Constantine VII Porphyrogennetos on Croatsin early medieval Southern Pannonia (DAI, C. 30, 75-78): a note on concept and method of Byzantine history writing

A ®

DOI:

Цитирование. Грачанин Х. Константин VII Багрянородный о хорватах в раннесредневековой Южной Паннонии (DAI, с. 30, 75–78): замечания о концепции и методе византийской историографии // Вестник Волгоградского государственного университета. Серия 4, История. Регионоведение. Международные отношения. – 2020. – Т. 25, № 6. – С. 24–43. – (На англ. яз.). – DOI:

Introduction. The Croatian early Middle Ages is in many aspects just a collection of research puzzles, a mosaic of various details that open more questions than they give answers. As it stands today, one can easily say that the entire early medieval Croatian history is a construct which makes it only possible to follow the historical processes in broad outlines since the literary and material sources available to scholars mostly offer the material for merely making general, more or less founded, assumptions (cf. remarks in [29, p. 248]). An example of such a scholarly puzzle, which comparably engages both Croatian historians and archaeologists, relates to a question of the early medieval presence of the Croat ethnie in southern Pannonia, i.e. modern Northern Croatia or, as it also called, the region between the rivers Sava, Drava and Danube.

Research problems and approaches. In a nutshell, the entire problem is centred, which is not a rarity when it comes to the history of the Croatian early Middle Ages, on an isolated record. In this case, it is the two lines from the famous treatise composed by the Byzantine emperor Constantine VII Porphyrogennetos (905–959) for his son and heir Romanus II, and known under the Latin historiographic title De administrando imperio , commonly abbreviated as the DAI , and, in Greek, ProFRwmano>n. The work is, in essence, a sort of a Byzantine Fürstenspiegel, a manual with guidelines and useful facts primarily for diplomatic purposes, which emphasized the importance of proper information or, at least, helpful tips as such knowledge would give obvious advantages in dealings with foreign states and peoples ². Not that the royal compiler was always able to adequately fulfill the task taken upon himself, since the collecting and arranging of the material proved to be, in some instances, too challenging and some information is clearly garbled ³.

Nevertheless, even though the final product – in its present form – clearly lacks a precise editing and finishing touches, it was a project that fitted well into Constantine Porphyrogennetos’s overall literary program (whether or not he was its actual instigator), and it was sufficiently sophisticated in the sense that it had an elaborated agenda and general structure, and aimed at offering new perspectives on the up to four century long past as well as recent past of the regions and polities (with conspicuous omissions, at least to a modern researcher’s eyes) relevant or of value to current Byzantine interests. That is to say, whatever one may think of how successful was Constantine Porphyrogennetos in his undertaking, the fact remains that the DAI was an attempt at understanding and explaining the geopolitical situation of the day in its historical perspective, and finding practical lessons with bearing on the challenges that the existing circumstances posed. Therefore, it may be contended that the DAI was perhaps the most complex of the Constantinian collections considering its goals and implications ⁴. In other words, notwithstanding its apparent haphazardness, the DAI seems to be a deliberate document containing purposely collected pieces of information with concrete meanings and intents. Hence – I shall venture to claim – if one manages to decipher the background, the nature and the context of these pieces of information, they might perhaps be able to unlock their actual meaning and specific ramifications.

In line with that, the following examination will depart from the approach that, at first look, might seem innovative, mirroring recent theoretical advances and devoid of any ulterior motives but, by its dismissal of historical phenomena, processes or events in the manner they are portrayed in the DAI as Constantine Porphyrogennetos’s fabrications and inventions resulting from narrative strategies, instead of considering and addressing them in their own right with adducing all the possible evidence, does not get us actually closer to understanding Constantine’s concepts and methods nor it contributes to the interpretative clarification of not easily explainable claims or pieces of information or textual places 5. The main research problem tackled here is whether the Croats or, better yet, the fundamental (eponym) ethnic group or groups, which participated decisively in the construction of Croat ethnic identity 6, already settled in south Pannonian area in the seventh century A.D., or this, in all its value, cannot be the case, that is to say, the presence of the Croat ethnie in the region between the rivers Sava, Drava and Danube in the early Middle Ages should rather be seen as a result of insufficiently founded conjectures of modern scholars based on not enough informed claim by the source. The paper will argue that the view supporting such an early presence of the Croats in southern Pannonia is misleading, and if there were any Croat name-bearing groups in that region at all – which cannot be entirely excluded nor corroborated – their number must have been negligible when compared to other Slav(icized) populations, which is also very much consistent with the circumstance that the area of the primary construction of Croat identity was much farther to the south, close to the Adriatic 7.

Presenting the source’s account. The two sentences narrating about the Croats migrating from Dalmatia to Illyricum and Pannonia are found in the Chapter 30 of the DAI , entitled The narrative about the theme Dalmatia , and in English translation they are as follows: And of the Croats who arrived to Dalmatia one part separated and ruled Illyricum and Pannonia. And they also had an independent ruler who was sending envoys, though only to the ruler of Croatia from friendship (’ Apo de t ® n Crwbatwn, t ® n elqontwn en Delmatia, diecwri>sqh me>rov ti, kai< ejkra>thsen to< E Illurikoan? ei+con de< kai< auj toi< a]rconta auj texou> sion, diapempo>menon kai< mo>non proav kata< fili>an ; DAI , c. 30, 75–78) ⁸.

This passage is preceded in the DAI by the story about the Croat’s arrival to Dalmatia from their ancestral land, their fight against and victory over the Avars, and their occupation of Dalmatia, as well as by the remark about the Croat’s ancestral land near Saxony (and beyond Bavaria), where they are subordinate (uJpo>keintai) to Franks and in friendly relationship with the Turks (i.e. Magyars), and is followed by the story about how, in Dalmatia, the Croats were also subject (uJpeta>ssonto) to Franks, but then rose against them because of their brutal reign, prevailed over them and became sovereign and autonomous (aujtode>spotai kai< aujto>nomoi) (c. 30, 61– 75, 78–88) 9.

The context in which the two relevant sentences appear is tightly related to a sort of setting of, in modern terms, a geopolitical stage of East-Central Europe from the Byzantine imperial perspective, where the Croats are given a prominent place as a gens that possesses three separate regions (one is Dalmatia, the other is Illyricum and Pannonia, and the third is their ancestral land), of which the two they control were of an immediate interest to Byzantines. The mechanism by which the Croats are said to have gotten into position to possess these two regions is migration, which, as the source tells us, was twofold and happening at two different times: the Croats first came to Dalmatia and conquered the Avars, and afterwards a part of them detached from the main group and crossed to Illyricum and Pannonia, where they established a separate polity (an archontry, in Byzantine phrasing), same as there was a Croat polity in Dalmatia, which is said to have been subject to Franks for an unspecified amount of time (how and when it came under the Frankish dominion is never explicitly explained), before it became sovereign and autonomous (again). The ruler of the Croats of Illyricum and Pannonia is also accorded an autonomy as he is called aujtexou>siov (having self-authority) and portrayed as possessing right to send envoys, but his status is never described as fully authoritative as that of the prince of the Croats of Dalmatia 10. Furthermore, since the claim about the Croats of Illyricum and Pannonia having an independent ruler is in the past tense (even though the verb’s imperfective aspect is indicative of duration (cf. [88, p. 95])), it may be surmised that, from Constantine’s perspective, this was something occuring in the past and no longer extant in his days. What may be further concluded from the order of storytelling is that the Croats of Dalmatia fell under the Frankish control after the establishment of the Croat polity in Illyricum and Pannonia, but that the cordial relations between the two polities continued.

