Contributions to the connections of the Vrap-Velino horizon and the Avar material

Avars, Bulgars and Magyars on the Middle and Lower Danube Editors Lyudmila Doncheva-Petkova – Csilla Balogh – Attila Türk Proceedings of the Bulgarian-Hungarian Meeting, Sofia, May 27–28, 2009 National Institute of Archaeology Pázmány Péter Catholic University and Museum BAS Faculty of Humanities and Social Sciences Sofia Department of Archaeology София − Piliscsaba 2014  The publication of this volume was funded by the generous grants from the National Cultural Fund (Nemzeti Kulturális Alap, NKA 3437/01022) and the Hungarian Scientific Research Fund (Országos Tudományos Kutatási Alapprogramok, OTKA PUB-K 111155, OTKA K 7106369, OTKA NK 72636) Front Cover Detail from Bowl No. 21 of the Nagyszentmiklós Treasure by Imre Huszár Editor-in-chief Lyudmil Vagalinski, Csanád Bálint, Endre Tóth Editors Lyudmila Doncheva-Petkova, Csilla Balogh, Attila Türk Translated by Hajnalka Pál, András Patay-Horváth, Tsveta Raichevska, Tatiana Stefanova, Vajk Szeverényi Illustrations Zoltán Pápai † © The Authors and Archaeolingua Foundation © National Institute of Archaeology and Museum BAS Sofia © Pázmány Péter Catholic University Faculty of Humanities and Social Sciences Department of Archaeology ISBN 978-963-9911-55-0 HU-ISSN 1215-9239 All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording or any other information storage and retrieval system, without requesting prior permission in writing from the publisher. 2014 archaeolingua alapítvány H-1014 Budapest, Úri u. 49. Desktop editing and layout by Gergely Hős Printed by Prime Rate Kft.  contents Csanád Bálint Foreword  ..................................................................................................................................................   7 Lyudmila Doncheva-Petkova Introduction  .............................................................................................................................................    9 Lyudmila Doncheva-Petkova Ethnic Changes in Present-day Bulgaria in the 6th–9th Centuries  ........................................................   13 Csilla Balogh Masque type Mounts from the Carpathian Basin  ..................................................................................   37 Miklós Makoldi Bulgaria – the Link between the Steppe and the Carpathian along the Danube  ...................................  55 Gergely Szenthe Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material  ................   61 Maria Hristova Similarities and Differences between the Pottery on the Lower and the Middle Danube (Based on Data Yielded by the Cemeteries)  ...........................................................................................   77 Ivo Topalilov – Kamen Stanev Two Bulgar Pagan Burials from Plovdiv  ................................................................................................   83 Nikolay Markov About the Characters on Jugs № 2 and 7 from the Nagyszentmiklós Treasure  ....................................   93 Pavel Georgiev The Abodriti-Praedenecenti between the Tisza and the Danube in the 9th Century  ............................   107 Valeri Yotov The Kunágota Sword and the Dating of Two Bronze Matrices for Sword-Hilt Manufacturing  .........  125 Stela Doncheva – Boyan Totev A New “Hungarian” Type of Saber from the Outer City of Pliska  ......................................................  133 Attila Türk Towards a Classification of Grave Types and Burial Rites in the 10th–11th Century Carpathian Basin (Some Remarks and Observations)  .........................................................................   137 Péter Langó Bulgarian Connections of the Find-horizon of the 10th Century in the Carpathian Basin: a Case Study  ..........................................................................................................................................   157 Tsvetelin Stepanov Bulgar, Avar and Khazar Aristocratic Names in the Early Middle Ages (Scytho-Sarmatian and Altay Heritage in Central and East Europe)  ...................................................   165 Miklós Takács Die Krise des Ungarischen Königtumsnach dem Tod König Stephan  ................................................   173 Ciprián Horváth New Data on Earrings with Beadrow Pendants: Grave 2 of the Cemetery of Kőszeg-Kőszegfalvi rétek  .................................................................................................................   189 Szabina Merva The Analysis of Pottery from 10th–11th-century Graves in the Carpathian basin. Technological and Typo-chronological Studies  ....................................................................................   197 List of Abbreviations  .................................................................................................................................   