The original homeland of the Slavs
Witold Manczak
In the late seventies, the author proposed a new method of linguistic and ethnogenetic research, namely, the method of comparing the vocabulary in parallel texts. Using this method, he examined relationships between different languages (Gothic, Old Church Slavic, and Lithuanian; German, Polish and Lithuanian; Polish, Old Prussian, and Lithuanian, etc.) and finally arrived at the conclusion that the original homeland of the Slavs was in the basin of the Oder and the Vistula rivers.
In the late seventies, I proposed a new method of linguistic and ethnogenetic research, namely, the method of comparing the vocabulary in parallel texts. Since some authors, e.g. Birnbaum or Salmons, have identified my method with that used by glot-tochronologists, I have to draw attention to the fact that while there is a certain similarity between the two methods (both are lexicostatistical), there are, however, numerous and essential differences between them:
(a)	glottochronologists distinguish between cultural and non-cultural words, whereas I do not;
(b)	glottochronologists investigate the vocabulary on word lists (prepared by them), whereas I compare the vocabulary in parallel texts (which are authentic material);
(c)	the number of words examined by glottochronologists is limited (e.g., 100 words), whereas the number of words taken into consideration by me is virtually unlimited;
(d)	glottochronologists claim that the rate of change of non-cultural words in different languages is stable, whereas I believe there are great differences between the rates of change of the vocabulary in different languages;
(e)	the only purpose of glottochronological research is to determine the date for the split of languages, whereas my method allows us to solve different problems.
There is an essential difference between the method of comparing the vocabulary in parallel texts and that of comparing the vocabulary in dictionaries (which has been practiced since time immemorial). Here are two examples.
The Oxford Advanced Learner's Dictionary of Current English begins with the following 17 words, which occur on the first page: a, aback, abacus, abaft, abandon, abase, abash, abate, abattoir, abbé, abbess, abbey, abbot, abbreviate, abdicate, abdomen, abdominal. Among these words, 3 are Germanic and 14 (printed in italics) of foreign origin.
Consequently, by counting words in dictionaries, we would arrive at the conclusion that English is not a Germanic language, but we will change our opinion if counting words in texts. In the preface to the above mentioned dictionary, there are the two following sentences: "In this revised impression, the representation of pronunciation differs somewhat
from that shown previously. The phonetic notation now conforms to that to be found in the majority of important English dictionaries used by non-native learners of English, and in particular to the latest (14th) edition of the English Pronouncing Dictionary." In this short text, the proportion of native and foreign words is reversed: 32 words are Germanic and 18 (printed in italics) of foreign origin. It follows that only counting words in texts allows us to justify the Germanic character of English.
Another example. Popovic (1960: 199) claims that "in Bezug auf den Wortschatz ist das Rumänische keine romanische, sondern eine slavische Sprache, da z. B. im Dakoru-mänischen ... in einem Verzeichnis von 5765 Wörtern, auf nur 1165 Wörter lateinischen Ursprungs die imposante Anzahl von 2361 Wörtern kommt ... also 2/5 des Wortschatzes slavisch sind". Popovic alludes to the etymological dictionary of Rumanian published by Cihac in the seventies of the nineteenth century, that contains 5765 words, among which 2361 are of Slavic origin, 1165 of Latin origin, 965 of Turkish origin, etc. We will, however, arrive at another conclusion if we count words in a text. Here is a fragment from a novel by Sadoveanu, where words of non-Latin origin are printed in italics:
In munte, la altitudinea aceea unde ne duceam sä gäsim cucofii sälbatici, seva primäverii incä nu pornise, defi ne aflam la inceputul lunii Mai. Subt brazi erau incä pete de omät §i, subt bältile mlaftinilor, gheturi. Tufäri^urile arbustilor päreau moarte in umbra cetinilor, soarele nou incä nu le fäcusä sä scoatä ace de muguri.
In this text, more than 80% of words are of Latin origin. Consequently, in order to arrive at the right conclusion that Rumanian is a Romance language, words have to be counted in texts.
Linguistic Kinship
I applied the method of comparing the vocabulary in parallel texts to different questions, which, as will be shown later, are connected with that of the original homeland of the Slavs.
