Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Alpine regiments of the Roman army wikipedia , lookup
Food and dining in the Roman Empire wikipedia , lookup
Culture of ancient Rome wikipedia , lookup
Early Roman army wikipedia , lookup
Roman agriculture wikipedia , lookup
Education in ancient Rome wikipedia , lookup
Roman economy wikipedia , lookup
Battle of the Teutoburg Forest wikipedia , lookup
Ancient Roman pottery wikipedia , lookup
Romanization of Hispania wikipedia , lookup
IX. THE BARBARICUM IN THE ROMAN PERIOD Historical overview | 265 HISTORICAL OVERVIEW Andrea Vaday The territory east of the Danube was the homeland of various Barbarian peoples in the Roman period. The ethnic composition of these peoples and the balance of power between them shifted periodically on the left bank of the Danube, as well as in the adjacent Baèka in Yugoslavia and the Banat in Romania. In the mid-1st century B.C., the Celts lost their hold over the Great Hungarian Plain and were ousted by the Dacians, led by their king Boirebistas. The remnants of the Celtic tribes survived only along the northern mountainous fringes of the Great Hungarian Plain. After the Romans occupied Transdanubia, they fortified the Danubian frontier of the empire to prevent the Barbarian peoples’ expansion. Roman policy was to ensure and secure the loyalty of the neighbouring Barbarian tribes to Rome, whether by political, economic or military means, this being the reason that Rome extended her authority over the Germanic Quadi who had founded their kingdom north of the province, but whose tribal territory and power also extended to the left side of the Danube Bend. Arriving to the Carpathian Basin from the Lower Danube region, the first groups of the nomadic Sarmatian Jazygians of Iranian stock settled in the northern part of the Danube–Tisza Interfluve in the early 1st century A.D. After consolidating their settlement, they forged an alliance with the neighbouring Quadi, an alliance that remained unbroken for over four centuries. The Jazygians expanded southwards between the two rivers; leaving the hilly region on the northern fringes of the Great Hungarian Plain, they crossed the Tisza at the close of the 1st century, at the time of Trajan’s Dacian wars and – with the exception of the Upper Tisza region – they occupied the area beyond the Tisza that had formerly been ruled by the Dacians. After defeating the Dacians, Rome founded the province of Dacia in the area that is now part of present-day Romania. The Sarmatians found themselves in an entirely new political environment, squeezed in between the Roman provinces of Pannonia in the west, Dacia in the east and Moesia in the south, with an area occupied by a mixed Celtic, Dacian and Germanic population to their north. Led by their king, the main tribe of the Jazygians soon joined the first Jazygian groups in the Great Hungarian Plain. The population growth, the increasing strength and expansion of the Barbarian population living in the Great Hungarian Plain did not escape the Romans’ attention. The Roman merchants who carried their wares across the Danube to distant territories, also acted as spies for the provincial army. The strategic importance of the roads traversing the Great Hungarian Plain and the crossing places on the Tisza increased. The war on the empire’s Danubian frontier broke out at the same time as the wars against Parthia in the east, in the late 2nd century. The Barbarians invaded the Danubian Ro- man provinces in two major waves from the Rhine region to the Black Sea. Marcus Aurelius had to face the fact that no matter how ingeniously Rome tried to force these peoples to become vassals of Rome, the alliance between the Barbarian tribes proved stronger. One possible solution to the crisis east of the Danube was the creation of new Roman provinces in the Germanic and Sarmatian territories that were to be called Marcomannia and Sarmatia. The Roman troops advanced deep into the Barbarian heartland, occupying the territory up to present-day Trencsény (Trenèín, Fig. 1. Coffin burial of a warrior from Mezõszemere 266 The Barbaricum in the Roman period Slovakia). After Marcus Aurelius’ death, however, Roman foreign policy again turned defensive, reflected in the linear protection of the frontier and the renewed efforts to forge an alliance with the Barbarians. The idea of creating new provinces was discarded. The war also affected the Barbarian peoples. Groups of Sarmatian Roxolani from the east settled in the Great Hungarian Plain, while the Vandals, a Germanic tribe, occupied a part of the Upper Tisza region. Parallel to the appearance of the Vandals, the Sarmatians expanded towards the northeast. The Vandal-Sarmatian border was established at this time, as was the Sarmatians’ settlement territory that remained unchanged until the last third of the 4th century. The so-called Devil’s Dyke (variously called Csörsz or Roman Dyke), a massive earthwork constructed in the late Roman period on the Romans’ initiative, marked the boundary of the Sarmatian settlement territory. The products of distant provinces found in Sarmatia reveal that Roman trade with the Sarmatian lands intensified. A number of northbound roads branched off the Aquincum–Porolissum road, along which Roman wares reached faraway areas in Poland through eastern Slovakia (Fig. 1). The profound changes that shook the Roman Empire in the 260s and 270s also affected the Carpathian Basin. The repeated Gothic attacks from the east weakened the Roman frontier defence to such an extent that Aurelian was forced to withdraw the Roman troops from Dacia, to completely evacuate the province and to resettle the Roman population south of the Danube. The road to Transylvania and the west lay open to the Goths. Driven from their homeland east of the Carpathians by the Gothic advance, new Sarmatian groups arrived to the Great Hungarian Plain; their majority eventually settled in the Baèka and the Banat in the south. That the Sarmatian territory became increas- Fig. 2. Sarmatian belt with late Roman military mounts. Mezõszemere– Kismari-fenék, grave 30 ingly important to the Romans is shown by the repeated peace treaties concluded with this people in the first third of the 4th century and the creation of a defence zone in the Barbaricum to protect the Danubian provinces of the empire against the Gothic onslaught. The middle line of the Roman defensive earthworks running along the boundary of the Sarmatian settlement territory was probably constructed at this time, in the wake of Marcus Aurelius’ campaigns (Fig. 3). That this immense earthwork system was patrolled by Barbarian troops supervised by Rome is reflected in the high number of armed male burials in the cemeteries lying along the entire length of the earthwork and the presence of late Roman belts with military mounts and Roman brooches, signalling their wearers’ ‘Roman’ military rank (Fig. 2). Fig. 3. Reconstruction of a Roman rampart at Fancsika