The Slavs and the Avars

Omeljan Pritsak















The name «Slav» appears in the Byzantine cultural sphere shortly after 550 in works written in Greek, Syriac, and Latin by both professional and amateur historians.





It refers to a new group of barbarian warriors roaming to the north of the Danube limes, approximately from Sirmium to the Danube delta, and frequently crossing into Byzantine territory.


Two professional historians dealing with contemporary matters, the secular author Procopius (d. ca. 572) who wrote in Greek, and Bishop John of Ephesus (d. 586) who wrote in Syriac [34], spoke of a new group of barbarians called Σκλαβην-οι or 'sqlwyn-w (esqlawin-ū).


Both in Procopius's «History of the Wars» and «Anecdota», the Σκλαβηνοι are first mentioned under the year 531, at the beginning of the rule of Emperor Justinian I (527-565) [35]. They appear along with two other groups of barbarians in two orders:


«History» : Οὖννοι, Ἄνται [36], Σκλαβηνοί [37],

«History», «Anecdota» : Οὖννοι, Σκλαβηνοί, Ἄνται [38].


An amateur, or at least unsystematic historian, who wrote in Latin rather than in Greek, was Jordanes from Moesia, a pro-Roman Ostrogoth. In 551 he wrote two works, probably in Ravenna: a gesta-type of barbarian «national» story under the title «De origine actibusque Getarum» (commonly referred to as «Getica»), and an



(34) John's three-volume Ecclesiastical History, is said to have covered events from 44 to 584, but only the last volume, starting in 575, has survived. See Nina Viktorovna Pigulevskaja, Sirijskie istočniki po istorii narodov SSSR, Moscow-Leningrad 1941.


(35) For the moment I am excepting an episode involving the Gothic Heruli in an area outside the Byzantine sphere, which scholars date to about 512, see vol. 3, p. 414 of the Loeb Classical Library edition of Procopius by H. B. Dewing (Cambridge, Mass. 1924 [reprint 1968]), but will return to it below in section IV.5.


(36) One passage in the History says that the Antai first crossed the Ister (Danube) and arrived in the vicinity of Naissus (Niš) during the reign of Justinian I (518-527), ed. Dewing, vol. 5, p. 38.


(37) History, ed. Dewing, vol. 4, p. 262; hereafter this edition will cited imply by volume and page-number.


(38) History, vol. 3, p. 252; Anecdota, vol. 6, pp. 216, 268.





outline of Roman history known as «Romana». He ended the latter work with a statement that his goal had been to relate Roman history proper and not to digress to discuss the activities of three latter-day groups of barbarians, whom he lists in the same order as Procopius did in his «History»: Bulgares, Antes, Sćlaveni [39]. This means that Jordanes was the first to identify the Bulgars correctly as Huns; it shows him as a reliable observer of contemporary events and relationships.


The professional historians of the time mentioned the barbarians only in connection with events along the Danube limes. Jordanes, however, is praised by modern scholarship for providing detailed topographical data about otherwise unknown regions that confirm a tripartite division of the Slavs at that time. His information has been considered invaluable, and has been viewed as a solid basis for classifying the ethnic and linguistic groups of Slavs of the mid-sixth century in the following way:


1) Venethi as Northern and Western Slavs;

2) Sclaveni as Southern Slavs;

3) Antes as Eastern Slavs.


But let us consult the original text of the «Getica» [40]:


34. iuxta quorum sinistruin ... near their left ridge [of the Alps,




(39) Romana, ed. Theodor Mommsen, MGH AA, vol. 5 : 1, Berlin 1882, p. 338.


(40) These passages are to be found on pages 136 and 150 of the edition by Elena Česlavovna Skržinskaja, Iordan. O proisxoždenii i dejanijax getov, Moscow 1960. Note that Jordanes, in dealing with non-Romans, uses terms denoting three levels of organization, though he is not always consistent (see Skržinskaja's commentary, p. 254, note 313, and p. 256, note 316): populus, gens, and natio. The highest unit I translate as «people», the intermediate is «tribe» or «kind», and the smallest group is then «band», although the term seems strange in view of other uses of natio by other writers and in other ages. Yet this translation is also justified for ἔϑνος in Procopius and some other Greek writers, as will become apparent in subsequent sections of this discussion.





latus, qui in aquilone

vergit, ab ortu

Vistulae fluminis

per immensa spatia

Venetharum natio populosa consedit,

quorum nomina licet nunc per

varias familias et loca mutentur,

principaliter tamen

Sclaveni et Antes nominantur.

i.e. Carpathians] which inclines toward the

north, and beginning at the source of

the Vistula river, the populous

band (natio) of the Venethi are

settled throughout a great expanse of land.

Though their names are now dispersed

amid various clans (familiae) and places

they are, nonetheless, chiefly

called Sclaveni and Antes.


35. Sclaveni a civitate

Novietunense et laco qui appellatur

Mursiano usque ad Danastrum et in

boream Viscla tenus commorantur:

hi paludes silvasque

pro civitatibus habent.

Antes vero, qui sunt eorum

fortissimi, qua Ponticum mare

curvatur, a Danastro extenduntur

usque ad Danaprum, quae flumina

multis mansionibus ad invicem absunt.