As far as I know, the German orientalist Ludolf, who lived in the seventeenth century, was one of the first to consider the nature of linguistic kinship. He arrived at the conclusion that "die Sprachwissenschaft offenbart sich nicht im Wörterbuch, sondern in der Grammatik". During the last 300 years, so many authorities approved of this method that it has become a dogma of linguistics. However, I confronted it with facts and showed that the dogma is false. Here are the arguments supporting this view.
First of all, it has to be stated that, for determining the degree of linguistic kinship phonetic criteria are useless. Not only for Slavicists, but even for laymen who have only a rough idea of Polish, Ukrainian and Russian, it is obvious that Polish is closer to Ukrainian than to Russian, but from the point of view of phonetics, there are many more similarities between Polish and Russian than between Polish and Ukrainian: (1) in Ukrainian g > h, while in Polish and in Russian g remains unaltered; (2) in Ukrainian e > i, whereas in Russian in all cases and in Polish in most cases e > e; (3) in Ukrainian i > y, whereas in Polish and in Russian i does not change; (4) prothetic consonants appear in Ukrainian more frequently than in Polish and Russian, cf. Ukrainian vin, but Polish and Russian on; (5) final voiced consonants preserved their original character in Ukrainian, but in Polish
and in Russian they turn voiceless; (6) in Ukrainian, if e or o is in a closed syllable, it becomes i, whereas in Polish and in Russian these vowels do not change; (7) in Polish and in Russian, consonants are palatalized before front vowels, which is not the case in Ukrainian; (8) at the end of a syllable, the consonant v undergoes a vocalization in Ukrainian, while in Polish and in Russian it remains unaltered; (9) in Ukrainian, a consonant before the cluster of b and j undergoes a duplication, which occurs neither in Russian nor in Polish; (10) in Ukrainian, the affricate which is explained by the second and third Slavic palatalizations preserves its palatal character, which is not the case with Polish and Russian. Altogether, I found 10 phonetic similarities between Polish and Russian and only 2 between Polish and Ukrainian. But it is sufficient to apply my method of comparing the vocabulary in parallel texts in order to come to another conclusion. In a fragment of a Russian novel translated into other Slavic languages, I found 13 lexical convergences between Polish and Ukrainian and only 1 between Polish and Russian.
For determining the degree of linguistic kinship, inflectional criteria are also useless. All scholars agree that Gothic is closer to English than to Old Church Slavic, because Gothic and English are Germanic languages, whereas Old Church Slavic is not. However, comparing a fragment of the Gospel (Mark VIII) in Gothic, English and Old Church Slavic, I found that, on the whole, there were 83 inflectional similarities between Gothic and Old Church Slavic and only 31 between Gothic and English. It is very easy to explain this result. Gothic and Old Church Slavic are synthetic languages, while English has an analytic character. E.g., the declension is rather rich in Gothic and especially in Old Church Slavic, whereas there are only some remnants of the declension in English. The same goes for the conjugation. In English one says I came, you came, he came, we came, you came, they came, whereas, in translation, Gothic and Old Church Slavic have six different forms each. As it was said, there were 83 inflectional convergences between Gothic and Old Church Slavic and only 31 between Gothic and English. But it is sufficient to apply my method of comparing the vocabulary in parallel texts in order to come to another conclusion. In the same fragment of the Gospel, I found that in total there were 93 lexical similarities between Gothic and English and 74 between Gothic and Old Church Slavic. That is to say that the unanimous opinion according to which the degree of linguistic kinship depends on grammatical structure is false. In reality, not phonetic or inflectional features but words allow us to determine the degree of linguistic kinship provided that words are counted not in dictionaries but in parallel texts.
Another example illustrates this view. All scholars agree that Polish is closer to Bulgarian than to Lithuanian because Polish and Bulgarian are Slavic languages, whereas Lithuanian is not. However, I have compared a fragment of the Gospel (John III-IV) in Polish, Bulgarian and Lithuanian, and have found that there were 62 inflectional similarities between Polish and Lithuanian and only 52 between Polish and Bulgarian. It is easy to explain why: Polish and Lithuanian are synthetic languages, whereas Bulgarian has an analytic character. E.g., there are seven declension cases in both Polish and Lithuanian, whereas only some declension forms remain in Bulgarian. There is an infinitive formed with the same suffix in Lithuanian and in Polish, but the infinitive disappeared in Bulgarian. As it was mentioned, there were 62 inflectional convergences between Polish and Lithuanian and only 52 between Polish and Bulgarian. But it is sufficient to examine the vocabulary in the same fragment of the Gospel in order to come to a different conclusion. There were 291 lexical resemblances between Polish and
Bulgarian and only 51 between Polish and Lithuanian. In other words, the true opinion that Polish is closer to Bulgarian than to Lithuanian is based not on morphology but on vocabulary.