The peoples of the Barbaricum during the Roman period | 267 Arriving from the east in the early months of 332, the Goths and the Taifali first attacked the Vandals living in the Körös region and then turned against the Sarmatians, who repelled the attack with Roman help. The late 370s heralded the dawn of a new era. There were regular clashes between the tribes arriving from the east and the Eastern Roman troops in the Balkans. Emperor Valens personally led the campaign against the Barbarian peoples. In the decisive battle at Hadrianopolis (Edirne, Turkey), the Gothic-Alan-Hun army dealt a crushing defeat to the Roman army. Some groups of this Barbarian army moved westwards and eventually settled in Pannonia with Rome’s permission. The wars became constant, leading to the collapse of the Sarmatian defence line. In 401, upon hearing of the Huns’ advance, the Vandals fled westwards from their homeland in northern Hungary; they were joined by the Quadi living north of Transdanubia and groups of Alani from the Tisza region. On December 31, 406, they crossed the Rhine, the first of the many successive waves of the Barbarian peoples’ migrations to the west. As a result of their flight, the population temporarily decreased in some parts of the Great Hungarian Plain. The migrations through the Carpathian Basin meant immense hardships for the Sarmatian population. Decimated by the constant wars, the Sarmatians withdrew to the Danube–Tisza Interfluve and the areas south of the Danube after the Gepidic takeover following Attila’s death. THE PEOPLES OF THE BARBARICUM DURING THE ROMAN PERIOD Andrea Vaday THE SARMATIANS The archaeological heritage of the Sarmatians was first identified in the late 19th and early 20th century. The first systematic overview of the Sarmatian finds from the Barbaricum was written by Mihály Párducz in the early 1930s. This study was followed by a string of others, publishing the find assemblages in the museums of Csongrád, Hajdú-Bihar and Bács-Kiskun counties. Mihály Párducz remained the leading expert in this field of research; in addition to publishing a number of excavated cemetery and settlement finds, he also wrote a second summary of the Sarmatian corpus of finds and determined the find types typical of individual periods (the early Sarmatian period, the finds of the 2nd and 3rd centuries and of the Hun period). Sarmatian finds were published but sporadically after World War 2 until the 1970s. Fieldwork was mostly restricted to rescue excavations; the single planned excavation was conducted at Madaras, where an entire Sarmatian cemetery was unearthed. The history of the Sarmatians in the Carpathian Basin was at the time based on András Alföldi’s works. János Harmatta contributed important studies on Sarmatian linguistics and the history of the eastern Sarmatians, an indication of the rigid separation of historical, linguistic and archaeological research in this field. The study of the peoples living on the fringes of the Sarmatian settlement territory, such as the Dacians, the Quadi and Vandals, was even more neglected in Hungary. Our knowledge of the archaeological heritage of the peoples of the Barbaricum is extremely patchy in Hungary. The long-time activity of the archaeologists working in the Szeged Museum can be strongly felt in the southern part of the Great Hungarian Plain, similarly to the intensive work of a handful of archaeologists in a few smaller areas. Very little work has been done in the Baèka in Yugoslavia, in the Banat, and in the northwestern and western areas of Romania. The number of published find assemblages is very low. Only a few new sites have been reported from the Banat since Bódog Milleker’s overview, published in the late 19th century. New sites from the Baèka have only become known in the past few decades. The research and the interpretation of the archaeological assemblages from the Roman period in these two countries was coloured by a political bias, especially after World War 2: in Romania, some Sarmatian assemblages were attributed to the Dacians, while in Yugoslavia, the archaeological heritage of this Iranian people was associated with the Slavs of the Roman period. The field surveys in the 1960s along the Devil’s Dyke marked a major milestone in the research of this period. The exact chronological position of this earthwork rampart was clarified by trial excavations. Two other important discoveries were made during these field surveys: one was the Roman fort at Felsõgöd, the other a smaller Roman military building at Hatvan–Gombospuszta, both indicating Roman activity in the Barbaricum. Fig. 4 shows the number of new excavations and publicaFig. 4. Number of excavations and publications 268 The Barbaricum in the Roman period Fig. 5. Site distribution of the Microregion Research Project of the Great Hungarian Plain tions. It is clear at first sight that the findings of only some 25 per cent of the excavations were published between 1951 and 1960, while only 19 per cent of the excavations conducted in the next decade, even though the number of archaeological investigations had almost doubled. The discrepancy between the number of excavations and the number of publications continued to rise in the ensuing decades. This state of affairs will hardly be remedied in the near future since the number of finds recovered from large-scale rescue excavations eclipses by far the assemblages known to date. The systematic cataloguing and evaluation of the known find assemblages in the past decades has increased the number of known sites by many hundreds, even in counties that had not been particularly well investigated. The field surveys conducted in the Great Hungarian Plain revealed that the Sarmatian settlement network was much more dense than the modern one. The number of sites is very high, even if they are proportionately distributed over four and a half centuries. In Békés county, for example, 109 of the 535 sites representing various periods identified during the topographic survey of the 35 km2 large area bordered by Gyomaendrõd, Szarvas and Örménykút were Sarmatian ones (Fig. 5). The conspicuous difference between the number of sites identified during field surveys and the number of excavated sites reflects the paucity of investigated sites in a given area. As a result, a general analysis and evaluation of the find material is rather difficult owing to the uneven state of research, the lack of excavations and the badly documented or undocumented assemblages in local museums still awaiting publication. At the same time, the findings of the excavations preceding major construction projects and the so-called microregion projects, involving both field surveys and excavations in smaller areas, have vastly enriched our knowledge owing to the wealth of information provided by the find material and the large excavated surfaces. Major construction projects are usually preceded by one of three excavation types: linear, block-like or a