Robert Elsie | AL Art | AL History | AL Language | AL Literature | AL Photography | Contact |


Robert Elsie

Texts and Documents of Albanian History

BACK  |  AL History

Georg Stadtmüller (1901-1985)

Georg Stadtmüller (1901-1985)


Georg Stadtmüller:
Research in Early Albanian History

German historian Georg Stadtmüller (1901-1985) studied classical philology and European history at the universities of Freiburg am Breisgau and Munich, where he finished his doctorate and worked for two years for the Bavarian State Library. Around 1935, he toured Albania on foot on his way to Constantinople, and, in the following year, finished his seminal habilitation thesis "Forschungen zur albanischen Frühgeschichte" (Research in Early Albanian History), Breslau/Wrocław 1936, for which he is best remembered. This thesis, published in Budapest in 1942 and republished in Wiesbaden in 1966, traces the origins of the Albanian people back to the Mat region and provides much thought-provoking material for the thorny issue of Albanian ethnogenesis. The following excerpt from the book provides the core of his thesis.


The Early Albanian Period (the Roman and Early Byzantine Age)

The primary event in the early history of the Albanian people was the major transformation they underwent when the pre-Albanian tribes were partially Romanised under the enormous influence of imperial Roman culture and of the language of imperial Rome. It was at this time that the diverse Albanian tribes first became a people and managed to preserve their language in the tidal wave of Romanisation that engulfed all the other ancient Balkan languages. What we know of the history of the Albanian language enables us to reconstruct the steps of the partial Romanisation of the early Albanians. This other main issue of early Albanian history, however, remains unsolved – that of where the early Albanians lived in Roman and early Byzantine times.

View of Burgajet in the Mat district (Photo: Robert Elsie, December 2008).

View of Burgajet in the Mat district (Photo: Robert Elsie, December 2008).

View of Burgajet in the Mat district
(Photo: Robert Elsie, December 2008).

For the period after the Slavic invasion (around 600), we can demonstrate that the focal point of Albanian settlement was the mountainous Mat district. However, the same cannot be said with certainty for the Roman and early Byzantine periods in view of the tremendous changes that took place in the migration period in the Balkans from the fourth to the seventh centuries. We cannot ignore the hypothetical possibility that, as a result of the mass migration of peoples, the Albanians immigrated to their subsequent homeland, either from another region of the Balkan Peninsula or from outside the Balkans. Their homeland in the Roman and early Byzantine periods must thus be investigated without reference to where they lived in later times. Linguists, ethnologists, archaeologists and experts in both political and ecclesiastical history must work together to answer this question.


Latin Loanwords in Albanian

The Albanian language possesses a substantial corpus of Latin loanwords which, however, because of the thorough phonological transformation they underwent as a result of the strong expiratory accent of Albanian, no longer look Latin at all. Latin influence on Albanian was so profound and so strong that one could pretty well call it a partially Romanised language. Indeed, in some fields of activity, most concepts were taken over from Latin.

For instance, almost all family terms (e.g. uncle, brother-in-law, parent, child, cousin, stepson) come from Latin, as do basic concepts of urban life (e.g. city, house, roof, street, mill, herd), agriculture (e.g. farmer, quince, olive, fruit) and governance (e.g. emperor, court, king). Also from Latin are terms for village life and winter grazing.

