14
THE BARBARIAN DANUBE The Roman Limes on the Lower Danube – a sophisticated and highly developed defensive and communication network which is testimony to the magnificent planning and organizational skills of Imperial Rome. Or was it? In fact, recent archaeological research in the Lower Danube area (Serbia, Bulgaria, Romania) clearly indicates that this network was not developed by Rome, but adopted from a previously established Iron Age network. In the eastern Syrmia area of today’s Serbia elements of this Celtic (Scordisci) fortification/communication system are gradually being identified by archaeologists – a series of forts/trading centres on the southern bank of the Danube stretching from Acumincum (Slankamen) to Singidunum (Belgrade) which testify to the highly developed infrastructure which existed in the pre-Roman period. The Scordisci fortresses, consisting of palisades built on earthen walls with broad trenches in front of them (see below), were constructed in the 2 nd half of the 2 nd c. BC, i.e. a century before Roman conquest of the area, and the standardization of

THE BARBARIAN DANUBE - WordPress.com · THE BARBARIAN DANUBE The Roman Limes on the Lower Danube – a sophisticated and highly developed defensive and communication network which

Embed Size (px)

Citation preview

THE BARBARIAN DANUBE

The Roman Limes on the Lower Danube – a sophisticated and highly developed defensive and communication network which is testimony to the magnificent planning and organizational skills of Imperial Rome.

Or was it?

In fact, recent archaeological research in the Lower Danube area (Serbia, Bulgaria, Romania) clearly indicates that this network was not developed by Rome, but adopted from a previously established Iron Age network. In the eastern Syrmia area of today’s Serbia elements of this Celtic (Scordisci) fortification/communication system are gradually being identified by archaeologists – a series of forts/trading centres on the southern bank of the Danube stretching from Acumincum (Slankamen) to Singidunum (Belgrade) which testify to the highly developed infrastructure which existed in the pre-Roman period.

The Scordisci fortresses, consisting of palisades built on earthen walls with broad trenches in front of them (see below), were constructed in the 2nd half of the 2nd c. BC, i.e. a century before Roman conquest of the area, and the standardization of

building material and structures clearly indicate that they were built according to a predetermined and comprehensive planning framework (Tapavički-Ilić 2007).

Archaeologically confirmed late Iron Age Celtic fortresses in eastern Syrmia

(after Tapavički-Ilić 2007)

Burial goods from the Celtic (Scordisci) grave No. 63 at Karaburma, Belgrade

(1. Duchov fibula; 2. Biconical iron fragment; 3. Silver earrings; 4. Glass beads; 5. Wheel-thrown bowl; - after Ljuština, Spasić 2012)

The ‘castella’ were mostly of relatively small size (dimensions ½ - 1 ha.), while a number of larger centres such as Acumincum (Slankamen), Taurunum (Zemun) and Singidunum (Belgrade) were situated at key geo-strategic locations. For example, the Celtic centre at Slankamen was built on a plateau from which the confluence of the river Tisa with the Danube, as well as Banat and parts of Bačka, could be perfectly observed. The later Roman castra adhered exactly to the Scordisci one, the only difference being that the walls were built of stone and new buildings erected. Similarly, the Gandoš plateau in Zemun, on which the Celtic centre of Taurunum was situated, has an excellent view of the confluence of the Sava river with the Danube, as well as the Banat and the area south of the Sava. 7 km. from Taurunum, on the Zvezdana Plateau, lay the Scordisci (and later Roman) center at Singidunum (Belgrade) (loc cit; on the spatial distribution of these fortresses see below).

Roman Taurunum was built on the remains of the Celtic fortification on the Gardoš hill and, as with the Celtic settlement, the Roman centre developed around the stronghold and quickly grew into a rich settlement due to the important trade, land and river connections.

An idea of the type of fortified Celtic settlement involved is to be seen from one of the few completely preserved Scordisci fortified settlements at Čarnok on the Crna Bara river near Vrbas (Vojvodina region), also built in the second half of the 2nd c. BC. Situated on a plain between the Danube and Tiza, two distinct phases can be distinguished at Čarnok – an initial open settlement and a later fortified settlement. The fortified settlement had an earthen rampart ca. 12 m. wide, and a 12 m. wide ditch. Archaeological material related to Phase II at the settlement, during which it was fortified, included evidence of storage facilities in which archaeological products – barley, wheat, rye, nuts, furs and meat – were stored, indicating that the settlement had the dual function of fortification and trading post (Jovanović 1988).