263 Avars, Bulgars and Magyars on the Middle and Lower Danube София − Piliscsaba 2014, 61–76 Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material Gergely Szenthe Recently the Byzantine origin of the Vrap treasure the manufacturing technique. Due to my research and of closely connected Bulgarian finds has become field, discussion is restricted to artifacts displaying an axiom in Hungarian and European research, since floral and geometric decoration. the theories of an Avar origin have become outdated In advance it has to be emphasized that the com- (Kiss 1995, 101–102; Garam 1997, summarizing the parison between the mounts and ornaments from Bulgarian and Western literature: Fiedler 2008, 218– the Carpathian Basin and the Balkan is hindered 220). In contrast, Bulgarian researchers emphasize by their different context. While the archaeolog- the local, Lower Danubian relationships and ascribe ical material from the Carpathian Basin abounds the treasure to the Danubian Bulgars (summarized in average quality belt mounts, these are relatively in Станилов 2006, 114–157). The direct Byzantine rare finds on the Balkans and it seems more likely connections of the Vrap–Velino horizon, which was that they have belonged to the elite compared to the denominated after the two main sites, are undisputed; Avar materials. The number of bronze mounts com- however, I suggest that the comparison of the hori- pared to the number of precious metal belt mounts zon with the archaeological material from the Car- in Bulgaria is relatively high, yet it pales beside the pathian Basin could yield intriguing data concerning material from the Carpathian Basin. The artifacts the internal connections of the two groups of finds from Bulgaria are mostly stray finds and therefore and on their Byzantine (or Mediterranean) links, and they are hardly comparable (or only certain aspects eventually on their origin. can be compared) – similarly to the Vrap-like elite In order to compare the Vrap–Velino horizon culture – to the artifacts from the Carpathian Basin, and the material from the Carpathian Basin first the majority of which were found in graves. Still, of all the characteristics of South European finds, it seems reasonable to correlate the Avar and Bul- then those of the Carpathian Basin are briefly gar finds, as besides their distinct contexts several introduced. First and foremost ornaments are resemblances can be discovered regarding their for- described, secondly the mount shapes, and finally mal features. General characteristics Concerning mount shapes, no underlying differ- The Vrap–Velino horizon (Figs. 2–5) is circum- ences can be detected between the Balkan material scribed by its ornaments: in all cases it is a bas-relief and the finds from the Carpathian Basin. In both that avoids giving depth to the representation and regions concave suspension mounts, thin, sheet belt employs a restricted spectrum of motifs. The motifs mounts can be found; however, traces on their back are cut out conically so the surface of the motifs plate imply that these were frequently cast. Further- determines the plane of the surface of the object (the more, cast, two-sided, open-work, U shaped belt bands are not interweaving, instead they break off). ends and their two-piece variants, rectangular and Bas-reliefs are traditionally associated with shield/or oval shaped mounts with pendant rings, carvings, in the examined time horizon their clos- hoof shaped belt hole guards, etc. are present (on est parallels can be seen on a vast number of carved the types and chronology of belt mounts see Daim bones (quivers, saddles and needle cases); neverthe- 2000, 184, Abb. 112). less, bas-reliefs appear in Mediterranean sculpture.1 The concurrences in shape in the two regions are The fundamental motifs of the Vrap horizon accompanied by considerable chronological differ- are the circular lobe ornaments, the succulent and ences (see under Chronology). sickle-like leaves organized into palmettes and half 1 On Avar bone carvings see Straub 1997; K iss 1996–1997, on Mediterranean sculpture see e.g. Wamser 2004, 76, 94, Cat. Nr. 98. 62 Gergely Szenthe palmettes with one to three leaves. These palmettes characteristic is that the borehole emphasizing the are arranged into symmetric compositions; they curve of the leaves is situated on the surface of the run along wavy lines and whirlings or two half pal- leaf and it does not separate the stock of the leaf and mettes build up a simple palmette-tree. A significant its folded-back tip.2 Absolute and relative chronology The chronology of the Vrap–Velino horizon is still and these items were used till the end of the middle not properly cleared. The chronology presented by third of the century at the latest. The Vrap, Erseke, Falko Daim is based on the accurately dated Avar Divdyadovo, Târgşor and Gledachevo finds in fact specimen’s formal parallels and on ornamental precede the Velino, Kamenovo belt ends (Станилов parallels. (Daim 2000, Abb. 112). The Vrap 2006, 92, Abb. 2) and their parallels. (Werner 1986; Garam 1997; Станилов 2006, The Avar material from the Carpathian Basin 108, Abb. 13), Erseke (Sotheby 1981; Stadler shows broader differences than the Vrap–Velino 1988–89, Taf. 1–3), Shumen, Divdyadovo quarter group. The Vrap-like main belt strap ends with a (Ataнacob et al. 2007, Fig.  2.) and Târgşor