Are they published or not?


B. Lukács


President of Matter Evolution Subcommittee of the HAS


H-1525 Bp. 114. Pf. 49., Budapest, Hungary



In 2003 Bérczi & Detre published a linguistic relict (Ref. 3 of this study) from 500-700 AD, which, for first sight, looks like a substantial amount of proto-Magyar words. Here I show that the language is not proto-Magyar, in the sense that after 1400 years of evolution the language would not have ended in recent Magyar. (By any chance the language is now extinct.) I think no more definite statement can be done now, due to the fragmentary nature of the publication. The material seems, however, an important linguistic relict, so I discuss some possible connections to Uralic & Altaic languages. As a possibility, I show up the mysterious σαβαρτοι ασφαλοι of Constantine Porphyrogenetus (Ref. 54), in genetic connections with Magyars but “near to Persia”.



What are Magyar & Hungarian? Magyar is a language and an ethnic entity (facts & beliefs will come soon), Hungary is a state. If you meet a language called "Hungarian", it is very probably Magyar. Magyar language is surely not Indo-European, and has some kinship with Finnish & Estonian; the exact degree of this kinship is a matter of argumentation. That degree is just depending on some Armenian texts on which this study has been written, so wait to the end. However an almost unequivocal academic consensus does exist and that will be recapitiulated early in the main text.

            Language. I wrote an English text, mainly for technical reasons. Some Magyar vowels turn out badly in some Netscape Navigators, and I do not like to type in Magyar under Windows. Also, the matter is not an internal Hungarian one: as you will see, Armenian language & culture is heavily involved, plus Uralic linguistics too.

            Transliterations. Languages involved here have at least 3 totally different alphabets (Latin, Cyrillic & Armenian), and orthographies widely differ even within one alphabet. So here I transliterate every word using in discussions of linguistic evolution to Magyar. Of course in many cases the Magyar transliteration is possible only in some approximation; let it be, however. Now, Magyar orthography is quite surprising, but very phonetic and logical.

            For consonants, most single letters are read more or less as in German; the only bigger difference is "s", which stands for German "sch", English "sh". As for digraphs, "softening" or "palatalisation" is indicated by a "y" (as in English "Kenya"). Only 4 consonants can get a "y": g, l, n & t, and in today's Magyar "gy" is rather "dy", but that change happened a mere 800 years ago. For other digraphs, I give approximate English equivalents: "sz" stands for "s", "cs" for "ch" and "zs" is some "zh". The only trigraph, "dzs" stands for "j", but it is rare.

            For vowels, you get an approximate picture via German. Vowels a, e, i, o & u are not far from German, Italian or Turkish; for "e" the letter really stands for 2 different sounds, one is more or less German, Slovakian or Finnish "e", the other is "ä". Another 2 letters, ö & ü, correspond to the same German Umlaut (and Turkish) sounds. However a special Magyar (& Slovakian; so Hungarian) trick is to indicate the length via primes. There is a minor difference other than length too between "a" and "á", but I will not discuss that; it is a matter of geographic variation too.

            Present Magyar "i" seems to be the inheritor of two ancient vowels, both of the front and the back unrounded high vowels. The two i's you can now found in many Turkish languages. If needs be, I will transcribe the back i as "ď".

            Long-range Correlations.At some points you will find a sign Ąx, where x is a number, and later a text (Go back to Ąx.). Of course you do not have to go back. But there are some long-range correlations in the text. Indeed this "Go back..." is almost the sign for a footnote; but the footnote is some previous part of the text. The footnote is the text just before the respective Ą.

            Ob Ugor peoples. Here I use the “international” (in fact, Russian) names of Ob Ugors: Vogul & Ostyak. But really they are (with Magyar orthography) Manysi & Khanti (or Hanti).



            Heribert Illig has become quite unpopular among Central European historians while not so in some folklorist &c. circles. I definitely decline to discuss this matter. However now a Hungarian site exists [1] which discusses his ideas about the "invented, non-historic" 297 years between 614 & 911 AD, and this site contains a "dictionary" of "Hunnish" words [2].

            This "dictionary" may be very important in Magyar linguistics; we soon shall see, why. However something is rather unexplained about the text.

            The text seems to originate from the Hungarian geologist Cs. H. Detre. However it tells that it is so far unpublished. However a booklet identical textually for some 95 % with [2] exists [3]. While on the book Year 2004 is indicated, the booklet had been in distribution on 24 Oct., 2003 already, when I got a copy.

            So I do not fully understand the circumstances under which [2] was created. However [2] & [3] are almost identical (except for some graphics in [3]). So I will use [3] here.

            [3] (or [2]) is extremely important in Magyar linguistics, especially in the study of the evolution of Magyar language (and I note that Sz. Bérczi, one of the two authors of Ref. [3]) was the Secretary of the Matter Evolution Subcommittee of HAS led by myself for 12 years). The reason of importance will be given in the next Chapter in a proper way. Here I only note that some circumstances are rather anomalous: e.g. the booklet is formally published as a painting book (!); and simultaneously [2] does not seem to know about [3].

            Since I am not in contact by the authors for years, although I wrote to one of them in the autumn of 2003, without any answer), I cannot discuss the problems with them, and that is not needed anyways. The material is published both as a booklet [3] and as an Internet material [2], and now everybody can use that material.

            One author of [3] is a physicist at a University, the other is a geologist, so both work in Science Proper. So I am not permitted to question their scientific honour and have to regard the published material genuine. I definitely would like to know more about the circumstances and details of the discovery, but perhaps, according to physicist jargon “they deserve a subsequent paper”. Just now it seems I have not much chance to get the extra information.

            But I can write something even based on the booklet. And indeed, the material is important for Magyar linguistics. Maybe professional linguists should discuss it first; but I waited 10 months already.



            Magyar is the biggest language in the Carpatian Basin, and spoken outside of it by numerous emigrants of last century (plus some 100,000 autochtonous csángó-ceangau in Roumanian Moldavia), altogether 14-15 million speakers. The language is not Indo-European; it may or may not be Nostratic.

            The most widely accepted opinion about its classification, definitely accepted by the Academies of Sciences of all Uralic nations having Academies at all, is that Magyar is genetically a Uralic language, and especially

            1) Uralic forked into proto-Samoyedic & proto-Finno-Ugric;

            2) then Proto-Finno-Ugric forked into proto-Finnic & proto-Ugric;

            3) then proto-Finnic on the West evolved into Finnish Proper, Lapponian, Estonian, Liv, and a multiplicity of other languages not spoken outside of Russia, including some extinct ones as well, while

            4) proto-Ugric on the East evolved into Ob Ugric in Westernmost Siberia and Magyar in the Carpathian Basin, plus maybe some unknown extinct languages too.

            This scheme is supported by many good etymologies which I will not repeat here, plus some (not fully unequivocal) data of archaeology. As for Magyar, however, it is not supported by physical anthropology & culture. In these points Magyar speakers are nearer to Altaic groups (mostly Turks). Also an extensive layer of early Turkish loanwords can easily be identified in Magyar, mainly r-Turk, or Bulgarian Turk (Chuvash) ones. These "Turkish" words are completely absent in Finnish Proper; a few words can be found in the Ob Ugric languages, but only a very few. Some Southern Finnish languages (e.g. Mari/Cheremys) also contain Turkish words, but less than Magyar does.

As for scholarly literature, it is big, but mostly written in Uralic languages. Here I give references in an indirect way. You can go hence to 2 Internet sites of me [4], [5], and there I give etymologies plus references to some textbooks.

            Of course, you cannot get scientific certainty about the past of Magyar language. In principle you can collect unlimited information about related languages in the present, but you cannot observe Past. You may accept old written records (or you may not; e.g. old legal documents were frequently falsified even after centuries), but even then millennia remain unobserved. But you may invent kinship patterns, can try to force all your data into such a picture and within it you will get the most parsimonious story. So finally most parsimonious stories of different paradigms will compete; and maybe you will be able to choose the most convincing one.

            The best story of the Uralic paradigm starts with a purely Uralic scenario up to the end of pre- (or proto-)Magyar times, and then applies heavy Turkish influence from Southeast in the middle of the first millenium AD. Finns were already too far to West and Ob Ugors too Northward. Then Magyars entered the Migration, and on horseback they traversed the space between the Ural River and the Carpathian Basin, while they adopted Turkish way of life and the words denoting its specific notions, plus a lot of fashion words too. Also, living in a confederation of tribes, some tribes of Turkish origin may have been incorporated.

            However a diametrically opposite Turkish paradigm exists too, in which Magyars are not Turkized Ugors, but Ugrized Turks. Supporters of such a view are in utter minority amongst linguists, but not so small minority amongst anthropologists, and the idea is fairly acceptable for Turks.

            Note that it is not easy to settle the question for ever. Namely, languages can be learnt; and words can be borrowed even more easily. Grammar is more conservative, until it breaks down in Pidginisation, as Anglo-Saxon & Norman French simultaneously lost almost all of their synthetic structures about 1100 AD in fusing into Middle English. Now, Magyar is moderately polysynthetic even now, so it has not undergone Pidginisation. However Uralic and Altaic grammars are rather similar for first principles.

            I am not arguing for or against Turkish origin of the Magyar language; I generally use the Uralic theory and would bet for it, if for any. I only tell that the opposite theory is not utterly nonsense. The next Chapter gives some chronologies, heavily depending on historical theories.



            While educated guesses do exist on the time data of the Uralic forkings, they are unimportant for us.

            Tacitus about 100 AD writes about some "Fenni" more or less at present Finland; however Finns rather believe that they were still moving to West, and Tacitus' Fenni are proto-Lapponians. Let them be.

            However proto-Magyars must have been involved into at least one of horserider migration. When they enter the Carpathian Basin (official date is April 896 AD, but really any date between 889 & 899 is possible), they are horserider people, rather similar to r-Turkish Bulgarians & Onogurs and z-Turkish Petchenegs, although Hungarian archaeology can distinguish all of them. However there were 3 big migration waves in the Migration Period on the Eurasian grasslands:

            A) Huns starting from the East cross the River Volga (this is the Uralic name: Volga ~ világ(os) = bright = Etil/Itil [Turkish], also used in Middle Ages) under Balambér in cca. 370. Ą1 They occupy the Carpathian Basin under Buda (German Bleda, but two consonants in starting position are almost unheard of in Altaic & Uralic), almost reach the Atlantic Ocean under Attila, but lose the Nedao Battle against a federation led by Ardaric the Gepid under Ellac in 454 and in some years they retreat until the Northern shores of Black Sea. Henceforth their history is complicated & multiple: a part enters the Bulgarian confederation (the first ruler seems to be Irnac, son of Attila), others enter z-Turk confederations, and Armenian sources speak about smaller Hunnish states at the plains at the Northern feet of the Caucasus. We shall have to return to this point later. This chronology is supported by many Late Roman sources.

            B) The Bulgarian Migration starts in cca. 463. Priscus Rhetor (earlier an envoy from Constantinople to Attila) is a contemporary source. As for the preceeding steps he repeats a tale about the migration of some semi-mythical people adjacent to the Icy Ocean, but 463 is the time when the migration reaches the Caucasus & Ural (River).

            C) The Avar Migration. Avars are mentioned as Zhuan-zhuans by the Chinese sources, but this seems to be a simple derogatory name (cca. "susurration", moving always to and fro, for a Chinese). In 551 they are defeated by the Turks in the Mtn. Altai, then in 17 years they traverse the whole grasslands to the Carpathian Basins, which they occupy on Easter Monday, 568.

            Hungarian archaeologists can detect two later waves. An immigration to the Basin is rather probable about 630, and the Codex Fredegarius gives some story about it; and about 680 certainly a new population arrived. This last wave coincides the foundation of Danube Bulgaria in 681 (Asparukh), and also with a long ago unexplained entry of the Vienna Illustrated Chronicle [6] written in cca. 1374 at the Hungarian Royal Chancellery (677 as the Magyar Conquest, too early by 2 centuries!). However now I neglect the late waves; you will see why.

            In the XIXth century in Hungary Wave A) was popular: so proto-Magyars were involved in the Hun migration until the Basin itself, then they retreated and again entered the Basin in 896. However lots of arguments were collected against proto-Magyars in the Basin in the Vth century (mainly linguistic ones). Proto-Magyars remaining at the Eastern side of the Hun Empire would not contradict too much anything, but see the next paragraph.

            However, (proto)-Magyars surely lived in something near to a symbiosis with Bulgarians (or Onogurs), see the linguistic traces. Everything gets an easy explanation if Magyars became involved in the Bulgarian Migration in 463. However they remained in Easternmost Europe, perhaps still in 750 just at the Western shore of River Don. Consider that in the earliest Hungarian chronicles the Maeotis (Azov Sea) is a place of origin; in the adjacent marshland Magyars meet and take "the daughters of Kings Dul & Belar", which is clearly the Bulgarian people and the Bulgarian royal dynasty, the Dulo: a mythic formulation of a mixed population of Ugors & Bulgar Turks.

            Then Wave C) is "too late". Of course, Magyars, having entered Easternmost Europe with Bulgars could participate later in Avar motions too. But proto-Magyar - proto-Slavic contact at 600 would be too early for the linguistic evolution (I will return to this point in due course).

            So in the Uralic paradigm of Magyar origin Wave B) yields the most parsimonious scenario. The southernmost Ugric population reached the neighbourhood of Ural River before 463. (Independently of any paradigm, theory & such the two Ob Ugor languages, Vogul & Ostyak, seem the nearest extant relatives of Magyar. So the Ugor group is established even without theories.) When the Bulgarian Migration started, they became involved, maybe accidentally, while the northern ancestors of Ob Ugors did not. Under heavy Bulgarian influence proto-Magyars accultured to steppe nomadism, borrowed a lot of r-Turkish termini technici, and maybe mixed anthropologically.

            Some details still have to be elaborated. Surely Ugors have their unique horse terminology (horse = ló, law, the Finnish word hevonen is clearly not a relative, but the Altaic words are also apart (although Sinor [7] tries with ulagh = post horse); also stirrup = kengyel < kengy-al, beneath  the "kengy", a commion Ugric word for a kind of boot), but Ob Ugors do not show any trace of horserider nomadism, although they do know horses.

            Now if a population without any nomadic tradition suddenly becomes involved in a migration of horse nomads, by any chance they will be employed as mere living targets; they cannot be used even as herding boys. So either proto-Magyars learnt with surprising rapidity, or their chiefs were extraordinarily good diplomats. But maybe the details will be worked out later.