Terms for animal husbandry and milk and dairy activities, on the other hand, are purely Albanian. This demonstrates that in the Roman and early Byzantine periods the early Albanians were transhumant herders, and that only after contact with Roman culture, did they settle down in towns and villages, and learn how to farm and to write, and how to organise themselves with a government. Many names of plants were also taken from Latin, which is an indication of the influence of Roman agriculture on the early Albanians. The word for ‘beehive’ is Latin, but the terms for ‘bee’, ‘honey’, and ‘wax’ are purely Albanian. As such, one can surmise that the early Albanians knew about wild bees before they had contact with the Romans, but they learned of bee-keeping from the Romans. All these loanwords show the strong cultural influence of Roman culture on the early Albanians, but they do not allow us to reach any conclusions as to their homeland at the time. But we do have some indication from Albanian terms for forestry. Words for the mountain forests and mountain trees (e.g. mountain pasture, slope, mountain forest, beech, linden, fir, oak, wood, spruce, hazel, elm, hornbeam) are purely Albanian, whereas the words for woods in the lowlands and the vegetation in them (e.g. grove, poplar, Mediterranean hackberry, Adriatic oak, ash, willow, wild olive, privet) are taken from Latin. Also from Latin are words for ‘plain, flatland’. Thus, the Albanians had contact with the plains and the woods of the lowlands when they were on Roman linguistic territory. At the same time, the summer pasturelands of the early Albanians must have been in the interior, far from the coast, which also explains the lack of genuine Albanian terms for navigation and fishing. Thus, based on the Latin loanwords in Albanian, we can draw the following conclusions:

The early Albanians in the Roman and early Byzantine periods lived as transhumant shepherds in the mountains of the interior. Their winter pasturelands were situated on the Romanised coastal plains. It was here, in sometimes friendly and sometimes hostile contact with the neighbouring Roman settlers, that they learned of social organisation (government) and sedentary ways of life, and that they acquired goods and, in particular, learned the art of bee-keeping and more advanced agriculture.

The massive influence of Latin on Albanian can be explained by the fact that the early Albanians and the late Roman settlers lived together for centuries within the borders of the Roman Empire. Had the early Albanians inhabited a neighbouring region, i.e. outside the borders of the empire, as did the Germanic tribes for instance, Romanisation would never have penetrated as deeply as it did here. In view of this, one can easily reject the earlier hypotheses that the early Albanians immigrated to the Balkan Peninsula during the migration period. We can safely assume that the early Albanians were living on the territory of the Roman Empire at the start of Roman rule.


Ancient Greek loanwords in Albanian

Of particular value in proving where the original homeland of the Albanians lay are ancient Greek loanwords in Albanian, that occur primarily in the names of fruit trees and cooking. These loanwords lead us to the following indubitable conclusion: the homeland of the ancient Albanians was near the Latin-Greek language border, on the Latin side of it.

The border between the Latin and Greek-speaking regions in the Balkan Peninsula can be determined by inscriptions, milestones and coins. As Jireček noted, “It corresponds more or less to the provincial borders of the first to third centuries, before the reforms of Diocletian. Starting from the Adriatic Sea, the language border ran past Lissus / Alessio [Lezha] and continued along the frontier between Dalmatia (later Praevalis) to the north and Macedonia (later Epirus nova) to the south. It ran south of the present road that links Shkodra to Prizren because, according to Ptolemy, Ἐπικαρία (called ad Picaria in the Tabula Peutingeriana), now Puka, was still in Dalmatia at the time. It also corresponded to the border between Moesia superior (later Dardania) and Macedonia. Here, the inhabitants of the Dardanian settlements of Ulpiana (Lipjan) in Kosovo and Scupi (near Skopje) in the upper reaches of the Axios (Vardar) river valley left behind Latin inscriptions. The nearby settlement of Stobi, on the other hand, left us Greek inscriptions. The language border then followed the old frontier between Moesia superior (later Dardania) and Thrace, such that Naissus (Niš) and Remesiana (Bela Palanka between Niš and Pirot) are on Latin territory, but Pautalia (Küstendil), Serdica (Sofia) and the region of Pirot are on Greek territory. The late Roman province of Dacia mediterranea of the fourth to seventh centuries, created from bits of Thrace and Moesia superior, with the settlements of Serdica, Pautalia, Naissus and Remesiana, was thus bilingual. The language border continued from the area around Bela Palanka and Pirot eastwards along the northern slopes of the Haemus, following the provincial border between Moesia inferior and Thrace, such that inscriptions from the area around Vraca and from Nicopolis (Nikjup near Trnovo) are mostly Greek, but those from the banks of the Danube down to its estuary are almost all Latin. At its eastern end, the language border met the territories of the Hellenic communities along the Black Sea coast up to the estuary of the Danube.”

Thus, the homeland of the ancient Albanians lay somewhere near the Latin-Greek language border.


