From Mahmud Kaşgari to Evliya Çelebi: Studies in Middle Turkic and Ottoman Literatures 9781463216931

Citation preview

From M a h m u d Ka§gari to Evliya Çelebi

Analecta Isisiana: Ottoman and Turkish Studies

103

Collections of thematic essays focused on specific themes of Ottoman and Turkish studies are brought together in Analecta Isisiana. These scholarly volumes address important issues throughout Turkish history, offering in a single volume the accumulated insights of a single author over a career of research on the subject.

From Mahmud Ka§gari to Evliya £elebi

Studies in Middle Turkic and Ottoman Literatures

Robert Dankoff

The Isis Press

preSS 2009

Gorgias Press LLC, 180 Centennial Ave., Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2009 by The Isis Press

All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of The Isis Press. 2009

ISBN 978-1-60724-090-7

Printed in the United States of America

Robert Dankoff grew up in Rochester, N.Y. After receiving his BA from Columbia University (1964) he served in the U.S. Peace Corps in Turkey and then went to Harvard University where he received his PhD (1971). He taught at Brandeis University and the University of Arizona before going to the University of Chicago (1979), where he is Professor Emeritus of Turkish and Islamic Studies. He currently resides in Philadelphia. His research has concentrated on linguistic and literary topics relating to Ottoman Turkish texts and to the earliest Islamic Turkish texts in Central Asia. He has published several text editions and translations, including: Wisdom of Royal Glory (Kutadgu Bilig): A Turko-Islamic Mirror for Princes (1983) and The Intimate Life of an Ottoman Statesman: Melek Ahmed Pasha (1588-1662), as portrayed in Evliya Qelebi's Book of Travels (Seyahatname) (1991). His most recent book is An Ottoman Mentality: The World of Evliya gelebi (2004).

TABLE OF CONTENTS

Part 1 : MIDDLE TURKIC 1.

2.

Qarakhanid Literature and the Beginnings of Turco-Islamic Culture. In: Central Asian Monuments (ed. Hasan B. Paksoy, Istanbul, 1992), 73-80

11

On Nature in Karakhanid Literature. Journal of Turkish Studies 4

(1980), 27-35 Three Turkic Verse Cycles Relating to Inner Asian Warfare. Harvard Ukrainian Studies 3/4, 1979-80 (= Eucharisterion Omeljan Pritsak, Part 1), 151-65 4. Inner Asian Wisdom Traditions in the Pre-Mongol Period. Journal of the American Oriental Society 101.1 (1981), 87-95 5. Kashgari on the Tribal and Kinship Organization of the Turks. Archivum Ottomanicum 4 (1972), 23-43 Kashgari on the Beliefs and Superstitions of the Turks. Journal 6. of the American Oriental Society 95.1 (1975), 68-80 7. The Alexander Romance in the Diwan Lughat at-Turk. Humaniora Islamica 1 (1973), 233-44 8. Baraq and Buraq. Central Asiatic Journal 15.2 (1971), 102-17 ... 9. Middle Turkic Vulgarisms. In: Aspects of Altaic Civilization II (ed. L. V. Clark and P. A. Draghi, Bloomington, Indiana, 1978), 59-64 10. Introduction to Wisdom of Royal Glory (Chicago, 1983) 11. Textual Problems in Kutadgu Bilig. Journal of Turkish Studies 3

19

3.

(1979), 89-99 12. Animal Traits in the Army Commander. Journal of Turkish Studies 1 (1977), 95-112 13. Some Notes on the Middle Turkic Glosses. Journal of Turkish Studies 5 (1981), 41-44

27 41 57 77 99 Ill

127 131 173 189 203

Part 2: OTTOMAN 14. The Lyric in the Romance: The Use of Ghazals in Persian and Turkish Masnavis. Journal of Near Eastern Studies 43.1 (1984), 9-25

207

6

F R O M M A H M U D K A § G A R i TO B V L i Y A

gELEBI

15. The Romance of Iskender and Gtilshah. In: Turkic Culture: and Change (ed. S. M. Akural, 1987 = Indiana continuity University Turkish Studies 6), 95-103 16. Inner and Outer Oguz in Dede Korkut. Turkish Studies Association Bulletin 6.2 (1982), 21-25 17. The Seyahatname of Evliya £elebi as a Literary Monument. Journal of Turkish Literature 2 (2005), 71 -83 18. Turkic Languages and Turkish Dialects according to Evliya (.'elebi. Altaica Osloensia: Proceedings from the 32nd Meeting of the Permanent International Altaistic Conference, ed. Bernt Brendemoen, Oslo, 1990, 89-102 19. The Languages of the World according to Evliya (,'elebi. Journal of Turkish Studies 13 (1989 = Gerhard Doerfer Festschrift), 23-32 20. Evliya £elebi on the Armenian Language of Sivas in 1650. Annual of Armenian Linguistics 4 (1983), 47-56 21. "Migdisi": An Armenian Source for the Seyahatname. Wiener

22.

23.

24.

25.

26.

27.

Zeitschrift für die Kunde des Morgenlandes 76 (1986 = Festschrift Andreas Tietze), 73-79 Marrying a Sultana: The Case of Melek Ahmed Pasha. In: Decision Making and Change in the Ottoman Empire (ed. Caesar E. Farah, Kirksville, Missouri, 1993), 169-182 An Unpublished Account of mum söndiirmek in the Seyahatname of Evliya Chelebi. In: Bektachiyya: Etudes sur l'ordre mystique des Bektachis et les groupes relevant de Hadji Bektach (ed. A. Popovic and G. Veinstein, Istanbul: Isis, 1995), 69-73 Establishing the Text of Evliya £ e l e b i ' s Seyahatname: A Critique of Recent Scholarship and Suggestions for the Future. Archivum Ottomanicum 18 (2000), 139-44 "Shall We Tear Down That Observatory?" Evliya £elebi and Philology, [unpublished English original of: "§u Rasadi Yikalim m i " Evliya £ e l e b i ve Filoloji. In: Evliya f e l e b i ve Seyahatname (ed. Nurhan Tezcan & Kadir Atlansoy, Dogu Akdeniz Üniversitesi, 2002), 99-118 Some Reflections on the Editing of Book 9 of the Seyahatname. In: Izzet Gündag Kayaoglu Hatira Kitabi: Makaleler (ed. Oktay Belli, Yiicel Dagli, M. Sinan Genim; Istanbul, 2005), 122-32 .. Some Reflections on the Editing of Book 10 of the Seyahatname. In: Journal of Turkish Studies 30/1 (2007 = In memoriam §inasi Tekin, I), 225-235

229 239 245

259

277 291

299

307

319

323

329

353

369

TABLE

OF

CONTENTS

28.

T w o Armeno-Turkish Texts: Lament for a Dead Daughter and Game of Chance. Journal of Turkish Studies 14 (1990 = Fahir iz Festschrift I), 151-162

29.

"The Story of Faris and Vena": Eremya ( x l obi's Turkish Version of an Old French Romance. Journal of Turkish Studies 26 (2002 = Essays in Honour of Barbara Flemming), I, 107-61

7

383

401

INTRODUCTION

All the studies gathered here can be considered notes toward a literary history of the Turks. Should such a history ever be written it will be based on the kind of spadework and microanalysis and interpretive sally offered here. They are, all of them, text-based and text-centered. If there is a scholarly discipline they share in it is philology, the attempt to understand texts in all their stunning detail as well as within their cultural and historical contexts. To the extent I have succeeded in this endeavor, I owe it to my teachers, Omeljan Pritsak and §inasi Tekin, and to my role-model in the field of Turkish studies, Andreas Tietze, all three of whom have only recently passed away. I dedicate this book to their memory. In writing these studies originally I was reluctant to stray too far from the texts on which they were based. I tried to steer a course between the Scylla of fashionable theorizing and the Charybdis of antiquarian pedantry. I hoped to achieve results that would be accurate and therefore both lasting and useful. The least successful of these essays, from my present point of view, is the earliest, 8. "Baraq and Buraq". In my graduate-student zeal and fresh from reading Eliade I had convinced myself — with an assist from reading Kopriilii — that shamanism was the key to understanding everything non-Islamic (and even some things that were Islamic) in the culture of the Turks. The article may serve here as a rash display of youthful learning, and perhaps as a warning to others starting out in the field. As for shamanism, my infatuation soon gave way to skepticism, as in 6. "Kashgari on the Beliefs and Superstitions of the Turks." Most of these studies are ancillary to what I conceived as my main philological task, the editing and translating of some of the major literary monuments of the Turks. These have turned out to be, first, the two major Middle Turkic or Karakhanid texts, Mahmud Kashgari's Compendium of the Turkic Dialects and Yusuf of Balasagun's Wisdom of Royal Glory, and second, one of the major Ottoman texts, Evliya £elebi's huge Book of Travels. Along the way I have delved into other texts as well, such as The Book of Dede Korkut, Ahmedi's iskendername, and Eremya Qelebi's Story of Faris and Vena in Armeno-Turkish. (To my mind, the "literary history of the Turks" cannot exclude Turkish works by Armenians or anyone else.) The articles, while roughly chronological, are organized here rather along thematic lines. There is inevitably some overlap and repetition. While I have not seen fit to abbreviate, I have occasionally expanded the text

10

F R O M M A H M U D K A § G A R I T O E V Li Y A Q E L E B I

(particularly in 21. "Migdisi") and updated the footnotes (signalled by "See now ..."). Also, I have corrected what I now see as errors; and I have tried to make the various transcription systems a bit more uniform. Especially articles 24-26, showing my evolving thoughts on textual philology with regard to Evliya £ e l e b i ' s Seyahatname, have a good deal of overlap. But especially in this case, since I changed my view of Books 9 and 10 (originally favoring ms. Y over the other two mss., subsequently convinced that all three mss. are on a level), it seemed important to leave the original articles as they were. Readers should be aware that I no longer fully subscribe to the views expressed in 24 and 25, and that my mature view of Books 9 and 10 is to be found in 26 and 27. I am grateful to Yiicel Dagli for much assistance in preparing this volume.

1. QARAKHANID LITERATURE AND THE BEGINNINGS OF TURCO-ISLAMIC CULTURE

The two major Qarakhanid literary monuments were the product of a unique moment in cultural history. The Diwan Lughat at-Turk by Mahmud al-Kashghari, probably completed in 1077, is an encyclopedic lexicon of the Turkic dialects, including citation of proverbs and poetry, with glosses and explanations in Arabic. The Qutadghu Bilig by Yusuf of Balasaghun, written in 1069, is a long didactic poem in the mirror-for-princes genre. The language which Kashgharl described and in which Yusuf composed is substantially the same language as that of the Turkic "runic" inscriptions dating from the eighth century; of the vast translation literature in Uighur Turkic, mainly of Buddhist content; and of the later efflorescence of Eastern Turkic Islamic literature known as Chaghatay, with its modern descendants, Uzbek and New Uighur. Taken together, the two monuments can be considered examples of an attempt by the Turks of Central Asia to lay the foundations for a TurcoIslamic literary culture. The Qarakhanid Turks converted to Islam in the middle of the tenth century. Unlike the Seljuks, who began their career as a band of freebooters, and the Ghaznavids, who started out as slaves, the Qarakhanids, led by their Khaqans, preserved much of their Central Asian aristocratic and cultural heritage. They traced their ancestry to the legendary hero Alp Ar Tonga, whom they identified with the arch-enemy of Iran, Afrasiyab. They cultivated Turkic language, and also continued to employ the Uighur script (which they called "Turkic" script) — a rare example of a Muslim people using a nonArabic script. By the eleventh century, while the Iranian component of Islamic culture was already well advanced, the Turkic one had yet to be created. The Qarakhanids played a cultural role for the Muslim Turks similar to that of the Samanids for the Muslim Iranians a century before. In this they again differed from the Ghaznavids and Seljuks, who both patronized Iranian and not Turkic culture. By "culture" here I intend something wider than literature. I mean specifically what is connoted by the Arabic term adab.1 Originally the word

' Sec F. Gabrieli, art. "adab" in Encyclopedia paragraph are from that article.

of Islam, New Edition. The quotations in this

12

F R O M M A H M U D K. A § G A R I T O E V L t Y A

£ELEBi

meant "custom." In early Islam it came to mean "high quality of soul, good upbringing, urbanity and courtesy." It represented an ideal corresponding to "the refining of bedouin ethics and customs as a result of Islam and the contact with foreign cultures during the first two centuries A . H . " In intellectual content, adab meant "profane culture ... based in the first place on poetry, the art of oratory, the historical and tribal traditions of the ancient Arabs, and also on the c o r r e s p o n d i n g sciences: rhetoric, g r a m m a r , lexicography, metrics." During the period of high Abbasid culture in the ninth century the concept was broadened to include non-Arab traditions as well, particularly Iranian epic and narrative, and Iranian gnomic wisdom ( a n d a r z ) , but also Indian fables and Greek philosophy. Finally, in the narrower sense of literature, adab meant belles-lettres; thus it became the basis of the term for literature (adabiyat) in several modern Islamic languages. For what concerns us here, which is the creation of a Turkic adab, we can see three outstanding elements in the Arab and Iranian adabs that served as models for the Turkic. These are, first, the mastery of language; second, the transmission of profane wisdom, particularly as attached to the royal courts; and third, pride in the national legends, customs and traditions. The Arabic philologists of the first few Islamic centuries, partially for religious reasons, made it their task to collect and record all the linguistic usages of the Arabs, especially as preserved and handed down in the poetry and proverbs of the Jahiliyya. The study and mastery of Arabic provided the basis not only of the profane culture, or adab, but also of the religious sciences, or k ilm. This might be the reason why grammatical and lexicographic scholarship lagged in the Iranian cultural sphere. From the early period we have only A s a d i ' s Lughat-i Furs, written c. 1070, with its limited aim of explaining difficult words used by Firdawsi and the other New Persian poets. W e shall see in a moment that K a s h g h a r i expresses a rather d i f f e r e n t orientation to the question of linguistic scholarship, one that hearkens back to the Islamic ideology which spurred on the Arabic philologists in their classical period. The great Iranian contribution to adab culture was the translation of the Sasanian royal traditions into a form suitable for the Islamic context. The works of Ibn al-Muqaffa' are pre-eminent here; but we may also mention the Kitab

at-Taj of pseudo-Jahiz, and the Javidan

Khirad

of Miskawaih. This

movement added a stock of Iranian andarz to the Arab amthal (proverbs); also a stock of epical and historical traditions which the chroniclers tried to coordinate with their inherited Arabian and Israilitic materials. The specific pride in Iranian, versus Arab, civilization, which had given rise to the Shu'ubiyya phenomenon, emerged triumphant with the Samanids, and is very

QARAKHANID

LITERATURE

13

clear in the Shah-nameh. It is characteristic that when al-Ghazali (d. 1111) set out, toward the end of his life, to write a mirror for princes, he chose to do so in Persian and not in Arabic. Returning to the Turks, let us briefly examine the Qarakhanid literary monuments to determine whether they can be interpreted, each in its own way, as an attempt to create a Turkic adab. In his introduction to the Diwan Lughat at-Turk, Kashghari states:1 When I saw that God Most High had caused the Sun of Fortune to rise in the Zodiac of the Turks, and set their Kingdom among the spheres of Heaven; that He called them "Turk," and gave them Rule; making them kings of the Age, and placing in their hands the reins of temporal authority; appointing them over all mankind, and directing them to the Right; that He strengthened those who are affiliated to them, and those who endeavor on their behalf; ... [then I saw that) every man of reason must attach himself to them, or else expose himself to their falling arrows. And there is no better way to approach them than by speaking their own tongue, thereby bending their ear and inclining their heart.

He then quotes the prophetic hadith: "Learn the tongue of the Turks, for their reign will be long;" and goes on to say: "if this hadith is sound ... then learning it is a religious duty; and if it is not sound, still Wisdom demands it." The peroration ends with his dedication of the work to the reigning caliph, al-Muqtadi. In explaining his methodology in drawing up his work, Kashghari writes: I have set it out according to the order of the alphabet; and adorned it with words of wisdom and elegant speech, proverbs, verses of poetry, and sentences of prose I originally intended to structure the book along the lines of al-Khalil ... in order to show that the Turkic dialects keep pace with Arabic like two horses in a race I have strewn therein examples of their verses, which they utter in their pronouncements and declarations; as well as proverbs which they coin according to the ways of wisdom, both in adversity and in felicity, and which are handed down from speaker to transmitter. And I have gathered therein much-repeated matters, and famous expressions. Thus has the book attained the utmost of excellence, and the extreme of refinement.

The immediately following section is an exposition of the "Turkic" (i.e., Uighur) script, which "is used for all documents and correspondence of the Khaqans and the Sultans, from ancient times to the present, and from

' Mahmud al-Käsgari, Compendium of the Turkic Dialects (Diwan Lugät at-Turk), edited and translated with introduction and indices by Robert Dankoff, in collaboration with James Kelly, 3 vols. (Duxbury, Mass., 1982, 1984, 1985), vol. 1, pp. 70-86.

14

F R O M M A H M U D K A § G A RI T O E V L i Y A

gELEBi

Kashghar to Upper Sin |China], encompassing all the lands of the Turks." The last sections of the introduction deal with grammar, dialectology, and linguistic geography, including the famous map. The verses which Kashgari cites are, like the proverbs, oral and anonymous. And they are all in the syllabic-counting meters of Turkic folk poetry. The scattered, isolated verses, which are cited to illustrate usage, can be grouped together in "verse cycles" — groups of stanzas sharing a common rhyming and metrical scheme and a common theme. 1 There are fifty or so such cycles, ranging from one to sixteen stanzas. A third of these falls in the class of "wisdom" poetry, and a fifth relates to warfare; the rest are concerned with love, nature, the hunt, etc. One frequently cited is a lament on the death of Alp Ar Tonga (identified with Afrasiyab). There are, in addition, two "narrative cycles" which can be pieced together from Kashghari's historical or legendary notes connected with the folk etymologies of geographical names and culture terms. All these materials relate either to Afrasiyab, or to a certain Shu, "king of the Turk," who defeated the world-conqueror Dhu'l-Qarnayn (Alexander the Great). The striking thing about both these cycles is their lack of resemblance to the Afrasiyab-Turan theme and to the Alexander romance theme as found, for example, in the Shah-nameh. Rather, we have indigenous Central Asian legends relating to Tonga Alp Ar and to King Shu as national heroes of the Turks. Thus: Afrasiyab was the Khaqan; the founder of the royal dynasty of Khans, Tegins, and Terims; and the father of Qaz, Barman, and Barsghan, who all founded cities named after them. (The city named after Qaz is Qazvin in Iran, originally Qaz Oyni meaning "Qaz's playground"!) He himself founded Kashghar (= Ordu Kand), while his residence was at Barchuq. Now none of this is known to the Iranian tradition. In the Shah-nameh. for example, Afrasiyab has a son named Shida and daughters named Manizha and Farangiz; while his residence is at Qunduz (= Gang), later at Baikand. In Kashghari's version of the Alexander romance, 2 Dhu'l-Qarnayn gives names to the three main Turkic groups of the age — the Chigil (= Qarakhanids), the Turkman (= Oghuz), and the Uighur — all of them provided with Persian (!) etymologies. He also furnishes the names for a place (Altun Qan), a title (Oga), another tribe (Qalach), and a food (Tutmach) — all these ' The task of grouping the stanzas in a coherent order was begun by Köprülü and continued by Brockelmann and, more recently, by Stebleva and by Dankoff and Kelly (see Dankoff and Kelly, vol. 1, p. 5, n. 9 and vol. 3, pp. 290-311). See also Talat Tekin, XI. Yüzyil Türk §iiri (Ankara, 1989). See R. Dankoff, "The Alexander romance in the Diwan Lughät at-Turk, " Humaniora Islamica 1 (1973), 233-44.

Q A R A K H A N I D

L I T E R A T H

RH

15

with good Turkic etymologies. But Dhu'l-Qarnayn is not the hero of the legend, for he is defeated by the Turkic king Shu, founder of the capital of Balasaghun. The idea of Alexander as the enemy reflects a pre-Islamic Iranian view (Iskandar-i mal'un); in the Shah-nameh and later 1skandar-namehs he is the invincible hero, even (under Koranic influence) a kind of Prophet. Finally, Kashghari at one point relates a ghazi legend, which in turn can be connected with four groups of verses that originally must have been part of one larger verse cycle, perhaps interspersed as songs within a prose narrative as in the later Turkish minstrel cycles of Dede Korkut and Koroglu. 1 So in the Dlwdn Lughat at-Turk we do have, in germ, Turkic epic materials. While it is a historical fact that these were not embroidered and developed in a pan-Turkic epic tradition, in the manner of the Shah-nameh, some elements did survive in different dress. Thus, some of the Alexanderromance material turns up again in the legends of Oghuz Khan. 2 Similarly the ghazi legend, relating the miraculous victory of a Muslim Turkic hero over a group of infidel Turkic tribes, survives to some extent in themes incorporated into the Tazkare-i Satuq Bughra Khan. From Kashghari's point of view, these poems and legends were only so many cultural materials, to be recorded in his Dlwdn along with data on Turkic ethnography and folklore, social organization and kinship structure, calendars, recipes, and folk remedies. He saw it as his task to present these materials in a coherent way, for he was convinced of the supremacy of the Turks in God's design, and of the need for non-Turkic Muslims to know the language and the lore of their Turkic brothers in the faith. Indeed, Kashghari succeeded in doing for the Turks what the Arabic philologists in the first centuries of Islam had done for the Arabs: namely, to organize and elucidate their linguistic, genealogic, and cultural traditions. We can say something very similar about Qutadghu Bilig:3 that Yusuf of Balasaghun attempted, with some measure of success, to establish the Central Asiatic Turkic tradition as a legitimate element within the parameters of Islamic culture, just as his counterparts from Ibn al-Muqaffa' to Firdawsi had done for the Iranian tradition. But unlike Firdawsi Yusuf took as his starting point, not the sagas and epics that were current at his time, but rather the heritage of "royal wisdom" (qutadghu bilig) preserved in Qarakhanid ruling circles, which he tried to amalgamate with the Irano-Islamic ideals of statecraft preserved in Arabic and Persian adab.

^See R. D a n k o f f , "Three Turkic Verse Cycles Relating to Inner Asian W a r f a r e , " Eucharisterion ... O. Pritsak (Harvard Ukrainian Studies III/IV, 1979-80, Part 1), 151-165. 2 S e e J. A. Boyle, "The Alexander Legend in Central Asia," Folklore 85 (1974), 217-28; and "The Alexander Romance in Central Asia," Zentralasiatische Studien 9 (1975), 265-73. "'.See R. Dankoff, "Introduction" to Yüsuf Khäss H ä j i b , Wisdom of Royal Glory (Kutadgu Bilig): A Turko-lslamic Mirror for Princes (University of Chicago Press, 1983); "Inner Asian Wisdom Traditions in the Pre-Mongol Period," Journal of the American Oriental Society 101.1 (1981), 87-95.

16

FROM

MAHMUD

K A§ G AR I TO E V Li Y A £ E L E B i

Thus as authorities for the wisdom sayings scattered throughout the text, Yusuf cites only various Turkic princes and poets, but also "an Arabic saying" (line 5809), "an Iranian sage" (line 3265), and the Sasanian king Nushirvan, the model of just sovereignty (line 290). "If you observe well," he states in the introductory portion of the work (lines 276-282), you will notice that the Turkish princes are the finest in the world. And among these Turkish princes the one of outstanding fame and glory was Tonga Alp Er. He was the choicest of men, distinguished by great wisdom and virtues manifold The Iranians call him Afrasiyab, the same who seized and pillaged their realm. The world-conqueror requires great virtue indeed, and mind and wisdom, in order to rule. The Iranians have written this all down in books — and who could understand it if it were not written down?

Among the proverbs quoted by Kashgharl in the Diwan Lughat atTurk (fol. 465) is: yas ot koymas, yalawar olmas "Fresh grass does not burn, the messenger does not die." Kashgharl goes on to say that this is so, even though his message may contain treachery or coarseness on the part of the sender. This is similar to the words of the Exalted [Koran 5:99] "It is only for the Messenger to deliver the Message."