spout, cast in one piece, and the supplementing rect­ (Станилов 2009, 147, обр. 2) strap ends and belt angular shaped mounts decorated with griffins, strap mounts and their parallels are assigned to the holders with cast cover and their closest parallels can first half of the 8th century, while the Velino type be dated to the first half of the 8th century; the over- strap ends consisting of two parts are dated to the whelming proportion of the two-part strap ends – second half of the 8th century (Fig. 6; Daim 2000, contrary to the Velino find – with attachment lugs Abb. 112). emerge in the second half of the 8th century.3 Although Falko Daim’s relative chronology is Presumably, spouted, one-piece cast strap ends logically adequate, the absolute chronology might were continuously manufactured from the middle be narrower. The sole difference between Velino of the 8th century, while at the end of the century and Vrap type of finds is that Vrap strap ends are – now with utterly different decoration and mainly cast in one piece, while Velino type strap ends con- with attachment lugs – their use is predominant sist of two identical plates. The two find groups again. The motif pattern applied by the two relief are chronologically correlated by the treatment of techniques (high and bas-relief) is entirely distinct. space and surface, their ornaments, and the use of Following the relative variety of – at least in intent – bas-relief technique (e.g. mounts with griffins, with high relief representations in the first half of the 8th identical framing, geometric and floral motifs – on century (the horizon of two-part belt mounts) dec- the latter two see below), despite of the fact that orations start to resemble to the ornaments of the one can rely only on the Avar material as a basis Vrap–Velino horizon; however, the variety of motif for comparison. So far sheet belt mounts have not patterns is reduced. The ornament range is based been found among the Velino type artefacts and on floral motifs cut from the plane (on decoration this fact implies that even though there is a chro­ and execution [in bas-relief] see below). In the Car- nological difference between the two groups, in the pathian Basin at the second half of the 8th century, light of the common features it cannot be half a cen- a previously unrivalled variety of motifs appeared tury. Due to this evidence the Vrap–Velino hori- (Szőke 1974, 45–63); these are again characterised zon probably dates to the first half of the 8th century by multi-dimensional depictions. Decoration, mount shapes Although mount shapes are generally similar in unknown in the Carpathian Basin; furthermore, the the Balkan and the Carpathian Basin, some shapes decoration system is somewhat different. and technical solutions present in the Balkan are 2 It has to be noted that a number of mounts from the Erseke find are significantly different, especially those pieces on which the thickness of the sickle shaped scrolls is equal to the thickness of the stem. The treasure’s origin is debated; it comes from an unknown provenance. 3 On the relative chronology of Avar material culture see Garam 1995. Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material 63 The asymmetrical belt mount from Sofia has be noted that these assemblages contain some ear- close formal and stylistic parallels only on the mar- lier belt mount shapes, which were in fashion in the gin of the Carpathian Basin (for a summary see first half of the 8th century.8 Consequently, a well- Daim 2000).4 Likewise rare, but not uncommon is documented time gap arises between the Vrap hori- the adaptation of attachment lugs (or rivets) on the zon, dating to the first half of the 8th century, and reverse of the mounts, that are cast together with the the appearance of similar ornaments in the Car- artefact. This feature is generally regarded as a sign pathian Basin. of Mediterranean provenance. In contrast to the riv- The list of exact parallels of the Velino type, eted attachment, artefacts with attachment lugs spo- two-pieced strap ends’ ornaments is apparently radically appear among the Avars. Although their shorter in the Avar settlement area (see footnote 8). use is not preferred, in some cases they appear on The number of objects that, alike the Balkan group, certainly locally manufactured specimens (e. g. exhibit bas-relief-like decoration, is higher, the Orosháza, Szeged, Áporka, Nagypall). The techni- spectrum of the applied motifs, however, is even cal background, the mount shape and the ornaments more restricted. The designs on the evidently uni- associated with some shapes display a number of formed belt fittings of the late flat scroll horizon concurrences. A case in point is the belt hole guard in the Carpathian Basin are dominated by the flat from the Püspökszenterzsébet ensemble (today scroll leaf folded back in circle and simplified into Erzsébet, county Baranya): its exact formal paral- an acanthus hook, that are occasionally integrated lel is the specimen from Izvorul (Rumania, north into two- or three-leafed palmettes or half-pal- of the Lower Danube), which belongs to the Balkan mettes (Fig. 7. 