black Sea and seized Greek cities from Olbia, at the mouth of the Dnieper, to Apollonia (Sozopol, Bulgaria). It is likely that before he launched this campaign, Burebista subjugated the Getian tribes along the lower reaches of the Danube, and also the Bastarnae, who lived to the north of the Getae, on the eastern side of the Carpathians. A people of Germanic or Celtic origin, the Bastarnae had already supplied troops to the kings of Macedonia in the 2nd century B.C., and they would continue to serve as mercenaries of foreign powers; after Burebista's death, they fought against the Romans as allies or mercenaries of the Dacians.

Much to Rome's alarm, Burebista's second target was Macedonia. His armies crossed the Danube and, plundering their way across the Balkan peninsula, headed for the Roman province of Macedonia and the Dalmatian coast, which was also under Roman control. In the last years of Julius Caesar's rule, fending off the Dacian menace was a prime concern of Roman foreign policy. In the event, the only tangible evidence of this conquest is that the Scordisci, defeated by Burebista, became his allies in subsequent campaigns, and that the Dacians established a lasting presence south of the Danube, in the northern part of today's Serbia.

The third, westward direction of Dacian expansion touched the neighbouring Celtic tribes. In the first half of the 1st century B.C., the Boian alliance encompassed Celtic tribes that had settled in northern Transylvania. Burebista first objective was probably to subjugate these tribes, the Taurisci and the Anartes; in the process, he confronted the Celtic tribal (Boian) alliance that dominated the {1-48.} entire northern half of the Carpathian Basin. The clash occurred when Burebista crossed the Tisza and headed towards the Boian tribe's heartland in northwestern Transdanubia and western Slovakia. His victory over the Celts led not only to the breakup of the Boian alliance, but also to the establishment of Dacian settlements in the southern parts of today's Slovakia. Evidence of this settlement includes characteristic hand-formed Dacian pottery, as well as the 'Dacianization' of Celtic names in the region during the 2nd century A.D.

The resulting change in the balance of power on the middle and lower reaches of the Danube worried the Romans, all the more because a people virtually unknown to them and living beyond their sphere of influence had suddenly emerged as a major political factor in the hinterland of Illyricum (the Dalmatian coast) and Macedonia. Thus the destruction of Dacian power became one of Julius Caesar's key political objectives, and he made plans to launch an offensive from Macedonia. Had Caesar not been assassinated, the campaign would probably have got under way in 44 or 43 B.C. Burebista, too, fell victim to his political enemies at about the same time. The plot (the sources speak of a rebellion) may have served the interests of dissident Dacian nobles, for the unification of the Dacian tribes no doubt required the dismissal of many independent chieftains. Nor can it be ruled out that the Romans had a hand in the plot; in the period after Burebista's demise, more than one Dacian ruler kept contact with Octavian and Mark Antony.

At the peak of his power, Burebista had 200,000 soldiers, although somewhat later the Dacians' military strength was estimated at no more than 40,000. The late ruler's domain was divided into five, then four parts. At the center of that domain, Decaineus managed for a time to rule over Burebista's tribe; the merging of the roles of king and high priest was probably made permanent by his successor, a certain Comosicus. The kingdom over which ruled monarchs from Burebista to Decebalus obviously encompasses the {1-49.} central territory, which lay in southwest Transylvania; the other Dacian or Getian kings mentioned during the reign of Augustus ruled over the isolated Getian tribes on the lower reaches of the Danube. Identification of the actual Dacian kingdom's rulers is not without problems, for the commonly cited list names Comosicus as successor to Burebista and Decaineus but ignores Cotiso; yet the latter is identified in a number of sources as the Dacian who ruled over the region around the Iron Gate, the Danube's mountainous gorge. Since there are other misspellings in the list, it may be that Comosicus and Cotiso are one and the same; in any case, all sources agree that the latter ruled over the central part of Burebista's onetime kingdom.

