Is Rumanian an indigenous development out of the Latin of the Roman colonists in the area of Dacia to the north of the Danube, and has it been spoken in the territory of present day Rumania ever since the times of the Roman Empire? B or is it a more recent importation from elsewhere, presumably south or west of the Danube? There has been a long debate over this point, not unrelated to nationalist territorial claims. The Aofficial@ view is that the former of these two hypotheses is valid, despite absence of concrete evidence to support it. The later hypothesis has been, in general, rejected.

The merit of the present book is that it marshals and presents the evidence on behalf of the second, or Anon-Dacian@ hypothesis. In the absence of detailed historical records for the post-Imperial period in the non-Byzantine regions of the Balkans, we must rely on the indirect evidence furnished by linguistic and other cultural relaitonships. Dr. Du Nay has sifted and weighed the relevant material with thorough and objective scholarship, making it evident that the Anon-Dacian@ hypothesis must be taken seriously and reckoned with.

We shall of course never know what actually happened in those centuries when Proto-Rumanian was being formed; nor exactly where it took shape, unless by some chance we come into possession of more detailed historical accounts than are at present known. In the meanwhile, it is good to have Dr. Du Nay´s reasoned presentation of the Anon-Dacian@ side of the argument, to set over against the dominant official view.

I feel honored to have been asked by André Du Nay to write a brief Preface to the new edition of his book The Origins of the Rumanians B The Early History of the Rumanian language.

The point I will try to make in this preface is one which is becoming more and more recognized in Linguistics both in Europe and in the USA B the fact that the planet´s natural languages are living eco-systems. What this means is that there are no isolated or random events in the history of a language and that, therefore, everything that happens in the course of a language´s development organically interdepends with every other part. Thus, phonemic change affects the morphology; the loss of morphological endings, in turn, affects the syntax and the semantics of a language. That this is so is a language universal regardless to where one does one´s research.

Think of English, for example. Between the arrival of the Angles, the Saxons and the Jutes and the Norman Conquest in 1066, the British Isles developed a number of versions of Anglo-Saxon, or Old English. Whatever the differences between these OE dialects were, they were all Germanic, even if one adds the Scandinavian loan words that were added by the several Viking raids on Britain. But in 1066 all of this changed B the conquering Normans brought Old French with them and since they were the dominant nobility, an overwhelming amount of French vocabulary entered the English language changing its character forever.

The Middle English Period lasted for a relatively short time B from the Norman Conquest till the beginning of the 16th century; a mere three hundred years. This was also the period that saw the AGreat Vowel Shift@ take place between 1400 and 1500, one of the most mysterious and least understood events in the history of a major Indo-European language. The greatest poetic genius of this Middle English period was, of course, Geoffrey Chaucer.

Nobody is seriously questioning in Linguistics today whether English is a Germanic or a Romance language. The answer is obvious to all: English is, basically, a Germanic language with a heavy overlay of Romance vocabulary. Yet if some extraterrestrial visitors were to scan the printed output of the English speaking world with some megacomputer for a month and send a report to their mother ship, they would probably come to the conclusion that English is one of the Romance languages. How wrong would such a report be? Not very B on the surface, at least. Words such as international, situation, atomic, bomb, revolution, abandon, masses, report B the kind of ´intellectual language´ newspapers and magazines are full of B is not of Anglo-Saxon provenance. Our extraterrestrials simply would have made the mistake of going by the printed media only and not listening to the conversations of families at home, where they would have heard words such as brother, mother, father, sister, son, daughter, foot, arm, hand, fire, water, etc.

Someone exposed to this kind of home-spun, basic English only, could understandably miss the Romance elements in English and say AEnglish is a purely Germanic language.@

Obviously the truth is that English is both Germanic and Romance with a very large number of additional loan vocabulary from Greek, Italian, Spanish, Swedish, Norwegian, Arabic, Hebrew, Chinese and Japanese B to mention just the few most obvious sources.

The United States is a complex culture of immigrants B including the 10 American Indian Phyla with the many languages therein that are no longer mutually intelligible, since they span from the Arctic to Tierra del Fuego and cover anywhere between 10.000 and 35.000 years of linguistic separation. Yet the Indian presence is tangibly there in state names (Delaware, Illinois, Iowa, Massachusetts, Oklahoma, Wisconsin) and in city names (Chicago, Wichita, Kansas, etc.). Our Black population speaks Black English and creates enchanting music and poetry in this particular version of English. Our public and private schools and universities cater to our Hispanic population in a wide variety of programs known as ´bilingual education´.

