Toponymy and Chronology

In light of the foregoing, an approach that combines etymology, typology, and chronology should make it possible to trace the evolution of the place-names linked to the various ethnic groups in Transylvania.

However, great caution must be exercised: unless the toponyms are analyzed according strict rules, the process can degenerate into unscientific speculation. The etymological derivation {1-350.} of a toponym is crucial only in the case of words that describe natural phenomena, and even then, it must be considered in the context of comparative linguistics. For example, the Hungarian word 'bükk' (beech) comes from the Slavic 'bik', but when it appears as the toponym 'Bükk' or its derivation, 'Bükkös', it must be regarded as a Hungarian place-name, originating with a Hungarian community that had been subject to some Slavic linguistic influence. Particular care must be exercised with regard to toponyms drawn from personal names, for the latter are commonly drawn from a multitude of sources and ages. Obviously, the Hungarian medieval village called Izsák (Isaac) did not get its name from a biblical community in biblical times; Alparét is derived from the German Albrecht, but this Hungarian village did not get its name through the application of German toponymic customs; similarly, Drág is derived from a Slavic personal name, but it is not the Slavs who applied the typically Hungarian, nominative singular form to designate a village. When the toponymic typology of different ethnic groups can be clearly distinguished, typological criteria must prevail over etymological considerations. Finally, chronology must take precedence over both typology and etymology. The Hungarian place-name Csesztve (Alsó-Fehér county) is derived from the Slavic proper name 'Čestivoj', but it displays the Hungarian toponymic pattern of place names. The Romanian name of the place, Cisteiu, is clearly derived from the Hungarian toponym; had it come from the Slavic 'Čestivoj', the Romanian equivalent of which is 'Cestivoiu', then the Romanian toponym would be 'Cestivoiul' or 'Cestivoii' — like Berivoii, in the Fogarasföld, which was adapted from a Slavic proper name in keeping with Romanian toponymic custom. In the case of Csesztve, the etymological and typological analysis is complemented by the chronological aspect: the Hungarians had taken a Slavic proper name and used it for a place name. The linguistic exchange between Slavs and Hungarians preceded — and did not coincide with — the naming {1-351.} of the locality. If a Hungarian toponym is adopted and adapted by Romanians, it must have been well-established before the arrival of the latter.

A different problem — and one which is highly instructive — arises with places that have multiple names, in two or more languages. The meaning of the several names may or may not be the same: thus the Hungarian/Romanian/German toponyms of two localities in Szeben county, Veresmart/Roşia/Rotberg and Fenyőfalva/Brad/Gieresau. It might be supposed (indeed, many believed) that these are simply translations. In fact, what commonly happened was that the ethnic group which arrived later did not translate the established toponym but took it over or adapted it phonetically. On the other hand, multiple names such as those cited above generally occur when at the time of the locality's establishment, there are several ethnic groups in the vicinity, and each of them formulates a toponym in its own language. Sometimes, the pattern in uniform: in the first example, all three toponyms are inspired by the site's red soil. Alternatively, the several toponyms may focus on different characteristics; in the second example, the Hungarian and Romanian toponyms evoke the site's vegetation (pines), while the German one (which means 'Gellert's grove') indicates the proprietor. Thus in the case of such multiple names, it may be assumed that the ethnic groups which gave rise to them were cohabiting in the district.

A study of Transylvania's toponyms was conducted in accordance with the previously-noted principles and completed some time ago. Thanks to its findings, the earlier and current names of the major rivers, along with the place-names that figure in pre-1400 documents (285 of them before 1200, 602 from the first half of the 14th century, and 1120 from the second half, for a total of 2007) can be analyzed and correlated with archaeological evidence to trace the geographical and historical links between the name-giving ethnic groups and to establish the chronological order of the latter. {1-352.} Since the documentary sources are sparse and often contradictory, they will be cited only when they contribute indisputable chronological evidence.

One of these uncontested facts is that in 892, the Bulgars were shipping salt to the Moravians, and that this salt could only have come from Transylvania. Thus, at the very least, some salt-yielding districts of Transylvania must have been under Bulgar control. The graveyards at Maroskarna and Csombord and other, more scattered Bulgar traces in southern Transylvania are obviously linked to this presence, and one can therefore expect to find {1-353.} Bulgaro-Slavic toponyms.

The second chronological fact is the Hungarian conquest of the Carpathian Basin in 895. The Hungarian pagan rituals revealed by many graves found along the river valleys of the central Transylvanian Basin survived no later than the first decade or two of the 11th century, when they were generally supplanted by Christian rites. One would expect to find in the vicinity of these 10th-century burial places toponyms that are derived from Hungarian personal names, without article or suffix. There was probably some earlier, sporadic settlement by Germans, giving rise to toponyms that included the word 'Németi', but according to a document from 1224, the massive settlement of Germans in southern Transylvania took place around the middle of the 12th century. The earliest German place-names in the region presumably date from the same period.