combination of the two. The first type characterizes road con- struction projects and the laying of gas and oil pipes. In these cases, the length of the excavated area is significantly larger than its width. One disadvantage of these investigations is that practically nothing can be learnt about the actual extent of the site, about the area falling outside the excavated territory. Another one is that the sections to be excavated are determined on the basis of the field survey preceding the excavation and the areas that appear to be ‘empty’ are not investigated, even if they fall within the planned line of the motorway or pipeline. One case in point is the Kompolt–Kistér site, where the surface features observed during the preliminary field surveys indicated the presence of two nearby sites. The rescue excavation conducted in the area between the two ‘sites’ revealed that the features identified during the field surveys were in fact part of the same site. The third disadvantage of linear excavations is the relatively small width of the excavated area. The planned course of a new road is usually 60 m wide in the case of motorways and 20–40 m in the case of smaller roads, meaning that in spite of the relatively large size of the excavated surfaces, a number of archaeological features cannot be interpreted. The second type of major rescue excavations is the block-like type. The excavations conducted on the site of future shopping centres, petrol stations, border crossing stations and the extraction pits of the motorway constructions fall into this category (e.g. Csengersima and Polgár). In contrast to linear excavations, the length and width of the excavated area is more proportionate, providing considerably more information on the stratigraphy of a site. The third type is the most advantageous, combining the advantages of the above two without their disadvantages. This type of rescue excavation, however, is far too rare: it is an option only in cases when there is a chance to excavate the planned exit areas and service station of motorways. Fig. 6. Archaeomagnetic survey of a Sarmatian workshop at Gyoma The peoples of the Barbaricum during the Roman period | 269 The research of the Barbaricum in Hungary entered a new phase during the past three decades not only in terms of the quantitative and qualitative increase of finds, but also as regards the use of various analytical methods in the investigation of a given site. In addition to traditional survey methods, aerial photography, geophysical surveys and subsurface probes are now also employed in site prospecting (Fig. 6). The reconstruction of the one-time environment of a settlement site has similarly become a more or less routine exercise. these names indicate a Celtic origin for these towns, suggesting that the Jazygians occupied these settlements and that the Celts probably continued their existence under Jazygian rule. The different lifeways of the sedentary Celts and the semi-nomadic Sarmatians no doubt eased some of the tensions between these two peoples. The surviving Celtic population contributed much to the distinctive material culture of the Sarmatians of the Carpathian Basin, whose original eastern culture gradually faded. THE CELTS THE DACIANS The study of the peoples who lived in Pannonia before the Roman conquest and of the spiritual and material culture of the native population has always been a part of Roman studies; in contrast, the study of the Celtic groups living in the areas east of the Danube during the Roman period has been largely neglected. This can in part be ascribed to the lack of excavations, and in part to the fact that while the survival of the local Celtic population can be traced rather accurately in Transdanubia, in the Barbaricum the Celtic population blended with the native population. Very little is known about this ‘mixed’ material culture and the few find assemblages can be dated within broad chronological limits only. As a result, the distinctive Celtic find types of the Roman period in the Barbaricum cannot be determined for the time being. We know from Greek and Roman sources that Celtic tribes lived in the Danube–Tisza Interfluve and on the northern fringes of the Great Hungarian Plain. Unfortunately, these sources are silent on the relationship between the immigrant Sarmatians and the local Celts. In his Geographike written in the 2nd century, Ptolemy lists the names of the Sarmatians’ towns in this region. Some of The study of the Dacians has also been neglected in the Great Hungarian Plain. While working on his monograph about the Sarmatians, Mihály Párducz also collected the Dacian assemblages and the finds he believed could be associated with the Dacians. The first study offering a historical interpretation of the Dacian find material in the Great Hungarian Plain was written by Zsolt Visy in the late 1960s. One of greatest difficulties was that the material in question was mostly made up of stray finds without any context that did not even allow a precise dating. In the early decades of the 1st century, the Jazygians conquered the northern areas of the Danube–Tisza Interfluve that had until then been occupied by the Dacians and they gradually moved into the Baèka as well. The Great Hungarian Plain, however, remained under Dacian rule until Trajan’s wars. Curiously enough, there were no finds that supported the information contained in the written sources. Authentic Dacian finds were lacking from the one hundred years following the mid-1st century. The occasional Dacian vessel recovered from the burials of the earliest Jazygian groups in the northern part of the Danube–Tisza Interfluve indicated the presence of a mixed population. The dating Fig. 7. Dacian and Sarmatian finds from Gyoma. End of the 2nd century 270 The Barbaricum in the Roman period and the archaeological interpretation of the finds from Jánosszállás and Hódmezõvásárhely–Kakasszék was practically impossible. No ‘pure’ Dacian finds were known, only mixed Sarmatian-Dacian assemblages. A Dacian settlement was unearthed at Szegvár. The finds from this site finally confirmed the historical picture projected by the written sources of the Great Hungarian Plain before the Sarmatian occupation. Finds dating from the last quarter of the 2nd century have been uncovered on the Sarmatian settlements at Újhartyán and Gyoma (Fig. 7). The Dacian finds from the second occupation phase of the Sarmatian settlement suggest that Dacian groups fled to the Sarmatian settlement territory in the face of the Vandals’ invasion of northeastern Hungary during the Marcomannic wars. Dacian finds again appear sporadically in the final decades of the 4th century (for example at Kardoskút), indicating that the population groups fleeing the Huns also included Dacians. THE QUADI Gábor Márkus Little research has been done on the Quadi in the Barbaricum, in part owing to geographical reasons since only the fringes of the areas occupied by this Germanic group