The Croats as being in Pannonia are also alluded to in Constantine’s biography of his grandfather Basil I [15, c. 52, 6–11, pp. 188–189]: “Moreover, Scythians dwelling in Pannonia, Dalmatia and beyond – I mean the Croats, Serbs, Zachlumians, Terbuniotes, Kanalites (i.e. Konavlians), Diocletians (i.e. Docleans) and Rentanians (i.e. Neretvians) – rebelled against the immemorial dominance of the Romans, and became independent and sovereign, led only by their rulers” (... eti ge mhn kai t ® n en Panonia kai Dalmatia kai t ® n epekeina toutwn diakeimenwn Skuq ® n _ Crwbatoi fhmi kai Serbloi kai Zacloumoi, Terbouni ® tai te kai< Kanali~tai kai< Dionklhtianoi< kai< F Rentanoi, thv anekaqen t ® n F Rwmaiwn ejpikratei>av ajfhnia>santev aujto>nomoi te kai< aujtode>spotoi kaqeisth>kesan, uJpo< ijdi>wn ajrco>ntwn mo>non ajgo>menoi ). This development was dated by Constantine to the reign of emperor Michael II (820–829) [15, c. 52, 1–3, pp. 188–189] ¹¹. It is important to point out that this passage has a parallel in the DAI (c. 29, 63– 66), where, however, there is no mention of Pannonia.

Variety of hypotheses ¹². The older Croatian scholarship – taking the account in the DAI at face value – saw the two sentences as a confirmation that the Croats were present in the region between the rivers Sava, Drava and Danube since their (assumed) arrival in the seventh century. Such a view was supported, among others, by Šime Ljubić [61, p. 117] and Tadija Smičiklas [84, pp. 108–109, 160–161], and it had already been maintained by the so-called father of the Croatian historiography Ioannes Lucius (Giovanni Lucio, Ivan Lučić) in his capital work The Kingdom of Dalmatia and Croatia ( De regno Dalmatiae et Croatiae ) [63, lib. I, c. 12, 21–34, p. 306], and repeated by Daniele Farlati in the second volume of his monumental work The Sacred Illyricum ( Illyricum Sacrum ) [25, Th. III, c. 3, § 2, p. 312]. The identical standpoint was adopted by Vjekoslav Klaić [54, p. 42], Josip Bösendorfer [10, p. 62], and, eventually, by the founding figure of the modern Croatian historiography Ferdo Šišić [79, p. 277]. The same opinion was also held by Kaspar Zeuss

[92, p. 612], Pavel Jozef Šafárik (Paul Joseph Schaffarik) [76, pp. 278–279], Hermann L. Krause [57, p. 4], Ernst Dümmler [20, pp. 373– 374], John Bagnell Bury [13, p. 275], Stanko Guldescu [42, pp. 92ff.], František (Francis) Dvornik [50, pp. 118–119, 121], John Van Antwerp Fine, Jr. [29, pp. 254ff.¹³], and Huw M. A. Evans [24, pp. 106, 270]. It has echoed strongly in part of the contemporary Croatian scholarship, especially in the works by Ivo Goldstein [32, pp. 9, 94; 33, pp. 220–221], and has been particulary insisted upon by Vladimir Sokol who has even assumed that the Croatian ethnie might have been present in the area north of the Drava, around the lake Balaton in modern Hungary [85, p. 194; 68, pp. 17–18]. Hence it has become a rule to speak of “Pannonian Croats”, and “Pannonian Croatia” or “Posavian Croatia” (the latter took its historiographic name after the river Sava), and the rulers (the duces of the Frankish sources) associated with the region between the rivers Sava, Drava and Danube (Liudewit, Ratimar, Brazlavo ¹⁴) have all been made into ethnic Croats. Moreover, as voiced by some scholars, even the Slav population(s) living in today’s Croatian Baranya in the tenth and eleventh centuries were Croat [75, p. 104].

Some researchers have pointed out that the presence of Croats in Pannonia does not necessarily have anything to do with the Croat migration in the seventh century, and formulated a hypothesis that the sentences actually describe the situation in the age of the Croat king Tomislav (first half of the tenth century) [39, pp. 30, 31; 52, p. 278 15]. However, there were opinions that the information about the independent Croat prince in Pannonia could only refer to the times before Tomislav [26, pp. 74–75], although some scholars raised objection that, chronologically, this would have been too early [53, p. 52].

There have also been conjectures about the existence of the autonomous Croat principality in southern Pannonia, the so-called “Pannonian Croatia”, in the early ninth century [64, p. 56; 67, p. 61], or that the rule of the Croat prince extended, after the suppression of Liudewit’s rebellion, into the region between the rivers Sava, Drava and Danube in the first half of the ninth century [3, p. 90; 12, p. 400]. On the other hand, some concluded that there is no evidence that there had ever been a unified Croat dominion encompassing both Slavonia and Dalmatia, and opted for an independent principality in Slavonia in the tenth century [32, pp. 284–285; 33, p. 231 16].

The cause for an equally fertile scholarly dispute was the area which is meant under Constantine Porphyrogennetos’s Illyricum and Pannonia. Thus, Ioannes Lucius had already believed Illyricum to be the interior of Dalmatia, and that Pannonia was Pannonia Savia, the late Roman province encompassing the western parts of the region between the rivers Sava, Drava and Danube. Some scholars assumed that Illyricum and Pannonia refer, respectively, to Bosnia and the entire region between the rivers Sava, Drava and Danube [61, p. 117], or that a part of northwestern Bosnia was meant under Illyricum [67, pp. 61, 84], or even the entire area between Dalmatinska Zagora (the mountainous region behind the coastal strip of Dalmatia) and modern Slavonia [66, p. 164]. This was decisively rejected by others, who even sought to introduce designations that are not corroborated in the sources, such as “the Illyrian Panonnia”, which, in their view, would encompass the area of the so-called “Greater Pannonia” along the edge of Illyricum or southern Pannonia, i.e. the region between the rivers Sava, Drava and Danube [88, p. 93; 53, pp. 40–42]. Others yet thought that both Illyricum and Pannonia refer exclusively to the region between the rivers Sava and Drava [57, p. 4; 39, p. 31; 26, p. 74; 12, p. 398]. Moreover, one scholar proposed that Constantine Porphyrogennetos’s Pannonia should be placed around the lake Balaton in modern Hungary, but this was refuted as unsubstantiated [41, pp. 93–95; contra 40, pp. 173– 176]. Some believed that either Noricum or Doclea (Duklja) were meant under Illyricum¹⁷. Yet another scholar thought that Pannonia refers to the area between the lower Sava and the Danube (modern Syrmia), and Illyricum to the territory encompassing Dalmatia up to the Sava in the north and the Drina in the east [20, pp. 373–374].