Of the fragments of Burebista's domain, only the one within the Carpathians, in Transylvania, can be considered a Dacian kingdom, for the rest encompassed Getian tribes. Sources indicate that a king named Dicomes ruled over one of those regions; he kept in touch with Mark Antony in the turbulent period following Caesar's death and offered him help before the Battle of Actium. Cotiso, meanwhile, tried to court favour with Octavian — at least according to Antony, who charged that Octavian even considered linking the two families by marriage. Other Getian or Dacian kings tried to sell their support to one or the other of the Roman factions. Such farflung diplomatic initiatives were uncommon among Barbarian peoples at the time, but each of these kings considered himself to be Burebista's legitimate successor and sought foreign help to realize his hegemonic ambitions. The reality of this struggle for hegemony is attested by the fact that the empire's five fragments were soon reduced to four; presumably a king had succeeded in eliminating one of his fellows. In these circumstances, the Dacian menace continued to worry the Romans even after Caesar's death. Octavian wanted to wage war on the Dacians in fulfilment of Caesar's intentions, although the timeliness and feasibility of such action became increasingly doubtful. Even the Iapodic war (35–33 B.C.) was {1-50.} launched on the pretext of preparing for a Dacian war, whereas its real objective control over the hinterland of coastal link along the Adriatic between Italy and the Balkans. One tangible gain was the town of Siscia (Sisak), in the Sava river valley; Octavian claimed that it would serve as an important base for a war against the Dacians.

The bellicose project faded as Caesar's successors intensified their struggle for undivided power, and in later years Rome's other concerns took precedence over a Dacian menace that was declining. Thus Rome contented itself with driving Cotiso out of his domains south of the Danube. The campaign was launched in 29 B.C., shortly after Octavian's victory at Actium. In the first phase of Balkan war that lasted several years, M. Licinius Crassus defeated Cotiso, who nevertheless hung on to the left bank of the Danube and probably continued to rule for some time over the heart of Burebista's kingdom.

Augustus' drive to the Danube threatened the Dacians, but Rome confronted them only later, in the context of a vast military and diplomatic operation aimed at stabilizing the political situation north of the Danube. In the first phase, the Romans occupied two regions south of the Danube, Pannonia and Noricum; formed a Moesian army in Macedonia; and prepared to extend their influence west of the Danube. In 10 B.C., during the Pannonian campaign, the Dacians had launched an attack across the Danube, and were repulsed. In retaliation, Augustus dispatched an army that 'compelled the Dacians to submit to the rule of Rome'.[10]10. Res gestae divi Augusti (Monumentum Ancyranum) 30 = Dobó, Inscriptiones... 769. The commander was probably M. Vinicius; it is recorded that in one of his campaigns, he had defeated an army of Dacians and Bastarnae, and then compelled the Celtic peoples living on the northern flank of the Great Hungarian Plain to join the Romans in 'alliance'.[11]11. Dobó, Inscriptiones... 769a. This northward thrust across the Great Plain probably coincided with Lentulus' campaign, in which 'a people difficult to reach' — the Dacians — were drive out of their highland homes, and north of the {1-51.} Danube; the Romans then established military outposts along the right bank of the river. One source notes that although this operation 'did not defeat Dacia, it kept that power distant'.[12]12. Florus, Epitome II, 28 = IV, 12. Of the subsequent Roman campaigns, only a few details survive, e.g., that Roman soldiers sailed up the Tisza and the Maros rivers to reach the Dacians. A poet's comment, that Dacia's Appuli tribe (presumably people in the region of Apulum [Gyulafehérvár]) lived within early reach of the Pontus, must have been prompted by some event. In fact, this is the only surviving indication that the Dacians sought to make contact after Burebista's death with people in the lower reaches of the Danube. The shortest route from Apulum to the Black Sea lies along the Olt valley; and it is hardly coincidental that Augustus, who saw no necessity in militarizing the new frontier on the Danube, chose a spot near the mouth of the Olt to establish one of the first Roman military camps in the Danubian region (Oescus-Gigen). The second camp (Carnuntum-Deutschaltenburg), probably also established during Augustus' reign, lay near the lands of Rome's other dangerous opponent in the Danubian region, the German king Maroboduus. All this indicates that despite the fragmentation and territorial losses suffered after Burebista's death, the Dacian kingdom remained one of the better-organized — as well as less accessible — political units in the region.

