Strata of Ethnics, Languages and Settlement Names in the Carpathian Basin 1

: When entering the Carpathian Basin in the 9 th century, the Hungarians found a decisively Slavic population on the territory, so toponyms were formed based on the already existing toponymic system. Hungarian toponymic research has been able to re - construct toponyms from the period prior to the Hungarian conquest only very scar - cely and ambiguously – as opposed to the names of larger rivers, which show strong continuity, going back to very early times. The toponyms of the Carpathian Basin, in connection with the formation of the settlement structures of Hungarians, can almost exclusively be documented from the period after the Hungarian conquest. However, the Carpathian Basin became a “meeting point of the peoples” in the centuries after the conquest in 896 and as such, numerous ethnics and languages could be found here: Slavic peoples and Germans settled in larger blocks, while smaller groups of Turkish people, such as Cumans and Pechenegs, and some Neo-Latin peoples (Walloons and later Rumanians) also contributed to the ethnic and linguistic diversity in the area. The layering of different peoples and languages influenced toponyms too, which also allows us to investigate language contacts of the time. This is the main concern of my paper, with special focus on the question of how these phenomena can be connected to issues of language prestige in the Middle Ages.

Ethnic relations experienced in the Carpathian Basin today basically become transparent in the period after Turkish occupation. Namely, a large part of Hungary (especially its southern region and the Great Hun garian Plain) was victim of the Turkish expansion wave in the 16 -17 th centuries (1526 -1686). The Turkish rule, which in the area lasted for almost 150 years, caused widespread panic, the depopulation of villages and emigration, and in this respect, the Turkish rule greatly diminished former settlement structures. The repopulation of the region, and its ethnic "reanimation" was the major event of population history in the century after the expulsion of the Turks and it was realised partially by relocating Hungarians from other areas, and partially by resettling other ethnic groups (especially Germans). Consequently, the present-day ethnic image of the Carpathian Basin was roughly developed due to these resettlements and it was followed by state formations in the 20 th century.

2
Ethnic diversity appeared in a slightly different manner in the early medieval period. Naturally, the ethnic relations in the Carpathian Basin became interesting from a Hungarian perspective only after Hungarians settled and formed a state in the 9 th century at the terminal point of their Eastern European migration. (We are not dealing with the period preceding it for this reason.) Within the early history of Hungarians, the period worth studying both from a historical and onomastic point of view is the earliest period, which includes the first long period of the newly developing Hungarian monarchy and the three centuries of the Árpád dynasty (1000 -1301). Namely, this is the first period about which we can make statements on ethnic relations based on straightforward linguistic facts -only with a degree of caution, of course. First, let us have a brief look at what history, using its own resources, could discover about this period, about the migration of different people to the Carpathian Basin (about the possible settlement place and time of certain people). We assume that Slavic peoples were predominant at the time Hungarians arrived to the Carpathian Basin. Slavic peoples in the period of the Hungarian Conquest settled mainly in Western Trans danubia, and the hills and mountains around the Great and Little Hun garian Plain, while after the Hungarian Conquest they mainly populated the unsettle d peripheral areas. Slavic people living in the central area became assimi-lated and were linguistically Magyarised shortly after the Hungarian Conquest. While we can show connected settlements of Slavic people, there are only fragments of Turkish people. The Pechenegs were the most numerous in the Carpathian Basin and we assume they settled from the second half of the 10 th century. However, as indicated on the map as well, there were only smaller, fragmentary settlements. The earliest settlement period for Cumans was probably the beginning of the 12 th century and primarily in the Danube -Tisza interfluvial region where they populated larger connected areas. Among Turkish people were also the Ismailite traders, Besermans and Khalyzians, who came in small numbers from the Hungarian Conquest until the 13 th century. There was no significant German population in the Carpathian Basin until the arrival of Hungarian people. They settled in larger numbers only after the Conquest: primarily Saxon settlers (hospeses), who formed significant settlement blocks in the north (Upper Hungary) and in the east (Transylvania). Among Neo-Latinspeaking people, Rumanians populated the eastern areas after the age of the Árpád dynasty, and there were smaller French groups at the beginning of the 12 th century, especially in the northern mountain areas. (For more details see Korai magyar tör téneti lexikon [KMTL. Dictionary of Early Hungarian History] , Györ ffy 1963-1998, Kristó 2000 For the map see also KMTL.,[432][433] 3 For an early ethnic picture of the Carpathian Basin, historical sourcesbecause of their scantiness -are insufficient, of course. The picture drawn in this article appears fragmented for two reasons: firstly from a chronological and secondly from a territorial perspective. This is why a long time ago scientists started using linguistic data, especially toponyms, as traditionally important sources to decide historical issues. The toponym source material from the age of the Árpád dynasty is not too rich either, but the number of name data from before 1350 is at least several tens of thousands, which is by any standard a significant quantity of data. The use of toponyms from charters in ethnic reconstruction procedures is well exemplified in the works of János Melich (A honfoglaláskori Magyarország [Hungary in the Conquest period], 1925 -1929) and István Kniezsa (Magyarország népei a XI-ik században [Hungarian people in the 11 th century], 1938) respectively from the first half of the 20 th century. Their work, however, needs to be reconsidered. Therefore, in what follows I am calling attention to a few methodological difficulties concerning one narrow stratum of toponyms -settlement names. In connection to ethnic relations, we may consult the settlement names in a twofold manner. Namely, there are two types of settlement names which are liable to serve as a basis for ethnic conclusions. Firstly, settlement names which refer to an ethnic group semantically may have a similar function: for example, the settlement name Németfalu ('German village'), of Hungarian origin, refers to 'a village inhabited by Germans'. The same role may be attributed to those name forms which are related to an ethnic group etymologically. For example, Visegrád is a name most probably given by Slavic people, namely, Visegrád settlement name is of Slavic origin and it means 'upper castle'.