Posing the research questions. The variety of opinions just surveyed clearly shows quite a perplex scholarly problem as the source’s succinctness only raises a number of possible interpretations, some of which are more or less probable, and some straight out unacceptable. In approaching the analysis of the source material one has to first determine if it is at all reliable before proceeding to decipher and interpret its actual meaning. For a long time, there had been a tacit opinion among the scholars that the account is not to be suspected, i.e. that it is based on a true tradition. More recently, however, there has been a tendency to treat the historical literary sources as mere narratives, stories that create events or even invent the reality based on literary techniques and literary topoi, thus taking away their power of evidence altogether. To be sure, it is important to bear in mind that the textual sources are multilayered and multifaceted products of specific periods of time, sociopolitical, ideological and cultural contexts, goals, experiences and opinions of the authors, and views, interests and needs of the targeted audiences, meaning that the information they offer cannot be trusted at face value and simply fitted into a reconstruction that seems most likely to a researcher 18. While fully taking this into account, the approach taken here is not to reject possible historicity of what was meant by the two sentences under scrutiny, but to try to go behind this particular representation of the past by avoiding pitfalls of postmodernist and poststructuralist hypercriticism. Therefore, the following analysis is conducted under the assumption of, mutatis mutandis, basic trustworthiness and usefulness of the DAI as an historical source, not the least as a testimony of how the past was perceived. After all, had Constantine Porphyrogennetos composed his treatise by knowingly and intentionally introducing the literary creations in his text in an attempt to be seen as following the established tradition of Greek historiography and ethnography, the treatise would have lost its practical and instructive – that is to say, its principal – purpose (however much imbued with ideology and Constantine’s own perceptions and rationalizations).

This brings us to the question of the account’s actual meaning. As it seems, the real background to what it describes is not clear on several different levels. Who were the Croats of Illyricum and Pannonia? In what sense did they separate from the Croats of Dalmatia? When might this have happened if at all? What do Illyricum and Pannonia refer to? What was this independent Croat principality? How are we to understand the claim about cordial relations between the two Croat princes? Before taking up an historical analysis and proposing answers to the questions as laid out, we shall investigate three separate issues that might prove valuable for better understanding of Constantine Porphyrogennetos’s approaches in explaining the past.

Movements of peoples and groups in the DAI. Up until recently the idea of Croat migration as an historical fact established by the DAI has never been challenged. Lately, a number of sharp observations has been made aimed at dismantling of the migration narrative [16, pp. 138–139; 17, pp. 41ff.; 18, pp. 116ff.; 22, pp. 112ff.; 23, pp. 158ff.; 9, pp. 210ff.; with 2, p. 148; 1, p. 190]. Therefore, it seems worthwhile to examine how the appearance of peoples or groups in some region or their shift to another area is depicted in the DAI. These movements may tentatively be divided into several categories based on their character as presented in the DAI: conquest; branching-off; refuge seeking; unsanctioned settlement; sanctioned settlement; resettlement by invitation; planned resettlement; forced resettlement; divinely inspired migration; seasonal movements (Table 1).

Table 1. Movements of peoples and groups in the DAI

Conquest	Branching-off
c. 25, 23–24, 40–41: the Visigoths, after taking Rome, conquered the Gallic provinces and later Spain as well	c. 30, 61–67: at this time (i.e. when the Avars conquered Dalmatia), the Croats lived beyond Bavaria, and a family of five brothers and two sisters separated from them, arrived with their folk in Dalmatia, which was possessed by Avars c. 31, 3–6: the Croats who are now settled in Dalmatia are descended from the Croats who are settled beyond Turkey and next to Francia
c. 29, 14–22, 36–49: the Slavs, also called Avars, were incited by a plundering incursion made by the Dalmatian Romans across the Danube to cross the river, took Salona by a strategem, settled down in Dalmatia and gradually conquered the places of the Romans c. 30, 18–61: the Avars conquered all the country of Dalmatia and settled down in it
c. 37, 2–14: the Pechenegs used to be settled on the rivers Atil (i.e. Volga) and Geich (i.e. Ural), but then were expelled from their country. After they had fled and wandered around in search of where to settle down, they arrived to the country where the Turks dwelt. After a battle they cast the Turks out and expelled them, and settled down in their country	c. 30, 75–77: of the Croats who arrived to Dalmatia one part separated and ruled Illyricum and Pannonia
	c. 32, 3–7: the Serbs are descended from the Serbs who are settled beyond Turkey and are neighbors to Francia, and this is where they originally lived
c. 38, 55–60; c. 40, 16–27: after the Pechenegs fell upon the Turks (for the second time) and expelled them, the Turks were on a search for a land to settle in, and they expelled those who dwelt in Great Moravia and settled down in their land, where they live to this day	c. 32, 16–18: one part of the Thessalonican Serbs decided to depart to their own homes and were sent off by the emperor
	c. 33, 8–10: the Zachlumians are Serbs from the time of that prince who came in refuge to the emperor Heraclius
	c. 33, 16–18: the family of the prince of Zachlumians, Michael, came from those who live on the river Visla
	c. 34, 3–6: the Terbuniotes and Kanalites are descended from the unbaptized Serbs, from the time of that prince who came from the unbaptized Serbia and in refuge to the emperor Heraclius
	c. 35, 1–9: the land where the Diocletians now dwell was populated in the time of the emperor Heraclius, just as were Croatia, Serbia, Zachlumia, Terbounia and the country of Kanalites
	c. 36, 5–9: the Pagani (also called Arentanians) are descended from the unbaptized Serbs, of the time of that prince who came in refuge to the emperor Heraclius, and their country was populated in the time of the emperor Heraclius
	c. 37, 50–53: one part of the Pechenegs, at the time they were expelled from their country, remained and united with the Uzes, and they are among the Uzes even to this day
	c. 38, 24–31: after the Turks were defeated by the Pechenegs (for the first time), they split in two parts, one part settled in Persia, while the other settled in the western region, where now the Pechenegs live

