Texts on the Early Hungarians in the Ǧayhānī tradition (original) (raw)
TEXTS ON THE EARLY HUNGARIANS IN THE ĞAYHĀNĪ TRADITION REVIEW ARTICLE 1{ }^{1}
István Ormos
Eötvös Loránd University, Budapest
István Zimonyi: Muslim sources on the Magyars in the second half of the 9th century. The Magyar chapter of the Jayhānī tradition. (East Central and Eastern Europe in the Middle Ages, 450-1450, 35). Leiden-Boston, Brill, 2016. ISBN 978-90-04-21437-8.
The work under review deals with an account of the Magyars in the ninth and tenth centuries, mainly before their arrival and final settlement in their present-day habitat in Central Europe. It is part of a succinct description of the peoples of Eastern Europe which has come down to us in several, slightly differing versions in Arabic, Persian and Turkish. It is generally assumed that they all derive from an Arabic work composed by the wazīr Ğayhānī in Buhārā in the Sāmānid Emirate around the beginning of the tenth century, the original of which has not survived.
The present work is the English translation of a volume originally published in Hungarian in 2005 (Zimonyi 2005a). It was also published in German in 2006 (Zimonyi 2006). It deals first with the Ğayhānī tradition, presenting an account of Ğayhānī’s person, his activities, his sources and the works which preserved his account of Eastern Europe. Then follow the versions of the Magyar chapter in Arabic, Persian and Turkish, accompanied by English translations. An interpretation of the contents of the Magyar chapter follows sentence by sentence, with a detailed philological analysis, in essay form, of the questions involved. Finally, the author offers a tentative reconstruction of the original text - in English translation - with a
[1]THE ARABIST. BUDAPEST STUDIES IN ARABIC 38 (2017)
https://doi.org/10.58513/ARABIST.2017.38.6
- 1{ }^{1} There is a growing interest in the subject and it can be assumed that readers from widely differing backgrounds will consult the present publication. Unlike its counterparts normally published in these pages, the review article offered here addresses a broader audience which is often unfamiliar with Oriental languages in general and Arabic in particular. Therefore aspects and details evident to Arabists will also be explained. Space constraints allow for only a limited number of examples illustrating the phenomena discussed. An extended version with numerous examples, more detailed analyses and more references appeared in print (Ormos 2017; 60 p.). It is accessible on the internet, too (see the Bibliography below).
2{ }^{2} In this review article the transliteration system of this journal is being followed, which is different from that of Zimonyi (e.g., g/j,h/kh,g/gh\mathrm{g} / \mathrm{j}, \mathrm{h} / \mathrm{kh}, \mathrm{g} / \mathrm{gh} ). ↩︎
presentation of the various stages of its growth. The present work is in fact a sequel to an earlier joint publication by the author and the late Hansgerd Göckenjan (d. 2005) of Giessen, which treated in a similar way the whole Ğayhānī tradition, i.e. his account of all the peoples of Eastern Europe (Göckenjan & Zimonyi, Berichte). It contained the texts in German translation only, without the originals. As a matter of course, it dealt with the Magyars in considerably less detail than the work under review.
These works by Zimonyi go back ultimately to a collection encompassing all the basic texts in Oriental languages (Arabic, Persian, Turkish) dealing with the nomads in Eastern Europe who were migrating in an East-West direction in the period in question. Among these were the Magyar tribes, also moving westwards in the steppe belt until they finally reached their present-day habitat. This collection of texts (the originals, their translations accompanied by commentaries) was prepared in the 1920s by Mihály Kmoskó (d. 1931), professor of Semitic Languages at the University of Budapest. He more or less completed his manuscript but did not succeed in publishing it. This was finally achieved by Zimonyi about seventy years later, between 1997 and 2007 (Kmoskó, Mohamedán). An eminent Syriac scholar, Kmoskó dealt with relevant works in Syriac, too. His unpublished manuscript was edited by Szabolcs Felföldi, one of Zimonyi’s students (Kmoskó, Szír). Numerous translations included in the present work were actually made from Kmoskó’s Hungarian versions.
Zimonyi’s book in Hungarian and its German version generated a discussion. I published an extensive review of the Hungarian original (Ormos 2005) followed by a separate publication containing further additions (Ormos 2009). András Róna-Tas published a one-page remark on my review (Róna-Tas 2006), while Zimonyi replied to the additions (Zimonyi 2010). My reply followed in two parts (Ormos 2010a; 2010b). I also published a succinct English summary of the controversy (Ormos 2010-2011). (I published altogether 148 pages, Zimonyi 9 pages and Róna-Tas one page.) Zimonyi leaves all of them, amounting to 158 pages, unmentioned in the present work, although he has tacitly accepted some of the criticisms and modified his text accordingly.
The author of the monograph is a specialist in Altaic studies and Turcology. As is clearly shown by his treatment of the Arabic texts, he knows some Arabic but his familiarity with it is not sufficient for dealing with the texts in a sovereign way. He relies on translations, without noticing when they contain omissions or mistakes, and he is often at a loss when different translations offer different interpretations of one and the same text. Every now and then, however, he modifies the translations he is quoting, yet without indicating his intervention.