1
Names belonging to the first group, the Németfalu-type, cannot be disregarded when drawing a picture of ethnic composition, although they are not the result of foreign but of Hungarian name giving, since they show the presence of a given ethnic group at the time of name genesis. Settlement names of ethnonymic origin are the oldest stratum of the Hungarian toponym system and the name entities belonging to this group can be of three types. The following names developed purely from ethnonyms, without additional formants: Cseh 'Czech', Tót 'Slovakian', Orosz 'Russian', Olasz 'Italian', Besenyő 'Pecheneg', Kun 'Cumans', Böszörmény 'Beserman' 2 , Német 'German' settlement names; with a toponym formant (especially with the derivational suffix -i expressing possession): Csehi, Tóti, Olaszi, Németi, and finally, attached to the ethnonym, a geographical common name second component bearing a 'settlement' meaning (for example, falu 'village', város 'town', telek 'estate') formed toponyms of the Tótfalu 'village inhabited by Slovakians', Olaszváros 'town inhabited by Italians', Besenyőtelek 'estate inhabited by Pechenegs' type. The ethnic group appearing in the names of settlements like these inhabited the given settlements and this is what motivated the Hungarian population when giving a name. Denominations like this, of course, only make sense in a foreign language environment, since where all settlements are in habited by Germans, a Német type of name does not have a distinguishing role. Settlement name of ethnonym origin therefore do not denote larger blocks of given nations, but rather smaller islands within a larger nation's (in this case Hungarian) sea. However, when talking about the source value of this name type, we have to bear in mind an important hindering factor. The majority of ethnonyms in the Hungarian language in the age of the Árpád dynasty was also used as a personal name: we have data about Cseh, Tót, Orosz, Besenyő, Német, Böszörmény personal names (cf. ÁSz.). It can be easily imag ined, therefore, that certain settlement names which seem to have ethnonymic origin actually became toponyms through personal names and do not refer to the ethnic affiliation of the inhabitants, but merely point to the owner of the estate: in other words, some villages called Németfalu are not 'villages inhabited by German people' but 'ones owned by a German person'. What is more, the ethnic identification role of personal names of ethnonymic origin is completely different from settlement names of ethnonymic origin: Német 'German', Cseh 'Czech', Tót 'Slovakian' personal names can barely be used to denote a given person's and not a larger ethnic group's ethnic affiliation.

2
Loanwords, names borrowed from a foreign language, belong to the second settlement name category which can be used to discover the early medieval ethnic relations in the Carpathian Basin. Names of Slavic origin, such as Csongrád 'black castle', Nógrád 'new castle', Kosztolány 'those belonging to the church', Tapolcsány 'those living at the small aspen grove', Zemplén 'made of soil [i. e. castle]', etc. provide evidence of Slavic people in the area at the time of name giving, the same way as Késmárk 'cheese market' provide evidence of German and Tállya 'clearing' of French pre sence. However, the source value of Csongrád, Késmárk, Tállya type of names is also limited, since they provide clues regarding a given ethnic groups' (Slavic, German, Walloon) presence only for the period of name genesis.
(However, we do not have knowledge of this; since centuries may pass afte r name genesis before the name is recorded in a legal case.) Namely, the Hungarian population started using these names and integrated them into its own name system, as a consequence of which these names referred to the memory of a former name giving nation even when it had been already long assimilated into the Hungarian population. For all these reasons, it is important to emphasise that loanwords, centuries away from the original situation of name forming, naturally lose their ethnic identi fication role and can be used in ethnic reconstruction procedures only with great caution.