In these circumstances, the Romans, who had barely consolidated their positions on the Danube, must have welcomed the gradual advance of two tribes of Sarmatian horsemen along the lower reaches of the river. The Jazyges, closely followed by the Roxolani, drove a wedge between not only the Getae and the Dacians, but also the Roman empire and the Dacians; moving westward, they came to serve as a buffer between Pannonia and the Dacians. The main body of Jazyges reached the northeastern part of the Great Plain in first decade A.D., expelling Dacians who had settled there in the time of Burebista. It is possible that the Sarmatian migration — which Rome may have sporadically encouraged — was a factor in {1-52.} Dacian incursions of varying scale, and that some of those were joint Dacian-Sarmatian operations. Already in Augustus' time, the Romans had resorted to resettlement measures in an attempt to pacify the region north of the Danube; to make way for the Sarmatians, they moved large numbers of Getae (or Dacians?) to Moesia. The turbulence took long to dissipate, for Sarmatians and Dacians were warring in Moesia even towards the end of Tiberius' reign. The subsequent period of peace coincided — according to the list of kings noted earlier — with the forty-year reign of the Dacian ruler Coryllus.

Since Coryllus is not mentioned in any other source, it is probably a misspelling of Scorilo, a relatively common Dacian name. An anecdote has survived about a Dacian king called Scorilo: To dissuade his people from meddling with the Romans, he set two dogs fighting, then introduced a wolf, whereupon the dogs joined forces to attack the wolf.[13]13. Frontinus, Stratagemata I, 10, 4. Such prudence must have been characteristic of the long reign of Coryllus-Scorilo; the example of the dogs was certainly apposite during the Roman empire's first great crisis (68–69 A.D.), when legions left the Danube frontier unguarded to join in the civil war. The Sarmatians repeatedly exploited this situation, annihilating several Roman legions and even governors. On one occasion, probably in the winter of 68–69, the Dacians, too, crossed the Danube into Moesia, where they seized a few Roman camps along the border. It is not known if this raid was mounted by Scorilo or by some independent Dacian group. If the attribution of the dog anecdote is valid, it is more likely that the attack was launched by some Dacians from Wallachia, and that it served as the stimulus for Coryllus/Scorilo's cautionary stratagem.

In reference to the raid, Tacitus noted that the Dacians were 'always unreliable'.[14]14. Tacitus, Historiae III, 46, 2. Admittedly, his observation also drew on subsequent experience of Roman-Dacian clashes, but the Romans had been wary of the Dacians since the time of Burebista. Their strategy of forging alliances (in which the 'allies' were actually {1-53.} client states) was generally successful, but the Dacian kingdom proved less tractable. It is reported that, towards the end of Augustus' reign, the Dacians had become less menacing and more ready to accept Rome's supremacy — presumably a reference to the relatively peaceable relations in the time of Scorilo. It seems, however, that the Roman-Dacian alliance (foedus) lacked a solid foundation. Dacia differed in several respects from the German and Sarmatian client states along Rome's Danubian frontier. For one thing, the Dacians were favoured by geography, for high mountain ranges impeded access to their kingdom's heartland from the direction of the Danube. Rome's armies had to make long detours along the Maros or Temes valleys in the west, and the Olt or Zsil valleys in the east; in either case, they confronted easily-defended narrows and passes. Thus tactical advantage lay with the Dacians, and particularly so at a key stretch of the imperial frontier, where the Danube passes through a narrow gorge at the southern extremity of the Carpathians. There, a road had to be cut into the cliff face to serve as a towpath for shipping; this, one of the great engineering feats of Antiquity, was completed towards the end of Tiberius' reign. It is probably not coincidental that Daco-Roman relations took a peaceful turn at this juncture. Rome must have resorted to bribing the Dacians in order to keep this fluvial route secure.