Zimonyi presents the Arabic, Persian and Turkish texts of the Ğayhānī tradition in the original with parallel English translations. In the Hungarian edition Zimonyi claimed to have presented “new critical editions” of the texts. However, it proved
demonstrable that the Arabic texts - I analyzed only them - could not be regarded as critical editions. In fact it could be shown that Zimonyi was not even familiar with the essence of a critical edition (Ormos 2010-2011:380). In the present English version, any claim to their being critical editions has been dropped and Zimonyi remains silent on the nature of his texts. In the meantime, however, he does not seem to have acquired a clear idea of the ways of publishing texts in general. Namely, he is evidently unaware that in the present case he is supposed to have presented “critical texts”, i.e. texts in the shape in which their respective authors may have written them. The major problem is the treatment of the name of the Magyars in these texts, which is nothing short of completely chaotic (see below). Another basic problem is that Zimonyi does not treat his texts in a uniform way. He copies them from a variety of editions, all prepared in different ways. He often modifies them, partly on the basis of manuscripts, but without following a clear principle. His own readings are unreliable, as are his so-called “critical apparatuses”. The result is a mess. There is one major improvement as compared to the Hungarian version. Namely, in Ibn Rusta’s text he has finally eliminated an atrocious copyist’s mistake (برناو موتهو) of the London manuscript with which he had believed to have improved on both Khvol’son and de Goeje by restoring it to his “new critical text”. In their turn, both Khvol’son and de Goeje had tacitly omitted the misplaced hamza from the waˉww \bar{a} w, correcting this form to باو موتهو, which Zimonyi evidently considered an illadvised and unjustified interference. However, another atrocious mistake still shines in Ibn Rusta’s account: النسمى as the passive participle (al-musammā) required by the context (Zimonyi 2006:34 3[Ar.1 { }^{\text {3[Ar.1 }}; 2016:38 3[Ar.1 { }^{\text {3[Ar.1 }} ). 3{ }^{3} This means that Zimonyi’s third effort within ten years at producing an acceptable text of Ibn Rusta’s relevant brief paragraph has also failed. Similar considerations are valid for the other Arabic texts as well. 4{ }^{4}
The name of the Magyars in the Ğayhānī tradition
There is one aspect of the present work which captures the reader’s attention early on: it is the name of the Magyars in the Gayhānī tradition texts. On account of the uncertainty of transmission, the Magyars appear under a wide variety of name-forms in the actual manuscripts: M.h.f.r.ya / M.ğ.f.r.ya / M.ğ.ğ.r.ya / M.ğ. '.r.ya / M.h.r.qa / Muhtariqa etc. There is a general consensus among Arabists that the correct reading
- 3{ }^{3} There are three possible explanations for this erroneous form, which does not even appear in the manuscript but represents Zimonyi’s own contribution and his own improvement on de Goeje: Zimonyi lacks a familiarity with the elements of Arabic writing, or with the basics of Arabic morphology - or both.
4{ }^{4} I did not analyze the Persian and Turkish texts. ↩︎
is Mağgariyya, which is based on the only precise form to be found in Arabic literature. 5{ }^{5} Just to name the most illustrious of these scholars: Defrémery, Khvol’son, de Goeje, Goldziher, Kunik, Rozen, Barthold, Barbier de Meynard, Kramers, Marquart, Wiet, Kmoskó, Németh, Minorsky, Czeglédy, Lewicki, Zakhoder, Martinez, Habī̄̄̄, Bosworth, Golden. In accordance with the rules of the art, they regard all the other forms as copyists’ errors, which are very common in Arabic manuscripts. Zimonyi rejects this communis opinio. The reader is eager to see what he has to offer instead. However, he can hardly believe his eyes, because Zimonyi fails to give a clear-cut, definite answer to this question, and it is impossible to find out what in his view the name of the Magyars was in the Ğayhānī tradition. In actual fact, Zimonyi does not seem to have realized the essence of the problem. In his work Ğayhānī mentioned the Magyars several times but we do not know the actual form because his work does not survive. We know his references to the Magyars only from relatively late manuscript copies of works which were copied or excerpted from his work. The name of the Magyars appears in them in a number of varieties, most or all of which must be considered scribal errors in accordance with the rules of Arabic palaeography. The modern scholar’s task is to reconstruct (from these erroneous forms) the original name - a single word! - which Ğayhān̄̄ may have used in his work. Instead of doing so, Zimonyi adopts a number of these forms, such as Mağfariyya, Muhafariyya, Mahğariyya, Mağgariyya, M.hr.f.h, M.hr.q.h, etc., in the original texts in his book. 6{ }^{6} It also happens that in one and the same text the Magyars appear under different forms, which must be considered utter nonsense. Thus, for instance, in his view Gardīzī used no fewer than four different words (Mahfariyān/ Muḥaffariyān, Maḥğariyān, Mağğariyān, Mağfariyān) indiscriminately, without any system, to denote the Magyars in his relatively brief account (Zimonyi 2016:4044; cf. Ormos 2017:10). Nobody in the possession of any amount of sound judgement will accept this absurd claim! Zimonyi bases the adoption of the form Muhaffariyya on Kmoskó’s idea that this latter form (meaning “depressed” and referring to the story of the miraculous “Depressed Land” in Arabic geographical literature) is in fact a folk etymology of the name of the Magyars (Kmoskó 1927:150-150; Zimonyi 2005b; 2016:62-66). Zimonyi is unable to present his thesis
- 5{ }^{5} Khvol’son and in his footsteps de Goeje accepted the reading Mağgariyya recorded by Abū l-Fidā` in his Taqwīm al-Buldān. However, the first to do so was Charles Defrémery who in 1849 adopted this reading for the apparently nonsensical Muhaffariyya form in a relevant passage by Bakrī. Defrémery also identified this form as the name of the Magyars. Abū l-Fidā (d. 1331) was an unoriginal, rather late compiler who, however, had access to important sources which have disappeared in the meantime. Defrémery 1849-1850:464, n. 3; 473. Cf. also Zakhoder 1962-1967: II, 48.
