The Avar Period (567-827)

The archaeological legacy of the one-time Avar Empire that extended over eight present-day countries in the Middle Danube region includes over two thousand sites with tens of thousands of burials — some of the completely excavated cem eteries contained several hundred or even over one thou­

sand graves — as well as hundreds of house remains from Avar villages.

Supra-national co-operation among archaeologists is admirable even if there is often a lack of consensus on minor details of chronology and ethnic attri­

bution. Co-operation of this sort, however, is sadly lacking in the case of the eastern part of the Avar Empire. Only thirteen graves of the Avar cemetery at M arosgombas in Transylvania — the first burial ground to be excavated by an expert — could be uncovered in the year preceding the outbreak of the First World W ar (M&rton Roska, 1913). (By that time, the num ber of excavated and published Avar graves was already in the thousands in H un­

gary.) Roska was never to publish his findings. The same fate befell the excavation conducted at TOvis by S&ndor Ferenczi in 1938, on the eve of the Second World War: its only result was about fifty, still unpublished burials.

Since then, Avar graves have at most been investigated as an unwanted

"by-product" of excavations conducted on sites thought to belong to an­

other period, or in the course of unavoidable rescue excavations on a few other sites.

It is not difficult to account for why archaeological research on this pe­

riod was neglected and later, was outright discouraged. The twelve sites that have also yielded horse burials — five of which are described as "soli­

tary" horse burials the likes of which are unknown outside modern Transyl­

vania proper — and which are listed as Avar finds in a recent summ ary are

interpreted as the evidence of a gossamer "m ounted nom adic" layer that ruled over the "autochtonous Roman" or, as other scholars would have it Slavic, population. It has been claimed that the archaeological finds in the Transylvanian Basin which can be associated with the Avars (finds which have been labelled as belonging to the "M ures group" or "G im bas group") must be dated to after the internal upheavals and transformations within the Avar Empire, that is, to after 670. In the same vein, the distribution of the Avar finds is restricted to an area not exceeding roughly forty by fifty kilom etres in size. These theories would have small Avar groups or com ­ munities living out their lives in some sort of splendid isolation until the fall of the Avar Empire, after which their fate is shrouded in mystery. Accord­

ingly, the period until 670 is filled out by these theories with "surviving"

groups of the above-mentioned B^nd-type Gepids and of the "rom anized"

population. Scarce and sporadic though they be, this interpretation of the Avar finds in Transylvania is not only contrary to the historical evidence, but is also inconsistent with the archaeological evidence that has come to light for that period.

Interestingly enough, Transylvania played a crucial role in the European history of the Avars well before this population of inner and central Asian origin overran the Carpathian Basin. After their conquest of the windy steppes to the north of the Black Sea littoral, the Avars who had by that tim e concluded an alliance with Justinian I, arrived to the Lower Danube region in 562 where their kagan, Baian, "pitched his tents in great numbers over the w ide plains".12 It is from this area that they asked to be admitted into the empire. Though the emperor refused to allow them to enter, he whetted their appetite with the possibility of seizing from the Gepids Pan­

nonia Secunda — a territory which would have made an excellent settle­

m ent territory for the Avars. Baian, however, sensed that it would be futile to even attempt to penetrate into the fertile lowland through the narrow passes of the massive southern Carpathians, or along the Iron Gates. The young, energetic kagan, therefore, decided on a bold and bizarre course of action. Leaving their camp behind, he and his horsemen skirted the eastern slopes of the Carpathian massif in search of a pass. However, they could find no route through which a people of migratory herdsmen and their cattle could cross the 1,500-2,000 metre high mountains covered with dense forests several hundred kilometres wide on either side. The few passes they might perhaps have tried were heavily guarded by the Gepids. W inding their way along the mountains, they finally reached the eastern fringes of the M erovingian Em pire (Galliae) w here Sigebert I, King of A ustrasia, checked their advance. The Avars were forced to turn back in 563. And since the new emperor of Byzantium, Justin II who ascended the throne at the end of 565, made no new promises and also stopped their supply of gold, the Avars made an abortive attempt to cross the Danube (winter of 565-566). It was at this point that Baian decided to try his luck again, and to try to penetrate into the fertile lowland beyond the Carpathians along the already tested, longer route.