            The alternative scenario is far less elaborated but it exists; it was more or less accepted by, e.g., archaeologist Gy. László. It is based on the fact that the present ratio of Ob/Danube Ugors is o(10-3), and it was in the same order of magnitude 400 years ago. So maybe Ob and Danube Ugors are not sister groups, but present Ob Ugors were originally a peripheric population having partly learnt proto-Magyar. Then, of course, Ob Ugors would give only a snapshot about proto-Magyar language, some 1500 or 2000 years ago, and no more evolutionary data could be got from their present languages. But then Magyar would be the only original and extant member of the whole Eastern half of the Uralic family (Magyar-Finnish correspondences are too heavy to simply take Magyar out of the Uralic group, see: hand=kéz~keszi, water=víz~veszi &c.).



            In the Uralic family/group the first extant written texts are rather late. In Finnish the date is cca. 1540 AD. On the wider Finnish side the first written texts come from St. Stephen of Perm, cca. 1370, a missionary bishop, who wrote some Biblical texts in Permian. The Ugric branch is even more diverse. For Ob Ugors the first texts practically come from a Hungarian linguist, A. Reguly, from cca. 1840. (Previously only isolated words were published by linguists.) On the other hand, the donation letter of the Tihany Abbey from 1055 contains a lot of isolated words and half a sentence in Magyar, and in another 150 years there is a substantial burial speech purely in Magyar.

            This Magyar is an ancient stage of the linguistic evolution (see later), but even this early stage was very similar for the present Magyar for the trial origin of words: Ugric words, r-Turkish words and Slavic words. From the early texts as well as from the XIXth century statistics we see the same picture. Words for elementary kinship & nature are more Ugric, less r-Turkish and lest Slavic. Agriculture words are maybe fifty-fifty r-Turkish and Slavic; pastoral terms are more r-Turkish, less Ugric and lest Slavic. Finally, Christian and state terminology is more Slavic and less r-Turkish. The Slavic is most probably proto-Slovakian and/or -Slovenian.

            This is easily understood in the Uralic paradigm. Proto-Magyars were first similar to Ob Ugors: mainly hunters, fishers & gatherers, but with some horses too. Then they accultured to Bulgarian r-Turks, so the older layer of agriculture words (wine, beer, wheat, barley &c.) are purely r-Turkish, and also lots of termini of statemanship & religion were too. But in the Carpathian Basin they learnt Christianism, early feudal society, and some more evolved agriculture.

            But I must tell something about the opposite scenario too, even if I do not recommend it. There "Ugric" simply means the layer which is not Turkish, neither Slavic. Such words do exist in great abundance. Numerals are neither Turkish nor Slavic. Also most parts of human body are such (but not all, i.e. knee=térd and neck=nyak are r-Turkish). Also many words for natural phenomena are such, i.e. Sun=Nap & Moon=Hold. This alternative viewpoint always has some difficulty about the Finnish connections of many "archaic" words, but let me continue.

            According to the traditional teaching the earliest Magyar text is the 86 isolates in Tihany donation document from 1055. This may be challenged for two points. The first is Cs. Detre's one in [2] and [3], and that will come later. The second is the observation that there is an earlier donation document (of the nuns of Veszprém Valley, maybe from 997) in Greek, also with isolates. OK; Graeca non leguntur.

            Now some examples must be seen; but in the next Chapter.



            Look at the text from 1055 [8], the Royal Chart of the Tihany Abbey. The longest Magyar isolate is:

feheruuaru rea meneh hodu utu rea

The meaning is clear (the term must define one border of the estate): in modern language & orthography it is

Fehérvárra menô hadútra

Of course, the Latin cannot show vowel length. The meaning is:

To the military way going to White Castle

the military way was built by the Romans and the White Castle is the present (Székes)fehérvár. Hence and from lots of shorter isolates we learn that in 1055

1) There are neither definite nor indefinite articles (the, a).

2) Lots of nominal endings are still separate words (as e.g. in present Japanese).

3) Lots of words ending in 2004 with consonants still ended in 1055 in a high vowel. Feheru>Fehér; uaru>vár; hodu>had; utu>út.

            To Point 2): nobody is surprised. In present Magyar there are 25-30 nominal endings and more postpositions (nobody counted them). Obviously in the last 950 years some postpositions agglutinated to the noun. To Point 3): less words end in 1055 in vowels than in consonants. Linguists always reconstruct vowel endings for the ancient stages: so then the end-vowels are just being lost in 1055. The 997 document also shows the vowel endings, a 1138 foundation document still shows a few, and there is some argumentation about the Regestrum Varadiensis in 1216 [9]; but never later.

            Now let us combine the two observations. In the modern language

way, street = út < utu

to the way = útra < utu rea

way, Acc. = utat < utu + -t (?!)

This shows that the "original" Nominative (meaning that much earlier than in 1055) was "uta" or "uto", not "utu". In that time the “-t” ending of the Accusative was already agglutinated, but the "-ra" ending of the Allative was still a separate postposition. When, after 1055 sometimes, the "-ra" fused, there was already no vowel at the end of "út" (the lengthening at the front compensated the loss at the end), so there is no vowel between the "t" and "r", but there is one between the two "t"'s. Lots of such examples do exist. Of course, the evolution before 1055 is theory (not having texts from that time), but modest enough a theory.

            So far I spoke independently of paradigms. But there are problems with the Accusative in any paradigm. Nowhere amongst Finnish, Ugric or Turkish languages can we find a “-t” Accusative suffix; and we cannot even guess the time of agglutination. We cannot know everything.

            Now let us see the burial speech from 1192 [10]. It is already almost modern. But some endings still does not show vowel harmony. Here some explanation is needed for non-Altaic and non-Uralic readers.

            It is not absolutely certain from an ancient text if two groups of sounds formed two words or one compound. But noun with ending can clearly be distinguished from noun + postposition in languages with vowel harmony. Vowel harmony is most evolved and regular in Turkish languages, from which you may take the Turkish of Turkey (the "Ottoman Turkish") as a good example. It is also quite strict in Mongolian. As for Uralic languages it is laxer, with exceptions (which have good explanations) in Magyar, while it is quite regular but somewhat simplified in Finnish. It is almost lost (?) in Estonian, while it is practically nonexistent in the two Ob-Ugric relatives of Magyar. However let us not dwell on theories about the evolution of Vowel Harmony.

            Let us much simplify the picture and tell that Vovel Harmony means the following. There are front vowels and back vowels. Now 1) in one word stem all vowels must be of the same type: either all front or all back. And 2) endings must have two forms: front and back. Front stems take front endings, back ones back ones.

            If this is too abstract, let us see examples from Modern Magyar. The Allative ending is -ra/-re. Then:


















your health




cat’s rhythmical sound




beating the door




Table 1: Magyar illustrations for Vowel Harmony


At the same time see the postposition


With it, the forms are: út felett and felett. So we can clearly distinguish par excellence endings obeying Vowel Harmony and par excellence postpositions not obeying it (being separate words). The cleavage at any snapshot is not so clear: just now in 2004 we have two endings, -ért=for and -képpen=(cca.) as it goes, which do not yet have back forms. OK; after some time they will. We can detect when scribes started to write them together with the stems. The number of nominal endings is growing in Magyar in the last thousand years.

            The situation is similar for verbs. The suffix for Pres. Ind. Pl3 Indet. is -nak/-nek, so they dorombolnak, but dörömbölnek. We cannot even guess how old is the fusion. But there is an ending forming a verbal form which is approximately the English can/may + Verb. This suffix is -hat-/-het-. Ą2 I write = (Én) írok. I can write = (Én) írhatok. I look = (Én) nézek. I can look = (Én) nézhetek. (The pronoun is in bracket, because it is redundant. You may use it, but that betrays that you are a foreigner. Or: there is some emphasis on it. I write; you do not.) Now, there is (I would like to write: there does be, but that is ungrammatical) the quite regular and common verb hat. (Én) hatok = I act, or, I have an influence. Hatás = influence. Hatalom = power. Even: hathatok = I can have an influence. We do have educated guesses when the verb hat started to act as a suffix -hat-/-het-.

            Presently the Inessive -ban/ben (in, but in the "within" sense; the other "in" would be rather an Adessive) has two forms and of course stems automatically choose the proper ending. But in 1192 it is still milostben, paradisumben, not malasztban, paradicsomban, as in 2004. This is OK; the original form of the separate postposition contained "e", as today the form "benne"; it took some time after agglutination to develop the alternative "-ban". So in 1055 the later Inessive was still a separate postposition; in 1192 it was (seems to have been) fused but still was not a par excellence suffix, not yet even in 1350; but in 1450 the texts show fusion and Vowel Harmony, so the evolution of the Inessive suffix was ready.

            Finally let us observe a more complicated suffix in 1192. Ą3 We found 2 nouns: hotolm (> hatalom in 2004), scerelmes > (szerelmes in 2004). Hatalom = power, szerelem = love, both are formed from verbs: hat = acts, szeret = loves. The ending is then -alom/-elem, already with fusion and Vowel Harmony in 1192. (And hence you can see how to distinguish the two "e"'s in Magyar: in -elem the first is the pair of "a", in the second of "o", so the first "e" is lower than the second. It is indeed so.) So -alom/-elem forms a noun from a verb. Let us try to form more: köt = bind, kötelem = a specific legal status to bind. Sír = cries, siralom = a sad thing/situation. This ending is so uniquely Magyar that even Ob Ugors cannot make it. Or is it uniquely Magyar? Look: Magyar szerelem = Komi muszalöm. The Perm Finn languages show some surprising similarities to Magyar as if they were as close as Ob Ugors. Still no explanation. (And you can see that Vowel Harmony in Komi is much weaker than in Magyar.)



            Sometimes in the 60's T. Cseres, (later for years the president of the Union of Writers; not the Pen Club, but a kind of a Trade Union) got an idea, I do not know, whence. Armenians collect any manuscripts of national affinity. Armenia was the first Christian country, therefore old religious manuscripts just might contain something in connection with proto-Magyars roaming not too far. Ą4

            Now, this was a rather desperate idea. In any Uralic scenario the main body of proto-Magyars were never nearer to Armenia than 800 km. Except one splinter group: around 700 one group detached itself and went through the Caucasus. We know almost nothing about them; but in 1329 they still existed because then Pope John XXII wrote a suggestion to Missionary Bishop Bernhardus to try to convert them [11]. They were quite near to Armenia.

            But T. Cseres got the usual Hungarian reactions. The plan was refused as "mirageous". This is the usual lazy Hungarian answer if somebody gets an idea: it is easier not to do anything and for reason you can tell that some hallucination misdirected the guy as a mirage in a desert. He seems to see something nice but it is not really there.

            In that time I read the suggestion in a literary weekly, then the argumentation, and finally the negative conclusion. OK, I told: the chances would not be too good, anyways. Of course in the sixties I was in no position to make any research; I could not read Armenian texts, and I would not have been admitted into Matenadaran anyways. (Cs. Detre is older than me and knows Armenian.)

            Remember this suggestion, however.



            Of course, the Matenadaran has Hunnish texts. Some groups called Huns roamed in the plainer regions of the foots of the Caucasus between at least 505 and 682. Pawstos Buzand reports that Huns killed Grigoris, grandson of Gregory, the Illuminator, sometimes in the IVth century about 340 [12]. (Then, of course, Huns were just going to West: they crossed the Volga (Rha, Itil) in 370 under Chief Balambér (this is the Magyar form) and in 375 the Don [13]. Then, after the Nedao battle various groups went to East, and Saragurs reached Daghestan at least in 505. And in the VIIth century Mowses Kalankatuaci (?) still reports Huns in the neighbourhood [12], [14]. (I omit such diacritics which are not too hopeful on Internet.)

            Are they really Huns, more than 2 centuries after the Nedao battle? Who knows. But Kalankatuaci tells that they burn horses for a giant monster named Tangari Xan. Now, Tängri Khagan, or in other style the Kök Tängri is the Eternal Blue Sky, the chief deity of all traditional Turks and Mongols, and of course religious horseriders, including old Magyars, offered horses to the chief god. Even Ob Ugors know that in great matters the gods would prefer horses. So Kalankatuaci's Huns are some horserider people indeed.

            Now, cca. in 682 Israel, Bishop of Mec Kolmank, "converts the Huns". Maybe not all of them; but at least from this time some Armenian priests are among the "Huns", so Hun texts may be in the Matenadaran, only we cannot predict, what was the language of these Huns. We know hardly anything even about the language(s) of Attila's Huns.

            And now it is proper to start to cite Ref. [3]. Immediately the subtitle on p. 2 states that "The Hun Little Dictionary and the Grammar Part was made according to the Grabar (Old Armenian) translated by Ödön Schütz & Csaba Detre (1978)." And near the end of the booklet one can find a reference "Edmund Sjuc: Hin hajeren - chuneren jerkulezú matjanner Iráni medzs - Matenadaran, Jereván, 1962". Fortunately a Magyar translation is also given (in this form even Armenians would be in trouble: an Armenian text with Magyar orthography cannot be understood for anybody except a Hungarian Armenian/Armenologist) "Ödön Schütz: Very old Armenian-Hunnish bilingual manuscripts in Iran, Matenadaran, Yerevan, 1962".

            I did not see this publication and shall not be able to read it in the close future (anyways, it is in Armenian), so I remain at [3]. [3] tells that two manuscripts contain Hunnish texts, the Isphahan Codex and the Codex Cretensis, the first was written cca. 500, the second in cca. 700, and now both are in the Abbey of Szurb Khács (Sacred Crux), in Isphahan, Iran. I do not know how they were dated; but if the dating is correct, then the Codex Isphahanicus belongs to the arriving Saragurs (or to the Sabirs; cf. Chap. 15), and the Codex Cretensis to the Huns of Bishop Israel.



            Detre or Schütz (or both?) applied an unexpected method when interpreting the texts. The Magyar orthography is very phonetic (of course, for the Magyar phonemes), and Armenian is too, for the Armenian ones. The Armenian alphabet contains 38 simple letters (one, for the open "o" sound, was not yet in use in 700), + some ligatures. Magyar, including di- and trigraphs, has an alphabet of 44 signs (of which, to be sure, 4 do not represent Magyar phonemes). Although the two phoneme sets do not coincide (there are more consonants in Armenian and more vowels in Magyar), it was obviously better to transliterate the "Hunnish" words into Magyar than, say, into non-phonetic English or into Latin of 25 letters. (Henceforth I omit the quotation marks of "Hunnish". OK, they were maybe Saragurs or Belendjers; it seems that Hun was not a language but a tribal federation, but Saragurs had belonged to the Empire of Attila the Hun, and the educated guess of historians about the nomads of Northern Daghestan between 500 & 700 is generally "some Altaic".