6{ }^{6} I have counted altogether seven varieties of this name of the Magyars in the main (Arabic, Persian and Turkish) texts of the Ğayhānī tradition as determined by Zimonyi in the present book. I left his so-called “critical apparatuses” out of the equation. - I.O. ↩︎
lucidly with all its details and complex ramifications. 7{ }^{7} There can be no doubt that the reason lies in the inconsistency of his thesis, which is not devoid of contradiction. The result amounts to total chaos. 8{ }^{8} One of the troubles with Zimonyi’s thesis is that he is unable to assess the degree of significance, in other words the relative value of the various, often serious, copyists’ errors that Arabic manuscripts abound in, as he has never worked with Arabic manuscripts. 9{ }^{9} He attributes great significance to forms originating with uneducated copyists, who sometimes did not even know Arabic properly, because they were Persians or Turks by birth. 10{ }^{10} In its present form, this section of the book gives the impression of an ignorant dilettante helplessly erring on the one hand among copyists’ errors (which he imagines to be endowed with arcane meanings) and on the other among his own contradictory statements, getting completely lost in the ensuing confusion. The only relief I can feel in this respect is to see that Zimonyi has been persuaded to abandon his Arabic etymologies of the name Maßgar and its various manuscript forms. In the Hungarian and German versions of the present work he listed among them mağfar, “an impediment to venery, a cause of diminishing the seminal fluid; anti-venereal food”, without offering any explanation as to why on earth the Arabs should have named the Magyars after “a cause of diminishing the seminal fluid” or a food that inhibits sexual activity (Zimonyi 2005a:54; 2006:53-54; Ormos 2005:745; 2010-2011:384385). It was regrettable that Zimonyi did not even feel the necessity to justify such a weird claim.
There is no relationship between the Depressed Land and the Magyars. Zimonyi is unable to adduce even a single instance from Arabic literature to prove his thesis: no Arab or Muslim author ever mentions it. The originator of this thesis, Kmoskó, was unable to produce a single instance of it, either: he merely referred to what the Arabs “might have thought”. The Arabs left us an immense literary legacy. There can be no doubt that somebody would have mentioned it if it had ever occurred to anyone. There is a further serious difficulty with this claim. Namely, that even if it
- 7{ }^{7} The summary on p. 66, for instance, is vague, using the verb “may” in key positions: it is a collection of suppositions and statements lacking any foundation and with many internal contradictions. In addition, they cannot be always harmonized with statements made elsewhere in his book.
8{ }^{8} For a detailed analysis of this subject see Ormos 2017:9-11, 48-51.
9{ }^{9} Indeed, Zimonyi compared the few lines of some of his texts, which were available to him in printed editions, with the relevant manuscripts. However, this can hardly be regarded as serious independent activity comparable to working on a previously unknown manuscript with the aim of making sense of a text with few and in some cases misplaced diacritical dots or none at all.
10{ }^{10} Khvol’son writes that the copyist of Ibn Rusta’s London manuscript evidently did not understand everything he was copying, because as a Persian by birth he never learnt Arabic properly (Ibn-Dasta, Izvestiya 10). ↩︎
existed, the popular etymology Muhaffariyya, “the people of the Depressed Land”, would only work in Arabic, but not in Persian and Turkish, where authors would have to explain it to their readers. There are seven authors in the Ğayhānī tradition who write in Persian and Turkish but none does so!
Zimonyi discusses the phonetic aspects of the name Mağgariyya, too, without being aware that the phoneme g˘\breve{g} represented by the letter g˘iˉm\breve{g} \bar{i} m, which is of central importance in our case, is perhaps the most unstable member of the phonemic inventory of Arabic with a wide variety of manifestations (Cf. Ormos 20102011:388-389). 11{ }^{11}
Translations
Zimonyi’s present translations of the Arabic texts of the Ğayhānī tradition contain numerous inaccuracies and errors. With a few exceptions, the texts were not translated from the originals into English directly, but came down through one or even two intermediary versions. It is also clear that one or perhaps more persons undertook a stylistic revision of the English translations without consulting the original texts. The adoption of such a multi-stage process is not devoid of problems. Even when stylistically good, the results are often inaccurate, free paraphrases of the original texts, which contain numerous errors.
Philological analyses (Essays)
Zimonyi’s philological analyses of the textual passages one by one are of varying interest. On the subject of the Magyars’ habitat in the vicinity of the Black Sea, he presents a twenty-eight-page essay on seas in Arab and Muslim geographical literature in general (Zimonyi 2016:202-230). For the purpose of the present book it would have sufficed to offer a summary of the information that is relevant to the book’s subject on half a page or one page at most, since there are only three seas of interest here: the Caspian, the Black Sea (with the Sea of Azov) and the Mediterranean. Another possibility would have been to write an exhaustive monograph on the subject. What we have instead, are long and difficult passages from geographical works where the textual transmission is problematic and thus the texts display many variants. This is because the Arabs’ and Muslims’ knowledge of the seas was quite vague and controversial at the time. Therefore the texts are in need of extensive commentaries if any use is to be made of them. However, commentaries are few and meagre here. The reader acutely misses a fruitful dialogue with some
- 11{ }^{11} For a few bibliographical items for further orientation, see Ormos 2009:1143, n. 57. ↩︎
important scholarly literature on the subject, too (Beylis 1962; Kalinina, Dzhakson, Podosinov, Konovalova 2007, especially Kalinina 2007). 12{ }^{12} In addition, the whole section is in awkward English, so that reading it is something of an ordeal. I have serious doubts that anyone besides me will ever read it from beginning to end.
Zimonyi’s long discussion of tents and his remarks on the subject elsewhere are severely handicapped by the terminological confusion they display (Zimonyi 2016:139-160). Several Arabic and Persian words (qubba, hayma, bayt, hargāh) occur in the relevant texts referring to “tents”, and several English equivalents (dome, tent, house, yurt, felt-huts, etc.) are used to denote them in the English texts. In addition to his own text, Zimonyi quotes a number of English translations from various scholars, who all use these words in their own particular ways and Zimonyi keeps the original wording in each case. The result is that a given Arabic or Persian word has different equivalents in English in the various translations, while one and the same English word or expression stands for different Arabic or Persian forms. In the ensuing confusion the helpless reader is totally lost, unable to guess what these words exactly mean and who writes exactly what.