It is well known that almost every place-name in America is either of English, French, Spanish, Italian, or Amerindian origin B with a few of the minor languages of the world also having contributed interesting data for toponymic research. Ypsilanti, Michigan is, for instance, named after the Greek word upsilon, the ancestor of the Latin letter [y], and the city of Philadelphia, where the Declaration of Independence was read on July 4, 1776, means ´City of Brotherly Love´ also from the Greek.

The English word new also occurs before French and Spanish place-names B consider New Orleans and New Mexico.

The city of Detroit, MI. pronounced /ditróyt/ derives from the French for ´straights´, ´narrow sea or fresh water passage´, D´étroites. One look at the map of the Great Lakes, and one can see why the French named it that. In the State of Wisconsin, a famous town is called Fondulac, pronounced /fá:ndñulaek/. It is also a French toponym, fond du lac [fo dü lak], meaning ´bottom of the lake´.

Many people mistakenly believe that Chicago and Michigan are also French B the reason for this belief is the pronunciation of ch as /š/, obviously a French habit. It turns out that both words are of Algonquian Indian origin. It was simply that the Indian words were first heard and recorded by French explorers who applied French orthography to the Indian words. Chicago means ´wild onion leaves´ or ´skunk weed´, and the morpheme spelled [-chi-] is the same in both Chicago and Michigan, referring to the sharp smell. Out of the five Great Lakes only Lake Superior has an English name (curiously following the French word order from an earlier Lac Supérieur,) the rest, Erie, Ontario, Huron and Michigan are all of Amerindian origin.

The Spanish language has contributed a tremendous number of Hispanic toponyms to the USA. Consider Colorado, Florida, San Francisco, San Antonio, Las Vegas, Palos Verdes, San Juan Capistrano B the list is a huge one. ´The colorful one´, the ´flowery one´, Saint Francis, Saint Anthony, ´The brushes´, ´green branches/sticks´, ´Saint John of Capistrano´ B these would be the awkward English translations of these Spanish place-names.

Russian has left a mark on North America as well. In Alaska the place-names were left in the original B Sitka, Alaska, was settled by Russians. But the capital of the State of Idaho is Moscow, not Moskva or Muscovy; i.e., the original Russian city´s name shows up in its anglicized version.

It is interesting to compare this paradigm to modern German, where one gets Hund, Hund, Hundes, Hunde in the singular paralleling the OE exactly; the plural has been levelled out as die Hunde. OE hund got respelled under French influence as hound, since [ou] stood for /u/, whereas [u] would have stood for /ü/. The word, together with most monosyllables containing stressed /u/ has diphthongized into [au] B thus today we have hound, /hawnd/, ´a kind of dog´ (´greyhound´, ´bloodhound´). The phenomenon is commonly known as semantic narrowing. The paradigm itself shrank and almost disappeared. The stress having settled down on the first syllable of English words, the second syllable vowels became a murmured Aschwa@ sound, and eventually dropped. The genitive singular survives as hound´s as in the hound´s leash broke; the nominative and the accusative plural, which already coincided in OE, survive as hounds, as in the police used the bloodhounds to track the criminal, (accusative usage), and the hounds are howling (nominative usage). All the other cases have disappeared. This, in turn, has caused English to change its typological character from an inflectional into an isolating language. Whereas in OE one could still distinguish the nominative and the accusative singular by the definite article se and pone, (se hund waes god ´the dog was good´ vs. we wyllad cweljan pone hund ´we wanted to kill the dog´), this is no longer the case in modern English. Thus the hound/dog chased the man and the man chased the hound/dog have identical forms for ´dog[nom]´and ´dog[acc]´, and it is strictly by the word order that o ne can tell which occurrence of hound/dog is nominative (subject) and which one is accusative (direct object).

Once upon a time the word queen simply meant ´woman´as it still does in Swedish kvinna. (The word is, of course, a well attested cognate of gyne- as in gynecology, the women´s name Gwen and Gwendolyn.) In the Middle Ages the word underwent ´semantic depression or pejoration´ and came to mean ´tavern wench´ B a girl who sells beer, or may even be a prostitute. Centuries later, queen, spelled with a capital letter as Queen, means ´ruling monarch´or ´ruling monarch´s wife.´ It has, therefore, obviously undergone ´semantic amelioration´or ´upward shift´.

AEtymologies are fossil poetry,@ as already Ralph Waldo Emerson observed in The Poet. What does this mean? All languages are full of idioms. Consider the saying fly off the handle ´become angry´ and kick the bucket ´die´. The first of these comes from falconry B the medieval habit of hunting with a falcon. When the falcon, sitting on T-shaped piece of wood, sees the rabbit, it becomes excited and literally ´flies off the handle´, i.e., becomes airborne. Hence when someone becomes overly excited or angry, he is comparable to a hunting falcon that just saw a rabbit.