Conceivably this Turkic proverb was in Yusuf s mind when, in an appropriate context in the Qutadghu Bilig (lines 3817-3819) he appeals to the authority of the Khan of the Turks (tiirk hani) for the following lines: yalawacqa bolmas olurn ya qi'yin esitmis sozin ci'n tagursa tilin yalawac tediikiim bu tilci turur bu tilci sozin aysa olmas qalur

"The messenger deserves neither death nor punishment, so long as he faithfully reports what he has heard. For this messenger is merely a spokesman, and when the spokesman transmits his message, he is not killed, but is left alone." What is interesting to observe here is that for Kashgharl and apparently for Yusuf as well, the Turkic proverb is on a par with the Koranic verse. The major issue in Qutadghu Bilig is the conflict between the political ideals of the community and the religious conscience of the individual. The conflict is dramatized in the form of a debate between two brothers, one of them a statesman and chief advisor to the king, the other a recluse and mystic. The statesman is called "Highly Praised" (Tk. Ogdiilmis, a translation of the

QARAKHANID

LITERATURE

17

Arabic name Muhammad), while the recluse is named "Wide Awake" (Tk. Odghurmis — cf. Ar. Yaqzán, which was already used as an allegorical name by Ibn Siná). Highly Praised knows what is best for the world's governance. In response to the king's queries, he describes the qualities and duties of the various courtiers: prince, vizier, commander, ambassador, secretary, treasurer, cook, etc. To Wide Awake, ignorant in the ways of the world, he explains how one must conduct oneself with the various classes of society: courtiers and commoners, scholars, physicians, diviners, astrologers, poets, farmers merchants, stockbreeders, craftsmen, beggars. He also gives advice on how to choose a wife, how to raise children, how to behave as host and as guest, and how to interpret dreams. He is the perfect adlb, the personification of worldly wisdom. Wide Awake knows what is ultimately best for a man's soul. In pursuit of complete devotion to God, he has adopted a life of poverty, renunciation, and solitude. He personifies, as Yüsuf tells us (line 357), 'aqibet: Man's Last End. In the mirror-for-princes scheme, Wide Awake provides a leaven of otherworldly goals and ideals, without which the ruler's life would be vain. The ultimate reconciliation of the brothers, in the king's presence, demonstrates one of the deep rooted themes of the Irano-Islamic statecraft tradition: that just sovereignty and right religion are twins, born of the same womb, and cannot be separated. I have tried to show that the two major Qarakhanid literary monuments, judged on their own terms, were successful in laying a foundation for a Turkic adab\ the one in the areas of linguistic scholarship and the recording of national lore; the other in the area of royal wisdom. But in terms of the historical development of Turkic culture, the efforts of Kashghari and of Yüsuf of Balasaghun were practically fruitless. No Turkic Firdawsi came along to celebrate the pre-Islamic exploits of Alp Ár Tonga. The lively epic and historiographic traditions developed later by the Ottomans in the west and by the Timurids in the east were entirely based, on the one hand, upon the Oghuz settlement of Anatolia, and on the other hand, upon the exploits of the Chingisids, Timurids, and Shaybanids. As far as we know, only one Islamic historian ever attempted to incorporate Kashgharl's legends about King Shu and Dhu'l-Qarnayn into a grander scheme — viz., Badraddin al-'Ayni in the first volume of his thirtyvolume world history, written in 1422.1 As for Qutadghu Bilig, the only one to quote it in later Turkic literature, again so far as we know, was Rabghuzi in his Qisas al-Anbiya, written in 1310, where we find a paraphrase of ^See Dankoff and Kelly, vol. 1, pp. 19-23.

18

F R O M M A H M U D K A § G A R I TO E V L I Y A

gELEBI

Yusuf's, chapter "On the Seven Planets and the Twelve Constellations." 1 It is true, judging by the three extant ms. copies, that Qutadghu

Bilig did enjoy a

certain vogue as late as the Timurid period. But it never served as the basis for an elaboration, or even an imitation; in contrast, say, to the Persian nameh,

Qabus-

a mirror for princes written in 1082, which was translated into

Ottoman Turkish no less than five times in the fourteenth and fifteenth centuries. To conclude: if I am right that Diwan Lughat

at-Turk and

Qutadghu

Bilig represent the budding of a Turkic adab; still, in terms of the historical development of Turco-Islamic culture, the labors of Kashghari and of Yusuf of Balasaghun did not bear fruit. To revert to my original image, they laid a foundation, but the edifice was not built — or, to be more exact, an edifice was built later on (especially by the Ottomans in the west and the Timurids in the east), but on a different foundation.

' S e c G. Clauson, "Early Turkish Astronomical Terms," Ural-Altaische 132-50.

Jahrbücher

35 (1964),

2. ON NATURE IN KARAKHANID LITERATURE

In this paper I shall present and compare the nature themes and imagery found in the two chief Karakhanid literary monuments, Dlwan Lugat al-Turk of Mahmud al-Kasgari (DLT; c. 1075) and Kutadgu, Bilig of Yusuf Khass Hajib (KB; 1069-70). In DLT, Kasgari cites isolated verses of poetry as illustrations of usage, in the same manner that he cites example sentences or proverbs. The verses are anonymous; they show features of oral composition; and they are in the syllable-counting meters characteristic of folk poetry. But they are not simply products of the popular culture. They are often worldly and sophisticated, and sometimes exhibit influence from the direction of the Arabic-Persian literary culture. We may assume that they reflect the kind of literature that was popular in Karakhanid court circles in the eleventh century. The scattered verses can be grouped together in what I call "verse cycles" or groups of stanzas sharing a common rhyming and metrical scheme and a common theme. There are fifty-four such cycles, ranging from one to sixteen stanzas each. The following breakdown provides a rough idea of the popularity of various themes. Of the fifty-four cycles, eighteen (33%) fall in the class of Wisdom poetry; eleven (20%) relate to Warfare; and eight (15%) are concerned with Love. Only six (11%) deal with Nature. The remaining eleven (20%) include various themes, Elegy, Eulogy, the Hunt, etc. Of the six cycles devoted to nature, two are Odes to Spring; one is a Debate of Summer and Winter; one a description of the Creation of the Heavens; one an appreciation of Stars and Bird at Night; and finally one is a description of the Volga River. We shall return to these in a moment. KB is a long didactic poem in the Perso-Islamic mirror-for-princes tradition. It is court literature and reflects the popular culture to a much lesser degree than the verses cited in DLT. Since it is an ethical tract there is little occasion for overt descriptions of nature. Still, we do find an Ode to Spring, also descriptions of the Creation, and of the Heavens. A good deal of the imagery too is drawn from the world of animals and from the natural world. Thus the king is likened at various points to a lion (line 784, also 4098f.), to the sun (5339, etc.), to fire (4087), to the sea (5356) and to a mountain

20

FROM

MAHMUD

KA § G A R i TO HV L i Y A

gHLEBi

(5357). Another example is this lovely quatrain which the author cites in approval of cooperation among men (3212-3): 1 yagiz yer ya§il suv yara§ti bile ara ming gigekler yazildi kiile yara§ik yorisa kada§ ya ada§ olardin bolur kap kada§ ka bile Brown earth and blue water made a pact: Between them a thousand flowers bloom. Friends who walk and work together Are like brothers born of a single womb.

In what follows, I shall compare the material in DLT and in KB with regard to three themes: Creation; Stars and Bird at Night; and Ode to Spring. Let us begin with the Creation cycle in DLT. It consists of two stanzas which, in view of the rhyme pattern, must be in the following order: 2 417 toriit

tengri ajun toriittl gign ediz tezginiir yulduzlari gerge§ip tun kiin iize yorgeniir

"God created the world and the heavenly sphere, lofty, (always) revolving; with its stars arrayed in ranks, and night turning on day." 166 f e j

yaratti ya§il ge§ sawurdi tirting ka§ tizildi kara ku§ tun kiin iize yorgeniir

"(God Most High) created a sky like turquoise in its greenness; He sowed therein the stars as though they were jade (this is a white stone used to make seals); He arranged from them Libra (a constellation called in Turkic Kara Kus [Black Bird]); that night and day would turn on one another."

^Citations from KB are to line number according to the critical edition by R. R. Arat (Istanbul, 1947). The transcription is altered slightly. ^Citations from DLT are to page number and entry word according to the facsimile (Divanii Lugat-it-tiirk Tipkibasimi, Ankara, 1941). The English reflects KasgarT's Arabic translation, modified to indicate what Kasgari adds (in parentheses) and what he omits (in brackets).

ON

NATURE

IN

KARAKHANID

LITERATURE

21

The images are striking. The constellations are viewed as regiments drawn up in battle array, reminding us that the Turks called one famous war tactic "the Pleiades line" (iilker cerig: DLT, 60). 1 Then the stars in the sky are tiny bits of jade set in turquoise. The final verse of both stanzas is the same, implying a possible fault in the transmission of the text or else in Kasgari's memory. However this may be, Kasgari's translation of the verse in both cases (yukawwiru l-layl 'aid n-nahar) immediately hearkens to a phrase in Q u r ' a n 39:5 (yatakawwaru

l-layl

'ala

n-nahar).

W e are dealing here with a Quranic

image. One of the introductory chapters of KB (V: On the Seven Planets and the Twelve Constellations) begins with language very similar to what we just saw in DLT (KB, lines 126-7): yaratti kor evren tugi evrilur arung birle tezging yime tezginiir yasil kok yaratti iizre yulduzi kara ton toriitti yaruk kiinduzi

"He created the sphere of Heaven ever revolving, and the firmament ever turning. The blue sky He created, with the stars upon it, and black night and bright day He brought into being." *

Coming to our second theme, an isolated stanza in DLT recounts a sleepless night in the following terms: 530 sayra§-

yaruk yulduz tugarda odnu kelip bakar men satulayu sayra§ip tatlig iinin ku§ oter

"When the bright star rises, I awaken from sleep and look (at the trees; and I listen to) the birds warbling their sweet melodies." This recalls the several descriptionss in KB of the sleepless nights suffered by Ogdiilmis. I shall cite three of these here. The first has stars only (4888-90): "After sleeping a little he awoke. Mars had descended from the zenith. Looking again he saw that the Pleiades had set, and the Scorpion was

' But cf. also Fouchécour (see note on p. 25 for the reference), p. 131 (citing M30:1083): "les Pleiades, paran, forment une troupe et sont comparées à des étincelles."

22

FROM

MAHMUD

KA § G A R 1 TO

EVLiYA

gELEBI

rising opposite in the east. He lay down once more, but could not sleep; his eyes, like the watchman's, refused to close." The second has a bird only (4963-5): "The yellow nightingale began to warble. He awoke with a start and lay awake a while, listening and marvelling at that sweet voice. Sleep had fled, the night was very long he lay wishing for day to come." The third has stars and bird together (5674-7): "He could not close his eyes again but lay awake thinking, sunk in melancholy. As he lay watching, Black Bird rose in the east like a flame rising from the ground. Then Gleamer shone along with Stallion, the Twins ranged next to them. Finally the Bird of Dawn soared up to the sky and warbled as though reciting a Hebrew psalm." The third of these passages offers several problems of interpretation. "Bird of Dawn" is Arat's translation of siigic, otherwise unattested. Assuming it is not meant to be a star, it is probably the nightingale, since elsewhere in K B that bird (sandvac) is said to "sing a thousand melodies, night and day reciting Hebrew psalms" (line 78). "Gleamer" (yildrik) and "Stallion" ( a d g i r ) are probably Procyon and Sirius. 1 "Black Bird" is probably Jupiter, although this identification is not certain. W e just saw Kasgarl explaining it as Libra (since kara kus otherwise means "eagle", we might have expected it to be rather the constelation Aquila). But Kasgari also tells us (DLT, 167) that it is the name given to Jupiter; furthermore that this planet "rises, according to them (or, in their country?), at dawn" (but there must be some confusion with one of the heliacal planets, of Ogdiilmis's sleepless nights (6219): "Black Bird rose in the east and went up toward the zenith, blazing like an enemy's firesignal" — this just before Gleamer, Stallion and the Twins set at the end of the night. (A wholly different Black Bird, not a star at all, is mentioned in connection with still another of Ogdiilmis's sleepless nights, this time in the context of the darkening of the day (3949): "The brooding eagle (kara kus) covered the sky and filled the world with its plumage." This fits in with the elaborate conceits for dusk and dawn borrowed from the Arabic-Persian tradition, and relates in particular to the A r a b conception of dusk as a besetting eagle or crow.) *

Before turning to the Spring odes, we shall just glance at the Debate between Summer and Winter, which has no counterpart in KB. Kasgarl cites eight stanzas relating to this debate, of which two are introductory in nature

1 See G. Clauson, "Early Turkish Astronomical Terms," Ural-Altaische p. 366-7.

Jahrbücher

35 (1964),

ON N A T U R E

IN K A R A K H A N I D

(DLT, 90 ut-, 317 karis-),

LITERATURE

23

there give the boast of Summer, and three the

boast of Winter. Summer says to Winter: "The bullfinch flees from you, the swallow rests in me, the nightingale sweetly sings his songs, male and female couple" (264 sanduvac = 523 sanduvac). "When the cold comes it covers the world, envious of blessed summer, the snow nearly hides the earth, bodies shake and shiver" (233 tumlig). "Clay and mud pile up, the wretched poor une crouches and hunches up, his fingers crack (from the severe cold), he comforts himself with a firebrand" (129-30 ogrul-). "Winter says to summer: (the flesh of) men and horses becomes hard in me, sickness decreases, men's bodies and flesh become firm" (566 tawra-). "Snow and damp fall in winter, because of it grain grows in summer; in me the hostile foe rests (from his raids), when you come (O Summer, then in you) he moves" (369 tepres-). "In you there arise scorpions, flies and gnats, harmful worms) and snakes, [thousand and tens of thousands], they tie their tails and attack" (605 singek). Thus, while Summer berates Winter for its foul weather, Winter retorts that the cold is not only salubrious but also that it delivers men for a season from the attacks of vicious creatures and of raiding armies (notice the image of the insects and vermin "tying their tails," just as horsemen tie their horses' tails on campaign). *

Of the two Odes to Spring in DLT, one has only five stanzas, and it is partially assimilable to the Wisdom verses in that the final verse in each quatrain reflects upon the first three or addresses the listener in a tendentious fashion. Indeed, in translating the final stanza Kasgari cannot refrain from drawing a moral: 60-1

72 ucmak

178-9 bulit

Spring dawned, the snowmelt flowed in flash floods, the shining star rose; hear my words (which are strange) without laughing! Varieties of flowers have split open (as though) a brocade carpet has been spread, and the land of paradise has been seen; (the season has become so temperate that) cold will never return. A cloud came up thundering, rain and hail leaping from it, the air blowing it about; no one knows what direction it will go.

24

FROM

4 4 6 - 7 yin

MAHMUD

K A § G A R i TO EVLl Y A

gELEBI

All the birds and beasts have c o m e to life (after death w h e n the spring breezes blew), and male and f e m a l e have c o m e together, f o r m i n g scattered herds; they will not re-enter their lairs.

515 yay

D o not d e p e n d on spring (or h o p e f o r good f r o m its b l o s s o m s and bright colors), d o not lean on water (for he w h o d e p e n d s on taking pleasure f r o m the good things of this w o r l d a n d f r o m spring breezes is like o n e w h o leans on water), d o not p r e p a r e y o u r s e l f f o r evil; but b r i n g f r o m y o u r t o n g u e w o r d s ( w i t h which people will be pleased).

T h r e e m a i n t h e m e s can be detected here: s n o w m e l t and f l o o d w a t e r s ; c l o u d - r i s i n g a n d storm; and f l o w e r s a n d a n i m a l s . T h e s e s a m e t h e m e s are p r o m i n e n t in the other O d e to Spring in D L T . T h i s has f i f t e e n stanzas, and is in the s a m e r h y m e as the D e b a t e b e t w e e n S u m m e r and W i n t e r ( - s u r ) . T h e f o l l o w i n g is not a translation but a paraphrase: With the coming of spring the snow and ice on the mountains thaw (120 akis-); the torrents rush down with a moaning sound (461 yagmur)\ the ponds swell and reflect the mountains (599 iler- = 407 kak). White and grey clouds rise from the valleys and sway in the sky like boats (102 akis-)\ or else they spread across the sky like a net (461 yagmur); then pile up and turn black, while the moon gets a halo presaging storm (133-4 orlen-)-, there are flashes of lightning, and the mist rises (124-5 okras-)\ finally the clouds discharge he rain which rushes down in a murmuring flood (133-4 orlen-)-, the people marvel at tne thunder-bolts (617 tanglas-); they go indoors, shivering from the blizzard-like breeze (378 kokres- = 508 kad). But spring also brings a warm breath to the world (99 iler- = 507 kak)-, the mountain-tops turn green as new foliage covers over last year's withered growth (308-9 koser)-, the myriads of buds and flowers, which endured so long beneath the ground, now burst forth and rank themselves in rows (123 adris-)\ or else they are piled on top of one another, their calyxes like ribbed knots intertwined and ready to burst asunder (219 ttigsin = 408 bokuk); now the calyxes split, revealing the blossoms like pearls from their shells, and spreading their fragrances as though mixing sandal-wood and musk (329-30 sucul-)\ the people marvel at the abundance and variety of color, red and yellow, brown and green (199 kizil). The animals react to the coming of spring in more dramatic fashion: the bulls and the oxen bellow at the abundance of water and greenery (308-9 koser-)-, the antelope and the wild ass are aroused, male and female pair off and form rows, leaping and running to the summer pasture (115 emit-)-, while the ewes are separated from the rams and allowed to mingle with the lambs instead of being milked (486-7 yamras-)\ the cattle graze and put on flesh, the Begs can now find fat mounts, and the horses bite at one another for joy (145 atla-)-, the stallions herd the mares together and they neigh and whinny when they see the signs of spring (124-5 okras-).

ON N A T U R E

IN

KARAKHANID

LITERATURE

25

Although the stanzas could be arranged to follow the sequence suggested here many of them touch on more than a single theme, piling image on image, or else juxtaposing images from nature with responses from the human or animal realms, often with masterful fitting of sense to sound. Here are two examples: 378 kokreg-

keldi esin esneytt kadka tiikel osneyii kirdi bodun kasnayu kara bulit kokre§iir

"The breeze came blowing, it was like a blizzard (in its blowing of wind and mist), the people [came in] shivering (from the cold), the block cloud thundered." 124-5 okra§-

ya§in atip ya§nadi tuman turup tii§nedi adgir kisir ki§nedi ogiir alip okraijur

"The cloud lightened, the mist was stirred up, the stallions and mares (saw the traces of spring and) neighed, they whinnied together (for joy) as each stallion took his mares as a herd." The Ode to Spring in KB is of quite a different character from the ones in DLT. It is not at all an occasional piece, but is fully integrated into the plan of the work as a whole. The description of spring is subordinated to the main theme of the ode, the praise of the author's patron. The ode occurs as the fourth of eleven introductory chapters, following chapters in praise of God, the Prophet, and the Four Companions, and preceding discourses on the heavenly bodies, on speech, virtue, wisdom and intellect, the author's apology, and his explication of the book's title. In its overall structure and in the details of its imagery, the ode is wholly indebted to Persian lyric poetry, in particular the poetry of the Ghaznavid court as it has been analyzed by Fouchecour. 1 Spring's herald is the zephyr (line 63; cf. Fouch. 95-101), which banishes the snow, compared to camphor, and exposes the dark rich and fragrant earth, compared to musk. The sun resumes the auspicious position between Aries and Pisces (66; Fouch. 29). Earth dons a green cover decorated with multicolored flowers; the covering is compared to stuffs of silk and brocade (68; Fouch. 51-94, 31 fof China]). Earth's renewal in the spring ^C.-H. de Fouchécour, La description de la nature dans la poésie lyrique persane du XIe siècle: Inventaire et analyse des thèmes. Paris: Klincksieck, 1969.

26

FROM

MAHMUD

K A § G A R i TO

EVLÌYA

gELEBi

is likened to an old woman suddenly becoming young again and arraying herself to be the prince's bride, the very image of his bountiful good fortune (84-5; Fouch. 30). The birds raise their voices in praise, albeit in an unfamiliar language (72-8; Fouch. 137-41, 145 [partridge laughing)). Heaven weeps in order that earth may smile (80; Fouch. 104). The spring storm in its raging is an image of the prince's might in battle and of his glory in maintaining his rule (86-8; Fouch. 17). Spring's grifts are transitory, but the prince's glory and fortune are permanent (112-5; Fouch. 17). May the prince continue to prosper, as long as nature continues in its course and the blessing of spring recur (118-21; Fouch. 19-20). Fouchecour's study is based on the divans of the three great Ghaznavid court poets, 'Unsuri, Farukhi and Manuchihri. In a book devoted to the third of these, Jerome Clinton discusses the structure of the Navruz Qasidah:' the naslb or prelude, which describes the coming of spring, is joined to the madh or eulogy of the patron by the gurizgah, a line introducing the patron's name; at the end of the poem comes the du'a or benediction on the patron. Clinton goes on (p. 94) to speak of the "absence of any overall metaphorical unity" in Munuchiri's qasidahs. "With few and possibly accidental exceptions, there is no continuity of images between the naslb and madh of a particular poem. No bright birds fly over the gurizgah to roost on the Sultan's throne...." In the Ode to Spring in KB the birds do exactly that! (lines 93-5): W h e n this Khaqan assumed the throne, the World breathed easy and began to s h o w e r him with kingly gifts. Birds d e s c e n d e d out of the lofty aether, some f r o m the Raja of India and others sent by Caesar, vying wth pride and j o y in praising him and calling out his name.

Possibly this image was available to Yusuf because of a custom that was practiced in the Karakhanid court but not in the Ghaznavid court — the giving of birds as royal gifts in the spring. One of the birds mentioned at the beginning of the KB ode (line 71) is the "wild swallow" (kil); and Kasgari speaks of the kil kus or kil kudruk, "wild swallow" (Ar. subad), in these words (DLT, 169-70 kil): "It is a duck-like bird which comes at the beginning of the spring, and the Begs give it to one another as a gift." Not only the birds carry the "metaphorical unity" between the nasib and the madh, but also the flowers, and also the personified World. Finally, before his benediction (which "unites the themes of naslb and madh," Clinton, 94), Yusuf adds his own gift, which is "wisdom of Royal Glory" (kutadgu bilig), to the series of royal gifts; and proclaims (112-3) that unlike the gifts of nature, his will last forever.

' J. W. Clinton, The Divari of Manuchihri Dàmghani: A Criticai Study. Minneapolis: Bibliotheca Islamica, 1972.

3. THREE TURKIC VERSE CYCLES RELATING TO INNER ASIAN WARFARE

Among the verse cycles that can be reconstructed from the scattered stanzas recorded by the Karakhanid philologist Mahmud al-Kasgharl in his Diwan Lughat at-Turk (c. 1075),1 there are three which relate to warfare among various Turkic groups. One of these has to do with a Muslim (Karakhanid?) campaign that took place in the first part of the eleventh century and resulted in a victory over a confederation of infidel tribes in the region of the Irtysh Valley. The second recounts a raid of Muslim Turks against Buddhist Uighurs. The third concerns the Khan of the Tangut and his putting to rout the troops of Qatun Sini, "a city between Tangut and China." Kasgharl traveled among the Turkic peoples, "throughout their cities and their steppes," as he tells us, "learning their dialects and their rhymes." 2 His aim was to provide a complete lexicon of the Turkic dialects. Following the model of the Arabic lexicographers, Kasghari often cited proverbs and poetry to illustrate the usage of words. The poetry invariably consisted of quatrains and couplets which followed the patterns of traditional Turkic folk verse, although there is evidence of some slight influence of Arabic-Islamic forms. When arranged according to contents and to metrical and rhyme scheme, most of the stanzas fall into fifty or so clearly delineablc cycles, relating either to warfare (as the ones treated here), or to the hunt, death (elegy), love, nature, or proverbial wisdom. In the reconstructions that follow, an attempt is made to arrange the stanzas in a plausible order and to identify as far as possible the geography and chronology of the three campaigns, especially the first. In studying the texts, the reader should keep in mind the following points: (1) All of the stanzas in the Diwan relating to a given campaign are presented, but there is no assurance that these represent all the stanzas in

' See the Encyclopaedia of Islam, new ed. (hereafter EI2), s.v. " a l - K a s h g h a r i " (Hazai) and "Ilek-khanilds" (Bosworth). See the phototypical edition of the unique manuscript, Divanii Lugatit-turk Tipkibasimi (Ankara, 1941), p. 3.

28

FROM

MAHMUD

K A § G A R 1 TO EVLIYA

£ ELEB 1

the original cycle. There is also no guarantee that the ordering of the stanzas reflects the original ordering.1 (2) The English translation does not reflect the Turkic directly, but rather, Kasghari's Arabic rendering. Places where Kasghari is translating from the Turkic are indicated by quotation marks in the English. Within the quotation marks, any additional comments by the author which are not direct translations from the Turkic are in parentheses. In places where Kasghari's translation is eliptical or inadequate, my rendering of the Turkic is given in brackets. (3) Citations are given according to manuscript page number and entry word.2

I. WAR OF THE MUSLIM TURKS AGAINST THE COALITION OF YABAQU, BASMIL, g O M U L A N D YEMAK

Besides the four groups of stanzas relating to this campaign presented below, there is also a prose account which allows us to place it during or just before Kasghari's own lifetime. This comes at ms. 545-6, s.v. boka "dragon": This w o r d is used as a name for warriors. For example, one of the Yabaqu chieftains w a s called Boka BudraC. God Most High put them to rout the day when Ghazi Arslan Tegin fell u p o n them with 4 0 , 0 0 0 Muslims, while the infidels under Boka BudraC were 700,000 strong. M a h m u d [al-Kasghari] says: "I asked one w h o witnessed this battle, ' H o w is it that defeat fell upon the infidels despite their n u m e r o u s troops?' He said: ' W e also were amazed at this, and w e asked the infidels how they were put to flight despite their great host. They told us: "When the drums began to beat and the trumpets began to blow, we saw just ahead a green mountain blocking the horizon, and in it there were gates too numerous to count, each of them wide open and shooting at us sparks f r o m fires. We were d u m b f o u n d e d , and so you defeated us." ' I said: " T h i s is one of the miracles of the Prophet (God bless him and grant him peace!) which persist among all Muslims."