2, lower part). According to the shape group.5 of the panel to be decorated the scrolls are orga­ Decorations display a definitely wider vari- nized into infinite friezes running along wave lines ety than belt mount shapes, still the overall pic- or they fill a circular or trapezoid shaped field in ture is similar but more complex. Specimens shar- axially symmetric pairs or in fours. The close par- ing common characteristics with the Vrap6 and allels of the Velino group are partly contemporary Velino7 group (floral ornaments) have appeared and even appear together. A scroll decorated belt set in several sites in the Carpathian Basin. Accord- was found during the construction of the Szeged- ing to the accompanying grave goods from undis- Fiume vasútvonal in the beginning of the twenti- turbed graves, those examples from the Carpathian eth century that contains a belt hole guard with axi- Basin whose formal features are identical with the ally symmetric scroll work similar to the Bulgarian Vrap belt mounts must have been buried in the sec- finds. This assemblage can be dated with certainty ond half of the 8th century. Nevertheless, it has to to the second half of the 8th century. 4 Áporka-Ürbőpuszta, Grave 20: buckle (Bóna 1957, XXXV. T. 2. The back plate of the mount not shown!); Orosháza- Bónum téglagyár, Grave 105: buckle (Juhász 1995, Taf. VII); rectangular belt buckle with griffin from the vicinity of Szeged (unpublished, HNM); Nagypall I, Grave 16: rectangular belt mounts with griffin and floral ornament (K iss 1977, Pl. XXVIII); Keszthely, stray find: buckle with rod-palmette (K iss 2005, with literature). 5 The belt hole guard from Püspökszenterzsébet (today Erzsébet, Baranya county; Hampel 1905, Taf. 254) is the exact parallel of the Izvoul find, however, it was attached by rivets instead of attachment lugs. The parallels collected by Stani- slav Stanilov extend the group to such an extent that several other types of finds from the Carpathian Basin would belong to it. I would add a few parallels to the presumed Vrap and Izvorul specimens enumerated by the author, that cover the entire Avar settlement zone: Szeged-Fiume vasútvonal, stray find (Hampel 1905, Taf. 95); Körösladány, Grave 10 (Fettich 1930, 209, 135. kép); Kaba-Bitózug, Grave 87 (Nepper 1982, 12. kép); Tiszafüred-Majoros, Grave 536/a (Garam 1995, Taf. 100); Szentes-Lapistó, stray find (Csallány 1934, 1. tábla). Therefore, I believe it is more appropri- ate that when the relationships of small and (consequently) simple objects are defined, only their exact parallels should be taken into consideration. 6 Alattyán-Tulát, Grave 170: main strap end (Kovrig 1963, Taf. XIV); Dalj (Dálya), broken, stray find: main strap end (Dimitriević et al. 1962, 111); Gyód, Grave 74: main belt strap (K iss 1977, Pl. X); Erzsébet (Püspökszenterzsébet), stray find: belt hole guard (Hampel 1905, Taf. 254); Leobersdorf, Grave 93: buckle (Daim 1987, 373. Taf. 95); Orosháza- Béke TSz-homokbánya, Grave 82: buckle, the exact parallel of the Leobersdorf specimen (Juhász 1995, Taf. XVIII); Tiszafüred-Majoros, Graves 199, 1084 and 1221: buckle, belt hole guard and wide, shield-shaped mount (Garam 1995, Taf. 74, 147, 161). 7 Regöly, Grave 119: buckle (K iss 1984, 77. t.); Szeged-Fiume vasútvonal: belt hole guard (Hampel 1905, Taf. 250). 8 Grave 199 in Tiszafüred-Majoros (Garam 1995, 30. Taf. 74) could be slightly earlier than the other two on the basis of its position and environment; still, horizontal statigraphy definitely dates the grave to the second half of the 8th century. Alattyán-Tulát (Kovrig 1963); Leobersdorf (Daim 1987) and Orosháza-Béke TSz-homokbánya (Juhász 1995) show a similar situation; however, the position of the graves with belt mounts is not as obvious as in Tiszafüred. 64 Gergely Szenthe From a formal point of view (taking into con- solutions (simple floral motifs in bas-relief) appear sideration style, namely bas- or high-relief, and with a substantial delay among the Avars. the variety of motifs) the difference between the The chronological difference in the two regions archaeological material of the first and second half can be explained by the location of the influential of the 8th century in the Carpathian Basin is nota- centres applying the enlisted motifs and designs bly broader than the discrepancy between the Vrap (Fig. 1) and by the different social background of specimens, traditionally dated to the first half of the owners of the Balkan and Avar finds. the 8th century, and the Velino specimens, suppos- edly from the second half of the 8th century. It can- Belt-mounts of the Vrap-Velino group not be a coincidence that the belt mounts from the Vrap-VStray finds of the Vrap-Velino Carpathian Basin reflecting the ornamental tech- Group in Bulgarian collections niques of the Vrap finds can be dated to the second 0 50 100 km half of the 8th century. Therefore, these exemplars are chronologically close to the Velino-like pieces from the same area regarding the design and motif pattern. Consequently, it is highly probable that the Vrap and Velino finds are not