Rome had to take into account not only Dacia's geographical defences but also the strong central organization of the kingdom. The 'royal' fortress on the heights of Sarmizegethusa, in the western Kudzsir Alps, was surrounded a ring of hilltop forts, and thus the center of power was easily defensible even against attack from the kingdom's outlying regions. These forts not only served as accommodation for troops but also as industrial centres, storehouses, treasuries and even shrines. The fortified emplacements, which covered several hectares, were protected by earthworks and generally had thick walls as well as towers; they served not only military purposes, but also as treasuries, industrial centres, warehouses, and {1-54.} even shrines. The ruler's ability to marshall great manpower is best attested by fortresses's walls, built of huge stone blocks on timber frames, by the paved yards and roads, stone steps, and by the drains hewn out of the rock. All this no doubt also served to make an impression on the ordinary folk, who lived in primitive conditions; a function that must not be underrated, for the great social gap between the 'capped' and 'long-haired' people probably necessitated an awesome display of kingly might.

Religion also served to buttress the monarch's authority and power. Religious sites include the fortresses and the more recently discovered shrines. The latter come in two shapes: round ones, marked by evenly laid stones, which probably also served as calendars, and quadrangular ones enclosed by four peristyles. All indications are that the cults had an astrological dimension. According to ancient sources, these cults originated with the Thracian Zalmoxis or Zamolxis, who was said to be a disciple of Pythagoras. The teachings of the legendary, deified Zalmoxis encompassed an ascetic way of life. This religion of Thraco-Getian origin was probably introduced by Decaineus, who had been invited by Burebista and became the latter's closest advisor. Burebista presumably perceived the potential utility of the cult and exploited it to consolidate his authority. To serve the cult was a privilege, and the priests were the guardians of its mystical truth; the head priest served in effect as mediator between the king and the transcendent sphere. The priestly privileges may well have included healing functions. The fact that ancient Greek treatises on medicine mention many Dacian medicinal herbs, along with their Dacian names, points to a high level of botanical and medical knowledge, and it can be surmised that the experience of sheep-rearing in a mountain environment contributed to the Dacians' knowledge.

The excavation of Dacian fortresses has uncovered not only locally produced objects but also Roman ones. Some of the latter are luxury goods that are commonly found beyond the frontiers of {1-55.} the Roman empire amidst the belongings of Barbarian leaders. It is noteworthy, however, that the goods of Roman origin include not only those attributable to a tribal aristocracy, but also high quality iron tools that probably served advanced crafts practiced in the fortresses. It is also probable that the Dacians had recruited Greek and Roman experts to design the fortifications. Some of the ashlars that have been unearthed, most commonly those from shrines, bear incised, Greek characters that were probably intended to guide the sequence of assembly; others may be linked to the calendar function of the shrines. Latin-character sealmarks bearing a remarkable similarity to those on Roman bricks were found on a huge, inverse conical dish which was probably used in religious functions: one seal reads DECEBALUS, and the other PER SCORILO.

Most Romanian experts interpret these two sealmarks to mean 'Decebalus, son of Scorilo'. The problem with this interpretation is that the names figure on separate sealmarks, as if one was the owner who commissioned the object, and the other, the artisan who produced it. It is likely that the first name refers to King Decebalus, although the omission of 'rex' (king) is puzzling. However, if the interpretation 'Decebalus, son of Scorilo' is taken literally, then a further problem arises, which is that another reign, that of Diurpaneus, came between those of Scorilo and Decebalus. The Dacians' new political orientation, that which led to the demise of the Dacian monarchy after the golden era of Decebalus, is generally ascribed to the same Diurpaneus.

As noted, Scorilo found acceptable the associated relationship (foedus) that Rome tried to establish with neighbours all along the European borders of the empire. He may have been motivated by the exceptionally high annual stipendium that he received from Tiberius after the construction of the road along the Danube at the Iron Gate. According to a historian, the late Count Jordanes, King Diurpaneus only launched his attacks across the borders of the Roman empire when the Dacians, 'after a long wait, grew wary of {1-56.} the greed of the reigning emperor Domitian and renounced the contract that they had concluded with earlier emperors.'[15]15. Jordanes, Getica, p. 76. Domitian may well have wished to reduce the exorbitant stipend, but it is hardly likely that he would have done so at a time when the Danubian Germans were preparing for war. It is more plausible that the Dacians timed their devastating surprise attack to exploit the tense situation on the German front.