The Sclaveni dwell from the city (civitas)

of Noviotunum and the lake called

Mursianns to the Dniester and

northwards as far as the Vistula:

they have swamps and forests

for cities (civitas).

The Antes, who are the bravest of those

[bands dwelling] in the curve of the

Sea of Pontus (Black Sea) extend from

the Dniester to the Dniepr, rivers

which are many days' journey apart.


119. ... Nam hi, ut in initio

expositionis vel catalogo

gentium dicere coepimus, ab

una stirpe exorti,

tria nunc nomina ediderunt, id

est Venethi, Antes, Sclaveni;


Now these [bands] - as we started

to say at the beginning of our

account or catalogue of tribes –

being off-shoots of one origin,

now have three names, that

is, Venethi, Antes and Sclaveni.






qui quamvis nunc,

ita facientibus,

peccatis nostris ubique deseviunt,

tamen tunc omnes Hermanarici

imperiis servierunt.

Though they now rage

far and wide [?in war]

because of our sins,

yet at that time they were all obedient

to Hermanarich's commands.



Unfortunately, this locus classicus for Slavic history and historical philology has never been subjected to critical analysis. Nineteenth-century scholars found the tripartite division of the Slavs into Southern, Eastern, and Western groups self-evident, natural, and probably therefore ancient. Jordanes was therefore read as confirming the obvious, and the quotation passed from book to book [41]. Slavists have been so pleased that a mid-sixth century author provided apparently unambiguous insider's information about the northern barbarians of his time that they failed to scrutinize the text with the rigor that good scholarly method requires.





According to Jordanes's own words, section 119 is merely a recapitulation of the catalogue of tribes he had already presented in sections 34 and 35. Therefore there is no reason to regard it as a self-sufficient contemporary statement about three branches of the Slavic people. We need, rather, to look closely at the two earlier, introductory passages in Jordanes.



(41) For a typical example, see the authoritative 1954 textbook Przegląd i charakterystyka języków słowiańskich, by T. Lehr-Splawinski, W. Kubaszkiewicz, and F. Slawski, three of the most important Polish Slavists in our century, pp. 19-20.








It is important to understand Jordanes's aims and the structure of the introductory part of his «Getica».


Jordanes, let me repeat, was not a fully professional historian. His work is poorly organized. Not only can we see various seams where he has patched together information from different sources, but the basic exposition is full of insertions, disgressions and associations.


Jordanes, as he tells us in his introduction, used four types of sources, three of them not known from other surviving writings. His major source was the twelve-book «Historica Gothica» by the Roman senator Cassiodorus (490-585), which has survived only in Jordane's meager abridgement. Second was the lost Gothic history by the otherwise unknown Ablavius (descriptor Gothorum gentis, § 28), which Jordanes used for information about events of the second and third centuries, C.E., and third was Gothic popular tradition. His fourth source was works by Greek and Roman authors, both classical (pagan) and Christian; here we can check to see how he used them.


Jordanes's identifications of places and peoples are often innacurate. From the beginning, he equated his main heroes, the Germanic Goths, with the ancient Getae, a people said to be of Thracian origin, akin to the Daci. This association led him to posit as the second habitat of the Goths the area which was in fact the land of the Getae, namely Moesia, Thracia and Dacia. This confusion probably was fostered by the fact that the Ostrogoths, although at a much later time (433-471), had lived in Roman Pannonia as foederati of the empire.


Jordanes then mingled the Gothic popular traditions with classical sources about the Getae, and thus was able





to view the Gothic kings as companions to Getic philosophers who were known from Greek literature. With this approach, Jordanes envisioned a high cultural niveau for his ancestors. But in «restoring» the alleged Getic past of the Goths, Jordanes made several chronological blunders, especially in his catalogue of learned Getae. Thus he fused a Zeuta who was alive in 424 B.C.E. with another who died in 383 B.C.E. into a putative Zeuta he pictured as a contemporary of Dicineus (whom Strabo places in the first century B.C.E.) and presents the two as predecessors of Zalmoxis, a learned Getic slave of Pythagoras. But that famous Greek philosopher belonged to a much earlier period, about 530 B.C.E. [42].


An assiduous reader, Jordanes picked up names from different sources without realizing just what they referred to. For instance, in his list of Greek colonies on the northern shores of the Black Sea, he equates Olbia, located on the Borysthenes River (i.e., the Dnieper), with a «Borysthenide» (§ 32), which simply did not exist [43]. The Dniester becomes two separate rivers, «Tyra, Danaster» (§ 30). The Olt, an important tributary of the Danube, is once called Aluta (§ 30) but reappers as flu[men] Tausis (§ 136) [44]. Jordanes's knowledge of the geography of the regions beyond the Mediterranean was not always accurate; his list of Greek colonies on the northern shores of the Black Sea included the Anatolian trading center of Trebizond (§ 32); he was probably misled by the existence of the Crimean mountain range with a similar



(42) Herodotus, Historiae, IV.95.1-3. According to classical authors, Zalmoxis was the god of the Getae, not merely a human.