3
The indications of the two settlement name types (the Németfalu type referring to the ethnic group via Hungarian name giving and the Csongrád type identifying the name giving community with a word of foreign origin) are far from being in accordance with each other. Namely, while the former group is quite diverse and gives evidence of many different ethnic groups (Slavic, German, Neo-Latin and Turkish people), the latter form a rather homogenous type: a huge majority of settlement names in the early centuries of the Carpathian Basin can be deduced from Slavic or Hungaria n name giving, and besides them, one or two German or Walloon names may randomly appear. Nevertheless, there are no Beserman or Cuman names from this age. I believe that behind the different characteristics of the two settlement name types there are two reasons: a chronological one and the specific procedure of charter-issuing practice in Hungary.

4
In connection to chronological factors, we may assume that each ethnic group represents a different temporal stratum. As I have already mentioned, the people of the Hungarian Conquest most probably found Slavic people whose way of life was a settled one. The large majority of Slavic origin settlement names dated in the early charters were probably given before the Hungarian Conquest, and the nomadic and semi-nomadic, as well as the settling Hungarian people borrowed them. There are numerous settlements among Slavic ones which gained enormous impor-tance in the life of Hungarians: for example, Csongrád and Nógrád, which became names of comitats, or Visegrád, which served as a royal residence for a while. It is important to emphasise that these settlements obviously had Hungarian population at the time of charter issuing; however, they never appear in Hungarian translation. The earliest settlement and settlement name stratum of the Carpathian Basin therefore is of Slavic origin. The Hungarian people built onto this early stratum; in a few centuries they developed the region's settlement structure and, in parallel, its name network, integrating a part of the Slavic settlement names into its own name system. What is more, evidence for the early presence of Slavic people can be found in the linguistic factors in connection with the appearance of bigger and middle-sized Car pathian rivers in Hungarian. That is to say that these river names, both those of early (old European) and of Slavic origin became a part of Hungarian corpus through Slavic mediation. This is important to mention even if these river names -because of their extended feature -do not have an ethniclocalisation value. These linguistic factors (whether we are looking at the borrowing of hydronyms or toponyms) provide evidence for a strong Slavic substratum in the early centuries. The settlement or rather population by other ethnic groups (Germans, Turkish people, Walloons) was realised through the effective contribution of a higher, royal power. Villages which were established by settling these people were not named by the settled population, but rather by the Hungarian-speaking environment or the higher power forming the village. This is why it is possible that none of the ten German villages established in Abaúj county in the 12 th century carries a German name. Their majority was named after a German personal name, but via Hungarian name giving: for example, Gönc (cf. German Gun(t)z < Kuntz < Konrad personal name, FNESz., Tóth 2001, 65), Korlát (cf. German Konrad personal name, FNESz., Tóth 2001, 92), Vilmány (cf. old German Wilman personal name, FNESz., Tóth 2001, 159), Céce (cf. German Zitze < Siegfried personal name, FNESz., Tóth 2001, 36), etc. In addition, what refers to a great block of Saxon settlers are not settlement names of German origin but the adjectival first component Szász-'Saxon' attached to a primary Hungarian name form (also given by the Hungarians): for example, Kézd > Szászkézd, Régen > Szászrégen, Sebes > Szászsebes, etc. These secondary name forms imply that other ethnic groups settled into the villages already settled (and named) by Hungarians as a new ethnic stratum. Nevertheless, we cannot exclude the possibility of certain privileged ethnic groups forming their own village -the already mentioned Késmárk and Tállya may be examples of this type. They are, however, by no means general phenomena.

