Selected Writings: West and East in the Unity of the Ancient World 9789634726197

Table of contents :
Preface
Orientalia
Der alte Orient und das Klassische Altertum
The Myth of the Stone Heaven in Eurasia
The Rise of the Old Persian Empire. Cyrus the Great
Die Schrift bei den antiken Steppenvolkern
The Origin of the Name ΣΟΓΔΟΙ
Alexander the Great in Central Asia
The Wall of Alexander the Great and the Limes Sasanicus
The Origin of the Huns
L'apparition des Huns en Furope Oriental
La Coupe de Xusro et I' origine de la legende du Graal
Les source iraniennes de la legende du Gral
Graeca
The Prehistory of the Greek Language
Zu den kleinasiatischen Beziehungen der griechischen Mythologie
Zur Ahhiyawa-Frage
Ahhiyawa Names-Mycenaean Names
Wort und Begriff λαός von der mykenischen Zeit bis ins 5. Jahrhundert v. u. Z.
Un peuple finno-ougrien dans la tradition litteraire de I' Antiquite
Sur l'origine du mythe des Hyperboreens
Rapporti tra Grecia e Bacino Carpatico tra VI e V secolo a. C
Herodot und die altpersische Novelle
Herodotus, Historian of the Cimmerians and the Scythians
Prolegomena to the Libykoi Logoi of Herodotus' "Histories''
Poseidonios uber die romische Urgeschichte
Pannonica
Landed Property in Early Roman Pannonia
Goten und Hunnen in Pannonien
Fragments of Wulfila's Gothic Translation of the New Testament
Hungarica
Remarques sur le lexique du latin medieval et le substrat hongrois
Erudition, tradition orale et realite geographique (Le recit sur l'exode des Hongrois chez Anonyme)
Altgriechische soziale Typenbegriffe in der ungarischen Sprache
Varia
Zum Illyrischen
Agri vacui et solitudo (Zu Tac., Ann. XIII 53 ff.)
Historische Vergangenheit und Identitatsbewusstsein

Citation preview

Ar A 0 A

ArA0A STUDIA AD PHILOLOGIAM CLASSICAlvl PERTINANTIA QUAE IN AEDIBUS UNIVERSITATIS DEBRECENIENSIS REDIGUNTUR XII. SERIEM EDE1\TDAM CURAT LADISLAUS HAVAS

JOHANNES HARMATTA: Scripta selecta

XII.

DEBRECENI EGYETEM BOLCSESZETTUDOMANYI KAR Klasszi ka~fi lol6giai Tanszek I

-

Janos

Harmatta

SELECTED WRITINGS WEST AND EAST IN THE UNITY OF THE ANCIENT WORLD

HOC VOLUMEN REDEGERUNT Ladislaus Havas et Emericus Tegyey

Edited by

Laszlo Havas & lnire Tegyey

Debrecini Anno MMII

Debrecen, 2002

I CONTENTS

CONTENTS ........................................................................................................... 5 PREF ACE .................................................................·.............................................. 7 PART I ORIENTALIA ......................................................................................... 9 Der alte Orient und das Klassische Altertum .......................... :........................ 11 The Myth of the Stone Heaven in Eurasia .......... :.............................................. 19 The Rise of the Old Persian Empire. Cyrus the Great ....................................... 27 . Die Schrift bei den antiken Steppenvolkern ..................................... :................ 40 TheOriginoftheNameior~o, ............................................ :........................ SI· Alexander the Great in Central Asia ............................................ ;.................... 56 The Wall of Alexander the Great and the Limes Sasanicus ... ;... ... ... .... ... ..... .... 64 The Origin of the Huns ....................................................................·................... 74 L'apparition des Huns en Furope Oriental ........................................................ 88 La Coupe de Xusro et l 'origine de la legende du Graal .. .. ......................... ....... 94 Les source~ iraniennes de la legende. du Gral ......... ..... ...... .............. ..... .. ......... 105 PART II GRAECA ........................................................................................... The Prehistory of the Greek Language . .. .. . .. .. .... .. .. ... .. .. .... ... .. .. .. .. .. .... .. .. .. .. .. .. .. Zu den kleinasiatischen Beziehungen der griechischen Mythologie .............. Zur Abbiya,va-Frage ....................................................................................... Ag.g.iyawa Names-Mycenaean Names ............................................................ Wo11 und Begriff Aa6s von der mykenischen Zeit bis ins 5. Jahrhundert v. u. Z. Un peuple finno-ougrien dans la tradition litteraire de I' Antiquite ................. Sur l'origine du mythe des Hyperboreens ....................................................... Rapporti tra Grecia e Bacino Carpatico tra VI e V secolo a. C ....................... Herodot und die altpersische Novelle ...... .... ......... ....... .... ... ... .. ......... ... .. ... .... .. Herodotus, Historian of the Cimmerians and the Scythians .. .. .... •... .. .. .. .... .. ... Prolegomena to the Libykoi Logoi of Herodotus' "Histories'' ......... ;.............. Poseidonios Uber die romische Urgeschichte .................. ., ..............................

·111 113 120 140 149

158 165 176 184 192 207 217 222

PART III P ANNONICA .................................................... ;............................ 229 Landed Property in Early Roman Pannonia ................................ :................. ;. 231

5

Landed Property in Late Roman Pannonia ............ .. ... .... ........ ................. ........ 240 Goten und Hunn en in Pannonien ..... ... .......... .. ... ..... .. .. ........ .... ..... ....... .. ..... ... ... 245 Fragments ofWulfila's Gothic Translation of the New Testament fron1 Hacs-Bendekpuszta ........................................................................ :.............. 249 PART IV HUNGARICA ................................................................................. -275 Rerilarques sur le lexique du latin medieval et le substrat hongrois ....... .. ..... . 277 Erudition, tradition orale et realite geographique (Le recit sur l'exode des Hongrois chez Anonyme) ... .. .. ... ... ... .... ..... .. .. .. ... .. .. ... ... .... ... .... .. .... ........ ... .... 286 Altgriechische soziale Typenbegriffe in der ungarischen Sprache ................. 304 PART V VARIA ............................................................ ·................................... Zurn Illyrischen ..... ......... .. ... ..... ... ...... .. ... .... ....... .. .. ..... .. ..... ... .. ............. ... ... ...... Agri vacui et solitudo (Zu Tac., Ann. XIII 53 ff.) ........................................... Historische Vergangenheit und Identitatsbewusstsein .~ .............................. ~ ...

313 315 322 335

ACKNOWLEDGMENTS ................................................................. -.............. 346

PREFACE

Debrecen University Press fulfils an old wish of the Hungarian Study of Antiquity and Linguistics by publishing the shorter writings of Janos Hannatta, a leading figure of Inda-European and Iranian Studies in Hungary. Taking into account the disGiplines taught at the Philosophical Faculty of the University of Debrecen we have collected ai1icles from the rich oeuvre of Janos Harmatta which are in some way connected with the author's central issues, i.e. the language and literature of the ancient Greeks and those, mostly Iranian people who vt1ere in contact with the Greeks·; To these we added articles dealing with different issues of the history of Pannonia, of the Hungarian language and culture and with several subjects of general interest. Wehad to omit some imp01tant articles on the more remote peoples and languages of the East. The editors' work was made especially difficult, as the articles we could · not include, being written by an author of such wide-ranging insight, contain several observations and many references that are relevai"lt to our field. The editors are grateful to Debrecen University Press for accepting these seminal articles for publication and to Professor Janos Harmatta for his valuable councel and for his help during the proof-reading process.

Laszlo Havas

Imre Tegyey

6 7

DER AL TE ORIENT UND DAS K.LASSISCHE ALTERTUM Die Frage der historischen Beziehungen zwischen dem Alten. Orient und dem Klassischen Alte11Um beschaftigt schon seit langer Zeit her die wissenschaftliche Forschung. Schon die antiken Schriftsteller erkannten die politische und kulturelle Bedeutung der orientalischen Staaten ftir die griechisch-romische Welt, und sie liberliefe1ien in ihren Werken zahlreiche Angaben Uber die Beziehungen der klassischen Antike zu dem Orient. Das antike Quellenmaterial beeinfluBte weitgehend auch Pliny Pterophoros appellata ·regio ), because there can be hardly any doubt that this Greek name goes back to his epic, written in Greek language. Consequently, by the mediation of the Scythians, Aristeas disposed of rather precise infon11ations c?ncerning the myth of "the hovering heaven" and about their ideas connected with the ]and of the Arctic Circle. Fortunately, Pliny the Eld.er had preserved his report for us. The similar concepts of the Palaeo-Asiatic tribes, but perhaps even the ideas of the Finno-Ugrian peoples also arrived at China at a very early date. According to the kind information by Professor Ma Yong, from a grave, dated to the Illrd century B.C., came to light the literary work entitled Mu-t'ien-ts 'i chuan "History-of Emperor Mu", which has been written after 970 B.C. This Emperor Mu was identified by Ma Yong with Mu wang; king of Chou, ruling from 1001 to 946 B.C. According to this romantic story, Emperor Mu made a great travel to the north-west. He started from Lo-yang and arrived through the Kun-lun (Altai mountains) up to the Black river(= Irtish) and in the course of this journey he gathered a load of 100 ca1is, consisting of feathers, shed by the birds. Another Pre-Ts 'in work, the Shan hai Jing, mentions "the land of the oneeyed men", lying far-away in the north. Apparently, this report reflects the myth of the Arimaspi, while the Chuang-ts 'i infonns of "the land of the bald-headed men", which obviously corresponds to the 'OpyEµ1ra101, described by Herodotus, but probably 23 mentioned in the Arimaspeia of Aristeas as well. Thus, we can state that the knowledge of all these mythical peoples and the mythical world concept created by the Finno-Ugrian peoples arrived at China almost at the same time, as they became described in the epic of Aristeas. As we can see, the idea of "the hovering stone heaven" fits a broad context of the history of civilization. On the basis of the evidence, discussed above, there can be hardly any doubt that the. archaic Finno-Ugrian, Palaeo-Asiatic and the still earlier 23

I thank very much Professor MA YONG for his informations.

25

Ind~-European myths permit us to presume the existence of a myth involving "the firm stone heaven" hovering over the earth and moving up and down. Obviously, this concept of a "hovering stone heaven" belongs to a very a1;chaic- layer of Finno"'Ugrian mythology and religion and surely it is connected with the similar lndo;..lranian ideas. Therefore, it seems to be very likely that Hungarian eg, Finnish sati, Siryan sjngd, going back t~ a Proto-Finno-Ugrian form *sdJJe, represents a borrowing from a Northern lr;iian *sang; 'stone, stone heaven' to be presumed on the basis of the passage, quoted above from th~ Dadastan zdenzg.

THE RISE OF THE OLD PERSIAN EMPIRE CYRUS THE GREAT I The foundation of the Old Persian Empire was undoubtedly an· important link in the development of the ancient wo_rld.- It happened for the first time in the history of humanity that the huge territory stretching from Libya to the Pamir, from Sudan to the Syr Darya was incorporated into a relative unity by an empire, which was practically a historical summing up of the results achieved by the earlier Near Eastern states in social-economic, technical anclcultural fields~ and which with its two centuries long existence exerted_ a significanf influepce on the further historical development. Although the historical importance of the Old Persian Empire was obvious-since long time to historical research, the problem-of its social-economic structure remained surprisingly· unclarified up to the latesttimes.· A few years ago in scientific circles it was still a generally spread opinion that tpe Old Persian Empire was a slave-holding state, and the major part of its poplllatio11 ~e11~iste~ of slave_s. In contrast to this opinion, represented by such eminent scientists like for."·example M; D'yakonov, F. Altheim, M. A. Dan1 damaev, W. B. Henning, ·Gershevitch, Yu.B. Yusifov, W. Hinz and LAliev, as from 1961 in severar lechlres and publications, and especially in my study entitled «The Problem of Slavery in the Old Persian Empire»2 I tried to show that the above mentioned historical conception is based on an erroneous interpretation of the social terminology of the Old Persian, Elamite, Babylonian, Aramaic and Greek sources, and that the relations of production in the Old Persian Empire do riot reflect the conditions of the slave-holding social formation, but they represent the conditions of the Asiatic mode of production. Since then this opinion has found a rather general acceptance, and today it seems already to be doubtless that in the Old Persian Empire, beside the back. ward smaller peoples living still in tribal society and the slave-holding Greek citystates, the largest part of the population were half-free producers living on the "royal land".

L

L

1

I. M. D'YAKONOV: VDI 1959/4. 70 ff.; F. ALTHEIM-R. STIEHL: Die aramaische Sprache unter den Ac-hameniden. Frankfurt am Main, 170 ff.; M. A. DANDA~1AYEV: Foreign Slaves on the Estates of the Achaemenid Kings and Their Nobles. Moscow 1960; I. GERSHEVITCH: AM NS 2 ( 1951) 13 9 ff. with interpretations by W. B. HENNING, Yu. B. YUSUFOV: VDI 1961/4, 32 ff.; \V. HINZ: ZDMG 110 (1961) 24 7; I. ALIYEV: HcTopllil MHJlHH. I. Bairn 1960, 281. 2 Neue Beitrage zur Geschichte der Alten Welt. I. Alter Odent und Griechenland. Berlin 1964, 3 ff.

26-

27

If in the clarification of the social-economic structure of the Old Persian Empire, for which we have a fairly rich stock of sources, historical research could step ahead only in recent years, then it is not at all striking that in the investigation of the historical circumstances of its rise an even greater backwardness can be observed. The reason for this is partly the fact that the sources (especially the Greek and Babylonian sources) are rather scanty and one-sided. Thus, taking a look at the historical literature of the last few decades, we can state that since the thorough study of F. H. WeiJ3bach 3 published on Cyrus in 1924 there has been hardly any progress. Investigation has stuck up in the clarification of the single events and chronology, and almost completely_ disregarded the internal historical motives of the development of the Old Persian Em4 pire and the appearance of Cyrus. In fact it did not even raise this question. Just there..: fore, ifwe want to promote historical research, then we must examine in the first place, what were the historical circumstances of the rise of the Old Persian· Empire, what factors rendered its development possible or necessary and in what it differed, from the viewpoint of historical development, from the earlier Iranian state, the Median Empire. The rise of the Old Persian Empir~ is closely connected with the name of Cyrus the Great (or II), and essentially it coincides with his reign lasting from 558 to 529 B.C., inasmuch as the process of its coming into existence can be regarded in many respects as closed down with the death of Cyrus, which took place in 529, that is~2500 years ago. The problems of this period comprising nearly three decades were divided in the earlier historical investigations as follows: 1. the origin· of Cyrus, 2 .. his uprising and victory over the Medes, 3. his campaign against Lydia, 4. the conquest of Eastern Iran, 5. occupation of Babylon and Mesopotamia, 6. struggle against the nomads of Northern Iran.

II Among these groups of ques~ions the problems of the origin of Cyrus, his uprising and his victory over the Medes deserve special attention. It can be stated from the sources written in cuneiform script with doubtless surety that Cyrus was the great-'grandson of Cispis, «the great king, king of Ansan», from whom through· the other branch of the 5 Achaemenides also Darius descended. The father of Cispis, Haxamanis was the· founder of the dynasty, who was obviously the first to adopt the title of a king. These data of the Old Persian inscriptions and the clay cylinder of Cyrus are supp011ed by the · fragment of an inscription of Assurbanapli, which mentions in 639 the grandfather of

. Cyrus of a similar name, «Kuras, the king of the country Parsumas». 6 Thus Cyrus came of an old ruling dynasty, which could found its kingdom· about 700 B.C., about the . same time as the ruling dynasty of the Med es. The Greek sources~ first of all the report of Herodotus (I. 107 ff.) also confirm this result inasmuch as they mention Cambyses as the father of Cyrus, and they also mention Cispis-Teispes as the son of HaxamanisAchaimenes (Herodotus VII. 11 ). Otherwise, however, the legend of origin of Cyrus told by Herodotus does not call Cambyses a king, nor does it mention that he would have been member ofan old royal dynasty. This peculiar interpretation of the legend of origin of Cyrus is easily understood, inasmuch as deemed on the basis of the names and the word spaka 'dog' occurring in the legend, it was of Median origin, and the whole was inspired by Median aspect: according to it Cyrus was the grandson of the last Median king and ascended the throne with the help of the Medes. This presentation obviously glosses over the events from Median view point and keeps silent about the fact that the dynasty of the Achaemenids was at leac;t as old as that of the Median De~ iocids. Of course, that much can be true from the legend that Cyrus 'stood in some sort of family relations with the overthrown Median royal family. According to Ctesias, Cyrus married the daughter ·of the last Median king.This is a step that.can be observed in numerous cases in connection with the coming to power of new dynasties. 7 Thus the origin of Cyrus can be clarified fairly well from historical view point. Ifin this respect still various hypotheses were raised in historical research,8 the reason for it ·can be looked for mainly in the circumstance that the. names of Cyrus and his predecessors, Cambyses and Teispes, create a strange impression among the names of the Achaemenian·rulers. In fact they could not be explained from Iranian and thus they can easily induce anybody to the conclusion that they are of foreign, non-Iranian, origin. In this present state of investigation the thorough examination of the names of Cyrus, Cambyses and Teispes is, therefore, an actual task. The name of Cyrus (Old Persian Kurus) was recently discussed by W. Eilers in detail in· a very thorough study. In this he convincingly pointed out that none of the ex9 planations given so far is acceptable. Eilers himself could not find or point out such an Iranian word either, from which the name Kun& could be explained.·. However, that much could at any rate be rendered likely by him that the name Kurus is connected with the Old Indian name Kuru- and that this connection is parallel with the correspon~ dence of the O Id Persian Kamb iijiya- (name of the father of Cyrus) and the Old Indian name Kamboja-. This helplessness· of investigation in connection with the explanation of the name Kurus is on the whole surprising, because in the Iranian languages there · 6

Cf. E. F. WEIDNER: AfO 7{I931-1932) I ff. On some similar cases cf. J. HAR.MATTA: Acta Ant. Hung.17 (1969) 404 ff. 8 See H. ZIMMER: Altindisches Leben. Berlin 1879, 102; F. C. ANDREAS: Verhandlungen des XIII. lnternationalen Orientalistenkongresses 1902 in Leiden. 1904, 93 ff., etc. 9 . .. W. EILERS; Kyros. BzN 15 (1964) 180-236. 7

3

F.H. WEISSBACH: ,,Kyros". PWRE SpBd. IV. Stuttgart 1924, 1128 ff. Cf e. gr. A. T. OLMSTEAD: The History of the Persian Empire. Chicago 1948, 34 ff 5 Cyrus, Cy!, 1. 25 (F. H. WEISSBACH: Die Keilinschriften der Achameniden. Leipzig 191 l, 4-5). 4

28

29

10

exists a well-known root, from which it can be explained satisfactorily. This is the root *kur- 'be born', the continuation and developments of which can be pointed out from Old Iranian, as well as from Middle Iranian and from New Iranian. From Old Iranian we can quote the Scythian name *Kula-x§aya- 'young king'. From Middle· Iranian we can refer to the following data: Saka kula- 'kinsman', Sogdian wkwr, wk'wr 'kinsman', Bactrian or Hephtalite kula- in the name Mihirakula- 'Mihir's begotten';Pahlavi kwlk, kwlg 'young of animal'. In New Iranian we find the following continuations of the root: Ossetic igurun 'be born\ Kurdish kurr, kur, New Persian korre 'young of 11 horse etc.'. The quoted data testify the earlier existence of the verbal roqt kur- on the one hand and of the three nominal derivatives kuru-, kura-, *kurnaka-, on the other hand. As regards the way of formation the most archaic of these is *kuru-, 12 while the forms *kura- and *kurnaka- were formed with formative syllables still productive in Old Iranian and partly in Middle Iranian. Accordingly, the meaning of the name Kuru- could be 'young, child, youth', and can reflect an aspect simi.lar to the Scythian name *Kula-xfaya- 'young king'. In the myth of origin of the Scythians and also in the name *Kula-xsaya,; the matriarchal aspect of ultimogeniture is reflected, and thus· it is not impossible that the name Kuruis the remnant of earlier matriarchate also with the Persians. Semantically the people's name Kuru- 'young' can best be explained by refen-ing to the Scythian legend of ori:. giri, according to which the Sc}1hians regarded themselves as the youngest people in the world.