The Conversion of the Early Albanians to Christianity

The history of the conversion of the Albanian people to Christianity provides clues that allow us to locate their homeland more specifically. The Albanians made their first appearance in European history when they were already Christians, in the high Middle Ages (11th to 13th centuries). We have no documentary information about their conversion, but the field of linguistics can help us out here. Church expressions that Albanian borrowed from Latin show a phonetic form that evinces a connection with the old Romance language of Dalmatia. A smaller portion of church vocabulary is of Greek origin.

The conversion of the early Albanians to Christianity thus took place primarily from Romanized Dalmatia and only to a much lesser extent was there contact with the Greek church.

Thus, the homeland of the early Albanians is located in a region neighbouring Dalmatia, but which was not too far from the Greek language border. This restricts the location to a smaller region, the basin of Old Serbia [Kosovo] and the mountain regions of northern Albania.


The Early Albanian Region as a Linguistic Enclave

To get closer to a solution of the homeland issue, we must clarify what we know about the character of early Albanian territory in the Roman and early Byzantine periods. Early Albanian territory was a moribund linguistic enclave, the reduced remains of a language once spoken throughout much of the Balkan Peninsula. The early Albanians are the only ancient Balkan people to have survived the general Romanisation that swept through the peninsula. Throughout the rest of the region, the superiority of imperial Roman culture and the imperial Latin language razed and levelled the peoples and languages of the Balkan provinces in a silent process that went on for a century. Latin, the language of the administration and of the army, took prominence everywhere, and the native idioms died out slowly. The process was more or less complete by the end of the third century. The native idioms only survived for a time in isolated regions, away from the major routes of communication.

Latin penetrated all areas of Bosnia and Bulgaria and replaced all the native languages spoken there, but in northern Albania and Old Serbia the early transhumant Albanian shepherds managed to retain their tongue. Of course imperial Roman culture did leave its traces here, too. The early Albanians first had contact with the more advanced culture of the Roman population of the plains when they descended from the mountains to their winter pasturelands. It was here that they absorbed much Roman culture. The modern-day Albanian language reflects their early dependence on Roman culture in the many Latin loanwords they adopted. Modern Albanian is a partially Romanised language, one might even say, a semi-Romanised language. Had Roman rule in Albania continued for another few centuries, it would have blotted out the early Albanian tongue, as it had the native languages of the rest of the Roman Empire. That is to say, the Latin language of imperial Rome would have melted it down, too, as it did the other languages. However, this ongoing process of Romanisation was interrupted by the Slavic invasion of the Balkans around 600 AD. This is what preserved the early Albanians and their language.

The fact that early Albanian managed to survive the onslaught of Latin, when all the other native languages had already died out, can only be explained by the fact that the early Albanians had less contact with Roman culture. As transhumant shepherds, they lived on the winter pasturelands of the Romanised coastal plain in the cold season, and here, were subject to the influence of Romanisation. But in the summer, they followed their flocks up into the mountain pastures. The homeland of the early Albanians was thus a two-fold region: winter pastureland on the Romanised coast and summer pastureland up in the mountains. Gradual Romanisation certainly took place in the winter months, but in the summer, the Albanian shepherds were more or less uninfluenced by any Roman influence for half a year. Romanisation spread with time. As the early Albanians were the only ancient Balkan transhumant shepherding people to preserve their native language for such a long period, it is evident that their summer pastureland had to be located in a mountainous region away from the influence of Roman culture. The early Albanian language can only have survived in an isolated mountain region which was situated at a distance from Roman towns and from major routes of Roman communication.

With this in mind, if we take a look at the potential locations of summer pastureland in Old Serbia and northern Albania that the early Albanian shepherds might have used, it becomes evident from the start that we must exclude Old Serbia. Because of their low elevation, the great inland basins of Old Serbia (Kosovo, Metohia and the Sandjak of Novi Pazar) would have served them as winter pastureland and not as summer pastureland. In the mountains surrounding these basins, there is extensive pastureland, but these regions were not isolated enough for a linguistic enclave to survive. These regions, easily accessible from all sides, played a major role as a nucleus of the early Serbian state (Rascia), but they were never prime locations for the native population to resist foreign rule, be it Roman, Byzantine or Ottoman. As such, the early Albanian enclave cannot have been located in Old Serbia.