The Dacian attack must have occurred no later than the winter of 85–86 A.D., and probably across a frozen Danube, as they had done in earlier raids. Oppius Sabinus, proconsul of Moesia, was one casualty of the surprise attack, which led to the capture of several settlements on the Moesian shoreline. The gravity of the situation is attested by the fact that Domitian rushed to Moesia and spent months preparing a retaliatory campaign. The operation was entrusted to Cornelius Fuscus, commander of the Praetorian Guard, who crossed the Danube and advanced steadily into Dacian territory. At this critical moment, Diurpaneus abdicated in favour of Decebalus, who started his reign by inflicting a crushing defeat on Fuscus' army in 87 A.D. Fuscus was killed in battle, and there were so many casualties that the Romans had to write off an entire legion. The third Roman to take charge of the campaign, Tettius Iulianus, finally managed in 88 A.D. to inflict a decisive defeat on the Dacians at Tapae, a mountain pass on the way to the royal capital.

It is difficult to discern, on the basis of the fragmentary evidence, the extent to which Daco-Roman relations were altered by the peace treaty and the new alliance concluded in 88 A.D. Roman historians, hostile to Domitian, regarded the peace terms as a victory for Decebalus: a sizeable stipend, and Roman experts whose knowledge he was evidently free to exploit for warlike as well as for peaceful ends. On the other hand, after the victory but before the peace treaty was concluded, a legion could already march unhindered 'across the kingdom of Decebalus' on its way to the German {1-57.} front on the Great Plain.[16]16. Dobó, Inscriptiones... 502 = 774a. Moreover, Decebalus had proposed peace on several occasions before his defeat at Tapae, at a time when it was hardly in Domitian's interest to prolong the war, given the growing problems with the Germans.

Nor did Decebalus take advantage of the Romans' difficulties subsequently, when Domitian was drawn into prolonged warfare in Pannonia with the Germans and the Sarmatians. In all likelihood, his ambitions were satisfied by the peace terms. He did not participate in the peace talks himself but sent an envoy, Diegis (possibly his brother); it is upon the latter's head that Domitian placed the diadem symbolizing the powers of a subordinate ruler.

The time of construction of the Dacian fortresses has yet to be ascertained. It is quite possible that at least some of the stone walls and towers date from Decebalus, for it is during his reign that Roman experts were made available to direct such projects. It might well have been a Roman military engineer who produced the seal bearing the name of Decebalus, in the form of a stamp used by the Roman army's brickmakers. The roof-tiles of the fortresses also suggest the presence of Roman architects.

While maintaining his alliance with the Romans, Decebalus significantly enlarged the territory of his kingdom in the decade following the peace treaty. The Romans tolerated such expansion, as long as it respected the foedus system, i.e., did not put at risk the agreements with other kings and the network of client states which had been so painstakingly constructed. The best indication of Decebalus's territorial gains is probably found in Ptolemy's geographical manual. The latter was written well after the conquest of Dacia, but the borders it describes are not those of the Roman province; and although, in reference to many provinces, it mentions the presence of Roman legions, it fails to do so in the case of Dacia. Moreover, since Ptolemy calls Sarmizegethusa 'royal', he must be referring not to the Roman colony established at Várhely but to the royal castle near Grădiştea Muncelului. According to Ptolemy, {1-58.} Decebalus' Dacia was bounded in the west by Tisza River, in the north by the Carpathians, and in the east by the Dniester River. This immense territory, the greater part of which was inhabited by Celtic, Sarmatian and other peoples as well as the Dacians, must have been conquered progressively, and not without bloodshed. Ptolemy lists the peoples under Decebalus' rule, though in a manner that makes it hard to identify their precise location. Other sources confirm that the northernmost regions were inhabited by the Celtic Anartae and Taurisci — peoples of the former Boian alliance living in northern Transylvania —, and the Thracian-speaking Costoboces who lived beyond the Carpathians. The other groups noted by Ptolemy are not mentioned in other sources; among them, a significant number bear the name of settlements (Predavenses, Ratakenses, Kaukoenses, Buridavenses, etc.). Since these groups were located at the center of the Ptolemaic map, one possibility is that Decebalus divided the people of Dacia into territorial units, each with its administrative center, thus erasing the traditional tribal delineations; it is significant that the one Dacian tribe previously identified, the Appuli, do not appear on Ptolemy's list even under the name of Apulenses, derived from the locality of Apulum. In this case, the structure of Decebalus' kingdom differed little from that of the conqueror Burebista, who had amalgamated a number of tribes and peoples; as noted, the latter kingdom was reduced after Burebista's death to its central region, which reached from the Maros to the Danube, and from the Banat to the Olt.