Ill Si1nilarly to the name Kuru-, the name Kambiijiya-~ name of the father and elder son of Cyrus, also withstands to the attempts of explanation. This was also discussed in detail 13 recently by W. Eilers. He pointed out the unsatisfactory character of the explanations given so far, however, without being able to give a better one. At any rate he deems to be doubtless that this name is in connection with the Old Indian tribal and country name Ka,11boja-. He also sees it well that this relation between the two names would demand that the Old Persian form should be reconstructed as *Kambaujiya-. However 10

As it . was~lrea dy pomtcd · out by V. I. ABAYEV: K JTHl\IOJJonm .llpem-renepc1rncK11x nMeH Kurus, Kambujiya, Cispis. 3TH~IOJ1onrn 1965. ~foscow 1967, 286 ff The interpretations by ABAYEV of the names Kambujiya and Ciipis are, however. unacceptable. 11 For the Middle and New lrai7ian data .sec H. \V. BAILEY: .TRAS 1953, 114 ff.; I. GERSHEVITCH: tSOAS 14 (1952) 493 ff.; H. W. BAILEY: TPhS 1953, 39 ff. Cf. also Ossetic kur 'bull-calf'. On the suffix -u- cf. J. WACKERNAGEL-A. DEBRUNNER: Altindische Grammatik. II/2. Gottingen 1954, 463 ff. 13

W. EILERS: BzN 16 (1964) 210-213.

on account of the Greek transliteration (au = ypsilon, j = sigma?) he holds also this problematical. But the Greek transliteration does not mean an insuperable difficultv. Also in the Greek transliteration of Zranka- sigma corresponds to the z, and in tl;e Greek t.ransliteration of haumavarga- the sound-group ava is rendered by ypsilon. If, however, we start out from the forms *kanibaujva-, *kambauja-, then not only the analysis kam-bauja-~ kam-baujya- of these is possible: but we can also regard these as compounds consisting of the elements kamb-azifa-, kamb-azifya-. In this case we get as first element the word *kamb- (eventually *kamba-) 'little, small\ and as second component the word *aujah- 'strength'. Thus the meaning of the whole compound will be 'of small strength, weak:. It is a question, whether a personal name of this kind can be presumed in the stock of Old Persian names: but it can be easily understood as an apotropaic name. In this case, of course, it has still to be regarded as the survivai of an archaic custom of giving names: and that it could really he that: is shown by the circumstance that after Cambyses II it was no longer used by the Achaemenids. The people's name kambazifa- 'weak, powerless' belongs to the category of people's names expressing negative valuation, and as such, it cannot be, of course, the name of the kamboja~s used by themselves, but can originate from one of their hostile neighbours; This is suppmied also by the circumstance that in Iranian and Greek sources no trace of the people's name kambauja- can be found, although if it has been the name of the people used by itself, then it ought to have spread in a broader circle. From the aforesaid it follows that in spite of the linguistic identity we cannot presume any historical relation between the personal name Kambaujya- and the people's name kambavJ°a-. Even otherwise a relation of such character could be presumed at the most in the case of Cambyses II, when as a result of the conquests of Cyrus the East Iranian tribes adjacent to the Indian territory also came within,Jhe horizon or under the rule of the Achaemenids. Howeve~, about 640 B.C., at the time of the birth of Cambyses I, this 14 could hardly be the case. Linguistic research is even more helpless in connection with the name Cispis (ca-i-fo-pa-i-sa) or Caispis. This can be the only explanation for the fact that even excellent scientists looked in it for the ijurrian name of god Tefop, or the Urartian name of god Teiseba and it was brought into connection even with the Cimmerian name Te15 us-pa-a. The psychological basis of these attempts was concisely summed up by W. 16 Eilers in his statement that the name Cispis by the king in his court. Exactly these are the ones, who later on turn away from Astyages first. The characteristic element of the Median court etiquette was the proscynesis, and that of jurisdiction was interrogation by torture? 24 In the case of the data of Herodotus, of course, it is possible that they reflect already the contemporary conditions of the Old Persian Empire to a certain extent. In ~ fact, the picture given above by the scanty data is fairly near to that one, which can be 21

Cf. S. LANGDON: Neubabylonische Konigsinschriften. Leipzig 1912, 220 (inscription no. I of ~abuna'id, I. 27). Cf. also I. M. D'YAKONOV: l1crnpm1 M11.zum, 336. ~~ F. H. WEISSBACH: Die Keilinschriften der Acharneniden. Leipzig 1911, 4. ;: Ancient Near Eastern Textes2• Ed. by J.B. PRITCHARD. Princeton 1965, 305. On the proskynesis see F. ALTHEIM: Proskynesis. Paideia 1950/5. 307 ft~

fact, the picture given above by the scanty data is fairly near to that one, which can be re~o~structed from the reports 1:ero~otu~ reg~ding the ·old Persian Empire. Partly . this 1s exactly the reason, why m histoncal mvestigation we meet frequently with such assumption, according to which the social and economic structure and state organization of_the Old Persian Empire ~id not differ at all from those of the Median Empire. The evidence of the data of Median reference preserved in the Old Persian inscriptions and other contemporary cuneiform sources, and Aramaic documents become in this 25 point of decisive importance. The terminology to be regarded of Median origin on the 26 basis of phonetic criteria from the viewpoint of subjects can be grouped as follows:

?f

_

1. ruling class: xsayaBiya- 'king'~ xsayaBiya xsaya8iyanam 'king of kings' xsoyaBiya vazraka 'great king', *vis-pu8ra- 'royal prince', *vois(a)-pu8ra- 'aristocrat', azota- 'nobleman, freeman' ,farnah- 'royal splendor', pati-zbay- 'to proclaim'. 2. state organization: *pati-xsiiya8ya 'majordomo', *hazahra-pati- 'chiliarch', ~sata-pati- 'centurion', *dasa-pati- 'decurion', *pasca-dasapati- 'vice-decurion', *ganza- 'store', *ganza-bara- 'store-keeper', vispa-zana- 'being from each tribe', paru-zana- 'being from many tribes'. 3. army: spii8maida- 'mobilized army, army in camp', uvaspa- 'having good horse'. 4. religion: magu- 'magus', zfira-kara- 'evil-doer', asman- 'sky'. 5. administration of justice: ufrastam prs- 'punish'. 6. craft (architecture): asan- 'stone', kasaka- 'semi-precious stone'. Let us examine now the historical lessons of this terminology of Median origin by comparing it partly with the Greek data and partly with the characteristic Old Persian terminology. The adoption of the royal titles points to the circumstance that the position of the Median king and the character of the Median kingdom differed as much from those of the Old Persian kingdom that after the overthrow of the Median Empire Cyrus did not keep the old Persian royal title. In a characteristic way Median *xsiiyaBya- made its way even into the title xsaya8iya Parsaiy »king in Parsa, although the original title of the Persian kings could have remained, since it is doubtless that before Cyrus the Great, already Cispis, Cyrus I and Cambyses I had also worn ti:le of a king, which, however, could hardly by xsaya8iya-. We can perhaps think of the word 25

Regarding these cf. R. G. KENT: Old Per~ian Grammar. Texts. Lexicon. 2 New Haven 1953, 8; W. BRANDENSTEIN-M. MAYR.HOFER: Handbuch des Altpersischen. Wiesbaden 1964. 12 ff.; E. BENVENISTE: Titres et names propres en iranien ancien. Paris 1966, 26 ff.;.M. MAYRH0FER: Die Rekonstruktion des Medischen. Graz-Wien-Koln 1968. 26 The attempt by I. GERSHEVITCH: TPhS 1964, I ff. to cancel a part of the phonetic criteria considered Median, cannot be regarded as convincing, cf. M. MAYR.HOFER: Die Rekonstruktion des Medischen, 8' ff: Concen'ling the Median character of the data quoted in the followings see the scholarly literature cited in note 25.

35 34

xsaya- as the Old Persian royal title, which in the name Xsayarsan- occurs in Persian also later on. 27 It is interesting on the other hand that the phrases *mana pasni- 'queen' (as a matter of fact 'lady of the house') and dux9 i- 'lady' are characteristically Persian. This shows that originally the Achaemenids had their own characteristic Persian ruler's terminology, which was of a rather archaic character, and on the basis of the above mentioned two words essentially it corresponded still to the conditions of clan or tribal society. This is in harmony with those Greek data, which regard the social differentiation of the Persians in the period of Cyrus to be still much smaller than that of the Medes. The lesson of the words vis-pu8ra- and vllisa-puBra-, as well as azata- corresponds to this fact, inasmuch as these show that the social position of the great aristocratic clans and families took already a much more definite shape with the Medes than with the Persians, and the class of the freemen was already clearly separated evidently first of all from the class of.the half-free. If we interpret the data of Herodotus cor28 rectly, than the title fratama- «first» too was already known in the Median royal court. Thus, as a whole, the data discussed point to the circumstance that the class differences of the Median society in the period of Cyrus were considerably greater than those of the Persians. The number of adopted Median terms is strikingly high also in the field of · state organization. This shows that the Median Empire had a highly developed bureaucratic organization, which later on was adopted also by the Achaemenids. In this tespect becomes important the report of Herod.otus, according to which the system of kataskopoi and katekooi was introduced in Media already by Deiokes. These two Greek terms very likely render the Median words *spasaka and *gausaka-. It is note.: , worthy that besides the kataskopoi «the eye of the king» appears as a separate office, in which we can see the equivalent of Median *patyaxsa-. Thus we must reckon with several categories of observers and supervisors. This explains the fact that already W. 29 Eilers could point out several such terms, which from the semantic viewpoint belong to this circle. Among these offices, «the eye of the king» could be denoted by the term *patyaxsa-, and this was the name of a high office. Beside him, however, the 27

Thus explained by R. G. KENT: Old Persian,2 182; W. BRANDENSTEIN-M .. MAYRHOFER: Handbuch des Altspersischen, 126. Recently W. EILERS: BzN 15 ( 1964) 181, note l supposes a form xsayat(i.e. a participle with the suffix -nt) as the first part of the compound. This presumption is possible but it cannot be regarded as necessary, because according to the testimony of Avestan a participle of the type xsaya- could also occur as the first member of compounds, see J. DUCHESNE-GUILLEMIN: Les composes de l 'Avesta. Liege-Paris 1936, 198 ff. The question cannot be decided on the basis of the Old Persian alone and the Avestan evidence permits both explanations. 28 On this title cf. W. EILERS: ZfA 17/51 (1955) 226 ff. 29 W. EILERS: Iranische Beamtennamen in der keilschriftlichen 'Oberliefe1 ung. I. Leipzig 1940, 23, 26, 119 ff.

36

*spasaka-s acted in large numbers, also in the villages. According to the testimony of the PFT- at least in the Old Persian Empire - there were also economic controllers who bore the denomination ti-ti-ya-kas < didayaka- 'inspector'. Much later; even in th~ Parthian documents ofNisa, the derivation 'wpdyt < Old Iranian *upa-dita- of the verb day- 'to see' is used for economic supervision. The Iranian (eventua11y Median) prototypes of the other occupations or offices mentioned by Herodotus can also be pointed out, viz.: masons - *raza-, spear-:bearer - *rstibara-, doorkeeper - *dvara-pana-, messengers - azdakara-. The report of Herodotus regarding the administration in writing . of the Median royal court is interesting and deserves further investigation .. Here the question can be raised, what script or written language was used by the Medes. 30 It is not worth while to enter into conjectures, but at any rate it can hardly be doubted that the state administration used some kind of a written language. Unlike the state organization, the number of the Median phrases relating to the · army is strikingly sinall in Old Persian. According to Herodotus the division of the army into branches.ofservice was the innovation of the Medes. Thus we could expect that the denominations of these were adopted also by the Persians. Among the three branches· of service, arstibara- 'spear-bearer' and Banuvaniya-. 'archer' could be also Median, but asabara- 'cavalry~man' is clearly of Old Persian origin. The phrases usabari- 'camel borne', va9abara- 'axe-bearer', *var9abara- 'shield-bearer' ,31 ma0ista- 'general' are similarly of Old Persian origin, whilepasti- 'foot soldier' could also be of Median origin, but this cannot be proved. It is especially significant that the Median word spada- 'army' was not adopted by Old Persian, but instead of this the word kiira- ·remained in use, which seems to be a more archaic phrase (cf. Pashto kor 'house, family', Kurdish kar 'fan1ily'). All this shows that the organization of the Persian army was either independent from the Median c-ne, or in comparison to the Median army organization Cyrus introduced such significant changes and innovations, which rendered the Old Persian army and tactics higher developed than those of the Medes and made the adoption of the Median military terms unnecessary. One important tenn seems to be of Median origin, viz. spa8maida- 'mobilized army or aimy in camp'. This could be adopted by Old Persian, when the Persian kings still as vassals had to join the Median «mobilized army», i.e. it was a special Median bearing of taking the field by the Persian army.

30

As it is well-known, according to the assumption of many scholars the Old Persian cuneiform writing is of Median origin, cf. recently L M. D'YAK0NOV: The Origin of the 'Old Persian' Writing System and the Ancient Oriental Epigraphic and Annalistic Traditions. W. B. HEl\TNING Memorial Volume, 98 ff. For the opposite view cf. recentlyW. HINZ: Neue Wege im Altpersischen. Gottingen 1970, I ff. 31 See J. HAR.MATTA: Ant. Tan. 11 (1964) 24. The objections by F. ALTHEIM-R. STIEHL: Geschichte der .Araben. II. Frankftut am Main 1965, 566 are based on simple misunderstandings.

37

The terms adopted in the field of religion are in hannony with the report of Herodotus on the impmtant role of the Magi in the Median royal court, while the Median material background of the term ufrastam prs- is weil illustrated by the interrogation by torture of the herdsman Mitradates. All this underlines in a-plastic way the difference of the social and economic structure and state administration of the Median Empire and the Persian kingdom before Cyrus the Great. However, it still does not give a reply on the reasons of the fall of the Median Empire and the characteristics of · the internal structure of the Old Persian Empire. Was the fall of !he Deiocid dynasty really caused by the accidental disunion of the Median aristocracy, and did the Old Persian Empire not, at all differ from the previous Median state?"

VI For the answer of this question we must still examine a passage of the Bisitun inscription. This is the much disputed text, in which Darius compares his own measures with those of Gaumata, the Magus, viz.: DB I 64--65 adam: niya9arayam: karahya: abicaris: gaiBamca: maniyamca: 32 Old Persian vi 0abisaca Elamite UDU-tas a-ak zjz a-ak Tkur-tas a-ak I>- UL.Hz1g rmar-ri' u-ma ap pi-li-ia33 «I entrusted to the military people the fields and the live-stock and the domestic people and whatever (in the fields) is the property of the royal house» The social and economic policy of the Old Persian ruler and that of the Magus are put in a sharp contrast. Darius undoubtedly restored the state of affairs which had existed during Cyrus and Cambyses, while in the case of the Magus we can think that with his measures he tried to restore the conditions before Cyrus and Cambyses. It is at any rate doubtless that Gaumata separated himself from the public just like the Median rulers did. It is also quite sure that the term niya9arayam is Old Persian, and not of Median origin. Thus it renders likely the Persian origin of one of the institutions of decisive importance in the Old Persian Empire. The major part of land was royal- property in the

Old Persian Empire. This condition arose as a result of the conquests of Cyrus. The social and economic policy of the Old Persian rulers was that pait of the «royal land» was under their direct administration, but a considerable prut of ft (the baga-s, estates) was entrusted to the Persian and Median military aristocracy - the technical term of this is 34 the verb ni-9iiraya- -, against certain obligations and payment of tax. The accessories of these estates were the cultivated fields, the live•stock, the working domestic people and in general everything, what on them was property of the royal house. This is the explanation for this much disputed passage of the Bisitun inscription. Gaumata obviously wanted to abolish this system, which made the military aristocracy interested. in the conquests and rule of the Achaemenids aild wanted to return to that state organization,· which was represented by the Median Empire, and in which a centrally directed bureaucratic state organization managed the «royal land». This was to be realized in a smaller state like Media, but it closed the possibility of development and of getting rich from the aristocracy, and thus in final conclusion it opposed the Median aristocracy to the dynasty of the Deiocids, and brought about the fall of the Median state. This way was possible for Cyrus to overthrow the Median state and to establish the Old Persian Empire. The Median aristocracy was ready to line up behind the warlike Persian king, who with the conquest of waste territories and with the development of the new social and economic organization insured for it unprec€dented possibilities. Thus, in final conclusion, we can state that, besides the personal prominence and other personal circumstances, the fall of the Median state and the rise of the Old Persian Empire were brought about in a decisive way by the fact that the policy of the Deiocids led to a deadlock, from which the way out was found by Cyrus the Great and the other Achaemeni~ rulers, who succeeded to crePte the highest developed ancient form of government based on the Asiatic made of production, which later on not only existed for two centuries, but became the basis for the rise and development of the later 35 Hellenistic, and then the Arsacid and Sasanian states.

32

I hope to give a detailed explanation of this passage elsewhere. Old Persian v; 0abisaca 'property of the royal house' can be interpreted as consisting of viB- 'royal house' and abi-isa- 'property'. Old Peraian isa-·(< *ixsa-) compares with Avestan aesa 'Habe, Eigentum' (< *aixsa-) in the same manner as Avestan is-·'Wunsch' does with Avestan aesa 'Suchen' 33 I quote the Elamite text on the basis of the reading by G. CAMERON: JCS 14 (1960) 63. Instead of the reading rmar-ri-zp-ma, however, recommended by him, it seems to be more correct to read the fonn rmar-rt•-u-ma.

G.