fall within Hungary’s border. It is therefore hardly surprising that in his study of the Quadic finds in Hungary written in 1963, István Bóna relied heavily on the finds from Transdanubia. New advances in this field were brought by the field surveys and excavations in the Ipoly valley, in the course of which several Quadic and Sarmatian sites were identified. One interesting observation made during these surveys was that the settlements of these two peoples were not spatially separate in the territory east of the Danube Bend, but often lay quite close to each other, sometimes even sharing the same site. The finds collected during the field surveys were unsuitable for clarifying their chronological position. The excavations at Ipolytölgyes revealed intensive trade contacts between the Quadi and the Romans in the 2nd and 3rd centuries. The large-scale excavations in the 1990s marked a major turning point in the research of this period. The settlements yielding Germanic or mixed Germanic-Sarmatian assemblages investigated along the planned course of Road 2/A must be mentioned in this respect. The sites around Kompolt, lying by the Heves county section of the M3 motorway, yielded information of a different nature. Even though closed assemblages of Germanic finds were not found on these sites, various Germanic artefacts were recovered from several features (Fig. 8). The Sarmatian settlement can be dated to the period after the Marcomannic wars, when commerce with the neighbouring Germanic communities flourished along the borders. Although we now have a fairly good idea of the archaeological heritage of the Quadi from the 2nd–3rd centuries, this is not the case for the late period. It is unclear whether this can be ascribed to a genuine lack of Quadic/Suebic finds from the 4th century or the inability to recognize the finds for what they are. The Roman historian Ammianus Marcellinus mentions that as a result of the long co-existence of the neighbouring Quadi and Sarmatians, their customs and lifeways became very similar and it is therefore possible that the assemblages from the late period lack the distinctive and well datable Quadic and Sarmatian traits. The problem is the same in the areas where the settlement territory of the Quadi, the Sarmatians and the Vandals overlapped. Fig. 8. Germanic finds from Sarmatian settlement features at Kompolt–Kistér The archaeology of the Sarmatian territories | 271 THE VANDALS Gábor Márkus The archaeological heritage of the Vandals, tribes of eastern Germanic stock who controlled extensive areas in Poland, Slovakia and the Ukraine, is known as the Przeworsk culture, after the eponymous cemetery unearthed in Poland in the early 20th century. As a result of the German, Polish and Slovakian studies in this field, we now have a fairly good idea of the internal development of the culture. In eastern Europe, the northern fringes of the Carpathian Basin marked the southern border of the culture’s distribution, roughly from the Tarna river, through the Tisza bend to the Szamos and Kraszna valleys. This southern border zone has hardly been investigated and only a few sites of the culture are known. The scanty Hungarian material is nonetheless important owing the fact that only on the northern fringes of the Carpathian Basin was there a direct interaction between the Vandal tribes and the Roman Empire. The first finds from Hungary came to light in the mid-19th century. All of these assemblages were recovered from the cremation burials of warriors (Lasztóc/Lastovce and Gibárt). At the time they were dated to the Iron Age. Comparable find assemblages were found in the 1930s at Apa, Árdánháza/Ardanovo, Szolyva/Svaljava and Kékcse. The advances made in this field of research eventually led to the correct dating and the ethnic attribution of these graves. The currently known graves of Vandal warriors number about two dozen. Most of these lie in Szabolcs-Szatmár-Bereg county (Tiszakanyár, Nyíregyháza–Árpád Street, Nagyvarsány, Kisvárda–TV tower and Vásárosnamény–Hajnal Street), with two known from Heves county (Terpes and Sirok) and one from Romania (Bujánháza/Boineºti). These burials can be dated to between the final decades of the 2nd century and the early 4th century. These graves were without exception cremation burials (either inurned or scattered cremation burials) and they all contained many weapons, mostly spears – often as many as two or three pieces –, shields and the occasional sword. In addition to weapons, the finds from these burials invariably included another important article of the equipment used by Vandal warriors, namely spurs that could be attached to the boots. Larger cemeteries of the Vandal population are known only from the neighbouring countries. Royal burials have not yet been found in Hungary, although István Bóna has suggested that the golden pin found at Vállaj had perhaps originated from a royal burial. Similarly to the other Germanic peoples, a shift from cremation to inhumation can be observed in the 4th century among the Vandals, probably under cultural influences from the Mediterranean. The study of the Vandals’ settlements began in the 1950s and has since then been continuous. Unfortunately, only smaller sections of the known settlements were excavated in earlier campaigns and even though larger areas have recently also been investigated at Csengersima and Beregsurány, their finds have not been published, meaning that these settlements cannot yet be set in a wider context. The distribution of the known settlements reflects the extent to which an area has been investigated, rather than the actual, one-time settlement network. Most of these settlement sites are known from the Miskolc area ((Miskolc–Sötétkapu, Miskolc–Szabadság square, Miskolc–Szirma, Sajókeresztúr). Beside various pits, the settlements features included the typical houses of Barbarian settlements, such as the sunken house with two to four posts and without an oven uncovered at Ózd–Stadion. The ethnic attribution of these settlements is problematic, especially in the Sarmatian border zone, owing to the mixed nature of the pottery finds. Vandal pottery reflects a strong Roman influence from the turn of the 2nd–3rd centuries; this influence was largely technological and the Sarmatians’ mediation can be assumed. The Vandal potters adopted the footturned potter’s wheel and the earlier hand-thrown pottery was soon supplanted by wheel-turned wares that had earlier been wholly absent. The pottery finds from the settlements are dominated by ‘Sarmatian’ type wheel-thrown vessels and the ratio of the typical, traditional Przeworsk pottery is very low. THE ARCHAEOLOGY OF THE SARMATIAN TERRITORIES BURIALS Eszter Istvánovits & Valéria Kulcsár The past twenty-five years have greatly enriched our knowledge of Sarmatian burial rites and, indirectly, of their religious beliefs. Although many graves had been earlier unearthed (even if very few larger cemetery sections), only a fraction