Still another example can be mentioned here. Latin is closer to French than to Gothic. Nevertheless, I have compared a fragment of the Gospel (Luke VII) in Latin, French and Gothic, and have found that there were 103 inflectional similarities between Latin and Gothic, but only 18 between Latin and French. The explanation of this fact is very simple: Latin and Gothic are synthetic languages, while French has an analytic character. As far as declension is concerned, there are six cases in Latin and four in Gothic, but the French nouns do not show any case forms at all. In Latin one says canto, cantas, cantat, cantamus, cantatis, cantant. The Gothic conjugation is similar, but in French one says je chante, tu chantes, il chante, nous chantons, vous chantez, ils chantent. In other words, in the Latin and the Gothic paradigms, there are six endings, but there are only two in the French paradigm. As it was said, there were 103 inflectional convergences between Latin and Gothic and only 18 between Latin and French. But it is sufficient to examine the vocabulary in the same fragment of the Gospel in order to come to a different conclusion. There were 222 lexical resemblances between Latin and French and only 47 between Latin and Gothic. In other words, the true opinion that French is a Romance language is based not on morphology but on vocabulary.
It is worth mentioning that my opinion on the nature of linguistic kinship was criticized by Salmons (1995), who wrote that
lexical data should be included in the overall picture of genetic relationships, in conjunction with phonological, morphological, syntactic, archeological, and (pre-)historical data... In an undertaking frought with so many difficulties as determining genetic relationships, it would seem prudent, if not absolutely obligatory, to draw on all available and reasonably reliable information. Even where such data are or might be available to support his argument, Manczak does not admit them, using one reductionist method in the place of all the diverse tools historical linguistics has evolved over the last centuries.
Unfortunately, it is impossible to follow his advice. Comparing texts in Gothic, English and Old Church Slavic, I have obtained the following results:
Inflectional convergences Lexical convergences Gothic and English	31	93
Gothic and Old Church Slavic	83	74
It follows from this that the true opinion that Gothic is closer to English than to Old Church Slavic is based only on vocabulary. Had we followed Salmons' advice and considered both inflectional and lexical similarities, we would obtain the following data:
Inflectional and lexical convergences Gothic and English	124
Gothic and Old Church Slavic	157
It would follow from this that Gothic is closer to Old Church Slavic than to English, which surely does not correspond to reality. If an Englishman knowing neither Gothic nor Old Church Slavic looks at a Gothic text and at an Old Church Slavic one, he will understand none of them, but he will guess the meaning of more words in the first one than in the latter one.
Ethnogenetic Research
Personally, for many years, I have drawn attention to the fact that ethnogenetic research should be made more scientific by applying the two following procedures:
(a)	One should distinguish arguments which are verifiable by statistics and those which are not. E.g., if somebody claims that Polish resembles Old Prussian more than Lithuanian and somebody else holds a different opinion on this subject, it is possible to count convergences between these languages and determine which of these two opinions is correct. But if a scholar thinks that Tacitus' Veneti in the first century A. D. are the same as Jordanes' in the sixth century, while another scholar denies that, it is impossible to resolve this question by using statistics.
(b)	One should rely only on arguments which can be confirmed by statistics.
By using statistics I succeeded in establishing the following facts:
(1)	Johannes Schmidt, well known for his Wellentheorie, has already realized that there is a connection between kinship of languages and their distribution on earth. Comparing the vocabulary in parallel texts has confirmed Schmidt's opinion that neighbouring languages are closer to each other than non-neighbouring ones. For example, Polish is closer to Czech than to Slovenian.