Reconstruction of house No. 3 at Čarnok

(after Jovanović 1988)

A similar picture is to be observed with Celtic fortresses in eastern Slavonia and western Syrmia. Important settlements at Dalj, Vukovar, Sotin and Ilok were constructed along the Danube, further at Osijek and Sarvaš along the Drava, and finally at Donja Bebrina on the Sava. The largest number of so far documented fortified settlements was located along the river Bosut and its tributaries in the area of Vinkovci. Characteristic of these fortified settlements was also the existence of a fortification system consisting of an earthen rampart and a wide ditch. There were workshops in these settlements, producing various goods to satisfy local demands, but also for trade and exchange with the neighbouring communities. The lowland-type settlements and farms, relying on agriculture as primary production activity, gravitated towards fortified settlements, although smaller workshops for the production of certain goods, especially ceramic vessels, also existed in them (Dizdar 2012:119).

Ceramic finds from the Scordisci grave (No.5) in Ilok (After Dizdar 2012)

Archaeological excavations at Celtic sites in this region, such as Gradina by Bosut, Orolik, Privlaka, Dirov Brijeg in Vinkovci, and Stari Mikanovici have confirmed these fortifications were also built in the late 2nd c. BC, and situated in strategic positions on the river terraces. As in Eastern Syrmia, their defensive systems consisted of earthen walls, palisades and broad trenches in front of them. The earthen rampart was constructed with horizontally placed beams which were covered with earth in such a way that air-flow was possible, and subsequently subjected to controlled fire. Thus, the interior of the upper earthen wall became rock hard (Tapavički-Ilić 2007).

Scordisci Fortresses along the Bosut River

A most interesting phenomenon to be observed in the Scordisci network in Eastern

Syrmia is that the distance between each Celtic fort is almost identical, measuring circa

7 km., with an approximate distance of 14 km. between the major centres (loc cit). This

spatial distribution thus enabled visual communication between the forts along the

Danube. Also interesting to note is the similarities between the Scordisci fortification

system and that of the Treviri tribe, who inhabited the lower valley of the Moselle,

which was constructed during the same period and on the same principles. The Treviri

system was composed of oppida constructed circa 25 km. from each other, with a series

of smaller castella arranged at intervals of 12 km. between them – a similar distance

ration as the Scordisci system (Koch 1988:169-182).

Reconstruction of Treviri dwellings at Altburg bei Bundenbach

Although best researched in the aforementioned areas of e. Croatia/Serbia, traces of

this extensive pre-Roman settlement/trade network are gradually being identified

further along the Lower Danube, stretching from Viminacium in eastern Serbia to the

Aquae area of Moesia Superior – around the confluence of the Timok and Danube

rivers in northwestern Bulgaria – where two Celtic settlements have been identified by

academics – Braiola / Βραίολα (Beševliev 1970:25; Duridanov 1997:134; probably todays

Bregovo), and Setlotes / Σετλοτες (Detschew 1957:434 – ‘sicher keltisch’; Duridanov

1997:134; Delamarre 2003: 272-273), the latter possibly at the Kula site where recent

excavations have uncovered Celtic material dating from the 1st c. B.C. – 1

st c. A.D under

the late Roman fortification Castra Martis (See ‘New Material (1)’ article – archaeology

section).

Slightly further along the Danube lay the Celtic settlements of Bononia / Βονωνία

(modern Vidin) (Holder I, 1904, 481ff; Der Kleine Pauly I, 928; Kazarow, 1919:62; Hubert

II, 43; Duridanov, 1980 (2); 1997, 134; Falileyev 2009:281), and Remetodia (modern

Orsojo – between Ratiaria and Almus – Lom district, Montana region) (Tomaschek, II,

2. 1894:69; Detschew 1957:391; Duridanov 1997:134; Falileyev A., Graham R. 2006,

Falileyev 2009: 282).

Topography of Viminacium

The remains of Viminacium, the capital of the Roman province of Moesia Superior, are located on territories of the villages of Stari Kostolac and Drmno, about 12 km from the town of Kostolac and about 90 miles southeast of Belgrade. Viminacium became one of the most important Roman cities and military camps in the period from 1st to 4th centuries. The earliest Celtic cemetery at Viminacium dates from the late 4th /early 3rd centuries B.C. Less than 1 km southeast from the ruins of Roman Viminacium, a large Celtic cemetery with 43 graves has been excavated (Jovanović 1984, 63-93; 1985, 13-17).