End of Table 1

Refuge seeking	Sanctioned settlement
c. 31, 8–10, 26–27: the Croats came in refuge to the emperor Heraclius before the Serbs did the same; the land, in which the Croats had settled, was originally under the authority of the Roman emperor	c. 25, 25–29: the Goths were allowed to dwell in Thrace, and then were permitted to conquer the Western Empire (i.e. Italy)
c. 32, 7–10: one of the two brothers who succeeded the reign over White Serbia took one half of the folk and came in refuge to the emperor Heraclius	c. 32, 10–11: the Serbs who came in refuge to the emperor Heraclius were granted the province of Thessalonica to settle in
c. 41, 20–25: the country of the Moravians was destroyed and conquered by the Turks who even presently dwell in it, while the Moravians were scattered and fled for refuge to the Bulgars, Turks, Croats and the rest of the nations	c. 32, 18–27: the Thessalonican Serbs who left the province of Thessalonica on a return migration were granted by the emperor Heraclius other land to settle in and settled down in what is now Serbia, Pagania, the country of Zachlumians, Terbounia, and the country of Kanalites
Unsanctioned settlement	Resettlement by invitation
c. 25, 15–21: the Goths, Visigoths, Gepids and Vandals, who were settled in the regions across the Danube, crossed the river and settled on the Roman soil	c. 25, 32–49: the Vandals were persuaded by the Roman general in Africa, Boniface, to cross the sea strait and settle in Lybia
c. 25, 24–25: the Goths first held Pannonia	c. 27, 32–37: at the time the Lombards lived in Pannonia, where the Turks presently dwell (c. 27, 30-31), they were invited by the patrician Narses to come to Italy and settle in it, and so they took their families and arrived in Beneventum
c. 25, 29–32: the Vandals joined with the Alans and Germans (i.e. Franks), crossed the river (Rhine) and settled in Spain
Planned resettlement	Forced resettlement
c. 29, 1–5; 33, 3–5; 35, 3–5; 36, 3–5: the emperor Diocletian brought families from Rome and settled the Romans in Dalmatia	c. 28, 4–16 (cf. c. 27, 71–75): those who are now called Venetians were Franks from Aquileia and from other places in Francia, and they lived on the mainland opposite Venice. But after Attila, king of the Avars, devastated and depopulated all the parts of Francia, the Franks from Aquileia and other cities of Francia began to flee and to come to the inhabited islands, and after the king Attila advanced as far as Rome and Calabria leaving Venice behind, those who fled to the islands decided to settle there and they have been settled there to this day
c. 47, 15–25: the emperor Justinian II decided to populate Cyprus, persuaded the caliph to dismiss the folk of Cyprus that were in Syria, and resettled those who lived in other parts of the empire	c. 29, 49–50: after the Slavs/Avars’ conquest the remaining Dalmatian Romans saved themselves in coastal cities c. 29, 230–235: there was a migration of a group of people from Salona to Rausium
c. 32, 135–142: the Serbs who were staying in Croatia, Bulgaria and the rest of the countries came back to Serbia, and those who escaped from Bulgaria to Constantinople were sent back by the emperor to Serbia	c. 39, 2–7; 40: 6–7: the Kabaroi, who were of the race of Chazars, were forced to escape and came and settled with the Turks in the land of the Pechenegs
	c. 32, 123–126: the Bulgars entered Serbia and took away with them the entire folk and carried them into Bulgaria, though a few escaped away and entered Croatia
Divinely inspired migration	Seasonal movements
c. 45, 10–12, 38–39: the Iberians originate from Jerusalem, and were warned by an oracular dream to migrate from there and settle near Persia, in the country where they now dwell	c. 8, 34–35: the Pechenegs cross the river Dnieper to spend summer on its far side
	c. 9, 105–111: at the beginning of winter, the Kievan Russians go off to other Slav regions of their tributaries and come back to Kiev in spring
	c. 11, 6–8: along their routes, the Chazars pass through to Sarkel

Of particular interest here is the manner in which the DAI explains the appearance of the new groups after the arrival of Croats and Serbs, namely, the Zachlumians, Terbuniotes, Kanalites, Diocletians and Pagani (Arentanians). The Serbs are said to have been allowed to settle by the emperor Heraclius in what is now Serbia, Pagania, the country of Zachlumians, Terbounia, and the country of Kanalites, which is then further corroborated by a statement that some are Serbs

(Zachlumians), while others (Terbuniotes, Kanalites, Pagani/Arentanians) are descended from Serb’s from the time of the Serbs arrival from their ancestral land. This would posit that the Serbs split off in several additional groups at a very early stage. The only excluded are the Diocletians, even though their land is said to have been populated in the time of the emperor Heraclius as well ¹⁹. The key to understand Constantine Porphyrogennetos’s logic behind these claims of branching-off, i.e. connecting what were obviously diverse identity groups under one ethnic umbrella label, seems to be his limiting of the duration of the Terbouniotes’ and Kanalites’ being descended from the Serbs: until the prince of Serbia, Blastimer, made the zupan of Trebounia independent (c. 34, 6–10). That is to say, Constantine equated the ethnic identity affiliation with the political overlordship. He seems to have followed the same “misconception” when he turned the Venetians into Franks (c. 28, 4–6). The DAI provides additional hints to such an interpretation: Pagania is indicated to have been subject to the prince of Serbia in the early tenth century (c. 32, 84–85), and the princes of Trebounia are said to have been always under the command of the prince of Serbia (c. 34, 11–12).

What also merits attention is the claim by some scholars that what separates all the migration stories in the DAI from the story of Croat (as well as of Serb) migration is that Constantine Porphyrogennetos did not provide any reason for their movement 20. That is not entirely valid observations since the DAI does not provide any reasons for initial movements of the Goths, Visigoths, Gepids and Vandals as well, the story of which was copied from the Chronographia of Theophanes Confessor [89, A.M. 5931, pp. 93– 95], who, in turn, excerpted Prokopios of Caesarea’s Wars [74, lib. 3, c. 2–3, pp. 311–324; lib. 5, c. 1, pp. 4–10], where much more detail is given. Furthermore, the Croats are also depicted as seeking refuge, albeit the details as to the reasons why are not specified. The reasons are perhaps alluded to in the passage where it is said that the Croats’ ancestral home is constantly plundered by Franks, Turks and Pechenegs (c. 31, 85–87) 21. Even though this information belongs to a different time plain, it may be that the exposure of both the Croat and Serb ancestral lands from external attacks is implied as the main reason for their migration.

Earliest chronology in the so-called Dalmatian dossier of the DAI . Recently it has been – rightly – inferred that the description of events in the Dalmatian dossier of the DAI does not represent an historical chronology, even though it may be presumed that the events closer to Constantine Porphyrogennetos’s own time are much more reliably dated [1, p. 190]. Therefore, the attempts to pinpoint the exact time of the earlier events amount to nothing more than sheer speculations (see, for example, attempts in [67, pp. 56ff.; 65, pp. 144ff.]). Of some use might be the order in which the events are arranged as it would suppose that this, at least, reflects in general the sequence in which they occurred (departures from this are due to narrative peculiarities). What may be termed as the Dalmatian dossier is comprised in three chapters of the DAI (c. 29, 30, 31). The order of the events related to the earliest history is given in the synoptic table below (see Table 2).

What may be surmised from the comparison is that, given its position in the master narrative comprising of all the three Dalmatia-related stories up to the point when the Croats are said to have settled down in Dalmatia, the account of the Dalmatian Croats’ split-off group’s arriving to Illyricum and Pannonia seems to refer to some later times. The intrinsic narrative logic of the integrated story would suggest that, since the entire Dalmatia with all its nations and cities is said to have been under Byzantine rule from the reign of Heraclius until the reign of Michael II, the creation of another Croat polity is – in the representation of the past as offered – to be rather connected to the time of the latter emperor.

Illyricum, Pannonia and Dalmatia in the DAI . The last of the three interim issues investigated here is the examination of what was understood in the DAI under the regions of Illyricum, Pannonia and Dalmatia. Illyricum is mentioned only once in the entire treatise ²², but since the perspective is eastern Roman, i.e. Byzantine, it is fairly certain that the Illyricum of the DAI refers to the late Roman prefecture of Illyricum. Furthermore, as it is mentioned in connection to Pannonia, it surely signifies the part of the prefecture that was geographically adjacent to Pannonia ²³.