34

As it was correctly recognized already bv E. l--IERzFELo: Altpersische Jnschriften. Berlin I 938. 51 ff. [Additional note. - On the Median ele~ents in Old Persian see recently M. MA YRHOFER: Neuere Forschungen zum Altpersische~. Donum Indogermanicum. Festgabe fi.lr Anton Scherer zum 70. · Geburstag. Heidelberg 1971, 46-50.] 35

38 39

Bibeliibersetzung war, die Johannes vor den Augen schwebte, uhd dafl er die in dieser Periode oft Skythen genannten Goten mit den antiken Skythen identifizierte, selbst dieser Umstand aridert nicht an der Tatsache, daB er eine Schdftlichkeit bei den antiken Skythen und Sauromaten angenommen hat, Diese Vorstellung steht jedoch in ausgesprochenem Gegensatz zu der Ansicht der sprachgeschichtlichen Forschung, wonach die Skythen und Sauromaten keine Schriftdenkmaler hinterlassen haben.

DIE SCHRIFT BEI DEN ANTIKEN STEPPENVOLKERN (Vorlaufige Mitteilung)

Nach einer allgemein verbreiteten Auffassung ist die Schriftlichkeit als ein wichtiges Kennzeichen der Hochkulturen anzusehen. Dementsprechend stellt man .die Kulturvolker, die -Uber die Schrift verfligen, den schriftlosen Stammen mit oraler Kultur oft gegenilber. Aber diese Gegeni.iberstellung von Schriftlichkeit und oraler Kultur kam erst in den letzten Jahrhunderten nach allgemeiner Verbreitung des Buchdrucks auf, den alten Griechen und Romem war diese Anschauung noch unbekannt. Mustert man die 1 Topoi der Barbarendarstellungen in der antiken ethnographischen Literatur cKHX u.apeii CKHnypa H Ila.JiaKa. KCCHHMK 12 (1946) S. 44. Da Skiluros der Rival von Mithridates VI war, ist die obige Erganzung seiner Titel wohl wahrscheinlichet, als diejenige, die man frOher vorgeschlagen hat. 13 V_ero··roent1·tcht von R. GHIRSHMAN: Notes iraniennes IV. Le tresor de Sakkez. Artibus Asiae 13 (1950) S. 187-188, Abb. 9-11. Vgl. noch R. GHIRSHMAN: Tombe princierl! de Ziwiye et le debut de l'art anirnalier scythe. Paris 1979. lO

auch die ilb~igen 15 Hieroglyphen konnen aus hieroglyphisch-hethitischen Schriftzei~ chen. abge!e1t~t w:rden: Aufgrund dieses Zusammenhangs -diirfte man auch die ldentitat oder die Ahnhchkett der Lautwerte beider Schriften annehmen. Substituiert nun die· hieroglyphisch-hethitischen Lautwerte in den Inschrifttext von Sakkez hin:::: so erhalt man den folgenden Text: 1 pa-ti-na-sa-na ta:•·pa w,ia-s 6-na-m5 XL was-was-ki XXX ar-s- tf:;.m 5 s3-kar-kar (HA) har-s6-tu 5 LUGAL . 2 par-ti-ta5-wa5- ki-s3-a4-a KUR-u-pa-ti QU-wa-a5 3 i5-pa- si-a-m2

:

~

p

~+ -

. 2~i

#~ tiD ~a

}I T8,ll.:iKHKCI\'.Oro 1-tapo.ua. I. MocKBa 1963, 255-256. J. HARMATTA: Sino-Indica. Acta Ant. Hung. 12 (1964) 18. 12

59 58

L I: '

I

.milarly his successors continued to launch preventive military expeditions nomads. Sl , • d ti b ti · t the nomads on the one hand,. and to fo11ify the frontier e. ence · y .· ortresses .agams . . . and fortification systems, transformed later into coherent tunes on some temtones, ·?n the other hand. Afterwards, these defence systems spread to Europe and E~stern Asia, too, for keeping off the nomads, and divided the Ancient and Early ~ed1eval World into a southern, more developed Lone and a northern undeveloped one. • Beside the well-known important centres such as Baktra (today Ball}), Marakanda (Samarkand) and Kyropolis ( on the Elamite Persepolis Fortification Tablets Ku17 16 risti, its Sogdian name being Kurkat "City of Cyrus" = today Ura-Ttibe ), Naut~ka played a great role as a strategic base in the military operations of t~e Mace~oman army. Historical research has not succeeded to clarify the geographical locat10n of Nautaka so far, but the opinion has become generally accepted that we have to look for it in Sogdiana. 18 In reality, however, we dispose of a whole series of evidences for ~t which has escaped the attention of historical research hitherto. By help of these evh dences the geographical location of Nautaka can be determined exactly. ' The earliest evidence for Nautaka occurs in the inscription of the Taxi la silver 19 scroll, dated from year 136 of the Old Saka Era. This date corresponds to 70/7LA.D. The passage concerning Nautaka runs as follows: Urajhakena l,ri.tavhriaputrana Bahaliena Noachae nagare vastavena. Its translation is the following: "By *Hurazaka, ~of the *]J.tafriya-sons, the Bactrian, the resident of the town of *Noachaka". The inscription was written in GandharI Prakrit, its author was, however, an Iranian and on the basis of his epithet Bahalia (< *Bahalika- < *Bahlika-) he was of Bactrian origin and resident of the town of Noachaa. In GandharT Prakrit, the fo1m Noachcia- goes back to the antecedent *Noachaka- and taking into consideration that this form is the name of a Bactrian, i.e. Iranian tO\vn, on the basis of Iranian historical phonology, we can restore its Old Iranian form as *Nava0yaka-. Without doubt, this form can be identified with· the Greek spelling Na,haKa as far as it reflects a form *Navlitaka- or *Nava0aka-.2° According to the testimony of the inscription on the Taxila silver scroll, Nautaka was lying- contrary to the assumption generally accepted- not in Sogdiana, but in Bactria. This mistake of historical research is connected with the cartographic ideas of the Greek geographers who regarded the Oxus river (=Amu-darya) as the frontier between· Bactria and Sogdiana. As the oriental sources prove, however, the Barsun-tau moun-

tains and the pass leading through them, the Dar-i Ahanin, ·the. "Iron Gate" was the fr ontier between the two lands, i.e. the territory of Sogdiana only included the· valley of the Zerafsan river .. The other evidence for Nautaka is furnished by the Middle Persian work Sahristaniha-i Eran (ch. 8) and mns as follows: BYN b'hl Y w'myk str'st'n' nw 'ck' spndy 't Y wst 'sp 'n BRH hi 1 = andar Bahl i vamig sahristan i Navacag Spandiyad i Vistaspanpus kird. Its translation is the following: "In the splendid Bactria, the town of Navacag was built by Spandiyad, the son of Vistasp''. The· Middle Persian form Navacagalso goes back to Old Iranian *Nava0yaka:-. The third occurrence of this place-name can be found in the geographical work of Pseudo-Mowses Xorenac'i inthe fonn Dzi Navazak, which is borrowed from a 22 Middle Persian phrase Diz-i Navazag "the fortress of Navazag". In Middle Persian, from the form Navacag developed Navazag regularly. Lastly, the fomth evidence comes from an enumeration of Ibn Ijurradaghih which describes the route leading along the valley of the Surxan-darya. 23 This route started from Termez, from where at a distance of 6 farsaxs was Car1?angan, from there again at 6 farsaxs lay DarzangI, then at 7 farsax,s was B.rngi, which is unidentified so 24 far. This place-name occurs, however, in the fonn Nauion in the Hudud aVAlam. Taking into consideration that the final alif and nun could come into. being by the break of the ve11ical · stroke of a final kof, we can restore the original, conect form of the place-name as · *Nauiak or *Navaiak. This forn1 could go back to the antecedent *Nava0yaka- as well. As c·oncerns the origin of this place-name, obviously, it is a compound, whose first component is Old Iranian nava- "new", while its second element is Old Iranian (Avestan) a0a- "habitation, settlement". Thus, the meaning of the place-name *Nava0aka- > GreekNm.1mKa will be HNew Settlement". Inthe case of the other data concerning this place-name, we have to start from the form *Nava0yaka-, inasmuch as the suffixed form a0ya- of the \:vord a0a- forms the base of the compound. According to the enumeration of Ibn Ijurradagbih, Nauiak ~ Navazak ( = Naurbxa) lay at a distance of 19 farsaxs (= c. 106 kilometres) from Tennez, somewhat to the north· of present-day Surci, at a place from where the pass, leading through the Barsun-tau, the "Iron Gate" of Central Asia was easy to access. At this place the goro25 disce Dalverzin-tepa lies today, where Albaum and Pugacenkova executed· archaeological excavations. On the basis of their results, this fortress dates back to the begin1

15

J. HARMATTA: The Wall of Alexandl!r the Great and the Limes Sasanicus. Bull. of the Asia Inst 10 (1996) 79 ff. [In this volume pp. 64 ff.] 16 W. BARTHOLD: Turkestan down to the Mongol Invasion. Oxford-London 1928, 166. 17 6. f. f AYPOB-E.A.JlHTBHHCKl1H: op. cit. 255. 18 6. f. f AYPOB-E.A.JIHTBHHCKl1H: op. cit. 245. 19 ST. KONOW: KharoshthT Inscriptions with the Exception of those of Asoka. Cllnd Vol. II. Part I. Catcutta 1929, No. XXVII. 20 J. i:IARMATTA: Sino-lndica. Acta Ant. Hung. 12 (1964) 18.

60

21

J. MARKWART: A Catalogue of the Provincial Capitals of Erai1shahr. Roma 193 l, 10; J. MARK.WART: Wehrot und Arang. Leiden 1938, 143 ff. 22 J. MARQUART: Eransahr. Berlin 1901, 82. 23 V. MIN0RSKY: Hudiid al-'Alam. "The Regions of the World". A Persian Geography 372 A.H.-982 A.O. Translated and Explained by~. Oxford-London 1937, 313. 24 V. MIN0RSKY: op. cit. 114. . 25 r. A. TTYfA1IEHKOBA: X~1qasni,. TalUKeHT 1966, 23 ff. 61

26

ninos of the Hellenistic Age. Accordingly, it may belong to the group of cities, fou~ded by Alexander. In any case: the Macedonian king ~ad chosen the valley of the Surxan-darya with an excellent sense for his strategic line-of-communications area. This is the sole valley of the right-side tributaries of the Amu-darya forming a plain c. 60 kilometres broad which could provide the Macedonian troops with agricultural products. From strategic viewpoint, Nauta.l(a was an excellent military base as well, because the Barsun-tau mountains provided shelter from the unexpected· raids of the Iranian nomadic Dahae and at the same time the pass Dar-i Ahanfn, the "Iron Gate" insured easy access to the more irnp011ant Sogdian cities. From among the Sogdian settlements, mentioned in the relation•of AITian, Kyropolis and Ale.-i:andreia Eschate have already been discussed above. The rest, ra(ct and Bayaf, couid have been identified by earlier historical research as welL Gaza was one of the seven Sogdian settlements, kept by the revolting Sogdians (Arrian,. Anab. Alex. IV. 2). From the report of Arrian it becomes clear that Gaza was lying between Kyropolis and the Sir-darya river. This settlement existed even in the Middle Ages, it was called razaq and its exact geographical location was indicated by the Arab geog.: 27 raphers. It was lying at a distance of 6 farsax,s from Xogand. · As concerns Bayaf, Arrian (A.nab. Alex. IV. 17, 4) describes it as .follows: ... ts Bayas. xevp!ov r'i}s Xoy81avfJs 6xup6v, tv µE8opfctJ rfjs Tc Xoy81a11c3v yfjs Kai~ rfjs Maaaaycrc3v XKv6c3v c/)K1aµf1,,;ov. On the basis of this passage, it is doubtless that Bayaf Jay on the \vestem frontier of Sogdiana in the neighbourhood of Buxara.

Consequently, it can be identified with the medieval settlement ofFagdfz or Fa,gcmcliz ( < Sogdian Bay-dfz or Bayfm-dzz "'Fortress of Bay or Bayan'), mentioned by the 28

rian, Anab, Alex. IV. 21 ), the name of a ~ogdian landlord, can be explained as a Sogdian compound *xwr-y 'n, consisting of the Sogdian words xwr '"sun~' and y'n "favour", hence *Xvaryiin ''Having the favour of the Sutf'. 2101µ!8pTJs (Strabo, 443, 52), the name of another Sogdian landlord, reflecting a Sogdian form *Sisimi0ra-, is again a compound whose first component may be Sogdian *sis- "to adhere, stick, 29 cling, be devoted" (cf.. Saka· $G§'-; §'i§'- "'to adhere" < Old Iranian sray-/sri-s- ), while its second element is the name of the god Mi0ra. Accordingly, the meaning of this proper name is "Devoted to Mi0ra". The third Sogdian proper name· 'Ap:aµciCTJs (Strabo, 443, 54) may be the Greek transcription of a Sogdian form *A_,ya-mazda..;, whose meaning might have been ''Noble Wisdom" or "Having noble wisdom~~. On the other hand, the name of the Scythian leader XaTpaKTJS-, killed in the battle, fought against Alexander beyond the Sir-darya, reflects a Saka form *Satrak or *Sa0rag which goes back to. Old Iranian *Xsa0ra-ka-, being· an abridged form of Old Iranian *xsa0ra-pati- "commander". The phonemic change· xs- >·s- points to a proper name ofSaka origin. The historical importance of Alexander's campaign into c;entral Asia can be seen in the fact that in his person ancient civilization came up for the first time against the equestrian nomads of Eurasia, and the experience: obtained in the course of this conflict, determined its .behaviour and defence against them for a millennium. It was Alexander· the· Great.who took . the first step in Central Asia to· construct the mighty system of limes, stretching froni the Atlantic through Eurasia as far as the East China Sea.

Arab geographers as lying somewhat to the north-west ofBuxara. The name 'OpfavTTJS of the Sir-darya may be particularly interesting from the viewpoint of lranjan dialectology. 'OpfavTf7S, reflecting an Iranian prototype *Vorxsant or *Vorx.ffmd, may go back to an Old Iranian form *Varu-xsayant- "ruling over wide (land)" and be connected ,vith the ethnic name ±avB101 or za118101 (Strabo 438 and 442,5 < Old Iranian *xsayant- '"ruling, royal"), the name of a Daha tribe, nomadizing to the north of Lake Aral. Thus, on the basis of its phonological development, 'OpfavrTJs(*Vorxsant-) may be the Daha name of the Sir-darya river~ The reliabiiity of the reports concerning the campaign into Central Asia, recorded by Alexander's historians and preserved by Arrian and other Greek authors, is proved beyond any doubt by the linguistic and historical analysis of the topographical data. The study of the Sogdian proper names, preserved similarly by Arrian and Strabo in connection with A_iexande(s campaign, can also confirm this result. Xop1rj1117s (Ar.: 26

27 28

f. A. IlYrAtJEHKOBA: op. cit. 22 ff W. BARTHOLD: op. cit. 167. W. BARTHOLD: op. cit. 123. 29

H. W. BAILEY: op. cit. 410.

62

63

II"""

THE WALL OF ALEXANDER THE GREAT AND THE LIMES SASANICUS I The first comprehensive description of the peoples living in Europe, Asia and Africla has been preserved in the work of Herodotus. If we project the many-sided characterization of the individual peoples, given by Herodotus, upon his geographical image of the world, a very interesting vision of human civilization emerges before us. In the centre, in the Mediterranean world and in the Near East, peoples (like Greeks, Egyptians, Babylonians etc.), creating a highly developed urban civilization, are dwelling. This central kernel is surrounded by a stripe of less developed, but settled agricultural peoples and tribes. Beyond these, more distant from the territory of the high cultures, follows the belt of pastoral, mainly nomadic, pasture changing· horse-riding· peoples. Lastly, on the fringes of the known world, in the most exterior zone are living the looting manner of life leading fisher and hunter tribes which often consume even their food in raw form -these are the ,,raw meat eating" barbarians, mentioned in the literatures of the high cultures. 1 ~ The image created by Herodotus about the development of civilization clearly reflects that recognition that about the middle of the first millennium B.C. the peoples and tribes, known at that time, stood at very different levels from the viewpoint of so~ cial and economic evolution. Surely, the unevenness in the development ofthe different human groups came into being already in palaeolithic times to a certain degree. But the neolithic revolution, the emergence of animal husbandry and agriculture increased the differencies in their development suddenly. At first, the backwardness of fisherhunter tribes became greater and of constant character, then the high cultures appeared in the fertile river-valleys. The differences, caused by the uneven development, increased later to an extraordinary degree when the specialized development of animal husbandry and agriculture led to the rise of separate pastoral and agricultural tribes. Afterwards, the emergence of the nomadic, pasture changing economy brought qualitative change in the relations between the pastoral tribes and the sedentary agricultural peoples. The agricultural tribes were essentially self-supplyiii.g, while, because· bf their one-sided economy, the nomadic peoples were necessarily dependent on the sedentary

populations and the mighty quantity of energy, being at their disposal in their great stock of horses which can be used not only for communica~ion and warfare, but also for the creation and functioning of state organization, made for them possible to force the ploughmen to deJiver them the necessary products in form of trade, taxation, gifts or even plunder, too. In the steppe belt of Eurasia, nomadic animal husbandry came into being eveth th rywhere until the 8 to the 7 century B.C., while, to the south of it, sedentary cultures, to the north, fisher-hunter tribes were living, as it was fixed by the description of Heroth dotus in the 5 century B.C. However, both the nomadic peoples and the sedentary ones gradually established state organizations, including great territories and the latters tried to defend themselves againstthe raids of the fo1mers, becoming more and more freqtI~nt and more and more extensive. Firmly defended frontier line against the nomads was at first established by the Old Persian Empire in Central Asia which became still better fortified by Alexander the Great during his campaigns against the nomadic Sakas, neighbouring upon the northern frontier of the Achaemenian Empire. In the 3rd century B.C., the Chinese also began to create an effective frontier defence against the northern nomads. The prevention of the natural movements towards the south of the nomads and the forcing them back gradually increased the tension between the nomadic tribal confederations in the steppe belt which· finally could only be solved by migrations, repeated from time to time , from the east to the west temporarily. The migrations from Orient to Occident of the nomadic peoples started and their first waves 3 can already be felt after the rise ofthe Old Persian Empire. After all, this migration of peoples crossed, then turned back the southeast movement of the German tribes and by· the pressure put on the Rhine;..Danube line it compelled the Roman Empire to build up the limes Romanus, Accordingly, until the 2nd century A.D., significant difference - a true dividing line between the northern and southern halves of Eurasia came into being from economic, social, political and ideological viewpoints and last but not least as concerns the living standards. The highly developed southern cultures, the Roman Empire, Iran and China defended their riches· and· urban civilizations within the frontiers· of their wellorganized states by their fortified borders against the nomadic empires or the ethnic movements brought about by their emergence. The western and eastern parts, viz. the limes Romanus and the Chinese limes of this mighty dividing line, stretching from the region of the Rhine up to the Yellow Sea, were rather well studied by archaeological and historical research. On the other hand, our knowledge about the Iranian limes is 2

I

.