Carl Brockelmann collected all the stanzas of the Diwan into their various cycles, arranging them within each cycle acording to the order the appear in the text: see his "Alttiirkestanische Volkspoesie," pt. 1 in Hirth Anniversary Volume, = Asia Major, Probeband, 1923, pp. 1-22; pt. 2 in Asia Major 1 (1924): 24-44. I. V. Stebleva rearranged the stanzas and discussed the verse in connection with her idiosyncratic theory of Turkic poetics: Razvitie tjurkskix poeticeskix form v XI veke (Moscow, 1971). Tahsin Banguoglu rearrranged some of the stanzas (corresponding to cycles II and I-d infra) and attempted to place them in a historical context: "Ka§gari'den Notlar, I: Uygurlar ve Uygurca iizerine," Turk Dili Ara$tirmalari Yilligi Belleten, 1958, pp. 103-106. 2 Text and translations are based on a complete edition of the Diwan by Robert Dankoff, in collaboration with James M. Kelly, to be published in the near future. [Mahmud al-Kasgari, Compendium of the Turkic Dialects (Diwan Lugat at-Turk), edited and translated with introduction and indices by Robert Dankoff, in collaboration with James Kelly, 3 vols. (Duxbury, Mass., 1982, 1984, 1985)]

THREE

TURKIC

VERSE

CYCLES

29

This conversation with a veteran of the campaign shows that a ghazi legend (that is, a legend about a Muslim warrior) had already grown up around the hero, Arslan Tegin. 1 In the verses, Budrac is associated with the tribes Basmil, Comiil, and Y e m l k , while the Yabaqu is not mentioned. According to Kasghari (ms. 25) the Yabaqu, Basmil, and Comiil, along with the Qay and Tatar, knew Turkic well, but each also had a language of its own; all are expressly stated as being nomadic peoples. 2 In the geographical enumeration of the tribes (ms. 20-21), the first ten, who are "opposite Rum extending toward the East," are listed as follows: Bacanak, Qifcaq, Oghuz, Yemak, Basghirt, Basmil, Qay, Yabaqu, Tatar, Qirqiz. On the other hand, the Comiil is enumerated with the other ten, who are "middling between South and North"; the list reads Cigil, Tuxsi, Yaghma, Oghraq, Caruq, Comiil, Uighur, Tangut, Khitay (Sin), and Tawghac (Masin). Also mentioned in the verses are the Artis (Irtysh) and Yamar (Ob) rivers. Both appear as entries in the Diwan, as follows: 61 artis, "name of a river in the steppes of Y e m l k which flows into a lake there and which has many arms and tributaries ..."; 456 yamar, "name of a place over which flows Yamar Suwi; this is a large river in the steppes of Yabaqu." The geographical disposition of the above-mentioned tribes and rivers on Kasghari's map (ms. 22-23) is hardly trustworth. The Artis and Yamar, together with the Ila (Ili), all flow into the same body of water, which must be Lake Balkhash. Just to the west of this lake are the steppes (fayafi) of Basmil, north of the steppes of Tatar and Basqirt. The steppes of Y e m l k are on the left bank of the Artis, whereas the habitation (maskan) of Comiil is found between the Artis and the Yamar, south and east of Qay. Yabaqu is not on the map. 3 The next of the verses begins with a group of seven stanzas describing the outset of the affair: Boka Budrac musters his troops; the enemy crosses the Artis, causing panic among the people, who call for help to the "Tarkan" or "Khan" (presumably titles of Ghazi Arslan Tegin; Kasghari's translation as "khaqan" or king possibly refers to his subsequent rank).

' Note also ms. 227 budrac "name of a chieftain of the Yabaqu; they fell into the hands of the Muslims in the time of BäkäC Arslan Tegin." And see R. Dankoff, "Käsgari on the Beliefs and Superstitions of the Turks," in Journal of the American Oriental Society (hereafter, J AOS), 95, no. 1 (1975): 68-80, 69. 2 S e e R. Dankoff, "Käsgari on the Tribal and Kinship Organization of the Turks," in Archivum Ottomanicum 4 (1972): 23-43. 3 O n the geography see V. Minorsky, Hudüd al-'Älam, Gibb Memorial Series, n.s. 11 (London, 1937), pp. 275, 285, 288, 303, 316. B r o c k e l m a n n , " A l t t ü r k e s t a n i s c h e Volkspoesie", pt. 1, p. 12, suggested that the Yamär w a s the Ishim. A. Herrmann, "Die älteste türkische Weltkarte," Imago Mundi 1 (1935): 24, identified the Yamär with Sarysu. The identification of the Yamär with the Ob is due to W. Barthold, e.g., in his Zwölf Vorlesungen über die Geschichte der Türken Mittelasiens (Darmstadt, 1962), p. 96.

30

FROM

MAHMUD

K A § G A R j T O E V L1 Y A

CELEBI

I-a 1.

2.

3.

4.

5.

6.

(83 alpägut) budrac yemä qudurdi alpägutin üdürdi susin yana qadirdi kälgälimät irkisuir (85 uruncaq) urunöaq alip yermädi alimlig körü armadi adasliq üzä turmadi qalin ärän tergäsuir (163 kür) ärtis suwi yemaiki sitgap tutar bilaiki kürmät anig yiiraiki kälgailimät irkisür (88 usitgän) usitgän quyäs qapsadi umunölug adas täpsädi ärtis suwin kacsädi bödun anin ürküsür {All yan-) yandi ärinö ugragi käldi bärü tugragi oizi quyi ogragri alplär qamug tergäsür (222 törkün = 371 cärgäs-) kälsä apan tärkänim etilgämät terginim yadilmagay* törkiinüm amdi öärig öärgäsür *371 tarilmagay

7.

(231 bicgäs) biögäs bitig qilurlar and käy yemä berürlär xändin basut tilärlär basmil öömül tergäsür

Budrac (the name of a former emir of the Yabaqu) [went beyond the limit]; he chose his warriors, turned round his troops, and assembled to approach. (Reproaching a man:) He was not too tired to take a deposit, and to put off the creditor when he saw him; nor has he continued in friendship; now he has mustered a huge army (and is heading toward me with it). (Yemak is a tribe of QifCaq.) 1 This tribe [lit. the YemSk of the Artis River] have rolled up their sleeves, their hearts strong and high-spirited, and now have mustered to come against us. The [thirst-causing] summer heat encompassed us, the longed-for friend envied us (?), the enemy was about to cross the Artis River; because of that the people were panic-stricken. (Describing the enemy:) Perhaps he has turned back from his intention; for his mounted messanger came to us; and in the bottom of the valley and on the slope, the warriors are lined up for battle. (Calling the khaqan to help fight the Yabaqu [thus 222 only]:) May the khaqan but come! (371: If the king comes and helps us); then my gathered goods will prosper, and my clan's houses will not go to ruin (371: and my clan will not be dispersed); now the battle lines are drawn up; (222 only: may he leach me!). (Calling the khaqan to help in battle with the Yabaqu:) My people write a pact of allegiance not to go against the king; [they even take a firm oath; they seek aid from the khan;] for now the tribes of Basmil and Comiil are gathered to do battle with us. Then the khaqan overtook them and took them captive.

T w o stanzas with a slightly different rhyme-ending are c l o s e l y connected with the preceding.

'Cf. ms. 456 yemak, "a tribe of the Turks; they are considered by us to be QifCaq, but the Qifcaq Turks reckon themselves a different party."

THREE

TURKIC

VERSE

CYCLES

31

I-b (421

qomit-)

basmil stisin qomitti baròa kàlip yomitti arslan tapa âmltti qorqup basi tàzginur 2.

(177

turk)

qaòan kòrsa ani turk ayga anig aydaCi 1 munar tàgir uluglug munda naru kàslinùr

(Describing the Basmil troops which warred against the Ghazi Arslan Tegin:) T h e Basmil roused their troops (to war against us); all of them banded together; they went after the lion; (but when they saw us their eyes were dazzled and) they were dizzy with fear. (Praising a man:) When the tribes of the Turks see him they will say: "Glory and pride befit this one, and after him will be cut off."

" T h e l i o n " in t h e f i r s t of t h e s e t w o s t a n z a s is u n d o u b t e d l y A r s l a n T e g i n (literally "Lion P r i n c e " ) ; and the r e m a r k about the e n e m y ' s eyes b e i n g d a z z l e d u n d o u b t e d l y r e f e r s to t h e g h a z i l e g e n d narrated a b o v e . It is less certain, but still likely, that t h e h e r o referred to in the s e c o n d stanza is also Arslan T e g i n . T h e third g r o u p of s t a n z a s is a battle s o n g in t h e c o h o r t a t i v e m o o d . " O u r " troops are to take revenge u p o n B o k a B u d r a c , p u r s u e the e n e m y a c r o s s the Y a m a r , and restore p e a c e to the realm. I-c 1.

2.

(600 tan) tan ata yortalim budrac qanin irtâlim basmil begin ôrtâlim amdi yigit yewilsiin (266 kôc-) tiinlâ bilâ kòcll i m yamâr suwin kacllim tarnuk suwin icâlim yuwga yagi uwulsun

We'll set out at the break of dawn; we'll seek the blook (and the blood-price) 2 of B u d r a d (a man of the Yabaqu); and we'll burn the emir of Basmil; now let the young men gather in the their squadrons. We'll set out at night, and cross the Yamar River; we'll drink seepage water; and the enemy, swerving from us, will crumble in our hand.

' In the text the word bodun "people" appears at the beginning of this line. When it is omitted, as here, the line is metrically correct and literally means: "a sayer of theirs will say." This translates the Arabic phrase qala qd'iluha ("one of them would say") in the verse of al Farazdaq on which this Turkic verse is based, according to A.-Z. Validi [Togan], "Mahmud Ka§gari'ye ait notlar," Atsiz Mecmua 17 (1932): 126. The Arabic verse in question is: ida ra'athu qurayshun qala qa'iluha/ila makdrimi hada yantahi l-karamu (When the Quraysh saw him, one of them would say: "Nobility reaches as far as this one's noble qualities"). 2

Brockelmann, "Alttiirkestanische Volkpoesie," pt. 1, p. 14, points out that "blood-price" does not fit the context, since it implies taking revenge for BudraC; however the Arabic can be construed to mean that we are taking revenge upon Bud rac and thus exacting the blood-price for a dead ally.

32 3.

FROM (221

MAHMUD

qalqàn)

qiqrip atig kamsàlim qalqan siinun comsalim qaynap yana yumsalim qatgi yagi yawalsun 4.

( 2 7 0 tus- = 3 3 6

tagra awip attin tiisiip arslanlayu kuTCi anin 5.

(66

kàwil-)

àgralim yiigràlim kokralim kàwilsun

àndik)

andik kisi tétilsun èl tòrii ètilsun toqli bòri yétilsun qadgu yema savulsun

KA § G A R Ì TO EVLlYA

gELEBl

D e s c r i b i n g the e n e m y : W e ' l l attack screaming and shouting; we'll thrust with spear and shield; we'll be violent in the fray — then quiet (for the sake of peace, if it be sued for), that the hard foe may relent. We'll surround the enemy; we'll get down from our horses and advance on foot; we'll roar like lions; that his strength become weak thereby. (We'll uncover trouble with the sword) so that the simple-minded will come to his senses, the realm will prosper, the wolf will walk with the lamb close behind, and care will depart from us.

A final group of three stanzas probably belongs with this cycle. I-d 1.

2.

3.

(30 us) kàldi mafia tàt aydim amdi yàt qusqa bolup at sani tilatr iis bòri (52 àr) kòriip naòiik qaòmadin yamar suwin kaèmadin tawàrinni saòmadin yèsii sani àr bòri (35 es) anin isin kaòurdum esin yemà qacurdum ofluim otin iciirdum iòti bolup ytizi turi

An Uighur infidel came to me; [I said: "Now lie down!" (or, "Now, foreigner!");] I killed him and made him prey for vultures [literally: Be meat for the birds; may the vultures and wolves seek you out]. (Describing one of the routed soldiers whom he has captured:) When you saw me, why did you not flee, and cross the Yamar tributary (xalij)? Why did you not cast off your belongings (and escape with your life)? So now (may you be killed and) may the hyena eat you. I put an end to his affair (by killing him), and I made his companion flee; I gave him the potion of death to drink, and he drank it with a sour face. (He means: I made him taste death.)

The mention of "an Uighur unfidel" (stanza 1) seems out of place in this cycle; 1 but the mention of a captured soldier (stanza 2) relates this group to I-a 7, where it is stated that the khaqan took captives; and the mention of the Yamar relates it to I-c 2, where "our" troops pursued the enemy across the Yamar River. 2

' o n the other hand, tat might mean simply "infidel" (as at ms. 265, s.v. cap-), and not "Uighur infidel" as Kasgharl translates it. (Banguoglu, "Ka§gari'den Notlar," p. 106, suggests that here it means "stranger, enemy.") See below, p. 35, n. 2. 2 N o t e that nacuk "why" (also in stanza 2) is a Yabaqu dialectal form (ms. 197, s.v.).

THREE TURKIC

VERSE

CYCLES

33

D e p a r t i n g f r o m t h e s u r e g r o u n d of the text a n d v e n t u r i n g into t h e u n c e r t a i n r e a l m of h y p o t h e s i s , w e can p r o p o s e the f o l l o w i n g . T h e historical kernel of the g h a z i l e g e n d and of the f o u r g r o u p s of stanzas m a k i n g up cycle I is the victory of t h e K a r a k h a n i d k i n g T o g h a n K h a n , j u s t b e f o r e his death in 1 0 1 7 , o v e r a n u m e r o u s a r m y of i n f i d e l T u r k s w h o h a d a r r i v e d f r o m t h e direction of C h i n a . T h e story of this e v e n t is given by t h e M u s l i m historians Ibn al-Athir and a l - ' U t b i . H e r e is Ibn a l - A t h i r ' s report: 1 In this year [408/1017-1018] the Turks emerged f r o m China in great numbers, more than 300,000 tents belonging to various tribes (min ajnas at-turk). Among them were the Khitay who [subsequently] ruled Transoxania and whose rule will be described in due course, God willing. The reason for their emerging was the severe and protracted illness which Toghan Khan suffered when he was ruling Turkestan. Because of this illness they coveted the lands and went after them, conquering some, and plundering, and taking captive, until there remained only eight days between them and Balasaghun. When he heard of this he was sick to the heart at the news, and he asked God to cure him, that he might take revenge on the infidels and defend the country against them; then God might do with him as He willed. God answered his prayer and cured him. Then he mustered troops and wrote to all the lands of Islam to rouse the people, and 120,000 volunteers (mutatawwi'ah) gathered to him. When the Turks learned of his cure and of his mustering so many troops they returned to their own country. He pursued them a distance of three months, and finally overtook them when they thought they were secure because of the great distance. He attacked them and killed more than 200,000 men, taking as captive another 100,000 and taking as spoil an unprecedented number of cattle and tents and vessels of gold and silver and goods of Chinese workmanship. When he returned to Balasaghun his sickness came back, and he died.... Some say that this event occurred in connection with Ahmad ibn 'All Qarakhan, the brother of Toghan Khan, in the year 403 [1012-1013], T h e m o r e c o l o r f u l a c c o u n t of a l - ' U t b i 2 g i v e s 1 0 0 , 0 0 0 as t h e n u m b e r of tents c o m i n g f r o m the f r o n t i e r s of C h i n a , t h e n u m b e r of m e n c o m i n g to the aid of 1 Kitab al-Kamil, ed. by C. J. Tornberg, vol. 9 (Leiden, 1863; reprinted Beirut, 1966), pp. 209210 (vol. 9, p. 297, in the reprint edition). ^Kitab-i Yamini, in the margin of al-Manini, Al-Fath al-Wahbi, 2 vols. (Cairo, 1287/1869), 2: 219-26. The following report of Bar Hebraeus (The Chronography, trans, by E. A. Wallis Budge [London, 1932], pp. 204-205) may refer to the same event: And in this year [483/1046] the Nestorian Metropolitan of Samarkand sent a letter to the Catholicus, which was also read in the palace of the Khalifah, saying, "A people who are like unto the locusts in their sawarms have made a gap in the mountain between Tebit and Khutan, the |one] which ancient writers say that Alexander the Great closed up, and they have sallied out and gone as far as Kashghar. There are seven kings, and with each king are seven hundred thousand horsemen. And the name of their great king is ' N a s a r a t h , ' which is interpreted, 'Ruling by the command of G o d ' . . . " Cf. O. Pritsak, "Two Migratory Movements in the Eurasian Steppe in the 9th-11th Centuries," in Proceedings of the 26th International Congress of Orientalists, New Delhi... 1964 (New Delhi, 1968), p. 162.

34

FROM

MAHMUD

K A § G A R Ì TO

EVLÌYA

Q ELEB Ì

the khan as volunteers, the number of infidels killed in battle, and the number of captives taken. Clearly, these figures are no more reliable than the 700,000 which Kasghari's informant told him were the number of infidel troops under Boka Budrac; the sources agree only that there were a very large number of enemy troops, far outnumbering the Muslims. More striking is the agreement in the overall course of events as narrated by the Muslim historians and as reconstructed from the ghazi legend and verse cycle presented above. The "emerging" of tribes of infidel Turks north of China corresponds to the Yabaqu coalition crossing the Irtysh; their terror-filled approach to the Karakhanid capital of Balasaghun, 1 and the delayed reaction of Toghan Khan, were echoed in the panic-stricken state of the people and the calling for help to the khaqan (I-a); the khan's appeal to Muslim fervor, and the characterizing of the campaign as a holy war (more explicit in al-'Utbi's account), are reflected in the ghazi legend (cf., also, I-b); and finally, the pursuit of the enemy a distance of three months, the resulting great victory, and the taking of captives and booty correspond to the Muslims' chasing the infidels back to their own country beyond the Yamar (Ob) River (I-c, I-d). If the event of 1017 was indeed the historical kernel of the legend and of the verses recorded by Kasghari, then the Ghazi Arslan Tegin can, perhaps, be identified with Toghan's brother, Arslan Khan, who became king when Toghan died. 2 Or he might have been an otherwise obscure figure who played Roland to Toghan Khan's Charlemagne and who then became the focus of pious legends and epic-like poems among the troops and the people witnessing or hearing of the marvellous events. The ghazi legend, at any rate is similar to the ghazi tales later told about other heroes among the Turks, such as Danishmend. 3 And the four groups of stanzas must originally have been part of one larger verse cycle, perhaps interspersed as songs within a prose narrative as in the later Turkish minstrel cycles of Dede Korkut and Koroglu. 4

^See EI2, s.v. "Balasaghun" (Barthold — [Boyle]). Or else another Arslan Khan, who ruled over Kàshghar, Khotan, and Balasaghun when his father Qadir Khan died in 1031. 3 See EI2, s.v.v. "Battàl," "Ghàzi," "Dànishmendids" (Melikoff); Dàni§mendnàme, ed. and trans, by I. Melikoff, 2 vols. (Paris, 1960). 4 See EI2, s.v.v. '"àshik" (Lewis), "Dede Korkut" (Iz), "hikàya — Turkish" (Boratav). 2

THREE TURKIC

VERSE

CYCLES

35

II. THE CAMPAIGN OF THE MUSLIM TURKS AGAINST THE BUDHIST UIGHURS

Five (or six — see below) stanzas describing a raid on the Uighur. The raiders, presumably Karakhanid troops, first cross the Hi River and conquer1 the "realm" or "province" (TK. el) of Minglaq — an ethnic or geographic name otherwise unattested — and then attack the Uighur cities and profane the Buddhist temples. 1.

(549

kemi)

kemi iòra oldurùp ila suwin kaòtimiz uygur tapa baslanip minlaq ëlin aòtimiz 2.

(218

3.

tiinlâ bilâ bastimiz tâgmâ yanaq bustimiz kâsmaîlàrin kâsttmiz minlaq ârin biètimiz (242-3 backâm) bâôkâm urup atlaqa uygurdaqi tatlaqa ogri yawuz itlaqa quslar kepi uCtimiz

4.

(237

kàsmâ)

qudruq)

qudruq qatig tiigdiimiz tànrig iikus ôgdumiz kâmsip atig tâgdimiz aldap yana qaCtimiz

5.

(173

kànd)

kàlginlâyii aqtimiz kàndlâr iizâ èiqtimiz furxan àwin yiqtimiz burxan iizà siòtimiz

We rode on the boat, and crossed the lia (a large river); then we headed towards Uighur, and conquered Minglaq.

(Describing their raid on the Uighur:) We attacked them at night; we lay in ambush on every side; then we cut their (horses') forelocks, and killed the men of Minglaq (the name of a place). We put badges on the horses; we headed for the Uighur dogs (meaning ' p e o p l e ' ) ; 2 we flew toward them like birds (until we fell upon them). We tied the horses' tails securely; we praised God Most High greatly (meaning the cry allahu akbar among the heathen ranks); w e attacked them driving the horses, and then feigned flight (that they would come after us and we might turn on them and rout them). (Describing their raid on the Uighur:) We came down on them like a flood; we went out among their cities; we tore down the idol temples; we shit on the idols' head. (It is c u s t o m a r y f o r M u s l i m s , w h e n they capture a country of infidels, to defecate on the heads of their idols to profane them.) 3

' N o t e that Tk. ac- "conquer" is a caique from Ar. fataha, literally "open." T h e line uygurdaqi tatlaqa literally means "to the Tats among the Uighurs" (note the elision of r in the plural suffix before the dative ending); the words modifying "dogs," not translated, mean "hidden and evil." According to Kasgharl (ms. 406-7), "Tat" is used by most of the Turks to mean "Persian", but to the Yaghma and Tuxsi it means "Uighur infidels," which is the meaning it has here. See H.H. Schaeder, "Tiirkische Namen der Iranier," in Fetschrift Friedrich Giese, = Die Welt des Islams, Sonderband (Leipzig, 1941), pp. 1-34. 3 S e e Dankoff, "Kasgari on the Beliefs," p. 69. 2

36

F R O M M A H M U D K A § G A R i T O EV LI Y A

gELEBi

There is another stanza with the same rhyme scheme which can be considered as belonging to this cycle. It mentions a skirmish with the Oghraq tribe, which is defined in the entry at ms. 72 as "a tribe of Turks inhabiting a frontier district called Qara Yigac"; it appears between the Yaghma and the Caruq in the list of tribes "middling between South and North" given in the geographical enumeration mentioned above. 1 W e may assume that the skirmish took place on the way back from the Uighur country, since it caused a delay in the return of the troops. (525 tog) agdi qizil batraq togdi qara topraq yetsii kalip ograq togsip anin keCtimiz

The red banner was raised (among the Muslim troops); the black dust rose up; the Oghraq (cavalry) reached us; they battled with us and therefore we were slow (in returning).

Although there are no clues as to the date of this campaign, we may suppose that it, too, took place in the first part of the eleventh century, and perhaps relates to the conquest of Khotan by Qadir Khan Yusuf.

III. WAR OF THE TANGUT AGAINST QATUN SINI

The Tanqut appear between the Uighur and the Khitay in the geographical enumeration of the Turkic tribes. As an entry in ms. 602 we find: tahut "name of a tribe of the Turks; they dwell near China, but claim to be of Arab origin." 2 Elsewhere, in ms. 24, we are told that "some of the Tangut" are settlers (nazila) in the lands of the Turks, like the Khotan and the Tiibiit (Tibetans). In ms. 504, s.v. sin "grave", we find: qatun sini ["the queen's grave"] "a city between Tangut and China." On Kasghari's map the city, spelled Khatun Sini, appears between the country ( b i l a d ) of the Uighur and Masin. In the verse cycle, composed of fifteen stanzas, 3 the Tangut are the "good guys" and the poet speaks in their person, while Qatun Sini and its

^The Oghraq are mentioned in two other heroic cycles, in each case as the enemy (257

taturgan [but not in the variant at 306 qatar-\ and 376 qadris-); they are described as generous

and brave in an isolated stanza at 235 tigraq. 2

T h e Tiibiit (Tibetans) are also given an Arab origin (ms. 179, s.v.). With regard to the Tangut, the phrase "but they claim to be of Arab origin" (wa-hum yaz'amuna anna aslana min al'arab) can also be interpreted too mean: "and they claim that we [i.e., we Turks?] are of Arab origin." ^There are five other stanzas in the Diwan with the same rhyme scheme that are usually included in this cycle but whose connection with it is most unlikely. One describes a night watch (554 sa~y, the other four describe a wolf hunt (82 iqilac, 93 ew-, 276 cat-, 541 keca).

THREE TURKIC

VERSE

CYCLES

37

p e o p l e are a l w a y s referred to as "the e n e m y . " T h e T a n g u t leader h a s the title beg (Ar. amir, stanzas 1, 2, 5), but o n c e is called khan (Ar. haqan,

s t a n z a 4).