Table 2. Synoptic table for the Dalmatian dossier in the DAI (c. 29, 30, 31)

DAI , c. 29	DAI , c. 30	DAI , c. 31
3–11: the resettlement of Romans to Dalmatia, and the foundation of Spalato and Diocleia in the time of the emperor Diocletian	14–17: the emperor Diocletian built Salona and Spalato	11–14: the emperor Diocletian brought the Romans to Dalmatia and settled them there
14–15: the territory possessed by Romans used to extend as far as the river Danube	8–14: the province of Dalmatia, which used to extend as far as the confines of Dyrrachium, the mountains of Istria and the river Danube, was under the rule of the Romans before it was taken by Slavs
15–18: beyond the river Danube lived Slav nations, who were also called Avars	18–21: a force of cavalry, collected from the other towns of Dalmatia, was dispatched from Salona to keep guard on the river Danube against the Avars
	21–23: the Avars had their haunts on the far side of the river Danube, where the Turks are now
18–22: the Dalmatian Romans made an incursion into the Slav territory, overcame the Avars and took booty and prisoners	23–30: the men of Dalmatia crossed over into the Avar territory, attacked the Avars when their men and youths were on a military expedition, and came back to Salona carrying booty
22–32: the Romans introduced frontier station at the river, and the starting point for advancement of the exchanging garrison to the station was the frontier pass called Kleisa near Salona
32–38: the Slavs, who were also called Avars, ambushed, and defeated the Romans after they once more crossed over to their territory	33–39: another garrison was dispatched from Salona, they also crossed over into the Avar territory, but were defeated and slain or captured
38–46: the Slavs proceeded to attack Salona, took it by a strategem and settled there	39–56: the Avars proceeded to attack Salona and managed to take it by a strategem
46–49: the Slavs began to make raids and took possession of the lands of the Romans	56–58, 60–61: thereafter, the Avars made themselves masters of all of Dalmatia and settled down in it	15–17: the Romans of Dalmatia were expelled by the Avars in the time of the emperor Heraclius
49–53: the remaining Romans escaped to coastal cities and possess them still	58–60: only the coastal cities held out against the Avars, and continued to be in the hands of the Romans
	61–63: at the time, the Croats were dwelling beyond Bavaria, where the White Croats are presently
	63–67: one part of the Croats split off, came to Dalmatia and found the Avars in possession of that land	8–11: the Croats of White Croatia arrived to Dalmatia in refuge to the emperor Heraclius, at that time when the Avars fought and expelled the Romans

End of Table 2

DAI , c. 29	DAI , c. 30	DAI , c. 31
54–58: the whole of Dalmatia and the nations in it were subject to the empire of the Romans since the reign of Heraclius	67–71: the Croats and Avars fought one another for some time, until the Croats prevailed, killed some of the Avars, and subjugated the remainder, and from that time Dalmatia has been possessed by Croats	17–21: by command of the emperor Heraclius, the Croats defeated and expelled the Avars, and by the mandate of the same emperor, they settled down in that same country of the Avars, where they now dwell; at that time the Croats had for prince the father of Porgas 58–60: the prince of Croatia had ever since the time of the emperor Heraclius been subservient and subject to the emperor of the Romans
		21–25: the emperor Heraclius had the Croats baptized by bringing priests from Rome; at that time the Croats had Porgas for their prince
	75–78: one part of the Croats who came to Dalmatia split off and possessed themselves of Illyricum and Pannonia, where they also had an independent prince 78–82: for a number of years the Croats of Dalmatia also were subject to the Franks
58–66: the cities and nations of Dalmatia became independent in the time of the emperor Michael II	82–88: the Croats of Dalmatia rebelled against the Franks and won their independence
233–236: 500(!) years has passed from the migration from Salona to Ragusa till the present year, which is A.D. 949	71–75: the rest of the Croats who stayed behind in their ancestral land are now subject to Otto, the king of Francia, and maintain good relations with the Turks	3–6: the Croats of Dalmatia are descended from the White Croats who live beyond Turkey and next to Francia

Pannonia is mentioned directly three times and implied three more. Apart from the mention in connection to the Croat split-off group, Pannonia is also said to have been in possession of Goths before their transfer to Thrace (c. 25, 24–25), and is identified as the region where the Lombards were dwelling, and where now the Turks (i.e. Magyars) live (c. 27, 30–31). Given the latter, it may be presumed that it is also implied as the territory of Gepids who are said to have been later divided off by the Lombards and Avars and previously lived in the area of Sirmium (as well as that of Singidunum) (c. 25, 21–23), and as the territory of Turks who are also singled out as holding Sirmium (and Belgrade as well) (c. 40, 29–32). Moreover, the Turks’ is also the land between the rivers Danube and Sava (c. 42, 19–

20). Sirmium (and Belgrade, too) is mentioned as one of the landmarks along the Danube as well (c. 40, 31).

Dalmatia, understandably, received much more attention. It is said to be the most illustrious of all the provinces of the West (c. 30, 11–13) and the most fair country (c. 30, 60–61). Dalmatia is described as extending from the confines of Dyrrachium24 as far as the mountains of Istria and the river Danube (c. 30, 8–11). Its territory is implicitly matched with the territory that is said to have once been – i.e. prior to the Slav conquest – possessed by Romans and extending to the river Danube (c. 29, 14–15). In the tenth century, the Byzantine theme of Dalmatia encompassed all the remaining Byzantine possessions on the eastern Adriatic save those that belonged to the theme of

Dyrrachium²⁵. The theme of Dalmatia is referred to in the DAI as the “whole of Dalmatia” separated from other nations (the Croats, Serbs, Zachlumians, Terbuniotes, Kanalites, Diocletians and Arentanians) that are about it (c. 29, 56–58; cf. also 32, 24–25), and as the cities of Dalmatia (c. 29, 61–63, 110) as well as the islands of Dalmatia (c. 29, 285–293). All these other nations are said or implied to live in the territory of the former (Roman-time) Dalmatia, and their countries, as presented in the DAI , extended in some cases beyond the boundaries of what is known as the late Roman province of Dalmatia (including Praevalitana). Accordingly, the Croats are said to border with the Turks on the latter’s southern side (c. 40, 43–44), and, simultaneously, the Turks are pointed out as possessing the land between the Danube and the Sava (c. 42, 19– 20), which would suppose that the Croats and Turks must have shared a border in this area and that this portion was meant under the Danube to which Dalmatia is said to stretch (assuming that the DAI has not confused the Sava with the Danube ²⁶). Consequently, it may be that the DAI has simply (but not particularly precisely) translated the contemporaneous geopolitical circumstances to earlier times.

There is however an alternative explanation. Constantine Porphyrogennetos may have had the Croats of Illyricum and Pannonia before his eyes, and since they are presented as a group that was – to put it in the phrasing of the DAI – descended from the Croats of Dalmatia but no longer extant as a polity in Constantine’s own time, he merely included them – following the logic of equating the ethnic identity affiliation with the political overlordship – under their Dalmatian counterparts which, by extension, meant that they also could be regarded as part of Dalmatia ²⁷. Finally, yet another explanation might lie in the sphere of the imaginary, since the Danube could have been used in this context as a symbolic marker that divided the Romans from the barbarians, a perceived landmark of liminality between the two culturally disparate worlds ²⁸. Therefore, in regard to the northern boundary of Dalmatia in the DAI , the Danube might have been used as the divide symbol rather than the concrete watercourse, even though both of these meanings have been merged²⁹.