            Note that in XVth century the Armenian alphabet was regularly applied to an Altaic language: a Kipchak Turk tribe converted to Armenian Christianism in Poland and of course took Armenian letters [15]. But that transliteration must not be used in this case. First there is the 800 years, then the Armenian spoken in Poland is not Grabar but Western Armenian, and, finally, Kipchaks had a rich vowel structure and poor consonantal one in their own language, while in the present case the scribes were Armenians.

            The exact details of the Schütz-Detre transcription seem to be unpublished, and of course, if I cannot see the originals, much can depend on the transcription. In Ref. [3] only a few sentences deal with the transcription. The most important statement is the list of the transliterated letters (the Armenian originals are not given). The "Magyar" alphabet of the Hun texts is then

a, á*, b, c, cz%, cs, csh%, d, dz, dzs, e, é*, f, g, gh%, gy%, h, ch%, i, j, k, kh%, l, ll$, m, n, o, ö@, p, ph%, r, rr$, s, sz, ssz$, t, tz%, th%, u@, ü@, v, w*, z, zs

Now I must take the responsibility to guess and explain something not published by the authors. If any in the following few paragraphs is contrary to the unpublished details, Dr. Detre can correct it, e.g. at [2].

            * denotes single letters of the Magyar alphabet which, however, pose nontrivial problems in the transcription. The prime in á and é denotes length. However Armenian does not denote length, so it is nontrivial, what is behind the two primed letters. I have no real idea, what is the duality a/á in the original Grabar texts (maybe a geminated a for á, or a ligature), but at least Western Armenian speakers in Hungary must know it, because when transliterating to Magyar, they use both a and á. And indeed there are 2 Grabar e's, one low, one middle. Now, in the present Budapest pronunciation of Magyar é is always one degree higher than e, so this may be the clue. As for w, this letter does exist in the Magyar alphabet but did not denote an independent phoneme, but in Armenian there is a letter denoting a (bilabial half?) consonant whose scientific description is w. It is also used in the ligature (ow) for u.

            $ denotes geminated consonants in the transliteration. Hungarian orthographical tradition do not regard geminated consonants long: an "ll" or "kk" denotes two consonants. (Slovakian has long versions of r and l, but this length is denoted by a prime just as for vowels, and primed "ŕ" is not equivalent with "rr". In Magyar, e.g., ssz stands for a geminated sz, and behaves differently in hyphenation, if the first sz is in the first syllable, the second in the second, or if both are in the same. Now, there may or may not be geminated Grabar letters behind ll, rr and ssz, but note that there are two different "r" sounds in Armenian, so that difference may be responsible for r/rr. As for the other two "geminates" we must wait until the publication of details.

            % denotes di- and trigraphs not used in the canonical Magyar alphabet (although some were used earlier). The solutions seem to easy, let us go from group to group.

            Surely there are the aspirated stops of Gabar behind kh, ph & th. Also, there is an aspirated version of cs (English ch) behind csh. However, beware; I think, ch is not the aspirated version of c (English ts), but that aspirated sound must be either cz or tz (see next paragraph). I am not sure about the difference between cz and tz, but I note that the Dictionary does not in fact employ dz. So maybe tz is the Grabar letter for dz. Indeed the Western Armenian pronunciation would be c in the present Hungarian orthography, but in some older texts rather tz (and in English ts).

            The digraph ch must be the similarly written Scottish or German sound. In the scientific transcriptions of Grabar it is either x or χ; observe the place of ch in the Schütz-Detre transliteration alphabet just after h (exactly the same place as in Slovakian). The digraph gh is surely the γ of Grabar. As for gy it seems to be an alternative reading of the combinations di-, gi- or gj-; or maybe of dzs-.

            @ denotes simple vowel letters in Magyar, two of which are ligatures in Armenian. The letter u is surely the Grabar combination OW; present Armenian writes "u" in this way. Note that in the international transcription of Armenian proper names "u" is often corresponds not to this combination but to no Armenian letter at all. Namely, in groups of consonants Armenian speakers often insert automatically a short, back, high unrounded sound, roughly similar to Turkish "ď". Western Europeans then seem to hear an "u", because for them a back, high, unrounded vowel is impossible, so their brain substitutes it with the rounded counterpart. As for ü, Ref. [3] explicitly notes the problem and tells that it can be written in Armenian by UJ or by Greek Y. Now, while the front high rounded vowel does not exist in Armenian, it does exist in Turkish place names with substantial Armenian communities up to World War I, as e.g. Kütahia. In such cases a trigraphic combination of o, w & i (j) is used, and it would be rather surprising to read the trigraph anything else. As for ö, that vowel does not exist in Armenian either, but I checked the recognition of my ö sound by some native East Armenian speakers. Now, some of them accepted my ö as their swa sound, which does have a letter.

            I cannot tell more at the present state of publication of the Detre-Schütz studies. But the general principles seems clear enough.



            With the transcription fixed, it is straightforward to compose the dictionary. Albeit I did not see the raw material, the Codices must have been composed in Armenian, with some references to the (Daghestan) Huns, say: for X the Huns say Y. Ref. 3 gives some 300-400 words, and then tries to find out the elements of Hun grammar.

            The result is rather unexpected. Some half of the words is rather closely related to Magyar. (I will make more definite statements later.) Ref. [3] claims to have recognised a layer from an utterly different language; but that is a hypothesis, and there are hypotheses enough here. In addition the extra language is represented mainly by astronomical terms.

            Let us see a demonstration of "closeness" to Magyar. I take very elementary words for comparison.




Magyar, 2004

Magyar, XI-XII c.



Turkish (Ott.)































































Table 2: Some Hun words on Uralo-Altaic background


            I kept “ä” for “open e” in Finnish & Vogul, but remember that Magyar orthography does not distinguish “ä” and “e”. According to Grabar, I guess “e” behind Hunnish “é”, and “ä” behind Hunnish “e”; but I cannot be sure. Vogul “v” is bilabial, but old Magyar “v” was either. Maybe in some rows I should have chosen another alternative; but if a relative of the Magyar word does exist, I took that. Some Old Magyar words are really after XIIth century; then I chose the oldest available. Also, Old Magyar orthography generally did not indicate length. Even then, you can see what is the kinship at least for that half of the Hunnish words which Detre considers related to Magyar. As for statistics, for Parts of Body he gives 21 words, of which 14 is "very similar" to Magyar (I cannot make a more linguistic statement for first approach), while for Animals it is 15 from 31. Roughly it seems as if the Hun language were not much farther from present Magyar than the XIth c. Magyar is; and the Magyar-Hun distance seems definitely smaller than the Magyar-Vogul one.

            As for the grammar, the scribes are Armenian, and a Hunnish language, be either Uralic, or Altaic, or anything between, must be very far from Indo-European Armenian for grammar. So we cannot expect the correct recording of details. However for first impressions I can tell the followings.

            The Hun language of [3] seems agglutinating. This is common Ural-Altaic feature. 3 dependent nominal cases + Plural are recorded, with quite Magyar-like suffixes (the Magyar Accusative and Plural is without parallel anywhere so far in Uralic and Altaic). Personal pronouns are also very Magyar-like. As for the verbs, [3] states that much verbs seem irregular. Still the verbal declension is agglutinative, and the endings are generally close to Magyar. Instead too much explanation I simply give the "regular" endings on an Uralic & Altaic background:




Magyar Indet.

Early Magy. Indet.

Vogul Indet.














































Table 3: Verbal declensions in some agglutinating languages; Praes. Indic.


            In this Table lots of simplifications have been done. Magyar & Finnish have multiple endings because of Vowel Harmony, and a variety of extra interposed vowels may appear too (sometimes indicated by -). But anyways: the general picture is the same for any Uralic or Altaic language, but the actual Hun suffixes are nearest to Old Magyar.

            I would not discuss too much the further details of verbal declensions. The reason is double. First, I would not expect too much understanding of an Uralo-Altaic verbal declension for an Indo-European Armenian monk. Second, Ref. [3] gives moods & tenses too, but with partially a bad terminology. E.g. [3] calls a tense "Near Past", but that is really Simple Past or Preterite, contrasted to Far Past, which seems to be the Present Perfect, from the endings. The reason is that Simple Past is now out of use except a small corner of (present) Roumania [16], and it is substituted with the Present Perfect as "Past".

            Also, [3] mentions a combined form which is called there as (the Magyar mirror translation of) the Present Conditional. Now it is given as

heti-(m,sz, ř, nkh, tekh, kh) + Infinitive

where the Infinitive is Root+in.

            Now, this combination is definitely not used in Magyar, it was unknown even in Old Magyar, but it is analogous with the

fog + personal ending + Infinitive

of the Formal Future. However, according to the Vocabulary,

hetin = to be able, to act, in the sense of German "mögen"

But then this cannot be a Conditional. Today's Magyar has a synthetic form

Verbal Root + (-hat-/-het-) + Personal Endings

expressing that the indicated person may/can do something; and that is not Conditional. (Go back to Ą2.) Since henceforth I should analyse synchronously the Hunnish grammar + the authors' individual ideas about Magyar grammar, I stop here with the verbs.

            As for Numerals, the Hun words until 11 seem good Uralic:







Turkish (Ott.)






























































tiz hen idzsi




on bir


Table 4: Some Uralo-Altaic numerals, from 1 to 11


Now until 7 the (Hun, Magyar, Vogul) triad is so close that there is no reason to analyse the differences. For 8-11 the general formation method is similar, for "8" Magyar and Vogul are really close, but for "9" Magyar & Hun are.

            I stop here with the comparison of words. The situation seems simple enough: Hun has a dominant part of vocabulary, which is only so different from Magyar as one would expect from the 1500 years of separation compared to the Old Magyar - Magyar distance of 900 years. Geographically & historically this would cause substantial difficulties, but first let us concentrate on the "other half" of words.



            Ref. [3] is rather brief on the other half of the vocabulary, telling simply that 1) another language, very distinct from the Uralic or Turkic languages, is present; 2) the next strongest influence comes from Armenian; and 3) that there are very few explicitly Turkish etymologies. Now, Point 1) is interesting but needs much more work; Point 2) is natural enough (e.g. the scribes being Armenian) and Point 3) does not seem to be true.

            E.g. look at Colours. Yellow is sarakh. While this is close to the Magyar "sárga", that is Turkish in Magyar (probably Bulgarian Turkish or r-Turkish). Interestingly enough "green"="zezild". That again seems Magyar ("zöld"), but in Magyar it is Osetian ("zäldä"="a kind of grass").

            I try to make a very approximate statistics, using Plants & Animals. There are 14 plant names in [3]. From this 14 10 is related to Magyar, but 2 Magyar words, "alma"="apple" and "árpa"=""barley" cannot be distinguished from Turkish (elma, arpa); and “dió” is Turkish too. 3 others (for "tree", "stock" and "grass") have good Uralic etymologies, and "tölgy"="oak" is something Iranian. But "alma", "árpa" & “dió” demonstrate that Turkish words are really present in Hunnish.

            It is even more so amongst Animals. 31 words, of which, as I told, 15 is probably related to Magyar. However of them some are clearly Turkish. Let us see some traditional etymologies; of course until Cs. Detre does not look for etymologies of Hun words, I can discuss only those of the Magyar ones.






Turk. etym.

Slavic etym.

Uralic etym.































Yes, B














buka, B







vögör, B






































































Yes, B














Yes, B




























Yes, B




Table 5: Comparisons for the origins of 23 animal names.


            Here – means the lack of accepted etymology. B indicates that the accepted Turkish etymology seems to be Bulgarian (r-) Turkish. I omitted indefinite or utterly exotic animals. As told, the etymologies are for the Magyar words, and I composed this Table according to usual Uralic linguistics. Still we can see that lots of the Hunnish words do have Turkish etymologies. 6 Hunnish words surely have Turkish kinship (the majority Bulgarian Turkish); observe that #28, bulbül, is clearly the Turkish bülbül, while the Magyar word is different, and for 2 others (#2 & #3) it is possible. So (Bulgarian or r-) Turkish words are quite frequent among Hunnish animal names, just as one would expect it from the general history of Migrations.



            This Chapter is a Variosa Curiosa. I am not a professional linguist, but at some places I would investigate some details if I could. I mention some points here.

1) The Status of Vowel Harmony

            The degree of Vowel Harmony greatly varies among the related languages. Generally it is a strong rule in Altaic languages (even proto-Japanese might have had it about 800 AD [17]). On the other hand, it is rather weak in lots of Uralic languages. It is a strong rule in Magyar & Finnish, but it is weak or nonexistent in some near relatives of these two. E.g. it is rudimentary (remainder?) in Estonian, next kin of Finnish, and it is nonexistent in Vogul & Ostyak, next kins of Magyar (except for endangered Tavda Vogul dialect, under Tatar (Turkish) influence).

            A simple explanation would be that Vowel Harmony is an Altaic import; it may be, but Samoyedic languages have something similar [18], and I cannot guess any place & time for strong Turkic influence on Finnish.

            OK, we do not understand it; still facts are facts. Now, Vowel Harmony in Hunnish seems much weaker than in Magyar. Of course, the language of the scribes did not know Vowel Harmony.

2) River and Sea

            River="volgjagh", sea="tengir". "Volgjagh" has an ending of the Part. praes., so "something flowing". Now, 4 rivers have this as second half of the name (rivers Danube, Volga, Don & Dnieper), but it has not gone into Magyar. On the other hand Russians (!) took that word. Turks call Volga, or sometimes other big rivers too, Etil.

            The "sea"="tengir" is rather similar to "tängri"="sky", which is then the name of the Supreme Turkic and Mongolian God. Borrowing?

3) The Magyar Name of Ödenburg

            The most Western Hungarian substantial city is Scarbantia (Latin); the Magyar name is Sopron, the German is Ödenburg. The German name was always quite obvious (the last city before the usual desolate territory at the border, according to habits of defence in the Xth century), but Sopron does not mean anything similar in Magyar, so it was routinely assumed that the first Lord Lieautenant was some Sopron. Now [3] tells us that "soprun" is cca. "foot of the mountain"; indeed Sopron is at the foot of the Alps, a mountain enough. Unfortunately, I cannot see the etymology. OK, it may come from the postposition “sopru”, but then what is the etymology of “sopru”?

4) Shanyü Balambér

            This Emperor led the European Huns through the River Volga (Etil) in cca. 370 according to age-old tradition. (Go back to Ą1.) But [3] gives the meaning of the name: "Welambiri"="Strong Ally". Now, whose ally could be Balambér, Emperor of All European Huns?