. . :

. '· I *x.§ivonu > *Jvonu > Jun of Old Iranian *x.§aivana- 'ruler, king' without any difficulty. The phonemic change x.§>*i,is well known in Eastern Iranian languages thus e.g. in Xwarizmian and Waxi (e.g. Old Iranian *x.§apa- > Xwarizmian 'XYb, 'xb xb 'night', Old Iranian *vaxsu- > Waxi ·. Wux 'Waxan'. This form Jvonu > Jun is still reflected by the Sogdian variant Xiln, the Greek form X0Dvo1, and the Latin Chuni (in Orosius). The two latters can obvi- _ 1

H. DEGUIGNE: Histoire generale des Huns, des Tures, des Mongols et des autres Tartares. Paris 1756-1758. 2 For the history of the Hsizmg-nu - Hun identification see L. LIGETI: Attila hunjainak eredete (Origin of · the Huns of Attila). In: Attila es hunjai (Attila and his Huns). Ed. by GY. NEMETI-l Budapest 1940, 11 ff. 3 Q; ~CHEN-HELFEN: Huns and Hsiung-nu. Byzantion 17 (1944-1945) 222 ff. 4 D. SINOR: The Hun Period. The Cambridge History of Early Inner Asia. Cambridge 1990, I 77 ff.

ously go back to a Sogdian source because the -y- disappeared after consonant in Sogdian. Consequently, the Sogdian form Xan developed from Xyiin while the form Xyiln was constantly preserved in Middle Persian up to the end of this linguistic period .. Because the Persians became acquainted with the name of the Hsiung-nu not by the. Chinese -but either immediately or by another Iranian language of Central Asia (in any case not by the Sogdian), this name was the self-denomination of the Hsiungnu in all probability. This fact may explain the striking phenomenon that the name Hsiung-nu < Xwm only appeared in authentic Chinese sources about 200 B.C., i.e. at 5 the time of the rise of the Hsiung-nu nomadic empire. Obviously, the names of the individual Hsiung-nu tribes might have existed long since but they could only assume the general name meaning 'ruling, royal' after the rise of theirc tribal confederation or nomadic empire. rd Later, probably during the 3 -4'h centuries A.D., the form >J:.'iun > Xiln developed into Hun in the Eastern Iranian languages (cf. e.g. xvarasan 'sunrise, east'> Hephthalite horsano) and this form became known in the spelling Hil~za- in India.and as Ouvvo,, Huni used by the Gi·eek and Latin authors, respectively. At the same time, however, the Persians used the former form Xyiln further on and the Latin Chionitae .represents its transliteration (through the intermediation of the Greek spelling *X1ovfra1). As one .can state, on the basis of a more precise knowledge of the phonological history of the Iranian languages, from linguistic viewpoint we cannot regard the consonance of the names Hsiung-nu and Hun as fortuitous coincidence but their identity and historical connection can clearly be proved, even though the historical interpretation of the form XorJvo, and the· corresponding Latin variant Chuni was not yet discussed for the time being. Before we would attempt to explain the latter forms from linguistic and historical viewpoint, it will be useful to examine a Chinese report, which is suitable to fill in the gap between the history of the Hsiung-nu and that of the Huns to a certain 6 degree. This report is to be found in the Wei-shu ( 102, 7a) and its text runs as follows : ,,The land of Su-te lies west of the Ts 'ung-ling (= the Pamirs). It is the ancient Yents 'ai. Its other name is Wen-na-sha. It lies on a great lake northwest of K'ang-chii. It is 16 000 Ii far away from Tai(= the capital of the Wei dynasty). Previously, the Hsiung5

G. HALOUN: Zur Oe-t~i-Frage. ZDMG 91 (1937) 306, note 1; According to L. LIGETI: Az azsiai hunok (The Asiatic Huns). In: Attila es hunjai (Attila and his Huns). Ed. GY. NEMETH, Budapest 1940, 32, the name of the Hsiung-nu occurs for the first time in 318 B. C. His opinion depends on the review written by G. HALOUN, OLZ 25 (1922) col. 435-438 about the book ,,Die Hunnen der vorchristlichen Zeit" (Berlin-Leipzig 1921) of M. DE GROOT. It escaped, however, his attention that HAL0UN reverted to this question in his study quoted above and rejecting his own former opinion, he refuted the authenticity of the first mention supposed to have been in 318 B. C. of the Hsiung-nu. 6 For the followings see 0. MAEN~HEN-HELFEN, op. cit.

75 74

nu had killed its king and taken the land. King Hu-yi made·the third generation. Merchants of this land used to go in great number to the region of Liang but on the capture of Ku-tsang (which happened in 439 A.D.), all of them were taken prisoners. Anhe · beginning of the reign of Kao Tsung (452-466 A.D.) the king of Su-te sent an envoy to ask for their ransom which was granted by cabinet order. Since then no envoy has ever arrived to offer tribute." It was Fr. Hirth who drew attention to this report, the historical interpretation, given by him, however, was entirely erroneous from historical view• 7 pomt. From among the three land, mentioned in the report, Su-te (Old Chinese sjuk d'ak = Suyday) was identified by Hi1th namely with the town Sugdaia lying on the Crimean Peninsula in order to transfer the scene of the events· into Europe and to iden,. tify the capture of Su-te with the defeat the Alans suffered by the European Huns. He tried to prove this theory by the argumentation that according to the Wei~shu, Su-te is identical with the ancient Yen-ts 'ai which has, on the other hand, changed its name into A-Ian according to another Chines source, the Hou-han-shu (c. 118). Thus, according to him, the Hsiung-nu defeated actually the Alans when they took Su-le. Since Hirth published, however, his study, making sensation in the scholarly · world at that time, the sinology also developed into a philological science and on the basis of recent researches concerning the quoted report of the Wei-shu an entirely different opinion was fom1ed about the events described in it. At first it should he stressed that the quoted chapter 102 of the Wei-shu had,been lost before the 11 century A.D. and its Sung age editors replaced it by the corresponding chapter of the Pei-shih ap-. peared in 644 A.D.). The author of the Pei-shih again might have borrowed the report from the Chou_;_shu published 8 years earlier. According to the text of the latter, however, Su-te was only ,,presumably" identical with the ancient Yen;..ts 'ai. But, according to the earlier Chinese historical sources (e.g. the Hou-Han-shu), Su-te was not identical with Yen-ts 'ai. All the three lands, Su-te, Yen-ts 'ai and K'ang-chu can be identified by help of different Chinese sources with absolute certainty. Su-te was Sogdiana, lying in the Zara:fsan-Valley in Central Asia, K 'ang-chii was the region of the rivers Cu and'Talas 7

FR. HIRTH: Ueber Wolga-Hunnen und Hiung-nu. SBA W Bd. JI/II. MOnchen 1900, 245-277 and D~e- Ahnentafel Attila's nach Johannes von Thur6cz. Bull. de l'Acad. d. Sc. de St.-Petersbourg. ye Sene, Bd. XIIl/2. St. Petersbourg 1900. The assumptions of HIRTH were already refuted by KURAKICHI SHIRATORI: A Study on Su-t'e or Sogdiana. MRD of the Toyo Bunko 2. Tokyo 1928, 81 ff., esp. 87 ff. in detail without clarifying, however, the historical connections of the report. The theory expounded by HIRTH about the family tree of Attila was convincingly rejected by L. LIGETI: Attila csalad~aj_a (The Family Tree of Attila). Eotvos-FOzetek N° 5. Budapest 1926. Later he also supported the opm1on of K. SHIRATORI concerning the report of the Wei-shu, expounded in his study quoted a~ove, see Attila hunjainak eredete (The Origin of the Huns of Attila). In: Attila es hunjai {Attila and his Huns). Ed. by GY. NEMETH, Budapest 1940, 19 ff . : ..·

76

I

.J

and o(the middle course of the Sir-Darya. Yen-ts 'ai was again the land lying northwest

of K'ang-chu and north of Lake Aral. Its name, Archaic Chinese *iam-ts 'ad < *Hamciid points to the fact that its population consisted of Alans. It w~ namely the language of the Alans in which the Old Iranian term *hamcyiita- 'rest, quietness, quarter' developed into *hamcad with the characteristic phonemic change cy- > c (cf. Ossetic ancad.'quietness'). The later name ofYen-ts 'ai was not identical·with Su-te. It happened only after 430 A.p. that some Chinese sources identified the two lands erro8 . neous1y. · Consequently, according to the Wei-shu, the Hsiung-nu took Sogdiana, lying in Central Asia, however, these Hsiung-nu were neither Huns; nor Hsiung-nu, but Hephthalites, to whom in this report and also in other sources the name of the Hsiungnu was transferred. The Hephthalites were called Xyiin by the Middle Persian reports, (Ovvvo, by the Byzantine sources, they were named even ,,White Huns' (Sped-Xyiln in Middle Persian, /twKoi Ovwo, in Byzantine Greek, Sveta HilJ,.za in Old lndian). On the other hand, because the Hsiung-nU: = Xyiin = Chionitae invaded Kusansahr from the territory of Sogdiana and conquered its capital, Balx, about 350 A.D., it becomes evident that the Hephthalites, mentioned in the repo1t of the Wei-shu under the name Hsi9 ung-nu, won Sogdiana from the true Hsiung-nu (Chionitae) over. The .statement that the masters of Sogdiana were the Hsiung-nu before the Hephthalites will be confirmed beside the conquest of Kusansahr by the Chionitae (Hsi?,mg-nu) even by the other ·name of Su-te: viz. Wen-na-sha. Namely, in the Old Chinese form *uan-na-Ja of this name, the ethnic name Hiln and the Eastern ·Iranian (Hun) outcome nasa of Old Iranian *nisaya- 'settlement, habitation' can he recognized. Acc()rdingly, the meaning of the whole name *Hiln-nasa might have been 'Hun 10 settlement, Hun country' Obviously, this denomination go.es back to the Hsiung-nu themselves because its phonemic form has not Sogdian character and its proves that the development Xiln > Hiin took place in their language in the meantime. This conclusion is also supported by the fact that the Hsiung-nu appear under the name H ilJ,.za in India in the beginnings of the 5th century ,t\..D. The remarkable phenomenon that the Persians still called the same Huns by their former name Xyfin, can be explained by the assump.. tion that thef beecune acquainted with the name of the Hsiung-nu at that time when it still had the form X--iin. Accordingly, it is true that the report concerning Su-te of chapter .102. in the Wei-shu does not verify the historical identity of the European Huns and the Hsiung~nu 8

.

.

.

.

Cf. 0. MAENCHEN-HELFEN, op.cit. J. 1-IARMATTA: Chionitae, Euseni, Gelani. Acta Ant. Hung. 31 (1985-1988) 43 ff., esp. 51. 10 0. MAENCHEN-HELFEN: op.cit. 231 interpreted the name Wen-na-sha as *Huna-sha 'the Kings of the Huna'. If Wen-na-sha would be a name of Chinese origin, then this interpretation could be accepted because the Chinese denoted sometimes a country by the name of its ruler. This practice does not occur, however, in Iranian name-giving.

9

71

of Eastern Asia or the ethnic continuity between them, notwithstanding it represents a source of vital importance for the westward migration of the latter and for the phonemic developmentx!-un > Hun of their name. After this consideration, now we can examine the evidence of Ptolemy, according to which the XorJ1101 would have lived in the region of the Dnieper river alnd ready in the first half of the 2 century A.D. (Ptolemy III.V,10: µEmfu Si BaarEpvc:511 Kai 'PC,,Jfo).a11&J11 XoD110,). It is a well-known fact among the researchers of Ptolemy that there are some inexactitudes in the localization on the map of the ethnic names. in his work. The reason for this phenomenon can be found in the working method of Ptolemy who also used beside the astronomic data and the itineraries even the reports of travellers (I.2,2 faropfa TTEp1001Krj) and geographical maps (I.19 tv rois aKpt/3Eartpo1s rrfvaf,) whose geographical system was, however, not identical with his own geographical projection. It was the consequence of this working method that the geographical localization of several peoples is erroneous even on the territory of the Roman Empire. It is self-evident that there are much more errors in the localization on the map of the tribes and peoples on the enormous territories of Eastern Europe and Central Asia which were known much lesser. If we examine the maps of Sarmatia Europaea and Sarmatia Asiatica in Ptolemy, we find a confused mixture of disappeared, mythical or really existing but elsewhere dwelling peoples on them. Fortunately, in the case of the XorJ1101, we have the possibility to clear up the source of the error committed by Ptolemy and to detennine the real geographical location of this people. It was already observed by J. Markwart that Ptolemy.used in his description of Eastern Europe a geographical map which considered the Tanais-Don and the Tanais-Sir-Darya to be identical. The consequence of this working method was that Ptolemy (II.5, 12) transferred the altars of Alexander the Great from the Tanais-Sir-Darya to the Tanais-Don. In consequence of the same error he redoubled the people of the Rhoboskai and he located them once in Sarmatia Eurapaea, and once more in Scythia intra hnaum in his description. 11 As it is wellknown, the en-oneous identification of the Don river with the Sir-Darya goes back to Alexander the Great and his historians and it resulted not only in the catastrophe of one of his generals and his army but it led to a whole series of erroneous cartographic location of several peoples in the works of later geographers (cf. e.g. Pliny n.h. VI.20-22 ~ n.h. VI.48.c...50). 12 Thus, it can clearly be proved that Ptolemy used one or several sources (maps) which identified the Tanais-Don with the Tanais-Sir Darya. As a consequence of it, he located on his map and in his description west of the Don river several peoples which were dwelling in the geographic reality 1101ih of the Sir-Darya. His source localized the

~0171101 north of_the Tanais-Don = Tanais-Sir-Darya imagined to be one and the same rt~_er but regardmg the ~on and the Sir-D~rya as tw~ different rivers, .. for lack of any ot~er foothold, Ptolemy mterpreted the evidence of his source incorrectly and located this ~eople west of the Don river in the region of the Dnieper on his map. The evidence o~ ~his source was also p:eserved by Orosius (1.45) a fantibus Ottaragarrae usque ad civztatem Ottaragarram znter. Chunas Scythas et Gandarides mans Caucasus . c onsequently, according to this source, the Chum· Scythae were dwelling north of the mans Caucasus whose name denoted the Hindukush, the Pamirs and the T'ien Shan. This I _ calization ~ncludes the region of the Talas and Cu rivers, i.e. the territory of K'ang-c:u (of the Chmese sources) and the mention of the Ch uni there points to the fact that the Hsiung-nu immigrated into this territory sometime before the compilation ofthe work by Ptolemy. On the basis of the Chinese sources, two dates can come into account for this historical event. I) On the one hand, the time about 40 B.C. when Chih-chih, the shanyu of the Hsiung-nu tried to establish his power-centre in K'ang-chil during some years. 2) On the other hand, 91 A.D. when, after the defeat he suffered by the Chinese the shan-yu of the Northern Hsiung-nu, crossed the Altai Mountains and settled dow~ in K'ang-chii, while one part ofhis people, not being ab]e to follow him, remained in the vall~y of~he u~per course of the Ili river and founded the state Yueh-pan there. The latter still existed m the 5 · century A.D. and preserved the tradition of its Hsiurza-nu origin.·From among the two dates, it is the first which seems·to be moreprobabl; because of the escape into K'ang-chu of the Hsiung-nu shan-yii in 91 A.D. Even the Chines: took notic~ much later from the historical tradition of the country Yueh-pan. To an earlier date pomt also the fonns of the Greek and Latin ethnic names XoOvo, and Chuni, respectively, with which .the Greeks and Romans could only be acquainted through Sogdian intermediation. Similarly, the appearance in the form Xviln of the name Hsiung-nu among the Parthians and Persians points to an early dat~ about 40 B.C. On the basis of the Chinese reports concerning Chih-chih, it was possible namely ~o estabHsh that the legionaries of the Roman commander Crassus, captured at Carrhae m 54 B.C. became settled as prisoners by the Parthians in Margiana, on the northeastern frontier of their empire wherefrom, however, as mercenaries they went over· to 13 Chih-chih shan-yii. It seems to be self evident that the fame and the name of the Hsi~ng-nu arriv~d by them at the Parthians, the Persians and the Roman Empire. Accordmgly, the evidence of Ptolemy for the XoD1101 proves not the early appearance of the Huns in Europe but it represents an important element of those data which bear witness to the migration towards the west into Central Asia of the Hsiung-nu. Thus,. in final

II

J. MARKWART: um 4 (1924) 269-270. J. HARMATTA: L'apparition des Huns en Europe orientale. Acta Ant. Hung. 24 (1976) 280. In this volume pp. 88 ff. 12

13

·

··

.

.

H.H. DUBS: An Ancient Milit~1y Contact between Roman and Chinese. AJPh 62 (1941) 3 ff and recently V.A. L1vs1c: Ilepe.iJ.Hea:maTCKHH c6opH11K 2 (1966) 151 ff.

78

79

analysis, it confirms the theory of the historical continuity between the Hsiung-nu and the Huns. In any case, some doubt were raised by the historicar research against the authenticity of the evidence to be found in Orosius. Pliny and other Latin and Greek historical sources mention namely the ethnic names T6xapo1, (/)poDvo1,X1JpEfand Phuni et Thocari in Central Asia at this epoch and some scholars presume that the ethnic 14 name Chuni in Orosius came into being simply by the corruption of the form.Phuni. Contrary to this opinion, however, it was possible to prove that all those sources, mentioning the ethnic names Phuni et Thocari, go back to Apollodorus and refer to the epoch before the migration of the Yueh-chih and the years after it, i.e. to the time about nd the middle of the 2 century B.C. while the description of the source used by Orosius reflects the historical situation about 40 B.C. It is also beyond any doubt that the corrupt form is Phuni and it is the ethnic name (/)poOvo, which has preserved the correct form of the ethnic name. This people had been dwelling in the northwestern neighbourhood of China, i.e. in the northeastern part of Eastern-Turkestan and had not to. do 15 anything with the Hsiung-nu. The text of Orosius does not mention either the ethnic name Thocari, or that ofthe Seres but represents an entirely other historical and geographical horizon. It is fully independent of the report of Apollodorus,. consequent1y, there can be hardly any doubt about its historical authenticity. Without doubt, on the other hand, there exists some chronological gap. be- .. tween the last period of the history of the Northern Hsiung-nu, the historical activity of· the Central Asiatic Xyiln or Xun and the appearance of the European Huns but this gap can be filled by a thorough study of the Chinese sources. The westward migration of the Northern Hsiung-nu already began by the short-lived western conquests of Chichchich shan-yu about 40 B.C. but the No11hern Hsiung-nu tribal confederation still existed up to 153 A.D. when it was mentioned by the Chinese sources last time. Its centre of power was, however, transfened after 91 A.D. more and more into Dzungaria, to the region of the Hi river where one part of the Hsiung-nu settled down constantly and established the state Yueh-pan. Its name, Archaic Chinese *diwat-b 'wan, Ancient Chinese iwat:-b 'uan, Northwestern Chinese dialect iwao-b 'uan, points to a Hsiung:.nu form *Varpan which may still originate from their Eastern Iranian language and its meaning might have been 'defenders of fortress', or 'dwelling in fortress', and bears witness to the settled manner of life of this Hsiung-nu group. It is a remarkable fact that Chichchich shan-yu, too, had a fortress built for himself in the territory of K'ang-chu. The Hsiung-nu population of the state Yueh-pan only consisted, however, of. 14

0. MAENCHEN-HELFEN: op.cit. 232 ff., IDEM: The Legend of the Origin of the Huns: Byzantion 1 · (1944-1945) 248 ff., IDEM, Pseudo-Huns. CAJ 1 (1955) 102 ff., IDEM: The World of::the Huns, · Studies in their HistOI)' and Culture. Berkeley-Los Angeles-London 1973, 444 ff. · · 15 J. HARMATTA:Sino-Indicii:ActaAnt. Hung. ,12 (1964) 9-12. ..