Analysis and discussion. Let us now return to the research questions related to the two sentences from the DAI that are the focus of this investigation. Before we proceed, we shall make a cursory point related to the recent state of research of the early medieval ethnic identities. Succinctly put, the ethnic identity was a quite different category in early medieval societies than it is in modern societies, even though some national historiographies still seem slow in escaping the grips of nineteenth-century perceptions 30. To sum up, the current socio-anthropological discourse is that the early medieval ethnic identity is a situational construct, a fluid and constantly created and re-created subjectivity of the group that is based on the perception of the group’s history and traditions. The ethnic identity is created by the group from within the group but also from outside the group, through the perception of differences toward other groups. At the same time, the ethnic identity is also imposed by external observers that control the discourse thanks to their political as well as ideological and cultural domination.

In other words, the feeling of ethnic affiliation in the early Middle Ages had much less to do with the awareness of common descent and shared cultural traits than with political, ideological and economic interests. It was primarily a matter of the élite and contingent on the success of the fundamental (eponym) group (or several such groups united by a strong appreciation of shared kinship and origin) in imposing their own traditions and values on other groups and populations 31. In the case of the Croats, there could have really been several groups since the DAI, our main source for the Croatian origo gentis, has preserved the tradition about seven brothers and sisters who purportedly led the Croat split-off group during their migration to Dalmatia (c. 30, 63–67). Therefore, being part of an early medieval people (gens, natio) was more than anything else an ideological political category, and especially in the perception of external observers. Consequently, the “Croats” are most likely to have been – in early medieval terms – an ethnic-identity label for various populations that began to identify themselves with the eponym group and, in this way, to form an ethnie. The label could also be applied only to members of the core group(s), i.e. the élite as holders of power and influence, through which the ethnonym would become a marker for authority, that is, a population or populations, or even another ethnie might have been perceived as “Croat” in the eyes of external observers if the Croats imposed their rule or exerted their influence over these populations, or if a member of the Croat élite became their ruler.

The last remark aims directly at the actual meaning of the story in the DAI about part of the Croats splitting off and possessing Illyricum and Pannonia. Taken at face value, it would mean a migration of a portion of the Croat ethnie from Dalmatia to a new abode (which Constantine Porphyrogennetos, strictly speaking, says), or it could signify the transfer of political dominion. Not discarding a priori the possibility that there really was another Croat migration (but certainly nothing like a large-scale movement, which may also be valid remark for their original, assumed, migration from the ancestral land), it must be said that the archaeological and anthropological research – notwithstanding methodological limitations – has so far not yielded any evidence to support such an interpretation. Craniometric analyses of skeletal remains discovered in graves located in the area from the Adriatic into the deep inland seem to indicate that the so-called “Old Croat” populations gradually spread from the coast into the interior to southern Pannonia in the tenth to thirteenth centuries [81; 82, pp. 106–107; 83]. The Dalmatian Croat(icized) populations have been determined as clearly differing by their anthropometric and craniometric traits from the continental populations from Vukovar and Bijelo Brdo – which is actually suggestive of the predominance of pre-Slav (and pre-Croat) population substrate in the post-Roman Dalmatia – while the populations from the site Gomjenica near Prijedor (based on archaeological finds, the site has been attributed to the so-called Bijelo Brdo cultural complex) are included in the Dalmatian Croat(icized) populations [80, pp. 105–106]. Slow penetration of the Croat influence to the north is further backed by jewelry finds from the same period, which may be taken as evidence for direct contacts between the Dalmatian Croat and south Pannonian Slav cultural circles [86, pp. 138–141; 87, pp. 136–142]. Under the assumption that these artifacts were specific to the Croat cultural circle to the point of being used as identity markers, both the anthropological and archaeological findings definitely suggest that the early Croats had not settled at all in southern Pannonia during their original (assumed) north-south migration, even though it is possible – if we accept that the Croat migration happened at some point and that the Croats traversed, on their southbound move, the region between the Danube, Sava and Drava – that there might have been minor Croat groups staying behind in this area and eventually fading away in the mass of other Slav(icized) populations. It would appear that the spreading of the so-called “Old Croat” populations from the south to the north belongs to the period from the tenth century onward and that it is related to the altered political situation, the strengthening and expansion of the early Croat state.

Based on all of these assumptions it is more likely that the ethnonym “Croats” and the purported reverse migration hide the circumstance of the transfer of political dominion. This would mean that Constantine Porphyrogennetos, as already argued, identified the political overlordship with the ethnic presence, i.e. ethnicity. When could this presumed transfer of political dominion have taken place? Certainly, before the tenth century since there could not have been an independent prince in “Pannonia and Illyricum” at the time when Constantine Porphyrogennetos was compiling his treatise (the mid-tenth century), which he himself alludes to by using the past tense in connection to this Croat prince. The ninth century is the only possible candidate given that southern Pannonia was still part of the Avar Kaganate in the late eighth century.

Let us now try to determine what Constantine Porphyrogennetos may have meant under Illyricum and Pannonia. A logical point of departure is the assumption that, in his use of provincial administrative designations, he had before his eyes the contemporary situation as well as the situation in Late Antiquity since he also had written sources from that time at his disposal, and he is known to have used them. This duality could explain apparent illogicalities that exist in his account regarding the use of the names of Pannonia, Dalmatia and Illyricum, i.e. their overlapping. As has been pointed out, under Illyricum the early Byzantine sources regularly understood the prefecture of Illyricum which was bound on the northwest by the Sava and the Upper Drina watercourses. The late Roman Pannonia encompassed the area from the northern slopes of the Bosnian mountain range just below the Sava to the Drava and Danube in the north. Dalmatia stretched from the southern borderline of Pannonia down to the Adriatic coast. Assuming, sensibly, that the Illyricum and Pannonia mentioned in the account have to be areas in immediate proximity to each other, since they stood under the authority of one and the same prince, at first there seems to be a plenty of possibilities for various speculations.

A proposed solution to the conundrum lies, as I see it, in the circumstances that existed in Late Antiquity. In principle, the label Pannonia could be applied to all of the territory included in the late Roman Pannonian provinces ( Pannonia Prima, Valeria, Pannonia Savia , and Pannonia Secunda ), or, in the broadest sense, could denote the diocese of Pannonia ( dioecesis Pannoniarum , also known as the diocese of Illyricum) ³². Following the collapse of the Roman administrative system in the Middle Danube region, southern Pannonia between the rivers Sava and Drava retained a certain value for the eastern Roman government, even though it had been almost continuously under the sway of various barbarian groups (the Ostrogoths, Gepids, Lombards) (for details, cf. [36, pp. 167ff; 38, pp. 193ff.]). The Eastern Roman Empire succeeded ultimately in preserving, until the Avar invasion, the southeasternmost part of the Second Pannonia ( Pannonia Secunda ) with the town of Bassianae (Donji Petrovci east of Sremska Mitrovica) as a forward stronghold. This area was also known in the early Byzantine literary sources as “the Herul land” ( hJ F Elou>rwn cw>ra ), after the Heruls who were first settled there in 512 ³³, and it was joined, together with Sirmium as long as the city was in Byzantine hands, with the prefecture of Illyricum and considered part of Illyricum. Therefore, it may be assumed that Constantine Porphyrogennetos’s Pannonia refers to the region between the rivers Sava, Drava and Danube (modern northwestern Croatia without Medjimurje, Slavonia and western Syrmia), and Illyricum to the Sirmium-Bassianae area, i.e. eastern Syrmia ³⁴.