5) Star names

            Some star names in [3] can be explained (maybe Antares is Vereni because it is red, i.e. veresi?), and some cannot yet. But I am disturbed by 4 names: "Aldebaran"="Aldebaran", "Altair"="Athair", "Bethelgeuse"="Betelgeuse" and "Fomalhaut"="Fomalhaut". Kulturwörte? Whence and in what way?



            Let us now continue the evolutionary discussion suspended at the end of Chapter 8. Some 50 % of the Hunnish vocabulary is strongly correlated with today's Magyar; the remaining 50 % has not been linguistically analysed. Some words may be Turkish (as "bulbül"), some Iranian (as "lady"="aszuni", "fortress"="vara" and "green"="zezild", which are Iranian in the usual way of explanation; for "gold"="zarani" the Iranian etymology is very strong and "zaranya" is "gold" already in Avestan times, so it must have come that way). Slavic words are not possible in 500, and a lot of words may have come from the numerous and not too well known Caucasian languages (e.g. Chechen "sziri" = Magyar "szürke" = Hun "szorild"). The 50 % of "unknown origin" indicates a lot of independent evolution for Hun and Magyar. In the same time the other half is very similar to Magyar. For that half Hun is indeed the nearest relative of Magyar, nearer than Vogul.

            There is a routine explanation in the Turkish paradigm to the analogous fact that some words are rather similar between Magyar & Vogul, while some ones are not related at all. The explanation is partial language acquisition; that Vogul was originally a small Palaeosiberian language, not similar to Magyar at all, and then proto-Voguls learnt a lot of Magyar words. Ą5 Good; but why so much Body Parts & Numerals?

            Really, in its Magyar-like half the Hun is "too modern". From 997 upwards we can follow the evolution of Magyar in written documents. In that time Vowel Harmony was strong but a lot of today's nominal endings had only one form, so could not obey Vowel Harmony, so surely they were still separate words (or agglutinated only recently). OK, in the Hunnish vocabulary we also do not find too much agglutinated endings.

            However there is another unidirectional change: the reduction of short vowels at end-position. We saw that Finnish preserved them at a degree similar to that where Magyar was in 1055, while present Vogul more or less lost them as well as present Magyar. Really it seems that this loss took 4 steps:

Oth step: Short end-vowels of 2nd syllable can have any degree of highness. We cannot verify this step from any record, but surely it existed, and maybe it is preserved in the Magyar Accusative: uta-t, helye-t; or in such forms as vére-s, where the "e" is middle degree of opening, as "o" is [19].

1st step: Short vowels at the end are always high (but full-fledged). That is similar to the present status of Finnish, but the change of the second vowel in the conjugation shows that in some words the original quality was not high: "käszi" but in Gen. "käden"="hand"="kéz" (Magy.), "veszi" but "veden"="water"="víz" (Hung.) &c., but: "laszi, laszin"="glass", "leikki, leikkin"="play". That may have been the Magyar state in 1055, but see also the next step.

2nd step: Short end-vowels are not only high, but also ultra-short; they start to vanish. This may or may not have been the Magyar stage of evolution in 1055; but surely it was reached in 1216, when the Regestrum Varadiensis [17] gives all the nominatives of proper names without the high vowels: Tunc, Legyn, Gyan; but the Accusatives still may contain the high vowel: Isepu-t, Egidu-t, Tadeusu-t. (In 2004 the Accusatives are: Izsép-e-t, Egyed-e-t, but Tádeus-t.)

3rd step: The short high vowels are neither heard nor written anymore. This is surely so in Magyar since 1300.

            Now, Vogul is also at Step 3, which is not a problem at all. In the Uralic scenario Vogul has at least 2000 years of independent evolution and could have reached Step 3 any times before 1800.

            However the problem is that Hun seems to be at Step 1 or 2 already in 500. Examples are:



Hun 500 AD

Magyar 1055

Magyar 2004










































Table 6: On the evolution of final high vowels.


We can see no great difference between the Hun evolutionary degree in 500 and the Magyar one in 1055, in spite of the 550 years between. The only possible evidence for Step 1 for the Huns would be Row 7, but the argumentation is complicated and not too stable. Let us, however, see.

            Modern Magyar "íj" (bow) is back for quality in Vowel Harmony, although the orthography does not distinguish here: "my bow" is "íjam", not "íjem", and "from bow" is "íjról", not "íjrôl". Let us assume that this was the original quality as well. Then probably in Step 2 already the reduced final vowel was automatic, so it would have been "*íjď". Now, as told earlier, Armenian does not write "ď" at all, so we would not expect "viju" in Step 2. So if we see “viju”…

            However there were unproven assumptions, and the 4th column may be Step 1 as well. The difference of Columns 3 & 4 is definitely not much in 550 years. It seems as if the language represented in [3] would result in 1055 in much more "modern" forms for the end-vowels.

            And note that the Hun words do not show 2 other Magyar features as well. As was already mentioned, Magyar and Perm Finns, in spite of the numerous Magyar-Ob Ugor and Perm Finns-Other Finns similarities, agree in 3 features (in which they differ from other Finns & Ugors), namely:

1) existence of voiced-unvoiced consonantal oppositions;

2) the "-ni" ending of Infinitive (it is indeed –nü in Permian, but a high vowel anyways, and the two suffices are homologous, both evolved from  Lative constructions);

3) the -VlVm deverbal nominative ending.

            Now, Hun agrees with Perm Finns and Magyar in Point 1), but does not exhibit Points 2) & 3)!. Ref. [3] gives the "-in" Infinitive ending, and does not give any "-alom/-elem" (or even -VlVm") deverbal nominative ending. Look: "szerelem" is "szirünild" in [3] and "diadalm-as" (from an original "diadalom") is translated as "dievild". (Go back to Ą3.)

            The nature and details of proto-Magyar and proto-Permian connections are rather obscure, still, the results show that the Hun-Permian interaction was weaker than the Magyar-Permian one.

            Finally observe the divergent evolution of "I can do" expressions resulting in the mirror meanings but quite different constructions:





You can drink

Hetitekh hümin



            From evolutionary viewpoint nothing at all suggests that Hun was an intermediary phase of an evolution ending in Magyar! Rather, it seems that [3] implies a language similar to Magyar (if at all) which expired sometimes between 700 and 1990. Now, we do know about a splinter group of proto-Magyars: the Savard Magyars detaching themselves from the main body and crossing the Caucasus about 700. Late mentions are the first Hungarian chronicles (where the formulation is rather obscure), the letter of Pope John XXII to Missionary Bishop Bernhardus in 1329 [11] and the excavations of Halikova at Bolshie Tigani [21], [22], [23]. I do not suggest that there would be the same event behind all the 3 records; the chronology is not clear. I only tell that there is against my educated guesses about evolutionary steps an evolution resulting in Magyar as we know it about 1200 through the stage shown by [3] in 500-700!



            In present years it is quite usual to tell that Schleicher's Stammbaumtheorie belongs to the XIXth century and is not valid. Sure, it cannot be Absolute Truth (argument is coming). However it is simple; and there is no better up to now. "Stammbaum" is "family tree" or "evolutionary tree"; but we should be careful, what is a tree.

            The trees around us are tree-sized specimens of Gymnosperms and Angiosperms. Their branching is monopodial: there is a main branch, from which secondary branches emerge. It seems that even tree-sized ferns follow this pattern.

            However, the first trees in Late Paleozoic showed two other branching patterns. The biggest trees of Late Palaeozoic were lycopsids & horsetails, the first group with dichotomous branching, the second with verticillate one [24], [25], [26], [27]. Now, monopodial, dichotomous and verticillate linguistic trees would differ a lot.

            A verticillate Stammbaum would show a language keeping its identity but at some "convulsions" (e.g. at migrations) emanating more than one daughter languages. Nobody ever suggested seriously such a linguistic tree, although Latin, the language of the Western half of the Roman Empire, produced her daughter languages in such a way after the decline of the Western Empire: after the decay of the unity in many regions divergent evolutions started cca, synchronously.

            In XIXth and early XXth century the monopodial trees were popular; also in biology. Man descended from Apes. Remote languages retained more or less the Ur-language. Or: the nation of the researcher kept the language in its purity. Or: Sanskrit is quite close to the original Indo-Germanic language.

            Now Cladistics prefers the dichotomous branching. There is a species. Evolution destroys that species and creates two daughter species instead, and so for infinity. Language A forks into B & C. Then B forks into D & E, and somewhat later C also forks into F & G; at that moment we have A, E, F & G...

            Try to describe Latin's decay and emergence of Neo-Latin languages in a dichotomous Stammbaum. Then you should find a first language detaching itself from the Imperial Unity. Westerners would tell Roumanian in 270 AD; however it is clear from Theophylactus' & Theophanes’ records of "torna, torna, frater" [28], [29] that this "proto-Roumanian" was uttered South of the Haemus (Balkan) Mountains, in Province Haemimontus, well within the Eastern Empire. Of course, that evolution might have led to Roumanian, because Roumanian formed South of the Danube; but the cited phrase was still Standard Late Vulgar Latin.

            OK, modern linguistics knows 10 Neo-Latin languages (one of them expired in 1898, when Tuone Udina Burbur died, but ample records remained). OK, the first Neo-Latin written text is the Strassburg Oath of King Louis the German in Old French language from 14th Feb. 842. Let us take as the first forking:

L -> (F,L')

Then in 960 the Placito di Capua is recorded: birth of Italian.

L' -> (I,L")

and so on.

            But this would be nonsense. The Late Vulgar Latin in Capua had not been influenced at all by the separation of French in France. I branched from the same stem than F. Latin's decay is best described via a verticillate branching, other cases may be monopodial, sympodial or dichotomic. In the undocumented case, of course, it depends on our belief how we draw the Stammbaum; and this ambiguity indicates the structural problems.

            The second important point (where linguistic evolution is not a Stammbaum, or any Baum either) is hybridisation. Branches of a tree do not unite. There is an irreversible nature of branching: the number of branches increase at any forking.

            For biological evolution it seems as if hybridisation were impossible, so biological evolution is drawn as a Stammbaum. This is correct now; but some billion years ago it was not. You may remember Lynn Margulis' endosymbiont theory [30], and you can see several old hybridisations in Ref. [31]. Even the mitochondria in us and the chloroplasts in plants were in ancient times separate organisms, and even now they are autonomous with their own genetical information & reproduction; in strict sense you cannot draw a Stammbaum with tree-like branches for the kingdoms of biology since the "body" of a plant has a kinship quite different from that of the chloroplasts. Chloroplasts are close kins of blue-green algae. It turns out that the fundamental barrier to hybridisation is now the double cell wall of eucaryotes.

            Languages can hybridise, although full hybridisation is rather rare. An idea is that the problem of the classification of Japanese comes from a hybridisation story: e.g. Northern, Altaic, horseriders and Southern, e.g. Austro-Asian agriculturists met [32] in Japan. Maybe this is remembered in Kojiki & Nihongi, when Ninigi, grandson of Goddess of Sun, Amaterasu, and Grandfather of the first Emperor, Jimmu, descends at Northern Kyushu.

            English is another example. The two parents, Anglo-Saxon & Norman French are both Kentum Indo-European, but the sub-families of them, German & Italoceltic, are not too close to each other. Still, you may believe that you could detect a linguistic hybridisation in the past. When Anglo-Saxon & Norman French hybridised, the synthetic grammar decayed to next to nothing. Modern English has a synthetic Sg3 Paes. verbal form ("asks"), a synthetic weak Praeterite ("asked"), a synthetic weak Part. Pass. (again "asked"), a synthetic Saxon Genitive of nouns ("Queen's"), a twin synthetic suffix (nobly, noble), and practically nothing more. Otherwise English is isolating, grammar is expressed via word order & such. Anglo-Saxon & Norman French pidginised, went down almost to word roots before hybridising. This can be detected even after 900 years, so if we detect this, we may conjecture hybrid origin. Otherwise we do not have any reason to conjecture, so we can draw nice, simple trees. Of course, languages do influence each other, but borrowed words only colour the picture and grammatical tools are hard to be borrowed.

            Now I draw a somewhat simplified Stammbaum for Uralic languages. See [20]:

U (Ur, the Uralic Ur-language)

Forking 1 (-4000)

Finno-Ugric (FU)                     Samoiedic (S)

Forking 2: (-2000) FU into

Fenno-Volgan-Permian (FVP)              Ugric (U)

(we do not follow Samoiedic)

Forking 3: (-1500) FVP into

Fenno-Volgan (FV)                  Permian (P)

Forking 4: (-800) FV into

Common Finnish (CF)              Volgan Finnish (VF)

Forking 5: (-500) Ugric into

Magyar (M)                 Ob Ugric (OU)


Table 7: A consensus tree of Uralic languages


            Henceforth I only give the final results. CF resulted in : Finn, Karjalan, Inkeri, Vepse, Votic, Estonian, Liv and Lapponian (Lapponian is somewhat anomalous). Volgan Finnish resulteed in: Cheremys & Mordvin). Permian (Finnish) resulted in Komi & Udmurt (or Züryen & Votyak). Ob Ugors resulted in: Vogul & Ostyak (or Man'shi & Khanti). And Magyar did not fork since Forking 5, or at least there is no other living language on the Western side of Forking 5.

            This is a Stammbaum liked in Hungary. Finnish trees maybe differ in small details, but the majority of Uralic linguists are Hungarian. The trees are constructed according to more or less important differences, more or less conservative words &c. For first approximation they are all at the same distance from the Uralic common ancestor; but according to the principles of Cladistics, they are not at the same distances from each other. Very, very roughly the mutual distances come from the numbers of forkings between. The unit is the distance between sisters, say D. Since languages within the same group (say CF) are near to mutual understanding, I stop at the groups still shown on the tree. Then the distances are:




CF - VF:


CF - P:


CF - M:


CF - OU:


VF - P:


VF - M:


VF - OU:


P - M:


P - OU:


M - OU:



Table 8: Distances within the Finno-Ugric groups. Very approximative cladistic numbers, based on the tree of Table 7. 1D is somewhat bigger than between sister languages, it is established in cca. 1500 years.


I do not know if these distances meet linguists' common opinion or not; but any cladist would agree with them in first approach, if the Stammbaum is good.

            As for the times of the forkings, the matter is far from trivial. But archaeology can be used with some caution, and borrowed words can be used for dating too, in both directions. Ref. [20] gives some data and they are written at the forkings. Very substantial ± errors should be expected there.

            If this evolutionary tree is true, then between Permian (Finnish), the Easternmost Finnish, and Magyar, the Westernmost Ugric, the distance is big, in spite of being neighbours, because the last common ancestor was Finno-Ugric, 4000 years ago. Analogy: although German & French are neighbours at the Rhine, French goes back to Latin, then Italic, then Italoceltic and thence to Proto-Indo-Germanic, while German to Western Germanic thence Germanic and only through that to Proto-Indo-Germanic. The last intimate closeness of Italoceltic and Germanic surely was not later than -1500.