In today’s north-central Bulgaria the Celtic settlement of Icacidunum has been identified at the confluence of the Danube and Iskar (Oescus) river, near the present day village of Gigen (Guleantsi district, Pleven region) (Beševliev 1952:#92; Duridanov 1997:136; Falileyev 2009: 282), while in the north-east of present day Bulgaria the most significant Celtic centre on the Danube was situated at Mediolana (modern Pirgovo, Rousse region) (see ‘Zaravetz Culture’ and ‘Celtic Hoards from Bulgaria’ articles). Mediolana was strategically situated near the confluence of the Danube and the Lom river, the latter connecting Mediolana with Celtic settlements in the interior such as Abritu (Razgrad). A vast amount of Celtic archaeological and numismatic material has been discovered in the vicinity of Pirgovo/Mediolana over the past century, (Paunov 2012; see also Balkancelts ‘Celtic Hoards from Thrace’ and ‘Mother Matrix’ articles), clearly indicating that Mediolana/Pirgovo was a key Celtic economic and coin production centre in the pre-Roman period.

Further Celtic settlements along this short stretch of the Danube included Ablana (today’s Gorno Ablanovo) to the west of Mediolana, as well as Tegris (today’s Marten), and Appiaria (placed XIV and IX Roman miles from Tegris (respectively TP and IA), to the east of Mediolana (see ‘Celtic Settlements in Northern Bulgaria’ and ‘Zaravetz Culture’ articles). As with Mediolana, Ablana was situated at a vital strategic point – in this case near the confluence of the Jantra river with the Danube. Extensive Celtic numismatic and archaeological material discovered along the courses of both the Lom and Jantra rivers indicate that these were vital trade arteries connecting the Celtic settlements on the Danube with those in the Thracian interior.

Kiln for the Mass Production of Celtic (La Têne ceramic) from the Krivina site, Russe region, Bulgaria

The recent excavations at Krivina (carried out by Lyudmil Vagalinski, Director of the National Institute of Archaeology with Museum at the Bulgarian Academy of Sciences (NIAMBAS), near the Celtic settlement of Ablana, illustrates that this was not just a trading post, but a major Celtic economic centre, as the discovery of a bread oven and a large ceramic kiln at the site indicates. The kiln, which is dated roughly to the end of the 1st c. BC/beginning of the 1st c. AD, is especially noteworthy for a number of reasons. Its unusually large size reveals a high capacity of manufacture, i.e. the mass production of Celtic ceramic, which included late La Têne painted ware. This type of ceramic was popular among the Celtic tribes from Normandy to southwest Germany in the west, to the Scordisci in the east, and especially along the Danube (Vagalinski 2011; see ‘Zaravetz Culture’ article, with relevant lit.), and is usually found in large settlements such as the Celtic oppida. It was produced by professional potters, and used by people of high social status. It is usually found together with late La Têne burnished pottery – exactly the case with the Celtic site at Krivina.

Hoards of Celtic Coins of the ‘Macedonian Types’ found along the Danube and Jantra

rivers in the Rousse/Veliko Tarnovo areas of n.e. Bulgaria

(see ‘Celtic Hoards from Bulgaria’ article)

Further to the east lay the settlements of Transmarisca (Proc. Aed. IV, 3, 7/9) (Holder, 1904, I, 431; Beŝevliev, 1966: 422; Duridanov 1980(2):6), Altina (Proc. aed. 4,7,9, see Detschew 1957:14) and Candidiana (modern Malak Preslavets (Glavnitza district, Silestra region) (Tomaschek II,2:72; DeLamarre ZCP 54, 2004: 262) while the Celtic origin of Durostorum – today’s Silestra (Dottin 1906:334; Tomaschek II, 2:73; Kazarov 1919:62; Hubert II, 43; Beševliev 1970:26; Duridanov 1997:139; Mac Congail 2008:38; Boyanov 2010; Ivanov R. Roman Cities in Bulgaria (In print) has recently been confirmed by the discovery of an inscription from Silestra which mentions two other settlements of Celtic origin – Gavidina and Arnumtum, in the vicinity of Durostorum (Boyanov, Ivanov op cit).

This network continued to the Danube Delta where major Celtic centres such as Noviodunum (now Isaccea in Romanian Dobruja) (Holder II:787, Duridanov 1997:137), ’Αλιόβριξ (opposite Noviodunum on the Danube Delta) (Duridanov op. cit), Νίσχονις (Detschew 1957:332, Duridanov 1997:137), Vergo[b]rittianus (Doruţiu-Boilă 1980:137-138 nr. 115, Gerov 1967:40, Duridanov 1997, with relevant lit), and Arubium (now Măcin by Galaţi; Duridanov op cit.; see Coralli article with relevant lit.) have been identified.