of these burials had been published according to modern standards. Grave drawings and cemetery maps, indispensable for the analysis of the burial rite, were rarely included and as a result there were many inaccuracies and, also, misconceptions in our knowledge of the various elements of the burial rite. One of the problems in this respect is the separation of the funeral customs that can be traced to the earlier, eastern Sarmatian homeland and the ones that only characterized the Sarmatians of the Carpathian Basin. The exact dating of the graves from the earliest Sarmatian settlement territory in the northern part of the Danube–Tisza Interfluve is rather difficult. The gold dress ornaments, earrings, torcs and gold and carnelian beads of southern Russian and Ukrainian origin found in female burials remained fashionable over a longer period of time. It seems likely that the grave assemblages lacking Roman articles from neighbouring Pannonia are the earliest ones. Graves with Roman brooches first appeared at the turn of the 1st–2nd centuries. The graves of the next group of Sarmatian immigrants included a number of more richly furnished male burials as well. The eastern Sarmatian sword with a ringed hilt and the golden strap end with a tamga, a 272 The Barbaricum in the Roman period mark of the individual or the clan, are typical finds from these graves (Figs 9–10). Considerably more female graves are known from the early period, in part due to the fact that male burials tend to be more poorly furnished and that the modest finds rarely enable a more accurate dating. One of the most characteristic traits of the burial rite is the orientation of the deceased. The Jazygians of the Carpathian Basin buried their dead in a south to north oriented grave pit, with the head to the south. The graves of the next wave of immigrants, arriving at the close of the 2nd century, were usually oriented to the north. It has been demonstrated that in addition to a number of other features, such as the position of the horse harness in the grave, the types of horse harness and weapons, this indicates the arrival of a mounted group from the Don delta at the time of the Marcomannic wars. It was earlier believed that the custom of raising a mound over the grave, so widespread among the eastern Sarmatian tribes, was not practiced in the Great Hungarian Plain, and that the newcomers adopted the custom of burying the dead in ‘unmarked’ graves. Burials mounds from this period have only been reported from two areas: the northern part of the Fig. 9. Sword with ringed hilt from Újszilvás Fig. 10. Golden strap end from Dunaharaszti Fig. 11. Sarmatian tumuli at Vácszentlászló–Harmincadhányás The archaeology of the Sarmatian territories | 273 Great Hungarian Plain and some areas of the Baèka and the Banat. Few burial mounds, so-called kurgans, have survived in Hungary since most have been ploughed away. Some can still be seen in areas that have not been drawn under agricultural cultivation, in marshy or forested areas, such as the burial mound groups in the Hortobágy, the Sarmatian kurgans in the Gödöllõ Hills and in the Baèka and Banat (Fig. 11). Graves without a burial mound that were enclosed by circular ditches open to the south, a practice recalling eastern burial rites, were first observed in the Great Hungarian Plain in the early 1950s. It is generally assumed that a mound had originally been raised over the burials enclosed by ditches. The cemeteries containing burials enclosed by a circular ditch show a rather even distribution in the Great Hungarian Plain. Only a handful of sites with such graves were known until the late 1970s. By 1980, their number grew to thirty and today some fifty sites with graves of this type have been registered. Almost every larger Sarmatian cemetery contains such burials, suggesting that this custom was fairly widespread and that these grave ditches had probably been missed during earlier excavations. The number of cemeteries in which the burial rite could be clearly be observed has increased greatly. It could be demonstrated in several cases that the graves enclosed by a ditch occupied a central position within a cemetery or a grave group, indicating that these were the burials of the paterfamilias or the ancestors of an extended family. The simpler burials were arranged around these graves. In other cases, the burials lay a little farther away and were arranged into rows. Cemeteries of this type were unearthed at Törökszentmiklós–Surány and Endrõd–Szujókereszt. At Lajosmizse–Kónya-major, the female and male burials lay in separate parts of the cemetery, while at Sárdorfalva–Eperjes the men and the boys were buried with their weapons in the central part of the small cemetery surrounded by a ditch, while the women and the lower ranking members of the community were buried around them. A number of previously unknown elements of the Sarmatians’ burial rites could be observed in the cemetery section excavated at Szõdliget–Csörög in the mid-1990s. The postholes uncovered in one of the graves indicated that the deceased had probably been laid on a bier, a practice that has also been documented among the Avars. The remains of fire beside or inside some graves, perhaps the remains of a ceremony to commemorate the dead, were noted at this site for the first time in the Carpathian Basin. Several differences can be demonstrated between the burials in the Carpathian Basin and those in the eastern steppe. The graves in the Great Hungarian Plain usually contain fewer finds than the ones unearthed in southern Russia and the Ukraine, and the burial rite itself is also less varied. One of the reasons for this is that the graves in the Great Hungarian Plain were often robbed and we can no Fig. 12. Female burial, with the reconstruction of the belt. Endrõd–Szujókereszt 274 The Barbaricum in the Roman period longer tell what articles had been originally deposited beside the deceased and what the grave had originally looked like – whether it had a side-ledge or had contained a bier – and it is equally difficult to reconstruct the customs associated with the funerary rites. Another reason for this apparent ‘poverty’ is that the Sarmatians of the Great Hungarian Plain had settled far from their eastern kinsmen and there was practically no contact between them, this being the reasons that eastern articles were no longer deposited in the graves after some time. Another reason for the differences in the burial rites can no doubt be sought in the cultural impact of the neighbouring peoples, first of all the Romans. The analysis of the burial rites offers many clues for ethnic attribution. The period preceding the Hun invasion saw the arrival of various population groups, as well as a tendency towards the ‘uniformization’ of the costume worn by these peoples, another difficulty when attempting the ethnic separation of grave finds. The burial rites, however, are the most conservative elements of a culture