(43) See Henryk Łowmiański, «Scytia», Słownik starožytności słowiańskich, vol. 5, Warsaw 1975, p. 115. (This encyclopedia will henceforth be cited as SSS).


(44) Friedrich Westberg, Zur Wanderung der Langobarden (Zapiski Imp. Akademii Nauk, 8 ser., vol. 6 : 5; St. Peterburg 1904), p. 11; C. Diculescu, Die Gepiden, Leipzig 1922, p. 73.





name [45]. It is clear that Jordanes must have misinterpreted his written sources. For instance, the «Chorography» of Pomponius Mela (fl. 44 C.E.) contains the river name Hypanis (Boh, Southern Bug) and the tribal name Callipidi [46]; Jordanes records both as names of alleged Black Sea cities (§§ 32, 46).


Jordanes was inconsistent even in dealing with the names for the Danube. Although he was clearly aware that the names Ister and Danubius referred to one river (§§ 32, 75, 114), he still uses Ister in its original Thracian sense, i.e., to designate the Danube from the confluence of the Sava and Tisza to the Black Sea (§§ 30, 33). Not uncommonly Jordanes, compiling information from different sources, places names side by side and sometimes uses several variants of the same name interchangeably, e.g., Vistula and Viscla; Mursianus lacus and Morsianus stagnus; Alani, Halani and Spali [47]; Sadagarii and Sadagis; Araxes and Abraxes, and the like.





A full critical analysis of the «Gethica» as a source would be out of place here, but it is important to look carefully at the first thirty-eight paragraphs in order to place the locus classicus concerning the Slavs in the context of the larger system, such as it is. Here is an outline, with some comments:


1) Author's preface (§§ 1-3), which is irrelevant for the present analysis.



(45) As suggested by H. Łowmiański, SSS, 5 (1975), pp. 354-355.


(46) Mela, Chorogr. II.7 is taken from Herodotus, IV. 17.


(47) Concerning the identification Alani = Spali, see Francis Dvornik, The making of Central and Eastern Europe, London 1949, pp. 279-280, and H. Łowmiański, «Spalowie», SSS, 5 (1975), pp. 354-355.





2) Division of the earth into a tripartite continent (Asia, Europe, Africa) and an ocean with islands; here Jordanes acknowledges the authority of Paulus Orosius (ca. 380-420) as a source (§ 4).


3) Information on islands, both generic (e.g., the «Cyclades» and «Sporades») and specific (e.g., Orcades, Thyle [probably Ireland], and Skandza [Scandinavia] (§§ 5-9). Main sources are Claudius Ptolemy (d. after 151), Julius Caesar (100-44 B.C.E.), Titus Livy (59 B.C.E.-17 C.E.), Tacitus (55-120), and Dio Cassius (155-240) (§§ 10-15).


4) Topography and geography of the island of Skandza. Following Ptolemy, Jordanes locates it opposite the Vistula river. He quotes Pomponius Mela and Ptolemy (§§ 16-19).


5) Anthropological data on Skandza. The only source Jordanes quotes here is Ptolemy, Geogr. II.11.33-35. But while Ptolemy knew of only seven Scandinavian peoples, Jordanes presents fresh and correct information which must come from Ablavius or oral tradition or both: his list contains about thirty Old Scandinavian tribal groups (§§ 20-24a).


6) Digression about the fate of the king of Heruli, Roduulf (d. 494), which was apparently taken from the «History» of Procopius or a common source (§ 24 bis) [48].


7) Very important and highly original is Jordanes's historiosophical concept of officina gentium 'the factory of tribes' and vagina nationum 'the vagina of bands' (i.e. the starting point for the planned migration of no-



(48) See Procopius, History, vol. 3, pp. 404, 406.





mads of the sea) and its identification with Skandinavia (Skandza). The story of the initial migration of the first Gothic king, Berig, to the Vistula Delta (Gothiscandza), is apparently taken from oral tradition (§§ 25-26).


8) Next follows the migration of Filimer (the fifth successor of Berig) to the northern shores of the Black Sea, in Gothic called Oium (lit. «bei den Auen», i.e. «Mesopotamia»), that is, the curved shores between the Dniester and the Dnieper, and his victorious encounter with the [Alanic] Spali. Ablavius is named as an authoritative source, but epic songs were probably also used (§§ 26 bis-28). The author also polemicizes with Josephus Flavius (37-95) about the relations between the Goths, the Scythians, and the biblical figure of Magog (§ 29) [49].


9) The topographical and geographical description of Scythia that follows (§§ 30-38) requires more detailed treatment. Here one must keep in mind that Jordanes's goal was to set down the history of the Goths; Scythia and its geography were of interest to him only because the Goths once lived there, after they had left the island of Skandza. It is important as a major stop during the long journey of the Goths. What Jordanes wanted to do first was to present the frontiers of Scythia as clearly as possible. Therefore, he delineated points of orientation first as viewed from the south (as did classical maps and both Christian and Arabic early medieval maps) and then from the northern perspective. From time to time he inserted associations, digressions, and glosses, but his data are always primarily concerned with Gothic affaires.