            As for Magyar and Hun (anything this latter be amongst Altaics), Cladistics + linguistic common opinion would suggest the following. Surely, Uralic & Altaic are kins [32], maybe sister families. The agglutination is common, the Vowel Harmony can be found in all Altaic and some Uralic languages (including the 2 biggest Uralic ones), and Collinder collected 69 common words in the proto-languages. However the common parent language could not have forked into Uralic & Altaic after -5000, so between any present Uralic & Altaic pair the distance must be cca. 10D; far enough.

            And now let us see first the scenario suggested by the best evolutionary tree for the formation of Magyar.

            There was a Forking 0 resulting in Uralic and Altaic. Then came the Forkings of Table 7. Some archaeologists believe that not much after Forking 2 they recognise Ugors as one half of the Andronovo culture; and the other half is Iranian, (see e.g. [33].) Until Forking 5 the Ugors live together, but not too compactly, and proto-Magyars are the Southerners (or Southwesterners). After Forking 5 proto-Magyar separated and started its individual life, surely in a generally Southern migration. Unfortunately there are no mentions of Magyars for the next 1300 years. Just after Forking 5 there is a very uncertain spoor: in the neighbourhood of Chelyabinsk the Sauromat graves are oriented to West, but there are graves also with Northern orientation. You may or may not regard this second group as proto-Magyars.

            The Ob Ugric results of Forking 5 left clearer remnants, Ust’-Poluy culture from the Ob Valley is a mixture of Arctic autochtones (Samoieds?) and (relatively) Southern horsemen [34]. Note that the eponymic site at Ust’-Poluy comes from 5th-3rd c. BC, and it is very Arctic, near to Icy Ocean. Ust’-Poluy culture occupies a substantial part of Middle & Lower Ob, and a millennium later Ob Ugors are there. So very probably the Ust’-Poluy bronzes give a snapshot from a process in which Ugors from the South and Northern autochtones (but probable Uralians too) started to merge in present Ob Ugric. The details are irrelevant here; but note that the many words in Ob Ugric languages which are not Ugric (i.e. cannot be found in Magyar) may come simply from the unknown autochtones. (Go back to Ą5,)

            After Chelyabinsk even the uncertain traces of proto-Magyars vanish, but linguists can detect first Iranian loanwords (they appear amply in other Uralic languages too, so they may come from nomadic Iranians as Sarmathians, Scythians, Alans &c.). Then come Turkish loanwords, mainly from r-Turkish. Then, finally, Magyars appear in records in the 800's, East of the Carpathians, maybe first bw. Rivers Don & Dnieper. In that time they are populous, they are horserider half-nomads (they have quite developed agriculture too, including grapes), and 5 of their 7 tribes have Turkish names (4 of 5 r-Turkish). However in the middle of the IXth century the supreme leader of the (con)federation is from the tribe Megyer; the word is very probably close kin of the word Magyar and it is hard to avoid relation to "Man'ch" = "Man'shi" = "Vogul".)

            But now we have to fill in the story between 500 BC and 463 AD. From this period we know only the linguists' statements about various Iranian horseriders' words and about the individual life of Magyar. Maybe some proto-Turkish tribe was in the neighbourhood too, because 3 Turkish words (for "swan", "beaver" and "word") are common with Ob Ugors, but we cannot identify that tribe and maybe it was not too strong. Maybe in some part of that millenium proto-Magyars were just neighbours of Permians and influenced each other; but the two languages were already at 3D distances, unintelligible without heavy learning. (Now a Magyar is quite lost in Finland, except that he can understand some Swedish inscriptions via German or English [35]. He is also quite lost in Syvtykvar.) But somehow, one learnt from the other the -VlVm deverbal suffix and the -ni/-nü Infinitive from Lative. (We do not know the direction of learning.) Then came the Huns of Balambér (or not). They crossed the Volga in cca. 370, so they could contact (influence? kill? they definitely did not rob them, because of having nothing important for a Hun) Permians about 340. We are in the time of Huns tearing Grigoris, grandson of Saint Gregory, the Illuminator, Katholikosz of Armenia into pieces [12]. OK, maybe Huns did not go North enough to meet Komis; but maybe they went. Proto-Magyars were nearby, and by any educated guess, Southward, on horses. (See the common Ugric horse vocabulary.) Surely the Huns found them. What happened then?

            We do not know! It is the consensus of Magyar historians that there is no genuine Magyar tradition about the Huns. My tribe knows otherwise, but my tribe can be Bulgarian Turkish quite well. Some Magyar historians admit that the ruling clan of the Magyars might have come from the Hunnish federation, so the rulers might have had some tradition, but not the people. Others admit that some tradition migh have diffused from the Bulgarian Turks, but centuries later. But no genuine, original Magyar tradition.

            Surely, had Huns teared the Supreme Shaman of Magyars into pieces, there would be a tradition about. However this in itself does not prove that Magyars and Huns avoided each other about 340. We do know (from Pawstos Buzand) why the Huns teared Grigoris, Katholikos of Armenia, into pieces. He spoke long about the virtue of self-abstaining (either from women, or from drinking or from eating; surely from all). Now, Magyars quite might have been horseriders inferior to Huns in 340; but the shamans of even inferior horseriders would not speak about such nonsense, and then they are not harmed substantially. Maybe Magyars (on horseback) made a wide detour (but whither? you cannot go too far North on horseback); maybe they submitted to the Huns, or maybe they were taken as parts of the army. After that they either herded the Hunnish animals, or guarded the Northern borders or went to the West. (As I told, Hungarian historians tell that they cannot detect any tradition about.)

            Then the main body of Huns returned to the East. If Magyars did not go West, they may or may not have met the retreating Huns at the Volga or Kama. If Magyars had gone to West, they either returned, or not, or both. But Magyar linguists cannot find any loanword in Magyar from the Vth century Carpathian Basin. (OK, how to distinguish an Vth century Alan word borrowed on East from one borrowed in the Carpathian Basin? But e.g. there are no Vth century Latin words in Magyar.)

            Then in a few years starts the second wave of migration recorded by Priscus Rhetor; the Bulgarian Turks are on the move, and until today Magyar tradition clearly remembers Bulgars at the Maeotis. (Hunor and Magor take the daughters of Kings Dul & Belar. Belar is Bulgar according to anybody; Dul is either Dulo, the ruling clan of the Bulgars, or a Dula clan of Alans.) Henceforth a strange Magyar-Bulgar (or Magyar-Onogur) symbiosis starts. Ugric Magyars lead a confederation whose majority is Bulgar Turk (at least this is true for tribe names), Ugric Magyar takes hundreds of Turkish words from Bulgarians, and in spite of this the Magyar language does not pidginise.

            I guess, why not; but this is still not usual. The old Turkish loanwords came into Magyar in their root forms, SgNom for nouns and Sg3 Praes for verbs. It is Pure Root in both languages. But then the Magyar speaker could use any Turkish root as a Magyar root. I tell an example from Modern Magyar & Ottoman Turkish. No doubt, the example itself happened many times between 1541 & 1686, when Ottoman Turks occupied some half of Hungary.

            You are speaking with somebody, and you want to be polite. Not deferential, but polite. Then you start with "Sir". Now, as a polite title, it is approximately "Ur" in Magyar and "Efendi" in Turkish.

            Of course, we cannot start a Magyar sentence (or a Turkish, either) with the mirror translation of "Sir"; that would be rather ungrammatical. You must use the construction "My Sir" (see "Milord" for analogy, or "Monsieur" from French). Then let us see the mirror translations:

(My) Sir (E) = Uram (M) = Efendim (T)

Even the Sg1 possessive suffix is the same! But the similarity of suffices is not important, see Pl1:

Our Sir = Urunk = Efendimiz

Now imagine that, e.g. for fashion, one wants to use "efendi" instead of "úr". No problem. He can put the Magyar possessive suffices at the end of the Turkish word, and gets a Magyar construction proper in any Magyar sentence. Indeed, "efendi" was borrowed, but with a geminated "ff", resulting in "effendim, effendink", &c. You may fill up half of the Magyar vocabulary with Turkish verbs, nouns or adjectives, and all the Magyar synthetic constructions continue to work.

            There is e.g. a triplet in the Magyar vocabulary for: military/police officer = tiszt = csôsz = csausz

The first is originally German, the second is originally Kipchak/Kumyk (or Petcheneg), the third is Ottoman Turkish; the Ugric word is not used today. Just now "csôsz" is a surveyor of parks, playing fields & so, and "csausz" is used only in Ottoman Turkish and historical contexts. If you want to add some suffix, let us take e.g.

from the officer = a tiszttôl = a csôsztôl = a csausztól

you can see the Vowel Harmony in work independently of the origin of the root. Similarly, Turkish "basď" = "head, leader". "Yüzbasď" is "Captain", but in mirror translation "head of hundred", "onbasď" is "head of ten" and so on. Now in occupied Hungarian cities the head of (military) administration was a "basď”; Magyar imported the word as "basa" (no "ď" in the Magyar system and back vowel was needed) and used without problems as "influential and rich Turk".

            An interesting verb is "öl"="he dies" (in Turkish). In Magyar, however, "öl" is "he kills". "He dies" there is "hal" (Vogul "hóli"). There might have been some complication or misunderstanding; but no grammatical problem. "Blue"="kök" (T)="kék" (M), and "yellow" = "sarď" (T) = "sárga" (M). I simply do not know what were the words before the r-Turkish connection: in Finnish "blue"="sininen", but that is clearly Russian. In Vogul "blue" is "atörhari" but that is unintelligible for a Magyar, of course.

            I stop here with Uralic & Altaic comparison. The essence is: Magyar is full with Altaic words, but that fact never disturbed anybody. I do not know how to detect if Magyar took half of its word roots from an unknown agglutinating language but I am sure the only trace would be lots of words without any etymology. Magyar easily hybridise. It can take Latin words too, but for Latin verbs Magyar puts some formative suffix to the root ("concurs"="konkurrál", and then this will be the Magyar root"), and then you see that the word was borrowed from somewhere; but not from Turkish. Now, you cannot be sure that you know about all agglutinating languages in the neighbourhood bw. 500 BC and 463 AD (including extinct ones).

            In 551 Turks defeated the Avars in the Altai Mountains. Avars went to West and occupied the Carpathian Basin in 568. In principle they might or might not have taken proto-Magyars with them. I rather would think they did not, because they did take lots of proto-Slavs. For example, proto-Slovakians; Serbians from the West (see: now Sorbians live in Eastern Germany); Croats from Northeast (White Croatians lived from Cracow to Ukraine for more centuries). And if they had taken also proto-Magyars, then Magyar would be full now with Old Slavic words. True, Magyar is full with Slavic words, but with not so old ones.)

            Then there start troubles in Avaria. A new wave arrives around 680. The direction is roughly the (feet of the) Caucasus; this is a commonplace in Hungarian archaeology (their artifacts clearly tells so), and look: some Bulgarian Turks found Danube Bulgaria in 681. And now again I call the attention to the Chancellery-supervised Vienna Illustrated Chronicle, which dates the Conquest of the Basin by the direct ancestors of Hungarian Kings to 677. It cannot be a pen error: the datum is given twice, written by words, not by Roman numerals, and the text gives the synchronous Pope and Eastern Emperor too. The given persons did rule in 677 according to our modern Tables.

            So the newcomers of the Carpathian Basin must be intimately connected with Bulgarian Turks; and Bulgarian Turks may have come from the general direction of the Caucasus. No Hungarian archaeologist or historian would be surprised about Bulgarians in the Basin in 680; and Onogur Bulgars are detected there in the next century. What is more, Modern Slovakian calls the Basin Uhorsko; and this "uhor-" is not "ugor" but "onogur". (Proto-Slovakian "on-" evolved in many cases into Modern Slovak "u-", and "g" into "h". E.g. in Modern Slovak "mushroom" is "huba", but Slavists know that until Xth century it was something "gomba, gonba", with a nasal "o", and it is exactly "gomba" in Modern Magyar.)

            But might have some Magyars come with the Bulgarians in 680? The symbiosis probably still went on, so they easily might. Indeed, it would be hard to see how not (not why not). But observe that no Magyars were detected in Danube Bulgaria. So some Hungarian historians/archaeologists tell: they might have but they did not. Others tell: they might have and did. We did not know for sure. There are indeed Old Slavic words in Magyar, but it is hard to say, too much or too few.

            So where are the Magyars from 680? Some may be in the Carpathian Basin, some outside. Between Don and Dnieper? Without problem; there are such theories in Hungary. At the feet of the Caucasus? There are such theories in Hungary too. We do not know.

            Western sources cannot be used too much. A few synchronous sources mention Huns or Hungarians; the first can well be archaism, the second can be Onogurs. The sources never mention Magyars, and do not mention clearly (proto-)Magyar personal or place names. However even during Xth century Magyar chiefs will have Turkish names.

            But Westerners do not observe even the great emigration wave in 680. For them the Carpathian Basin is Avaria, and it is near to a black hole whence information does not emerge at all. In Byzance the last Avar embassy was accepted in 678. Later Magyar chronicles may preserve some Avar traditions (or they may be imaginations), but they always mention "Huns".

            Then in 860 King Louis the German of the Eastern Franks issues a donation letter to the Melk monastery (in present Austria [36].) It defines the Eastern (or Southeastern) border of the lands as "Uuangariorum marcha", i.e. "the Vangari Marches". It would be a tricky accident f any other people were meant here than the "Ungari". Nowadays the German "Ungarn", the French "Hongrie" and the Latin "Hungaria" is "Hungary".

            But please do not conclude that the "Vangari" are (proto)-Magyars. They may have been. But they may have been Onogurs either. Or even Slovakians living in Onogurian lands. The most probable interpretation is "some Onogurs", so badly defined Bulgarian Turks. They may or may not have their friends and stepbrothers, Magyars, with them. Even it is possible that Louis the German calls Magyars as Onogurs. Linguists should look after proto-Magyar names in the Basin. They look, indeed, and the result is eqivocal.

            However the situation is not much better even outside the Basin. Nobody records Magyars on the Eurasian steppe, until 870. Then a Persian statesman called Jayhani recorded something in his book Kitab al-mamalik va’l-masalik [37]; this book is lost, but other travellers copied/modified the text, and now we are interpreting the final results. Anyways, Gardizi & Ibn Rusta (Zayn al-akhbar –t) write [38], [39] (maybe after Jayhani) that the Khazarians erected fortresses against the Magyars. (Sometimes they write something M.j.f.r, which is maybe a pen error instead of, sometimes rather; the latter may be Baskhirians, maybe misspelled Majars~Magyars, maybe a third nation. Who knows?) But in the Magyar tradition there is both alliance and enmity with the Khazarians. So then the fortress of Sarkel was built about 830 against Magyars; but note that Sarkel is not too far from Northern Daghestan.