since they are bound to religious beliefs by a thousand strands. The analysis of the burial rites and the grave goods enabled the separation of a late Sarmatian/Alan group on the northern fringes of the Great Hungarian Plain. The cemeteries of this group were unearthed at Tiszavalk, Tiszakarád and Tiszadob. Beside the traditional costume articles and the burial rites, various artefacts from the east and articles modelled on eastern prototypes occurred among the grave goods. New eastern burial rites could also be observed. Comparable finds were made on some of the sites that were investigated during the rescue excavations preceding the construction of the M3 motorway. The finds from these excavations will no doubt contribute to a better understanding of this period. The Sarmatians’ costume can be reconstructed from the carefully excavated and documented graves. In the early phase, round carnelian beads were worn around the neck and sometimes on the arms together with gold jewellery. Roman brooches (fibulae) and the distinctive bead-embroi- Fig. 13. Straps of a buckled boot dered costume of Sarmatian women appeared around the late 1st and the early 2nd century. The neck, the sleeves and the hemline of the overgarment was richly embroidered with colourful beads, as was the shift worn under it and the lower part of wide-legged trousers. The dress was fastened on the left side with a textile belt fitted with a metal ring. This belt was also profusely decorated with beadwork, rattles, bells and pendants that were believed to ward off evil (Fig. 12). Smaller knives were suspended from the belt. The gold jewellery of eastern origin was replaced by silver and bronze ornaments (earrings, torcs, bracelets, lunular and axe shaped pendants). In addition to the customary grave goods, such as spindle-whorls and vessels, metal mirrors too made their appearance – the latter were often broken before being deposited in the grave. Male burials were rather modest compared to female ones. Roman brooches fastened the upper garment, the leather belt was fastened with an iron, bronze or silver buckle and a leather pouch containing a strike-a-light, flints, an awl or other smaller tools, was suspended from the Fig. 14. Detail of a female burial with jewellery. Tiszaföldvár–Téglagyár The archaeology of the Sarmatian territories | 275 belt. The knife was kept in a wooden or leather case. Very few graves with weapons are known from this period. Eastern costume articles again appear with the new population groups arriving at the close of the 2nd century. Female burials yielded belt pendants of Cypraea shell, a symbol of fertility, while the warriors’ graves usually contained boot straps ornamented with narrow buckles and strap ends, as well as the occasional pair of spurs (Fig. 13). Eastern articles again disappeared from the burials within a few generations and were replaced by Roman or local products. The appearance of funeral obuli in male graves, usually placed in a beaded pouch, can be ascribed to Roman cultural influence. Women’s costume continued to be embroidered with beads and the popularity of silver and bronze jewellery, as well as of bead necklaces and bracelets remained unbroken. Articles of eastern origin appeared again with the arrival of groups fleeing the Huns at the end of the 4th century: marine shell talismans and gold jewellery in female burials and a wide array of weapons, buckles and strap ends in male ones. Several bracelets were worn on both arms and dresses were fastened with several brooches (Fig. 14). SETTLEMENTS Andrea Vaday The Great Hungarian Plain is not a uniform region in terms of its geography and natural resources, this being one of the reasons that the settlements uncovered in various areas differ. Another reason for this diversity is to be sought in ethnic and economic differences. Earlier excavations usually uncovered smaller settlement sections and only a fraction of the finds was published. A conceptual change could be noted in the reports on the smaller settlement sections investigated in the Middle Tisza region: in addition to the publication of the entire find material, the zoological finds were also included. The settlement features unearthed on these sites, however, were mostly storage and refuse pits, with the occasional house. The overall layout and the nature of these settlements could hardly be reconstructed from these pits, and neither could the internal chronology of the settlements be established. A multi-period site was unearthed at Gyoma as part of the Microregion Research Project of the Great Hungarian Plain. In contrast to the earlier practice, the finds from the sites, representing various periods, were published in one volume, together with the zoological material and the results of the pollen analyses. It has by now become a standard procedure to publish not only the archaeological finds from an excavation, but also the analytical results and the findings of geophysical surveys and subsurface probes, if these techniques had been used for investigating a site. The information gained from more recent excavations have added a number of details to the rather sketchy picture of Sarmatian settlement patterns in the Carpathian Basin. The semi-nomadic Jazygians who arrived here in the 1st century probably established temporary campsites at first, none of which have yet been identified. Their lifeways only changed in the earlier 2nd century: while preserving their nomadic economy based on stockbreeding, settlements engaged in agricultural cultivation also appeared. Settlement structures show a great diversity, depending on the region, the function of the settlement and the ethnic composition of its occupations. It was earlier believed that the Sarmatian settlements in the central and southern areas of the Great Hungarian Plain were characterized by a few houses and many refuse pits, while the settlements on the fringes had fewer refuse pits and more houses. This general picture was based on the findings of rescue excavations conducted over small areas and the comparison of the few known sites from the southern and northern part of the Great Hungarian Plain. The layout of a settlement is greatly influenced by the lifeways of its occupants. Settlements occupied over a longer period of time that retained their original layout usually have storage and refuse pits dug beside the houses. Since it was near-impossible to set up a finer internal chronology for the different settlement features owing to the smallness of the excavated area and the fact that the find material had not been analyzed in detail, settlements occupied for a long time that preserved their original layout became ‘multi-pit’ sites compared to more briefly occupied ones. Well-documented excavations conducted over larger areas offer more information on the settlement structure and its changes. The Sarmatians established their settlements by watercourses and natural waters, on elevations rising slightly