Consequently, what remains are the mountains of Northern Albania. Three different geographical regions are involved here: the Northern Albanian Alps, the highlands of the Drin River, and the district of Mat.

The Northern Albanian Alps (Malësia / Prokletija) are a range of solid mountain peaks which constitute the last remnants of an earlier limestone plateau. Torrents flow in all directions from the massif of the Maja e Jezercës down the mountains and open into large valleys which offer access to the region. Roman culture was able to penetrate this mountain region from all sides. From the plain of Shkodra, the population centre of the Roman province of Praevalis, the valleys of the Proni i Thatë and the Kir Rivers lead directly up into these Northern Albanian Alps. On the southern side, the mountains are accessible from the wide Valbona Valley. To the north, the valley of the Lim River provides access from what is now the Sandjak of Novi Pazar. Accordingly, this region cannot be regarded as an enclave because we must assume that it was open on all sides to the advance of Roman culture and the Latin language.

The highlands of the Drin River encompass a wide range of territories, the best known of which are Mirdita, Dukagjin and Cukali. There are no natural lines of communication here. All transportation is conducted on donkey paths. Nonetheless, the highlands of the Drin River have always been full of movement as they are situated between the plain of Shkodra to the west and the plains of Old Serbia to the east. This region thus is even less suitable as an enclave than the Northern Albanian Alps.

What is left, therefore, is the district of Mat. It possesses the natural qualities for an enclave and we can assume that the early Albanians had their summer pastureland here. The broad basin of the Mat River makes a relatively large region for settlement. Some 24,000 people live there at present. The surrounding mountains offer rich pastureland in the summertime, and the valleys are cultivated. The district of Mat is surrounded by a chain of high mountain ranges that make it a natural fortress. Contact with the outside world is only possible on difficult donkey paths. The only other routes out of the valley [aside from the valley of the Mat] are at Ungrej in the northwest where there is a path leading down into the plain of Shkodra, and over the Qafa e Bulqisës to the east which leads to the valley of the Black Drin. The natural isolation of the valley has made the district of Mat an mountain canton of its own that, in Albanian history, always played the role of a centre of national resistance to foreign intrusion. The district of Mat has all the natural qualities that the early Albanians would have needed to settle there.

Based on the general hypothesis we have set forth about the process of Romanisation, we can reach the following conclusion: in the Roman and early Byzantine periods, the summer pasture region utilised by the early Albanians was the district of Mat.

As any hypothesis based solely on history and comparative development, this one can only be a probability, even though it is in my opinion a very high probability. To convert this probability into certainty, we must prove that the district of Mat was the only region between northern Albania and Old Serbia that was not affected by Roman settlement and Romanisation.

We must therefore endeavour to determine the extent of Roman settlements in the whole of the region in question, on the basis of place names and archaeological remains.


The Latin Place Names

Latin place names in northern Albania are rare. The Slavic conquest of the end of the sixth century put an end to Roman culture in the interior. However, the names of rivers and larger towns survived the catastrophe, but these names are almost all pre-Roman in origin (e.g. Drilon-Drin, Scodra, Lissus, Dyrrachion, etc.). Very few place names in the Albanian coastal region, the mountains of the Drin River and the Northern Albanian Alps can be shown with certainty to be of Latin origin because it is unclear whether the terms for villages and locations named after saints extend back before the Slavic period and because the Latin names for the many Justinian fortresses have not been preserved.

Among the Latin place names on the Albanian coast are: Domni (Lat. domini), a village about 16 km northeast of Shkodra; Pëdhana (Lat. pedaneus), a village on the coastal plain where the Mat River leaves the mountains; and Vjerdha (Lat. virida), a village on the coastal plain where the Drin River leaves the mountains.