separated by half a century and their latest specimens are from the mid- third of the 8th century. As mentioned above, the use of bas-relief is almost unknown in the first half of the 8th century among the Avars (or appears only on simple, small- scale objects probably due to the limits of forming), whereas it can be found on the contemporary Balkan belt-mounts frequently. Consequently, the floral pat- terns have either sharp or rounded stems, while the surface of the leaves appear to be three-dimensional; Fig. 1: Sites of the Vrap –Velino group their middle is lowered most of the time. The range of floral and the supplementing, or even combined, geo- The motif pattern exhibited on the Balkan finds metric ornaments is wider in the first half of the 8th can be subscribed to the direct cultural influence of century. Beside the dominance of the flat scroll other the Byzantine Empire or of the Dalmatian coastline kinds of leaves, several types of flowers, cornucopia, (Dyrrhachium). Due to the geographical distance and curling and blooming stems make their appearance the political situation unfolding in the 8th century, as as well, sitting along wavy lines or along a straight described in the written sources, there is nothing to axis (Fig. 7. 1. 3; upper part of Fig. 7. 2). Palmette suggest direct influence on the Carpathian Basin. trees with two or three leaves become general. However, the Avars could have been affected from The decrease in the range of motifs and the exe- the southwest (from Italy and Dalmatia, Daim 2000, cution in bas-relief (stylised scroll work cut from the 180), nevertheless, and primarily Mediterranean and plane, with the hole on the leaves’ surface stress- late antique influences were transmitted. The altered ing the curve of the leaf tip) appear later among use of shapes and motif system in the Balkan and the Avars, in an even more reduced form. It seems the Carpathian Basin are the result of divergent local that there is a chronological difference between the tastes. A further interpretation of the chronological Vrap–Velino horizon and the material from the Car- delay could be that the Balkan finds reflected Byzan- pathian-Basin: the latter follows the Balkan finds tine luxury industry as these belonged to an elite cul- with a short delay and in a distinct system. While ture, whereas the majority of the Avar bronze arte- the shapes of the uniformed Vrap group appear in facts were possessed by lower layers of society and the Carpathian Basin at the beginning of the 8th innovations in their material culture appeared later. century or even earlier (the earliest ones are the The underlying difference between the Avar sheet belt mounts from the 8th century: Vrap type, material dated to the first half of the 8th century and spouted mounts, mounts with griffin, strap holders, its South European counterpart is the vast amount of wide shield-shaped and oval mounts), other formal Avar belt mounts and the variety of motif patterns.9 9 Because of their vast number only a few motifs are introduced here. The illustrations – the motifs and the designs – are part of my doctoral thesis. In the Avar material floral motifs appear predominantly on (simple or symmetrical) curling stems or on simple palmette trees. Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material 65 A number of exact formal parallels draw our atten- structural or originate from the divergent dynam- tion to the fact that dividing the two groups sharply ics of cultural development in geographically dis- is improper: the essential distinctions are rather tant regions. Manufacturing technique Traces of the manufacturing technique can be textile stiffening the wax model that was retained studied singularly on the semi-finished pieces of on the mould. This textile imprint can be discov- the Vrap treasure (Figs. 8. 1–2). The edges of the ered on the contemporary lower Danubian material mounts are framed by thin welds; the surface of the (Инкова 2007, 238),13 however, this feature is also unchased mounts as they come out of the mould present on the geographically and chronologically form is smooth in contrast to the moulding chan- close Biskupije press mould (7th century) (Korošec nel’s highly porous surface transmuted into metal. 1958, T. 3) and on the reverse of Viking, Scandi- The section of the sprue is triangular and its end is navian cast metal objects (9th century) (Korosuo stained, rounded. 