(2)	Linguistic kinship shows a startling stability. It is obvious that, for geographical reasons, Latin used in Dacia about 270 (when Roman troops left Dacia) resembled Latin spoken in Italy more than Latin used in Gaul or Spain. Since the evacuation of Roman legions from Dacia, 1700 years went by and during this time there was no linguistic contact between Dacia and Italy. Nevertheless, to-day's Rumanian resembles Italian more than any other Romance language.
Let me give another example for the surprising stability of linguistic kinship. I once compared parallel texts in Gothic, Old Church Slavic, Lithuanian and Latin, and found out that, from a lexical point of view, Gothic, first of all, resembled Old Church Slavic, then Lithuanian and finally Latin. Then I compared the same fragment of the Gospel in German, Polish, Lithuanian and Italian, and this comparison revealed that German, first of all, resembles Polish, then Lithuanian and finally Italian. These two comparisons show that relations between Germanic, Slavic, Baltic and Romance languages did not change during the last 1500 years.
(3)	Gothic resembles Old Church Slavic more than seventeenth-century Lithuanian, which proves that the Slavs, in the prehistorical period (as in the historical period), lived between Germanic and Baltic tribes.
(4)	German resembles Polish more than Lithuanian, which also indicates that the Slavs must have initially lived between Germanic and Baltic tribes.
(5)	Polish resembles Old Prussian more than Lithuanian, which means that the Slavs must have initially lived closer to Old Prussians than to Lithuanians.
(6)	Polish resembles German more than Ossetian, which is used in an area nearer to Poland than any other Iranian language is. The distance between Hamburg and Kiev is
more or less equal to that between Kiev and Vladikavkaz, capital of the Ossetian republic, which is situated on the northern slopes of the Caucasus. It follows from this that the original homeland of the Slavs could not be in the Dnieper region because, had it been there, the number of lexical resemblances between Polish and Ossetian should have equalled that of the lexical similarities between Polish and German.
(7)	The Italian vocabulary resembles the Polish vocabulary more than the Lithuanian one, which indicates that the Slavs, in the prehistorical period (as in the historical period), lived closer to Rome than the Balts.
(8)	The Germanic languages resemble the Slavic languages more than the Romance ones. If we bear in mind that the distance from Hamburg to Rome amounts to 1300 km, from Hamburg to Poznan 500 km and from Hamburg to Kiev 1500 km, the fact that there are more lexical convergences between Germanic and Slavic than between Germanic and Romance is also an argument for a western homeland of the Slavs.
(9)	Irish resembles Polish more than Lithuanian, which means that the Slavs, in the prehistorical period (as in the historical period), lived between Celtic and Baltic tribes.
All these facts prove that the original homeland of the Slavs was in the basin of the Oder and the Vistula (for statistical data upon which this opinion is based, see Manczak 1992).
Godlowski's Conception
Until the late seventies, most Polish scholars were convinced that the original homeland of the Slavs was in the basin of the Oder and the Vistula rivers. But in 1979, the archaeologist Godlowski arrived at the conclusion that the Slavs originally lived in the area of the upper and part of the middle Dnieper river. In a short time, his conception gained many adherents. Among them, there were also linguists (Popowska-Taborska 1997, Rze-telska-Feleszko 1998 or Slawski 2000), which is strange because Godlowski's thesis is contradicted by certain linguistic facts.
If it were true that the original homeland of the Slavs was in the area of the upper and middle Dnieper, statistical data which I succeeded in establishing should be quite different. E.g., the Germanic languages could not resemble the Slavic languages more than the Baltic ones, but, conversely, they should resemble the Baltic languages more than the Slavic ones. Polish could not resemble Old Prussian more than Lithuanian, but, conversely, it should show more lexical convergences with Lithuanian than with Old Prussian. Irish would not resemble Polish more than Lithuanian, but, conversely, it should agree more frequently with Lithuanian than with Polish.