Trajan probably did not need Decebalus's conquests to convince himself that the Dacian kingdom should be eliminated. As for Decebalus, his style of statesmanship makes it unlikely that he ever entertained ambitions to conquer parts of the Roman empire; after the peace treaty, he avoided any action that might have jeopardized his relations with the Romans. Yet there are indications that in the years following Domitian's death, the Romans were already preparing for a showdown with the Dacians. Trajan initially concentrated {1-59.} on forging alliances with the Danubian Germans, and then, in the fourth year of his reign, moved against the Dacians.

There were numerous factors favouring a showdown: The already excessive stipend might have to be raised again; the terms of Domitian's treaty were difficult to sustain; further expansion of Dacia would threaten the alliance network along the Danube; the Dacians might resort to diplomatic blackmail; opposition elements were mobilizing in Decebalus's realm; and the security of navigation on the Danube had to be assured. The only trouble spot along the Danube that Rome was unable to neutralize permanently with money and diplomacy was the Dacian kingdom.

Few details survive concerning the two short but bloody campaigns, in 101–102 and 105–106, by which Rome conquered Dacia. A commemorative monument, Trajan's column, does bear an ascending, serpentine relief that evokes the war, but no amount of expert analysis has succeeded in translating this into a plausible account of the historical highlights. Trajan had marshalled considerable military force in the two campaigns, and these armies advanced on the Dacian capital from several directions, along the river valleys and through mountain passes. In the first war, the Romans, having reached the fortresses, forced their opponents to make a stand, and defeated them. Trajan stationed garrisons in the fortresses and laid down peace terms that spelled the political demise of the Dacian kingdom and aimed at a final solution to the Dacian problem: disarmament and the surrender of weapons; demolition of the fortresses; surrender of the territories annexed by Decebalus; the release of refugees; and loyalty, or indeed passivity in foreign policy. After this victory, the Romans built the first permanent bridge over the Danube, at Drobeta (Turnu-Severin); it allowed assured access to the territory of what was considered a defunct kingdom. Although, during this first war, Decebalus had shown a disposition to negotiate, after defeat he turned obdurate and prepared to continue the struggle. He rebuilt his forces, sought {1-60.} help from neighbouring peoples, and retook some of the lost territories; he even tried blackmail, taking as hostage a Roman officer who had come at his invitation to parlay.

The second Roman expedition ran into a well-planned counterattack led by Decebalus, who was determined to fight to the last. Nevertheless, the fortresses fell to the Romans in the spring of 106. If the triumphant reliefs on Trajan's column are to be believed, suicides, beheadings, precipitous flight, and attempts to parlay marked the disintegration of the ruling elite of 'capped ones'. Hunted down by the Roman cavalry, probably in eastern Transylvania, Decebalus took his own life; his head was brought back to Trajan.

With the demise of the kingdom, Dacian society also fell apart. Some of the common folk emigrated, and of those who remained, some were taken into slavery. The rigidly stratified society's highest caste, the 'capped ones', were decimated, and the survivors lost their status. The priesthood's authority vanished along with the kingdom; there is no surviving trace of a Dacian religion during the Roman period. Those Dacians who remained in place had great difficulty in becoming integrated into the society of a Roman province. In other conquered lands, the Romans could work with the only interlocutors that they were prepared to acknowledge, the aristocracy; but in Dacia, the aristocrats had disappeared, leaving the people without representation. It was this social vacuum, as well as the loss of population, that prompted Trajan and his successors to encourage people from other parts of the empire to settle in conquered Dacia.