O r d e r i n g t h e s t a n z a s is d i f f i c u l t b e c a u s e they a r e m a n y a n d o f t e n o b s c u r e (the p r o n o u n s in stanza 3 are particularly d i f f i c u l t to sort out). In t h e f o l l o w i n g reconstruction the stanzas are s o m e w h a t arbitrarily arranged in three groups. Ill-a: T h e Attack and the R u s e 1.

(586 caglia-) qatun sini cogiladi tanut begin yagïlâdi qani aqip Zagiladi boyin suwin qizil sagdi

2.

(152 ogurla-) begim ôzin ogurlâdi yarag bilip ugurlàdi ulug tânri agirlâdi anin qut qiw tozi togdi

3.

(155 alukla-) tanut susin usiklaîdi kisi îsin àluklïïdi ârin atin belàglàdi bulun bolup basi tigdi

4.

(587 yopila-) tanut xâni yopïlâdi òltìm birlà tôpiilaîdi qadaslari tapâlâdi òltìm koìriip ytìzi agdi

5.

(592 qonuqla-) yagi begdin udiqlâdi kôriip suîni aduqlâdi ôliim ani qonuqlâdi agiz icrà agu sagdi

(Qatun Sini is a city between the Tangut and China; the people of Qatun Sini and the Tangut were warring, and the Tangut routed the people of Qatun Sini.) (The people of) Qatun Sini shouted, and warred against (the people of) Tangut and their emir, until their blood flowed like gurgling water, and their necks milked red water (meaning blood). (Describing the emir of the Tangut who prepared an ambush for the enemy:) My emir stole away from the army and lay in ambush, waiting for the right moment; God Most High honored him with victory, and thus arose the dust of Luck and Fortune. (The enemy) attacked the troops of Tangut (a realm or province near China) at night in the bitter cold so that he would not gain victory); then he [the Tangut beg] mocked them, until they presented to us their horses and men; he [the king of Qatun Sini] bowed hs head f r o m the hardships he suffered [literally: he fell captive and hid his head]. The khaqan of the Tangut tricked (the king of Qatun Sini), and struck him on the top of his head with death; his b r o t h e r s reproached (the routed one and his enemy took pleasure in his misfortune); when he saw that death (was inevitable) his face turned pale. The enemy was overcome with sleep in the presence of the emir; when he saw the army (had attacked him unawares) he was amazed and found it strange; Death had him as a guest; poison milked into his mouth (and he died).

38

FROM

MAHMUD

K A § G A R I T O EV LI Y A Ç E L E B t Ill-b: Battle Scenes

6.

(101 urus = 181 qilic) àrân alpi oqistîlàr qinir koîzin baqistilâr qamug tulmun toqistilâr qiliô qinqa kiïôiin sigdi

7.

(200 tàlim) tàlim baslar yuwuldimat yagi andin yawaldimat kiiôi anin kâwildïmàt qilié qinqa kiiCiin sigdi (156 orïla-) ôzin ôgniip orilâdi yorig yêriq qarilâdi atig kâmsip orilâdi uwut bôlup tôpii agdi

8.

9.

10.

11.

(229 bôktir) aga bôktir uza yordim yiraq bâqip qara kôrdim ani biliq taqi turdim tiikàl y agi tozi togdi (310 qiqir-) qodi qiqrip ogus terdim yagiqàru kiris qurdum toqis iCrâ urus berdim aràn kôriip basi tigdi (620 mànilà-) ani yëtip siinùlaîdi basin yandru yanïlâdi ârân bâyup mânïlldi anig alpin qira bogdi

12.

(547 sôka) àràn idip sôkâ turdi basi boynin sôkâ turdi uwut bôlup bukâ turdi udu qama tebàn yigdi (?)

13.

(590 satigsa-) âwin barqin satigsàdi yulug bërip yazigsâdi tirig ârsâ turugsàdi anar sàqinô kiini tugdi

(Describing battles [thus 181 only]): The warriors called out to one another; they looked at one another askence; they fought with all their weapons; until the swords hardly fit in their sheaths (because of the great amount of dried blood on them). (Describing a battle:) The warriors' heads rolled; the enemy's rage was stilled because of it; their strength waned; until the sword hardly fit in the sheath (because of all the gore clotted on it). (Describing a warrior whom he attacked:) He was proud and boastful and puffed up with pride; he came measuring the ground in cubits; he spurred on his hores [and shouted]; (then he turned back) in shame, climbing up the hill. I claimed a hard o u t c r o p p i n g of the mountain; I saw a black spot from afar; I stayed in my place after I recognized it; then the dust of the enemy all rose up. I shouted out to the foot of the mountain and gathered the clan with my call; then I strung my bow against the enemy; I shot and fought with him; when he saw the men of my army he hid his head (and fled). ( D e s c r i b i n g a man w h o routed s o m e troops:) He caught him up and speared him; he opened anew his old wound; the men savored their plundered wealth; he strangled the enemy warriors. III-c: The Ransom (Describing one who was routed:) He sent men who sat kneeling near me (referring to the chief of the people); he continually reviled his head and neck (for what he did); he was ashamed and skulked and hid from the people; he held me back saying "You must not follow me" (so I stopped). The enemy wanted to sell his houses and his landed property, to ransom himself thereby; he wanted to stay alive; the sun of sadness arose for him (meaning the day).

THREE

TURKIC

14.

( 1 1 3 äsüt-) idu berip bosuttum tawär yülug tasuttum ärän äsin äsüttüm yiikin baröa özi ci «di

15.

( 2 0 1 bulun) apañ qolsa udu bärip tutar ärdim süsin tärip bulun qlllp basi yärip y u l u g baröa maña yigdi

VERSE

CYCLES

39

( D e s c r i b i n g a captive:) I let him g o free; (he ransomed himself for s o m e t h i n g and) I had the r a n s o m c o n v e y e d to its place; I ordered the m e n ' s corpses to be c o n c e a l e d (underground); their chief tied his load with his o w n hand (not f i n d i n g a n y o n e to help him). (Describing the routed enemy:) Had I w i s h e d I w o u l d have f o l l o w e d him, taken him and dispersed his troops, made him captive and split o p e n his head; but he g a v e m e m u c h ransom (so I granted him freedom).

Little can be said with certainty about the geography and the chronology of the events described in this cycle. Qatun Sini is perhaps the K'o-tun Ch'eng which Chinese sources place on the Etsina-ghol. 1 The Tangut are undoubtedly the Hsi-hsia Tangut, who occupied Kan-chou in 1028. It is possible that Karakhanid foreign policy paralleled that of the Hsi-hsia for a short perod after ca. 1012, when the Tibetan king Kio-sseu-lo in alliance with the Uighur prince of Kan-chou opposed the Hsi-hsia and diverted the caravan traffic running between Central Asia and China. It is hard to know, however why a Karakhanid poet celebrated a Tangut victory and attributed the victory to God (stanza 2: ulug tanri, translated with the Islamic phrase alldh ta'ala).

'See V. Minorsky, Sharaf al-Zamän Tähir Marvazi on China. The Turks and India (London, 1942), p. 73. There was also a K'o-tun ch'èng in Mongolia on the Orkhon, and one near the northern bend of the Yellow River. See, further, E. Esin in Türk Kültürü El-Kitabi, Cilt II. Kisim la (Istanbul, 1972), p. 304, fn. 239.

4. INNER ASIAN WISDOM TRADITIONS IN THE PREMONGOL PERIOD

I. SOURCES Our knowledge of Inner Asian wisdom traditions before the thirteenth century is scanty indeed. Indigenous sources may be reduced to four: 1. The Orkhon inscriptions, particularly the monuments of Kul-Tagin erected in 732 and of Tonyukuk erected some years earlier (Talat Tekin, 1968). Although it is clear that they reflect a tradition of royal wisdom, it is too early to judge whether this was indigenous or whether it depended on Chinese and/or Iranian models. 2. The Tun-huang and Turfan manuscript frangments. These are of uncertain provenance and date, though they probably belong to the tenth century. The fragments in runic script were originally published by Thomsen (1912: 215-217) and were recently re-edited by Hamilton and Bazin (1972). The Uighur fragment, consisting of thirteen proverbs actually labeled tiirk savi ("Turkic proverb"), was published by Arat (1936: 53ff; 1965: 272-275). 3. The Dlwdn Lugat at-Turk of Mahmud al-Kasgari. The final redaction of this encyclopedic dictionary, modeled after the Arabic lexicons, was probably made in 1077. It makes clear that the Turks were in possesion of a large stock of indigenous proverbs. Kasgari cites roughly 270 proverbs, several given in more than one variant; also numerous wisdom verses. The unique manuscript of the Dlwan was published in facsimile (Kasgari 1941), and is cited below as DLT plus page number (according to that edition) and entry word. (See the forthcoming Compedium of the Turkic Dialects, ed. and tr. R. Dankoff in collaboration with J. Kelly.) The proverbs were collected by Brockelmann (1920; see also Hommel, 1923) and Birtek (1944); the wisdom verses again by Brockelmann (1924: 34-44). 4. The Qutadgu Bilig by Yusuf Khass Hajib, of Balasagun, written in 1069-70. This is a long didactic poem in the "Mirror-for-Princes" genre which combines Turkic Inner Asian with Arabo-Persian Islamic traditions. The text is cited below as KB plus line number according to the critical edition of Arat (Yusuf, 1947). The shorter and later (thirteenth century?) work entitled Atebetii'l-hakayik (Yiikneki, 1951) is wholly within the Islamic Iranian framework and so is not considered here, although it does furnish illustrations of the process of transforming proverbs into verses, discussed below (e.g. line 328; cf. DLT 514 qdn, 554 yu-).

42

FROM

MAHMUD

K A § G A R i TO EVLIYA

gELKBi

Finally we should mention non-indigenous sources. Both the Chinese and the Arabs tended to belittle the cultural traditions of the Inner Asian peoples, with only occasional and reluctant admission of their sagacity. In the early years of the seventh century, the counsellor P'ei-kii advised the emperor that the Turks were really simple and uncomplicated people, and it would be easy to control them, were it not for the presence among them of malicious and cunning Sogdians who acted as their instructors and guides (Liu, 1958: I, 87; cf. I, 194). The T'ang-shu provides a potrait of the wise Tonyukuk (Liu, 1958: I, 171 ff.) which nicely parallels the self-portrait in his monument (see below). The ninth century Arab writer al-Jahiz, in the course of an encomium on the Turks, quotes a contemporary of his as saying: "And if in their part of the world there had been prophets and wise men in their country, and they had happened to think of such things and had had leisure to attend to them, they would have made you forget the learning of Basra and the wisdom of Greece and the industry of China" (Walker, 1914: 676). Despite the implication, at least one Turkic wisdom tradition did enter the mainstream of Arab belleslettres (Dankoff, 1977).

II. CONTENTS A. Royal Instruction. Although the main functions of the Orkhon inscriptons are commemorative and propagandistic, they may be considered "wisdom literature" insofar as these functions are carried out through the technique of royal instruction. This is clearest in the first part (south side) of the Kiil-Tagin monument. The ruler, Bilga Khagan, harangues the people, reminding them first of his conquests, and second of his having saved them from the ignominy of Chinese subjection. He lays great stress on the idea that the legitimacy and the power of his rule are bound up with its being centered in the Otiikan (i.e. probably the Tannu Ola) range (KT S 3, 8). And in particular he cites what sounds like a hoary maxim: "The land from which to govern the realm is the mountain stronghold of Otiikan" (KT S 4: el tutsiq yer otiikan yis armis). This agrees with what the Chou-shu tells us of the Turk Khagan, that he always dwelt on the mountain Tu-kinshan (Liu, 1958: I, 10); but it also agrees with Tonyukuk's policy of Turk self-government and rebellion against the Chinese (Ton S 10). We unfortunately know too little about the religious and strategic significance of Otiikan to judge whether the above maxim indeed reflects an age(s)-old tradition of royal wisdom (cf. Gabain, 1949; Czegledi, 1962).

INNER

ASIAN

WISDOM

TRADITIONS

43

Then, while warning against the seductions of the Chinese, Bilga Khagan contrasts the unwise instruction of "an evil man" who counseled the people to associate with the Chinese, and his own good instruction to keep their distance. The evil saying is: "If you are distant they give bad silks, if you are near they give good silks" (KT S 7). His own admonition is: "When you are hungry you do not recall being full, and once you are full you do not recall being hungry" (KT S 8). Both counsels are couched in the parallel and contrastive form typical of Turkic wisdom sayings: iraq arsar yavlaq agi bertir yagaq arsar adgii agi beriir acsiq tosiq omaz san bir todsar a£siq omaz san

It is significant that the same title of Bilga, meaning "sage", is adopted by the Khagan and by the royal counsellor Tonyukuk. The latter quotes what must be considered the earliest attested Turkic proverbs. The first occurs while Tonyukuk is wondering whether or not to make Elterish the Khagan: "If you try to distinguish from a distance between a lean bull and a fat bull, you cannot say whether it is a fat bull or a lean bull" (Ton W 5-6).The second relates more clearly to its context in which Tonyukuk is trying to convince the khagan to attack his enemies separately before they have the opportunity to form an alliance: "What is thin is easy to bend and what is weak is easy to break: if thin becomes thick it is hard to bend and if weak becomes tough it is hard to break" (Ton S 6-7). The same notion lies behind the single-arrow versus joined-arrows motif in the later Chinggis Khan and Oguz K h a n traditions. B. Proverbs. Concerning the early Turkic proverbs, there is remarkable uniformity between the few Tun-huang and Turfan examples on the one hand, and the rich collection provided by al-Kasgari on the other. W e may illustrate this by two examples, beginning with a pair of closely related proverbs in the Diwan: DLT 470 yaz-: yazmas atim bolmas yanilmas bilga bolmas

There is no marksman who does not miss, there is no sage who does not err.

610 yanqu: yazmäs atim yagmur yanilmäs bilgä yanqu

The marksman that does not miss is rain, the sage that does not err is echo.

44

FROM

MAHMUD

K A § G A R I T O E V L1 Y A

gELEBi

N o w , the best preserved of the T u n - h u a n g f r a g m e n t s contains a series of five proverbs with the verbs az-mayerci

"guide") and yanil-ma-

" n o t to stray" (always connected with

"not to err" (always with bilga " s a g e " or

bitkaci

"scribe"). T h e one which is closest in sense to the a b o v e pair is ( H a m i l t o n Bazin, 1972: 37): ( y a n i ) l m a ( z b)itka£i a(zmaz yer£i) yoq

yoq

There is no scribe who does not err, there is no guide who does not stray.

A l t h o u g h the use of yaz-

rather than az-

in the f o r m e r e x a m p l e s

f u r n i s h e s the typical initial r h y m e , it would be wrong to c o n c l u d e that these are "better" or " m o r e original" in terms of the oral tradition as a whole. O n e of the oracle texts provides another e x a m p l e of az- and yanil-

together (Arat,

1965: 286): yol azsar av tapmaz kisi yanilsar is butrnaz

If you stray from the road you won't find the house; if a man errs his job won't get done.

A n d Kasgari furnishes an example of az- parallel to yaz-: D L T 5 8 ula: ula bolsa yol azmas bilig bolsa sOz yazmas

If there is a waymark you won't lose the way; if there is wisdom you won't err in speech.

O u r second e x a m p l e m a y also serve to illustrate the state of the art with regard to reconstructing the oral tradition. Hamilton and Bazin (1972: 34) were confronted with the following fragment: (S) LN: TUT (g)ii: TUTmz :: They compared this with the f o l l o w i n g proverb in the

Diwan:

DLT 622 arslan: alin arslan tutar kuCin sicgan tutmas

By trickery you can capture a lion; by force you can't capture a mouse.

T h e y then reconstructed the above f r a g m e n t as follows: alin arslan tutar kiiCin kiisgii tutmaz

(same meaning)

INNER

ASIAN

WISDOM

TRADITIONS

45

Although this reconstruction is unexceptional so far as it goes (kiisgii being a synonym of sicgan), three further matters ought to have been pointed out. First, Kasgari himself cites two other variants of the same proverb, both with a word meaning "scarecrow" instead of "mouse" (DLT 410-411 kosguk', 53 al — here with oyuq, the Oguz dialectical equivalent of kosguk). Second, Kasgari also cites two other proverbs dealing with a lion and a mouse: DLT 50 alim, 206 berim\ allmöi arslän bërimëi siögän 560 qari-: arslän qarisa siögän ütin ködlzür

The creditor is a lion, the debtor is a mouse.

When the lion mouse's hole.

grows

old

he watches

the

Based on these first two points we may surmise that the "original" proverb had sicgan, that this was replaced by kiisgii, and that this in turn was replaced by kosguk and then, among the Oguz, by the equivalent of the latter, oyuq. Third, Hommel (1923: 185) had suggested virtually the same reconstruction (but with kalagii "gerboa" instead of kiisgu) fifty years before Hamilton-Bazin. It has often been pointed out that the typical qualities of primitive Turkic verse — assonance, syntactic parallelism, initial rhyme, etc. — are also characteristic of Turkic proverbs, riddles, oracular sayings, and the like; and the conclusion has been drawn that verse has its origin in these other kinds of highly charged speech (Gandje'i, 1957; Gabain, 1963: 208 ff.; §inasi Tekin, 1965: 59 ff.) In the Diwan we can, as it were, see the process at work; for there we find four examples of wisdom verses embodying proverbs which are also cited independently. Here we give the proverb first, followed by the verse: 1.

DLT 224 taqägu: yazidaqi siiwlin edärgäli äwdäki taqägu iëginma 449 yawas: qoldas bilä yarasgil qarsip adin üdürmä bäk tut yawas taqägu süwlin yazin edärmä

When you go out to hunt pheasants in the field, don't let the hens in the house escape. When you make a friend show deference to him, do not oppose him or choose another over him; hold fast to the hens in your house, don't go after pheasants in the field and let the hens get away.

46

F R O M M A H M U D K A § G A R Ì TO E V L Í Y A 112 a tit-: tày atitsa at tinur ogul aradsa ata tinur 112 tày, 514 tày: tagiir mànig sàwimi bilgàlakà ày tinur qali atitsa qisraq sani tày 59 urna: urna kàlsa qut kàlir

4.

55 oyuq, 193 qonuq: bardi aran qonuq kòriip qutqa saqàr qaldi yawuz oyuq kòriip àwni yiqàr 46 anuq, 278 san-'. anuq utru tutsa yòqqa sanmàs 59 urna: kàlsa qali yarlig bolup yunòig urna kàldiir anuq bolmis asig tutma um-a

£ELEBi

When the colt is reckoned a horse, the horse rests; when the boy is counted among the men, the father rests. Convey my words and say to the sages, that the mare rests when her colt is counted among the horses. When a guest comes good luck comes. Gone are those who when they saw a guest, reckoned him as good luck; left are the evil ones who when they see a scarecrow, tear down the tent. If one offers the guest what is available, hospitality is not reckoned as nought. When a guest comes tattered and distressed, offer what food is available, don't keep him waiting and hoping.

The tendentious quality of such verses is clear in the second example above ("Convey my proverb to the sages and say ..."), as also in the following: DLT 221 talqàn: oglum ogüt algil biligsizlik ketlr talqàn kimnig bolsa añar bàkmas qatàr

My son, take my counsel and drive off ignorance, for he who has barley gruel mixes it with syrup (and similarly, he who has intelligence will accept advice).

211 qardu: qarduni yinCü saqinmàfl tuzguni manèu sezinmSñ bulmaduq nañka sàwinmlfi bilgalàr ani yerlr

Consider not ice grains as pearls, consider not a gift of food as a wage, rejoice not at what you have not found — at all these the sages scoff.

464 yüksak: iis às kòriip yüksak qaliq qodi òoqàr bilgà kisi ògiit berip tawraq uqàr

When the vulture sees prey he swoops down from high in the air; when the wise man is given counsel he understands it immediately.

T h e practice of c o m p o s i n g such verses must have been widespread. It is presupposed in the Qutadgu Bilig (see below).

INNER

ASIAN

WISDOM

TRADITIONS

47

The large majority of the proverbs in the Diwan embody what may be called "standard wisdom," often couching a homely truth in an image from nature or daily life (e.g., D L T 271 tug-: mus ogli muyavu tugar "The kitten is

born miaowing," coined about a son who takes after his father), and counseling the virtues of modesty, generosity, patience, respect for elders, etc. It is noteworthy that the wisdom verses illustrate this standard wisdom exclusively. More interesting are those proverbs which reflect the lore and customs of the Inner Asian Turks. Elsewhere I have gathered those proverbs pertaining to tribal and kinship organization and to beliefs and superstitions (Dankoff, 1972; 1975). Related to the latter are sayings involving such originally alien culture terms as suburgan and toy in (cf., Dankoff, 1975: 69): DLT 257 suburgän: suburgända äv bolmäs topurgända aw bolmäs

There is no house for a dwelling among old tombs; there is no game for hunting on bare ground.

140 agri, 519 toyin: bïr toyin basi agrïsa qamug toyin basi agrïmâs

If one priest has a headache, not all priests have head-aches.

608 tänri: toyin topugsäq tänri säwinösiz

The priest is worshipful, but God is not pleased with his devotion.

Several proverbs reflect the relations of the Turks with the Chinese and the Iranians: DLT 215 torqu: tawgäc xännig torqusi tälim tänlamädip biëmâs

176 börk, 407 tat: tatsiz tiirk bolmäs bassiz börk bolmäs 406 tat: tatig kozrä tikäniq tüprä

407 taf. qiliö tatiqsa is yunöir är tatiqsa ät tinöir

The Khaqan of China has much silk but he measures it out before cutting (advises moderation and thrift).

No Turk without a Persian, no cap without a head.

(Strike) the Persian on the eye, (cut) the thorn at the root.

W h e n the sword gets rusty the w a r r i o r ' s condition suffers; when a Turk assumes the morals of a Persian his flesh begins to stink.

48

FROM

MAHMUD

K A § G A R i TO E V L l Y A

gELEBI

Another group transmits lore regarding the nomadic habits, the medicinal practices, the flora and fauna and terrain of the Inner Asian peoples: DLT 164 quz: quzda qar agsumas qoyda yag agstimls

The shaded slope lacks not for snow; sheep lack not for fat.

234 qislag: Sz k5z ir qislag

Oneself (or, the valley?) is the eye; the sunny slope is the w i n t e r quarter.'

39 agir: agir bolsa ar olmls

If a man has galingale he won't die.

70 ahduz anduz bolsa at olmls 179 sdgut, qadin sogut sBlina qadin qasina

502 sol, 510 qds,

If there is elecampane the won't die.

horses

605 The willow for its sap, the birch for its bark.

458 yalga: qara mun kalmaginCa qara yalga kitcma

As long as black misfortune, does not came, do not cross Qara Yalga [an i n a c c e s s i b l e pass between Farghana and the Turk country].

220-1 qazgan: qus yawuzi sagizgan yigad yawuzi azgan yer yawuzi qazgan bodun yawuzi barsgan

The worst bird is the magpie, the worst tree is the Azgan, the worst land is the eroded, the worst people are those of Barsgan.

Finally, there is a group of proverbs which illustrate the royal-wisdom tradition. One of those can be related to the tradition only tangentially, and because Kasgari connects it with an anecdote that must have been current in Karakhanid court circles: DLT 238 qisrdq: qiz birla kiirasma qisraq birla yarisma

Do not wrestle with a virgin. Do not race against a young mare.

Exact meaning uncertain. Note KasgarT's interpretation: "If a man does his own job and does not entrust it to someone else he is like one who takes as his lot the winter quarter on the sunny slope of the mountain, for that side keeps green longer and has less snow in winter." Cf. (?) KB 6577 dzka koz tag; also the proverb (Arat, 1965: 274): tagda o'z yoq say yazida bel yoq "No valley on the mountain, no slope on the plain."

INNER

ASIAN

WISDOM

TRADITIONS

49

"This," says Kasgari, "is one of the proverbs of the Khaqaniyya which pertain to the wedding night of Sultan Mas'ud, when she tripped him with her foot." (For the historical background, see Bartold, 1968: 279-80). It is an example of what Boratav calls "les expressions proverbiales dont on explique l'origine par une anecdote a laquelle elles font allusion" (Boratav, 1964: 69; for another example see DLT 342 sucin-). More significant are those having to do with beg "chief' and el "realm, the territory under a Beg's control": DLT 234-5 basruq: yër basruqi tag bodün basruqi bëg

The counterweights of the earth are the mountains, of the p e o p l e , the chiefs. [See Dankoff, 1972: 27.]

453 yarin: yarin bulgansa el bulgänür

If the shoulderblade is beclouded, the realm will be disturbed. [See Dankoff, 1975: 73.]

276 qal-, 542 törü: ël qalir törü qalmäs 273 ciq-, 4 9 4 kuc: küö eidin kirsä törü tünlüktin ciqär

The realm may be left behind, but custom may not be left behind.

When v i o l e n c e (or injustice) c o m e s in by the courtyard, custom (or law, justice) goes out be the window.