1946). Chronologically remote examples are the Based on the fact that the welds surrounding South Siberian Scythian bronze casts, still, these the mount10 are situated in one plane it could be illustrate the pristine origin of the technique (e.g. concluded that the mounts were produced in two- Wagner–Butz 2007, Cat. Nr. 1). piece moulds.11 Other traces reveal that the mate- To my knowledge, utilizing a piece of textile rial of the two-piece moulds might have been sand. during model-making has no precedent in antique The sprues also support this idea: their porous sur- casting tradition. On the contrary, the enumerated face and triangular section inevitably refers to the examples suggest that this know-how was applied practice of sand casting, where the mould chan- by “steppe cultures” and appeared in their contact nels are “cut on”. The section of the mould chan- zones. Its use was required to multiply the wax mod- nel is triangular or rhomboid, seldom len-shaped, els in order to produce identical series by lost-wax depending on the shape of the tool. The surface at casting. When making a wax model in a two-piece the removed material becomes more porous than clay mould, a piece of textile was placed into the wax the surrounding, stamped surface. In the case of filling of the negative front, and then the reverse was lost-wax casting, a spruing system consisting of pressed onto the front side. After the wax solidified, wax pipes is attached to the wax copy, therefore the textile strands stiffened the thin, fragile model. their section is circular or oval, and the fineness of The finalized model (following surface smoothing, the surface is standard. chasing) was covered into moulding clay (Brepohl On the belt mounts of the Vrap–Velino group 1987, 61–68) – in the 12th century the clay was mixed – maybe because these are elaborate objects of an with horse manure and hairs –, then fired so the wax elite culture12 – a wide-spread peculiarity present melted and its empty place was filled with bronze. in the 8th-century Carpathian Basin and the Bal- The overwhelming proportion of Avar cast kan cannot be observed. On the unchased reverse bronze artefacts was undoubtedly produced by lost- of cast bronze Avar objects a bulging, positive tex- wax casting, which agrees with our knowledge tile imprint can be noticed that is the imprint of the on early medieval casting techniques:14 practically 10 It is important to emphasize that only welds running along the longitudinal axis of the object implicate two-pieced moulds. Welds come into existence during lost-wax casting as well, if the pressure of the melted metal cracks the mould. These welds are – generally – not as regular as on artefacts cast in two-piece moulds, since int he latter case welds appear at the juncture of the mould halves. 11 The moulds used for the manufacturing of the discussed artefacts are always closed. 12 E.g. the smoothing marks on the back plate of the Velino main strap end. 13 Although I hardly know marks of casting on Bulgarian finds – partly because of their chased surface apparent on pho- tographs – it is probable that these were produced with the same technique as their Avar counterparts. In the light of the formal similarities it cannot be a coincidence that on the back plate of a number of Avar main strap ends the same longi- tudinal carving and chasing marks can be noted as on the Velino find. 14 Some works written in the second third of the 12th century that deal with Early Medieval casting suggest that two-piece clay and sand moulds were in use, however, their arguments are not satisfactory. I. Erdélyi (Erdélyi 1958, 69–73) and N. Fettich (Fettich 1962, 105; Fettich 1990, 129–130) suggest that Avar objects were cast in sand moulds. Still, it is unlikely that sand or fired clay two-piece moulds would have been applied – even though in the early medieval Car- pathian Basin so far moulds appropriate for producing belt mounts or other Avar objects have not been discovered. 66 Gergely Szenthe only lost-wax cast artefacts are known. However, Unlike lost-wax casting and the use of two- besides the Vrap mounts in certain cases the marks piece clay or talc moulds to produce certain arte- of the producing technique are ambiguous, like facts, applying two-piece sand moulds is unknown the Szentes-Lapistó belt mount set with griffins in the antique casting tradition. The same statement (Erdélyi 1959, 72). Even though on the back plates can be made concerning the textile imprints on the textile imprints characteristic to lost-wax casting reverse of artefacts; however, this – contrary to the can be noted (Fig. 8. 3a–b), the back plate of the belt previously probably unknown sand casting – is the mounts are identical (because of the direction of the side product of lost-wax casting since the Scythian textile strands and their place). This phenomenon is Period. The textile imprints appearing in Central so far unique in the Avar material. If it means that and Southern Europe in the Early Middle Ages must this set was transmuted of one original, then lost- be the result of a technique coming from sources wax casting as a producing technique could be different from the antique metalworking tradition; excluded (see above).15 presumably these originate from the steppe. Despite of this feature, other traces on the sur- If the cases in point suggest the use of sand face of the artefacts suggest that the mount series moulds, this technique emerges in Southern and were cast by lost-wax technique. Central Europe without having traces in either the An archaeological feature from Mikulčice can antique or the steppe metalworking tradition. A prove the existence of sand casting in the Early probable interpretation is that the provincial Byzan- Middle Ages, in which three kinds of sand were tine Vrap finds were