Zimonyi offers an essay on the fortresses which the Slavs built against the Magyars according to Gardīz̄̄. Its central piece is an account of Slav fortressbuilding technique as related by the Andalusian traveller Ibrāhīm ibn Ya’qūb, who - as Zimonyi explicitly mentions - “also visited Prague”. This famous traveller hailing from Tortosa in Catalonia visited many places all over Europe in the second half of the tenth century, e.g., Utrecht, Tours, Verdun, Rome, Pavia, Verona, Prague, Fulda, Mainz, Schleswig, Dorf Mecklenburg, Schwerin, Nienburg (Saale). 13{ }^{13} If Zimonyi singles out Prague from all the places Ibrāhīm mentioned in his travelogue, his readers will inevitably conclude that the building technique Zimonyi is quoting refers to this famous Slav city. However, this is not the case. It is now well known that Ibrāhīm’s account offers an astonishingly precise description of a particular building technique applied by Slavs living in the vicinity of the Baltic Sea in the area of present-day northern Germany before it was invaded and conquered by Germanic tribes. It has also been convincingly demonstrated that the account in question refers in fact to two Slav fortresses: Michelenburg-Mecklenburg in the vicinity of modern Wismar and the earlier Slav fortress on the site of modern Schwerin. It is also known that the Slavs did not bring with them a common building technique when they dispersed from their original habitat. Instead, each tribe developed its own technique
- 12{ }^{12} Beylis 1962 deals with the Black Sea and the Sea of Azov in Arabic sources, while Kalinina 2007 is dedicated to the rivers, seas and lakes of Eastern Europe in Muslim sources. Kalinina, Dzhakson, Podosinov, Konovalova 2007 examines the waterways of Eastern Europe in antique and medieval sources. These works came out long before the present book.
13{ }^{13} These are the modern names of these places. Ibrāhīm visited several localities which cannot be identified. ↩︎
in close interaction with its new neighbours. Thus it is evident that the building technique used by Slavs in modern northern Germany in the second half of the tenth century has no relevance to fortresses built by Slavs against the Magyars in Eastern Europe at a distance of one thousand miles to the east approximately one century earlier (Zimonyi 2016:364-365. Cf. Ormos 2010-2011:392-394). 14{ }^{14}
Ibn Rusta informs his readers in a famous passage that at some earlier date the Khazars surrounded themselves with a moat as protection against the Magyars and other peoples. According to the generally accepted interpretation, this passage refers to the building of the fortress of Sarkel on the Lower Don, which is known from Byzantine sources. Zimonyi first addressed this passage in 1996, declaring it to be out of the question that the Khazars would have defended themselves by a moat around a fortress in the steppe, and that therefore it was impossible to establish any connection between this passage and the Magyars. Rather, Zimonyi declared, the whole passage was a literary topos relating to the famous Battle of the Moat at Medina in 627, in which the Prophet Muhammad played an outstanding role (Zimonyi 1996:57). I pointed out in a review at the time that it was difficult to see why the Khazars could not have constructed a moat around a fortress in a plain. After all, moats were usually constructed around fortresses located in plains and not on mountain peaks (Ormos 1996-2002:282-283). Zimonyi’s statement was all the more remarkable because he was living in the city of Szeged in southern Hungary, a location which is geographically strikingly similar to Sarkel. Namely, Szeged lies on the river Tisza in the Great Hungarian Plain, and in medieval times a fortress was built on the banks of the river surrounded by a moat, which was connected to the river and filled with its water as an additional defensive measure. When he wrote the Hungarian original of the present book, Zimonyi was unaware that extensive archaeological excavations had been carried out on the site of Sarkel before the Tsimlyansk Reservoir waters submerged it in 1952. Indeed, both moat and rampart were found. In the interim, Zimonyi seems to have been informed of these facts. Yet he does not give here an adequate account of the present state of our knowledge concerning this question, but gets lost in unimportant details as well as offering an account of the Battle of the Moat, which is totally out of context here. 15{ }^{15}
Ibn Rusta mentions in a famous passage that the Magyars regularly conduct raiding parties against the Slavs, seizing captives from them whom they take to a Byzantine port, trading them with the local residents for various luxury articles (Ibn
- 14{ }^{14} I am not happy with the use of the word “castles” by Zimonyi in this context, because it sounds anachronistic to me. Perhaps “fortress” or “stronghold” would describe better the defensive structures which the Slavs built against the Magyars in areas bordering on the South Russian steppe in the ninth century.
15{ }^{15} It is evidently a remnant of his wholly untenable earlier thesis that the reference to Sarkel in Ibn Rusta is in fact a literary topos (cf. Ormos 2010-2011:390-392). ↩︎
Rosteh, AA 'lāk 142nih −1435142^{\text {nih }}-143^{5} ). The name of the port appears as Karh in de Goeje’s critical edition and its identification is hotly debated among specialists, although the majority tends to agree that it refers to the city of Kerch in the Crimea. In an essay on this problem Zimonyi repeats the argument which he adopted from one of his students, Szabolcs Polgár. This argument was based on Polgár’s claim that there is a common noun, karh, of Aramaic origin, meaning “city”, “town”, in Arabic. Thus the sentence “they take them to the Byzantine port which is called Karh” simply means that “they take them to the town”. Thus the word could refer to any town in the area. Without entering into the moot question of the identification of this port, I pointed out at the time that there was no such common word in Arabic. Polgár and Zimonyi seem to have overlooked the fact that the Arabic sentence is unequivocal in indicating the name of the given port (yuqālu lahu Karḥ; “[which] is called Karḥ”). In addition, it is hardly believable that this port should possess an Arabic name, because the local population did not speak Arabic and the Arabs living in distant lands had no particular interest in it. This untenable theory is repeated here in a rather vague and scarcely comprehensible way.