C. Key terms. We may use the last two proverbs cited above as the point of departure for a discussion of el and its relation with tor and torti. These are key terms in the Turkic/Inner Asian political tradition, as illustrated particularly in the Orkhon inscriptions and in Qutadgu Bilig. The first point of significance is the semantic spread of the word el. There are four distinct meanings (cf. DLT 36): 1. courtyard; 2. vile or ignoble (as in el qus "vulture," lit. "ignoble bird"); 3. realm or province (Ar. walaya; as in beg eli "the Beg's realm"; also el basi "head of the realm" and secondarily "groom for horses"; 4. peace. As a hypothesis we may take "courtyard" as the core meaning from which the others developed by regular semantic transfer. If we picture in our minds the tent or yurt of an Inner Asian chieftain, el is the area before the entrance. In the above proverb it is contrasted with tiinluk, originally the smokehole at the top of the dwelling. In Qutadgu Bilig we find el contrasted

50

FROM

MAHMUD

K A § G A R I TO EVLIYA

CELEBl

rather with tor "seat of honor," the area within the dwelling which is opposite the entrance and near the fireplace. KB 2253: "Two things increase the fame of Begs: their banner in the courtyard (el) and their feast-table in the place of honor (tor)." 262-3: "If a fool has a seat in the place of honor (tor) that place is reckoned the courtyard (el), while if a wise man has a seat in the courtyard then the courtyard is far ahead of the place of honor." 844 (citing the Chief of Ili): "He who knows how to serve will attain the place of honor (tdr); he who does not known will be thrown out of the place of honor into the courtyard (el)." (Cf. also 614-5.) It is clear that tdr is the noble part of the chieftain's dwelling, while el is the ignoble part. This explains the second meaning of el above. But it is also clear that el, the courtyard, symbolizes the entire area under the chieftain's authority. The Beg has not only to entertain the nobles in the place of honor; he has also to maintain his banner in the courtyard. He has to maintain control over all his subjects (just as a groom maintains control over the horses in his charge — cf. Ar. siyasa "managing a horse; administering a country"). The realm or empire that is under the firm authority of the ruler is a Pax. Hence the third and fourth meanings of el above. As a second hypothesis, we may assume that tor "seat of honor" fell together in some contexts with torii "custom, rule, justice" (Ar. rasm, insaf in DLT). 1 If the courtyard symbolized the territory under the ruler's command, the seat of honor symbolized the command itself, the law and order provided by the ruler's authority or, more narrowly, the custom and protocol of his court (e.g. KB 3997 ff.). A phrase such as KT E 3 elig tutup toriig etmis originally could have signified. "He captured the outer court or courtyard (el, metaphor for the land) and organized the inner court or place or honor (tor, metaphor for the government)." Then toriig could have been analyzed as toriig. By the time of the Orkhon inscriptions, however, it is clear that the word parallel to el was torii, not tor, for the political organization of a country. In the Kul-Tagin inscription, for example, el and torii are used synonymously (object of the verbs icgin- and qazgan-: KT E 6, 13; 9, 30), or else are used together as a paired expression or hendiadys (KTE 1, 8, 22, 31). A like usage is found in a verse in the Diwan: el torii etilsUn (so read), translated "that the realm (Ar. walaya) may prosper" (DLT 66 andik). Alongside this usage, however, we find el and torii contrasted, as in the proverb cited above (el ' I,ess likely, torii is a secondary form of tor. In the Diwan we find another secondary form, 542). From Kasgari's example, tora yoqladi "He assumed the seat of honor," and from the fact that yoqla-like ag-, meaning "to ascend," requires the dative (e.g. DLT 537 -qa,

torii (DLT

KB 1002, 1661), this must have been originally *tdr with the rare dative in -a (or possibly *torra with the directional suffix-ra). On tora in post-Mongol Turkic texts see Clauson 1972: 528-9, 531-2.

I N N E R

A S I A N

W I S D O M

51

T R A D I T I O N S

qalir torii qalmas). This is the normal situation in Qutadgu Bilig, where torii is a synonym of konilik meaning "justice" (e.g. KB 355, 800) and is opposed to kiic "injustice" (e.g. 3121, 3206-7, 5576). Thus, the ideal ruler "wielding the sword lops off the neck of his foe, and with law and justice (torii) governs his realm and his people (eli bodni)" (KT 286; cf. 2015, 2034, etc.). Furthermore, the contrast el:tdrii allows Yusuf to make plays on words between torii and tdr: KB 2196: toriiliig kisi boldi el

tor basi

830: qayu elkä tegsä mänin bu törüm ol el baröa etliir tas ersä qorum 4604: törü bilsä yalnuq bulur tör basi

The s t a t e s m a n is head of both the courtyard (or realm) and the seat o f honor (or court proper).

Over w h a t e v e r land m y extends, that land prospers, it be stones and rock.

justice though

If a man k n o w s proper protocol attains the h e a d o f the p l a c e honor.

he of

D. Authorities. Islamic mirrors for princes leaned heavily on the citation of authorities which added prestige to the wisdom imparted. These authorities included Muslim prophets and saints, Greek sages, Iranian rulers and princes, etc. (cf. Bagley, 1964: lvi-lxxiv). Yusuf, the author of Qutadgu Bilig, took as his starting point the Islamic-Iranian ideals of statecraft and of ethics which he found in Arabic (and perhaps also Persian) literature. But his aim was to make a new home for these ideals as part of an Inner Asian Turkic literary heritage. Stated otherwise, he aimed to establish a Turkic wisdom tradition alongside the Arab and the Iranian traditions within the larger frame of Islamic culture. The way he accomplished this may be illustrated in the following example. The king has sent his chief advisor on a mission to summon a certain holy man to the court. The latter refuses, saying that he has renounced the world and also that he is ignorant of royal protocol. He gives the advisor a written and an oral message to this effect, and states: "A messenger's job is only to deliver the message and wait for an answer; you have heard my answer, now convey it..." (KB 3814). Now any Muslim hearing this line would immediately think of the Koranic verse (5:92/3, 99): "It is only for the Messenger to deliver the Message." 1 But the holy man, instead of citing the Koran, continues as follows: ' Nizam al-Mulk, in a similar context (1960: 98) quotes Koran 24: 54/3: "The messenger has only to convey the message plainly."

52

FROM

MAHMUD

K A § G AR Í TO E V L Í Y A

gELEBÍ

The khan of the Three Camps has said, There is no fault in the messenger so long as he does not mince his words but delivers the message whole as it was entrusted to him. And the Khan of the Turks has put it even better: Do not punish the messenger for the message; he deserves neither punishment nor death so long as he truly reports wht he heard; rather a messenger, as the ambassador of peace ( e l c i ) , is inviolable, and if he transmits the message that was entrusted him, then praise and reward are his due. Let u s g l a n c e at a list of the cited a u t h o r i t i e s f r o m the T u r k i c / I n n e r Asian background: ÜC (Uö?) Ordu Khani 2966, 3815, 5569 Üö (Uö?) Ordu Begi 1594 Türk Khani 3817 Khaqan 3126 Türk Buyruqi 1163 Ögä Buyruqi 2941 Cigil 3491 Kök Ayuq 2644 Bökä Yavgusi 5043, 5523 IIa Atligi 841, 1629, 2319 IIa Irkini 4752 IIa Sir Teñi 2696 IIa Begi 1779 Ötükän Begi 1962, 2682 Yagma Begi 1758, 4947 Toña Alp er 5861 O n l y t h e last of t h e s e c a n b e related to a T u r k i c h i s t o r i c a l or l e g e n d a r y tradition (cf. K B 277; D L T 3 3 alp, 5 0 9 qäz). T h e rest a r e a p p a r e n t l y s i m p l y titles (cf. K B 4 0 6 7 - 9 ) , s o m e t i m e s a t t a c h e d to g e o g r a p h i c a l or tribal n a m e s . O n e also f i n d s c o m p o u n d s with B e g which are not attached to a proper n a m e : Beglär Begi "the Beg of Begs" 893 Bodun Begi "the Beg of the people" 6138 El Känd Begi "the Beg of the realm and the city" 216 Ulug Känd Begi "the Beg of the great city" 5354 S i m i l a r l y , t h e r e are general titles or o f f i c e s , such as Bodun Bascisi t h e P e o p l e " ) , Elci

Beg,

( " C h i e f of

a n d t h e like. F i n a l l y , t w o cited a u t h o r i t i e s f r o m

o u t s i d e t h e T u r k i c s p h e r e (note also t h e appeal to N ü s h i n R a v ä n , K B 290): Tazik Bilgäsi "the Iranian sage" 3265 Sartlar Basi "Chief of the merchants" 2745, 3002, 5754

INNER

ASIAN

WISDOM

TRADITIONS

53

There is no discernible pattern to the citing of these authorities. The wisdom attributed to any one is usually of such a vague or general nature that it could just as well be attributed to any other, or to the standard "wise sage" who is in fact cited the most often (see below). There is also little if anything to distinguish the wisdom attributed to the above authorities from that embodied in what are called "proverbs" (masal, kicki sóz: 110, 1623, 3514), "Turkic proverbs" (türkca masal: 273, 319, 880, 1798, 1826), or "Arabic sayings" (tazi tili: 5809). The same holds true for the scattered quatrains attributed to various anonymous poets. Except for their proverb-like character, and their peculiar rhyme-scheme, there are no other characteristics which distinguish them from the surrounding text. The reason is clear. All of this material — proverbs, quatrains, and wisdom-sayings alike — was composed by Yüsuf himself, using the Islamic-Iranian wisdom and ethics tradition as his model. This does not mean that he did not also draw on a living Turkic/Inner Asian tradition; although in the one case where this is demonstrable, it is also likely that his access to that tradition was through the medium of Arabic literature (Dankoff, 1977). E. "Wisdom." A final matter to be discussed is the term bilig. It is comparable to Ar. hikma in that it means both "wisdom" (cf. Pers. khirad) and "wise sayings" (cf. Pers. andarz). The title of Qutadgu Bilig means "the wisdom that conduces to royal glory" (qui meaning "fortune" or "the charisma of rule," corresponding to Pers. farr; see Bombaci, 1965-1966). We find the doctrine that wisdom (bilig) is learned while intellect (uqus) is innate (KB 1824 ff.) Through plays on words wisdom (bilig) is connected with the removal of illness (ig: KB 156) and also with the rule of the prince (beg: KB 1953). The wise sage (biigii, bilga, biliglig, uquslug) who hands down wise sayings (sóz: 292 f, 1048, 3383, 3594, 4135, 4177, 5142, 5980; sav: 973, 1807; bilig see below) is assimilated to the ruler who hands down law and royal custom (tórü: 269, 4446-51). Both bilig and tórü are embodied in words, which constitute a legacy that the sage or the ruler lays down, and that is then passed on (qumaru "legacy": 270, 1341 f.; cf. DLT 634 qód-). To impart counsel in the form of a maxim or wise saying is "to give wisdom" (bilig ber-: KB 778, 1727, 2617, 3382, 3670, 3687, 4240, 4691, 5084, 5171, 5419, 6081, 6448). The term bilig is found in hendiadys with sóz (4221) and hikmat (2698). Occasionally the word is personified and it is bilig itself which utters the maxim (1673: te-; 4075, 4451: ay-; 1385, 5076, 5606: sózlá-; cf. 904, 1728, 1844 where uqus "intellect" utters the maxim). The plural biliglar is only found in the verse prologue which dates to a century or so after the original text:

54

FROM

MAHMUD

K A § G A R Í TO E V L l Y A

KB VP 10-11: bazanmis agirlig biliglar bila qamug barCasina biigular sozi

CELEBÍ

It is adorned with wise sayings most precious, words of the sages for one and all ...

III. C O N C L U S I O N T h e f o r e g o i n g survey should suffice to establish t w o points: that there was a strong indigenous w i s d o m tradition a m o n g the Inner Asian T u r k s ; and that a m a i n c o m p o n e n t of this tradition was royal, or political, w i s d o m . W e have d w e l t here on certain terms — bilig, bilga, el, torii — w h o s e s e m a n t i c spread and mutual interrelations illustrate these points clearly. Other w o r d s that c o u l d be studied f r o m this point of view are ardam

" v i r t u e " and the

d e r i v a t i v e s of o- " t o u n d e r s t a n d " (dg "intellect; m a t u r e (of an a n i m a l ) ; mother," oga "mature (of a man); counsellor (title)," ogiit "advice," ogriin- "to learn," ograyiik

" c u s t o m " ) . T h e question of (mutual?) influence between the

Inner A s i a n and the C h i n e s e a n d / o r Iranian w i s d o m traditions is left open. Finally, w e need studies of the relations between the pre-Mongol and postM o n g o l materials, particularly in c o n n e c t o n with the biligs

and yasas

of

C h i n g g i s K h a n and his d e s c e n d a n t s ; the inculcation of traditional w i s d o m illustrated in such texts as the Secret History Dede

Korkut;

of the Mongols

and the Book of

and the vast s t o r e h o u s e of p r o v e r b s extant in oral tradition

a m o n g the Turkic and Mongol peoples.

BIBLIOGRAPHY AND ABBREVIATIONS Arat, Res id Rahmeti (see Yüsuf; Yiikneki) 1936 Türkische Turfan-Texte. VII. [APAW Phil.-hist. Kl. 12]. Berlin. 1965 Eski Tiirk §iir. Ankara. Bagley, F.R.C. 1964 Ghazáli's Book of Counsel for Kings. London. Bang, W. and von Gabain, A. 1929 Türkische Turfan-Texte. II. SPAW. Berlin. Barthold, W. 1968 Turkestan down to the Mongol Invasion. Third Edition. London. [Gibb Memorial Series.] Birtek, Ferit 1944 En Eski Tiirk savlari. Ankara. Bombaci, A. 1965-66 "Qutluy Bolzun," Ural-Altaische-Jahrbücher 36, 284-291; 38, 1343. Boratav, P.N. 1964 "Les Proverbes," Philologiae Turcicae Fundamenta, II, Wiesbaden.

INNER

ASIAN

WISDOM

TRADITIONS

55

Brokelmann, Carl 1 9 2 0 "Altturkestanische Volksweisheit," Ostasiatische Zeitschrift 7: 50-73. 1 9 2 4 "Altturkestanische Volkspoesie," II. Asia Major 1: 24-44. Clauson, Sir Gerard 1 9 7 2 An Etymological Dictionary of Pre-Thirteenth-Century Turkish. Oxford. Czegledl, K. 1 9 6 2 "Cogay-quzi, Qara-qum, Kök-öng," Acta Orientalia Hungarica 15: 5569. Dankoff, Robert 1 9 7 2 "Käsgari on the tribal and kinship organization of the Turks," Archivum Ottomanicum 4: 23-43. 1975 "Käsgari on the beliefs and superstitions of the Turks," JAOS 95: 6880. 1 9 7 7 "Animal traits in the army commander," Journal of Turkish Studies 1: 95-112. DLT see al-Kasgari von Gabain, Annemarie (see Bang) 1 9 4 9 "Steppe und Stadt im Leben der ältesten Türken." Der Islam. 15: 3062. 1 9 6 3 "Zentralasiatische-Türkische literaturen, I: Vorislamische Alttürkische Literatur," Handbuch der Orientalistik. Fünfter Band: Altaistik. Erster Abschnitt: Turkologie. Leiden. Pp. 207-228. Gandjei, Tourkhan 1957 "Uberlick über den vor- und frühislamischen türkischen Versbau," Der Islam 33: 142-156. Hamilton, J. and Bazin, L. 1 9 7 2 "Un manuscrit Chinois et Ture runiforme de Touen-houang." Turcica 4: 25-42. Hommel, Fritz 1923 "Zu den alttürkischen Sprichwörtern", Asia Major Probeband. [Hirth Anniversary Volume]: 182-193. al-Jähiz see Walker al-Käsgari, Mahmüd 1941 Divanü Lügatit-türk Tipkibasimi. Ankara. KB see Yüsuf KT Kül-Tägin inscription; see Talät Tekin Liu, Mau-tsai 1 9 5 8 Die chinesischen Nachrichten zur Geschichte der Ost-Türken (T'u-küe). 2 vols. Göttinger Asiat. Forschungen 10. Wiesbaden. Nizam al-Mulk 1 9 6 0 The Book of Government or Rules for Kings. Tr., H. Darke. London. Rahmeti see Arat Tekin, §inasi 1965 "Uygur edebiyatinin meseleleri (§ekiller - vezinler)," Türk Kültürü Ara§tirmalari, 2/1-2: 26-67. Tekin, Talät 1 9 6 8 A Grammar of Orkhon Turkic. Bloomington.

56

FROM

MAHMUD

K A § G A R i TO E V L l Y A

CELEBI

Thomsen, Vilhelm 1 9 1 2 "Dr. M. A. Stein's manuscripts in Turkish 'runic' script from Miran and Tun-huang," Journal of the Royal Asiatic Society. 181-227. Ton Tonyukuk inscription; see Talat Tekin Walker, C. T. Harley 1915 "Jahiz of Basra to Al-Fath ibn Khaqan on the 'Exploits of the Turks and the Army of the Khalifate in General'." Journal of the Royal Asiatic Society: 631-697. Yusuf (Khass Hajib) 1947 Kutadgu Bilig I: Metin. Ed. R. Rahmeti Arat. Istanbul. Yiikneki, Edib Ahmed b. Mahmud 1 9 5 1 Atebetu'l-hakayik. Ed. R. Rahmeti Arat. Istanbul.

5. KÄSGARl ON THE TRIBAL ORGANIZATION OF THE TURKS

AND

KINSHIP

The following pages present a view of Turkic tribal and kinship structures that emerges from a careful study of the Diwan Lugat al-Turk. This "Divan" is the earliest known Turkic lexicon. It was written in Arabic, c. 1075, by Mahmud Kasgari. The author, born in Barsgan near Lake Issik-kul, was a Turk, and a member of the Qarakhanid dynasty that was ruling in Central Asia and, since 999, in Transoxania. He travelled among the various Turkic peoples, "throughout their cities and their steppes, learning their dialects and their verses" (3,3/i, 3,8).' Eventually he came to Baghdad, where he wrote the Diwan, and dedicated it to the caliph al-Muqtadi (reigned 1075-1094). It is written for an Arab audience, presumably in order to explain to the court at Baghdad the language and customs of their overlords. The Turks at this time were becoming the dominant political force in the Islamic world, the Seljuks having wrested the Sultanate from the heretical Buwaihids in 1055. Kasgari calls the language he is describing "Turklya". It is basically the dialect of the important Cigil tribe and of the other tribes of the Qarakhanid confederation. In addition, Kasgari clearly cites various dialectcal usages, notably "Ghuzziya" or " T u r k m a n i y a " — the dialect of the Oguz Turks, who included the Seljuks. With the various entries he gives a large number of illustrative sentences, proverbs, and verses of folk poetry. He also gives much valuable inforrmation on the geography of the Turkic tribes, on court customs and popular beliefs, and on social organization. 2

1 References to the Diwan give page and line number of the phototype edition of the unique manuscript (Divanti lugat-it-Turk tipkibasimi, Ankara, 1941); and the volume, page and line number of the Editio Princeps (Kilisli Rif'at, ed., Mahmud al-Kasgarl, Kitab diwan lugat atTurk, 3 vols., Istanbul, 1333-1335 [1915-1917]). On the author's life, see O. Pritsak, "Mahmud Ka§gari kimdir?", Ttirkiyat Mecmuasi 10 (1951-1953), pp. 243-246. Pritsak also elucidated the history of the Qarakhanids in a series of articles; these are summarized by Bosworth in the new Encyclopaedia of Islam, s.v. "Ilek-khanids". 2 Much of this material was gathered by Carl Brockelmann in a series of articles, most of which, however, need updating. Brockelmann's glossary based on the Diwan ( M i t t e l t t i r k i s c h e r Wortschatz ..., Budapest-Leipzig, 1928) has been superseded by that of Besim Atalay, based on his Turkish translation (Divanii lugat-it-Turk terciimesi, 3 vols., Ankara, 1939-1941; Dizin [Index], 1943). An English edition of the Diwan is being prepared by the author of this paper and James Kelly, under the aegis of Omeljan Pritsak.

58

FROM

MAHMUD

KA § G A R I TO EVLiYA

^ELEBl

Concerning the period roughly between the tenth and fifteenth centuries, Lawrence Krader states:1 Data for kinship and social analysis during this period are difficult to extract from the record.... We obtain a few hints, such as those from Turkic and Mongol dictionaries of the period, but these are insufficent for our purposes.

I believe the present study shows that from the Diwan, at least, we obtain more than "a few hints". Since the Diwan is primarily a lexicon, we cannot expect to find connected passages describing the contemporary socal structure, of the sort that we find in anthropological texts. Nor are there extensive literary passages, as offered, for example, by the Old Turkic inscriptions, or by the Qutadgu Bilig? Our method, therefore, was as follows: to examine closely the usage of key words, both Turkic and Arabic; to arrange together material that in the Diwan is widely scattered; and to draw up the pattern which the data reveal. The primary aim has been to gather disparate data into a discrete and reliable form, so that they may be readily available for comparative research. All the material in the body of the article consists of texts drawn from the Diwan itself. Comparative material is confined to the notes and used only to throw light on points in the Diwan that might otherwise remain obscure.

I. TRIBAL ORGANIZATION

Kasgari uses the following Arabic terms for various tribal and kinship groupings (alphabetical order): ahl, batn, jama'a, jil, hizb, hilla, raht, sirm, ta'ifa,

'asira,

qablla,

qawm.

Merely noting a dictionary equivalent in any given case would be unsatisfactory, since the Arabic usage is by no means precise: many of the words overlap in meaning; also, they are not used altogether consistently in the Diwan. Therefore, in each case the actual usage of the word has to be established first.

^L. Krader, Social Organization of the Mongol-Turkic Pastoral Nomads, The Hague, 1963, p. 189. This is a long didactic poem, written in the same language as Kasgari is describing, and at almost the same date (1069). It sheds some light on certain of the subjects under discussion here, and will be cited according to the critical edition of R. Arat {Kutadgu bilig I: Metin, Istanbul, 1947). An attempt has been made recently by two anthropologists to use this work as a source for social analysis. See Z. Eglar and P. J. Magnarella, "A View of Social Classes in the Eleventh Century Karakhanid State," Anthropos 66 (1971), pp. 232-238.

THE TRIBAL AND KINSHIP ORGANIZATION OF THF

TURKS

59

We will consider the larger divisions first, and then proceed to the smaller. 1. Bodun The largest — or rather the most general — division is qawm "people, nation", sometimes used in the meaning of nas "people; mankind". The Turkic equivalent is bodun (or bodun)1 which, in context, usually means people in general, or some unspecified group of people. As an entry (201,3/i,334,4) bodun is given as a Cigil word (therefore "Turkiya"), and defined as "subjects, commoners" (ar-ra'lya wa-'awamm annas)-, the corresponding Oguz form ("the dialect of those who change d to / ' ) is boyun (519,6/iii,127,l), defined simply as "people" (al-qawm). In its Turkiya form, ra'iya reminds us of qara bodun in the Orkhon inscriptions, meaning the mass of people as opposed to the ruling elite, and of qara 'amm bodun in the Qutadgu Bilig (henceforth QB\ lines 4320 ff.), meaning the mass of people, the lowest social class, people concerned only with filling their bellies. A number of illustrative examples in the Diwan bear out this particular meaning of bodun, the most striking being a proverb (235,1 /i, 388, 10): yer basruqi tag bodun basruqi beg "The counterweight of the earth are the mountains, the counterweight of the people (an-nas) are the princes." Kasgari explains this to mean that the stability of men depends on rulers who restrain them, just as the stability of the earth depends on the mountains. 2 The opposition bodun: beg is definitely one of class distinction — the masses as opposed to the ruling estate. 3 Beg is invariably rendered by the The word appears alternately with d and d; also with long or short vowel in the second syllable. The Turkic transcription used here includes notation of length, exactly as this appears in our ms. of the Diwan. The i is assumed to be neutral, and appears the same in words of front and back vocalization. See J. M. Kelly, "Remarks on Käsgarl's Phonology, I", Ural-Altaische Jahrbücher 44 (1972), pp. 179-193, esp. 184 ff. 2 I n the Arabic translation of the proverb, basruq in the first instance is rendered by awtäd "underpinnings", then by tiqäl "counterweights". The likening of mountains to stakes or tentpegs awtäd is an Islamic commonplace, based on Q u r ' ä n 78,7. Note also Diwan 308,2/ii, 64,11: tänri tag birlä yerig basurdi. 3 Bodun are the subjects, as distinguished from el which is the state as a political entity (alwiläya 36,3/i,49,13). Compare the phrase el tüz nätäg "How are the state ( w i l ä y a J and the subjects (ra'iya)?" (496,11/iii, 88,12). For a discussion of the relation between el and bodun, see A. A. Valitova, "O nekotoryx terminax v 'Kutadgu Bilig' ", Kratkie Soobscenia Instituta NarodovAzii 63 (1963), pp. 111-123.