It is no coincidence that this area matches with the so-called Principality, or Duchy, of Lower Pannonia, or the regnum inter Savo et Dravo flumine [7, a. 884, p. 113], which was created by Franks after they had toppled the Avar Kaganate in the early ninth century. Hence, it seems that Constantine Porphyrogennetos had precisely the

Principality of Lower Pannonia in mind, whose duke enjoyed a certain independence under the Frankish overlordship. One of the features of an independent position was the right to send embassies. Frankish sources suggest that it was customary for subject peoples to send envoys to the Frankish emperor. Consequently, Constantine Porphyrogennetos’s remark about the “Croat” prince of Illyricum and Pannonia dispatching envoys is in essence accurate. Especially close relations between the duke of Lower Pannonia and the prince of Croatia, as surmised from the DAI , should be seen as nothing more than an amicable relationship resulting from the same concerns in line with overarching interests of their Frankish overlords as they both governed their respective territories under the tutelage of the Franks (the duke of Lower Pannonia remained a Frankish vassal until the collapse of his regnum due to the advent of Magyars ³⁵, while the Croat prince seems to have enjoyed a significant freedom of action from very early on).

Proposed reconstruction for a conclusion. Based on the analysis and discussion conducted above I argue that the separation of Croats and their move into southern Pannonia, as depicted in the DAI , should not be understood as an actual migration but as a case of the identification of political dominion with ethnicity. Furthermore, the Illyricum and Pannonia of the DAI seem to fully correspond with the territory of the so-called Principality of Lower Pannonia during the ninth century. Finally, the cordial relations between the duke of Lower Pannonia and his counterpart in Croatia were – at least at first – the result of the political unity imposed by their Frankish overlords, since both of these polities were Frankish creations.

The identification of political dominion with ethnicity suggested here means that, in Constantine Porphyrogennetos’s view, the Croats were actually in control of southern Pannonia, i.e. the so-called Principality of Lower Pannonia. To be sure, it is usually believed that, during the first quarter of the tenth century, the Croat polity of Dalmatia managed to extended its rule and influence over some portions of southern Pannonia with the center at Siscia (see [35, pp. 200ff.]). Thus, it may well be that Constantine had this in mind. However, such an explanation cannot account for the claim that the Croats possessed

Illyricum and Pannonia, that is, the entire Lower Pannonia, unless he made a blunder, which is not likely considering that the developments mentioned above were close to his own time and age. Consequently, we have to look somewhere else for an answer. It is known that the Franks intervened twice against the local dukes in Lower Pannonia: there was a proper war fought against Liudewit (819–822), while Ratimar was overthrown in a single campaign (838) (apart from [35, pp. 157–168, 175–178], cf. also [91, pp. 307– 310, 312, 314; 11, pp. 60–71, 99–100]). The sources are tacit on what happened with the Principality after Ratimar’s removal (for that matter, they have nothing to say about what happened immediately after Liudewit had been ousted, although we may assume that Ratimar became Liudewit’s successor at some point, possibly in 831), but the Franks would certainly have been still very interested in finding a permanent solution to stabilize the southeastern frontier area of their empire. If this suggestion holds water, it is quite possible that, looking for a suitable candidate to fill the position of the duke of Lower Pannonia, they turned their attention to the Adriatic coast.

I would thus propose that the Franks chose as Ratimar’s successor a member of the Croat élite from the nearby Croat polity in Dalmatia, which, at that time, had been faithfully adhering to the Franks for quite a while. Attracting this presumed member of the Croat warrior élite and installing him as the duke of Lower Pannonia would not have been an unusual course of action for the Franks. Similarly, the former duke of Nitra, Pribina, after he had been forced to flee from the duke of Moravia, Moimir I, was eventually established as the Frankish vassal duke (dux) of Lower Pannonia north of the Drava and centred on Moosburg (modern Zalavár). The new duke of Lower Pannonia between the Sava and the Drava would have doubtlessly been loyal to those who had appointed him, the more so since he would have lacked a power base in the region. Moreover, it may be assumed that the south Pannonian Slavs would not have perceived the rule of a duke of the Croat descent as something alien since there was no language barrier or conflicting customs or practices. The presumable Frankish support for their appointee would have certainly suppressed any opposition to such an arrangement. From the perspective of a Byzantine emperor writing in Constantinople one century after the fact, with no clear understanding of the situation in distant times and with no access to accurate sources, and in an attempt to elucidate the past as best as he could, it might have indeed appeared as if the Croats had also established their rule in southern Pannonia. If the proposed reconstruction is correct, the move was a success for the Franks given that there is no more news in the sources of troubles in the Frankish southeast that were caused by unruly and disobedient dukes. The last known duke of Lower Pannonia, Brazlavo, was to the last a faithful follower of the Frankish emperor, and even entrusted with the control of Lower Pannonia north of the Drava as his fief as well [35, pp. 189– 193; with 91, pp. 92, 119; 11, pp. 245, 271].

For the final note I would like to suggest that the designations “Pannonian Croatia” and “Posavian Croatia”, which are commonly used by modern scholars to denote the early medieval Slav principality in southern Pannonia, be dispensed with, since such a Croat polity had never existed. The renown Croatian medievalist, Nada Klaić, consistently argued against the use of both of these labels, emphasizing that they are a scholarly invention by modern historians and not based on contemporary sources (cf., for example, [53, pp. 40, 45, 51]). They may perhaps only be used in a broad geographical sense, with a clear idea of their modern context, where Posavian Croatia has a narrower meaning than Pannonian Croatia. Instead of these anachronistic and unwarranted names, Lower Pannonia or the Principality or Duchy of Lower Pannonia, or perhaps even the Sava-Drava Principality (after the regnum inter Dravo et Savo flumine from the Annales Fuldenses ) should be preferred for the early medieval polity of south Pannonian Slavs that existed in the region between the Sava, the Drava and the Danube under Frankish tutelage during the better part of the ninth century.

NOTES

• ¹    This paper is a re-written, expanded and updated version of the article first published in [34], but the basic arguments are unchanged.

• ²    For a brief survey on the DAI and its purpose, with references to the relevant scholarship: [37, pp. 25– 27]. Now see also: [71, pp. 130–137; 56, pp. 41–58].

• ³    About the production process of the DAI , cf. [49, pp. 306–310]. However, his overall appraisal of Constantine Porphyrogennetos’s effort seems to be too harsh. For less critical assessment, see [78, pp. 97– 98]; with [8, pp. 553–554]. For a humorous but to-the-point justification of obstacles that the emperor might have met in the formation of the text: [77, pp. 189–192].

• ⁴    Cf. also [71, pp. 135–136], and [71, pp. 54–60] for the compilation method.

• ⁵    For an example of such an approach, despite some interesting assertions: [9]. A critical appraisal of Borri’s approach and methods: [5, pp. 45–49].

• ⁶    The idea behind this assertion is inspired by the approach of the so-called Viennese school to the construction of early medieval identities, but the concept of a firm group identity embedded in an immutable binding core of tradition is implicitly abandoned, since the acceptance and maintaining of an ethnic identity had to be negotiated and reaffirmed, especially under external ideological-political and cultural influences and in adverse situations, however, a group ethnic identity could be inherited and exercise a sufficiently powerful constraint upon its individual members. Cf. [73]; with [48, pp. 26–47; 59, pp. 87–96].