Further considerations
Zimonyi repeatedly refers to the Hungarian chronicler “Simonis de Kéza” (thirteenth century). However, the correct form is “Simon de Kéza”. Zimonyi is not aware that the form he regularly uses is the genitive of the name, which appears on the title page of the relevant printed edition in accordance with accepted practice for editions of Latin and Greek authors.
One of the most important and at the same time most difficult texts treated here is Gardīzī’s version, because we have only two, relatively late and corrupt manuscripts at our disposal. It is a serious shortcoming of the present book that Zimonyi did not make use of the new critical edition by Raḥīm Riḍāzāda Malik, which came out in Teheran in 2005, that is eleven years before the publication of the present work (Gardīzī, Zayn). Zimonyi appears to be unaware of the existence of this important publication although he might have read about it in Bosworth’s preface to his translation of Gardīzī’s work, too, which he seems to have consulted (Gardīzī, Ornament). 16{ }^{16}
- 16{ }^{16} I am indebted to Ēva Jeremiās for drawing my attention to this new edition and for putting it at my disposal. Bosworth says it does not supersede Habībī’s earlier critical edition, yet he also mentions that he has not been able to compare the two texts carefully by the time of writing, and on occasions he also quotes better readings by Riḍāzāda Malik. Gardīzī, Ornament 8. Cf. also ibid., 116, n. 27; 117, n. 6. This means that it should have been consulted by all means. I have briefly checked the account on the Magyars and I have found one alternative reading worth of consideration. ↩︎
One might ask: what relevance do the many minor details have to early Hungarian history which I subject to criticism? Do they affect the overall importance of Zimonyi’s overarching work? We may retort: Of course, such minor details do not affect the great strands of Magyar history. But why does Zimonyi treat them at all then? The present work is not of a theoretical nature, offering a new overall view of early Magyar history, where a few minor details may not count and may not affect the validity of an entirely new theory. The book is of a philological nature, discussing many such questions of detail (in actual fact, it is a collection of such details) which, though important in themselves, do not add up to a qualitatively higher entity. It is precisely these details for which readers will consult it. And if those very details are not reliable, then the whole work loses its viability.
At the end of his work Zimonyi offers a tentative reconstruction of the original text of Ğayhānī’s chapter on the Magyars in English translation. This is an endeavour that obviously suggests itself, yet it is at the same time highly problematic, especially if we consider all the previous - often quite controversial - observations and commentaries that have been made on the subject. The result is no less problematic. First of all, it is a problem of a theoretical nature that Zimonyi should be undertaking this on the Magyar chapter alone, treating it as an independent unit, whereas it in fact forms part of a greater corpus, Ğayhānī’s account of the peoples of Eastern Europe. Thus it stands to reason that any effort at a reconstruction would first have to consider the wider context, i.e. Ğayhānī’s whole account. There are many cross-references among his data on these peoples! It is only on the basis of the findings of such an undertaking that any reconstruction of the Magyar chapter can be considered with a view to special features. Second, there are many more details to analyze and elucidate before such an attempt can be undertaken.
English style
In general, I regard it as inappropriate that persons who are not native speakers of a given language should comment on the style of a publication in that idiom. However, in the present case I cannot refrain from infringing this rule, because its linguistic shape is an essential feature of the book under review, deeply affecting its scholarly value. Three parts can be distinguished in the book in this respect. One minor part is in idiomatic American English. However, the problem with this part is that the person who undertook the stylistic revision apparently did not check the original Arabic etc. texts but relied on intermediary versions, allowing the translations to become free paraphrases under his pen. 17{ }^{17} A good example of this approach is the following sentence from Ibn Rusta’s description of the Magyars: lahum qibäb. It can
- 17{ }^{17} His name appears in the Preface. (Zimonyi 2016:XII). ↩︎
be perfectly translated into English: “They have dome-shaped tents.” However, here we read: “They are tent-dwelling people.” (Zimonyi 2016:39). The general idea is of course correct, yet this cannot be considered an accurate translation: the words “dwelling” and “people” do not even appear in the original. Another problem is that the person who undertook the stylistic revision had little familiarity with the subject matter. Such an approach is not devoid of pitfalls, as can be shown in the sentence mentioning the Magyars in the steppe in the vicinity of the Black Sea: “The relevant section on the Magyar capital must have borrowed from a source which al-Ĝayhānī did not improve upon with the knowledge of his contemporaries” (Zimonyi 2016:227-228). 18{ }^{18} There is no such section. In any case, it is anachronistic to speak of a Magyar capital in the south-Russian steppe in the accepted meaning of this word. What happened? Zimonyi or his translator mixed up the English word “capital” with the German Kapitel (“chapter”) when preparing the first (rough) translation of the book. The person undertaking the stylistic revision found the result awkward and adjusted it to produce an acceptable sentence, but without being familiar with Ibn Rusta’s text or being aware that there was no Magyar capital at the time. There can be no doubt that what Zimonyi originally meant was “the Magyar chapter”. 19{ }^{19}
The second part, which constitutes the bulk of the book, was translated by someone whose English was mediocre at most. The text is often clumsy and difficult to read. Indeed there are many sentences which I could only understand by translating them into Hungarian in order to work out what the author might have had in mind. Some elementary errors: “Paragraphs 2 and 3 can be connected with one another context, as the first border of the Magyars east of the Volga is the consequence of their Turkic origin”. Recte: “Paragraphs 2 and 3 can be connected with another context, as the first border of the Magyars east of the Volga is the consequence of their Turkic origin.” This is a literal rendering of Hungarian egy másik szövegkörnyezettel, where the translator mixed up the singular indefinite
- 18{ }^{18} My italics.