over the plain. The settlement layout was adapted to the natural terrain, with the settlement usually extending along the banks of a river or lake. Many settlements lay by streams that were probably still active during the Roman period, but have by now been filled up. Settlements lying farther from water were supplied by water artificially and in these cases there was no need for the settlements to follow the watercourse, allowing it to grow and spread more freely. Roadside settlements probably followed the line of the road, resembling modern villages. The field surveys conducted in the Middle Tisza region revealed that the Sarmatian settlements lay fairly close to each other. Some of these settlement chains were coeval, others were not. The former are only known from the late Sarmatian period, from the phase preceding the arrival of the Huns and from the Hun period. Most of them appear to have been larger farmsteads surrounded by arable lands and pastures, resembling the present-day clusters of farmsteads. The situation is slightly different in the case of settlement chains with sites dating from different periods. These settlements were most often established in the 2nd and 3rd centuries and remained occupied until the close of the 4th century or the earlier 5th century. This settlement pattern can most likely be explained by the practice of abandoning the settlement when the nearby fields became exhausted, with the occupants of the settlement moving closer to the new fields. In these cases, the successive layers of the settlement do not accumulate over 276 The Barbaricum in the Roman period Fig. 15. Section of a Sarmatian settlement with the animal pens. Northern junction of the M5 motorway at Kiskunfélegyháza each other, but simply ‘shift’ to a new location, meaning that the structure and internal layout of settlements with a longer use-life changed over time. Although the overall size of these settlements can rarely be determined since not one single site has been completely excavated, size estimates can nonetheless be made for a few sites. On the testimony of the aerial photos and the field surveys, the settlement at Gyoma extended over an area of some 35–40,000 m2, of which about 14,700 m2 was excavated, while at Polgár–Kengyelköz, the aerial photos suggested that the settlement extended for another 6–700 m beyond the investigated 30,000 m2 large area. At Kompolt– Kistér the field surveys suggested that the settlement extended for an additional 350–400 m to the west and some 200–250 m to north and the south beyond the 28,700 m2 large excavated area, suggesting that the overall size of the settlement was around 190,000–192,000 m2. At the nearby site of Kompolt–Kistéri-tanya, the western boundary of the settlement fell into the 32,370 m2 large investigated area, but the settlement itself extended well beyond this area. The above clearly show that even though much larger settlement sections are now unearthed than previously, we still know little about the overall layout of these settlements. Nothing is known about the size and layout of the Jazygian ‘towns’ listed by Ptolemy, of which ‘Partiskon’ can be identified with present-day Szeged. We know that many Sarmatian sites were clustered around the Roman watchtower by the crossing place on the Tisza. A similar cluster of settlements could be observed on the Barbarian side of the Roman border near the Roman forts. The network of villages was no doubt fairly dense along the roads leading through the Great Hungarian Plain. A chain of roadside settlements has been identified in the Törökszentmiklós area, where settlements and cemeteries lay along a 7–8 km section of the road leading eastwards from the crossing place on the Tisza at Szolnok. Ditch systems were observed on a number of sites. Some of these were no doubt defensive in nature, while others functioned as animal pens or as drainage ditches, and some no doubt enclosed individual homesteads (Fig. 15). The relationship between the ditches and the settlements could be observed during the rescue excavations near Kiskunfélegyháza and Dusnok. The settlement at Polgár–Kengyel-köz, occupied between the later 3rd century and the turn of the 4th–5th centuries, was bordered by a north to south oriented system of parallel ditches. The outermost ditch was the widest. A series of postholes was found on the floor of the inner, much narrower and shallower ditch, suggesting a ditch and palisade structure. Comparable settlement features were unearthed at the Polgár–Csõszhalom-dûlõ site. At Nagymágocs, a village occupied from the close of the 4th century to the mid-5th century was bordered by two ditches, one facing the Mágocs brook and one perpendicular to it. At Szentes–Berekhát, a 3rd–4th century settlement enclosed by ditches was unearthed during the rescue excavation preceding the construction of the bypass road. The roughly 5 m wide and 2 m deep trench had steep walls. It bordered the village towards the marshland, and together with the rampart constructed The archaeology of the Sarmatian territories | 277 The sunken houses with wattle and daub or adobe walls were usually built on the higher-lying part of the settlements. At Polgár, smaller streets ran between the residential and the economic buildings. At Tiszaföldvár, at the Kompolt sites and at Gyoma, the settlement structure was looser. The location of the houses and their outbuildings did not appear to follow a systematic plan. The workshops usually lay beyond the houses, near the boundary of the village, and the direction of the prevailing wind appears to have been an important consideration in their siting. Some workshops were found beside the craftsman’s house. Wells and cisterns were dug either near watercourses or between the houses. They are often vital to determining the internal chronology of a settlement since shorter and longer periods can be distinguished in their fill (Fig. 16). The houses were surrounded by beehive shaped or cylindrical storage pits. Agricultural settlements usually had Fig. 17. Human remains thrown into a refuse pit. Kompolt–Kistéri-tanya Fig. 16. Sarmatian well. Kompolt–Kistéri-tanya from the earth removed during the digging of the ditch, it protected the village from floods. A similarly wide and deep ditch was identified at the Tiszaföldvár–Téglagyár site on the side facing the Tisza. Besides offering adequate protection against floods, these ditch and rampart systems also had a defensive role, as the one with the palisade at the Polgár–Kengyel-köz site. A settlement and a late Sarmatian cemetery from the period after the abandonment of the settlement was found at Mezõszemere–Kismari-Feneke, together with a section of the Devil’s Dyke. Another earthwork rampart ran parallel to the Devil’s Dyke in the Sarmatian territory and behind this second earthwork there was a palisade construction with gates and ramps. 