80

the tribal group who ~ould not unde11ake the faraway migration while after the defeat he _suf~ered by tl~e Ch1~ese the__ ~orthern Hsiung_-nu shan-ya escaped according to the Pe1-:sh1h (c. 97) mto K ang-chu m 91 A.D., obviously together with the warlike tribal aristocracy. Afterwards, in 158 A.D., according to the Wei-shu c. 102, Pei-shih c. 97 (Yueh-pan chuan), the Northern Hsiung-nu abandoned even Eastern Turkestan and migrated to the west into K'ang-chu and the Kirghiz steppes lying north of it, i.e. into the terri_tory lying n01th of Lake Balkash in present-day Kazakhstan. Their place in the possession of Eastern Turkestan and the territory of the Wu-sun became occupied by the Hsien-pi in 166 A.D. 16 st

~rom t~~ end of _the 1 century -:1'.D. up to the seventies of the 3rd century A.D_-, K ang-c~m ':as an important _state, mdependent of the Hsiung-nu and dwelt by Iraman nomadic tnbes. The first wntten monument of their language in form of an inscription scratched on two bone plates came to light recently. 17 The power of K'ang- . chualso extended over Sogdiana, therefore the latter could not be conquered by the Sasanian. dynasty coming into power in Iran in 226 A.D. Because ·approximately a century later this territory, including also Sogdiana, was taken by the Hephthalites, at the same time, however, starting from the territory of Sogdiana, the Xyun (Chionitae) invaded and conquered the co~ntry of Kusansahr, belonging to Sasanian Iran 18 , it seems to be beyond any doubt that K 'ang-chu and Sogdiana was occupied by the Hsiung-nu = Xyiln = Chionitae about 270 A.D. and had been in their power about up to 370 A.D. Probably, the course of the events can be reconstructed so that the Hsiung-nu immigrating into K'ang-chu in large numbers from 91 A.D. on, became gradually stronger and about 270 A.D. they seized the power over the land. ·, Accordingly, the historical continuity between the southwestern group of the Hsiung-nu and the Xyiln (= Chionitae) invading Eastern Iran (Kusansahr) about 350 A.D. can be presumed with great probability. At the same time, however, there can be no doubt that the European Huns cannot be identified with the Xyiln (= Chionitae) because the Hephthalites conquering Sogdiana about 370 A.D. became wedged between the Huns living in the region of the Volga river and the Xyun (= Chionitae) occupying Kusansahr and even if they might have had contacts with one another, this contact became interrupted after the invasion of Sogdiana by the Hephthalites. · In any case; we have even material proofs for the assumption tbat the Huns, th appearing in the second half of the 4 century A.D. in the region of the Volga, had rd contacts during theJ and the beginning of the 4th century with theXyzm (= Chionitae) 16

On the basis of G. UCHIDA: Chronologische Ubersicht Ober die Westwanderung der Nord-Hiungnu. Kyodoshi Kenkyu (Studien Ober die Geschichte der Hunnen, Tokyo 1953) 151..,..153; FR. ALTW'M~H.W. HAUSS!G: Die Hunnen in Osteuropa. Pala~ologia 6 (1957) 1..,..14, esp. 5, note 5. J. HARMATTA: Eltiint nyelvek nyomaban Kozep-Azsiaban (On the Tracks of Disappeared Languages in Central Asia). MNy 92 (1996) 397. · 18 J. HARMATTA: Chionitae, Euse~i, Gelani. Acta Ant. Hung. 3 I (1985-1988) 43 ff., esp. 51.

81

rrn · g Kus"a-ns"ahr• Such a material conquerm . . proof is a bronze . coin of Sahpuhr kusansah {about 281-293 A.D.) which came to hght from Grave 5 m ~ cemete1:' o~the ~un Age in Gyor-Szechenyi Place. 19 Because the Kusano-Sasanian coms were m circulat10n only . on the territory of Kusansahr, this specimen could arrive at the hands of th_e ancestors, living in Kazakhstan, of the European Huns, bringing it wi~h th~msel:es mto Central Europe solely through the Xyiin (= Chionitae) who had Sogdianam their power. The other material proof is furnished by the famous golden necklace-pendant of the find from Wolfsheim which is inscribed with the Middle Persian inscription_ 'rt!J,str ( = Ardaxsahr) and which, came to light together with the unworn gold coin of 0 Emperor Valens (364-378 A.D.t This necklace-pendant can~ot have been ~he property of Ardaxsahr, the first Sasanian sahansah because acc_ordm_g to the testni:iony of the bust on his coins, this king bore a pendant in form of a s1x-pomted star on his neck. The idea also arose that the specimen of Wolfsheim was the necklace-pendant of 21 Ardaxsahr II, Sasanian sahansah (379-383 A.D.). This presumption encounters, however three insurmountable difficulties. 1) On the basis of the Parsiy letter forms, rd the inscription cannot be later than the end of the 3 century A.D. 2) On the basis of the testimony of the bust on his coins, Ardaxsahr JI did not bear any necklace-pendant. 3) During his reign, not any war activity between Huns and Persians happened. ~ Among the Kusano-Sasanian kings, however, an Ardaxsahr by name is also known who ruled approximately from 3 03 to 310 A .D. According to the tradition pre- .· . served in the Middle Persian Xvaday-namag, when Ohrmazd Il was followed on the throne by his son Sahpuhr II still under age, the ,,Turks" and the Arabs, i.e. the north22 ern and southern enemies yearned for the territory of Iran and invaded it. Denoting the Xyiin (Chionate), the name Turk represents a modernization in the historical tradition. On the basis of this report, it seems to be very likely that the Xyiin invaded Kusansahr when Sahpuhr II still under age mounted the throne and Ardaxsahr kusansah fell in the fight against them in 310 A.D. and thus his golden necklace together with the pend.ant bearing his name came as booty in the hands of the· Xy iin. As far as it can be discerned on the bust of his little bronze coins, Ardaxsahr kusansah wore a necklace-pendant ending below in a triangular peak, reminding of the form of . the specimen from Wolfsheim. 23 Very likely, some Hun tribes living·more to the north, :, on the territory of Kazakhstan, could also take part in the invasion into Kusansahr of ·: the Xyiin and one part of the booty, the necklace and the pendant, representing an in119 20

M. PARDucz: Die ethnischen Probleme der Hunnenzeit in Ungaro. Budapest 1963, 25. N. FETTICH: A hunok regeszeti emlekei (Archaeo!ogical Remains of the Huns). In: Attila es hunjai

(Attila and his Huns). Ed. by GY. NEMETH. Budapest 1940, 242-243, Pl. VII. Figs I-la.· I. B6NA: A hunok es nagykiralyaik (The Huns and their Great Kings). Budapest 1993, 174. 22 TH. NOLDEKE: Geschichte der Perser und Araber. Aus der arabischen Chronik des Tabari libersetzt von ~. Leyden 1879, 53. · 23 A.D.H. BIVAR: The Kushano-Sassanian Coin Series. JNSI 18 (1950) 32, Pl. IIl.22, 22a.

21

82

,1 \.

~. portant symbol of dignity among the Persians, became the property of a· Hun tribal head who migrated into Europe together with one part of the Huns later. The historical presence of Hsiung-nu, the descendants migrating westwards of , the Northern Hsiung-nu, even on the territories, lying north of K'ang-chu, is also nd rd proved by their archaeological remains from the 2 _3 centuries A.D., coming to light in the region of the upper course of the Ob, the Irtish and Ishim rivers and standing in close relation with the finds of the European Huns. The more important finds and sites, respectively, are the followings: Kos-Agac, Borovoe, Kara-Kengir, Barmasnoe, Biisk, 24 Akmolinsk. On the basis of the above-said, it seems to be very likely that the Hsiungnu, migrating into Central Asia split into two parts about ·the middle of the 4th century A.D., probably under the pressure of the Hephthalites: one part of them, living in Central Kazakhstan: shifted to the region of the Volga while their southern group was expelled the Hephthalites from Sogdiana into the territory of Kuso.nsahr, lying in the eastern prut of Sasanian Iran. It seems to be, therefore, beyond any doubt that the historical continuity between the Hsiung-nu and the European Huns can be proved; On _the other hand the question arises whether this continuity also implies full linguistic, ethnic, social, political and anthropological identity .or the language, etlmic composition, social organization and anthropological type of the Hsiung-nu underwent some changes during their migration to the west. The split into Northern·. and Southern Hsiung-nu of the Hsiung-nu tribal confederation, the gradual migration of the Northern Hsiung-nu into Central Asia, then their fmther split into the Southern Huns (Xyiln, HiJJ.za) invading Eastern Iran and Northwestern India and wielding power over great territories and into the Huns of Kazakhstan moving into Eastern Europe, already indicates as a matter of course that the organization of the Hsiung-nu tribal confederation underwent profound changes. We do not ·find any traces of ·the flsiung-nu leader tribes, the Luan-ti ~ Hsu-lien-ti (*B1,avantig ~ Xorantig 'ruler, king' from which the Hsiung-nu ruler, the shan-yu originated, either among Western or the Southern Huns, or of the tribe Hu-yen (*Xogan), whose head was honoured with the title wang 'king' by the Chinese. The . name.of this tribe goes back to Old Iranian *xsavan 'king' and it was borrowed in the form qayan by the Hsien-pi, then later by the Tiirks. All these phenornena point to an important rearrangement in the balance of power of the Hsiung-nu tribes during their westward migration. The leading role was taken by other tribes. One of them might have been the Hsiung-nu tribe Lan (Archaic Chinese *Gian < Eastern Iranian *Grein < *Vrcin < Varhran) which is mentioned as 1puva101 XKuBm by Ptolemy (VI,133) located northeast of Sogdiana and north of the Pamirs. In this spelling, the Hsiung-nu tribal name *Gran appears in the form Grun 24

S.I. RUDENK0: C11611pcKrui K0JIJieKu11x IleTpa I. MocKBa-JleH11Hrpa.u. 1962.

SJ

with the characteristic Eastern Iranian (Saka) phonemic change a>o/fi. The ending -aioi of the name points to the intermediation of Syrian merchants. Surely, the leader

tribe of the Hsiung-nu, conquering K'ang-chu and Sogdiana, -was the tribe Grun be-· cause the same tribal name can also be recognized in the name Grumbates, king of Chionitae (*Grun-pati 'head of the (tribe) Grun') who occupied Kusansahr in Eastern "5 Iran about 350 AD.The disintegration of the Hsiung-nu tribal confederation and the westward migration _of one part of the Hsiung-nu tribes might have brought about, of course, some change also in the ethnic composition of the Hsiung-nu. While the names of all. Hsi:.. ung-nu shan-yu, the few Hsiung-nu terms and the single Hsiung-nu sentence recorded by the Chinese, all these Hsiung-nu linguistic monuments go back to an Iranian -lan6 guage of Eastern Iranian type (likes Saka)2 and only the tribal name T'u-ko (Archaic •. Chinese *d'o-kldk < *Toyrak) seems to be of different origin and the earliest Tilrk linguistic data only originate from the time of the Hsiung-nu N011hern Liang dynasty, about 422 A.O. from the region of Kao-ch 'ang (Qoco)27, at the same time, beside the names of Eastern Iranian (Hsiung-nu) origin, brought from the East, there appear some names which represent other Iranian languages and chiefly names of Turkic origin. This phenomenon points to the fact that the northern branch of the Hsiung-nu tribal confederation, settled in Kazakhstan, became reorganized in Central Asia and joined with different Iranian and Turkic tribes. Accordingly, the problem of the origin of the Huns can be formulated with some simplification in the following manner: which other Iranian and which Turkic elements had been brought along by the :Huns into Europe. In this respect, we find valuable evidence in the legend of origin of the European Huns which was recorded by Priscus and preserved by Iordanes in the fullest form. According to this legend, the hunting Huns were led by a hind through the marshes of the Maeotis. In this legendary framework, however, some historical events . became also included and this historical tale runs inlordanes (c. 126) as follows: nam mox ingentem illam paludem transierunt, ilico Alpidzuros, Alcildzuros, ltimaros, Tuncarsos et Boiscos, qui ripae istius Scythiae insidebant, quasi quaedam turbo gentium rapuerunt. ·Halanos quoque pugna sibi pares, sed humanitate, victu formaque dissimi- . les, frequenti certamine fatigantes, subiugaverunt. The historical authenticity· of this passage in Iordanes' tale is proved on the one hand by Ammianus Marcellinus (XXX:3, 1) who also tells that the Huns defeated the Alans, dwelling in the Don region (Alanos Tanaitas), and on the other hand by the fact that the peoples mentioned in Ior25

J. HARMATTA: Acta Ant. Hung. 31 (1985-1988) 51. · H.W. BAILEY: Iranian in Hiung-nu. Acta Iranica 21 (1981) 22, 26; J. HARMATTA: Preface to the new edition of ,,Attila es hunjai" (Attila and his Huns), published in 1986. Budapest 1986, XIV-

26

XVII. I. EcsEDI: Ancient Turk (T'u-chueh) Burial Customs. Acta Orient. Hur}g. 38 (1984) 263-287; J. HARMATTA, op. cit. in note 26, XVII~XVITI. 27

84

danes also occur - even though. in another context - in a fragment· of Priscus (Exe .. de leg. 121 De Boor): or, 'Poua (3ao1AEvovros Ouvvc.Jv, 'Aµ1A(oupo1s Kai '!T,µcipois Kai Touvaovpa, Kai Bo10Ko1s Kaifripo1s l8vco1 rrpooo1Koua1 rov "/arpov Kai ts TTJV 'Pc.Jµaf{J.)V 6µmxµfav KaTa({Jvyyavoua,v ts µaxT]v tABciv rrpoT]p17µivos ... In these reports, two phenomena are remarkable. One of them is that four ethnic names only occur in Priscus instead of the five ones, enumerated by Iordanes. For this phenomenon several explanations are possible. According to the textual criticism, originally it was only one name which stood.in the text of Priscus instead of the first two ones in Iordanes, both similars to one another and also to the form 'AµfA(ovpo, in Priscus, the other name was written as variant reading or alternative transliteration above it or on the margin 28 but later a copyist inserted it into the text. The other possibility to explain the difference in the number of the ethnic names in the two authors may be that originally more names were also enumerated in Priscus' text but the writer, preparing the excerpt, left these aside and abridged the text with the phrase Kai frtpo1s l8vc01. The other more important observation is the following. According to the tale narrated· by Iordanes, the Huns defeated the enumerated peoples and attached to their tribal confederation before the Alans. Because before the invasion of the Huns, it was the Alan tribal confederation which held the whole Eastern Europe in its power, the Huns defeated the mentioned peoples obviously still in Western Asia and brought along them therefrom. Because the legend about the mythical hind was connected to the Maeotis, Priscus had to change the geographical scrne of the events but the historical order of them remained unchanged. In the peoples, enumerated before the defeat of the Alans, we have to look, therefore, for those ethnic elements who became included into the Hun tribal confederation in·Western Asia. Accordingly, their ethnic identification has decisive importance from the viewpoint of the origin of the European Huns. As it was said above, two ethnic names, Alpidzuri and Alcildzurii in the text of· Iordanes, correspond to the name 'AµfA(ovpo1 in Priscus. Because the text of Priscus used by Iordanes must still have been written in Roman uncial script wherein the A, /\ for one another and for M as well as for AK could easily be mistaken, from among the two Latin ethnic names it may be Alcildzuri which can be regarded as the erroneous transcription of the Greek spelling 'Aµ!Ji(ovpo,. On the other hand, if an ethnic name, corresponding to the Alpidzuri of Iordanes, also existed in the text of Priscus, then its spelling could have been *AMINZOYPOI in all probability. Because in the Roman uncial script, the iota could be joined to the letters having ve11ical strokes, the ligature IN could also be read as TTL Thus, the spelling *AMINZOYPOI was interpreted by Iordanes erroneously as Al(l)pidzuri. The ethnic names hiding behind the Greek spellings 'AµfA(ovpo1 and * 'Aµfv(ovporcan be restored most likely in the form *Emil-cur 28

On the basis of M. KRASENINNIKOV: E:maHTHMCKoe o6o3peH~e 1 ( 19 I 5) 0. MAENCHEN-HELFEN: The World of the Huns. Berkeley-Los Angeles-Lqndon 1973, 402.

85

and *Em in-cur, respectively. The existence of the latter ethnic name in the Hun Empire is also proved by the fact that a relative of Attila was bearing the name Emnetzur (Iordanes c. 266) = *Emin-cur. The second element of these ethnic names represents the Turkic dignitary name car/cur of Iranian origin while their first element Emil and Emin, respectively, can be identified with the tribal names Yen-mieh (Ancient Chinese ·ien-miet, N01thwestern T'ang inmiJ< *!mil) and Yen-mien (Ancient Chinese -ien-mien, Northwesterrt, T'ang ·in-min < *!min) to be found in Chinese sources which mention the ,,three Yenmieh" or ,,three Yen-mien" tribes east of Lake Balkhash. The hesitation of the Chinese -sources concerning the name of the three tribes can he explained by the fact that the names of the first two tribes were !mil and !min and the Chinese used sometimes the 29 first, sometimes the second to denote the whole group of the three tribes. Surely, it was only one part of these two tribes which advanced together with the Huns into Europe because they occur even in the Kimak tribal confederation whose first two 30 tribes are denoted 'ymy and 'ym 'k ~ 'ymk, respectively, by Gardizi. By changing the punctuation of this spellings in Arabic script, we can restore without any difficulty the correct forms 'ymn and 'yml (to be read !min and !mil respectively). The tribal name !min was still known even in the Mongol Age.31 From among the peoples mentioned by Priscus, interpreting the .basic form Touv- of the name TovvaoupES as *Dun-, we can identify it with the Central Asiatic 32 tribal name Dun:z- , occurring in Saka documents several times, as far as we presume that its original form was *Dun-cur but in the Greek transcription * TO YNZOYPEI, the uncial letters NZ was erroneously read by a copyist as NC which was possible beth cause of the uncial form of the Zin the 5 century A.D. indeed. Very likely, the tribal nanie Boi'aKot can also be identified with the ethnic name Bosikiitti ·mentioned on the · 33 Stael Holstein scroll. This form of name· in the- Saka text· was transcribed from Chinese and it reflects the archaizing pronunciation of the Chinese characters in the Impe~ 34 rial court. Taking into consideration the Northwestern T'ang dialect and the Saka orthography, we can restore the true form of this tribal name as *Bosku or *Busku. The Greek diphthong 01 already developed into the monophthong u, u at that time (cf. the Greek transcription Ko1afaTOpos of Latin quaestor), thus the Greek transcription

Bol'aKot reflected rather exactly this Central Asiatic tribal name.