19{ }^{19} Indeed, in one bibliographical item English “capital” is indicated as the equivalent of Hungarian fejezet “chapter” (Zimonyi 2016:391, line 1). English and French use the expression “false friends/faux-amis” for the phenomenon when two words of the same origin have different meanings in different languages or dialects: both the English “capital” and the German Kapitel share a common origin: Latin capit[is] (<caput) “head”. The best example of this phenomenon I have ever encountered is the Hungarian “parizer”, which is of German origin. It entered colloquial Hungarian around 1881 from the dialect of Vienna, where it meant a sort of sliced sausage and it was borrowed into Hungarian with the same meaning (Pariserwurst, with the short form Pariser). A friend of mine on a visit to Berlin wanted to display his knowledge of German when going to do some shopping and told his hosts that he wanted to buy half a kilo of Pariser. Whereupon they burst into laughter and it took some time before they could tell him that in northern Germany this word meant “condom” (probably from Pariser Brief, cf. “French letter”). ↩︎
article with the numeral “one” in Hungarian (Zimonyi 2016:367). 20{ }^{20} Elsewhere we read: “They used to travel with the luggages, tents…”. Recte: “They habitually travel with luggages, tents…”, because the sentence is in the present tense (Zimonyi 2016:160-161, n. 513). Hungarian szoktak is, as is well known, an exceptional case of a past form possessing the grammatical meaning of the present. The basic difference between “used to” and Hungarian szokott/szoktak is so well known even among Hungarians with a limited familiarity with English that the authors of a guide to avoid the typical mistakes committed by Hungarians learning English did not find it necessary to discuss it: “Used to referring to a habit or state in the past is mostly well known and causes few problems” (Doughty & Thompson 1987:130).
The third part seems to originate with the author himself, who, giving the manuscript its final shape, evidently tampered with it in many places, modifying the text or adding new sentences or expressions. The level of his English is perhaps best characterized by the sentence in the Preface in which he expresses his thanks to the series editor and the native speaker of English who undertook the stylistic revision: “I thank to him and Mikael Thompson to read my text and polishing my English version” (Zimonyi 2016:XII). The reader comes across totally unintelligible sentences every now and then, even in quotations from English sources. Zimonyi’s treatment of grammatical agreement signals a boldly innovative approach to English syntax. The innocent reader encounters unorthodox forms, even in quotations, e.g., “The bride-price [they pay] for a women is wild animals…” in a quotation from Martinez, though the singular indefinite article is of course absent in Zimonyi’s source (Zimonyi 2016:362; Martinez, Chapters 127). Elsewhere we read of " aa historical phenomena" (Zimonyi 2016:67). Further examples: "The Slavic-Magyar relations is discussed … (Zimonyi 2016:309). “Khazars merchants were active among …” (Zimonyi 2016:314); “… the death of the three brothers (the legendary founder of cities Kiy, Shchek and Khoriv) …” (Zimonyi 2016:315); recte: “… the death of the three brothers (the legendary city founders Kiy, Shchek and Khoriv) …”. It may not be evident from Zimonyi’s rendering that the text is about three brothers who founded one city, Kiev. In one place Zimonyi mentions Ibn Rusta’s chapters on the Khazars (Zimonyi 2016:28). In actual fact, there is only one such chapter. Thus the last example is possibly another case of the erroneous use of the plural. Usually the reader can quickly work out what went wrong, but this is not always the case: “The place in which the Turks used formerly to be is called after the names of the river that run through it, Etel and Kuzu, and in it the Pechenegs live now” (Zimonyi 2016:282). Now, is Constantine Porphyrogenitus speaking of one river or two rivers? This is an important question! One cannot guess: the answer can be found out only
- 20{ }^{20} We disregard here the awkward construction of the sentence in general. It takes some time to work out what Zimonyi actually wants to say. ↩︎
if one looks up Zimonyi’s source, which he was unable to copy correctly: " after the names of the river that runs through it…" (Constantine, De administrando 117). 21{ }^{21} In one place the reader is captivated by the idea of a number of miraculous fish “that can be red”. There is a strand of medieval Arabic geographical literature that abounds in all sorts of miraculous beings and phenomena ('ağā’ib wa-ğarā’ib), therefore the idea seems to fit into the context. Before the reader’s eyes the image of changeant fish is conjured up, fish resembling fabrics with changing colours and hues: “There then follows a story of miraculous, meat-giving fish that can be red which were sent to the peoples of Gog and Magog to feed them.” But alas! A cursory check of Zimonyi’s source reveals that there is only one fish and the source says nothing about its colour: Zimonyi simply misspelt “read” as “red” in an otherwise infelicitous sentence (Zimonyi 2016:65).
The sheer quantity of misprints, orthographical and grammatical errors in the book under review is horrendous. I cannot remember ever having come across a publication which contained even a fraction of the number found here. 22{ }^{22} To publish anything in such a condition is an insult to the reader. It is a disappointment to see that we have reached an age when a publishing house such as Brill, formerly of such repute, apparently sends a manuscript to the printer without anybody having read it. It is beyond a doubt that the author has done a formidable amount of work, especially in view of his insufficient familiarity with most of the languages involved. Yet the volume of the work he accomplished was not commensurate with the task he had set himself. He miscalculated, gravely underestimating the amount of work to be done and the difficulties inherent in the task ahead. In actual fact, much more work needs to be done in terms of carefully elaborating and clarifying many details before such a comprehensive treatment of the subject can be attempted with any reasonable prospect of success. This was a premature undertaking, ill-conceived and misbegotten. We can state that the book under review is in general utterly inaccurate and unreliable. No piece of information can be trusted unless the reader checks it for himself. In assessing the present book, the words of Mihály Kmoskó may be quoted, which he wrote in another context in 1927: "Most of our specialists in the early history of the Magyars will be familiar with the so-called Oriental sources, i.e. the relevant places in the works of Arab and Persian authors, on the basis of the present publication deluding themselves in the false hope that the heuristic part of the scholarly work pertaining to the Oriental sources has been definitely completed once and for all and there is nothing left to be done. Yet in actual fact the situation is such
- 21{ }^{21} On this problematic passage, see Moravcsik, Fontes 47-48, n. 37.