60

F R O M M A H M U D K A § G A R I TO E V L I Y A Q ELEB I

word amir "commander" in the Diwan. A popular chieftain (ra'is on the other hand, is called bodun

basgani

(220,7/i,366,2), from

al-qawm), basgan,

which means a type of large fish! In another proverb (220,16/i,366,15) it is stated: bodun yawuzi Barsgan "The worst people (nas) are those of Barsgan" — since, as Kasgarl tells us, they are unsociable and miserly. Bodun in this instance is still an undifferentiated concept, referring to the entire populace of a city. This generalized usage of bodun is the most common one and is found in the following contexts: o n e ' s f a m e spreading "abroad" ( b o d u n g a lit., "among the people" — 34,15/i,47,10); people standing amazed at a storm (617,13/iii, 295,6); people shivering f r o m the wind (378,9/ii,176,8; 508,11/iii,108,15); people being crowded or uncrovvded (446,5/iii,3,8; 374,5/ii, 170,8; 372,5/ii, 167,2); people gathering together (331,17/ii,101,4; 478,1/iii,57,11); people screaming (126,10/i,203,14); people raiding each o t h e r s ' stocks (126,1 /i,202,4); people b u r n i n g each o t h e r s ' houses (122,11/i,196,16); people laughing at each other (324,3/ii,88,17); or people panicking at the approach of the enemy (88,16/i, 137,7; 626,8/iii,310,8). Especially revealing are two verses of admonition; in one of them the people (bodun / al-qawm) scorn a miser (391,5/ii, 198,3); the other states that one ought to camp with one's people ( b o d u n I al-qawm), wherever they settle, during a famine year (473,13/iii, 50,16). The last example illustrates a more primitive sense of bodun, in which the word stands for "one's own people" as opposed to "people in general". (The same duality is found in the corresponding English "people", German "Volk", Arabic "qawm", etc.) Kasgarl gives a special phrase — blr yartim bodun (464,12/iii, 35,8) — to indicate "a group of people w h o are distinguished from the whole" ( t a ' i f a min an-nas infaradu min bayn al-jumla) One of the more interesting examples is the phrase bodunlug boqunlug kisi (249,13/i, 413,1), defined as "one who has kinsmen" (insan du raht wa'asira). Here also bodun connotes only one's immediate kin or clan. W e have, in these last examples, an older stratum of linguistic usage, in which bodun is clearly related to bod "tribe" (with the collective suffix -un).1 This last word is attested in the Diwan only in its Oguz form, boy (see below, entry 4), and usually rendered by the Arabic qabila. With at least one 1 Compare the following sentence in the Tonyukuk Inscription (W, 3-4): tiirk sir bodun yerinta bod qalmadi. This is translated by T. Tekin (A Grammar of Orkhon Turkic, Bloomington, Ind., 1967, p. 283): "In the land of the Turkish Sir people there was no longer any ordered group of people left".

THE T R I B A L AND KINSHIP ORGANIZATION OF THE TURKS

61

verbal root — yuli- "to raid" (agara) — the object of the action is either bodun or boy, interchangeably. Thus we find: beg bodunug yulidi "The prince raided the people (qawm)" (482,1/iii, 64,8);' beg boyni yulitti "The prince ordered a tribe (qablla) to be raided" (423,11/ii, 255, 10); and bodun blr ekindini yulusdi "The people (qawm) raided each other" (476,6/iii,54,14). In addition, there is one passage in the Divvan where bodun is translated al-qablla (526,5/iii, 139,13): bodun qdndi "The tribe settled after migrating." Significantly, it is again the more primitive meaning of the word that is here illustrated. Bodun, then, stands for people in general, or for some group of people, whether or not specified by tribal or kinship ties. 2. Turk Are the Turks as a whole to be considered a bodunl Nowhere in the Diwan do we find that Turk is referred to by this word, or by its usual Arabic equivalent, qawm. There is indirect evidence, however, that the Turks are a bodun or qawm. The only people specifically named a qawm in an entry of the Diwan are the Sogdaq (237,6/i, 391,17). Kasgari tells us that they "belong to" or "are from" (wahum min) Sogd, but that they dress and act like Turks. The Sogdaq, then, are not a tribe distinct from other Sogdians; they are simply a group of the Sogdian "people" (qawm) who happen to have settled in Balasagun — a trading community. 2 Another non-Turkic people, who also live "in the lands of the Turk", are the Tibetans — Tubut (179,2/i, 296,11). They are not called qawm, however, but jll "tribe" (see below, entry 3). As for the Turks, Kasgari speaks of them as follows (176,13/i,293,3): Turk. The name of the son of Noah ( G o d ' s blessings be upon him). This is the name by which God called the sons of Turk, son of Noah; just as " M a n " (al-insari) is the name of Adam (peace be upon him) ^ It is a collective (ism jam'), since there is no one who can be excluded f r o m this singular. In the same way "Turk" is the n a m e of the son of Noah, in the singular; but when it refers to his sons it is a collective — like the word " h u m a n " (albasar) — it is used for singular or plural. Similarly, " R u m " is the name of R u m , son of Esau, son of Isaac ( G o d ' s blessings be upon him), and also his sons were called by that name.

^Correcting ms. a'ana to agara, along with B. Atalay, Terciime, iii, 90; and ms. yulaydi to yulidi (instead of Atalay's yuludi). See R. Frye and A. Sayili, "Turks in the Middle East before the Saljuqs", Journal of the American Oriental Society 63 (1943), pp. 200-201. •2 ^Kasgari here cites some Quranic verses, which we may omit . "Tiirk son of Noah" is in this passage a mistake; it should read "Tiirk son of Japheth son of Noah". See, infra, entry 3, and below, note 13.

62

FROM

MAHMUD KA§GARI

TO E V L i Y A

CELEBI

Later in the same entry he cites a verse (177,14/i, 294,10), which begins: qacan korsa ani tiirk bodun... aydaci Kasgari's meaning given to these lines is, "When the groups (.tawa'if) of the Turk see him, they say...." Although this interpretation of the lines is probably wrong, 1 it nevertheless points up that in Kasgari's own mind the term "groups of the Turk" is equivalent to a bodun. Immediately after this verse, Kasgari returns to his discussion of the word Turk in both its singular and its collective usage: One of them is "Tiirk", and so is a group of them. If one says kim san, meaning "Who are you?", the answer is tiirk man, "I am a Turk". You also say tiirk siisi atlandi, "The troops of the Turk mounted".

We may conclude from these passages that the Turk, like the Sogd, Rum, Rus, Furs, 'Arab and other neighbors of the Turk, are most likely to be considered as bodun. T o be sure, these same names are used also to indicate geographical areas, and even political groupings. As for the word "Tiirk", Kasgari includes in this term all of the "groups of the Turk" — as we would say, the Turkic peoples. Oguz, for example, is defined as "a tribe of the Turk" (.qablla min at-turk) (40,4/i, 56,2). Klsewhere in the Diwan, however, "Tiirk" is used synonymously with "Cigil" io indicate a group of related dialects as opposed to those of the "Oguz" (e.g. at 171,4/i,284,15) = "Turkman" (e.g. at 422,13/ii, 253,17). This duality can be explained partly by the analogy of other tribal or national names used as geographical terms, and partly by a then prevailing consensus according to which "Tiirk" designates the ruling clan with all its allied and subject tribes, and is thus a political concept.

3. The "tribe" In the introductory section of the Diwan, Kasgari writes (20,8/i,27,8): The Turk are, in origin, twenty tribes {qablla). They all trace back (ya'tazuna) to Turk, son of Japheth, son of Noah (God's blessings be upon them) — they correspond (bi-manzila) to the children of Rum, son of Esau, son of Isaac, son of Abraham (God's blessings be upon them). Each tribe (qablla) has branches (butun), whose number only God knows, I shall mention only the great tribes (qaba'il) and leave the little ones, except for the branches (butun) of the Oguz-Turkmln.... ''I he interpretation of Atalay (Terciime, i, 352) seems correct, taking bodun as the subject, and tiirk as the second object of kor-: "When they see that he is a Turk, the people say...".

THE TRIBAL AND KINSHIP ORGANIZATION OF THE TURKS

63

He then goes on to list the tribes in geographical order, as follows: Bacanak (near Rum), Qifcaq, Oguz, Y e m l k , Basgirt, Basmil, Qay, Yabaqu, Tatar, Qirqiz (near Sin); then Cigil, Tuxsi, Yagma, Ograq, Caruq, Comiil, Uygur, Tafiut, Xitay (Sin), Tawgac (Masin). About the first ten of these Kasgari states: "All of these tribes (iqaba'il) are opposite Rum, extending toward the East". About the last ten he says: "These tribes (qaba'il) are middling between South and North". He then states: "I have shown each of them in the following circle". On the map, however, which immediately follows his statement, only the first ten names appear (roughly in the geographical order of the list). In place of the last ten tribes we find, in the middle of the circle, the names of cities like Balasagun, Barsqan, Taraz, Bes Baliq, Ku£a, etc. (but also Bilad Uygur and Masin), the implication being that these ten tribes are considered to be sedentary. The first ten, in contrast, are considered mainly nomadic, and on the map, the different areas they occupy are given as Bilad al-GhuzzIya, Maskan Qifcaq, Fayafi Tatar, etc. This distinction is maintained in the linguistic classification which follows the map (24-25/i, 29-30). Masin, Sin and Uygur (along with Jabarqa and Tiibiit) are "sedentary peoples" (ahl al-amsar) (25,1/i, 30,6), while the Comiil, the Qay, the Yabaqu, the Tatar, and the Basmil, who each have a language of their own in addition to "Turkiya", belong to the "nomadic peoples" {ahl al-wabar). Thereafter, however, the linguistic criteria take over completely, and Kasgari groups Qirqiz, Qifcaq and Oguz together with Tuxsi, Yagma, Cigil, Ograq and Caruq, as speaking "pure Turkiya". "Approaching these", in addition, "is the language of Yemak and Basgirt". Finally, "the language of Bulgar, Suwar and Bacanak, approaching Rum", is a type of "Turkiya", with certain peculiarities of its own. In the body of the Diwan, the following entries are all defined as "a tribe of the Turk" (jil min at-turk): Aramut (81,15/i, 124,7), Adgis (61,5/i,89,7) Basmil (242,2/i,399,16), B a c l n a k (245,3/i,404,13), Bulaq (191,2/i,317,7) = Alka Bulaq (77,8/i, 116,7), Caruq (191,15/i,318,16), Comul (199,l/i,330,16), Kancak (241,13/i,399,6), Kiicat (180,2/1,298,6), Ograq (72,9/i, 108,6), Qarluq (238,7/i,393,10), Qay (514,9/iii,118,4), Qifcaq (239,2/1,394,12), Qirqiz (230,ll/i,381,15), Tanut (602,16/iii,268,2), Tatar (207,6/i,344,9), Tawgac (228, right margin/i, 378, omitted), Tuxsi (213,6/i, 354,7), Yabaqu (460,3/iii, 27,10), Yagma (459,4/iii,25,17), Yemak (456,15/iii,22,1). The two most important Turkic tribes in Kasgari's day were the Oguz (to whom belonged the Seljuks) and the Cigil (to whom may be reckoned the Qarakhanids). Kasgari refrains from calling either of them jll. The Oguz are

64

F R O M M A H M U D K A § G A R 1 TO E V L i Y A

gELEBI

called qablla (40,4/i, 56,2). As for Cigil, Kasgari says (198,7/ ¡,329,17) that this is the name for three different "groups" (.tawa'if) of the Turk. The first of these is a nomadic people ( q a w m min ahl al-wabar) inhabiting Quyas. The second and third are named after geographical locations known as "Cigil" — one is a small district (bulayda), and a fortress (hisn), near Taraz; the other a group of villages (qura) in Kasgar. We are also told that the Oguz, who used to fight with the second of these three groups, erroneously call all the Turk from the Oxus to Upper Sin "Cigil". It is this "erroneous" Oguz usage which Kasgari himself follows when he contrasts Cigil — Turk with O g u z = Ttirkmln. The above information then leads us to the following conclusions: (a) The Turk are traditionally held to be composed of twenty tribes, of which ten are nomadic and ten sedentary. The eponymous ancestors of these tribes were the sons of Turk, who himself descended from Noah along the Japhetic line. 1 The Arabic term which Kasgari uses in the introductory section for these tribes is qablla — no Turkic equivalent is given. (b) Of the nomadic tribes Oguz is the most prominent 2 and in the entry devoted to them they are consistently referred to as qablla. Basgirt has no separate entry in the Divvan, but in the entries devoted to the other eight is referred to as jil. (c) Of the sedentary tribes Cigil is the most prominent, and is considered, at least by K a s g a r i , to be the Turk par excellence. Uygur is (68,13/i,101,15) defined by the term wllaya, which we might here render "principality". Xitay (550,7/iii, 180,12) and T a w g a c (228,6/i,378,5) are basically terms for geographical regions, though Tawgac is also referred to as jil, as are the other seven in the entries devoted to them. (d) The term jil, then, is used by Kasgari in the body of the text for nearly all the groups which are reckoned as belonging to the T u r k conglomeration, including not only the twenty "tribes" (qaba'il) mentioned in the introductory section, but also other groups, some even of non-Turkic

In the Sajare-i Tarakime of Abu-l-Ghazi (A. N. Kononov, Rodoslovnaja Turkmen, MoskvaLeningrad, 1958, p. 10 of text), "Turk son of Japheth son of Noah" is one of eight brothers, the ancestor of the Turks through Mogol and Oguz, and of the Mongols through Tatar. Kasgarl's version reflects the common Arabic or Islamic view, according to which Japheth (Yafit) was given the territories to the North and East (ultimate source: Genesis X:2). A parallel Iranian version, in which the Turks derive from Tur, one of the three sons of Feridun, is found in Ibn Hassul, d. 1058 (A. Azzavi, "Ibni Hassul'iin Tiirkler hakkinda bir eseri," Turk Tarih Kurumu Belleten 4 (1940), p. 256; text, p. 37). 2 B u t cf. 450,16/iii, 11,12: "There is a class of Oguz, in their own land, who never nomadize or go on raiding expeditions" wa-jins min al-guzzlya fl biladihim Id yarhaluna ila mawdi' wa-la yagzuna. They are called yatuq, meaning "lazy ones, ones left behind" al-kasala almatruhun. F. Siimer (Oguzlar..., Ankara, 1967, p. 41) is probably correct in identifying the yatuq with the sedentary Oguz. Kasgari, of course, is our chief source for the Oguz cities. See Siimer, ibid:, and O. Pritsak, "The Decline of the Empire of the Oghuz Yabghu," Annals of the Ukrainian Academy of Arts and Sciences in the U.S. 1952, pp. 281 f.

THE T R I B A L AND KINSHIP ORGANIZATION OF THE TURKS

65

origin (Kangak, Tiibiit),1 living among the Turk. No Turkic word is given in the Diwan as an equivalent of jll.

4. The "subtribe" Kasgari uses the term qabila for the twenty "original" Turkic tribes; and once, as we have seen as the equivalent of bodun in its primitive meaning of "tribe" or "kinship group". Most frequently it is found used synonymously with batn, meaning "subtribe" or "branch", in connection with the OguzTiirkmain.2 In the "Oguz" entry (40,4i ,56,2 ff.), Kasgari first states that they are composed of 22 branches (batn), which he lists, along with their distinctive brands, in the following order: 3 Qiniq ("the chief of them (surratuhum) ... from whom are our present Sultans" — i.e., the Seljuks), Qayig, Bayundur, Ewa or Yewa, Salgur, Afsar, Begtili, Biigdiiz, Bayat, Yazgir, Aymiir, Qara

'Tiibiit (179,2/i, 296,11) are defined as "a numerous tribe in the lands of the Turk" jil fi diyär at-turk katir, and an Arab origin is attributed to them. C f . 24,2/i,29,2: "The best speech belongs to those who know only one language, who do not mix with Persians ( a l - f ä r i s i ) , and who do not associate with foreign settlers (lam yata'awwad nuzül al-bilad). The ones who have two languages, and who mix with citypeople, have a certain slurring (rikka) in their speech —for example, Sogdäq, Käncäk and A r g u . The second category are such as Xotan, Tiibiit and some of Tanut — this class are settlers (nazila) in the lands of the Turk." The Käncäk, who lived in the countryside around Käsgar (25,12/i,31,5), were a remnant of the ancient Saka population (see H. W. Bailey, "Saka Studies: The Ancient Kingdom of Khotan," Iran 8 (1970), p. 67). There are five other jll min at-turk who are not named among the twenty qabä'il in the introductory section: Arämut ("near Uygur", also a place-name), Adgis (y and kh>gh, which Kashghari also remarks on, cause no trouble within Turkic. Cf. G. Clauson, "The Name Uygur," IRAS, 1963, pp. 140-159; Doerfer, no. 626. 4 C f . Minorsky, Hudud al-'Alam, p. 94: "Near it [Chinanchkath = Qocho = Turfan] is the mountain T.fqan, behind (az pas) which are five villages: Kuzar. K.j.mlkath, *PanjIkath, Bar]ugh, Jamghar." Except for Panjikath = Besh baliq, this list does not correspond to our five cities, as Minorsky points out (p. 272). It is significant, however, that in each case we have five cities associated with a mountain. (Compare, in the Turkman story: " . . . this is the name of a mountain, called Altun Khan. Afterward Dhu-1 Qarnayn... built the cities of Uighur,") In light of this, "the mountain T.fqan" must be the same as our Altun Qan (T.fqan is easily explained paleographically from a form such as *Altunqan) — and not, as Minorsky suggests (p. 271), a misspelling of Turfan (which is an oasis in a depression and not a mountain). Altun Qan then, in Minorsky's words, "is certainly Eastern T'ien-shan (Boghdo, highest peak 12,080 f.)."

THE

ALEXANDER

ROMANCE

105

vanguard defeats him near Sin = Kashghar in a battle in the T'ien-shan (Altun Khan); he remains in the vicinity to build the five cities (Uighur). There is one further Alexander story in the Diwan, in connection with the naming of the food known as Tutmach (227,13/i, 377, 10):1 Tutmach. A famous dish of the Turks. According to them, it was one of the provisions of Dhu-l-Qarnayn. Thus: W h e n Dhu-l-Qarnayn e m e r g e d f r o m the Lands of Darkness the p e o p l e ' s f o o d supply was short, and they complained to him of hunger, saying: bizni tutma ach, m e a n i n g , " D o n ' t keep us here hungry (la tumsikna hahuna jiya'an) — but let us go, that we may return to our homelands." So he consulted with his advisors about this, and they came up with this food. It strengthens the stomach and reddens the cheeks, and is not quickly digested. A f t e r eating the T u t m a c h one then drinks the broth, double (ad'afuhu). When the Turks saw this they called it T u t m a c h . Its root is: tutma ach, meaning, " D o n ' t leave h u n g r y " (la tujawwi') — the two a ' s were elided, for lightness [of pronunciation]. The sense is, " D o n ' t leave yourself hungry, but take this food and eat it!"

Within this relatively short entry there are actually two different attributions of the origin of the name. First, Alexander's soldiers themselves say tutma ach — "Don't keep us hungry" — a logical thing to say under the circumstances of the story. Second, when the Turks see the wonderfully nourishing qualities of the dish, they call it tutma ach — " D o n ' t leave yourself hungry." It is as if Kashghari, or his informant, remembered that such a name could only be given by Turkic-speakers, and not by Alexander or his troops, who spoke — Persian. Can this episode be fitted into Alexander's route across Central Asia, as we have reconstructed it from the other stories in the Diwanl In other words, where are the "Lands of Darkness"?

^Cf. A Caferoglu, [Abfi Hayyan) Kitdb al-Idrak li-lisan al-Atrak, Istanbul, 1931, p. 64: "Tutmach. Bits of dough that are thrown into meat broth. The word means 'hunger-restrainer'. Tutma is 'don't keep' (la tumsik) and ach is 'hungry' (jayi'an). The name arose by combining these words." See also Doerfer, no. 876; and H. Z. Kogay, "Tiirkiye halk dilindeki yemek adlari," Ne'meth Armagani, Ankara, 1962, pp. 299, 315. From most of the descriptions of this dish it appears to be a kind of dumplings, usually served in broth. Unfortunately, Kashghari never describes it; but note the definition of sish (497, 16/iii, 91,3) as a spit or skewer Cminzam) of Tutmach! The Arabic equivalent of the word in the Diwan is itriyya, a kind of noodles. With Kashghari's Tutmach story compare the contemporary one given by Gunnar Jarring in his Materials to the Knowledge of Eastern Turki, 4, Lunds Universitets Arsskrift N.F. 1,47:4, 1951, p. 145 f. The origin of a special pilav is traced to Iskender — he ordered the wise men, or doctors, to prepare it as a medicine, and as 'food to eat and drink', when all his people became sick during one of his sea campaigns!

106

FROM

MAHMUD

K A § GA R i TO

EVLlYA

£ELEBi

Before answering this question let us note that Alexander is mentioned at one other place in the Dlwan, namely in Kashghari's map. 1 In the extreme Southeast, blocking off the land of Yajuj and Majuj (Gog and Magog), we find the Barrier (sadd) of Dhu-l-Qarnayn. 2 Now these two fabulous places — the lands(s) of Darkness and the Land of Gog and Magog — are the only ones mentioned in the Dlwan, in connection with Dhu-l-Qarnayn, which are also found in the more literary forms of the Alexander Romance, either in its classical sources or in its later, Islamic versions. In all cases they are connected with Alexander's fabulous wanderings in East and West, though they are generally thought of as being North of the civilized world. Alexander builds a Barrier around Gog and Magog in order to keep these fearsome tribes from ravaging civilization. He enters a Land of Darkness during his personal quest for immortality. There are certain other features of the above stories which tie them to the classical versions of the Romance. In each of these versions there is at least one episode describing the lack of food among Alexander's followers, and their complaining of hunger — as in the Tutmach story. A more striking motif is that of king Shu's silver tank, which he uses to divine the movements of the opposing army. In the Pseudo-Callisthenes and the versions of the Romance dependent upon it, king Nectanebos of Egypt employs just such basin-gazing in order to ward off enemy troops. The very first incident of the Romance relates how the barbarian nations came against Egypt; how the sorceror-king Nectanebos practiced his basin-gazing against them, but found the gods in this instance to be siding with the enemy; how he therefore took flight to Macedonia in the guise of an Egyptian astrologer; and how, using his magic wiles, he seduced Philip's wife Olympias and thus became the natural father of Alexander himself. 3 ' Sec A. Herrmann, "Die älteste türkische Weltkarte," Imago Mundi, 1, 1935, 21-28. There is nothing on the map that corresponds to "Lands of Darkness." But there is a "Land of Women" (balda an-nisä') in the extreme Northeast, a land associated with Alexander's wanderings in other versions of the Romance. (Cf. Brockelmann's remark on Jäbarqa in Kö'rösi Csoma Archivum 1 II], 1921, p. 37, n. 18.) 2 At 24, 10/i, 29, 14 Kashghari says that the language of Yäjüj and Mäjüj is unknown "because of the Barrier, and the interposition of the mountains, and the sea which is near Mäsin [China]". Cf. R. Dankoff. "Baraq and Buräq," Central Asiatic Journal, 15 (2), 1971), p. 107'. 3 This is found, among others, in the Greek, Syriac, Armenian, Hebrew and Ethiopic versions. (A. Ausfeld, Der Griechische Alexanderroman, Leipzig, 1907, p. 30; E. A. W. Budge, The History of Alexander the Great..., Cambridge Univ. Press, 1889, p. 1 [see also Budge's discussion, p. xxxvii f.]; A. M. Wolohojian, The Romance of Alexander the Great..., Columbia Univ. Press, 1969, p. 23; I. J. Kazis, The Book of the Gests of Alexander of Macedon, Cambridge, Mass., The Mediaeval Academy of America, 1962, p. 59; E. A. W. Budge, The Alexander Book in Ethiopia..., Oxford Univ. Press, 1933, p. 2.) It is not found in the Iranian versions, where instead we find quite a different legend, explaining Alexander's origin to be from the last Achaemenid Shah. For a full discussion, see E. E. Bertels, Roman ob Aleksandre i ego glavnye versii na Vostoke, Moskva-Leningrad, 1948.

THE A L E X A N D E R

ROMANCE

107

Water sorcery of this type, to be sure, implies that the enemy is approaching by sea. King Shu nevertheless uses it to some effect in Central Asia. Also, his divining is by means of ducks and geese rather than the wax figures of men and ships that Nectanebos had used. In our story, in fact, the meaning of the silver tank seems to have been forgotten, since the narrator, though including the motif, implies that the people did not understand it, and thought that their king was simply passing his time. Shu also sends out a vanguard to spy out Alexander's movements, which would not be necessary for the story if his basin-gazing was known to be a type of sorcery. The narrator has here borrowed the motif of the vanguard that appears in its proper place somewhat later, durng the battle at Altun Khan. In a more original version, we may surmise, there was no vanguard at this point. The king simply practiced his basin-gazing, saw that this was not a propitious time for battle, and therefore quickly retreated, failing to warn all of his troops. That Shu was considered a sorceror-king, however, is clear from the final motif of the story, the attaching of a powerful talisman to his new city of Balasaghun. All of these elements then, reveal a background of romance and legend that were current among the Turks in Kashghari's time, and that revolved around the fabulous wanderings of Alexander the Great. In the Dlwan, the legend and romance come to the surface only in connection with folk etymologies — this is not surprising in a work that is fundamentally a lexicon. What is striking is the lack of similarity between the Dhu-l-Qarnayn episodes in the Dlwan and the story of Iskandar in the Shah-nameh and in other contemporary Persian and Arabic sources. This fact speaks for a longestablished Alexander tradition among the Turks, though we cannot yet say whether they adopted this tradition originally from an Iranian version, from Islamic lore, or from some third source. That Alexander speaks Persian implies, as the Russian scholar Bertels argues, that he "appeared in Turkic territories already in the role of Shahanshah of Iran and Aneran and seemed to the Turks to be not a Macedonian, but an Iranian." 1 On the other hand his epithet, in the Diwan, is Dhu-l-Qarnayn, not Iskandar. This, to be sure, could be the result of Kashghari's translating into Arabic stories which his informers told him in Turkic. But there are also elements in these stories which, as we have seen, appear to derive more directly from the PseudoCallisthenes, and are not present in any of the Arabic or Persian versions. Outside of the Diwan, where the narrative is not a continuous one, we have no full-blown examples of the native Turkic tradition until the Mongol

1

Bertels, p. 177.