The earliest evocation of Transylvanian Romanians is found in a document, dating from 1222, which refers to an 'Oláh land' (terra Blacorum) in the Fogaras district. Some Romanian toponyms could date from this period, and perhaps from earlier, but it is difficult to establish their precise date of origin. Romanian shepherds may have started centuries earlier to practice transhumance and drive their flocks from the Southern Carpathians to winter pasture on the Lower Danube. Historical and archaeological sources indicate that Bulgaro-Slavic toponyms appeared in the 9th century and Hungarian ones in the 10th; German ones did not appear in any number before the middle of the 12th century. It is therefore in this context that one must attempt to date the Romanian toponyms.

The names of rivers are usually the best indicators of the language spoken by a larger area's indigenous population. In Transylvania, the names of the three largest rivers — the Szamos, Maros, and Olt — go back to ancient times. The sole reference to the Szamos that survives from Roman times names it 'Samu(s)', and the earliest Hungarian version is Zomus (1231). The Maros is first mentioned in Herodotus, then appears as 'Morisos', 'Marisus', and 'Marisia'; its name is rendered in early Hungarian sources as 'Moris' (Saint Gellért's Deliberations, ca. 1044) or 'Morus' (Anonymus' chronicle from the early 13th century). The Olt appears in Ptolemy, then as 'Alutus' on the Peutinger tablets, as 'Aluta' in Jordanes, as 'Alt' in 1211, and finally, in 1233, as 'Olt'. Two more rivers may belong to this list. One is the Abrud: the earliest reference, dating from 1271, is not specifically to a river but to a terra Obruth. Some linguists believe that this name is derived from a hypothetical Dacian word, 'obrud', meaning gold. The name of the other river, Ompoly, may have a connection with that of the town — 'Ampei(um)' — noted in a Roman inscription; in 1299, the river is referred to as the 'Ompey' or 'Ompay'.

The main difference between the ancient and the medieval form of these river-names is the vowel shift a > o, or, with Abrud, o > a. If Transylvanian Romanians had adopted and adapted the Roman names, then Romanian linguistic rules would have produced 'Sameş', 'Mareş', 'Alut', 'Aurud', and 'Împei'. In fact, the rivers' modern Romanian names are Someş, Mureş, Olt, Abrud, and Ompoi or Ampoi. Instead of coming directly from the Roman, these names must have been adapted from those applied by other ethnic groups that cohabited with the Romanians in Transylvania. The first three are derived from the likely Slavic forms, 'Somuš', {1-354.} 'Moriš', and 'Olt' (the a > o vowel shift is typical of all Slavic languages), either directly or via the Hungarian adoption of these forms. The Romanian 'Abrud' must have been borrowed from the Hungarian, for the transformation of Obruth into Abrud, which occurred in the 14th century, reflects the typically Hungarian vowel shift o > a (as in the case of the Szamos and Maros). The Ompoly is a similar case, reflecting the typically Hungarian vowel shift ej > oj (cf. 'olej' > 'olaj'). The Romanian variant 'Ampoi' is derived from the German 'Ampoj', which is an adaptation of Ompoly by the typically Transylvanian-German vowel shift o > a (cf. 'Omlás' > 'Hamlesch').

Thus these ancient names were adapted into Romanian through the linguistic mediation of the Slavs, Hungarians, or Germans. The same can be observed in the case of other Transylvanian rivers that pass through the more populated areas, where there was a greater likelihood that their names would be lodged in the collective memory. In what follows, the names of tributaries of the three major Transylvanian rivers will be traced from the Middle Ages to modern times, in three versions: (a) Hungarian, (b) Romanian, and, where applicable, (c) German. The Hungarian forms, and those borrowed from Hungarian by (b) and (c), will be written in normal characters. Slavic forms, and those borrowed by the others from Slavic, will be written in italics. German forms, and those borrowed from German, will be marked by capital letters. Names of unidentifiable origin will indicated by question marks.

First, the Szamos river-system. The tributaries of the Kis Szamos, from its source to its confluence with the Nagy-Szamos, are, on the right, the Fenes — Feneşul and Füzes — Fizeş; and, on the left, the Kapus — Căpuş, Nádas — Nădaş, Fejérd — Feurd, Borsa — Borşa, Lóna — Luna, Lozsád — Lojardul, Ménes — Valea Chicedului. Of the nine tributary names, seven are of Hungarian, and two of Slavic origin (from Lovina and Luža), but even the latter were adapted into Romanian from the Hungarian. In the case of the {1-355.} 'Ménes', the Romanians did not adopt this medieval name but adapted, instead, the more recent Hungarian variant 'Kecsedi patak (brook)'. The tributaries of the Nagy-Szamos are, on the right side, the Rebra — Rebra, Szalva — Salva, Széples — Ţibleş, and Ilosva — Ilişua; and, on the left side, the Leses — Leşes, Ilva — Ilva, Sajó — Şieu — Schogen, ?Budak — Budac — Budak?, Beszterce — Bistriţa — Bistritz, Tiha — Tiha, Lekence — Lechinţa — Lechnitz, and Mélyes — Patac. In this mountainous region, some names are of Hungarian origin, but most of them are Slavic, and some of these were borrowed by the Romanians directly from their Slavic neighbours. The tributaries found between the confluence of the Kis- and Nagy-Szamos and Transylvania's western are, on the right, the Beregszó — Bîrsău, Lápos — Lăpuş, Szőcs — Suciul, and Debrek — Dobric; and on the left, the Deberke — Deberche, Hagymás — Hăşmaş, and Almás — Almaş. Thus two names are of Slavic, and the rest, of Hungarian origin; one of the Slavic names passed into Romanian via the Hungarian, while the other may have been borrowed directly by the Romanians. The more imortant fact is that not one of the Maros' significant tributaries bears a name of Romanian origin. The majority are of Hungarian origin, one comes from the German, and those of Slavic origin passed into Romanian, in roughly equal proportion, directly or via Hungarian.