22{ }^{22} It is not easy to differentiate between misprints and grammatical errors. At first I was inclined to regard most unorthodox forms as misprints. However, later on I came to the conclusion that Zimonyi’s English was simply miserable. ↩︎
that we have to start everything from scratch again" (Kmoskó 1927:149; with slight modifications). 23{ }^{23}
REFERENCES
A. Primary Sources
Constantine, De administrando = Constantine Porphyrogenitus, De administrando imperio. Edited by Gyula Moravcsik, transl. by R[omilly] J[ames] H[eald] Jenkins, new, revised ed. (Corpus Fontium Historiae Byzantinae, I). Washington, D. C.: Dumbarton Oaks Center for Byzantine Studies, 1967; second impression 1985.
Defrémery, Charles. 1849-50. “Fragments de géographes et d’historiens arabes et persans inédits, relatifs aux anciens peuples du Caucase et de la Russie méridionale”. Journal Asiatique 4. ser., 13.457-522; 16.50-75, 153-201.
Gardīzī, Ornament = Abū Sa īd 'Abd al-Ḥayy Gardīzī, The ornament of histories. A history of the eastern Islamic lands AD 650-1041. The original text of Abū Sa īd 'Abd al-Hayy Gardī̄̄. Transl. and edited by C[lifford] Edmund Bosworth. (I. B. Tauris and BIPS [= British Institute of Persian Studies] Persian Studies Series, 4.). London-New York: I. B. Tauris, 2011.
Gardīzī, Zayn = Abū Sa īd 'Abd al-Ḥayy ibn Daḥḥāk ibn Maḥmūd Gardīzī, Zayn alahbār. Edited by Raḥīm Riḍāzāda Malik. Teheran: Anğuman-i Āṭār va Mafāḥiri Farhangī, 1384 [2005].
Göckenjan & Zimonyi, Berichte = Hansgerd Göckenjan & István Zimonyi, Orientalische Berichte über die Völker Osteuropas und Zentralasiens im Mittelalter. Die Ğayhānī-Tradition (Ibn Rusta, Gardī̄̄̄, Hudūd al-'Ālam, al-Bakrī und alMarwazī). (Veröffentlichungen der Societas Uralo-Altaica, 54). Wiesbaden: Harrassowitz, 2001.
Ibn-Dasta, Izvestiya = Abu-Ali Akhmed ben Omar Ibn-Dasta, Izvestiya o khozarakh, burtasakh, bolgarakh, mad’yarakh, slavyanakh i russakh. Edited by D[aniil] A[vramovich] Khvol’son. St. Petersburg: Kozhachikov i Eggers, 1869.
Ibn Rosteh, A 'lāk = Abū Alī Ahmed ibn Omar Ibn Rosteh, Kitāb al-a 'lāk an-nafisa VII. Edited by M[ichael] J[an] de Goeje. (Bibliotheca Geographorum Arabicorum, VII). Leiden: E. J. Brill, 1892.
Kmoskó, Mohamedàn = Mihály Kmoskó, Mohamedàn írók a steppe népeiről. Földrajzi irodalom [= Muhammadan authors on the peoples of the steppe. Geographical literature] I/1-3. Edited by István Zimonyi. (Magyar Östörténeti
- 23{ }^{23} Kmoskó referred here to Kuun, Kútfök. ↩︎
Könyvtár [= Library on Early Magyar History], 10, 13, 23). Budapest: Balassi, 1997-2007.
Kmoskó, Szír = Mihály Kmoskó, Szír írók a steppe népeiről [= Syriac authors on the peoples of the steppe]. Edited by Szabolcs Felföldi. (Magyar Östörténeti Könyvtár [= Library on Early Magyar History], 20). Budapest: Balassi, 2004.
Kuun, Kútfök = “Keleti kútfők [= Oriental sources] coll., transl., comm. by Count Géza Kuun”. A magyar honfoglalás kútföi. [= Sources on the Hungarian Conquest]. Edited by Gyula Pauler & Sándor Szilágyi. 137-284. Budapest: Magyar Tudományos Akadémia, 1900.
Martinez, Chapters = A[rsenio] P. Martinez, “Gardízí’s two chapters on the Turks”. Archivum Eurasiae Medii Aevi 2 (1982) 109-217.
Moravcsik, Fontes = Gyula Moravcsik, Az Árpád-kori magyar történet bizánci forrásai. Fontes byzantini historiae hungaricae aevo ducum et regum ex stirpe Árpád descendentium. Budapest: Akadémiai Kiadó, 1984.
B. Secondary Sources
Beylis, V[olf] M[endelevich]. 1962. “Svedeniya o Chernom more v sochineniyakh arabskikh geografov IX-X vv”. In: Blizhniy i Sredniy vostok. Sbornik stat’ey. 2128. Moscow: Izdatel’stvo vostochnoy literaturi.
Doughty, Susan & Geoff Thompson. 1987. Problem English. A practical guide for Hungarian learners of English. 4th ed. Budapest: Tankönyvkiadó.
Kalinina, Tat’yana. 2007. "Vodniye puti soobshcheniya Vostochnoy Evropi v predstavleniyakh arabo-persidskikh avtorov IX-X vv. In: Ead.; Dzhakson, Tat’yana; Podosinov, Aleksandr Vasil’evich; Konovalova, Irina. «Russkaya reka»: Rechniye puti Vostochnoy Evropi v antichnoy i srednevekovoy geografii. 45-76. Moscow: Znak.
Kalinina, Tat’yana, Tat’yana Dzhakson, Aleksandr Vasil’evich Podosinov & Irina Konovalova. 2007. «Russkaya reka»: Rechniye puti Vostochnoy Evropi v antichnoy i srednevekovoy geografii. Moscow: Znak.