We must also keep in mind the relative chronology of



(49) On these polemics see Pritsak, The Origin of Rus', vol. 1, p. 527.





major events in Gothic history which Jordanes sets up (§ 38):


a) Preparatory migration from Skandza to Gothiskandza;


b) The migration to the Ukrainian Mesopotamia (Oium);


c) The shift to the commercially important region around the Azov Sea (iuxta Paludum Meotidem) which he called the «first habitat» of the Goths;


d) Moesia [I and II], Thracia, and Dacia as the «second habitat» of the Goths (grouping together the classical Getae and the early medieval Ostrogoths, 433-471);


e) The «third habitat» of the Ostrogoths (the empire of Hermanarich, d. 375, and his successors until 433), placed again in the Ukrainian Mesopotamia between the Dniester and the Dnieper.





Jordanes did not follow Ptolemy, who had replaced Herodotus's concept of Scythia with the «modern» one of Sarmacia. He kept the name Scythia, but, like Strabo (63 B.C.E.-23 C.E.) [50] and Pliny the Younger (23-79) [51], extended its western frontiers as far as Germania [52].


Jordanes's treatment of the frontier of Scythia is extremely important for resolving the issue we are concerned with. Therefore our analysis must be scrupulous. Let me repeat that Jordanes gave two sets of orientations, one from the southern perspective and the other from the northern perspective.



(50) Ed. Müller-Dübner, VII.21.


(51) See Łowmiański, «Scytia», SSS, 5 (1975), p. 115.


(52) See Julia Zabłocka, «Germania», SSS 2 : 1 (1964), p. 97.





I. The western frontier of Scythia, i.e. the frontier between Scythia and Germania:


A. In the south:


a) the «birthplace» of the Ister river (ubi Ister oritur amnis, § 30). Accepting the views of Pliny [53], who regarded Illyricum as the place where the river Danubius ended and the river Ister began, Jordanes pinpointed the birthplace of the Ister at the confluence of the Danube with the Tisza (north) and the Sava (south). Three important Roman frontier cities were located there: Sirmium (Mitrovica), Singidunum (Belgrade), and Novae (Euscia) [54].


b) Stagnus Morsianus (§ 30), which Jordanes elsewhere calls Lacus Mursianus (§ 35), is surely the Neusiedler Lake in Austria [55]; Jordanes chose it because it was a clear marker of the northwestern end of the Roman province of Pannonia.


B. In the north: the delta of the Vistula River (§ 30).



(53) Natural History, III.79; cf. also III. 150. See also Skržinskaja's commentary on p. 199.


(54) On these cities see Limes u Jugoslavii I. Zbornik radova sa simpozijuma o limesu 1960 god, Belgrade 1961; Miroslava Mirković, Rimski gradovi na Dunavu u Gornoj Meziji, Belgrade 1968; Franjo Barišić, «Vizantijski Singidunum», Zbornik radova, knj. XLIV, Vizantološki institut, knj. 3, Belgrade 1955, 1-14; Božidar Ferjančić, Sirmium u doba Vizantije, Sremska Mitrovica 1969; Sirmium. Archeologic investigations in Syrmian Pannonia, 3 vols., Belgrade 1971-1973.


(55) This identification was first proposed by František Vitazoslav Sasinek (1830-1914), Czech medievalist, in Sbornik musea slovenskej společnosti, Prague 1896, 15, and later, independently, by Friedrich Westberg (1864-1920), a historian from Riga, «Anten», in Zur Wanderung der Langobarden (cf. note 44 above), pp. 12-14. On the history of this discussion see E. Č. Skržinskaja, «O sklavenax i antax, o Mursianskom ozere i gorode Novietune», Vizantijskij Vremennik, Moscow 1957, pp. 3-20, esp. 5-18.






II. The eastern frontier. Scythia extended eastward to the land of the Seres (China) and its northernmost boundary was the Ocean (Oceanus, § 31).


A. From the southern perspective, from east towest:


a) Persia, Albania, Hiberia (Iberia, Caucasian Georgia).


b) Pontus (Black Sea) and Ister (Danube, § 32).


B. From the northern perspective:


a) Caspian Sea (Mare Caspium), which Jordanes, following Eratosthenes, describes as being shaped like a mushroom (§ 30) [56].


b) The lands of the (historical) Hunni, the Albani, and the Seres (from west to east; § 30).


Jordanes's Scythia was divided by the Riphaean Mountains (a mythical range he took over from Eratosthenes, Ptolemy, and other ancient authorities) [57] into European and Asian parts; the frontier was the river Tanais (Don), emptying into the Maeotis (§ 32).


In the European part of Scythia, where the Goths lived, there were the rivers Tiras, Danaster (both names for the Dniester), Vagosola (probably the Boh) [58], and Danaper (the Dnieper) [59], as well as the Taurus Moutains



(56) See Skržinskaja's commentary on p. 201.


(57) On the Riphaean mountains see Aleksajstder Krawczuk, SSS, vol. 2 : 1 (1964), pp. 146-147.


(58) Jordanes uses Hypanis, the classical designation for the Boh, but in connection with a fictional Black Sea Greek colony: Hypannis (oppidum [§ 46]), mentioned along with another alleged Black Sea city, Callipolida [§ 32]. Both «cities» appear solely because Jordanes misunderstood a passage in the work of Pomponius Mela, (Chorogr. II.1.6 and II.7). See also Skržinskaja's commentary, notes 145-146, pp. 226-227.