            Khalikova believed that the Bolshie Tigani group had been Magyars, ancestors of the Conquerors of the Carpathian Basin, as just migrating to the general area West of Don. Sarkel is just between Volga and Don, where they are nearest to each other. Fodor [40] does not believe this: he thinks that the Bolshie Tigani people remained behind. If you like, you can invent different Magyar groups wandering to and fro on the steppe. Maybe the Daghestan Huns of [3] are such a people in close kinship with Magyars in 700. We do not know.

            As for the name Hun, [3] gives the translation "chun"="man". Now, if one is fanatic, he can interpret this in the Uralic paradigm, either as "man, male" = "hím" (Magyar), "chum" (Vogul), or as "host, army" = "had" (Magyar), "chant" (Vogul), "kant" (Ostyak), "konydä" (Mordvin), "kunta" (Finnish). Identification is not impossible, but it is better to try to stop sooner or later.

            And then comes the last act (so far). In April 21, 896, Prince Árpád and his tribal federation crosses the Verecke Pass at the Northeast of the Carpathian Basin (surely other passes too), and takes the Basin. However there are professional historians telling that Prince Árpád's people spoke Turkish and the Magyars had already been in the Basin (of course, from 680). They are excellent historians, and their main argument is that the newcomers' cemeteries are too small; they were not enough to fill up the Basin with the Magyar language. Maybe; maybe not. But it seems that Árpád led the Megyer tribe, and it would be tricky if the Megyer tribe would have spoken Bulgarian Turkish. Of course, in principle anything may happen in a symbiosis.

            In more 104 years the direct descendant of Prince Árpád, Vajk (the name is Turkish) takes the name Stephen (István, Stefán &c.), gets a crown from Pope Sylvester II, and founds the Apostolic Kingdom of Hungary. In more 55 years we have the foundation letter of the Tihany Abbey, and look: in 86 isolates there are good and genuine Magyar words there. No other languages than Latin & Magyar. Of course there are words of Turkish & Slavic origins among these Magyar words: but then what? There are words of Turkish & Slavic origins in the Magyar language in 2004 too, and it is neither Turkish nor Slavic language but Magyar. So no problem after 1055.



            I told a story in the previous Chapter. A story leading to present Magyars. This is the most elaborated scenario, and also most accepted in scholarly circles. However the story is rather awkward at some points, e.g. there is a millenium gap in it. Also, the words of the Vocabulary of Refs. [2] & [3] are rather against the story. And the story does not explain at all the "pecularities" of Magyar history.

            Look, Magyar history (between mid-1st millenium and 1000 AD; at that Christmas Magyar history merged with others, and is now a part of Hungarian history; bene docet qui bene distinguit) was a story of success, without explanations up to now. A hunter-gatherer + fisher community detaches itself from other hunter-gatherers. Both groups have horses, but no cattle, no sheep, no agriculture. You can live without cattle & corn; you can live well even; but you cannot go then to the mild South, where populous neighbours defeat you in one battle and take your lands. Now the Ob Ugor brothers are 25,000.

            And a handful of Manychars (this is the archaic form) did just this. At the subtropic 50th parallel about 500 BC there roamed the lords of the grasslands: the various tribal alliances of North Iranian Scythes and Sarmathians. There were somewhere in the neighbourhood also some Old Turks (see, the words, "word", "swan" & "beaver" were spoken near to  Ugors), but they may not have evolved further than the Manychars. (The knowledge of "swan" & "beaver" does not prove economic & military excellence.)

            And then we do not have any idea where Manychars went for a millenium. We know that they interacted with horserider Iranians. At the end of that millenium they had already cattle and sheep. They maintained a full horserider way of life somewhere on the borderless Eurasian grassland. But they kept their Eastern Uralian language. By other words, they were not assimilated and not subjugated on the borderless grassland where slightly better horseriders win in a single battle.

            And then, maybe from Vth century, proto-Magyars learn agriculture from Bulgarian Turks. (The words are Bulgarian Turk in Magyar.) In this time already the Turks were the lords of the grasslands. The Huns may or may not have been Turks (they may have been also proto-Mongols or an undifferentiated stock between Turks and Mongols as well), but ignoring them I could list here dozens of r-Turks and z-Turks roaming there. Bulgarian Turks were abundant on the grasslands in the V-VIIIth centuries; and Magyars were the only ones with Uralic words for Body Parts, Numerals & Horses.

            And still in 2004 the only Bulgarian Turks are 2 million (the Chuvash); Magyars are 15 million. (To be sure, z-Turks are some 100 million. There are no big miracles.) Magyars not simply have learnt agriculture from r-Turks; they have become more successful in 1500 years.

            But how? We have not got any answer to this question. But we should. What is the use of a history not explaining the most important events of Past? (Sorry, for a Magyar this question is more important than the discussion of the success of Alexander the Great. First, because when his father Philip started the anti-Persian campaign, the Western forces were better equipped, the society & economy behind was more versatile; the only shortcoming was the less population. Even without Alexander's genius the defeat of the Persian Empire was sure, although not her elimination. And second: we have been just admitted into the European Union...)

            Now, this excursion is not the answer. It is not even an analogy. It will simply demonstrate that in some cases the confusion originates from the wrong classification.

            Look; I avoided the worst misclassification. I do know that Hungarian and Magyar are not synonymes, and I must confess that I learnt this indirectly from Slovakian (a parent of mine had some Slovakian schooling). Accordingly in 1992, particle physicist Pisút, Minister of Education of Slovakia, suggested the introduction of a new term into Magyar and was totally refused...

            But there may be other misclassifications as well. We speak about Huns but we do not know the language of Huns. We speak about Avars but we do not know the language of the Avars. The Late Avars (from 680 upwards) were almost surely different from the Early Avars. Did Bulgarian Turks immigrate into the Carpathian Basin about 680? Surely, they did. Did Caucasians immigrate into the Carpathian Basin about 680? Surely, they did. Were these two different from each other? We do not know. Did Onogurs immigrate into the Carpathian Basin about 680? We do not know. Were Onogurs & Bulgarians different from each other? We do not know. Did Onogurs immigrate into the Carpathian Basin in any time before 860? They definitely did. Did proto-Magyars immigrate into the Carpathian Basin together with Bulgarians about 680? We do not know. Did proto-Magyars immigrate into the Carpathian Basin together with Onogurs in any time before 860? We do not know.

            And: were Huns utterly defeated in the Carpathian Basin in the Nedao Battle in 454? Yes, they were. Did they leave the Basin? Yes, they did. Were they North of the Black Sea assimilated into the Bulgarian Empire led by Irnak, son of Attila up to 469? (The sentence is deliberately involved.) Yes, they were. Did then Huns vanished in 469? No, they did not. Armenians continue to refer to Huns near the Causasus. (See also a recent Daghestan expert [41].) What is the relation between Carpathian & Caucasian Huns? Ask something easier. And so on.

            Now comes the example of wrong classification. The respective classification was definitely wrong; it is now proven. It did not touch such sensitive points as national pride; it touched something more sensitive: humanity itself.

            If one does not accept biological evolution, he may jump to the next Chapter. But he does not have to; the case is interesting even for creationists.

            What is/was hominisation? It was an august process leading from (brute? soulless? mindless? simple? anybody can use the most proper adjective) animals to (again some adjective) Homo, the top of the Evolution (so far). The last step of hominisation was of course separating Homo from Pongidae. When did this happen?

            Until the 80's the answer went in the following steps. First there was the improved Linnean taxonomy. In it both Man and the Great Apes were in the Superfamilia Hominoidea, but the Hominoidea superfamilia divided to 3 Familiae: Hylobatidae (gibbons), Pongidae (great apes) and Hominidae (man). Familia Pongidae contained 8 genera, of which 3 was extant, Pongo, Gorilla & Pan (orang, gorilla & chimpanzee). In contrast, Familia Hominidae contained 3 genera, Ramapithecus, Australopithecus & Homo, of which one species of genus Homo is extant, Homo sapiens [42].

            Now, hominization is a process (in taxonomic language) in which something having a name manufactured from the Latin word Homo became existent. Then palaeontology would tell, when.

            But on what taxonomic level should humaneness be reached? Obviously the Superfamilia Hominoidea is not enough. We should not call a process hominisation if it ends in a gibbon. Familia Hominida was already something, but generally to develop into Genus Homo was considered crossing the Rubicon.

            Now, several Pliocene & Miocene fossils were classified Pongidae rather than Hominidae, on the grounds of quadrupedalism & dentition. The family Pongidae seemed to have existed 15-20 My ago, so then Hominida too. As for Genus Homo some 2-3 My past was estimated.

            Obviously (continued the argumentation), the 3 Pongidae sister genera, Pongo, Pan & Gorilla are much nearer to each other than to Homo. The last common ancestor of all of them lived 15-20 My ago, while the common ancestor of the 3 living pongids may have lived 7-8 My ago in the Dark Ages without fossils. So European dryopithecids 10 My ago may still have been undifferentiated pongids, or (since Pongo is rather divergent) at least common ancestors of chimpanzee & gorilla.

            There was a single observation contradicting this picture: Sarich & Wilson studied antibody/antiserum reactions of bloods. Via such reactions indices of some biochemical dissimilarity can be measured. Table 9 shows these indices [43].


Sp. of albumin

Homo antiserum

Pan antiserum

Hylobates antiserum

Homo sapiens




Pan trogdolytes




Pan paniscus




Gorilla gorilla




Pongo pygmaeus




Symphalangus syndactylus




Hylobates lar





Table 9: Some dissimilarity indices among apes & man [43].


The index is exactly 1 if the bloods are identical on species level. While for untutored eye the Table is harmless enough, the numbers indicated that Pongo is disjoint from the group {Homo, Pan, Gorilla}, while on anatomic grounds Homo should oppose the group {Pongo, Pan, Gorilla} (Hominids vs. Pongids). Sarich & Wilson guessed the divergence times on the grounds of immunologic dissimilarities, and the result was:


Common ancestor of

Time, My

Old Word monkeys/apes


Man & All Apes


Great Apes & Man


Man, Chimp & Gorilla


All Gibbons



Table 10: Divergence times from immunology [43].


            And this result was surprising enough. It told a quite new story. That there was a common stock of Man & Great Apes until 8 million years ago, and then the ancestors of the orang detached first. Then for 3 more million years a common stock of African apes & man lived, and then this stock trifurcated. By other words, Sarich & Wilson told that bloods behave as if Man, Chimp & Gorilla were closer to each other than any to Orang.

            OK, this was a single datum. Maybe we would understand it later. But this result excited other biologists, and in 1975 King & Wilson repeated the study for dozens of alleles [44]. From allele frequencies a genetic distance can be calculated. The result was again surprising: the distance, which is routinely normalised to cca. 1 for sister species, was 0.62 between man & chimpanzee. As usual for twin species, pairs of non-hybridizing ones which however, cannot be told apart by classical criteria. In 1978 Bruce & Ayala got even smaller distances [45].

            Then evolutionists became excited, and revisited the similarities among man, chimp, gorilla & orang. They found [46] that indeed Pongo stands apart. A quite recent study [47] gives the following data for times when the human line detached itself from the respective ancestors of living apes:



Divergence time, My

Standard error, My














Table 11: Divergence times, 2001


            So Familia Pongidae contains only Pongo + extinct relatives. But then Pan & Gorilla must be transferred into Familia Hominidae! Pongo keeps the long separation period; indeed Pongidae are far, “only” Pan & Gorilla do not classify together with him. But what is hominisation in the new paradigm?

            If the man-chimpanzee distance is such than for sister species or even smaller, then, in addition, chimps could not be classified into their own Genus. This step was the hardest, but now I can refer to [48], where the taxonomy is as follows:

            Familia Hominidae contains Subfamilia Homininae. This latter divides into 2 Tribes: Hylobatini & Hominini. Tribus Hominini consists of 2 Subtribes: Pongina & Hominina. And Subtribus Hominina divides into 2 Genera: Gorilla & Homo. True, Genus Homo contains 2 Subgenera; but in first approximation subgenera can be forgotten. Namely classical Linnean names are Genus (with capital initial) + species (small letters) So man is Homo sapiens, common chimp is Homo trogdolytes and pygmy chimp  is Homo paniscus. Is an evolutionary process, then, hominisation if it ends in Homo paniscus?

            But confusion in Latin does not mean confusion in facts. I give a full Table of distances. These distances are percent distances in the DNA: simply the percentages of different base pairs.







Monkeys (OW)

























Monkeys (OW)







Table 12: Percentage differences in DNA among some Primates, according to Ref. [48]. Between any ape & mouse the value is cca. 40.


            Divergence times were also calculated between any pair [48]. So I could answer the question about the hominisation process, if you defined what is homidisation. As you see, the meaning is far from being trivial; it is rather paradigm-dependent. And if you form inhomogeneous groups, distances, divergence times &c. will be mere artefacts.

            Now, Ref. [3] is not a linguistic monograph, but a painting-book. Also, it is not a historic monograph, so it leaves us ignorant about the origin & migration of Daghestan (Belendjer) Huns. So, in contrast to palaeoanthropology, where, from morphology, we did have a classification, a priori one with respect to divergence times, although later it turned out to be wrong, now we must operate paradigm-independently.

            I demonstrate. Is "üker"="ox" in the Hun Dictionary [3] a (Bulgarian) Turkish word, or not? Detre tells that words of explicitly Turkish origin are rare in the Dictionary. I, in contrary, can detect a lot of tem. My definition for an explicitly Turkish Hunnish word is  that it should be found in Turkish languages but not in Uralic ones, of course ignoring Magyar and the Ob Ugors. The word "üker" is such. It is "öküz" in Ottoman Turkish (which is z-Turk), "vokor" in Chuvash (which is r-Turk), and similar words are absent in Finnish languages. I know that "üker" is good Mongolian too, but maybe the Turkish words come from proto-Altaic, common ancestor of Turkish & Mongolian. In contrast, Hunnish "szöm"="eye" is "explicitly" Uralic. Namely, we must again exclude Magyar from the comparisons, since we do not know the relation of Magyar & Hunnish a priori. (For any case, the Magyar word is "szem".) OK, it is "szam" in Vogul, but in the "Turkish paradigm" Vogul took it through Magyar, so we should rather ignore it, otherwise the result will not be paradigm-independent. But I can give very similar forms from Western Uralic languages with practically no Turkish contacts as: Finnish "szilmä" and Lapponian "csalbme". In the same time in Ottoman Turkish it is "göz", not related at all. So "üker" is an explicitly r-Turkish word in [3], while "szöm" is explicitly Uralic. (Accidentally both are closely related to Magyar.)