In the mountains of the Drin, a number of Latin place names have been preserved: Gushti (Lat. angustus) at the outer end of the Drin gorge, Bisaku (Lat. Bit(h)us-iacium) in the Fan Valley near Orosh, Qelza (Lat. cella) and Puka (Lat. publica) in the Dukagjin region, the water source Kron-i-Valung(ë)s (Lat. vallis longa) in Berisha territory, and several place names in Mirdita such as Groftat-e-Gojanit (Lat. crypta), the village of Skortul, the mountain Maja Volpul (Lat. vulpes), several peaks known as Kunora, the mountain M(u)nele-a (Lat. montellus), the creek Proni Sift (Lat. exsuctus), and the ruins of Kastri (Lat. castrum). In the region of Lurja to the east, we find the cave Shutrrija-e-Selitës (Lat. subterraneum).

In the Northern Albanian Alps, we find: Valbona (Lat. vallis bona), a tributary of the Drin; Kastrati (Lat. castrum), a tribe living on the edge of the Northern Albanian Alps north of Shkodra, possibly Plava (Lat. Flavia) in the valley of the upper Lim; the nearby mountain Maja Romanit, as well as the village of Pjani (Lat. planum) in the Malësia e Gjakovës and the mountain Kumul (Lat. cumulum) in the same area.

Of the pre-Slavic place names in the region of the Black Drin, on the eastern edge of modern Albania, only one of them could possibly be of Latin origin.

We do not know the exact location of the ancient fortress of Clementiana, but the name has survived in the tribal designation Këlmendi (from Clement).

In the neighbouring basin of Old Serbia, that was inhabited in Roman times, only two Latin place names have been preserved: Tauresium, modern Taor, and Ulpiana that was transformed into Lipljan as a Slavic folk etymology(from Slavic lipa “linden tree”).

The few place names that are clearly of Latin origin are thus distributed over coastal Albania, the mountains of the Drin River, the Northern Albanian Alps, the region of the Black Drin (?) and Old Serbia.

In addition to place names of Latin origin, many pre-Latin place names were phonologically altered by the process of Romanisation. The names of the larger settlements (Scodra, Lissus, Dyrrachion) are of pre-Roman origin. They survived both the linguistic transformation they underwent during Romanisation and the cultural break caused by the Slavic invasion. In many cases, the phonological forms of these words can only be explained by the assumption that a Roman population inherited them from the pre-Roman population and that, after transformation, they were inherited by the Albanians and immigrant Slavs. Such names provide indirect evidence of Romanisation. Related to these are place names that were not continuously in Albanian use. Such place names in the Balkan Latin area that did not undergo normal Albanian phonological development can be regarded as proof of Romanisation because the whole population of the Balkan Peninsula, with the exception of the early Albanian transhumant shepherds, had been Romanised (or Hellenised) by the eve of the Slavic invasion. They are situated in the basin of Shkodra, in the Valbona Valley, in the district of Ljuma in the valley of the Black Drin, on the coastal plains and in the valley of the Shkumbin River.

These place names which, by their etymological origin or phonological form, demonstrate Romanisation in the regions in question are insufficient in number to provide absolute proof that the regions were completely Romanised, but they can be used as an indication of this.

There are only two regions where there are no traces of Romanisation in the place names: the high mountain areas of the Northern Albanian Alps and the district of Mat. We would not expect Romanisation in the high mountain areas of the Northern Albanian Alps because it is obviously not a region that is conducive to settlement. Villages were and are rare there. At any rate, any genuine Latin place names would have been wiped out by the flood of Slavic invasion. The situation is quite different in the fertile district of Mat where, even today, there are sizeable towns and villages. It is here that we would normally expect to find Latin place names. The fact that there are none here leads us to the conclusion that this was indeed the linguistic enclave in which the early Albanian transhumant shepherds found their abode.



[excerpt from: Georg Stadtmüller, Forschungen zur albanischen Frühgeschichte, 1936, second edition printed in Wiesbaden: Harrassowitz, 1966, p. 76-96. Translated from the German by Robert Elsie. Readers are advised to consult the original text for the copious footnotes.]


Georg Stadtmüller (1901-1985)