According to Godlowski, the Slavs appeared only at the end of the fifth century in the area of the Oder and Vistula rivers, which, until then, was inhabited by Germanic tribes. In this connection, one has to mention that, in Proto-Germanic, there was a phenomenon called the Germanic sound shift, which, among others, consisted of the fact that the phoneme d remained in Slavic unaltered, whereas in Proto-Germanic it changed into t (cf. Polish dwa, woda, but English two, water). In the Slavic languages, the phoneme t was preserved, whereas in Proto-Germanic it changed into th (cf. Polish trzy, ten, but Engl. three, the). In the Slavic languages, the phoneme p still exists, but in Proto-Germanic it was shifted to f (cf. Pol. pigc, pierwszy, but Engl. five, first). The Germanic sound shift explains also why Engl. h corresponds to Pol. k, cf. Pol. kto, but Engl. who, Pol. kamien,
but Engl. hammer (the Germanic word for 'hammer' arose in the stone age, when hammers were produced from stone). And so on. In addition, in Proto-Germanic, there was a phenomenon explained by Verner's Rule, that is to say that, under certain circumstances, the fricatives became voiced. E.g., the equivalent of the Polish word bosy is in English the word bare, where the phoneme r arose from an earlier *z. Consequently, if it were true that, in the area of the Vistula and Oder rivers, the Slavs learnt the river names from a Germanic population, the Polish river names should sound differently from the way they do, they should show traces of the Germanic sound shift as well as traces of Verner's Rule, but such traces do not exist at all. E.g., in Upper Silesia, there is a river called Drama, and in northern Poland, there are rivers called Drawa and Drwgca. If the Slavs had learnt these hydronyms from a Germanic population, these river names should have begun with t, and not with d (according to the rule which explains why Engl. tree and to correspond to Pol. drzewo and do).
Some of the adherents to Godlowski's conception claim that the Slavs settled in the basin of the Vistula and Oder rivers half a century after this area was deserted by Germanic tribes. In other words, the Slavs are supposed to have settled in an uninhabited area. This opinion is improbable, too. The Polish river names fall into two categories: (1) river names which are understandable to a Pole, hydronyms like Kamienna, Bystrzyca or Prqdnik, which are of relatively recent origin; (2) river names which are not understandable to a Pole, hydronyms like Wisia, Odra, Raba, Sola, Nysa, Nida, Bug, etc., which arose in remote times, hundreds or even thousands of years before the fifth century A. D. It is important to stress that in the fifth century there were no geographic atlases, not to speak of the fact that the Slavs who lived then were illiterate. Therefore, we may wonder how the Slavs, who are supposed to have appeared in an uninhabited country, could learn the river names used before their arrival.
In order to be able to espouse Godlowski's conception according to which the Slavs appeared only in the fifth century in the Vistula and Oder region, which was inhabited by a Germanic population, one would have to admit that, among those illiterate Slavs, there was a genius who, 1400 years before the nineteenth-century scholars, discovered the Germanic sound shift as well as formulated Verner's Rule and, moreover, succeeded in convincing his fellow countrymen that they expurgate all traces of Germanic influence from river names they had learnt from the Germanic population. Those who believe that the Slavs settled in the Vistula and Oder region half a century after this area was left by Germanic tribes would have to admit that, among the Slavs settling in an uninhabited country, there was a genial clairvoyant who guessed the names of many Polish rivers which arose before the fifth century and, moreover, succeeded in forcing his fellow countrymen to use just these hydronyms. Is this possible? Unfortunately, none of the numerous adherents to Godlowski's conception wanted to answer this fundamental question. Personally, I think that this is impossible.
Godlowski, who recently died, often mentioned an argument brought forward by Moszynski. According to Moszynski (1957), the original homeland of the Slavs must have been in the Dnieper region because trees growing in more eastern zones have Slavic names, whereas the names of trees growing in more western zones are of foreign origin. It is the question of the following trees: buk 'beach', cis 'yew', jawor 'maple', modrzew 'larch' and swierk 'spruce'. In order to show the weakness of this argument, it is enough to compare the Latin names of these trees with the French ones:
beach	fagus	hêtre (Germanic)
yew	taxus	if (Celtic)
maple	acer	platane (Greek)
larch	larix	mélèze (Celtic or Pre-Indo-European)
spruce	picea	sapin (Celtic)
So the French people		give beach, yew, maple, larch and spruce names of foreign
origin although the Romans, who, 2000 years ago, conquered Gaul, knew these trees. Consequently, even admitting with Moszynski that all these names (buk, cis, jawor, modrzew, swierk) are of non-Slavic origin, does not allow us to establish on such a basis that these trees did not grow in the original homeland of the Slavs.