108

FROM

M A H M U D

K A § G A R Í TO

EV LtY A £ ELEB I

period. In these however, the material has been reworked, and either Oghuz Khan or Chinggis Khan has been substituted for Alexander. 1 One final puzzle is the personality of king Shu, who is known to us only from the Dlwan. At this point we enter the realm of speculation. On the basis of congruity of narrative alone, and without taking into account the similar appearence of the proper name, we may guess that the figure of Shu is based on the historical king of the Tiirgesh, or Western Turks, who ruled from 716 to 738. The Chinese sources call him Su-Lu; the Arab historians speak of him simply as Kháqán. 2 His capital was at Süyáb or at Nawákath, both places in the Chu river valley near Toqmaq. 3 A few years before Su-Lu's rise to power, the Arab commander Qutayba ibn Muslim conquered Samarqand, Khojend and Káshghar, in a devastating campaign which a modern historian describes in the following terms: "After clearing the enemy from Transoxiana, he crossed the Jaxartes in 712 and exacted the submission of Shash and Khujand; two years later he pushed forward into eastern Turkestan, a land which even Alexander had never seen, and he entered, if they did not hold, the town of Kashgar, on the frontier of China." 4 There then began in 721, what Gibb called the "Turkish counterstroke" under Su-Lu's commander Kürsül (Kul Chor). About the year 724, a son of Su-Lu routed the Arab troops advancing on Farghána and chased them back across the Syr Darya, in a famous battle which the Arabs knew as the "Day of Thirst". For fifteen years the Turks remained supreme in Transoxiana, until the Arabs finally defeated Su-Lu at the battle of K h a r i s t á n , west of Balkh, in 737. A year later Su-Lu was killed by Kürsül, and his kingdom ' T h e Oghuz Romance is found in three sources: t h e J ä m i ' at-Tawärikh of Rashid ad-Din (K. Jahn, Die Geschichte der Oguzen des Rasid ad-Din, Ö s t . Ak. d. Wiss., Pli.-Hist. Kl., Denkschriften, 100. Band, 1969, p. 20 ff.); the so-called "Uighur" version of the Oghuznameh (W. Bang and G.R. Rachmati, "Die Legende von Oghuz Qaghan", SPAW, 1932, pp. 683-724; and the Shajare-i Taräkime of Abu-1-Ghäzi (A.N. Knonov, Rodoslovnaya Turkmen, MoskvaLeningrad, 1958, p. 17 ff.) In all of these Oghuz Khan follows a course of conquest of the habited world analogous to that of Alexander in the Diwan, giving names, as he goes, to various Turkic tribes. All of the etymologies, be it noted, are Turkic — this accords with the more nationalistic tone of the Oghuz legends. The Mongol version is known especially from the recently discovered "Tartar Relation" (R. A. Skelton et al., The Vinland Map and the Tartar Relation, Yale Univ. Press, 1965, p. 62 ff.) It is in this form that the Turkic version of the Alexander Romance entered European literature. (For a suggestion of an unknown Uighur written version, see F .W. Cleaves, "An early Mongol Version of the Alexander Romance," Harvard Journal of Asian Studies 22, 1959, p. 27.) ^E. Chavannes, Documents sur les Tou-kieu (Turcs) Occidentaux, St. Petersburg, 1903, pp. 44f., 81 f., 284 f. Tabarl, Annales, 2, pp. 1421 f., 1478 f., 1512., 1593 f. The Arabs nicknamed him Abu Muzähim, "Charging Bull". Note in the Uighur story above that Shu is referred to as "the Khäqän of the Turks." 3

S e e V. V. Barthold, "History of Semirechye," Four Studies on the History of Central Asia, tr. B. and T. Minorsky, Leiden, 1962, p. 83 f. In his Zwölf Vorlesungen (see p. 101, n. 1), Barthold derived the name Shu from the name of the river, Chu (p. 80). See also Minorsky, Hudüd al'Älam, 300-304. J. Saunders, A History of Medieval Islam, London, 1965, p. 89.

THE

ALEXANDER

ROMANCE

109

collapsed. With it, according to Gibb, "disappeared the last great Turkish confederation in Western Asia for more than two centuries to come" 1 — i.e., until the rise of the Qarakhanids. Whether or not the identification of Shu with Su-Lu is correct, it is clear that Kashghari's narrative embodies a national memory of a powerful Turk king, his victory over a conqueror from the West, and his establishing his capital on the Chu. The fortress of Suyab is connected with Balasaghun, in the same vicinity, which the Turks conquered in 942, before they had converted to Islam, and which later became a dynastic seat of the Qarakhanid empire. 2 We remain in the dark as to the steps by which the conqueror from the West became identified with Alexander the Great. In line with the foregoing hypothesis, two possible intermediaries suggest themselves: the Arab foe, and the Soghdian ally, of the Tiirgesh. It is not impossible that the Arab troops under Qutayba and the other commanders in Central Asia looked on themselves as heirs to Dhu-l-Qarnayn, the paradigm of a "Muslim" warrior hero; and that the story grew up in camp circles of both parties that a sorceror-king of the Turks had, for a time, defeated the World-Conqueror. The relations of the Turks with the Soghdians, at this period, were much more fruitful than with the Arabs. It is well known that the Turks had close contacts with the Soghdians ever since their first appearence in Central Asia, in the sixth century, and that the Soghdians influenced them in trade, diplomatics, religion and art. It is now also known that the Soghdians exulted in the Iranian heroic past. Furthermore, in their wall paintings, for exemple at Pendzhikent, they depicted heroes of the past alongside of contemporaries. 3 If the Turks borrowed the Alexander Romance from the Soghdians, they did so at a point in time when the Alexander cycle had not yet been fully assimilated into the Iranian epic. 4

' H.A.R. Gibb, The Arab Conquests in Central Asia, London, 1923, p. 85. S e e art. "Balasaghun" in the new edition of The Encyclopodia of Islam. In the Uighur legend given by Juvaini, Balasaqun was founded by Buqu Khan, identified with Afràsiyàb (J. A. Boyle, The History of the World-Conqueror ... 2 vols., Harvard University Press, 1958, 1, pp. 54, 58). 3 See, e.g., A. M. Belenitski and B. I. Marshak, "L'art de Piandjikent à la lumière des dernières fouilles (1958-1968)," Arts Asiatiques, 23, 1971, p. 22. Soghdian was still spoken in Balasaghun In Kashghari's time (25, 10/i, 31,1). 4 T h e problem of the Alexander Romance in the Dlwàn Lughàt at-Turk must ultimately be discussed in connection with the problem of Afrâsiyâb, another figure from the Iranian epic, who also appears in the Dlwàn. See, for the time being, R. Devereux, "A Note on Afrâsiyâb," Muslim world, 1963, pp. 141-144; and the present writer's doctoral thesis, Harvard, 1971, p. 346. Also: R. N. Frye and A. M. Sayili, "Turks in the Middle East before the Saljuqs," J AOS, 63, 1943, pp. 194-207. 2

8. BARAQ AND BURÄQ

In the Diwan Lugat al-Turk ofMahmud Kasgari (circa 1075), there is the following entry:1 Baraq. "Hairy dog" (kalb 'ahlab). The Turk claim that when the eagle (vulture? al-nasr) grows old, it lays two eggs, then hatches them. From one of them emerges this dog, called baraq. It is the swiftest-running of dogs, and the most reliable ( ' a h f a z ) in hunting. From the other emerges a chick — the last of its chicks.

This intriguing passage suggests a number of questions. What is the meaning of the hairy dog born of an eagle, and why should it be called baraq? Is baraq a Turkic word, and if not, what is its origin? And finally, is there any connection with Buraq, the fabulous steed of the Ascension in Islamic belief?

I

In a note written in 1948, Abdiilkadir inan mentioned this notice of Kasgari, and cited two contemporary versions of the same folk-belief: 2 Another version of this legend is gien by Yakovlev for the Soyot-Urenha fUryankhai-Tuvinian] Turks (NE Mongolia). According to the belief of the Urenha, the bird Megaloperdix altaica gives birth to three things in the course of its life: a dog {baraq), a rope, and an axe. According to a legend I have heard among the Saljuvut, Katay and Barin clans of the Baskirs (E plain of the Urals), the raven, when it grows old, lays two eggs; from one of them emerges a hairy dog called tuygun, and from the other a hunting bird called sonqar. These creatures are said to be suitable as a hunting-dog and hunting-bird for princes. As far as I know, these are the only reflections among contemporary Turkish peoples of this "Baraq" legend, which is transmitted by M a h m u d Kasgari in the eleventh century.

1 Facsimile (B. Atalay, Ank. 1941) p. 190, line 3; Editio princeps (K. Rif'at, 1st. 1333) 1, p. 315, line 13; Tercüme (B. Atalay, Ank. 1939) I, p. 377-8. ^Belleten XIII (1948), p. 49; reprinted in Makaleler ve incelemeler (Ank. 1968), p. 219.

112

FROM

MAHMUD

K A § G A R i T O EV Li Y A

gELEBI

This is, however, not the last word on the subject. If we look separately for a moment at the mother-bird and the hairy dog, we find that each has a well-defined role in contemporary Altaic beliefs. L. Sternberg, in a discussion of the eagle cult among Siberian peoples, 1 wrote of the eagle which hatches eggs, "from which is then born the one chosen to be shaman." In the conception of the Yaqut, 2 the eagle swallows the soul (giir) of the child chosen to be shaman, whereupon it flies out into fields "with sun and moon," and there, on a holy birch planted for this purpose, hatches an egg, which it then breaks open. The child ... is placed by it in an iron cradle, standing at the foot of the tree. He is guarded and nourished by the so-called "mother-bird," and raised to be the Ayu Oyuna, i.e. shaman.

The association of the eagle with the first shaman provides our first clue to the mystery of baraq. M. Eliade writes: 3 "The majority of myths concerning the origin of shamans posit the direct intervention of God, or of his representative the eagle, the bird of the sun." He cites the Buryat myth, among others, in which the first shaman was the son of a woman and an eagle. In speaking of the shamanic costume, 4 he notes that it most commonly imitates an eagle. In the Buryat conception, according to U. Holmberg, 5 "the 'animal' (eagle| which inspires the 'first shaman woman' of the clan is also looked upon as a possible cause of her pregnacy. In this manner tales might arise of the animal-like ancestors of a clan." Holmberg goes on to say: One of the shaman's protective spirits in animal form is commonly regarded as being intimately connected with the shaman himself ... The Yakuts call a shaman animal of this description ija-kyl ('mother-animal'). These may be of varying species. The mightiest shaman animals are said to be the stag, the stallion, the bear, the eagle, etc. Unlucky the shaman whose ija-kyl is a wolf or a dog. The dog, it is said, never leaves the shaman in peace, but 'gnaws with its teeth at his heart and tortures his body.'

At other times and places, of course, the dog might have been regarded in a kindlier light, especially under Iranian influence.

' "Der Adlerkult bei den Völkern Siberiens," Archiv für Religionswissenschaft XXVIII (1930), p. 130ff. Sternberg, p. 133; citing Castren, Klein. Sehr. 232. 3 Shamanism (Bollingen, N.Y. 1964), p. 69. 4 Eliade, p. 156. 5 "Finno-Ugric, Siberian," The Mythology of All Races, IV (Boston, 1927), p. 506-7.

B A R A Q

AND

B U R A Q

113

Just as the eagle is associated with the first shaman, the hairy dog plays a role in the creation of the first man. 1 In the myth of the Black Tatars, the naked dog, set to guard over the yet lifeless people, is promised some golden hairs if he give them over to the devil Erlik. In a Yaqut legend, "Man" is promised an indestructible garment; but, as a punishment, he is changed into a dog. In these stories, man's defilement is brought about by the spittle of the devil, while his guardian receives hair, as reward (from the devil), or as punishment (from the Creator). "In another cycle of tales, in which the devil soils the people whom God had created by spitting on them, these people had originally some covering, hair or nail-matter." In this case, the creator curses the dog, on whose body he sees the devil's hair-covering, and cleanses the peoples' bodies of the hairs which the devil had soiled (Buryat). In a variant, the removal of the hair is the devil's work, and introduces mortality to man. In another version still, man loses his protective covering becuse he eats the forbidden fruit (Altai). Holmberg justifiably suggests a Near Eastern origin for these ideas, since they are also found in Jewish and Arabian tales. 2 In a Manichaean version, 3 hairy people are one of the ten races of mythical men, grown on the tree from which the first couple, Masya and Masyoi, were detached. Kasgari's notice of baraq, as a hairy dog born of an eagle, represents a conflation of these two themes of the first shaman and the first man. Now, what about the word baraq itself?

II If we examine first the living Turkic languages, we find a range of meanings for baraq that is bewildering. The new Derleme Sozliigii4 gives no less than ten separate entries under baraq, the first of which alone covers the following animals: hairy man, hairy dog, black dog with white spot; guinea pig (? kobay), mole; chaffinich (ispinoz ku§u [= fringilla coelebs according to Redhouse, new edition|); a yellow, hairy catterpillar found in orchards. The other entries are: (2) pockmark; (3) child; (4) bewildered; (5) lookout, informer; (6) dry, barren; reed bed that has gone green; dirty lake; (7) crooked; (8) flower or decorated handkerchief sent from a girl's house to a boy's house; gilt paper; (9) mixed, of mixed colors. Only the last — (10) a cushion etc. — brings us back to the hairy quality of the animals mentioned.

1

Holmberg, p. 373 ff. Holmberg, p. 384. For the Jewish version, cf. Pirke de Rabbi Eliezer (tr. G. Friedlatidet, N.Y. 1965), p. 98. 3 A.J. Carnoy, "Iranian," The Mythology of All Races VI (Boston 1917), p. 298. 4 Turkiye'de halk agzmdan derleme sozliigii II (Ank. 1965). 2

114

FROM

M A H M U D

K A § G A R Í TO

EVLÍYA

£ ELEB 1

The new Redhouse 1 has: barak 1

obs.[olete] plush; long-piled cloth.

barak 2

prov.[incial| long-haired dog.

R a d l o f f 2 knew baraq,

meaning "with thick (long) hair," only in

Ottoman and "Kirgiz" (=Kazakh). He gave the following compounds (the first two apparently belong to Ottoman): qil baraq "a race-horse with long mane and tail," qara baraq "a species of hunting dog," baraq qoy (Kir. |(Kazakh|) "a sheep with thick wool," and baraq it (Kir. [Kazakh]) "a dog with curled hair." K i r g i z 3 has: baraq

qulaq

at " h o r s e with shaggy ears

(non-

thorougbred)," and baraq el "(myth.) a hairy people: ilgeri baraq el bolot decii ele, tula boyun tiik basqan (folkl.) In former times, they say, there was a people [called] Baraq — their bodies were all covered with hair." 4 In Uzbek 5 baraq means "shaggy, wooly," referring to a dog (it), a cat (musuk) or an eye-brow (qas). In Yaqut 6 barax is a certain bird (snipe, scolopax gallinago). Baraq is a bird, too, in Turkmen 7 ( k a s k a l d a k , coot, water-hen). The word was borrowed from a Turkic language (Uzbek?) into Persian. 8 Hai'm gives buraq (read baraq), designating a cat "with bristling hair and attacking aspect." Similarly, colloquial Tadjik has baraq meaning "white and black (animal color), hairy (cat)." 9 The evidence of the modern languages sufficiently proves that baraq signifies "hairy" and not "hairless dog" as Kopriilu thought. 1 0 As for the older written languages, outside of Kasgari, we find baraq as a type of dog in two of the Qipcaq glossaries 11 and Baraq as the proper ^New Redhouse Turkish-English Dictionary (1st. 1968). I V , 1477. K. K. Judaxin, Kirgizsko-russkij slovar' (Moskva, 1965), p. 109. ^Judaxin also gives: baraq "fertile land" (Southern dialect), and baraq celek "wooden bucket." 2

^ B o r o v k o v , Özb. Rusk. Slovar', 55, as cited in G. Doerfer, Türkische und Mongolische Elemente im Neupersischen, (Wiesbaden, 1965), II, p. 280 (no. 728). 6 E . K . Pekarskij, Slovar' jakutskogo Yazyka (1958) I, p. 374. ^Baskakov et. al., Turkmensko-russkij slovar' (Moskva, 1968), p. 71. "Kaskaldak" is the same as K ä s g a n ' s qasgalaq (1434,8). ®The One Volume Persian-English Dictionary (Teheran, 1961). For the typical confusion of baräq and buräq, see p. 117, n. 2. 9 C i t e d in Doerfer, op. cit. Here, and in one or two of the meanings in the Derleme Sözlügü above, there is an apparent contamination with the Arabic labraq\ see below. 10 M . F . Köpriilüzade, Influence du Chamanisme Turco-Mongol sur les Ordres Mystiques Muslumans (1st. 1929), p. 14-15, n. 26 "chien sans poils." This mistake is probably due to a secondary meaning of the Arabic 'ahlab in Käsgari's definition, "hairless or clipped (tail)." 11 Ettühfet-üz-zekiyye 30 b 11; El-Idräk ha^iyesi (V. tzbudak, 1st. 1936), p. 7.

BARAQ

AND

BURÄQ

115

name for a dog in Dede Korkut.] For Cagatay, Pavet de Courteille 2 gives bduq (read baraq), meaning "long-haired dog" and "animal with a fetid odor and of which the lion is afraid."

m Baraq is mentioned at least twice in the Jami' al-Tawarix of Rasid alDln. Once is in the context of a Turkish folk tale, in which the hero's marvelous dog Qara Baraq plays an important role. 3 The other is in the context of the (mythical) conquests of Oguz. 4 After Oguz subdues the Qipcaq, he goes into the dark regions beyond the Itil (Volga) and conquers the people called K1 Baraq. Of this people, the information is given that the men resemble dogs, while the women are noted for their beauty; that their chief is the invulnerable It-baraq; and that in this region arose the Qipcaq. The only Turkic version of this entire episode is in the Sajarah-i Tarakimah of Abu-1 Ghazi, 5 where the name of the ruler is given as It-baraq Xan. The same Itbaraq is also mentioned in the Sanglax,6 Kasgarl apparently knew nothing of the Oguz legend, which in fact became current only during the Mongol period. All the folk-etymologies of tribal names (Qalac, Qanli, etc) which in Rasid al-DIn and Abu-1 Ghazi are connected with Oguz, in Kasgarl are connected with Du-1 Qarnayn (Alexander the Great). On Kasgari's map there is no Land of Darkness; while the Wall of Du-1 Qarnayn and Land of Yajuj and Majuj (Gog and Magog) are found in the extreme south-east! 7 Pelliot, in discussing the Land of Darkness in Marco Polo, 8 remarked : "In my opinion, both Rasld's source and Polo are in fact indebted, at least indirectly, to some Oriental rifacimento of the romance of Alexander." In fact, the entire episode of Oguz in the Land of Darkness parallels the Alexander legend, including the motif of the smuggled old man and the ruse of the suckling foals.

' D 114,1 "itiingiiz adi Baraq degiil mi-y-di." Dictionaire Orientale. For the correction of bduq to baraq, cf. P. Pelliot, Notes sur Vhistoire de la Horde d'Or (Paris 1950), p. 57. K. Jahn, "Die ältesten schriftlich überlieferten türkischen Märchen," Central Asiatic Journal (XII 1, 1968), p. 32. K. Jahn, "Zu Rasid al-DIn's 'Geschichte der Oguzen und Türken'," Journal of Asian History (I 1, 1967), p. 45ff.; ders., Die Geschichte der Oguzen des Rasid ad-Din (Denkschriften, d. Osterr. Akademie d. Wissenschaften, 100. B., Wien 1969), pp. 24-25. in the latter work, only as Qi'l Baraq, "shaggy dogs." "'A.N. Kononov, ed., Rodoslovnaja Turkmen (Moskva, 1958), p. 18. 9 5 v 13. For the geographical tradition lying behind this, cf. A. Herrmann, "Die älteste türkische Weltkkarte," Imago Mundi (I, 1935), p. 22. ^ Notes on Marco Polo II (Paris 1959), p. 618. 6

116

F R O M M A H M U D K A § G A R I TO E V L I Y A

gELEBI

K. Jahn reads R a s i d ' s Kl-baraq as Kilbark and says, "perhaps it is a matter here of a corrupt writing of Ka/cucrtpioi, as, according to Ctesias, the Ki!VOKE(|)«Xoi were called by the others." The identification of K1 Baraq with Kynokephaloi is felicitious, though 1 do not think it is a matter of a "corrupt writing," at least as far as Baraq is concerned. Rather, Baraq must have been identified with Kynokephaloi within the Oguz tradition. The Kynokephaloi are included as one of the twenty-two nations imprisoned by Alexander within the northern gate in some versions of the Romance. 1 It was evidently natural for the "Oriental" version lying behind the Oguz legend to identify the "dogheaded men" with Baraq. "The Buriats tell of a land in the north-east, where the men are born as dogs, larger, however, than ordinary dogs, while the opposite sex are born as ordinary women." 2 Another parallel is the "dogs the size of oxen" beyond a land of darkness, in the Qirqiz story according to Marvazi. 3 Minorsky, in his note to this passage, identified It-baraq as *iti-baraq, "one whose dogs are hairy;" he went on to say, on the basis of the passage in Marvazi, "If my suggestion has some truth in it, we should look for It-baraq on the middle Yenisei." 4

e.g., in the Greek "Pseudo-Callisthenes" — cf. H.R. Anderson, Alexander's Gates, Gog and Magog, and the Inclosed Nations (Cambridge, Mass., 1932), p. 41; and in the Syriac "Book of the Bee" — cf. W. Budge, The History of Alexander the Great (Cambridge 1889), p. 150, n. 3. ^Holmberg, op. cit., p. 504, cf. Yule's comments on Polo's description of the Kynokephaloi of the Andaman Islands (Book of Ser Marco Polo, 3rd ed., London 1929, II, p. 311): "The Doghead feature is at least as old as Ctesias. The story originated, I imagine, in the disgust with which 'allophylian' types of countenance are regarded, [quoting Ibn Sa'id:] 'The Boriis (Prussians) are a miserable people, and still more savage than the Russians.... One reads in some books that the Bonis have dogs' faces; it is a way of saying that they are very brave.' Ibn Batuta describes an Indo-Chinese tribe on the coast of Arakan or Pegu as having d o g ' s mouths, but says the women were beautiful. Friar Jordanus had heard the same of the dog-headed islanders. And one odd form of the story, found, strange to say, both in China and diffused over Ethiopia, represents the males as actual dogs whilst the females are women...." (I am grateful to Joseph Fletcher for the following information [written communication, June 25, 1970]:) Liang Fen (1640-1728), in his Ch'in-pien chi-lüeh (Sketch of the Northeastern Frontiers), written ca. 1685 and contained in Li P'ei, ed., Hui-hua chi (Collection on Strategy), published by the Hsiin-lien tsung-chien pien-chi chii, ca. 1910, has a biography of Galdan of the Zunghars (pp. 1242-1249). In it, he says that Galdan heard that west of Zungharia (the Kirghiz and Kazakh territories?) "there were men with dog-like bodies who could run several hundred li in a day and whose women were extremely beautiful," that Galdan attacked them, and that he carried off several of their women. See further: W. Eberhard, Lokalkulturen im Alten China II (Peking, 1942), p. 498; also, A Caferoglu, "Türk onomastiginde 'Köpek' Kiiltü," Belleten 1962, p. 3f. 3 V . Minorsky, Marvazi on China, the Turks and India (London 1942), p. 30. 4 p . 105. In the late Uygur Oguz-name (ed. Bang and Rachmeti, SB A W 1932), line 296, one of the lands which Oguz subdues is called Barqan, which, as the editors note (p. 714), can also be read Baraqa. It is questionable whether this is the same as It-baraq or KL B R ' Q , especially since it is located in the South (kündünki bulungda).