manufactured with a technique found, with different fineness, and were separated still unknown to the north and northeast of the by dark layers (Profantová 1992, 652). In order to Mediterranean, and that counted as a novelty even build a two-piece sand mould, sand types of differ- at their site of production. If the marks on the Vrap ent quality are beaten onto a box nowadays as well. pieces are results of sand casting, the technique The sand is sorted according to its granularity and must have been an innovation in the Mediterranean. is stored in caskets covered with damp leather to It could be a reason for the fact, that the small num- keep it evenly humid. ber of examined or examinable original Byzantine artefacts were produced by lost-wax casting.16 Summary The comparison of the archaeological material from patterns according to one’s taste, nonetheless, it is the Carpathian Basin and from Southern Europe apparent that the Vrap elite culture could imitate the reveals close analogies in form, execution, the use Byzantine elite’s costume directly, in contrast to the of the bas-relief technique, the used motifs and more distant and poorer Avar material. probably the production technique, although chron- It cannot be a coincidence that the know-how of ologically these are not completely parallel. Never- artisans, which is transfered slower than the range theless, the dynamics of change and the structure of belt mount shapes, shows resemblances in the of the archaeological material are fundamentally Carpathian Basin and on the Balkan Peninsula. divergent. Only the Vrap mounts may differ from this picture Due to the political situation in the region (the as visible production traces imply sand casting. In dominance of the Byzantine Empire) the artis- this case these finds may illustrate how (provincial) tic sources of the Vrap–Velino group and the Car- Byzantine culture influenced the bordering, Bar- pathian Basin could not have been fundamentally barian regions: applying a two-piece sand mould is different; hence the dissemblances derive from dis- uncommon in steppe metalworking traditions. tinct regional preferences and social reasons. One In the case of the Albanian and Bulgarian could select from the canonised mounts and motif mounts it is probable that their owners belonged 15 It is also possible that the textile was already there in the original and its imprint was transferred via model-making to the negative and then to the wax models. 16 Examinations were carried out by M. Fecht (Fecht 1988, 309–312). It can be confusing that referring to the cover- ing of the model the author uses the term “Formsand” (“moulding sand”) because of surface fineness, instead the usual “Formerde” (moulding clay). The examined Byzantine gold buckle was produced by lost-wax technique because of the overlapping details of its surface. Several implications to sand casting in the early medieval Period come from the doubt- ful interpretation of difficult-to-understand archaeological phenomena (e.g. Capelle 1974, 295–296). Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material 67 to the elite and imitated Byzantine military cos- manufacture could have taken place in the provin- tume (Daim 2000, 180), or – similarly to 7th cen- cial Dyrrhachium, while that of the Bulgarian finds tury gold pseudo-buckles – ready belt mounts (sets) in the Lower Danube region.17 were obtained from Byzantine regions. However, the majority of the mounts must have been pro- duced locally. With regard to the Vrap treasure, its Translated by Vajk Szeverényi Bibliography Ataнacob et al. 2007 Daim 1987 Г. Aтанасов – C. Венелинова – C. Cтойчев: F. Daim: Das awarenzeitliche Gräberfeld von Ранносредновековен некропол в Шумен Leobersdorf. Studien zur Archäologie der (квартал Дивдиядово). – Early medieval Awaren 3. Wien 1987. cemetery in Shumen, Divdyadovo quarter. Daim 2000 Aрхеология 48 (2007) 57–66. F. Daim: „Byzantinische” Gürtelgarnituren des Станилов 2006 8. Jahrhunderts. In: Die Awaren am Rand der C. Станилов: Художественияат метал на byzantinischen Welt. Hrsg.: F. Daim MFMA 7. българското ханство на Дунав (7–9. век.). Innsbruck 2000, 77–204. – Die Metallkunst des Bulgarenkhanats an der Dimitriević et al. 1962 Donau (7–9. Jh.). София 2006. D. Dimitriević – J. Kovačević – Z. Станилов 2009 Vinski: Seoba Naroda. Arheoloski nalazi C. Станилов: Към идентификацията на Jugoslovenskog Produnavlja. Zemun 1962. гроба от Тиргшор. – About the Identification Erdélyi 1958 of the Grave from Turgshor. B: Laurea. In Erdélyi I.: A jánoshidai avarkori temető. honorem Margaritae Vaklinova I. Ред.: Е. RégFüz Ser. II. № 1. Budapest 1958. Василева – Б. Петрунова – А. Аладжов Fecht 1988 София 2009, 145–149. M. Fecht: Untersuchungen zur Herstellung­ s­ Инкова 2007 technik und Gestaltung einer byzantinischen M. Инкова: Ранносредновековни накити от Goldschnalle in der Prähistorischen Staats- НИМ. – Early Medieval Jewellery from the sammlung München. BVbl 53. München 1988, National Museum of History. B: Иследвания 309–312. по българска средновековна археология. Fettich 1930 Ред.: П. Георгиев. София 2007, 224–239. Fettich N.: A Magyar