            And be careful. There are words either coming from the common ancestor of Uralic and Altaic, or being very old borrowings. "Mother" is "maja" in the Hunnish Dictionary, but it was "emä" in both proto-Uralic and in proto-Turkish [32]. And "stock" (of a plant) is "tüvi" in [3], but the corresponding word is "tünge" in both proto-Uralic and in proto-Altaic.

            Now palaeontology begins to fill in the gap of the Dark Ages of Hominisation between 4 & 8 My; e.g. an Ardipithecus remnant is estimated to cca. 5.5 My [49], and Orrorin tugenensis is cca. 6 My [50]. In contrary, the lack of archaeological finds connected with proto-Magyars between Vth c. BC and Vth c. AD remains complete.



            This will be a desperate Chapter. But we do not understand the situation, so I must cover even improbable scenarios. So: is it possible that Belendjer Huns carried their Magyar-like words from the West?

            Belendjer Huns' origin is far from being fully clarified, but very probably they arrived at Daghestan after a West-East migration. In 454 they must have been in the Carpathian Basin, or adjacent territories. Magyar historiography & archaeology does know about different tribes retreating to the East, or even remaining in or near to the Basin for decades [51], [52]. The Martinovka culture around Kyev survived until cca. 550, although of course we do not know the language of the population. We know about a tribe Sadagar remaining in Lower Pannonia in 454 and fighting later with the Ostrogoths; then they went to the East [51], [53]. The name is fairly similar to Saragurs in Daghestan since 505.

            In contrast, we do not know about tribes emigrating to the East either during the Avar troubles around 630 & 680, or at the end of independent Avaria, in 795-803. The Fredegarius Codex writes about emigration to Bavaria in 630, and Greek Orthodox tradition in Sirmium knows about emigration of Kuber, close kin of Khan Tervel of Danube Bulgaria, some time after 680, but this second emigration ended in Macedonia.

            Even if Magyars' wandering on the Eurasian grasslands is rather obscure until now for us, surely the last steps are recorded. In 948 a Magyar embassy went to Constantinople and talked to the litterary Emperor Constantine Porphyrogenetus, who then incorporated some information into his book [54]. Now Chapter 38 writes that just before the Conquest Magyars lived in Atelkuzu (Ατελκουζου).

            The usual Magyar interpretation is "the territory between 2 (big) rivers", and then rivers are selected according to theory, but clearly the area is adjacent with the Carpathians, on the East, because the Conquest starts with crossing the Eastern Carpathians in any Magyar and Hungarian tradition. However "between 2 rivers" is improbable.

            The usual interpretation starts with reading the Porphyrogenetus in Magyar: Atelkuzu ≈ Etelköz ≈ (a 2) Etel között = between the 2 Etels, “Etel" is of course "Etil" which is Volga, but maybe it may mean Don too, or Dnieper; and "között"="between". In 948, one century before the Tihany Foundation Chart, of course words end in high vowels, so we could expect not "köz", but "közü" or "küzü". QED.

            But this reading is impossible. Then the Porphyrogenetus would not have written "κουζου". In Greek "u" is written with the digraph "ου", while "ü" is a single letter "υ". Had the Emperor heard front vowels "küzü", he simply would have written " κυζυ". In the text once he writes even "Atel & Kuzu". So Kuzu is a river. It is not clear, which one; but clearly the Dnieper-Bug-Dniester-Prut-Seret area is meant. Indeed when the Emperor goes into details, he names exactly five rivers, and some of them are recognisable. And near to Dniester 2 graves were found with quite (if not fully) Magyar-like weapons & ornaments [55].

            The Porphyrogenetus recorded the previous Magyar lands too, called Lebedüa. According to him, the river of this land is the Khingilos/Khidmas. Now, various Hungarian historians suggested emendations for both names. According to them, Khidmas would be a right-hand tributary of River Bug, the Kodďma, while the Khingilos is a left-side tributary, the Ingul; the two confluences are near to each other.

            Consequently in the IXth century Magyars moved from East to West. Of course, earlier they might have moved in any direction; we simply do not yet know.

            By the way, the Porphyrogenetus tells that sometimes earlier the Magyars forked; one part went to the borders of Persia, and they were then called σαβαρτοι ασφαλοι. Nobody exactly knows what name is this; the usual interpretation is "unmovable Sabirs" or something such (but you may even read here a misunderstood Iranian "asva"="horse", so perhaps "horserider Sabirs".

            Czeglédy tried with a Central Eastern viewpoint, centered in Kazakistan/Eastern Turkestan [56]. In his narration the Daghestan Huns are first remainders of  Balambér’s & Attila’s Huns. Then Saragurs arrive from the West, but in 466; and in 506 Sabirs (σαβαρτοι?) from the East, and Balangars (Belendjers?) are Sabirs. There is then Khazarian overlordship from 557, however with a transient Bulgarian independence in the first half of VIIth century. Indeed, lots of details are uncertain about Caucasian Huns. I personally would expect a Bulgarian or r-Turk language, but the traditions might be Western, Eastern or mixed as well.



            But the Hun Dictionary of Detre [2], [3] comes from 500 or 700 or both. (Two codices and two data are reported [2], [3].) Cca. half of the reported words are very similar to Magyar words (also for meaning), the other half is not. What does this mean?

            I do not know. 3 things are sure.

            1) Well documented and careful observations must have precedence to any nice theory.

            2) If observed facts clearly contradict the predictions of existing theories then those theories are wrong and must be substituted by new ones.

            3) However there is a big trouble if we cannot even imagine the form of the new theory.

            Now, to Point 1) I must state that [2] & [3] are not well-documented. On the contrary, they are extremely ill-documented, at the end published in a painting-book. On the other hand, in [3] two decent scientists state that one of them, together with a good Armenologist, simply took an Armenian manuscript, collected Hunnish words from it, and there were Armenian definitions, translations or what besides them and they took them too. Now, surely Ö. Schütz (sometimes written as E. Schütz too) was a competent Armenologist and could do this easily. I do not know the extent of Armenian of Cs. Detre, but I do know that he knows some Armenian, and surely he knows all the letters. So in any time he can transliterate any part of an Armenian manuscript. Of course it is not trivial to read medieval manuscripts, so some errors are possible; but 10 % error would not help too much about the problems raised by this Hun Dictionary. Unfortunately Dr. Detre did not publish the rules of transliteration; I guessed them, but I definitely do not know how consequent they were. However I understand from [3] that Ö. Schütz took an importrant part in the transliteration and he was a professional Armenologist. It is meaningless to ask if these Huns were "really Huns". The Armenians reported Huns from Northern Daghestan for at least 200 years, and in the Dictionary a very similar word means "man". It is possible that some underlings of Huns took the name of their former masters and lords, but we know little enough about the ethnic composition of the Huns of Attila either. We know only that the majority of the Classical leaders bore Turkish names, and one of them z-Turkish [13], [57]. And observe that, if this dictionary is a record of proto-Magyar, then it is earlier by some 500 years than the oldest Magyar texts known so far. (Go back to Ą4.) What is more, then this is the first record from before the Conquest of the Carpathian Basin, so this could show the stage of Magyar earlier to the Slavic influence. Such a document would be a treasure for linguists and historians but of course after a more details, careful and consequent publication.

            2) Now, do the facts clearly contradict to the existing and accepted theories? For simplicity's sake, let us take the Uralic (or Finno-Ugric) paradigm for "theories". I do not repeat it, because I argued in the Uralic paradigm throughout this study when I did not state definitely otherwise.

            Not, they do not contradict clearly. Mainly, they do not, because the publication is sketchy, we cannot know the details, and at any point there may be a mistype, misprint or bad reading. There is another reason too, but that can wait. However this Dictionary poses difficulties enough for the Uralic paradigm indeed.

            3) OK, then take the Turkish paradigm. And now comes the biggest problem: the Turkish paradigm does not help either.

            Let us accept as working hypothesis that Magyar was originally not an Uralic language, but Turkish. Then obviously Bulgarian r-Turkish. Then very probably the Ob Ugric languages were also not Uralic either. But they could not have been Turkish, Altaic roots are very rare in them. No problem. In Siberia lots of Palaeo-Siberian languages exist; maybe there were even more millenia ago. And language families there like the valleys of the big rivers. The Eastern neighbour is the Yenisei, and we know that a century ago a few kins of today's only Ket existed there. Ket is an isolate in Eurasia now, although possibly the relative to North American Na-Dene languages [58]. Now, imagine that originally a family lived along the Ob too. Then Magyars came and the Ugors learnt proto-Magyar partially. Then we now cannot find the original family, and we mistake the donor Magyar and the small recipients as an Ugric sub-family of Uralic. (And this is really near to the scholarly interpretation of the Ust’-Poluy bronzes [34].)

            So far so good. But in the Turkish paradigm Magyar is originally Turkish. Still for body parts, numerals, some kinship terms & such Magyar has the words whose roots are similar to e.g. Finnish!

            This is possible only if Finns, Lapponians &c. learnt these fundamental words from Magyars too. Maybe then Finns were not yet at the Baltic, but surely they were at least halfway from the Volga to the Baltic. Then Magyars should have been everywhere from the Dnieper to the Ob, below, say, the 51th parallel, in long enough and intimate enough connections with lots of peoples to teach them words as "eye", "heart", "two", "three", "arrow" and "goose". And, finally, these words are not even Turkish in the donor language!

            So indeed: the Turkish paradigm, which at least exists, does not help at all. I cannot imagine the hopeful third paradigm, and Cs. Detre did not suggest any either.

            So I think just now the only hopeful work would be to produce a clear, detailed and well-documented critical edition of the texts. But even from the present one it is possible to draw the conclusion, as I formulated already in Chap. 12, that the language of [2] & [3] could not have been in the evolutionary chain of languages leading to present Magyar from anywhere. In the best case it is a sister language of Magyar, or a half-sister via hybridisation, and the language is now extinct.



            I would like to recapitulate 2 cases from hard Natural Sciences when respectable scientists got interesting results via legitimate and correct observational methods. In both cases the researchers wanted the actual results and from the beginning it was felt that something was wrong; but it took a substantial time to disprove the results.


The atmosphere of Mars

            Since Galileo it was clear that Mars has an atmosphere; but the actual surface pressure was unknown. Science reached the needed accuracy even to hope an answer cca. 100 years ago. Table 13 shows the results of some measurements [59], [60], [61], [62].




Pressure in Hgmm





Earth, from terminator




Earth, telescope




Earth, polarisation




Balloon, polarisation

de Vaucouleurs



Balloon, polarisation




Balloon, telescope

Owen & Kuipers



Earth, telescope










Table 13: Some measurements for Martian surface pressure.


            The first observations were, of course, telescopic, from beyond the terrestrial atmosphere. An old method used the diffuseness of the Martian terminator. It was obvious that smaller Mars would keep more dilute atmosphere. Now, take for example a spectroscopic method. Solar light would be absorbed on, say, oxygen two times, in & out, in the Martian atmosphere, and then one in the terrestrial atmosphere. If the Martian atmosphere is denser than the half of the terrestrial one, Martian absorption dominates and the effect is easily measurable. But it was clear soon that it was not so dense. Then observatories on high peaks in clear atmosphere were looked for, and then balloons. There was an ever decreasing tendency; still values 50-100 Hgmm pressure remained for decades. The values were good for other scientific hopes. With 100 Hgmm pressure fluid water, life &c. are possible, and since probes were in the imaginable future, everybody could keep the hopes for extremely important discoveries.

            On the other hand, everybody knew that Mars' escape velocity is only 5 km/s, while the average velocity of, say, O2 molecules is cca. 0.45 km/s. While this is clearly smaller than the escape velocity, according to evaporation & diffusion calculations it was not enough for 4 billion years, except if very fortunate mechanisms cannot maintain the atmosphere via volcanoes & such.

            Then a terrestrial measurement gave 15 Hgmm when Mariner-4 was already on the way, and then Mariner-4 herself on the spot got some 8 Hgmm. Obviously the small pressure was the difference of two much higher numbers, the measurement of the difference was not stable, and errors generated the erroneously high values; and the astronomers hoped for the higher value and maybe helped the errors, of course not deliberately. They all were good and, of course, honest astronomers. In physical sciences frauds are unheard of, because the results can be checked by anyone, so it is hopeless to falsify. This is the reason that I accept the mere statement of the authors of [3]: one is physicist, one is geologist.


Pluto as asteroid

            Flagstaff Observatory was established and maintained by Percival Lowell for looking for Martian channels, as signals of intelligent life. At the end of XIXth century that was conform with the expectations of Science; and Flagstaff Observatory mapped the channels. However with better and better telescopes observations were less and less reproducible, so in the 20's it was less and less chic to merely map channels. In addition, the founder died in 1916, although the Foundation continued. Then the idea came: there was another favourite idea of Lowell, namely to find the Trans-Neptunian Planet. And there were signals for a trans-Neptunian planet indeed.

            In 1846 Neptune were first predicted from calculations. Namely, the revolution of Uranus was not exactly conform with the predictions of Gravity Theory. Now, it was more hopeful to assume some disturbing factor than to throw away Newton's Gravity Theory being in accordance with all other observations in the Solar System.

            The simplest disturbing factor was one more planet orbiting beyond Uranus, so dimmer, so not seen yet. Two researchers, independently, calculated where must be the planet producing just the exact pull to modify Uranus' revolution to the observed one; they mailed the calculations to 2 astronomers, and one of them found the new planet very near to the predicted place.

            Then in more half a century some observations indicated minor discrepancies in Neptune's motion too. Pickering and Lowell, independently, calculated the data of Planet X. Pickering's results show great variance, but Lowell's ones were much more unequivocal. Finally he got an orbit with average distance cca. 40 Astronomical Units, a position at a fixed time, and a mass 5-7 terrestrial mass.

            The Lowell Foundation employed a young guy called Tombaugh to photographe the neighbourhood of the predicted position. Tombaugh observed excellently, improved the technique originally developed for asteroids and after a substantial time he found something. OK, it was at 6° distance from the predicted position, a rather big discrepancy, but maybe that happened because small differences in distance, so in revolution period can cause position differences linearly growing in time and then there were already some 30 years since the original calculations. For the story see [63].