(59) Jordanes also preserved the post-Attilan Hunnic name for the Dnieper: Var ( = vär) (§ 174). On this name, see my article, quoted in fn. 30 above.





(on the Crimea), which the author - this time - clearly differentiates from the Anatolian range of similar name (§ 30). Whereas the name Danaster for old Tiras was first introduced by Ammianus Marcellinus (ca. 330-393) [60], Jordanes is to be credited with the first use of Danaper/Dnieper, the «local» name for the ancient Boristhenus (quem accolae Danaprum vocant, § 44). The name Vagosola is a hapax and remains enigmatic.


Maeotis, the Azov Sea - so important in Jordanes's vision of Gothic history - is mentioned several times, usually in connection with the Thanais (Don), Bosporus (Kerch Strait), Caucasian Mountains, and the Araxes River in Transcaucasia (§§ 30, 32, 44, etc).


To stress the importance of Gothic Scythia, Jordanes inserted a list of Greek colonies of the polis type on the northern shores of the Black Sea. Some were real (Olbia, Cherson, Theodosia, Myrmician) and some íictional (Boristhenide, Callipolida, Careon); he also mistakenly included the famous Anatolian trade center of Trebizond here (§ 32).





10) Anthropological description of Scythia (§ § 33-37).


This material is presented along the same lines as the geographical description and must be analyzed in reference to it.


Jordanes again began with the western frontier from the southern perspective. The first people he named were naturally the Germanic Gepidae (§ 33) who from about 269 until after his own time (567) lived in Dacia (Panno-



(60) Rerum gestarum, ed. John C. Rolfe, vol. 3 (Loeb Classical Library, Cambridge, Mass. 1939 [repr. 1958]), p. 396).





nia). He described their habitat as being where the river Tisia (Tisza) embraces the land from the north and west, the river Tausis (Olut/Aluta) from the east, and the river Ister from the south (§ 33). Dacia is presented as having the shape of a crown, enveloped by the Alps (Carpathians: § 34).


In recent times, from Jordanes's point of view, i.e. during the first decades of the sixth century, another race had emerged - the Sclaveni. Their habitat he described as a territory between the city of Novietunum and the lake of Mursia. The latter, as I said above, has been identified as the Neusiedler Lake. The context shows that Novietunum cannot be the same as Novodunum/Noviodunum on the lower Danube, near modern Isaccea/Isakča, as Elena Č. Skržinskaja has pointed out [61]. The territory Jordanes describes is at the source of the Ister/Danube, not its mouth. But neither can Novietunum be equated with Noviodunum (modern Dernovo) on the Sava River near Ljubljana, as Skržinskaja would have it. Emona (Ljubljana) was not located in any of the three Roman Danube frontier provinces - Moesia (which he mentions first), Thracia and Dacia - connected with the history of the Goths, termed by Jordanes the «second Gothic habitat» and indeed the historical home of the Ostrogoths between 433 and 471 [62]. Since Jordanes was concerned precisely with the Gothic past, he had no reason to use Emona as a point of orientation. Moesia (= Pannonia) was where his immediate ancestors came from, and, as I have said, it was in this province that the three Roman centers on the Danube limes were situated: Sirmium, Sin-



(61) Iordan, pp. 213-218. See also «Novietunum», in SSS, 3 : 2 (1968), p. 418.


(62) See Ernst Schwarz, Germanische Stammeskunde, Heidelberg 1956 p. 90 and map on p. 84.





gidunum, and Novae. Therefore we must look for a town in Moesia that could habe been a suitable reference point. The easternmost Moesian stronghold (established, no doubt, to hold the frontier) was Novae (also called Euscia; today the village of Golubac) [63], which Jordanes mentioned elsewhere (§ 101). I stressed above that Jordanes often used two or more variants of a geographic name. Novae can be interpreted as the Latin abbreviation of the original Celtic name, and tunum is a Germanic variant of the Celtic dunom 'city'. I suggest, then, that Jordanes's Novietunum is none other than Novae. This would explain why Jordanes gave as the second point of orientation Lake Neusiedler, near the city of Carnuntum (not far from Vienna). Carnuntum was the northernmost and westernmost city of the Ostrogothic habitat in Moesia (Pannonia), just as Novae was the city lying furthest east and south in that habitat. It is hard to find a better way to define the extent of the southwestern border of that province than by giving these two points of orientation.


The original and therefore very precious information provided by Jordanes was that the territory of Moesia, once the habitat of the Getae and then the Goths, was now under the control of the Sclaveni; Sclaveni a civitate Novietunense et laco qui appellatur Mursiano. To this he made two additions, since he associated the classical Venedae (= his Venethae) with the Sclaveni and Antes of his time. One insertion relates to the habitat of the Antes: usque ad Danastrum (cf. Antes. . . a Danastro extenduntur § 35), and the second insertion connects the information on the Sclaveni with that on the Venethae: in



(63) See F. Barišić, Vizantijski Singidunum (1955), p. 11, fn. 44, and M. Mirković, Rimski gradovi (1968), pp. 103-107.





boream Viscla (cf. ab ortu Vistulae. . . Venetharum natio. . . consedit § 34).