In a functional eco-system B and natural languages are just such eco-systems B everything has a role. In Africa the tiniest insect counts as much as a large elephant. In an area covered by a language, phonetic change may seem like a small thing, yet it is of considerable importance B it may reveal the actual origin of a word. The same goes for etymologies.

It is common knowledge that whereas North America is geologically the oldest continent with the Anewest humanity@ inhabiting its soil, Europe, on the other hand, is laid out the other way around: It is the youngest continent geologically, with the Aoldest humanity@ inhabiting it. One would think that the older the humanity, the better known its history would be but, unfortunately, it doesn´t quite work that way. Whereas the history of Canada and the USA is reasonably well documented, in Europe, and especially in the Southeastern part of the Continent, history can be extremely complex and shrouded in the mystery of unrecorded migration patterns. In such cases the history of the language B replete with phonemic change, etymologies and toponyms B is almost the only tool that scholarship can use in an attempt at reconstructing the lost portions of history. This is one of the major contributions of the present volume: It fills a number of historic gaps through objective linguistic means.

Traiana existed between 106 and 275 AD. The extent to which this province became Romanized is not clear, and there is no historical mention of any Roman population there after the province was abandoned by the Roman Empire. It was about 800 years later, in the second half of the 11th century that Vlachs (Rumanians) were reported north of the lower Danube. Towards the mid-thirteenth century, Vlach political organizations (small principalities) appeared in Muntenia. The first known written Rumanian text is a letter written in 1521 to the judge of Brassó (Kronstadt, BraÕov), Hans Benkner.

Dacia Traiana was part B less than 40% B of the territory in which Northern Rumanian (ADaco-Roman@) is spoken today; it was one of the most distant provinces of the Roman Empire over which Roman domination lasted at most 169 years. In spite of this, the hypothesis that Latin continued to be spoken in Dacia Traiana after 275 AD and that it developed into present day Rumanian there was presented long ago: In the mid-fifteenth century, Italian humanists travelling in eastern Europe discovered a people whose language contained many Latin words also existing in Italian. Knowing that the Roman Empire once dominated Dacia Traiana, these humanists assumed that the Vlachs were the descendants of the Romans, who once subdued the Dacians.

Particularly during the last four decades, Rumanian archaeologists and historians have made great efforts to find remains in present day Rumania of a Roman population between the end of the 3rd and the end of the 11th centuries. Reports of such a population were based on material remains of Roman style. However, some Rumanian scholars, (A. Philippide, I. Iordan, I.I. Russu) expressed the opinion that the material finds from earlier periods are not sufficiently specific for such conclusions. This is because of the powerful influence of Roman culture and civilization on the material culture of all European peoples outside the Roman Empire. Thus, there is no historical or archaeological evidence of the ancestors of the Rumanians in the territory of present day Rumania before the 11th century. One may claim that this is not decisive, the ADaco-Romans@ were humble people, not important for the chroniclers. However, the question is inevitable: if the Rumanians are not the descendants of Traian´s Romans, where did they come from?

Large territories in the Balkan peninsula were Roman provinces for six centuries and were strongly Romanized. Records and finds such as ruins of churches show that Christianity was propagated in that part of the Roman Empire as early as the 3rd century. We have a considerable amount of information about the Church, whose language was Latin. There are numerous descriptions of incursions of Goths, Huns, Avars, Slavs, and other Abarbarian@ peoples in the Balkan provinces. However, records end in the 6th century; by the mid-seventh century, almost the entire peninsula was taken over by the Slavs. Roman and Byzantine historians no longer had access to these territories, and there are no records about the destinies of the Roman population in these times. It is probable that most of the remaining Romans were in a short time assimilated to the Slavs. From the 8th century, there is a mention of Vlachs (a Romance population) living in the valley of the Rhinos, towards the end of the 10th century, beginning with 976 AD, Vlachs are repeatedly recorded in different places of the Balkan peninsula.

While the surrounding populations designated them by the equivalents of the name Vlach, they always called themselves rumîn, Arumanian ar(u)mân, arr|m|nu, armân, which is an inherited Latin word (from Latin romanus). In English, it corresponds to Rumanian, in French, to roumain, in German, to rumänisch, and in Serbo-Croatian, to rymyn. The etymological form, rumîn, is still used in our days; it appears in the texts of Coresi (16th century), as well as in the Palia de la Or|Õtie (1581B1582), in which also the form român appears for the first time. This is not a popular but a learned form, constructed with the purpose of making it more like Latin romanus. Since the mid-nineteenth century, the Rumanian literary language uses exclusively this form.