The other Hun tribal names, mentioned in the Greek and Latin sources cannot be identified in the eastern sources for the time being but all these names ca~ be explained by help of the Eastern Iranian (mainly Saka) languages: 'frfµapo, < *Hitimaru 'strong, heroic', O1J;frfv(o11po1 < *Ultin-cur ,-majestic-cur', Angisciri < *Hangi-skir'entirely red', Bittugures CKHee .upeaHocrn ("Sarmatian and Indo-Scythian Antiquities"). In: Recueil d'etudes dedies a la memoire de N. P. Kondakov (Prague 1926), 144. Recently, in both Bulgaria and Rumania further objects of Old Persian origin came to light (according to the kind informations of Professor V. Velkov and Professor P. Alexandrescu). · 31 J. HARMATTA: A Recently Discover~d Old Persian Inscription. Acta Ant. Hung. 2 (1953/54), 1. ff.; M. MAYRH0FER: Supplement zur Sammlung der altpersischen Inschriften (Wien 1978) 16. Linguistique Balkanique 9 (1966) 7 ff.; K. OLZSCHA: IF 72 (1967) 152-6. [See now: J. HARMATTA: An Inscription Written in an Unknown Language. (In Hungarian). Ant.Tan 41 (1997) 13~24.] · · 32 V. BESEVLIEV: Inschrift in unbekannter Sprache aus Nordbulgarien. Glotta 4 (1965) 517 ff.; V. GE0RGJEV: Die Deutung der altertilmlichen thrakischen Inschrift aus Kjolmen. Linguistique Balkanique 9 (1966) 7 ff.; K. OLZSCHA: IF 72 (1967) 152-6. . 33 M. V ASMER: Untersuchungen Ober die altesten Wohnsitze der Slaven, I: Die Iraner in Sildrussland (Leipzig 1923), 65 ff.

°

PROLEGO11ENA TO THE LIBYKOI LOGO/ OF HERODOTUS' "HISTORIES" In Book IV of his "Histories", ~fter the narration of the expedition ·led by D · · · h · · anus agamsU e Pantie Scythians_, as_ the military operations of Megabazus were going on in the_~ellespontus ~d~hrac1a, Herodotus also makes mention of the Persian campaign which took place mL1bya at the same time. The immediate object of this expedition was to take Barca. and,_ therefore, Herodotus describes the foundation of Cyrene and Bare~ and :he ear_her history of both towns in a detailed manner (IV 145-164). From t~e v1ewpomt of its contents,. this part forms an almost independent whole with particular problems, even though it belongs, no doubt, into the framework ofthe Persian campaign against Barca. Herodotus only makes the reader acquainted with the reason o~ t~e Persian expedition in Chapters 165-167. Arcesilaus, King of Cyrene, being afraid of the fate predicted by an oracle, escaped to Barca where he had good relationships, viz.his wife, _being also otherwise his kinswoman, was the daughter of Alazeir King of B~rca. His flight became, however, his• destiny because refugees from Cyren~ together with some Barcaeans killed both him and his father-in-law. Thereupon, his mother, _Pheretime, resting at Cyrene where she enjoyed royal rights, fled to Aryandes, the Persian satrap of Egypt and referring to the services Arcesilaus rendered to Cambyses as well as to the fact that he was killed because of his pro-Persian policy, she asked the sa:rap for revenging himself on the assassins. Aryandes fulfilled the request of Pheretnne and placed great land and naval forces at her disposal. When the Barcaeans, warned before-hand, did not extradite the assassins, the Persian army left for Barca, in Herodotus' opinion with a farther intention, viz. for the conquest of the Libyans. . _Herod?tus describes the scene of the campaign, Libya and the peoples living there m a detailed manner (IV l 68~ 199). In his narration, he starts from the western border of Egypt· and, passing along the seashore, he takes the peoples one after the other, detennines the geographical location of every single tribes and gives short ethnographical characterization of them. Afterwards, describing the interior of the continent he mentions that farther in from the inhabited littoral belt there exists a zone rich in wild animals but otherwise uninhabited and farther in from that a desert zon~ extends in which oases are lying at a distance of ten days march one from another. These enumerations are followed by a summarizing part in which Herodotus describes the customs of the nomadic Libyans. Then he enumerates the peoples dwelling the.western part of Libya and draws a synthetic picture of the ethnic relations and of the fertility of the soil. .

216

217

Between the limits of the described geographica! area tak~s place the story of · (IV 200-205) Arriving at Barca the Persians lay siege to the town but the campaign · ' . . · ak · · t p stubborn resistance The Persian siege techmcs m e experiments the Barcaeans pu u · . . . with the digging of underground mines but by h~lp of a skilful mvent1on the Barcae~ns always discover and by digging of counter-mmes d~stroy th~m. At last, after nme month of siege, the Persian commander resorts to a femt: he· feigns _to conclu~e p~ace with the Barcaeans and when the latters already feel secure and admit the Persians ~nto the town, they take it by an unexpected attack. Now follows the v~ngeance: Pher:t1me has the criminals impaled. The other Barcaeans - with the except10n of t?e Battrndae and the non-accomplices in the murder - were taken prisoner by the Persians. On the way backwards, the Persian army makes an unsucces_sfu~ attemp~ at the capture of Cyrene. The Barcaeans taken prisoner are settled by Danus m Bactna. . . It becomes clear from the course of the narration that the whole consis~s- of three parts, rather sharply separating one from another. 1) The first part, ~ompnsmg Chapters IV 145-167, is again divided in two sections: the long first sect10n, rep~esented by Chapters IV 145-164, relates the earlier history of ~yrene and_ Barca, while the short second section (IV 165-167) expounds the causes which gave nse to the Persian campaign against Barca. 2) The second part, sharply separating from the former ' contains the geographical and ethnographical description of Libya (IV 168-199). 3) The third part relates the proper history of the Persian ex~edi:ion (IV 200~20~} N_ow, the question arises what is the function of the whole narra_ti_on i~ Her~dotus Histories. Without doubt, the histqry of the Persian expedition mto Libya represents ~n excursion from the viewpoint of Herodotus' whole work. Even the framework of thi_s digression burst, however, by another excursion viz. the geo~raphi~al and e~hnograph1cal description of Libya. On the basis of these and other -d1gress1ons ~arher research presumed that Herodotus originally wanted to write a work of geographical a~~ ethnographical character in the form of~ periegesis, simi!ar to Hecataeu~' co1:1pos1t1on, entitled Periegesis. According to this theory, at first his work_crystalhz~d m the form of monooraphs dealing with the single Barbarian peoples. Thus, the Libyan logos, too, would be the survival of such a monograph. Later on, when the Persian War got into the centre of his interest, he incorporated these earlier materials into his description of it. Recently, it became, however, dear that Herodotus wanted to describe the comprehensive picture of the oikumene and the general historical and cultural development of it. Correspondingly, Libya and the Libyans appear not only in Book IV Chapters 145205 but in one or another historical context also in Books I, II, III, V, and VII, i.e. altogether in six from among the nine Books of the Histories. They occur in diffe~ent ~istorical and cultural contexts whereby Herodotus created a whole network of h1stoncal and cultural connections between Libyans and some other peoples of the oikumene, demonstrating herewith the unity of mankind. _ . . The excursion on the geography and ethnography of Libya cannot be Justified

by the Persian expedition against_!3arca alone, because it describes even peoples and lands which were lying far from the theatre of war. Similar phenomenon could already be observed in the description of Scythia, too. Obviously, contrary to Hecataeus' world concept which did not sharply distinguish between real and mythical geographical elements, Herodotus laid great stress on the distinction between really existing and mythical peoples. Thus, his perception of the world explains why he dealt with the peoples located on the fringes of the oikumene, i.e. with the borderlands of myth and reality, in a more detailed manner. Similarly to his Scythian logos, he probably polemized also in the description of Libya against Hecataeus (without naming him) who-:according to the scanty evidence - mentioned at least two mythical peoples, the Pygmaejand the Sciapodes (FrGrHist 1 F 327 and 328a-b) in Libya in his Periegesis. Be~ide the Periegesis of Hecataeus, Herodotus also used his observations and informations, obtained in Cyrene, for the description of Libya. His visit in Cyrene may be proved by the fact that he speaks (II 181) of a statue erected in Cyrene which still existed in his time. Visiting Cyrene, Herodotus became acquainted first of all with the oral tradition about the foundation and history of the city, the fate _and vicissitudes of the Battiadae, but perhaps also with some written history of Cyrene. In favor of the latter assumption speak the dorisms in the text of the oracles of Delphi, given to Battus Ithe Founder (IV 157) and Battus II the Happy (IV 159) and the exact chronological and genealogical data. In Cyrene he also had the possibility to control the earlier descriptions of Libya and to acquire new informations about the Libyan tribes by his inquiries. During his stay in Cyrene he probably made even excursions in the adjacent region and could meet foreigners, thus e.g. Carthaginian or Phoenician traders. Herodotus himself refers to Carthaginians as his source for the island Kyrauis and the trade beyond the Pillars of Hercules (IV 195-196). If the term (cyiptES for some species of rats can really be_ regarded -as of Poenician origin, its interpretation by the Cyrenaean word f3ouvof, may prove that at least part of his Phoenician informations, Herodotus received at Cyrene. For the early history of Thera Herodotus might also have used a local historical work. of genealogical character. Such an assumption would well explain the abundance of genealogical data in IV 147. Perhaps this historical work was accessible to Herodotus during his stay on Samus which maintained friendly relations with Th era. The data on the archaic period of Thera going back up· to the Argonautae and the Trojan War may derive of the san1e logographic work while the similar references to Iason and Troy in connection with the Libyan tribes possibly come from the local history of Cyrene. In spite of the rather general opinion according to which Herodotus Libyan history cannot be controlled by help of contemporary sources, this possibility does exist to a certain degree. According to Herodotus' narration, the Persian expedition against Barca took plac~ at the_ same time when Megabazus led a campaign for the conquest of Hellespontus anq Thracia. If Darius' expedition against the European

219 218

Scythians can be dated to 514/513 B.C. (cf. J. HARMATIA: Darius~ ~xpedition ~gainst the Saldi Tigraxauda. Acta Ant. Hung. 24 [1976] 17), then the m1htary operations of Megabazus began in 512 B.C. and lasted at least the whole year, i.e. 512. B.C. It foilows that the Persian army also left for the conquest of Barca an9 Libya in 512. B.C. and the campaign ended in 511 B.C. Comparing Herodotus' chronology of these events with the evidence of the Achaemenian inscriptions, we arrive atthe following result: DB 520-518 B.C. mentions Maka but the Saka: tayaiy: paradraya "the Sakas who are (dwelling) across the sea", Kusiya "Nubia", Hi(n)duf"Sind" are still missing. It is a difficult problem where the Maka could be located. In· Ancient Greek geographical literature two peoples named Ma Km are known: 1) the Macae dwelling .. according to Herodotus (IV 175) in Libya and 2) the Macae mentioned by Strabo (XVI 3,2) living in Oman. The choice between the two Macae for the identification with OP Maka is easy. The Macae of Oman probably represent the descendants of the ancient population of the land Makan (in Old Babylonian) or Magan (in Sumerian). They must be distinguished from the MvK01 (Herodotus III 93, VII 68) who were living on the opposite Iranian seashore. There exists no evidence for the Macae of Oman as belonging to the Old Persian Empire and taking into consideration that the list of lands in DB still represents the extension of Darius' realm in the state at the end of Cambyses reign not even Hi(n)dus "Sind" belonged to it-, it seems to be beyond any doubt that the Q,P Maka of DB cannot be identified with the Macae of Oman. On the contrary, the Macae of Libya can be identified with OP Maka in all probability. According to Herodotus (III 13), Cyrene and Barca as well as the Libyans surrendered to Cambyses after his victory over the Egyptians. In this context, however, the name Libyans obviously refers only to a part of them because Herodotus explicitly says later (IV 167) that merely a little part of the Libyans was subject to the Persian King. Cons~quently, Old Persian Maka represents without doubt Cyrene, Barca and some Libyan peoples, first ofall the Macae. DPe 511 B.C. already enumerates in the list of lands beside Maka also dahyava : tayli : para : draya "the lands which are (lying) across the sea" without any specialization and first mentions Hi(n)dus. This evidence reflects the historical situation after Darius' expedition against the Pantie Scythians when the conquered European territories Were not yet organized definitively from administrative viewpoint but they were already included into the list of lands under Persian rule. The news of the successful end of the Libyan campaign did not yet reached Persepolis at this time while the conquest of Sind was already accom- · plished. DNa 500 B.C. enumerates among others the following lands: 13) Hi(n)dus 24) Saka: tayaiy: paradraya 25) Skudra 27) Putliya 28) Kusiya 29) Maciyli. In this inscription, beside Hi(n)dus "Sind", Kusiya "Nubia", Maciya "Macae" and the recently organized two European satrapies, the Saka : tayaiy : paradraya "the

220

Sakas wh~ are (living~ across the sea" and Skudra "Macedonia and Thracia" a new name Putaya (Babyloman !u-tt--tu) also appears. This name, sometimes identified with the land. Pwnt of the.b earlier E.• gyptian hieroglyphic texts in the scholarly literat ure, . sure ly denoted th e L1 yans as 1t 1s proved by the Septuaginta in which Hebrew Put is re~dered b~ "~iby~n~" in some ~assag:s, ~d by its equivalent pnhw in the Egyptian Hieroglyphic mscnpt10ns of Danus which 1s an ancient Egyptian term for Libyans. It follows that between 511 B.C. (DPe) and 500 B.C. (DNa), beside Maka, a further Libyan territory by the name Putliya was annexed to the Old Persian Empire. . DSz/Egypt!an 495 B.C. mentions in the list of lands 17) s~k~trw = Skudra (Macedoma and Thrac1a) 20) qmt ,,Egypt" 21) tmhw "Libya" 22) n~hs "Nubia" 23) mcg "Macae" 24)hndw~y "Sind". DZc/Egyptian 495 B.C. The name ofSkudra became illegible, the items 21-24 are the same as .in DSz/Egyptian. XPh 483 B.C:. enumerates 18) Maciya 21) Hi(n)dus 26) Skudra 28) Putaya 30)Kiisiya. These three Ac.haemenian inscriptions already reflect another historical situation: the "Sakas who are (living) across the sea" (= European Scythians) are missing from the list of lands because the Scythians revolted and shaked off the Persian rule in 496 B.C. On the basis of this historical evidence it becomes clear that Herodotus' narration about the Persian expedition against Barca and the Libyans can fully be verified. In particular, the following points are to be emphasized. 1. Herodotus' account on the surrender of Cyrene, Barca and part of the Libyans to Cambyses is confirmed by the appearing of the name Maka (Hieroglyphic mCg) in the Old Persian list of lands from the inscription D"ij (520-518 B.C.) on. The use of the name Maka (= M can be seen and after it - unless it is not a break on the surface - the nght-hand side stroke of the head of an A can be observed. Frg. II/3. The remains of two characters can be discerned on this little piece. On the right-hand side an almost wholly preserved F can be seen which was already identified by Ebbinghaus as well, while on the left-hand side, the slanting stroke, bending slightly to the left and crossed by an almost horizontal stroke, can hardly be interpreted otherwise as the right-hand side lower part of an A. Frg. II/4. On this fragment, two letters can clearly be discerned: S and I. The latter was correctly identified already by Ebbinghaus as well. Besides, the break-line of the plate, to be seen to the left from the S, has such a shape as if a cha~acter A would have been broken down from the rim of the lead plate. Some. traces of letters can be discerned even on the damaged part of the plate above the letter S which perhaps may · be the rest of an A and an I. Frg, II/5. The remains of three lines can be observed on this fragment. In line 1, the lower ends of the strokes of letters have only remained. On the basis of their po- ~ sition, they might have been an A and an I or a T. The first character in line 2 was interpreted by Ebbinghaus as H but the two vertical strokes are not connected above so that it is better to read the right-hand side long stroke as an I, while the left-hand side short stroke perhaps may be the upper part of an L. Then a clearly legible I> follows while in line 3 the preserved head of a I> can again be discerned. Frg. II/6. On this fragment, some remains ofletters can be seen in two lines; At the end of line 1, an N becomes rather clearly distinct. One can guess a faint cursive B before it, while to the left of it, a clearly visible vertical stroke of letter appears whose upper end bends slightly to the left. Surely, this is the rest of a U. In line 2, the upper right-hand side part of a B can be discerned.

IDENTIFICATION AND RESTORATION OF THE TEXTS OF THE FRAGMENTS Examining the fragments discussed above, it will be at once striking that two types of script can be observed on them: one is the Gothic uncial script whose letter fonns, however, slightly differ from the shapes of the signs used in the Codex Argenteus, the other type again represents the Gothic cursive script whose characters appear in different variants and they also differ in some details from the Gothic cursive letter ·forms, known so far. It is remarkable that this striking phenomenon has escaped the attention

of b_oth S~ekely _and Ebbinghaus. From among the fragments, the followings are written m uncial scnpt: 1, 2-3, 4, Sb, 7, 9, 12, 13, I/1, 1/2, 1/3, 1/4, 11/3, and Il/5, while the fragments Sa, 6, 8, 10, 11, 1/5, 11/l, II/2, II/4, II/6 were engraved in cursive script The importan~e of ~his _fact is obv~ous, because there can hardly be any doubt that the fragments wntten m different scnpt represent preserved pieces of different texts. Accordingly, they offer invaluable help for the identification and restoration of the orioinal texts. This observation is, however, of great importance even from the viewpoint that it testifies to the fact that the uncial and cursive variants of the Gothic script· had come into being already at a rather early date and both variants were used parallelly even for !he ~vriting of biblical texts. But from the palaeographic viewpoint the greatest surprise 1s still caused by the fact that the Gothic fragments from Hacs-Bendekpuszta, both the pieces written in uncial alphabet and the ones engraved in cursive script, use the same form of the letter S which developed from the Greek sigma and which·. was only preserved in the later Gothic documents written in the cursive script, while it was replaced 5 in the uncial alphabet by the Latin S after the immigration of the Ostrogoths into Italy. It follows from this fact that the text of the Gothic fragments from Hacs-Bendekpuszta had still been written before the march of the Ostrogoths into Italy, i.e. before 488 A.D .. Accordingly it is much_ earlier than the other preserved texts of Wulfila's translation of the Bible. The. following letters occur in the fragments with uncial script: A D E Z H I> I M N R S T W F H/ 0, while the cursive alphabet is represented on the pieces written in this script by the following characters: A B D EH PIN J UR S W F. The early date of these texts is also indicated by the fact that the differences between the uncial and the cursive letter forms are not so great as in the documents of the VIth century A.D. The typical letter fonns of the later, Italic cursive script of the Vlth century can only be observed in the case of A B D HP UR S w; while in that of S, the form borrowed from the Latin S does not occur in the uncial script of these fragments which still uses the letter form developed from the Greek sigma and preserved in the cursive script even later on. The identification of the context of the fragments becomes possible by two factors: l) all letters on the fragments could be identified, 2) some fragments consist of several lines whereby the identification of the context can be restricted in between narrow limits. Let us begin the identification and·restoration by fragment I whose reading is the following: rA1 UST line 1 2 ATTArw1 3 MINI>E 4 'MISEIS1

5

262

W.STREITBERG:

Gotisches Elementarbuch. Heidelberg 1920, 42.