Jordanes associated the classical Venedae with Vinid-, the Gothic designation for the Latin/Byzantine Sklavin-, on the same «linguistic» grounds that he identified the Goths with the Getae: simple similarity in sound [64]. On the other hand, he included the group he called Venethae = Vinid- in his introductory section for good reason: it was part of the Gothic tradition, where the great ruler Hermanarich (§ 119) had dealings with the Venethae. We assume, then, that Jordanes in this case is not speaking of his own time, but is merely putting together information from his heterogeneous sources. We cannot trust his remarks about the Venethae (= classical Venedae). Further, Jordanes had learned about the populous band of the Venedae from classical sources, especially Ptolemy, who connected them with the Vistula River, even giving the Gulf of Danzig the designation κόλπος Οὐενεδικός. Tacitus wrote that the Venedi's «plundering forays take them over all that wooded and mountainous country» (§ 46) [65]. It was apparently on that basis that Jordanes gave his description of the habitat of the Venethae: hi paludes silvasque pro civitatibus habent (§ 35). Jordanes here is not providing contemporary, sixth-century, data, but a rehash of material from older sources.


The word Vinid- is of Old Germanic origin; it is related to Old Norse vinr-, which had the meaning (established by Jost Trier) «der Genosse im Ring» (' comrade in the ring [of the warriors ']) [66]. But there was still another



(64) Since virtually all later Germanic sources use the terms Veneti, Venedi, Wenden, Winden, and the like to refer to neighboring Slavs, scholars assume (wrongly) that this equation was also used already by the Goths.


(65) Tacitus on Britain and Germany, tr. H. Mattingly (Penguin Books, Maryland reprint 1965), p. 139.


(66) «Zaun und Mannring», Beiträge zur Geschichte der deutschen Sprache und Literatur 66, Halle a.S. 1924, 232-264. esp. 251. See also Jan de Vries, Altnordisches etymologisches Wörterbuch, Leiden 1943, pp. 666, s.v. vinr, p. 665, s.v. vindr 2.





reason why Jordanes identified Venetae (= classical Venedae) with the Germanic term Vinid-. Like the Goths, who took over the former habitat of the Getae, the Vinid-, of his time lived on the territory of the ancient Venedae. Jordanes's data on the Vinid- (~ Vindi) are original and extremely important.


He tells us that the Vidivarii (this word has correctly been recognized as Vindivarii) live «on the shores of the ocean (Baltic Sea) where the waters of the Vistula river stream through three throats». They are not an ethnic unit, but people «congregated from different bands (or races)» (ex diversis nationibus adgregati, § 36). It is certain that in Jordanes' Vi[n]divarii we are dealing with a form of a name which Jordanes incorrectly took to be a supposed illustration of the original name of the people concerned: Vindi = Sclavi, plus the Latin word varii, meaning 'different'.


It is fortunate that Jordanes provides another form of this name: Viuidarii, gens Viuidaria (§ 96, which specialists have also corrected to Vinidarii). The second element in the name, -varii > -arii, stands for the Germanic *vari-ōs 'defender' [67]. The term *vinid-[v]ari-ōs, therefore, should be translated as «defender of the comrades in the [warriors'] ring» (or possibly: «inhabitant of the ring ») [68]. The remaining information that Jordanes furnishes corroborates this proposed etymology. He writes:


«The same Gepidae were bursting with envy as long as they lived in the region of Spesis (Spesis provincia,



(67) Adolf Bach, Die deutschen Personennamen, 2nd ed., vol. 1 (Deutscho Namenkunde I, 1) Heidelberg 1952, pp. 190-193.


(68) This designation became the name of an Ostrogothic king ca. 400; Vinitharius (*Vinidvariōs).





unidentified hapax) [69] on the island surrounded by spits of the Vistula which in their own language they called Gepedoios (Germ. ojos «island »). Now, I was told, this island is inhabited by the kind (gens) *Vinid-ari, since the [the Gepidae] left (ca. 250) for better lands.


It is well known that the *Vinidarii [are those] who had congregated together from different bands [or races] (ex diversis nationïbus) as if to one refuge and had formed one kind (in unum asylum collecti sunt et gentem fecisse noscuntur, § 96)» [70].





This definition of the Vinid-ari- is of profound importance for European history. Jordanes was in large part a compiler, and he patched together facts or statements from many different sources with no regard at all for the times and places these sources were written. Yet when he reports about groups and events for which he has contemporary information, he seems to have been accurate and therefore we can trust his statements.


During his lifetime, an institution that would subsequently have worldwide importance was in the process of formation. In Germanic it was called Vinid-, in the Byzantine cultural sphere Σκλαβην-/Sclavīn-, and (later) in the Islamic world aṣ-Ṣaqlab. From Jordanes we can deduce that these terms were originally neither ethnic nor linguistic in meaning, but organizational. What was invol-



(69) See Skržinskaja's commentary, note 311, p. 254. Concerning the migration of the Gepidae, see E. Schwarz, Germanische Stammeskunde (1956), map on p. 101.