The same goes not only for tree names, but also for other loan words. E.g., the Polish name of dance (taniec) is of German origin, but this does not mean that the ancestors of to-day's Poles learnt to dance only from the Germans. In reality, they knew how to dance earlier, but they used another verb for 'dancing, namely, plqsac.
In Latin, peace is called pax and war, bellum. The word pax exists until now in all Romance languages, whereas bellum survived in no Romance language or dialect. The Rumanians call war razboi, a word of Slavic origin, whereas all other Romance languages use a word of Germanic origin like French guerre or Italian guerra. But this does not mean that the Romans were a peace-loving nation. On the contrary, they were one of the most belligerent peoples of antiquity.
In English, the names of domestic animals are Germanic, while those of different kinds of meat are of French origin: ox and beef, calf and veal, pig and pork, sheep and mutton (also deer and venison). But this does not mean at all that the English people at one time raised cattle, but did not eat their meat because they were vegetarians.
All these examples show that the fact that the names of some trees growing in more western zones are of foreign origin is no proof that the original homeland of the Slavs was in the Dnieper region.
Before ending my paper, I would like to say a few words about the issue of prehistoric migrations of the Slavs. Among all the languages of the world, Romance languages occupy a unique and privileged position in that we perfectly know the protolanguage from which the Romance languages emerged, namely Latin, while all other protolanguages, e.g. Proto-Slavic, Proto-Baltic, Proto-Indo-European, etc., are only linguistic reconstructions. As a consequence, the latter have a more or less hypothetical character. The same goes for ethnogenesis. As far as the original homeland of the Slavs or of the Indo-Europeans is concerned, there are innumerable hypotheses, whereas it is quite sure that Latin was originally used in Latium, that is to say in the region of Rome.
Preparing a book on the classification of Romance languages, I tried to resolve this problem with my method of comparing the vocabulary in parallel texts. It turned out that, in most cases, there is a correlation between the chronology of Roman conquests and the number of lexical convergences which particular Romance languages show in relation to other Romance languages:
Language	Beginning of the conquest
Italian	7498	Italy 396 B. C.
Portuguese	7159
Spanish	7114	Spain 226 B. C.
Catalan	6985	
French	6851 *	Gaul (125) 58 B.C.
Provençal	6560 1	i
Romanche	6318 1	Rhaetia 15 B. C.
Sardinian	5333	Sardinia 237 B. C.
Rumanian	3564 J	' Dacia A. D. 101
The earlier a province was conquered by the Romans, the more lexical convergences in relation to other Romance languages exist in a language or languages used in a given province. The earlier a province was conquered by the Romans, the more archaic is the vocabulary of a language or languages spoken in the given province. Italian shows the most archaic vocabulary and Rumanian the most innovative one because Italy was the first province conquered by the Romans, while Dacia was the last one (Manczak 1991).
The results I obtained are noteworthy because if we knew the Romance languages only in their twenty-first-century form, we could reconstruct, with a high degree of probability, the relative chronology of Roman conquests, which took place between the beginning of the fourth century B. C. and the beginning of the second century A. D. Therefore, I decided to compare fragments of the Gospel in all modern Slavic languages in order to find out how many lexical convergences exist between each Slavic language and all others. Here are the results of this comparison:
Polish	9228	Russian	8328	Slovene	8434
Czech	8728	Bielorussian	8251	Serbo-Croatian	8126
Slovak	8695	Ukrainian	7710	Bulgarian	7186
Upper Lusatian	7802				
Lower Lusatian	7514				
As far as lexical convergences with other Slavic languages are concerned, Polish occupies the first position, which is a further proof that the original homeland of the Slavs was in the Oder and Vistula region.
Moreover, in order to obtain commensurable data, I considered three languages from every group of Slavic languages:
Western	Eastern	Southern
Polish	7607	Russian	7131	Slovene	6856
Czech	6756	Bielorussian	7040	Serbo-Croatian	6917
Slovak	6734	Ukrainian	6489	Bulgarian	6150
21097	20660	19923
These numbers show that the Western languages are the most archaic, the Eastern ones are less archaic and the Southern ones are the least archaic. In other words, the migration moved first to East and then to South.