Kmoskó, Mihály. 1927. “Gardízí a törökökről [= Gardízí on the Turks]”. Századok 61.149-171.
Nyitrai, István. 1996. “A magyar őstörténet perzsa nyelvű forrásai” [= Persian sources on early Magyar history] A honfoglaláskor írott forrásai [= Written sources on the age of the Conquest]. Edited by László Kovács, László Veszprémy. (A honfoglalásról sok szemmel [= The Conquest as Seen from Many Perspectives], II.). 61-76. Budapest: Balassi.
Ormos, István. 1996-2002. [Review of] A honfoglaláskor írott forrásai [= Written sources on the age of the Conquest]. Edited by László Kovács, László Veszprémy. (A honfoglalásról sok szemmel [= The Conquest as Seen from Many
Perspectives], II.). Budapest: Balassi. 281-285. Keletkutatás Autumn 1996Spring 2002.
2005. "A magyar őstörténet arab forrásainak újabb irodalma. Kmoskó Mihály, Hansgerd Göckenjan és Zimonyi István müvei [=Recent literature on the early history of the Magyars. Works by Mihály Kmoskó, Hansgerd Göckenjan and István Zimonyi]. 733-781. Hadtörténelmi Közlemények 118. (http://militaria.hu/hadtorteneti-intezet-es-muzeum/hadtortenelmi-kozlemenyekletoltes)
\qquad 2009. “Kiegészítések ‘A magyar őstörténet arab forrásainak újabb irodalma. Kmoskó Mihály, Hansgerd Göckenjan és Zimonyi István müvei’ címủ irásomhoz [=Additions to my publication ‘Recent literature on the early history of the Magyars. The works of Mihály Kmoskó, Hansgerd Göckenjan and István Zimonyi’].” 1129-1157. Hadtörténelmi Közlemények 122. (http://militaria.hu/ hadtorteneti-intezet-es-muzeum/hadtortenelmi-kozlemenyek-letoltes)
\qquad 2010a. "Rövid viszontválasz Zimonyi Istvánnak [=A brief answer to István Zimonyi’s reply]. 699-700. Hadtörténelmi Közlemények 123. (http://militaria.hu/hadtorteneti-intezet-es-muzeum/hadtortenelmi-kozlemenyekletoltes)
\qquad 2010b. Adalékok a magyar östörténet arab forrásainak kutatásához. Tüzetes viszontválasz Zimonyi Istvánnak [= Contributions to the study of the Arabic sources on early Magyar history. An exhaustive answer to István Zimonyi’s reply]. Budapest: [author]. (https://eltearabszak.hu/kiadvanyok/; https://elte.aca demia.edu/IstvanOrmos)
\qquad 2010-2011. “Remarks on the Islamic sources on the Hungarians in the ninth and tenth centuries.” Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften 19.377-395. (https://elte.academia.edu/IstvanOrmos)
\qquad 2017. The early Magyars in the Jayhānī tradition. Reflections on István Zimonyi’s book “Muslim sources on the Magyars”. Budapest: [author]. (https://eltearabszak.hu/kiadvanyok; https://elte.academia.edu/IstvanOrmos)
Róna-Tas, András. 2006. “Észrevételek Ormos István ‘A magyar őstörténet arab forrásainak újabb irodalma’ című írásához [= Remarks on István Ormos’s publication ‘Recent literature on the early history of the Magyars’].” 120. Hadtörténelmi Közlemények 119. (http://militaria.hu/hadtorteneti-intezet-esmuzeum/hadtortenelmi-kozlemenyek-letoltes)
Zakhoder, B[oris] N[ikolaevich]. 1962-1967. Kaspiyskiy svod svedeniy o Vostochnoy Evrope. Moscow: Izdatel’stvo Vostochnoy Literatury - Izdatel’stvo Nauka (Glavnaya Redaktsiya Vostochnoy Literatury).
Zimonyi, István. 1996. “A 9. századi magyarokra vonatkozó arab források. A Dzsajháni-hagyomány [= The Arabic sources on the Hungarians of the 9th Century. The Gayhānī tradition]”. In: A honfoglaláskor írott forrásai [= Written
sources on the age of the Conquest]. Edited by László Kovács & László Veszprémy. (A honfoglalásról sok szemmel [= The Conquest as Seen from Many Perspectives], II). 49-59. Budapest: Balassi.
2005a. Muszlim források a honfoglalás elötti magyarokról. A Ğayhānīhagyomány magyar fejezete [= Muslim sources on the pre-Conquest Magyars. The Magyar chapter of the Ğayhānī tradition] (Magyar Őstörténeti Könyvtár [= Library of Early Magyar History], 22). Budapest: Balassi.
2005b. “Das eingegrabene Land. Eine arabische Volksetymologie der ungarischen Selbstbezeichnung.” Ural-Altaische Jahrbücher 19.50-64.
2006. Muslimische Quellen über die Ungarn vor der Landnahme. Das ungarische Kapitel der Ğaihānī-Tradition transl. Tibor Schäfer. (Studien zur Geschichte Ungarns, 9). Herne: Gabriele Schäfer.
2010. "Hamzától a magyar kultúra káros befolyásolásáig. Válasz Ormos István recenziójára [= From hamza to detrimental influence on Hungarian culture. A reply to István Ormos’s review]. 329-337. Hadtörténelmi Közlemények 123. (http://militaria.hu/hadtorteneti-intezet-es-muzeum/hadtortenelmi-kozlemenyekletoltes)
2016. Muslim sources on the Magyars in the second half of the 9th century. The Magyar chapter of the Jayhānī tradition. (East Central and Eastern Europe in the Middle Ages, 450-1450, 35). Leiden-Boston: Brill.