(70) Gerard Labuda devoted a special study to the problem of Vidivarii and Vindivarii: «Vidivarii Jordanesa», Fragmenty dziejów słowiańszczyzny zachodniej, vol. 1, Poznan 1960, pp. 96-109. In general I agree with his results.





ved was a new kind of military organization, but one similar to groups which repeatedly were formed in the course of the rise of steppe or sea paces. Paesants from different hamlets, often diverse in ethnic origin but always similar in parochial outlook, after having spent some time together in a refuge - like the Vinidarii - and having undergone military training, developed into a band of professional warriors using the same lingua franca.


Let us recall two other equivalent terms used by Jordanes: the phrases vagina nationum and officina gentium, i.e., 'the vagina of bands', and the 'factory of tribes' (§ 25). These are revealing metaphors, surely, meaning the whole system of attracting raw recruits, gathering them into a training ground whose location was, for the observers in the sedentary empires, mysterious or even fantastic, and then se[n]ding out terrifying masses of trained soldiers in sudden and devastating attacks against distant territories. Even from Jordanes's data, it is evident that Scandinavia was not the original home (Urheimat) of the Goths. Berig had assembled his people there to train them for his planned military campaigns on the European continent. Both Scandinavia and the Vistula delta - like Mongolia earlier and later, and like the Zaporogian Sič in the sixteenth-seventeenth centuries - were training centers for the transformation of peasants (or pasturalists, or even fishermen) from the surrounding territories into a class of professional warriors.


This is, then, another instance of a phenomenon which kept repeating in Eurasian history. The non-historical pastoralists or peasants beyond the limes of the existing historical empires (Rome, Iran, China), who had no experience with the larger world, and whose parochial interests therefore did not in any way predispose them to larger political bodies, were more often than not forced into





undergoing a period of training that absorbed them into one larger body. This process, which usually lasted over successive generations for at least one century, created an upper class among the trainees that was cognizant of larger political bodies. That class became ready to take part in strengthening a pax and in forging the parochial dialects into a standard medium of communication for the entire pax. Linguae francae developed that embraced diverse linguistic entities into a «common language», whether based on Turkic or Slavic (or other) materials. Upon the demise of the pax, it was possible for several full-fledged «daughter languages» to emerge. This involves a concept of language development often ignored by those who take too literally the model of the genealogical tree of language as it was elaborated during the age of Romanticism. Rather than seeing only branches that continually sprout new branches, we are saying that a lingua franca which has evolved in order to serve large areas itself becomes a new and fairly uniform «tree» that then slowly puts forth new branches.


I spent four decades studying all twenty-two living Turkic languages, along with all the extinct forms that are known, with the aim of uncovering a Proto-Turkic stage (or perhaps more than one). I could not escape the conclusion that the oldest reconstructable common Turkic is the stage which directly preceded the oldest Turkic written texts, about 550-650, that is to say the time when the Turkic pax with its lingua franca, essentially free of dialectal diversities, was created.


My friend and colleague, Horace G. Lunt, has recently told me that he has had essentially the same experience with Slavic material. The oldest reconstructable Slavic differs so little from attested Old Church Slavonic, whose normalized form can be put in the ninth century, that





OCS itself must be considered a dialect form of Common Slavic, and a dialect-free stage could be envisaged for as late as 750-800.


Historians have generally used linguistic abstractions, such as the notion of Common Slavic, for their own purposes, without trying to discover what objective reality was behind them. We need to try rather to study concrete peoples in concrete situations, insofar as this is possible. It is my conviction that only a method of historical sociolinguistics, such as we are suggesting here, can produce valid answers to our valid questions.





Let us turn to another equation made by Jordanes in his introduction (§ 35). He presents the Antes as the same, apparently meaning the same society, as the Sclaveni, although they are stronger than the Sclaveni. There are two points to be considered before we turn to the etymology of this name. The first is the minor but perhaps significant fact that Jordanes has the plural form Antes [71] (Antium), which is unexpected as an equivalent of the Greek Ἄνται (sg. Ἄντης) of all other sources [72]. The second is that this group appears and disappears during the brief span between 535 and 602 in Procopius, Menander, Agathias, Pseudo-Mauricius, Theophylact Si-



(71) All three examples are in the passages quoted above, and the corresponding genitive Antium occurs in the «Romana». However, in «Getica» § 247 Jordanes uses Antorum, which implies a nominative *Anti.


(72) The Byzantine Greeks put this foreign name into the normal classwith singular in -ης and plural in -αι, like στρατιώτης : στρατιῶται. The singular Ἄντης is attested once in Procopius (vol. 4, p. 268) and once in Agathias (ed. L. Dindorf, HGM, p. 275), while the plural Ἄνται is universal in Greek. Jordanes, who was heavily dependent on written sources, as we have seen, apparently found the singular Ἄντης in a Greek text (Agathias?) and took it over.