263

A

the lines: line 1

2 3 4 5 6

(11) [NI I>ANASEH>S IM IN I> AMIY1A FA-] [IRH/AU; II> I>AI IN I>AMMA FAIRH/-] rAlI' Sr"J1 [ND, JAH IK DU I>US GAGGA,] ATTA rw [EIHA, FASTAI INS INNA-] MIN I>E[INAMMA, I>ANZEI ATGAFT] ~IS EI S1 [IJAINA AIN SWASWE WIT]

22 letters 22 22 23 24

eral lines. The reading of fragment 2-3 runs as follows:

~nr [. ]F

1

NAM~

rir

As one can state at once, the text of fragment 2-3 represents the immediate continuation of fragment 1. Between the two fragments, at the very most, 2-3 millimetres are missing which could have crumbled when the lead scroll was broken. After all, lines 4-6 of verse 11 in chapter XVII of St. John's gospel will be the following: line 4 5 6

ATTA f"\VEilf[A] ff'1 [ASTAI INS IN NA-] MIN I>E[l]NAM1M1 [A, I>ANZEI ATGAFT] rMIS EI SII'[AINA AIN SWASWE WIT]

Thus, almost half of the lines 4-6 of verse 11 in chapter XVII of St. John's gospel can be restored from fragments 1 and 2-3. Further restoration will be possible

264

;

(\

// J JI 8

1-' . .::-

ls'

B

D

0

()

t

~.e-

0 E

z

·:;..

H

h

z n

p

YJ

'}J

I

I

.:,

..

M

MM

1 H

N

N NN

N

'

'l

A

e,

;

21

Accordingly, there can be no doubt that fragment 1 belongs to the text of Christ's prayer to his Father beginning with verse 1 of chapter XVII in St. John's gospel. Because this fragment was written in the uncial alphabet, it seems to be very likely' that all fragments with uncial script belong to the text of the same prayer. Taking into consideration that the four lines of fragment 1 determine the length of the lines with the accura~y of 1 to 2 letters, we can easily check the correctness of the identification of the text by help of the other fragments written in uncial alphabet and consisting of sev-

line 1 2 3

'

~

HBK A {1 CA

CR

Ht1U

It is to be noted that the broad empty stripe, to be observed before the readings MINI>E and MISEIS on the left-hand side rim of the fragment, ind~cates that the. frag"' ment preserved the left-hand side rim of the text. Now, if we examine the preserved text of Wulfila's translation of the New Testament, we arrive at the result that the remains of these four lines in this arrangement cannot be put in any other context as in verse 11, chapter XVII of St. John's gospel and at that in the following making up of

e

JV I

rp ANrzE1

'NSWA1

One can state at first sight that this fragment also belongs to the same three lines as the former ones but somewhat more was lost between fragments 2-3 and Sb than between fragments 1 and 2-3 when the scroll was broken. lf we also fit this fragment in lines 4-6, then this passage will be the following: line 4 5 6

ATTA 'WEIH1[A] lf1 [ASYTAr IN[S IN NA-] 1 MIN I>E[I]NAM'M7[A] I>ANrzE [I A TGAFT] 1MIS EI SIJ7[AINA AI]'N SW A7[SWE WIT]

Now, if we take the other fragments in uncial script, enumerated above, one after the other and try to find their place in Christ's prayer to his Father, then we arrive at the result that one or several fragments are preserved from almost every line up tQ verse 19 of chapter XVII. If we perform this restoration work, then we obtain the following text. At first, I publish the text of Christ's prayer to his Father restored by fitting together the fragments, then I quote the Greek text in Streitberg's edition 6, and finally I give the English translation of the Gothic text: St. John's gospel, chapter XVII, verses 11-19 line 1 2 3 4 5 6 7 8 9 10 11 12 13 6

(11) [NI I>ANASEII>S IM IN I>AMMA FA-] [IRH/AU, II> I> AI IN I> AMMA FAIRH/-] rAlI' sr[ND, JAH IK DU I>US GAGGA.] ATTA 'WEIH1[A] lf1 [AS]'TAI IN7[S IN NA-] MIN JJE[I]NAM'M7[A] JJANrzE1 [I ATGAFT] 1MIS EI SIJ7[AINA AI]'N SWA7[SWE] WIT (12) [I>]AN rw7[AS MIi> I]'M7 [IN JJAM]MA [FAI-] [R]H/A[U IK FAST]AI[DA INS] IN [NAMIN] [I>EINAMMA, I>ANZEI ATG]AF[T MIS] [GAFASTAIDA, JAH AINSHUN US IM NI] [FRAQISTNODA, NIBA SA S]UN[US FRA-] [LUSTfAr[s, EI I>ATA GAMELIDO US-] [FUL]1"L7II> [WAU]RI>[I (13) II> NU DU I>US GAGGA]

Die gotische Bibel. Hrsg. van

W.STREITBERG.

[JAR] \f>,[ATA RODJA IN MANASEDAI EI] [HABAINA FAHED MEINA USFULLIDA] [IN SIS (14) IK ATGAF IM WAURJD [I>EINATA] [JAH so MANASEI>S FDAI]f))1 [A I]NS, [UNTE] ' [NI SIND US I>AMMA -FAI]RH/[AU SWASWE] [IK US I>AMMA FAIRH/AU NI IM (15) NI BIDJA] [EI US]rwIM[AIS INS US JJAMMA FAIRH/AU] [AK EI] IJ3AI7[RGAIS IM FAURA I>AMMAJ [UNSELJIN (16) US JJ AMMA FAIRH/AU] [NI SIND! SWASWE IK US I>AMMA] [FAIRH/AU NI IM (17) WEIHAI INS IN] [SUNJAI, WAURD I>EINATA SUNJA] [IST (18) SWASWE MIK INSANDIDES] [IN MANASEI>, SWAH IK INSANDI-] [DA IN]S IN 1>0 M1[ANASED (19) JA FRAM] [IM IK] WETHrA7 [MIK SILBAN, EI S-] [IJAI] NA J[AH EIS WEIHAI IN SUNJAI.]

Greek text: , ,, 11. K~l OVK~:l Eiµl EV T2;J KOCJµ!+), Kal OVTOl EV T(9 KOOµ!+) Eio(v, Kai Ey oeoeuK6:s µ01, Yva waiv ~v ~ae~s l"}µEl:: 12, On: ,l")µT]V µET' avTc:7:,v ev T(9 KOOµ!+), EYW ETTJPOVV avi"ovs EV T(9 ovoµaTL CJOV ovs OEOCuKOS µ01 ETc:7) Yva O:PlJS aVTOVS EK TOV ~o~µ~u, a\A' Y~a TfJPTJ~Tls.a~rnvs EK Tov TTovnpov. 16. 'EK Tov K6crµov ouK Eioiv, Ka8e0s ~yeu E~ TOV Kooµov OUK ElµL. 17. 'Ay(aoov miTovs EV Tl] O:AT]8Efc;t oov· 6 Myos 6 cros a~T]8E~a EOTIV 18. Ka8e0s EIJE 0:1TEOTELAas Eis TOV Kocrµov, Ka:yw 0:1TEOTEL/\a mhovs Eis : 0 ~ KOO~OV. 19. Kal VTTEP O\JTWV EYW 6:y16:l;eu EIJOVTOV, '(va WOIV Kal mhol fiy1aoµevo1 '

I

"

.

'

\

EV O/\fj8Etc;t.

English translation of the Gothic text: 11. I am no more in this world, but these are in this world and I come to thee Holy Father, keep in thine name those whom thou hast given me, that they may be one: as we two (are one). 12. When I was with them in the world, I kept them in thy name; those that thou gavest me, I hc!.ve kept, and none of them is lost, unless the. son of per-

I.Heidelberg 1906, 76. /"

267 266

among them the first. one has· its place before the restored . part . , the three o th er ones on the otl:er hand after 1t. One can also fit into. t ~ context even fragments IV4 and IVS, but bemg short · pace I . (1. to 2, at the. most 3 letters) and consisting of only one line , th e1r ~an_ als_o be 1magmed to be m other contexts. At first, I publish again the Gothic text, md1catmg the places of the fragments, then the Greek text of Streitbero's edition 7 d lastly the English translation of the Gothic text. b an

dition, that the scripture might be fulfilled. 13. But now I come to thee; and these (things) I tell in the world, that my joy might be fulfilled in them. 14. I have given them thy word and the mankind hath hated them, because they are not of this world, even as I am not of this world. 15. I do not ask that thou shouldest take them out of this world, but that thou shouldest keep them from the evil. 16. They are not of this world even as I am not of this world. 17. Sanctify them in the truth, thy word is the truth. 18. As thou hast sent me to the mankind, even so have I also sent them to the mankind. 19. And I sanctify myself for them, that they also might be sanctified in the truth.

St. Matthew's gospel VI, 7-13 line 1 2 3 4

As we can see, all fragments in uncial script could be fitted into the text of Christ's prayer to his Father, i.e. into the verses 11-19 of chapter XVII in St. John's gospel. That the lead plate contained the text of this prayer, is proved, without doubt, by the fragments consisting of several lines which determine the making up of the lines on the one hand, and cannot be fitted into other contexts on the other hand. It should be noted, however, that the fragments only consisting of one line or only containing 1 to 2 letters as AF, UN, RH/, can be fitted even into other passages. Thus AF can also be set in verses 2, 4, 7, 9, 24, UN can be fitted into verses 1, 6, 25, too, and RH/ can be put into verses 5, 6, 24 as well. Nor can one exclude even the possibility that the lead plate originally contained all the 26 verses of Christ's prayer to his Father (in chapter XVII of St .John's gospel). Because, however, any fragments consisting of several lines did not remain only from verses 11 to 19, be this assumption very likely as it may be, cannot be proved factually. Now, if we make an attempt to identify the context of the fragments written in cursive alphabet, it will be expedient to start again from the fragments consisting of several lines. Such fragments are Sa, 6a, 6b, 6c, 8, II/1, 11/2 and 11/6. Let us take at first fragment 8 whose text runs in the following way: line 1 2 3

-s 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

SA 'RJ>1 AINrA,

ro IN 1

The obvious restoration of this fragment with three lines may be the following: line 1

2 3

[ATTA UN]SNR I:>1 [U IN HIMINAM] [WEIHN]AI WA7[MO I:>EIN. QIM-] [AI I>IUfD1 IN[ASSUS I:>EINS WA-]

20 letters 18 20

This is the beginning of the Lord's Prayer in St. Matthew's gospel (VI,9 ff.). On the basis of this fragment, the length of the lines can be determined with certainty. Thus, we can easily find the places of the fragments Sa, II/1, II/2 and 11/6 as well. From

268

-(7) [BIDJANDANSUI>-I>AN NI FI-] [LUWAURDJAII:> SWASWE I>AI] [I:>IUDO. I:>UGKEII:> IM AUK] [EI IN FILUWAURDEIN SE-] [INAI ANDHAUSJAINDAN.] (8) [NI GALEIKOI:> NU I:>AIM: WAIT] [AUK ATTA IZVAR I:>IZEI JUS] [I>AURBUI>~ F]A ru1[R I:>IZEI JUS BID-] [JAIi:> IN] rA7(9) SWA N[U BIDJAII> JUS:] [ATTA UN]SNR I>1 [U IN HIMINAM,] [WEIHN]AI NrA1 [MO I>EIN. (10) QIM-] [AI I>IU]DIN[ASSUS I>EINS. WA-] [IRI:>AI WILJA I:>EI]N[S, SWE IN HI-] [MINA JAH ·ANA AIR]I>AI. (11) ri{1 [LAIF UN-] [SARANA I:>ANfA1 SI[NTEINAN GIF] UN[S HIMMADAGA. (12) JAH AFLET UNS] 1>A7[TEI SKULANS SIJAIMA, SWASWE] [JAH WEIS AFLETAM] I:>AI[M SKULAM] [UNSARAIM. (13) JAH NI BRIGGAI]S UN[S] [IN ?RAIST] 'UB1 N[JAI, AK LAUSEI UNS] [AF J:>AMMA U]ri31 [ILIN. UNTE I:>EINA] [IST I:>IUDANGARDI JAH MAHTS] [JAH WULI>US IN AIWINS. AMEN.]

19 letters 20 17 18 18 20 20 22 22 20 18 20 22 23 22 23 25 24 24 23 22 22 21

Greek text: 7. TT pocrevx6µevo1 Se µn [3aTToA.oyi]criiTE womp oi e8v1Ko1· bOKOVO"I yap OTI EV Tij TIOAVA.oyiq: OVTWV EicraKovoBi]ooVTat. 8. Mn ovv 6µ01c..)87iTE OVTOIS" oIBev yap 6 .

7

Die gotische Bibel. Hrsg. von W.STREITBERG.I. Heidelberg 1906, 6.

269

, • ... v xpeiav EXETE TTpO TOV vµos aiTnom O\JTOV. 9. Qi.he.JS ovv TTpOOEVTTaTT)p vµc.JV c.J , , ,, . , . 10. , '0, , TT , Ep n' µwv 6 EV Tois ovpavois 6:y1ao8T)Tc.J TO . El\ T) XEO 8E uµEIS aT , , , ovoµa , . _ oov. _ , ETc.J ,, r.>. , oov· yevT)0TjTc.J TO 8eAT)µa oov c.0s EV , ovpavQ t-'ao1 Ae1a ,, , _KOi ,ETTI, TT"JS, ,'(TIS• 11., Tov _ apTOV , , , - TOV ETTIOVOIOV cos iiµiv OTjµEpov. 12. Km aAI] rA1 NI> rAR1

[AI ........................................... ATTA] UN [SAR, l>U IN HIMINAM, WE~AI NAMO l>EIN, QIM]AI l>IU [DINASSUS l>EINS ........... SWE IN HIMINAM] JAH ANA [AIRI>AI .................. : ................... ] JAH AF8 _[LET UNS . , .

• -

[3a01/\EIO KOi T) uo':;,a EIS TOVS Olc.JVas.

'Aµnv. Enolish translation of the Gothic text: :::, 7. When ye pray, be not loquacious as the heathens (are), for they think that they shall be heard by their loquacity. 8. Be not ye, therefore, l~ke unto the~. For your Father knoweth what things· ye have need of, before ye ask him. 1 In this manner, therefore, pray ye: Our Father, thou in heaven, hallowed be th~ name. 10. ~ome thy kingdom, be done thy will as in heaven, so in earth, too. 11. Give us our dally bread this day. 12. And remit us our debts, as we forgive our debtors. 13. And lead us not into temptation, but deliver us from evil. For thine is the kingdom and the power and the glory for ever. Amen.

°-~

As we can see, these fragments belong to verses 8-13 of chapter VI in St. Matthew's gospel and the length of the lines alternates between 20 and 25 letters. This making up of the Jines roughly corresponds to the length of lines in the. Codex Argen-:teus which fluctuates between 21 and 26 characters on Fol. Sr (St. Matthew's gospel, chapter VI, verses 9-16). 8 From verse 7 of chapter VI in St.Matthew's gospel no fragment has been preserved, but the coherence of the contents renders probable that the text of the lead plate contained this verse as well. There remains still to examine the text of fragments 6a, 6b, 6c which contain the ends of seven coherent lines. Line 1 2 3

YSWA1SWE ]ANS SA I> YA1Nl> rAR1

4

] ... UN

5 6 7

]Al J:>IU ]JAH ANA ]JAH AF

Without doubt, this restoration represents again the text of a Lord's Prayer which cannot be identified, however, with the text of verses 7-13 of chapter VI in St. Matthew's gospel, restored above and fragments 6a, 6b, 6c cannot be fitted into it, for the phrase [AI I>IU]DIN[ASSUS] falls there on the beginning of the line, here again on the end of it, on the one hand, and because the text of the Gothic Lord's Prayer to be restored from fragments 6a, 6b, 6c, is made up in much longer lines than the variant of St. Matthew's gospel. Therefore, we have to assume that fragments 6a, 6b, 6c belong to the text of that variant of the Lord's Prayer, which is to he found in St. Luke's gospel. In this context, the difficulty arises that this part of Wulfila's Gothic translation of the New Testament is lost. This fact does not cause any insoluble problem because all Gothic phrases, corresponding to the Greek text of the Lord's Prayer in St. Luke's gospel, can be found in the preserved parts of Wulfila's Gothic translation of the New Testament and this fact permits us to restore the lost text of the Lord's Prayer in St. Luke's gospel. If we restore, however, the Gothic translation on the basis of the critical edition by E. Nestle of the Greek New Testament 9 we arrive at the result that the Gothic variant, restored from fragments 6a, 6b, 6c, does not coincide with the Gothic text of the Lord's Prayer translated from the Greek original of St. Luke's gospel. As a matter of fact, this negativ result is nc~"'-surprising. According to earlier 10 researches , it was the text of the New Testament of koine-type, influenced by the Palestinian recension of Jerusalem, established by Chrysostomus, which furnished the basis for Wulfila's Gothic translation of the gospels. This was, of course, not identical with the archetype, re~tored by the critical editions. Besides, we have to reckon even with the fact that the shorter variant of the Lord's Prayer in St. Luke's gospel became completed on the basis of the longer text of the Lord's Prayer in St. Matthew's gospel. This completion was performed to the greatest extent in the codex Bezae Cantabrigensis (D) which represents properly a mixture of the Palestinian recension of Jerusalem 11 and of the koine one. Very likely, Wulfila used some forerunner of this codex for his

9

From among these ends of lines, several can be restored with great probability.

8

W.STREITBERG: Gotisches Elementarbuch. Heidelberg 1920. Between the pages 256 and 257.

270

Novum Testamentum Graece et Latine. Imp. cur. D. EBERHARD NESTLE, nov. cur. dab. D. ERWIN NESTLE. 12 th edition, 180-18 L 10 Die gotische Bibel. Hrsg. von W. STREITBERG. I. Heidelberg 1906, XXXVIIIK ff.; W. STREITBERG: Gotisches Elementarbuch. Heidelberg 1920, 29 ff. 11 Die gotische Bibel. Hrsg. vonW. STREITBERG. I. Heidelberg 1906, XXXIX.