Thus, in the immediate pre-Roman period the Lower Danube region was controlled by a sophisticated network of Celtic settlement/communication centres, the Danube forming a vital economic/trade artery linking the population of Central Europe with the Black Sea/Pontus area, and by extension the ‘Barbarian’ and Hellenistic spheres.

Literature Cited

Beševliev V. Keltische Ortsnamen in der Kastellverzeichnissen bei Prokop // Actes du Premier congres international des etudes balkaniques et sud-est europeennes. T. VI. Sofia, 1968. S. 415–423.

Beševliev V. Zur Deutung der Kastellnamen in Prokops Werk „De aedificiis”. Amsterdam 1970.

Dizdar M. (2012) The Archaeological Background to the Formation of Ethnic Identities. In: The Archaeology of Roman Southern Pannonia. The state of research and selected problems in the Croatian part of the Roman province of Pannonia. BAR International Series 2393. P. 117-136

Detschew D. Die thrakischen Sprachreste. Wien 1976.

Delamarre X. Dictionnaire de la langue gauloise. Paris 2003.

Duridanov I. (1980) Les Rapports Linguistiques Entre Les Thraces Et Les Celtes. In: NOI TRACII VII, # 68, 69, 1980. Mai (1) Juni (2)

Duridanov I. (1997) Keltische Sprachspuren in Thrakien und Mosien. ZCP 1997. B. 49/50. S. 130–142.

Falileyev A., Graham R. (2006). Remetodia. Acta Onomastica, 47 (1), pp. 173 – 177

Фалилеев А. (2009) KЕЛЬТСКИЕ ЛИНГВИСТИЧЕСКИЕ ОСТАТКИ ЮГО-ВОСТОЧНОЙ EВРОПЫ. In: Acta Linguistica Petropolitana – Transactions of the Institute for Linguistic Studies, Russian Academy of Science. Vol. V, part 1. Ed. N. N. Kazansky. St. Petersburg. 2009. P. 275 – 299

Falileyev A. (in collaboration with A. E. Gohil and N. Ward). Dictionary of Continental Celtic Place-Names. Aberystwyth 2010. (DCCPN)

Falileyev A. (2010) Ancient Place-Names of the Eastern Balkans: Defining Celtic Areas. In: In Search of Celtic Tylis in Thrace (III c. BC) Sofia 2010. P. 121- 129.

Георгиев В. (1977) Траките и Технят Език. София.

Holder A. Alt-celtischer Sprachschatz. B. 1–3. Leipzig 1896-1910.

Jovanović, B. 1984. “Les sépultures de la nécropole celtique de Pećine près de Kostolac (Serbie du nord).” Études Celtiques XXI: 63–93

Jovanović, B. 1985. “Nekropola na Pećinama i starije gvozdeno doba Podunavlja.” Starinar XXXVI: 13-18

Jovanović M. (1988) Čarnok, Titov Vrbas, keltsko utvrdenje. AP 25

Koch K.H. (1988) Existierte ein eisenzeitlichen Befestigungssystem im Gebiet der Treverer? In: Archäologisches Korrespondenzblatt 18. p. 169-182

Ljuština M., Spasić M. (2012) Celtic Newcomers between Traditional and Fashionable: Graves 63and 67 from Karaburma. In: Iron Age rites and Burial Rituals in the Carpathian Basin. Proceedings of the International Colloquium from Târgu Mureş 7-9 October 2011. P. 367-376

Mac Congail B. (2008) Kingdoms of the Forgotten. Celtic Expansion in south-eastern Europe and Asia-Minor. 4th – 3rd c. BC. Plovdiv

Paunov E. (2012) From Koine To Romanitas: The Numismatic Evidence For Roman Expansion And Settlement In Bulgaria In Antiquity (Moesia and Thrace, ca. 146 BC – AD 98/117) Phd. Thesis. School of History, Archaeology and Religion. Cardiff University. November 2012

Tapavički-Ilić M. (2007) Fortification System of the Scordisci. In: The Lower Danube in Antiquity. 6-7.10.2005, Tutrakan. Sofia. P. 83-90

Tomaschek W. (1894) Die alten Thraker. Teil II.2 // Sitzungsberichte der Wiener Akademie der Wissenschaften, Phil.-Hist. Klasse. Band 131. Wien 1894. S. 1–103.

Mac Congail