mocattes and Theophanes [73]; it is only Jordanes that mentions Antes outside this period. He brings them up in connection with the Ostrogoths at the end of the fourth century. In a passage explicitly derived from old sources (and perhaps we may assume here the Gothic songs Jordanes mentions in his introduction), Jordanes relates how the resistance of the Antes was crushed by the Gothic king Vinitharius (fl. about 400). He captured Boz, «rex» of the Antes, and crucified him along with his sons and seventy nobles (primates, § 247). At that time, namely during the «third habitat», the Goths lived super limbum Ponti, « above an arm of the Pontic Sea », that is, in the curve north of the Black Sea, between the Dniester and Dnieper rivers (§ 82), called the Lukomorie in sources from Kievan Rus' [74]. Hence Jordanes never actually defines the current habitat of the Antes in his own time, but merely recounts at second hand tales about the glorious past of the Goths. A reasonable hypothesis is that the Antes/Antai, like the Vinid- (who, as we already know, merged in Jordanes's mind with the Venedae of classical times)



(73) Here is a complete list of all occurence of the name Antai: Procopius,vol. 3, p. 252; vol. 4, pp. 262, 264, 268, 270, 272, 342, 344, 350; vol. 5, pp. 38,84; vol. 6, pp. 132, 216, 268; Menander, ed. Dindorf (cf. fn. 30 above), pp. 5-6; Agathias, ed. Dindorf in HGM, vol. 2, p. 275; Pseudo-Mauricius, ed. H. Mihăescu, Mauricius. Arta militară, Bucharest 1970, pp. 40, 262, 276, 286,342; Theophylact Simocattes, Historiae, ed. Carl de Boor/Peter Wirth, Stuttgart 1972, p. 293; Theophanes, ed. Čičurov (cf. fn. 26 above), p. 34.


Though the first two books of the history by the Syrian John of Ephesus (d. 586) were lost (cf. fn. 34 above), scholars believe that certain passages have survived in the later Syrian compilations by Michael the Syrian (d. 1199) and Barhebraeus (d. 1286). The passages containing the name Anṭiyū (for Greek Ἄνται) have been treated recently by the Semitist Ruth Stiehl (in Franz Altheim, Geschichte der Hunnen, vols. 1-2 [Berlin 1959-60]). The name occurs in Michael, vol. 1, p. 88, and in Barhebraeus, vol. 2, p. 29 (cf. Michael's Chronicon, ed. P. Bedjan [Paris 1980], p. 90).


(74) In the «Hypatian Chronicle» s.a. 1193 the Right Bank Polovcians are called Lukomor'skies Polovci s Loukomor'skimi»), Polnoe Sobranie russkix letopisej, vol. 2, 2nd ed. by Aleksej Aleksandrovič Šaxmatov (St. Peterburg 1908), col. 675.





were present in the Gothic oral traditions. The Vinid- were linked with the great Ostrogothic king Hermanarich (d. 375) who, having overpowered the Heruli, forced them and the Baltic Aesti to submit to his rule (§ 119) [75].


As for the other historians, speaking of near-contemporary events of the sixth century, their statements are remarkably vague. In Pseudo-Mauricius Σκλάβοι and Ἄνται are always linked, in that order, while Procopius has Σκλαβηνοί and Ἄνται four times, but Ἄνται and Σκλαβηνοί three times, though does mention the Ἄνται alone in two extensive passages. In fact, it appears that this name had no precise meaning to the sixth-century historians.


This analysis shows clearly that §§ 35-36 do not contain precious information about the topography of the putative three branches of the Slavs, contrary to the belief of many scholars. Rather, apart from the current location of the Sclaveni in Jordanes's former homeland, Pannonian Moesia (a civitate Novietunense et laco qui appellatur Mursiano, § 35) and the information on the non-Slavic Vistula Vinidarii, all the data are only various insertions the compiler took from different sources, whether classical writings or oral traditions, of the Goths themselves. Jordanes put the Vinid-, Sclaveni and Antes together not on the basis of ethnic or linguistic criteria, but because all three terms refer to institutions of military colonists on frontier territories. Although this findings may dismay Slavists, it will help historians understand the process of nation-building in medieval Europe and Asia [76].



(75) In § 36 in the introduction, the Vinid- (Vidivarii) are mentioned just before the Aesti; surely this is part of the traditional association with Hermanarich.


(76) Parenthetically, let us complete Jordanes's catalogue of the people of Scythia. He named as living in the northern portion, but south of the Aesti, a pastoralist people called Acatziri, whom he knew about from Priscus (d. ca. 472 [«Fragmenta», ed. Dindorf, HGM, vol. 1, Leipzig 1870, pp. 298-99, 306, 310, 341, 346]). The Hunnic Bulgars he placed supra mare Ponticum, (which seems to be the same area he assigned to the Antes in § 35) in the curve of the Black Sea; their recent raids (550-551) on Byzantine lands (across the Danube limes), also described by Procopius (vol. 5, pp. 234, 236, 238, 240, 242) are termed «punishment for sins» by Jordanes (§ 37). The catalogue of nomadic peoples is enlarged by three «Hunnic» peoples, again taken from Agathias (Historiae, HGM, vol. 2, p. 365) and Priscus («Fragmenta», HGM, vol. 1, p. 341) and, possibly, from Gothic tradition: the Altziagiri, (var. Vltinzures § 272), Saviri and the commercially active Hunuguri (§ 37).


[Previous] [Next]

[Back to Index]