The tributaries of the Maros, to the right and down to the point where the river veers south, water part of the Mezőség plain; they are toponymically similar to the tributaries of the Szamos, which water another part of the same plain. The river-names Lucs — Luci, Komlód — Comlod, Ludas — Ludoş, and Aranyos — Arieş indicate that the mountains were inhabited by Slavs and the plain by Hungarians. The pattern changes in the valley of the Aranyos, whose tributaries are the ?Abrud — Abrud? (see above), Torockó — Trăscău, Jára — Iara, Hesdát — Heşdate, and Tur — Tur. These names indicate the presence of Slavs and Hungarians, who were later joined by Romanians. The same observation applies in the {1-356.} case of the tributaries farther downstream along the Maros: to the right, the Gyógy — Geoagiu (1), Ompoly — Ompoi and AMPOI (see above), Gyógy — Geoagiu (2), and to the left, the ?Görgény — Gurghiu? (a brook with the same name, in Transdanubia, is mentioned in 1211), Nyárád — Niraju, the two Székás — Secaş streams, Sebes — Sebeş, Bisztra — Bistra, Dobra — Dobra, Sztrigy — Streiu — Strehl, Farkad — Farcadin, Cserna — Cerna, and, again, Dobra — Dobra. The major tributaries are the Kis-Küküllő and Nagy-Küküllő, whose Hungarian appellation is of Turkic origin, and has the same meaning — 'full of blackthorns' — as the Slavic name borrowed by the Romanians, Tîrnava. When the Hungarians arrived in this region (and its southern vicinity), they encountered a mixed Slavic and Turkic population. It can be concluded (with a degree of certainty that varies by district) that the Romanians came into contact with Slavs along the middle reaches of the Maros; there, they borrowed river names from the Slavs, Hungarians, and — in one instance — from the Germans. The absence of Romanian-derived river-names confirms that the Romanians arrived after the Slavs, Hungarians, and Germans. But whereas, in the Szamos region, it is merely a possibility that the Romanians adopted some Slavic river-names without Hungarian mediation, in the Maros valley, it is a certainty. This makes it likely that the cohabitation of Romanians and Slavs began earlier and lasted longer in the south than in the north.

A different ethnic composition is revealed by the names of the Olt's tributaries. A striking proportion of the names is of unknown (i.e. definitely not of Slavic, Hungarian, German, or Romanian) origin; some are said — more or less plausibly — to be derived from a Turkic language (Barót, Tömös, Barca, Tatrang, Brassó, Zajzon). The Slavic, Hungarian, and German river-names (and those with variants in all three languages) are intermingled. The tributaries to the right of the Olt are called Árapatak — Arpatac, ?Barót — Baraolt?, Kormos — Cormoş, Vargyas — Varghiş, Homoród — {1-357.} Homorod — Hamruden, Kozd (< Kövesd, 1206) — COSDU — Kosbach, Hortobágy — HÎRTIBAV/Hîrtibaciu — Harbach, CIBIN — ŢIBIU — Zibin, Feketevíz — Cernavoda — SCHWARZWASSER, Sebes — Sebeş — Schewis, and Cód — Sad — Zoodt; the ones of the left at the Feketeügy — Fechetig, Kászon — Caşin, Kovászna — Covasna, ?Tatrang — Tîrlung?, ?Zajzon — Zizin?, ?Tömös — Timiş?, ?Brassó — Braşov?, VIDOMBÁK — GHIMBAV — WEIDENBACH, ?Barca — Bîrsa — Burze?, Hamaród — Hamarud — SCHELLENBACH, Sebes — Sebeş, Árpás — Arpaş, BESINBÁK — BEŞINBAV — BESCHENBACH, and ?Porumbák — Porumbac?. These names suggest that the Olt region's earlier inhabitants were Slavs and another, perhaps Turkic people (just like in the Küküllő region to the north), both of whom were still present when the Hungarians, Germans, and Romanians arrived. It appears that the Romanians were the last to arrive, for they invariably borrowed river-names from the Germans and the Hungarians.

Thus the analysis of river names confirms the message borne by archaeological finds: the Hungarians who settled in Transylvania during the 10th century encountered Slavs throughout the region, along with a small Turkic group in the southeast, near Küküllő and Olt rivers.