271

translation, be~ause it is only in this codex that the variant oi Ao~Tioi occurs ins:ead of the phrase oi eSvtKoi to be read in the text ~f_the Lord's Prayer m _St. Matthew_-s gosfavour of this assumption speaks dec1s1vely the fact that the fragment 6a offers pe1· In · · ' b fi · d · the Gothic translation [I>AI] ANI>AR[AI of the Greek phrase ot Aomot to e oun m the codex Bezae Cantabrigensis. Now, if we make the attempt to restore the Gothic translation of the variant of the Lord's Prayer in the codex Bezae Cantabrigensis, it becomes clear that it will be more detailed than the text to be restored from fragments 6a, 6b, 6c. Consequently, we have to start from the fact that the _text, restored f~om fragment 6b and the first line of fragment 6c, is absolutely certam, wherefrom 1t follows that this variant of the Lord's Prayer consisted of lines, almost two times longer thari the text in St. Matthew's gospel whose verses 8-13 of chapter VI consist of lines containing 18 to 25 letters. In dependence upon the opinion whether we regard the 37 letters of the line restored with certainty from linguistic viewpoint as minimum or as maximum instead of 7 we can reckon with a fluctuation of 10 to 14 of the number of letters in tl1is variant consisting of long lines. Because the number of letters in the lines restricts the restorations in between narrow limits, we can make an attempt to restore the Gothic translation of the variant of the Lord's Prayer in St. Luke's gospel by help of the Greek text of the codex Bezae Cantabrigensis as well as on the basis of frag,.' ments 6a 6b 6c moreover of fragment I/5 which could not be fitted into the texts of ' ' ' . the two former prayers, restored above. Then on the basis of the restored Gothic te~t, we can reconstruct even the original Greek text, furnishing the basis for the Gothic translation. Therefore, at first I publish the restored text of the lost Gothic translation of the Lord's Prayer in St. Luke's gospel, then-the original Greek text, reconstructed on the basis of the Gothic translation and lastly the English translation of the Gothic text. St. Luke's gospel, chapter XI, 2--4 line 1 JAH WARI> Mil>] 2 [I>ANEI WAS rs IN ST.ADA SUMAMMA BIDJANS: SWE I>AHAIDA, QAI>] 3 [AINS I>IZE SIPONJE IS: FRAUJA, LAISEI UNS BIDJAN] rsw A7SWE 4 [JAH IO HANNES LAISIDA SII>ONJANS SEINfA7NS . SA I>5 [AN QAI> DU IM: NI FILUWAURDJAII> SWASWE I>AI] rA7NI>rAR,,. 6 [AI JAH NI GALEIKOl> I>AIM; AK BIDJANDANS QII>II>: ATTA] UN7 [SAR, I>U IN HIMINAM, WEIHNAI NAMO I>EIN, QIM]AI I>IU8 [DINASSUS I>EINS,WAIRI>AI WILJA I>EINS SWE IN HIMINAM] JAH ANA 9 [AIRI>AI. HLAIF UNSARA.NA GIF UNS DAG.A H/AMMEH] JAH AFIO [LET UNS FRAWAURHTINS UNSAROS, UNTE WEIS SILBANS] 11 [ AFLETAM A] re L[AIM S,KULAM UNSARAIM ... ] St. Luke's gospel chapter XI, 1--4 (Greek variant, restored on the basis of the

272

Gothic text): ~ 1. Kai EYEVETO EV Tep EIVOl OVTOV EV TOlTCp TlVI lTpooevx6µevov,

w) bravoaTo, eTmv Tl) TWV µa611Twv OVTOV lTpO) avTov· Kvp1e, o(oasov T]l.10:) lTpooevxeo6at, K06G.)) KOi 'lc.vavvn) .eo(oa~EV TOV) µa8T)TCI) OVTOV. 2. ETmv OE avTOi)· 1.111 !3aTToA6yem w) oi Aomoi Kai µ~ 6µ01c.v6f}Te mhoi), a.AM Tipooevx61.1evo1 My1:Te· TTaTep i}µwv 6 EV Toi) ovpavois, ay1ao6f}Tc.v TO ovoµa oov· eMaTc.v Tl !3a01Ae(a oov, yev118iiTc.v TO Se:\nµa: oov ws EV ovpavc;, KOi ElTI TTJS yfis. 3. Tov apTOV iiµ&v { TOV ElTlOVOlOV?} o[oov iiµiv To Ka6' n1.1epav. 4. Ka\ aANA SINTEINAN 'every daily' after the words HLAIF UNSARANA into the restored text of the Lord's Prayer in St. tuke's gospel, then the number of letters in line 9 would be exceptionally high (54!). Therefore, either we can assume that Wulfila had known the interpretation by Chrysostomus of the word emovcrris in the sense 'eq,riµepos' and he regarded it as tautology beside the phrase Ka8' iiµepav used instead of oriµepov in the variant furnishing the basis for his translation and for this reason he left it untranslated or he had known and used such a variant of St. Luke's gospel, from· the text of which the phrase Tov Emovo1ov was missing. If we neglect the Gothic translation of this Greek phrase, then the number of letters in the line will be 41 which well represents the average number of letters in the lines. We can still mention that fragment II/4 (ASI) can also be fitted into line 2 of this variant of the Lord's Prayer, while the fragment I/5 (AL or LL) can only be put into line 11 of the prayer. Accordingly, after all, three important prayers of Wulfila's. Gothic New Testament were putinto grave 5 of the cemetery at Hacs-Bendekpuszta, the inhumation of a probably Ostrogoth noble: Christ's prayer to his Father from St. John's gospel, the text of tl~e Lord's Prayer from St. Matthew's gospel and the other variant of the Lord's Prayer from St. Luke's gospel. The palaeographic analysis of the preserved fragments proved that these texts had still been written before the immigration of the Ostrogoths into Italy and before the borrowing of the Latin letter S into the Gothic uncial alphabet.

273

Thus, these fragments represent the earliest monuments of Wulfila's Gothic translation of the New Testament known so far. Additional note. In the meantime the finds of the cemetery at Hacs-Bendekpuszta were published by A. KISS: Das gennanische Graberfeld von Hacs-Bendekpuszta (Westungarn) aus dem 5.-6. Jahrhundert. ·Acta Ant. Hung. 36 (1995) 275-342. At the request of colleagues reading my manuscript, I supply the restoration of the Gothic texts with data concerning the size of the lead inscribed with the Lord's Prayers, calculated on the basis of the photos (scale I : 1) : 12 x 24 cm for the Lord's Prayer from St. Mathhew's gospel, 12 x 24 cm for the Lord's Prayer from St:Luke's gospel, and 12 x 30 cm for the restored part (aboutthe half) of Christ's Prayer to his Father fonn St. John's gospel.

274

PART IV HUN GARICA

~

REMARQUES SUR LE LEXIQUE DU LATIN MEDIEVAL ET LE SUBSTRAT HONGROIS

Les recherches des monuments medievaux de la Iatinite de Hongrie ont obtenu, au cours des demieres decennies, de remarquables succes dans l'etude de la formation, du developpement et des formes stylistique de la litterature latine enHongrie. Considerables efforts peuvent egalement etre enregistres dans le domaine de l'examen de Ia formation et de I'histoire de l'orthographe hongroise. Cependant le rapport et les liens entre I'usage du latin ecrit et celui du hongrois de la Hongrie medievale, et par consequent les problemes que soulevent )'interpenetration et la coexistence de la latinite de Hongrie et du Hongrois medieval sont restes presqu'au dehors de l'interet de la recherche scientifique. Dans ce domaine deux groupes de questions ont seulement attire !'attention: d'une part le probleme de l'origine de la prononciation du latin medieval et de l'orthographe hongroise, d'autre part le probleme des mots d'emprunt et des suffixes d 'origine latine dans le hongrois. La cause de ceci sera certainement a chercher dans ce fait que ni en Hongrie, ni dans tel autre pays, ni a l'echelle de !'Europe entiere il n'existe aucun ouvrage Iexicographique du lexique du latin medieval lequel repondrait aux exigences modemes. Le sentiment de cette lacune a-t-il justement donne !'impulsion aux travaux du Novum Glossarium Mediae Latinitatis de meme que des dictionnaires de Ia latinite medievale de la plupaii des pays d'Europe. A cote de cet inconvenient objectif ii n'est pas cependantdifficile d'apercevoir un autre facteur de caractefr! theorique. 11 s'agit d'une conception qui sans etre jamais enoncee est meme invisiblement presente dans presque tousles ouvrages y relatifs: une conception qui, considerant la formation de !'usage du hongrois ecrit comme un phenomene allant de soi, ne trouve meme pas necessaire de s'en occuper. En suite de quoi le probleme de la coexistence et des rapports du latin ecrit et du hongrois parle n'a pas ete souleve non plus, alors que dans les recherches d'histoire litteraire les problemes du rappo11 entre la litterature latine medievale de Hongrie et la poesie hongroise orale jouent un role bien remarquable. Or, la formation de )'usage d'une langue ecrite ne vajamais de soi, pas meme dans les pays ou le sermo vulgaris genetiquement rapproche du latin est a un certain moment lui-meme devenu Iangue ecrite. Le systeme d'ecriture, la langue et Ia connaissance de l'ecriture sont inseparablement lies l'un a l'autre, aussi est-ii fort difficile d'emprunter n'importe quel systeme d'ecriture sans la langue ecrite, ou les deux sans des personnes sachant ecrire. Une_ longue serie d'exemples historiques nous montre que l'emprunt d'un systeme d'ecriture - d'un alphabet - sans la langue ecrite passe

277

pour tout exceptionnel. Ce qui est historiquement typique c'est que l'ecriture est chaque fois empruntee avec Ia Iangue ecrite et Ies personnes sachant ecrire, a un moment determine par le developpement social et economique. II s'ensuit de la que l'emprunt de l'usage de l'ecriture engendre normalement une dualite Iinguistique, a savoir Ia coexistence de la langue etrangere ecrite et de la langue (des langues) parlee(s) du pays en question, coexistence qui se manifeste de differentes fayon dans le domaine de l' administration, de la litterature et de la religion; Au point de vue historique ce phenomene pourra etre le mieux etudi~ dans le cas de I'empire achemenide, puisqu'apres quelques antecedents de l'Orient ancien c'est ici que se formerent la premiere fois dans l 'histoire un systeme et une pratique de chancellerie embrassant un immense territoire et dont plusieurs elements se sonttransmis -cpar I' intermediaire des Etats hellenistiques et de Rome - au moyen age. Les gouverneurs et les commandants militaires qui se trouvaient a la tete de I' administration de !'Empire achemenide etaient des Perses qui donnaient Jes ordres et les decrets en vieux-perse. L'usage de l'ecriture cuneiforme vieux-perse ne se pretait cependant pas a des fins de l'administration, et, en plus, ii n'y avait pas assez de personnes sachant l'ecrire. Pour cette raison dans les chancelleries de l'ancienne Perse on a employe la langue et l' ecriture arameenne, ce qui revient a dire que la langue de chancellerie etait l'arameen. Ces chancelleries fonctionnaient de la sorte qu'a Ia residence du Grand Roi les lettres et les ordres donnes oralement en vieux-perse furent chaque fois traduits en arameen, mis en ecrit et envoyes ainsi aux residences provinciales dans les chancelle;.; ries desquelles ces documents furent verbalement retraduits de l'arameen en vieuxperse. Le fonctionnement de ce systeme de chancellerie avait deux exigences. fondamentales. Il etait d'une part absolument necessaire qu'il y eut assez de scribes sachant l'arameen aussi bien que le vieux-perse; d'autre part i1 fallait qu'entre le vieux-perse et l'arameen, et plus precisement dans le domaine du lexique et de la structure syntactique eut lieu une compensation ou une adaptation linguistique permettant a leur tour une traduction rapide et exacte. C'est que, au point de que d'histoire des langues, la traduction n'est pas, elle non plus, une chose allant de soi. Dans les cas ou nous avons, en matiere de traduction, des exemples historique pour les rapports naissants entre deux langues, nous pouvons chaque fois constater que les premieres traductions sont tres inexactes et meme qu'elles ne peuvent guere etre considerees comme de veritables traductions. La naissance des premiers textes effectivement bilingues est precedee d'une longue periode de pratique de traduction. Ce n'est done que l'elaboration et l'emploi d'un systeme de correspondances lexicale et syntactique qui permettent l'emploi d'une pratique de chancellerie basee sur le systeme de traduction langue parlee - Jangue ecrite - Iangue parlee. Ce processus a eu en effet lieu dans ]e cas du vieux-perse et de ]'arameen. Entre ces deu_?( langues qui lexicalement et syntactiquement s'influenyaient l'une l'autre,

s'etab!it un systeme _de corr~spondances qui nou~ permet, apres 2600 ans et malgre la· conna1ssance lacunaire du v1eux-perse de retradmre en vieux-perse Ies documents arameens des chancelleries de l'ancienne Perse, c'est-a-dire. de restituer Ieur original vieux-perse qui n'a jamais existe en ecrit. Voi1a un fragment de la Iettre arameenne d' Arsama (Nr. 2, lignes 1-2), gouverneur perse de l 'Egypte:

I mn 'rsm 'I 'rtwnt §Im wsrrt sgy hwsrt lk wk't d § n"o/ mn mlk' wmny yhbl'bf;py . 2 'lym 'zyly zy pqyd hwh b g y' zyly zy b 'lyt' wtbtyt' ... I

La restitution de !'original vieux-perse:

I haca: Arsama: abiy: Artavantam: druvatamtaiy: uta: tava: vasiy: fi·astiiyiimiy: utii: nilram: dasnam: taya: haca: xsaya&iyii: uta: · hacama : diitam : Ahhapais : 2 haya : mana : badaka : haya : frataraka : atar : baga : taya : manii : hayii : Upariyaiy : utii : Adariyaiy : abava ... Traduction: I D'Arsama a Artawanta. Je t'envoie beaucoup de salut et de vigueur. Et maintenant la terre concedee qui par le roi et par moi fut donnee a.A~api, 2 a mon serviteur, qui de mes proprietes se troTlvant a Ia Haute et Basse (Egypte) etait le gouverneur ... Comme on peut le voir, le systeme ·de correspondances entre Ies deux textes est meme du point de vue de la structure syntactique parfait. La ou il s'agissait d'une categorie propre a l'ordre social de l'ancienne Perse, comme dans le cas des termes techniques diisna- 'terre concedee' et baga- 'propriete reyue de la terre royale' dans le texte arameen figurent tout simplement les termes administratifs du vieux-perse. II pouvait par consequent se produire qu'une structure syntactique entiere a passee dans le texte arameen, comme par ex. dans le cas de !'expression hd'bgw = hada abigava en vieuxperse 'grossi d' interets'. Pendant les deux siecles de !'existence de !'administration de l'ancienne Perse et malgre cette pratique developpee de chancellede !'usage de l'ecriture du vieux-perse employant l'alphabet arameen restait sans s'introduire. Les chancelleries arameennes continuaient a fonctionner. dans l'empire des Seleucides jusqu'a Ia formation du Royaume parthe, ensuite, pendant presqu'un siecle, dans ce nouve] Etat avec leur pra-

278 279

tique invariable basee sur ce systeme de traduction langue parlee - langue ecrite - langue parlee. Mais pendant cette periode aux scribes arameens succed_erent des iraniens (des parthes) qui connaissaient fort defectueusement l'arameen. Aussi ont-il commence a employer de plus en plus des mots et des expressions iraniens, et malgre que p. ex. les formules de protocole sont restees invariables, les moyens syntactiques arameens tombaient dans l'oubli. De cette sorte des mots arameens ne servaient plus qu'a la representation graphique des mots iraniens (parthes) correspondants, fait qui donna ·naissance a l'heterographie: on a ecrit des mots arameens, mais on les a lus en parthe. Voyons un exemple: MN KRM' 'wzbry 'rtbnwkn QRY (en majuscules les mots arameens, en minuscules les elements parthes) Ce texte se lisait de la fac;on suivante: ai raz uzbar Artabanukan xvont »du vignoble assujetti a la livraison appelle Artabanukan«. Cinq siecles durent se passer pour que la nouvelle tendance religieuse du manicheisme put se detacher de l'heterographie et etablir sur la base d'une nouvelle orthographe modeme l'usage des ecritures .purement, iraniennes (moyen perse, parthe et sogdienne). Ce processus pourra nous aider a divers egards a reconnaitre les problemes compliques concemant les rapports entre l'usage du latin ecrit de la Hongrie medievale et l'ancien hongrois parle. II parait premierement hors de doute que l'emprunt de l'usage du latin ecrit fut une necessite historique, etant donnee l'impossibilite d'etablir tout d'un coup l'usage de l'ancien hongrois ecrit, et quant a l'emprunt de l'usage du latin ecrit ii n'aurait pu bien entendu se passer sans l'aide des scribes etrangers. Nos premieres chartes latines furent en effet ecrites, comme on la constate, par le notaire Heribert C. qui etait arrive en Hongrie de la chancellerie de l' empereur allemand Henry II. La premiere question qui s'y impose est de savoir quel etait le. mechanisme de langue de cette nouvelle pratique de chancellerie. Les ordres qui devaient etre mis en ecrit ont ete tout probablement donnes en hongrois : le notaire, Iui, les a ou bien compris, ou bien il fallait les lui traduire. Le notaire etant de langue etrangere il est exclu qu'une traduction grammaticalement exacte ait pu jamais etre faite, et sans doute aussi chaque charte se divisait au point de vue de son langage en deux parties de qualite differente. Les protocoles initiaux et finaux, eux, ne dependaient pas d'antecedents hongrois oraux, tandis que le contenu effectif de la charte s'appuyait d'une fac;on directe ou indirecte (par traduction) sur un texte hongrois verbalement formule. Pour nos premieres chartes de langue latine un vague rapport grammatical sera seulement a supposer entre le texte latin ecrit et !'original hongrois oral. Il n'etait pas par consequent possible qu'un systeme de correspondances lexicale ou syntactique commence a se former comme dans le cas des chancelleries arameennes ·de I'Empire achemenide. La formation d'une pareille interpenetration du latin et du hongrois se trouvait

280

empechee par ceci aussi que le latin, langue Iitteraire employe'e d • d'E , ans. p1us1eurs pays urope et remontant a une remarquable tradition n, a s b. , • . u 1 qu acc1dentellement I'·rn fl uence de la langue locale parlee. Or apres Ia renaissance r . . . , caro me et ensmte a l ',epoque d e l'humamsme naissant, pour Iequel .le Iatin antique voi·re I · . . d d, I . . c ass1que servmt e mo e e, cette m~uence ne_ ~ut plus _ope~er de changements radicaux et decisifs dans la structure syntact1que du latm. La situation historique et sociale. du l t· E , · d l' . . Y • a 111 en urope etait one comp etement d1fferente de celle de l'arameen dan's l'E ·. h, , · • 1 • . · mp11 e ac ememde, pmsque angue de la chancellene, le latin s 'employait en meme te 111 ps com I d I 1· , . . me angue e a 1tterature et de la rehg10n. La formation d'un systeme de corresponda ' 1·d · , , . , , . nces pare1 1 ace m u v1eux-perse et de I arameen ava1t ete certamement entrave'e au ·. , d"f£' ss1 par 1e caractere 1 erent de la _str~ct:tre du hongrois. A cause de ces deux facteurs la formation d~ rapport entre le latm ecnt et le hongrois parle a suivi un chemin autre que celui du developpement du rapport entre le vieux-perse et l'arameen et !'usage du h · , ·t , , bl. • . , ongro1s ecn .s eta 1t, 1m auss1, d 'une fa