This means that it is not possible to detect anything about the early history of the Rumanians using the historical method : analyses and comparison of different texts, B not even by the study of material remains revealed by archaeology. The historian who sticks to these methods can only state: ignoramus et ignorabimus. As a reviewer of E. Illyés´ monograph argued:

However, modern historical science uses complementary disciplines: physics, chemistry, archaeology, numismatics, sociology, etc. In the case of early Rumanian history, the most important B in fact, decisive B complementary field of research is linguistics. Language is unconsciously transmitted historical evidence as is human remains, written business or military records, customs and institutions, etc. The Rumanian language contains elements of great significance regarding its early history. The present book displays and summarizes this vast corpus compiled to a large extent from Rumanian sources. These are facts, which must be explained in some way or other. Not all historians may be willing to do this, or to deal with these problems at all, allegedly because of the lack of written sources. For linguists, of course, it is easier to notice the historical significance of the facts of language (cf., for example, Herbert Izzo, 1984; in Rumania, cf. the works of O.Densusianu, A. Philippide and, more recently, I.I. Russu).

The problem is: how to interpret linguistic facts? I. Coteanu, in his introduction to Istoria limbii române, II, 1969, pp. 16B17, has stated clearly that linguistic research alone is unable to solve problems beyond the scope of linguistics and that it should only occupy itself with its own field, i.e., language. To be able to reach conclusions about historical problems, the scholar must use other methods than linguistics:

Many examples will be found in this book of the necessity of using known, well-established historical circumstances as a framework in the interpretation of facts revealed by linguistics. In this way, these facts receive the status of historical evidence, giving indications about the Asocio-historical frontiers [...] within which the Latin language [which forms the basis of Rumanian] was spoken@. Thus, for example, the pre-Latin substratum of the Rumanian language is to a significant part identical with Albanian: about one hundred lexical elements, of which more than half pertain to the life of shepherds living in high mountains, and there also are correspondences regarding the Latin elements of these languages. This suggests that the ancestors of the Rumanians spoke the same language as those of the Albanians. One must infer that the ancient areas of the Rumanians were the same (or close to) those of the Albanians. If it would not be known where the latter were living before the Roman occupation of the Balkans, these facts would only have relevance for linguistic research. Fortunately, the ancient areas of the early Albanians are quite well known.B Similarly, the presence in Rumanian of the innovations of Late Latin indicates that the ancestors of the Rumanians lived, during the Late Latin period, in close contact with other areas where Latin was spoken (particularly with Italy). However, in order to determine the territory in which they were living, the study of the historical circumstances of that period in southeastern Europe, the shifting frontiers of the Empire, etc. is also necessary. B The significance of the very powerful South Slavic influence upon the Rumanian language can only be appreciated if one takes into account historical data from a large part of southeastern Europe after the fall of the Roman Empire; namely, the Slavic colonization of large areas of the Balkans starting in the 6th century, the emergence of differences between Bulgarian and Serbo-Croatian towards the end of the first millennium A.D., Bulgarian dialects being also spoken in Macedonia; the Christianization of the Bulgarians in the second half of the 9th century, etc.

This historical problem became at an early stage the subject of political consideration and struggle. The Supplex libellus Valachorum, written in 1791 by Rumanian intellectuals living in Transylvania to king Leopold II in Vienna, was a petition for more rights. To strengthen their demands, the authors affirmed that the Vlachs were the most ancient population in the country, since they originated from the Roman veterans who were brought to Dacia by Emperor Trajan in the 2nd century AD. The affirmation of autochthoneity was here a means in a political struggle. The theory was developed in detail by Petru Maior, one of the members of the ATransylvanian School@ (Ôcoala ardelean|).

While the writings of the Transylvanian School may scarcely be regarded as scientific studies, later, extremely valuable work was done by many eminent Rumanian linguists, such as Alexandru Philippide (1859B1933), Ovid Densusianu (1873B1938), and Sextil PuÕcariu (1877B1948), to mention only the most brilliant ones among them. Their extensive study of all dialects of the Rumanian language has thrown light upon most of the problems of the early history of this language, showing its intimate relations with several Balkan languages. Mostly on the basis of these interrelationships, Philippide concluded that the Rumanian language developed, at least until the 7th century AD in the Balkan peninsula, south of the Danube. This view was also expressed by several scholars outside Rumania, of whom we here only mention R.Rösler (Romanische Studien. Untersuchungen zur älteren Geschichte Rumäniens, Leipzig, 1871) and L. Tamás (ARomains, Romans et Roumains dans l´histoire de Dacie Trajane@, Études sur l´Europe Centre-Orientale I, Budapest, 1936).