Finally, I would like to stress that an essential advantage of my method of comparing the vocabulary in parallel texts is that the results which I obtained are always based on statistical data and, therefore, can be verified.
References:
Godlowski, K.
1979 Z badan nad zagadnieniem rozprzestrzenienia Siowian w V-VII w.n.e. Krakow: In-
stytut Archeologii UJ. Manczak, W.
1991	La classification des langues romanes. Krakow: Universitas.
1992	De la préhistoire des peuples indo-européens. Krakow: Uniwersytet Jagiellonski. 2004 Przedhistoryczne migracje Siowian i pochodzenie jçzyka staro-cerkiewno-siowianski-
ego. Krakow: PAU. Moszynski, K.
1957 Pierwotny zasiqg jçzyka prasiowianskiego. Wroclaw: Ossolineum. Popovic, I.
1960 Geschichte der serbokroatischen Sprache. Wiesbaden. Popowska-Taborska, H.
1997	'The Slavs in the early middle ages from the viewpoint of contemporary linguistics.' In: Origins of Central Europe. Wroclaw: Scientific Society of Polish Archaeologists, pp. 91-96.
Rzetelska-Feleszko, E.
1998	'Nazwy miejscowe.' In: Polskie nazwy wiasne. Encyklopedia. Warszawa: Wydawni-ctwo Instytutu Jçzyka Polskiego PAN, pp. 191-230.
Salmons, J. C.
1987 'Another word on lexical data and genetic relatedness.' Journal of Indo-European Studies 15: 381-384. Schmidt, J.
1872 Die Verwandtschaftsverhältnisse der indogermanischen Sprachen. Weimar: Böh-
lau. Slawski, F.
2000 'Do artykulu Witolda Manczaka.' Jçzyk Polski 80: 332-333.
Witold Manczak Praojczyzna Slowian
Witold Manczak
Z koñcem lat siedemdziesi^tych autor zaproponowal now^ metod^ badañ j^zyko-znawczych i etnogenetycznych, a mianowicie metod^ badania slownictwa w paralelnych tekstach. Stosowanie tej metody doprowadzilo go przede wszystkim do odmiennego spoj-rzenia na istot^ pokrewieñstwa j^zykowego. XVII-wieczny orientalista niemiecki Ludolf twierdzil, že o pokrewieñstwie j^zykowym decyduje nie slownictwo, ale gramatyka, i pogled ten do dzis jest powszechnie aprobowany. Tymczasem obliczenia wykonane przez autora wykazaly, že faktycznie o pokrewieñstwie j^zykowym decyduje zgodnosci nie fo-netyczne i fleksyjne, ale leksykalne. Na tej podstawie autor przedsi^wzi^l zakrojone na duž^ skal^ badania zgodnosci slownikowych mi^dzy rožnymi j^zykami, które to badania przede wszystkim wykazaly, že zgodnosci leksykalne wykazuj^ zdumiewaj^c^ stabilnosc. Gdy okolo r. 270 legiony rzymskie opuszczaly Dacj^, jest rzecz^ oczywist^, že ze wzgl^dów geograficznych lacina užywana w Dacji nawi^zywala bardziej do laciny mówionej w Italii niž do laciny užywanej w Galii czy Hiszpanii. Przez 1700 lat mi^dzy Daj a Itali^ kontak-tów j^zykowych nie bylo, a pomimo to dzisiejszy j^zyk rumuñski jest bardziej podobny do wloskiego niž do francuskiego czy hiszpañskiego. Bior^c to pod uwag^, autor przebadal zgodnosci leksykalne mi^dzy róžnymi j^zykami (mi^dzy gockim a staro-cerkiewno-slo-wiañskim i litewskim, mi^dzy niemieckim a polskim i litewskim, mi^dzy polskim a staro-pruskim i litewskim, mi^dzy niemieckim a polskim i osetyñskim, mi^dzy wloskim a polskim i litewskim, mi^dzy j^zykami germañskimi a j^zykami slowiañskimi i romañskimi, mi^dzy irlandzkim a polskim i litewskim) i ostatecznie doszedl do wniosku, že praojczyzna Slowian ležala w dorzeczu Odry i Wisly. W koñcowej cz^sci artykulu autor polemizuje z pogl^dem archeologa Godlowskiego na praojczyzn^ Slowian.