• ⁷    The basic conclusions of this paper have been reproduced in [1, p. 300].

• ⁸    This translation differs somewhat from the standard translation by R.J.H. Jenkins in [14, p. 143]. For the peculiarities related to the translation of this spot, cf. [62, pp. 352–356].

• ⁹    For an interpretation and a possible meaning of the story about the Croats of Dalmatia revolting against Franks, cf. [37, pp. 29, 31–32]. For a different take, see [2, pp. 144–146, 157–158, 163–164], who explains the story as an invented event that served the military élite as a structural ideological element to open a way for their political emancipation from the Franks and the formation of a new independent ethnopolitical organism in the Frankish Dalmatia – the regnum Chroatorum . However, it seems far-fetched to conclude that the story of a hard-won war against the Franks was entirely a fabrication, and the only other incident of such magnitude in the wider area was the anti-Frankish rebellion of the Lower Pannonian duke Liudewit.

• ¹⁰    M. Suić [88, pp. 94–95] has interpreted this place as indicating a foedus ( symmachía ) existing between the two polities, where the prince of Croatia was primus inter pares and, in recognition of his primacy, the prince of the Croats in southern Pannonia sent him embassies with gifts. It might be worthwhile to address here a question of the English rendering of Slavic ruler titles. In the Croatian scholarship, it has been customary to use the word “knez” ( dux or comes in Latin) as a title for early medieval Croat rulers before Tomislav who is usually taken to be the first king of

  Croatia, as well as for early medieval Slavic rulers in southern Pannonia. The word “knez” is better to render in English as a “prince”, since a “duke”, which is commonly used in English-speaking scholarship (see, however, [29, pp. 253ff.]), translates in Croatian as “vojvoda” and does not necessarily denote a ruler in his own right. The problem which an English-speaking scholar might encounter when trying to properly translate these titles into English is shown, for example, in F. Curta’s two monographs about history of southeastern and eastern Europe in the Middle Ages. In his Southeastern Europe in the Middle Ages 500– 1250 , Curta uses the title “duke” for Borna [16, p. 135], Liudewit [16, p. 135], Trpimir [16, p. 137], Branimir [16, pp. 140–141], and Muncimir [16, p. 140], and “prince” for Tomislav [16, p. 137] and Domagoj [16, p. 140], which is not consistent. His somewhat of a dilemma is alluded to in the remark “early dukes or princes” [16, p. 141]. Similarly, in his Eastern Europe in the Middle Ages (500–1300) , Curta uses the title “duke” for Borna [17, p. 109], Liudewit [17, p. 109], Brazlav [17, p. 110], and Tomislav [17, p. 325], but “chieftain” for Ratimir [17, p. 110] and “prince” for Branimir [17, p. 591]. In German, the word “Fürst” is used, which is more appropriate. Therefore, in this paper the word “prince” is used for the Greek archon, whereas “duke” denotes a ruler under tutelage. For a discussion on Latin and Slavic titles of early medieval Croat rulers, cf. [31].

• ¹¹    The translation by I. Ševčenko has been slightly modified.

• ¹²    A partial survey of Croatian scholarship on this matter is also offered in [1, pp. 298–301].

• ¹³    J.V.A. Fine Jr. [30, p. 20] rightly expresses reservation with regard to the term “Pannonian Croatia”, but he uses it nevertheless later in his text [30, pp. 21, 49, 72].

• ¹⁴    In the Croatian scholarship, the names of these south Pannonian Slav dukes are regularly rendered in their Slavicized / Croaticized forms as Ljudevit, Ratimir, and Braslav.

• ¹⁵    This is followed in [60, p. 371 (note 19)].

• ¹⁶    Similarly to this was already maintained in [61, pp. 136–138].

• ¹⁷    Noricum: [76, p. 279; 43, p. 256; 44, p. 340]. Doclea: [45, pp. 22–23; 46, pp. 337–338; 47, pp. 112– 113]. Cf. also [32, p. 91].

• ¹⁸    This is a remark that has already been stressed by the author in several of his papers. Cf., for example, [38, p. 186, with note 5].

• ¹⁹    Simplistic is the claim in [55] that Constantine Porhyprogennetos considered the Diocleatians to be Serbs, even though he does not mention them as such. Komatina’s arguments are all quite circumstantial and the fact that Diocleia is not mentioned as a country settled by Serbs is explained as the emperor’s failing to follow strictly to the principle of listing countries and peoples in details.

• ²⁰    This was first observed in [62, pp. 328–329] (even though Lončar mentions that the words “in refuge” might allude to a reason), and accentuated in [22, p. 112, note 83].

• ²¹    In the previous chapter of the DAI (c. 30, 71– 75) it is, however, said that the Croats of White Croatia are subject to the king Otto of Francia and in friendly relations with the Turks (i.e. Magyars), which would suppose that the attacks by Franks and Magyars predated these new political arrangements.

• ²²    The Illyria mentioned in the chapter devoted to the Iberians (c. 45, 159) is the Armenian province of Aghori (cf. [50, p. 177]).

• ²³    That this Illyricum was adjacent to Pannonia had already been suggested in [39, p. 31].

• ²⁴    These confines are, on their northern side, set by Constantine Porphyrogennetos at Ulcinium ( DAI , c. 30, 95–96).

• ²⁵    On the theme of Dalmatia in the late ninth and the tenth centuries, see [28, pp. 165–191]. On the theme of Dyrrachium in the ninth and the tenth centuries, see [51, pp. 337ff.].

• ²⁶    For misidentifications of the Danube in ancient literary sources and Sirmium as a switch point between the two hydronyms for this watercourse (Danubius/Ister), cf. [19, pp. 138–139].

• ²⁷    It is perhaps instructive as well that the 13^th-century chronicler, Thomas archdeacon of Split, when describing the extent of the kingdom of Dalmatia and Croatia, its northern border also sets at the banks of the Danube [90, c. 13, pp. 60–61].

• ²⁸    For the construction of the Danube as a sociocultural frontier in Antiquity, see [70, pp. 751–755].

• ²⁹    It is worth noting that B. Ferjančić [27] does not dwell on these issues related to the boundaries of Dalmatia.

• ³⁰    For this problem in the context of modern Croatian historiography, cf. [4].

• ³¹    For a discussion on identities in post-Roman Dalmatia cf. [21; 22, pp. 64–73].

• ³²    Main sources for the administrative division of the later Roman Empire are the Notitia Dignitatum [72] and the Laterculus Veronensis [58] .

• ³³    The phrasing is by Menander Protector [69, fr. 5.4, pp. 52–53]. Cf. [38, p. 200, with note 60, p. 219].

• ³⁴    Constantine Porphyrogennetos’s Illyricum and Pannonia are reminiscent of the Illyricum and Pannonia mentioned in the Annales Laureshamenses’ account of the Frankish campaign against the Avars in 791 [6, a. 791, p. 34]. However, since the perspective is here decidedly western, it is most likely that southwestern parts of the former diocese of Pannonia, i.e. Noricum mediterraneum (the territory of modern Slovenia), were meant under the Illyricum. Cf. [35, pp. 150–151].

• ³⁵    For details on political developments in southern Pannonia during the ninth century, cf. [35, pp. 154–194].