5
The trace of foreign ethnic groups may also be carried by toponyms of their origin, only in a different name form: not at the level of settlement names, but at the level of microtoponymic systems. If we look at the present-day regions' bilingual toponymic systems -as an analogy of old relations -we experience that although both communities named objects in their environment in their own languages, the secondarily settled ethnic group rarely changes the early Hungarian settlement names: at most, it transforms it to make it more suitable to the phonology and phono tactics of their own languages (cf. Póczos 2008). The language sociological differences between settlement names and microtoponyms may explain the different attitude of secondarily settled ethnic groups to certain place classes and their names. The basic difference between them is that settlement names belong to the category of cultural names (in other words, they are names of objects formed by humans) whereas the majority of microtoponyms belongs to the category of natural names (i. e. names of natural items). Furthermore, in the creation of na tural names, it is the linguistic-communicational needs that play a vital role, whereas in the case of cultural names the significance of social motivation increases (cf. Hoffmann 2007, 101). For this reason, the conscious, intentional nature (and sometimes the inclination to assert the interests) could be realised to a greater extent than with natural names, as a consequence of which the name giving community can be distinguished more definitely from the broader layer of the name using community (Hoffmann 2007, 110 -111). This also means that even if a village inhabited by Germans and Hungarians had a German name besides a Hungarian one, for reasons of prestige, the former very rarely got to the level of written records and even then it was always mentioned together with the Hungarian name. What is more, we can also see in the early medieval charter issuing practice that in the areas of the Hungarian Monarchy populated exclusively by Germans, besides German names of the settlements there are Hungarian names appearing (Loránd Benkő draws our attention to this phenomenon). This may refer to Hungarian name forms (and consequently to Hungarian ethnic environment) but we may also consider linguistic interference of charter writers, i. e. artificial creation of Hungarian name forms (basically a Hungarianisation procedure) (cf. Benkő 2003, 130, see also Kristó 2000, 27, Hoffmann 2007. In connection to ethnic relations this means that even charter data of otherwise key importance can be misleading; they may hide the actual relations and we may assume the Hungarian population was not present in all regions indicated by the data. The Latinised procedures are excellent instances of linguistic interference of charter writers: namely, certain toponym classes (names of bigger rivers, mountains, comitats, and names of more significant settlements) and toponym types (for example, patrociny settlement names) frequently occur in charters which were either translated into Latin or which went through a Latinisation procedure (cf. Hoffmann 2004).

6
In charter issuing procedures like this, early linguistic prestige relations probably played an important role. In the Hungarian history of toponyms research it was István Hoffmann who was the first to emphasise that the model developed by Abram de Swaan for the present day global linguistic system can be applied to the linguistic relations of the age of the Árpád dynasty Hungary. According to de Swaan, languages form a hierarchical system on the basis of their prestige relations in which the lowest level is occupied by peripheral languages grouped around certain central languages. Among the central ones are the supercentral, connected by a hyper central. Today the only language with a hypercentral role is English (2001). In István Hoffmann's opinion, in Hungary in the age of the Árpád dynasty the hypercentral language was Latin, which on the one hand connected the region to European international communication and on the other hand served as the language of internal communication in certain spheres (education, state administration, law, certain areas of church life). Furthermore, in the beginning, it had exclusive role in written records as well. In the multicultural Carpathian Basin, a supercentral role extended to the whole of the country was that of Hungarian: this language was used in secular official issues connected to ownership (for example the enclosure and record of estates' borders, hearing of witnesses, etc. were all conducted in Hungarian). In written records, besides Latin, Hungarian also spread in a limited manner. The German language had a central role in its own area (in other words, in the region of settlements in age of the Árpád dynasty period and maybe in the western region). Besides Latin, German gradually gained ground in written records and it was also used at certain levels of oral communication. Other languages of the Carpathian Basin (different Slavic and Turkish languages, Rumanian and several other languages used by less numerous populations) were peripheral languages used almost exclusively in interpersonal communication. (See also Hoffmann 2007, 90 -91). All these factors influenced the Hungarian clerks of the royal office conducting the issue of charters to include the Hungarian names of settlements into legal documents when the given settlement had different language variants. At other times, they did not hold back from creating a temporary Hungarian variant of the foreign name form at the level of written records. This kind of procedure affected the Visegrád, Csongrád type of Slavic origin settlement names the least, because they were -due to the fact that they belonged to the earliest stratum -fully integrated into the Hungarian settlement name system and therefore -since they are considered full-fledged elements -we never come across their Hungarian equivalent. The Slavic settlement name stratum from a later period naturally went through the same dual usage as the one referred to in connection to German names. During the use of early linguistic records in the ethnic reconstruction procedure and during the settlement name strata of the Carpathian Basin, researchers did not take into consideration the latter factor, although this is an aspect which should not be left out when exploring the real situation.