However, in spite of the above-mentioned results of exhaustive linguistic research, the theory that the Rumanian language developed (mainly) from Latin spoken in Dacia Traiana (the theory of continuity) was the official thesis of the Rumanian Communist Party. Thus, this theory has been adopted in all publications from Rumania during the last four decades; or example, in the following works: Daicoviciu, C. (red.), Istoria Romîniei [The History of Rumania], Bucharest, 1960; Daicoviciu, C., Petrovici, E., & Ôtefan, G., Istoria României. Compendiu, Bucharest; 1969 (first edition), 1974 (2nd edition, red. Ô. Pascu); Condurachi, E. & Daicoviciu, C., The Ancient Civilization of Romania, London, 1971. Constantin Daicoviciu (1898B1973) has conducted extensive research into the period of the Dacian states, the Roman colonization north of the lower Danube, as well as the Migration Period in south-eastern Europe. His most important articles were published in one volume in 1969 (Dacica, Cluj). Dumitru Protase (born in 1926), published in one volume the more important archaeological material found in the territory of the former Roman province of Dacia Traiana known until the mid 1960s, trying to find arguments for Daco-Roman continuity (Problema continuit|Ûii în Dacia în lumina arheologiei Õi numismaticii [The Problem of Continuity in Dacia in the Light of Archaeology and Numismatics], Bucharest, 1966). Also the articles published in Dacoromania I, Jahrbuch für östilche Latinität, edited by P. Miron, Freiburg-München, 1973, argue for the theory of continuity. (These are, however, only some more important examples, because the literature pertaining to the problem of continuity, especially of archaeological material found in Rumania published during the last decades is enormous.)

The sensitivities of Rumanian historians, intelelctuals, B and also of the general populace B do not permit an objective discussion of the theory. A recent example will throw some light upon this: a Hungarian archaeologist said recently in an interview that the excavations started in central Cluj (Hungarian Kolozsvár, German Klausenburg) in the summer of 1994 will not yield any evidence of Daco-Roman continuity. It has been stated long ago by Rumanian archaeologists that on the site of Roman Napoca, there are no signs of life after the end of the 4th century AD until the appearance of the earliest Hungarian remains, from the 10th century. However, in a major Rumanian newspaper, an indignant article retorted with the question: Ahow far should our tolerance go?@ B [the Hungarians are now] Aquestioning the historical rights of Rumanians in Rumania.@

This is not an isolated opinion of a single journalist, but a typical example of the intellectual atmosphere today in Rumania. The Hungarian archaeologist did not question the theory of Daco-Roman continuity, he only said that there, in Cluj, there are no signs of it. If this statement evokes such a fierce response, one can imagine what the reaction would be if the archaeologist would declare that there are no signs of continuity at all! The situation is thus similar to that before 1989, when continuity was the official thesis of the Communist party and the government. Rumanian critics of The Early History of the Rumanian Language referred mainly to details or to material not taken into consideration, but without going into a real, comprehensive debate. Their conclusion was nonetheless that the text was biased and tendentious. The hope expressed by Blair A. Rudes, that ADu Nay´s work will, by bringing the ´theory of discontinuity´ back into discussion@ stimulate research, was not fullfilled. Glanville Price stated: [Du Nay demonstrates] Athe difficulties raised by the theory of continuity whose supporters will now have to present an equally well documented and argued answer.@ To the best of my knowledge, such an answer has not appeared in the 18 years that have passed since the publication of that book.

It is absurd to claim that anyone denying or questioning this thesis, such as R. Rösler, G. Weigand, A. Philippide, would do so because of political reasons. Also senseless is the accusation that Hungarians in Transylvania, if they are unwilling to accept this concept, Aquestion the historical rights of the Rumanians in Rumania@. All three main nationalities now living in Transylvania have been there for many centuries, B a sufficiently long time to establish the historical right to live there also in the future.

When preparing this book, I have conducted research in an objective way, without prejudice. The problem is primarily scientific and has nothing to do with considerations outside science: it is about the development of the Rumanian language from Latin. I mention the political side of the the question reluctantly, and only because I consider that the reader should be informed about a reality which impedes a free, unprejudiced discussion of the problem in the country which it concerns most B Rumania. Even if national consciousness is often based upon some measure of myth, the enormous gap between the theory of autochthoneity in Rumania and the result of a thorough analysis is much too large. It amounts to a serious deformation of the history of the Rumanians, who should have the right to be offered as objective an account as possible about their past.

THE ORIGINS OF THE RUMANIANS