12. C o rip p u s, In laudem Iustini III.

This time, his efforts were crowned with success. He defeated the Frank­

ish army and even captured Sigebert himself, who was, however, set free in the hope of his acting as mediator in a desirable alliance. The alliance in question was to be forged with the Pannonian Langobards which, given the particular political constellation, did not appear to be an overly difficult task. For the queen of the Langobards, the wife of King Alboin, was none other than Sigebert's sister. This temporary alliance was as m uch desired by the Langobards as by the Avars, pursued as they were by the wrath of their Turkic overlords. For in the Langobardic-Gepid war that had broken out in the previous year, Justin II supported Kunimund, the king of the Gepids and this Byzantine-Gepid league posed a threat to the very exist­

ence of the Langobards. It was thus that the Avar-Langobardic alliance (in which King Alboin undertook to surrender the lands of the Gepids to the Avars) was concluded by the two panic-stricken parties at the beginning of the year 567.

The Avar cavalry that was escorted by the Langobards into the Carpathian Basin through the Moravian Gates and was then shown the way to the Great Hungarian Plain from the Danube Bend inflicted a devastating defeat on the Gepids. The contemporary Byzantine chroniclers were fully aware of the fact that "the state of the Gepids had been crushed by Baian"13 — and not by the Langobards, as their chroniclers were later fond of boasting. Af­

ter the first unsuccessful attack on Sirmium in the autumn of 567, Baian occupied the Gepid territories on the left bank of the Danube. The first Avar warriors penetrated Transylvania from the west, along the Maros, and they finally cleared the way into the Olt Valley through the Vorostorony Pass for their kindred who, for the past five years, had been patiently w aiting on the other side of the mountains. The Avars devastated the country far and wide, and the majority of the Transylvanian Gepids probably perished at this time.

This is not as far-fetched a statement as it might first seem: it only appears to be nonsense to m odem , twentieth-century man who can only think in terms of countries with millions of inhabitants. The ruling elite of the na­

tion s of that age rarely exceeded a few tens o f thousands, and their populations rarely exceeded a few hundred thousand souls — which made the countries of that period very fragile indeed. At the close of the seventh century, the Geographer of Ravenna was pretty certain — for he mentions the fact twice — that the territory of one-time "D atia prima et secunda" or the older (magna) "D atia" which was also called "Gepidia ... is now popu­

lated by the Uns, also known as Avars''.14

Transylvania — but not the eastern parts of the Great Hungarian Plain, and the Temes region — practically lost its importance after the Avar con­

quest in 567-568, especially once the Langobards also ceded Pannonia to their formidable allies in the spring of 568. The pyre finds (burnt horse har­

ness, cheek bits, a pair of stirrups and a spearhead that were thrown onto the funeral pyre and later buried at a shallow depth) of inner Asian type which mark the earliest Avar occupation of the Carpathian Basin have only

13. M e n a n d e r P r o t e c t o r , Excerpta ... 27-28,195-198, 456-458.

14. Ravennatis anonymi Cosmographia ... 1 ,11.; IV, 14.

been found at the lowland exit of the Maros-Aranka (Nem etszentpeter), and in the Kis-Kukull6 Valley (Dicsoszentmarton). These stirrups can be classed among the oldest in Europe, similarly to their counterparts from H ungary. A num ber of early Avar guardposts can be assum ed on the Transylvanian side of the passes, especially in southern Transylvania, a fact suggested by a gold coin find of Khosrau I (531-579), the Persian king, which had obviously made its way to Brassb together with the Avars, as well as by the gold coins of Justin II (565-572) and of Mauritius Tiberius (582-602) found in the Sebes and Olt valleys. These Byzantine coins had in part been uspd as grave oboli by the Avars; the majority of these coins probably origi­

nates from looted graves whose find circumstances have not been adequately recorded. One proof for this is a solidus of Justin II found in a plundered burial of the small Avar cemetery at Isp^nlaka which is still unpublished, as are the other Avar graves that were uncovered there in 1970, and which contained also horse burials with gold or gilded grave goods. That an Avar centre had existed somewhere near the confluence of the Aranyos, Maros and Kukull6 rivers already before 600 is suggested by a splendid pair of gold earrings with a large gold bead pendant in a burial that was found at Torda; only fifteen similar pairs of comparable size and workmanship can be quoted from the entire Carpathian Basin.