278 The Barbaricum in the Roman period more storage pits and the same holds true for the settlements from the periods when the population of the Great Hungarian Plain increased significantly owing to new waves of immigrants, and agricultural production was more intensive in order to provide for the population. Several oven and kiln types can be distinguished on the basis of the finds, including baking and drying ovens, as well as pottery kilns. Some settlements were abandoned by their occupants, while others were destroyed during times of war as shown by their destruction layer. The human remains thrown into refuse pits can be linked to these troubled times, corresponding to the Marcomannic wars and the military clashes at the close of the Hun period (Fig. 17). THE ARCHAEOLOGY OF THE NORTHERN AND NORTHEASTERN FRINGES OF THE SARMATIAN SETTLEMENT TERRITORY Eszter Istvánovits & Valéria Kulcsár The earliest settlements in the northeastern part of the Barbaricum in the Carpathian Basin are known from the SzatmárBereg plain. The Beregsurány settlement was probably established sometime in the final decades of the 1st century. The settlement at Csengersima, investigated in 1998–99, can be dated to approximately the same period. Traces of metalworking were also found at this site. The ratio of wheelturned pottery is negligible in the rich ceramic material recovered from the settlement. The settlement structure and the various settlement features differed from the ones observed in the Sarmatian heartland, owing to the differing geographic environment. The construction of timber-framed buildings is one indication of this. The best analogies to the archaeological finds from these two sites are to be found partly in the Germanic Przeworsk culture and partly in the Dacian territory, suggesting that the Dacians who previously occupied this area had encountered and mingled with the Germanic groups arriving from the north at a fairly early date. The immigration of Germanic groups before the mid-2nd century is also indicated by the finds from the burial ground uncovered at Malaya Kopania in the Sub-Carpathians. It has been suggested on the basis of the written sources that the early finds of the Przeworsk culture can be associated with the Lugii or the Buri, both Germanic tribes. According to historical data, the Vandals/Victovali settled in the Upper Tisza region in the mid-2nd century, before the Marcomannic wars. The archaeological record confirms their presence in this area: this Germanic group can be identified with the second wave of the Przeworsk culture. The newcomers occupied the northeastern corner of the Carpathian Basin, including the plainland in the Sub-Carpathians. The cremation burials of this population, such as the ones uncovered at Kécske, Tiszakanyár and Kisvárda–TV tower, contained iron shield bosses, spurs with large spikes, heavy double-edged swords and spearheads. Unfortunately, none of the find assemblages discovered earlier come from systematic excavations and not one single burial that could be associated with this population has been reported from the past two decades; the control excavations on sites such as Kisvárda–TV tower did not yield any results. The Vandals’ southern and southeastern expansion was checked by the Sarmatians who, on the testimony of the archaeological finds, also reached this region sometime in the Fig. 18. Charcoal kiln (?). Csengersima The archaeology of the northern and northeastern fringes of the Sarmatian settlement territory | 279 Fig. 19. Timber-framed sunken house. Beregsurány–Barátság-kert mid-2nd century. Intermingling between the two populations proceeded quite rapidly in the contact zone between the two groups, as shown by a number of Vandal articles, such as shield bosses, found in Sarmatian graves. The mixed Daco-Germanic assemblages of the type found at Beregsurány and Csengersima were supplanted by a new culture in the later 3rd century. A high number of grey, wheel-turned wares with stamped decoration was found. Even though the late Roman period settlement Fig. 20. Pottery kiln. Csengersima partly overlapped with the earlier one at both sites, the distinctive stamped pottery was entirely absent from the ceramic inventory of the latter, indicating the chronological differences between the two, as well as the sharp break between the two material cultures (Figs 18–19). The late Roman period settlement at Beregsurány was excavated by Dezsõ Csallány in the late 1960s. One of the largest potters’ centres in Europe was uncovered on the bank of the Mic stream. Some forty thousand sherds were 280 The Barbaricum in the Roman period recovered from fifty-two grated kilns. Six similar potter’s kilns were unearthed at Csengersima (Fig. 20). In addition to the grey stamped vessels, many Roman wares, including wheel-turned pots and a variety of painted vessels were also found. The stamped pottery has much in common with similar wares turned out by the Roman pottery workshop of Porolissum in the nearby province of Dacia and the date of the sites too suggests that Roman potters were also active in the Barbaricum (primarily at Csengersima). Their appearance in the Barbaricum can be associated with the gradual deterioration of the situation in Dacia and its later evacuation and abandonment by the Romans. Grey pottery with stamped decoration had a fairly wide distribution, reaching even areas in Poland. The ethnic attribution of the peoples making and using this pottery is controversial since many different peoples are known to have lived in the area where it was distributed. No burials of the late Roman period have yet been found in the Szatmár-Bereg plain, adding yet another difficulty to resolving this question. Neither do we know when life came to end on these settlements. The vessel forms would suggest that the potters were still active in the Hun period. RESEARCH PERSPECTIVES Andrea Vaday & Gábor Márkus Until the 1970s, our picture of the Barbaricum was essentially based on the information provided by burials. The few inves- Fig. 22. Sarmatian-Quadic settlement. Vác–Csörögi-rét Fig. 21. Distribution of Sarmatian sites excavated between 1971 and 1995 tigated settlement sections did not play a decisive role in the interpretation of the archaeological heritage of the peoples who lived here. Even though the burials do offer a fairly good idea of attitudes to death, the funerary customs of different social groups and the changes in the burial rite over time reveal very little about day-to-day life, of which a better understanding can only be gained from the investigation of settlements. Until the mid-20th century, the ratio of the settlements was less than 1 per cent among the known sites of the Barbaricum. This ratio has since changed significantly (Fig. 21). The publication of the find assemblages recovered from settlements and cemeteries, whose number has increased vastly owing to the more recent large-scale excavations, calls for the elaboration of new analytical methods and, obviously, for new ways of looking at the finds (Fig. 22).