            Anyways, the planet was there as a fact, and its average distance was 38.8 AU, agreeing quite well with Lowell's calculation. It could not be anything else than Planet X, so it was recorded and named as Pluto, commemorating the initials of Percival Lowell. Flagstaff Observatory restored its scientific prestige, and Tombaugh entered the smallest group of Heroes of Astronomy. Only 3 men discovered planets, because up to Saturn each planet is clearly seen by unaided eye. Uranus was discovered in 1783 by Herschel, Neptune in 1846 by Leverrier, and Pluto in 1930 by him.

            There was only one discrepancy between prediction and observation: the brightness. Pluto was dimmer by some 2.5 magnitudo than Planet X in the prediction. So, probably, it was smaller; and then there was the danger of a smaller mass too. Lowell calculated, as I told, 5-7 terrestrial masses for the mass of Planet X, of course directly from the disturbances of Neptune. However the mass gives an indirect prediction for brightness. Assuming a reasonable albedo and density (e.g. Neptune's ones), from the distance (known already) one gets a brightness. Now, the observed Pluto was some 2.5 magnitudo dimmer than the prediction. That means, caeteris paribus, cca. 30 times smaller mass.

            Now a planet with 0.2 terrestrial mass simply could not generate observable disturbances on Neptune's revolution (gravitational disturbances being roughly linear with mass), so a Pluto of this mass could not be Planet X! However another, more substantial, planet cannot be in the neighbourhood; Tombaugh very carefully observed all the neighbourhood too. So Pluto has to be Planet X, because nothing else can produce the observed disturbances of Neptune.

            For 1940 the problem was clear; the solution was not. An idea was an extremely dark surface, but that was highly improbable: at the very low temperature frozen gases were expected and they reflect quite well. Another desperate idea was an extremely high density. Then the size is smaller, so also the total reflected light. Estimations gave 50 g/cm3 density, ten times Earth's one, necessary to give the observed low brightness. Nobody could imagine the composition of Pluto giving this density (in normal terrestial environment the densest metal, extremely rare osmium, is only 22.4 g/cm3). But it is one thing not to understand something, and it is quite another to doubt facts.

            There were, however, calculations then, having the orbit known, for the mass from the observed disturbances. That is Table 14.


Disturbed Planet

Disturbing Mass (Pluto)/Earth










Table 14: Measurements for Pluto's mass from planetary disturbances in the 30's


So Pluto seemed to disturb Neptune too slightly, but relatively much stronger far Jupiter. Strange enough.

            Then, again, somebody finally was able to get a disk with a big telescope, and got cca. 2400 km diameter. Then the volume is cca. 1/120 terrestrial volume, and with 5 terrestrial mass the density was already much above 100 g/cm3. Then D. Alter (University of Kansas) got the idea that maybe we see the reflection of Sun himself on a smooth reflecting surface of frozen gases. The idea was desperate enough, but the discussion faded away.

            Until 1978, when J. W. Christy started to improve Pluto's orbital data with a 155 cm telescope. With new evaluating techniques it was possible to improve the photograph until a moon of Pluto was recognised. So its distance was measurable. Repeating the photographing several times an orbital period of moon Charon was measured too, and then Gravity theory directly gave the mass: 0.0023 terrestrial mass. (Density is cca. the one expected for frozen gases.)

            So Pluto could not make the disturbances from which its position was  predicted and so being discovered! Then how?

            The present answer seems to be that the disturbances did not exist at all! Neptune revolved quite accordingly to the attraction of the 7 inner planets + Sun. However some observations in XIXth century were slightly erroneous. This was possible because Neptune in 1930 did not yet finish his first total revolution from discovery, and only the first full circle diminishes the observational errors automatically. (Now there is a full circle, however, because somebody found an observation from 1795!) Plus some errors may have happened in the calculations of Pickering & Lowell, still performed without even powerful mechanical calculators. Maybe the very much varying predictions of Pickering indicated that the discrepancies were not real; also Table 14 shows this. Maybe Lowell was more fanatic to force the calculations into a unique scheme (the Planet X paradigm), leading to a unique but fictitious result. I do not know; one should repeat Lowell's calculations from the slightly erroneous observations, and, of course, anybody could imagine a nicer project.

            Then it turned out that there is a second asteroid belt from Pluto outwards. Pluto is the biggest member, but not a major planet at all. Pluto is still on the list of major planets because of 70 year tradition, but it will be left out sooner or later; and instead it will go to the asteroid list more than 5000 long and increasing.

            Pluto was discovered because the Lowell Foundation wanted to discover Planet X very much. Maybe for any roughly ecliptical position sooner or later a member of the second asteroid belt could have been discovered in a neighbourhood of 6°. It is true if the belt has at least 60 members.

            It was known since 1940 that something serious problem was about Pluto. However it was rather surprising that Planet X was a byproduct of random errors. But nobody was responsible. Only: maybe in astronomy it is dangerous to observe something interesting. If nobody had wanted to find Planet X, Pluto would not have been discovered instead. But, of course, it is better to discover asteroid Pluto, than nothing at all.




 [1]       ***: Kitalált Középkor.

 [2]       Cs. Detre: Hun szavak, szövegek.

 [3]       Sz. Bérczi & Cs. Detre: A hunok művészete. TKTE & Uniconstant, Budapest-Püspökladány, 2004.

 [4]       B. Lukács: On the Border of Two Worlds.

 [5]       B. Lukács: The Great 7.

 [6]       This codex is called The Vienna Illustrated Chronicle, but originally it had nothing to do with Vienna. The codex is amply illustrated, indeed. According to some opinions, stronger than guess and weaker than fact, the author is Márk Kálti, from the Royal Chancellery, and the gorgeous book was sent to Paris in 1374 as a dynastic gift. It was surely a prestige product, and the Royal Chancellery supervised the text. The book is in the Országos Széchényi Könyvtár and for a modern text edition see e.g.: I. Szentpétery: Scriptores rerum Hungaricarum tempore ducum regumque stirpis Arpadinae gestarum, Vol. 1, Budapest, 1937, pp. 239-505.

 [7]       J. Sinor: Central Asiatic J. 10, 307 (1965)

 [8]       It is called “Tihanyi Alapítólevél”. The original of the document is in the archive of Tihany (Benedictine) Abbey, Tihany, Hungary, as (of course) Fasc. I, 1. Lots of books reproduce it somehow; the most authentic is: L. Erdélyi: A Pannonhalmi Szent-Benedek-Rend története, Vol. X. Pannonhalma, n.d.

[9]        It is called Regestrum Varadiensis; the manuscript belonged to the Bihar Bishopry. In 1550 it was in so bad states that the Bishop (Georgius) made it printed as: Ritus explorandae veritatis…, Claudiopolis, G. Haffgreff, 1550.

[10]      In the so called Codex Pray, p. 154a, in the Országos Széchényi Könyvtár, as MNy 1. First printing in: J. Sajnovics: Demonstratio idioma Ungarorum et Lapporum idem esse. Tyrnavia, 1770.

[11]      I am looking for the letter of John XXII, but until I find that, I tell the following. To my knowledge the letter was written on 29th Sept, 1329, to Missionary Bishop Bernhardus, in Tbilisi, Georgia. According to the Pope’s knowledge, in that time the Caucasian Magyars (“Ungari Asiatici”, but this term is a self-contradiction, so they must have been Magyars, not Hungarians) lived in the neighbourhood of Bernhardus and were led by some Yeretamir, kin of the old Kings of Hungary (in XIth century). (The name is clearly Turkish.) See also Chap. 13. About the letter see e.g.: L. Tardy: Régi hírünk a világban, Gondolat, Budapest, 1979 & Gy. Gyôrffy: Krónikáink és a magyar ôstörténet, Budapest, 1948

[12]      Pawstos Buzand: Patmutiwn Hayoc. Tiflis, 1912

[13]      We know that Balambér led the crossing at the Don, and that he led the battles against the Eastern Goths; see e.g. I. Bóna: A hunok és nagykirályaik, Corvina, Budapest, 1992. As for the Volga this is an extrapolation. 

[14]      Mowses Kalankatuaci: Patmutiwn Aluanic asxarhi. Tiflis, 1912.

[15]      E. Schütz: Acta Orient. Hung. 12, 139 (1961)

[16]      Márta Vámszer: A kalotaszegi nyelvjárás igeragozási rendszere. Kriterion, Bucharest, 1972

[17]      S. Ono: The Origin of the Japanese Language. Kokusai Bunka Sinkokai, Tokyo, 1970. But the idea comes from earlier; see in: Collected Works of Hashimoto Shinkichi, Iwanami Shoten, Tokyo, 1946-1959

[18]      P. Hajdú: Chrestomathia Samoiedica. Tankönyvkiadó, Budapest, 1968

[19]      G. Bárczi: A magyar nyelv életrajza. Gondolat, Budapest, 1966

[20]      P. Hajdú: Bevezetés az uráli nyelvtudományba. Tankönyvkiadó, Budapest, 1966

[21]      E. A. Halikova: in Arheologicheskie otkrytiya g. 1974. Nauka, Moscow, 1975, p. 178

[22]      E. A. Halikova: in Voprosy ëtnogeneza tyurkoyazychnyh narodov Srednego Povolzh’ya, Kazan’, 1971, p. 64.

[23]      J. A. Halikova: in A Hajdúsági Múzeum Évkönyve I. Hajdúböszörmény, 1973, p. 21

[24]      W. Zimmermann: Die Phylogenie der Pflanzen. Jena, 1959

[25]      A. J. Eames: Morphology of Vascular Plants. R. E. Krieger, Huntington, 1977

[26]      K. Mädgefrau: Die Geschichte der Pflanzen. In: G. Heberer (ed.): Die Evolution der Organismer. Stuttgart, 1967, p. 551

[27]      Sz. Bérczi, Agnes Holba & B. Lukács: In B. Lukács & al. (eds.): Evolution of Extraterrestrial Materials and Structures, Proc. 5th Symp. on Matter Evolution 1994; KFKI-1994-22, p. 147

[28]      Theophylactus Simocatta: Historiae. Ed. C. de Boor, Lipsiae, 1972. (The story is at II.15.)

[29]      Theophanes Confessor: Cosmographia. See e.g.: C. de Boor (ed.): Theophanis Cosmographia, Lipsiae, 1883-85

[30]      Lynn Margulis: Symbiosis and Cell Evolution. Freeman, San Francisco, 1981

[31]      T. Cavalier-Smith: Ann. N.Y. Acad. Sci. 503, 17 (1987)

[32]      Some very, very strict Hungarians try to call the kinship of Uralic and Altaic also "mirageous". However the kinship is commonplace, only now linguists do not put them into the same family. For the kinship I give here B. Collinder: Acta Univ. Upsal. 1, 109 (1965); but you can check it in any good handbook.

[33]      K. V. Sal’nikov: Sov. Étnogr. 5, 118 (1966)

[34]      V. N. Chernecov: Mater. Issledov. Arheolog. SSR 35, 121 (1953)

[35]      I am not joking at all. "Airport" is written at the airport, but city street signs to the airport are Finnish/Swedish "Lentokenttä/Flygstation". The Magyar word is "Repülôtér"=cca. "Flying area". Now, from a vocabulary I was able to learn that "lenttä" is "repül", so "fly", and "kenttä" is "field", but the roots are not genetically related to Magyar and cannot at all interpreted from Magyar. However "Flygstation" is obviously a station for flying.

[36]      P. Kher (ed.): Monumenta Germaniae ex stirpe Carolinorum. Berolii, 1934, Vol. 1, p. 145

[37]      S. Janicsek: BSOAS 5,  15 (1928-30)

[38]      Ibn Rusta: Kitab al-‛alaq al-nafisa, Ed. By M. J. de Goeje. BGA VIII, Leiden, 1892

[39]      V. Barthold: Mem. Acad. Sci. Russ. Cl. Phyl.-hist. 1, 78 (1897)

[40]      I. Fodor: in A. Bartha, K. Czeglédy & A. Róna-Tas (eds.): Magyar Őstörténeti Tanulmánok, Akadémiai Kiadó, Budapest, 1977, p. 79

[41]      M. G. Magomedov: Obrazovanie hazarskogo kaganata. Nauka, Moscow, 1983

[42]      see e.g. G. H. R. von Koenigswald: Die Geschichte des Menschen. Springer, Berlin, 1960.

[43]      V. M. Sarich & A. C. Wilson: Science 158, 1200 (1967)

[44]      Marie-Claire King & A. C. Wilson: Science 188, 107 (1975)

[45]      Elizabeth J. Bruce & Francisco J. Ayala: Nature 276, 264 (1978)

[46]      M. Blanc: La Recherche 15, 655 (1984)

[47]      R. L. Stauffer & al.: J. Heredity 92, 469 (2001)

[48]      D. E. Wildman & al.: PNAS 100, 7181 (2003)

[49]      Y. Haile-Selassie: Nature 412, 178 (2001)

[50]      M. Pickford & B. Senut: C. R. Acad. Sci. Paris II A332, 145 (2001)

[51]      E. Ferenczy: A magyar föld népeinek története a honfoglalásig. Gondolat, Budapest, 1958. The title means "The history of the peoples of the Magyar lands until the Conquest", and indeed gives that concisely, with lots of references.

[52]      J. Harmatta: MTAK 2, 152 (1952)

[53]      L. Schmidt: Geschichte der deutschen Stämme bis zum Ausgang der Völkerwanderung. Die Ostgermanen. 1934.

[54]      Constantine Porphyrogenetus: De administrando imperio. New revised edition by Gy. Moravcsik & R. J. H. Jenkins. Dumbarton Oaks Books, Washington, 1967

[55]      J. Pasternak: in N. Fettich: A honfoglaló magyarság fémművessége. ArchHung Vol. 21, 1937, p. 137

[56]      K. Czeglédy: Nomád népek vándorlása Napkelettűl Napnyugatig. Akadémiai Kiadó, Budapest, 1969.

[57]      Gy. Németh: in Gy. Németh (ed.): Attila és hunjai. Magyar Szemle Társ., Budapest, 1940, p. 217

[58]      M. Ruhlen: PNAS 95, 13994 (1998)

[59]      B. Lukács: Utazások térben, idôben és téridôben. Akadémiai Kiadó, Budapest, 1990

[60]      P. Hédervári, M. Marik & T. Pécsi: A Vénusz és a Mars ostroma. Gondolat, Budapest, 1976

[61]      P. Lowell: Mars. London-Bombay, 1896

[62]      A. Dollfus: Annales d'Astrophysique. 28, 4 (1965)

[63]      C. Tombaugh & P. Moore: Out of the Darkness: The Planet Pluto. Stackpoole Booke, Harrisburg Pa, 1980






My HomePage, with some other studies, if you are curious.