Nemzeti Múzeum Gróf Bóna 1957 Vigyázó ásatásai. Népvándorláskor. ArchÉrt Bóna I.: Az ürbőpusztai avar temető. – Le 44 (1930) 205–211. cimetière avar de Ürbőpuszta. ArchÉrt 84 Fettich 1962 (1957) 155–174. N. Fettich: Das awarenzeitliche Gräberfeld von Brepohl 1987 Pilismarót-Basaharc. StudArch 3. Budapest E. Brepohl: Theophylus Presbyter und die 1962. mittelalterliche Goldschmiedekunst. Leipzig Fettich 1990 1987. Fettich N.: A bánhalmi avar leletekről. SzMMÉ Capelle 1974 7 (1990) 123–137. T. Capelle: Die karolingisch-ottonische Bronze­­- Fiedler 2008 giessersiedlung bei Kückshausen. FMS 8 U. Fiedler: Bulgars int he Lower Danube (1974) 294–303. Region. A Survey of the Archaeological Csallány 1934 Evidence and of the State of Current Research. Csallány D.: A Szentes-lapistói népván­ dor­ In: The Other Europe. Ed.: F. Curta. Princeton láskori sírlelet. – Der Grabfund von Szentes– 2008, 151–236. Lapistó aus der Völkerwanderungszeit. Dolg 9–10 (1933–34) 1934, 206–214. 17 According to the latest publications, this is conditionally possible, see Fiedler 2008, 219–220. 68 Gergely Szenthe Garam 1995 Nepper 1982 É. Garam: Das awarenzeitliche Gräberfeld Nepper I.: A Kaba-bitózugi avar temető. – von Tiszafüred. Cemeteries of the Avar Period Das awarische Gräberfeld von Kaba–Bitózug. (567–829) in Hungary Vol. 3. Eds.: Kiss, A. – ComArchHung 1982, 93–123. Garam, É. Budapest 1995. Profantová 1992 Garam 1997 N. Profantová: Awarische Funde aus É. Garam: Über den Schatzfund von Vrap in den Gebieten nördliche der awarischen Albanien. Acta ArchHung 49 (1997) 23–33. Siedlungsgrenzen. In: Awarenforschungen II. Garam 2000 Hrsg.: F. Daim. Studien zur Archäologie der É. Garam: The Vrap Treasure. In: From Attila Awaren 4. Wien 1992, 605–801. to Charlemagne. Arts of the Early Medieval Sotheby 1981 Period in the Metropolitan Museum of Art. The Catalogue of the Avar Treasure. London 1981. Metropolitan Museum of Art Symposia. New Stadler 1989 York 2000, 170–179. P. Stadler: Argumente für die Echtheit des Hampel 1905 „Avar Treasure”. MAG in Wien Band 118–119 J. Hampel: Alterthümer des frühen Mittelalters (1988–1889) 1989, 193–217. in Ungarn I–III. Braunschweig 1905. Straub 1997 Juhász 1995 Straub P.: Avar kori tegezdíszítő csontleme- I. Juhász: Awarenzeitliche Gräberfelder in der zek. Motívumtipológia és kronológia. – Awa- Gemarkung Orosháza. MAA 1. Budapest 1995. renzeitliche Beinplatten als verzierungen von K iss 1977 Köchern. Typologie der Motive und Chronolo- A. Kiss: Avar Cemeteries in County Baranya. gie. MFMÉ – StudArch 3 (1997) 117–151. Cemeteries of the Avar Period (567–829) in Szőke 1974 Hungary Vol. 2. Ed.: Kovrig, I. Budapest 1977. B. M. Szőke: Über die späthellenistischen Wir- Kiss 1997 kungen in der spätawarenzeitlichen Kunst Kiss A.: Az avar kori tegezek csontlemezei- des Karpatenbeckens. Eine kritische Untersu- nek kronológiájához. – Zur chronologie der chung. DissArch Ser. II. No. 3. Buda­pest 1974. Beinbeschläge der Awarenzeitlichen Köcher. Wagner–Butz 2007 ArchÉrt 123–124 (1996–1997) 1997, 75–84. M. Wagner – H. Butz: Nomadenkunst. Ordos- K iss 1984 bronzen der Ostasiatischen Kunstsamm- Kiss G.: A regölyi temető. In: Kiss. G. – Somo- lung Museum für asiatische Kunst, Staatli- gyi P.: Tolna megyei avar temetők. DissPann che Museen zu Berlin. Archäologie in Eurasien Ser. III. Vol. 2. Budapest 1984, 123–150. Band 23. Mainz 2007. Kiss 2005 Wamser 2004 Kiss G.: Egy bizánci övcsat Keszthely-Dobo­ I. Wamser: Die Welt von Byzanz – Europas góról. – A Byzantine belt buckle from Keszt­ östliches Erbe. München 2004. hely-Dobogó. ZM 14 (2005) 203–214. Werner 1986 Korošec 1958 J. Werner: Der Schatzfund von Wrap in J. Korošec: Ostava brončanih matrica za Albanien. Studien zur Archäologie der Awaren otiskivanje u Biskupiji kod Knina. – The 2. Wien 1986. Depository of Bronze Matrices at Biskupija near Knin. Starohrvatska Orsvetja III/6 (1958) 29–44. Korosuo 1946 O. Korosuo: Gjuttekniken under järnåldern i Finland. FM 53 (1946) 5–30. Gergely Szenthe Kovrig 1963 Hungarian National Museum I. Kovrig: Das awarenzeitliche Gräberfeld von 1088 Budapest, Múzeum krt. 14–16. Alattyán. ArchHung 40. Budapest 1963. e-mail: szethe.gergely@gmail.com Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material 69 2 3 1 4 5 6 7 8 Fig. 2: Vrap (Albania) (after Garam 1997, Abb. 1–2) 70 Gergely Szenthe 1 4 2 5 3 6 7 Fig. 3: Vrap (Albania) (after Garam 1997, Abb. 1–2) Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material 71 1 2 4 3 5 6 8 10 7 9 Fig. 4: Albania (Erseke?) (after Sotheby 1981) 72 Gergely Szenthe Fig. 5: Belt mounts (Bulgaria) (after Stanilov 2006)  Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material Fig. 6: Chronology of the Vrap–Velino Group: “Byzantine” belt sets (8th century; after Daim 2000) 73 74 Gergely Szenthe 1 2 3 Fig. 7: 1–2: Leaf motifs (Carpathian Basin, 8th century); 3: Floral motifs (Carpathian Basin, 8th century) Contributions to the Connections of the Vrap–Velino Horizon and the Late Avar Material 75 1 2 3a 3b Fig. 8: 1–2: Semi-finished or spoiled casting with welds (Vrap, after Garam 1997, Abb. 2); 3–4: Belt mount with a griffin, front and back (Szentes-Lapistó)