The Avar commoners who occupied the windy plains and gently rolling hills of the Carpathian Basin could hardly have found the smallish Transyl­

vanian Basin hemmed in by mountains of dark woodland attractive. After the close of the first great wars against Byzantium (601-602), when the Avars finally realized that they had no choice but to make the Carpathian Basin their permanent home, traces of a new Avar centre in Transylvania have also been identified. The Szentendre and Deszk type gold earrings with pyram idal pendants "from Transylvania" — the exact findspot is not known

— suggest the appearance of nobles around and after 600. One centre can be located to Gyulafehervar, where silver harness ornaments and a contem ­ poraneous Byzantine buckle have come to light. Equally im portant are the press moulds from the Kiikulld valleys (Erzsebetvaros and Korond) that had belonged to Avar goldsmiths and which definitely prove that ornate belt and harness mounts testifying to the rank of their owner were m anu­

factured locally in Transylvania. Equestrian Avar groups who had buried their dead together with their horses settled in Nagyenyed and in the form er Gepid village of Mezdband where the Avar community interred its dead at the edge of the Gepid cemetery. Their burial rites and their costumes match those of the Pannonian Avars and the Avars of the Great Hungarian Plain down to the smallest detail. Aside from the above-mentioned sites, pre-630 Avar settlements are indicated by finds of Avar-type handmade pottery (Malomfalva, Marosnagylak, and Csap6szentgydrgy). However, not even the faintest traces of an early Avar habitation have been detected in the Szam os Valley. Obviously, this does not apply to the eastern part of the Great Hungarian Plain, the Er Valley, the M aros-Aranka triangle and the Temes region, areas that had been integrated into the Avar settlement terri­

tory at an early date and which yielded some highly informative assem ­ blages. Example of these are the horse burials and the grave group from Ermihalyfalva; the burial of a goldsmith together with his horse from Fonlak;

the warrior and horse burial from Nemetszentpeter into which were placed a helmet, a coat of mail, and a sword as well as a gold coin of Heraclius minted before 625; and in the south, the burials of women wearing silver earrings with large globular pendants from Oravicab&nya and the early Avar burials from Orsova.

The defeat suffered at the siege of Constantinople in 626, and the ill- fated Italian expedition in 628 led to a prolonged crisis in the Avar Empire beginning with 630. In the west and the southwest, a number of Slavic tribes shook off the Avar rule and gained their independence by fighting, whilst in the east their former Bulgarian ally turned against them and founded a new steppean empire. The internal strifes did not spare Transylvania ei­

ther, and the events of between 630-638 can be traced in the archaeological record. The MezSb&nd cemetery fell into disuse and almost all graves — including the Avar burials — were plundered. The only large hoard of gold coins from the Avar period (Korond-Firtosvcir) which contained at least 237 solidi, the last of which was minted in 625 in Byzantium and which gives the approximate date of the catastrophe, was hidden during these troubled times. Finally, there are those burials which came to light only in Transylvania within the Avar Empire, graves which yielded lance and spear­

heads w ith pierced blade of a form traceable to the Pontic steppeland, and graves of eastern Slavic women easily identified by their Slavic costume who had been the wives or relatives of the warriors alluded to above. East­

ern Avar and other, allied Turkic elements had probably fled beyond the mountains to Avaria when the new Bulgarian Kaganate was organized in 635 (Marosgombds and Tovis).

Only a handful of seventh-century Avar settlements are know n from Transylvania (Szdszsebes, M alom falva-Bors6fold, R ad n 6t-C sap 6szen t- gyorgy and Lapos, Segesvar-Sz<516k, and Bozod). The huts with sunken floors and their stone ovens hardly differ from the houses in the eastern part of the Great Hungarian Plain (Bihar), or the large Avar settlements else­

where in Hungary, nor are there substantial differences between the wheel- thrown and handmade pottery found in the fill of various settlements.

During the reign of Constantine IV, a little before the arrival to the Dan­

ube region of the Bulgarians, who had been pressed westwards by the em er­

gence and expansion of the Kazar Empire, the territory of the Avar Empire was flooded by eastern population groups. Among the new arrivals were O nogur-Bulgarians (Wangars), but mostly they were various Turkic groups

— including some Kazars — with whom the Avars shared a com m on an­

cestry, and who had been uprooted from their former abodes. In the north and the northwest the boundaries of the Avar settlement territory, as well as the political boundaries, shifted to beyond their former line. In the heart­

land of the empire, a number of settlements fell into disuse, and new ones, along with new cemeteries, were founded by the incoming population. From these it is evident that the new burial rite, the horse harnesses, the weapons and the styles of dress had no links whatsoever to the types current in the early Avar period. This new period is marked by grave finds dated with gold coins of Constans II and Constantine IV in the Temes region. No sim i­

lar coins, however, have been reported from Transylvania.

Still, the newcomers to the Avar Empire did, eventually, make their way to Transylvania, brandishing their novel weapon, the sabre, and wearing a new type of stirrup for their flat-soled boots; the horses had distinctive bits and phalerae, and pendant-ornam ented harnesses (01£hgorb6, Sz6kely- keresztur, Felenyed, Marosnagylak, and Csak6, where houses featuring a stone oven have also been uncovered). The cemetery at To vis was used by the newcom ers and they also started a new cemetery at Aranyosgyeres. On the testimony of a gold finger-ring, one of their leaders, a prince, estab­

lished his residence and was buried at Gyulafeherv^r. All in all, not much is known about this transitional period of about a quarter century. There are only indirect references to it in the Byzantine and Frankish sources or none at all. It is only recently that its real significance has been recognized. In fact, the archaeological evidence of this period reflects the em ergence of a new artistic style: a skilful blend of the traditions of earlier and later popu­

lation groups, a style that has aptly been termed "griffin-tendril" style after the motifs m ost frequently occurring on the cast bronze buckles and belt ornam ents. It was at this time that permanent villages were established;

under internal and external pressures, the peoples of the Avar Empire had turned to crop raising and animal husbandry, with semi-migratory pasto- ralism being the exception, rather than the rule. The majority of the known large Avar period cemeteries (about seventy to eighty per cent) contained the dead of four to five generations: village cemeteries with over one thou­

sand burials are by no means exceptional. This clearly reflects a change to a more stable way of life from the close of the seventh century onwards.

The presence of sim ilar late seventh-century and eighth-century cem ­ eteries in Transylvania is proven by horse burials described as "solitary graves" (Hari, Muzsnahaza, and Magyarlapad), as well as by griffin-tendril ornamented belt mounts, carved bone needle cases, and other contem pora­

neous ornaments such as earrings with star-shaped or lunular pendants (Tovis, Aranyosgyeres, Lesnyek, and Szentgyorgyvalya). Equally scanty is the evidence from the eastern Temes region (Tem esvar-M odosi Bridge), from the eastern fringes of the Koros-M aros interfluve (Szekudvar) and from the Er Valley. A settlement with huts featuring sunken floors and stone ovens, as well as timber-framed above-ground houses has been uncovered at Gyulafehervar-Stadion. The occupants of this settlement were mostly engaged in cattle raising; their vessels were partly handmade and partly wheel-turned, and bronze earrings with star-shaped or lunular pendants were also found. Late Avar huts with sunken floor and containing a stone oven have also been reported from Bihar. Unfortunately, only the m ere ex­

istence of these settlements and cemeteries is known, but even these slen­

der facts verify an important point, namely that in the eighth century — but only then! — the Avars withdrew to the Maros Valley and to the surround­

ing fertile lowland. In other words, at a time when the Avar settlem ent territory w as expanding throughout the Avar Em pire, it drew back in

ing fertile lowland. In other words, at a time when the Avar settlem ent territory w as expanding throughout the Avar Em pire, it drew back in