B A R A Q

A N D

B U R A Q

IV As a proper name, Baraq is attested in the Islamic sources only for the Mongol period and later. 1 Pelliot showed that Baraq is the correct form for the "arabizations" Buraq ~ Boraq in such names as Buraq Xan. "Besides," he wrote, "the name must have existed already among the Qara-khitai, since there is no reason to read 'buraq' as did Barthold, Turkestan, 364, the Baraq of Juwaini, II, 211." 2 According to B. Lewis, "The name is not infrequent among Mongols and Turks in the 13th-15th centuries (for some examples see G. Moravczik, Byzantino-turcica, Budapest 1942-3, s.v. papaKoa and jtapaK

j t a p a K . . ,)."3

Perhaps the best known figure by this name is Baraq Baba (ca. 1300) who, according to Kopriilii, "offers a fine example of the influence of Mongolian shamanism on Sufism." 4 Kopriilii discussed three aspects of Baraq Baba's personality, pointing to the shamanic origin in each case: 1) his habit of wearing a two-horned head-dress, ornamented with bird feathers and other objects; 2) his practice of riding on birds and other wild animals; 5 and 3) the style of his tonsure, which was to shave the beard but let the mustache grow long. Kopriilii also noted the report in Rawdah al-safa that the Mongolian shaman who gave the title of Djengiz to the Mongol conqueror counted, among his other prodigies, that of having mounted to heaven on a horse. 6 Of greater interest at this point is how Baraq Baba received his name. The fullest version of this story is in the Tarix-i Al-i Selcuk'J Saru Saltuk goban iken Seyyidi Mahmud, ana ugrayugeliip yogurd He bir lokma agzina virmi§idi, ol lokma, bir bakla mikdari bez gibi olup damagina yapi§up dururidi. Saru Saltuk'a kusmak vaki' oldi, kusunduyila ol lokma dii§di. §ehzadeye Saru Saltuk'un sohbetine abdallik vaki' olmu§idi, o kusunduyu yidi; Saru Saltuk, am baragum diyip ok§adi ve Sultaniye tarajina gonderdi. ^The Hypatian Chronicle for the year 6691 (1183 A.D.) names Barak as one of the Polovtsian (QipCaq) chiefs taken prisoner by the Kievan prince Svjatoslav VsevolodiC (Letopis po ipatskomu spisku, Spb. 1871, p. 427). I owe this referenced to Professor Omeljan Pritsak, to whom I am indebted for much valuable assistence in preparing this paper. 2 Notes ... Horde d'Or, p. 57. Cf. also his remarks in Notes on Marco Polo I (Paris 1959), p. 75: " B o r r a q is one of these later 'arabizations' which have for instance given birth to a Saiban and a Saibanid dynasty, when the original Mongol form was Siban > Siban.... " 3 E / 2 I, 1031. 4 Chamanisme, p. 14 ff. See further A. tnan, Makaleler ve incelemeler, p. 476-7. ^Eliade (p. 403) suspects a Southern influence on the choice of the ostrich as Baraq Baba's mount. 6 p . 18; cited at length in his Edebiyat Ara§ttrmalari (Ank. 1966), p. 62. ^quoted in A. Golpmarh, Yunus Emre ve Tesavvuf (1st. 1961), p. 18-19. We are reminded of Proverbs x x v i . l l , "As a dog returneth to his vomit, so a fool returneth to his folly."

118

FROM

MAHMUD

K A § G A R I TO

E V LÌ Y A

CELEBÌ

While Saru Saltuk was yet a shepherd, Seyyidi Mahmud paid him a visit, and placed in his mouth a morsel of f o o d along with some yogurt. This morsel was like a piece of cloth [or, a gland or cyst? — bez\, the size of a bean, and stuck to his upper palate. Saru Saltuk once had a fit of vomiting, and this morsel fell out with his vomit. The prince had fallen into a trance at Saru Saltuk's discourse, and he now ate up the vomit. Saru Saltuk carressed him, and called him ' m y Baraq', and sent him toward Sultaniye.

The Tarix-i Berzali1 gives the further information that baraq means "dog" in the language of Qipcaq. A third source, the Jami' al-Duwal,2 gives a second tradition as to why Saru Saltuk called him barakum "my dog": Sebebi de Saltuk'un yamna geldigi zaman rahipleri gibi saglarinm uzun olmasiydi. The reason is that when he came to Saru Saltuk, his hair was long like that of Christian monks.

It should be remembered that he was a Christian at this time. 3 P. Wittek 4 cites classical sources that the saliva was considered as a means for conferring spiritual power, but says that "the account of how our Baraq received supernatural powers may be suspected to be nothing more than an explanation of the name." However, there is evidence 5 that spitting in the month was part of a shamanic initiation.

V Drawing some of these threads together, we may say that the word baraq is widely disseminated in the Turkic languages, particularly associated with Qipcaq, and not attested outside Kàsgari (who had relations with the Qipcaq as well as with some proto-Mongolian groups) before the Mongol period. The basic meaning seems to be "hairy" as applied to various animals, but there are several other more or less related meanings. Long hair is a mark ^Golpinarh, p. 20. ^Golpinarh, p. 18. 3 C f . Gòlpinarli, p. 17ff. 4 BSOAS 14 (1952), p. 650. ^Eliade, p. 114; citing A. Friedrich and G. Buddruss, Schamanengeschichten aus Siberien, p. 169 ff. Hakim Ata (d. 1186-7), one of the successors of Ahmed Yesevi, is said to have begun reciting verse after the Prophet Khidr had spit into his mouth. Cf. Koprulii, Turk EdebiyatVnda ilk Mutasavviflar, Ikinci Basim (Ank. 1966). p. 75. This is cited by llhan Ba§goz in "Dream Motif in Turkish Folk Stories and Shamanistic Initiation," Asian Folklore Studies (Tokyo), 1967, p. 12. In another tradition (ilk Mutasavviflar, p. 22), Ahmed Yesevi himself received inspiration from the Prophet Muhammed through the intermediary of Arslan Baba, in the form of a date mixed with some of the Prophet's spittle; Arslan Baba, one of the 'Ashàb (contemporaries of the Prophet), preserved it in his mouth for 400 or 700 years, until he could pass it on to Ahmed Yesevi.

B A R A Q

A N D

B U R A Q

119

of shamans in particular, 1 and of culture heroes generally. 2 It is safe to conclude that baraq was a favorite descriptive word for Turkic shamans and shamanic animals especially, and for other more or less religiously regarded culture heroes as well, in various times and places. 3 The connection with "shaman" would be particularly apt because of the well-known habit of shamans of assuming the personality, and even the frightening visible aspect of certain hairy animals, notably the bear. The complex of belief in shamans born of eagles, and in certain "shaman animals," lies behind Kasgarl's notice about baraq. In Kasgari, apparently, the shamanic content has been lost, and what remains is simply a folk-belief. On the other hand, baraq as a descriptive term is applied to certain mythical peoples, either, as among the Kirgiz, to the first people, who originally had hair as a protective covering over the entire body, but lost it as they passed out of a state of grace; or to the dog-men living somewhere in the north. In the latter case they were indentified with the Kynokephaloi of the Alexander legend.

VI The word baraq itself must be of non-Turkic origin. This is evident under the assumption that any Turkic word must be explained as monosyllabic root plus suffixes: there is no root in the Turkic languages that might serve as an etymon for baraq. (The root bar\verbal | means "to go"; bar [nominal] means "existing; all," etc.) The most likely origin for a Turkic word of cultural significance is in Iranian. And, in fact, we find: Avestan bar- "to ride"; Old Persian asa-bara"rider, horseman"; S o g d i a n f i ' r ' k ~fi'ryh "rider"; Saka bargya- "rider", baraa (< *baraka) "riding-animal", assabdraa "rider"; Pahlavi barak, New Persian baragi "horse". 4

^Eliade, p. 152, n. 30; p. 407. C f . the description of Oguz in the Uygur Oguz-name (line 1 Iff.): qiriq ktindin song bddiiklddi, yiiridi, oinadi, adaqi ud adaqi tag, billari bori billdri tag, yagri kis yag ri tag, kogiizu adug kogiizu tag irdi. badaninung qamagi tiik tiiluklug irdi (Forty days later he was grown up, he ran and danced. His feet were like those of the bull, his loins like those of the wolf, his shoulder-blades like those of the sable, his breast like that of the bear. His body was all covered with hair.) o A Inan, in Tarihte ve Bugiin §amanizm (Ank. 1954), mentions Kirgiz ceremonies revolving about the recent figures of Baraq Tore (p. 194) and Baraq Batir (p. 61). 4 C f . V. I. Abaev, Istoriko-etimologiceskij slovar' osetinskogo yazyka, I (Moskva-Leningrad 1958), pp. 232 (bayrag "colt"), 236-7 (baraeg "rider, horseman"). 2

120

F R O M M A H M U D K A § G A R i T O EV Li Y A Q E L E B I

Either the Sogdian or the Pahlavi form of this word is a likely candidate for our baraq. But a question remains as to the semantic shift from "horse, mount" to "hairy dog" and the other meanings we have found in the Turkic languages. It should first be noted that barak in Middle Persian signifies no ordinary horse. Rather, it is the sobriquet for a fantastic mount, either the devil Ahriman, whom one of the mythical Persian kings, Tahmorap, subdued and rode upon "as his charger" (pat barak) for thirty years; 1 or else the "fiery horse" (ataxsik ... *barak i 'asp) of the apocalypse, which "will be seen by night and in the atmosphere... conceived by the spiritual gods ... and fearful (to behold)." 2 We may offer the hypothesis that it was in this religious complex that the word was first borrowed into the Turkic milieu — that is, as a name for the shamanic mount; that from this point the usage expanded to cover other shamanic animals, and even the shaman himself; and finally (because of the identification: First shaman = First man ?) came to designate the hairy dog, around which many of the later usages crystallized. 3 If baraq was originally the shamanic mount, this would explain why it means "bird" in two or three of the modern languages. In the Altai ritual described by Radloff, 4 the shaman twice changes mounts from a horse to a goose. Flying horses and birds are favorite shamanic mounts in many traditions, and so we can understand how the word baraq could settle on any of these particular animals in the various languages.

VII Is there a connection between Baraq and Buraq, the fabulous beast of the Ascension of the Prophet in Islamic tradition? Let us first note the many shamanic analogues in the description of the Ascension ( M i ' r a j ) . These do not, of course, imply a direct Siberian or Altaic origin: Eliade shows that such elements are nearly universal in world religions, and the immediate sources to their appearance in Islam may be in Christian hagiography (Syrian and Ethiopian), Indian oral tradition (Hindu or Buddhist), Zoroastrian or Jewish eschatology, and even pre-Islamic Arabian magic. But the same can be said for most of the language, concepts and ' R. C. Zaehner, Zurvan, A Zoroastrian Dilemma (Oxford 1955), p. 398. Ibid„ p. 354 (text, p. 348). 3 Note also the close association of vulture and dog in later Zoroastrian and especially Mithraic myths: cf. Zaehner, The Dawn and Twilight of Zoroastrianism (London 1961), p. 131. 4 Eliade, p. 191ff. 2

B A R A Q

A N D

B U R A Q

121

imagery in Islamic religious texts. The later Mi'raj traditions, however, tend, in their selection and arrangement of details, to conform to the strict shamanic pattern which Eliade outlines. In the mature recension of the story, 1 the Ascension proper is preceded by the Prophet's dismemberment and the purification of his internal organs. This is the shamanic initiation experience par excellence. 2 The Ascension itself takes place on a white riding beast named Buraq, commonly described as "smaller than a mule yet bigger than an ass, whose every bound carried him as far as his eye could reach." There is disagreement as to whether Buraq is male or female. He is later depicted, especially in iconography, 3 as a winged horse with a woman's face, but also as one of a number of other animals, including a duck (see below). We have noted the shamanic practice of mounting wild beasts, especially the horse, in order to ascend to heaven or descend to the underworld. In another version, 4 the vehicle of the Ascension is a green rafraf, usually taken to be a type of silk scarf, though the word also means "skylight," and is perhaps originally the rainbow. In any case, there is a shamanic analogue. 5 In the course of Muhammad's tour of heaven and hell, there are at least three important eschatological elements that also strike us as shamanic. The first is the sidrah al-muntaha, the "lotetree of the boundary," which corresponds to the shamanic World Tree. The second is the bayt alma'miir or Frequented Fane, which is the celestial Ka'bah, the navel of the celestial world, situated directly above the earthly Ka'bah. The third is the bridge sirat, narrow as a hair, which only the faithful succeed in crossing. 6 If we compare this version with the earliest and the intermediate descriptions of the Mi'raj, we find that the shamanic cast of the story is a result of a process of accretion of foreign material and of rearranging to conform to a pattern. In the earliest version, that of Ibn Ishaq (d. 768) according to Ibn Hisam (d. 833), the cleansing ceremony takes place early in the Prophet's life, preliminary not to the Ascension proper but to

' K.g., in the Masablh al-Sunnah of al-Baghawi (d. circa 1120); I am following the translation of A. Jeffery in his Islam: Muhammad and his Religion (N.Y. 1958), p. 35ff. Cf. the version of al-Ghaiti (d. 1574), given in Jeffery's A Reader on Islam (Mouton 1962), p. 62Iff.; or the summary of it in R. Hartmann, "Die Himmelsreise Muhammeds und ihre Bedeutung in der Religion Islams," Vorträge der Bibliothek Warburg 1928-29, p. 42ff. Eliade, ch. iv. T. Arnold Painting in Islam (Oxford, 1928), ch. "Burak." The first known picture of Buräq is in a Rasid al-DIn manuscript. 4 Found in al-La'äli al-Masnü'a of al-Suyüti (d. 1505), given in Jeffery, Islam, p. 42ff. ^Eliade, Horse: p. 154 ("The courser is pre-eminently the Shamanic animal."). Scarf-, p. 136. Rainbow: p. 132ff. It should also be noted that the word rafräf (with long second vowel) is lossed in the dictionaries as "male ostrich"! cf. Eliade, p. 202; p. 484; and passim.

122

FROM

MAHMUD

K A § G A R I TO E V L i Y A

Q ELEB i

M u h a m m a d ' s entire prophetic career; 1 the Night-journey (Isra) and the Ascension ( M i ' r a j ) are clearly distinguished; 2 and Buraq is associated only with the former, while the latter takes place by means of a ladder (mi'raj), and is from Mecca, not Jerusalem. 3 It is only later, in al-Bukhari (d. 870), that the Purification immediately preceeds the Ascension, and that Buraq is found as the prophet's mount in the Mi'raj as well as the Isra, though the former is still from Mecca, not Jerusalem. It is unneccessary to trace the process through all its stages here. The accretions and reargements were evidently the result of continual impingement of Iranian (and Turkic?) concepts, but it is difficult to see the exact line of development through the maze of hadith literature. 4 E. Blochet 5 was the first to point out the strong Iranian influence upon Islamic traditions of the Mi'raj. He adduced the book of Arda viraf (7th century?) 6 as possibly the immediate source, though we would do well to think of both this book and the Islamic traditions as deriving from the same oral sources within the Iranian milieu. Blochet warned 7 that "in none of the forms of the Iranian legend do we find a winged horse, transporting, to the foot of the throne of God, the mortal called to contemplate before his death the mysteries of the other world." He did mention another Iranian legend, however, in which the hero Tahmurat mounts Ahriman, and together they travel as far as the bridge Cinvat; and noted that in two Pahlavi writings Ahriman in this circumstance is called the barak.s Blochet concluded that Buraq is the same as this barak. Blochet's etymology has not been universally accepted. Horowitz, in particular, 9 argued for the conventional Muslim interpretation, which sees buraq as the rare diminutive pattern fu'al plus the root b-r-q, and so an

^cf. A. A. Bevan, "Mohammed's Ascension to Heaven," Studien I. Wellhausen ... gewidmet (Gießen 1914), pp. 51-61; also the article "mi'rädj" in £ / ' by J. Horovitz. 2 I n Ibn Sa'd, according to Bevan, the Ascension took place eighteen months before the Hijra, the Night-journey a year before the Hijra. 3 I b n H i s ä m , Das Leben Muhammed's, ed. F. Wüstenfeld (Göttingen 1858), p. 268ff. The ladder, of course, is also a shamanic vehicle. 4 M . Asin Palacios has conveniently gathered and sifted many of the sources. He is inclined to see the Persian Maysara (d. 786) as one of the prime culprits (La Escatologia Musulmana en la Divina Comedia, 2nd ed., Madrid 1943, p. 437-8). For a thorough study of one important episode, see H. Birkeland, "The Legend of the Opening of Muhammed's Breast," Avhandlinger utgitt av det Norske Videnkaps-Akademie i Oslo II. Hist.-Filos. Kl. 1955. No. 3. ^"L'ascension au ciel du prophete Mohamond," Revue de l'Histoire des Religions 40 (1899), pp. 1-25, 203-236; see esp. p. 10 ff. °Cf. The Book of Ardä Viräf (ed. Haug and West, 1872). 7 P . 203. Q P. 204 ff. See above, n. 51. For the legends about Tahmuraf ~ Tahmorap, cf. Carnoy, op. cit., p. 301 ff. Muhammeds Himmelfahrt," Der Islam IX (1919), p. 179 ff.

B AR AQ

AND

B URÀQ

123

alternate of the attested proper name Burayq. The word would then mean "little lightning flash". To complete the picture ol' possible Arabic etymons, we might glance at some other words containing the radicals b-r-q: burqa'

veil (women; men [to cover the beard]; horses)

birqi'

heaven

abraq

wearing a mixture of white and black

faras mubarqa'

horse with white face

baruq

camel feigning pregnancy by raising her tail

birqis, abu

a bird (chaffinch)

baraqis

That some of these are less fanciful than others when thinkning of etymons for Buraq is apparent from the Liber Scalae Mahometi (13th century), in which Alborak is identified as a duck or goose. 1 E. Cerulli offers three possible explanations for this: 2 1) there is a contaminaton with the abjective abraq meaning "(animal) having a cloak of two colors, white and black"; 2) the identification of Buraq as a duck is due to a (misinterpreted?) iconographic representation (of the heavenly cock, which Muhammad sees on his journey? or of Buraq himself, often depicted with wings?); 3) there is a contamination with the word burakah, pi. burak3 meaning "duck". This last explanation appears the most likely. To return to the question: should we accept an inner Semitic etymology over an Iranian borrowing ? I think not. Horowitz himself, in fact, cited evidence which makes Blochet's etymology virtually unassailable. This evidence is in the Talmud, 4 in the context of a discussion between King Shapur I and Samuel, the leader of the academy at Nhardea (d. 254 A.D.): King Shapur said to Samuel: "You say that the Messiah [will come| on an ass. Let me send him a horse barqa that I have." He said to him: "Have you indeed a xar hazar gunahT

The last three words are Persian, meaning "an ass of a thousand colors." The form barqa, then, is simply the Persian barak plus the Aramaic definite article-a. This is confirmed by the alternate form baraq found in Rashi's ' E . Cerulli, ed., II 'Libra delta Scala' ... (Vatican 1949), p. 42-43: "masculus de anatibus vel anseribus" in the Latin; "malard des annes" in the French. 2 P. 226. rt Dozy I 76 b; the Spanish pronounciation according to Alcala, 1505. ^Sanhedrin 98 b. For the conventional understanding of this passage, cf. The Babylonan Talmud (London, Soncino, 1935), Sanhedrin, II, p. 663-4.

124

FROM

MAHMUD

K A § G A R I T O EV LI Y A Q E L E B I

quote of this Talmudic verse. 1 This last word, baraq, is used elsewhere in the Talmud to modify the Persian word for wine, may. Rashi (11th century) explained it as a Persian word meaning "excellent"; and he explained barqa as "a combed, groomed, handsome horse." Barqa meaning "horse" appears once elsewhere in the Talmud in a proverbial expression. 2 The form barqa is a likely intermediary between P. barak and Ar. buraq first of all on linguistic grounds. Barqa would be analyzed as baraq + a, and the first vovel would appear as u in Arabic due to the labial consonant. A Jewish Aramaic intermediary is also likely on semantic grounds, since as the Talmudic quote shows, the Persian barak, which already carried apocalyptic overtones in the Zoroastrian milieu, was borrowed into the Jewish milieu specifically as the name of the ass upon which the messiah will enter Jerusalem. The Jews readily indentified the ass of the Messiah (Zech. ix, 7) with the ass which Abraham rode (Gen. xxii, 3), and with the one used by Zipporah and her sons (Ex. iv. 20). 3 The earliest attestation of Buraq in Arabic seems to be as the riding beast of Ibrahim in a verse of the poet 'Ajjaj. 4 And in the first notice of Buraq in Ibn Hisam 5 we read: "This is the beast upon which the prophets before him were borne."

' Pace M. Jastrow, A Dictionary of the Targumin, etc. (N.Y. 1967), p. 136: "Deriv. fr. Pers. bärah is refuted by barqä being used as an equivalent of our w[ord|." 2 C f . J. Levy, Wörterbuch über die Talmudim und Midrashim (2nd. ed., Berlin-Wien, 1924) I, p. 270. Levy thought that barqä in the context of two medicinal rituals meant "hair". This is a false lead. Jastrow corrected the meaning to "white" in these passages. Cf. also H. L. Fleischer's " N a c h t r a g " to Levy, I, p. 284: "Das neupersische bära dessen ehemaliger palataler Schlußconsonant durch das g in bärag dargestellt sein kann, wird in der That von einem edeln Rosse gebraucht, und susyä bärag wäre demnach eine Substantiv-Apposition; aber die allgemeine adjectivische Bedeutung 'vorzüglich, vortrefflich' hat es wenigstens nicht mehr...." I would like to thank Rabbi Ben-Zion Gold for his assistance with the Talmudic sources. 3 1 Cf. Horowitz, EI , op. cit. ^Diwän al-'Ajjäj XXXV, 51 (ed. Ahlwardt, Sammlungen alter arabischer 1903). ^P. 263: wa hiya al-däbba allati tuhmalu 'alayhä al-anbiyä ' qablahu.

Dichter

II, Berlin,

B ARAQ

AND

BURÄQ

125

VIII F r o m t h e a b o v e c o n s i d e r a t i o n s , w e m a y , in c o n c l u s i o n , d r a w u p the f o l l o w i n g tentative s c h e m e : bàrak

(Middle Iranian) "a fabulous steed"

bärag -barqä (Judaeo-Persian/ Jewish Aramaic) "mount of the Messiah"

Buraq (Arabic) "mount of the prophets; Mount of Muhammad in the Isra"

"Mount of Muhammad in the Mi 'raj"

baraq

(Turkic)

"mount of the shaman"

"shamanic animal; the shaman himself; any hairy animal"

"hairy people; hairy hunting dog born of an eagle; etc."

9. MIDDLE TURKIC VULGARISMS

In an essay on "Semantic Universals," the British linguist Stephen Ullmann writes: 1 Language taboos seem to spring from three main causes. First, there are those inspired by fear ...: religious restrictions on the use of the name of God, and also superstitious avoidance of any direct reference to the dead, to the devil, and to evil spirits, and the widespread and varied taboos on animals. A second group is dictated by a sense of delicacy: when we have to talk of such unpleasant topics as illness and death, physical or mental deficiencies, and such criminal acts as cheating, stealing, or killing, we often have recourse to euphemisms.... Third, taboo bans may result from a sense of decency and propriety: references to sex, names of certain parts and functions of the body, and swear-words are particularly subject to this form of taboo. While all three types are of wide currency, none of them is an unrestricted universal since they are governed by social and cultural factors and will arise only in certain environments. The first type is bound to become rarer with the progress of civilization, though it will not disappear altogether. The second and especially the third type, on the other hand, will be encouraged, up to a point, by the development of higher moral standards and more refined forms of social behavior, though some of these more sophisticated taboos may be subsequently rejected as prudish and hypocritical: we no longer say limbs or benders instead of legs, or waist instead of body, as did some Boston ladies a hundred years ago. The growth and decay of the various forms of taboo, in relation to social and cultural development, could be systematically studied in various languages Tracing the history of such taboo terms, however — especially those of the third type, which w e are concerned with here — is by no means an easy task. For one thing, w e are unlikely to encounter them in literature. For another, lexicographers tend to be prudes; or, as Sir Gerard Clauson repeatedly states in his Etymological

Dictionary

of Pre-Thirteenth-Century

"Vulgar words like this are usually omitted from dictionaries."

Fortunately, Mahmud al-Kasgari, the author of the Diwan 3

Turk,

Turkish:

2

Lugat

at-

did not omit these words — probably because he w a s following the

model of the Arabic lexicographers, who, contrary to the above generalization,

' i n J. H. Greenberg, ed., Universals of Language (The M.I.T. Press, 2nd edition, 1966), pp. 245-6. 2 = ED (Oxford, 1972), 702 s.v. kotif, cf. 155, am; 818, sit, 396, (ubek\ 495, tila:k; etc. ^Written c. 1075; page numbers according to the facsimile edition, Ankara, 1941. Translations of Koranic quotes, infra, are those of A. J. Arberry, The Koran Interpreted, Toronto, 1969.

128

FROM

MAHMUD

KA § G A R t TO EVLÌYA £ ELEB Ì

were never prudish. But Kàsgari was scrupulous. Thus in the article in the Dlwan

s.v. sik (168), translated "the male member" ('ard ar-rajul),

w e read

the following: Mahmud states: It shows good manners and respect for the book of God, for a Reader among crude Turks and their women to mufle his reading of the following verses: wa-àtat kulla wàhidatin minhunna sikklnan (Koran 12:31, "then she gave to each one of them a knife"); and: ma yaftahi llahu li-n-nàsi min rahmatin fa-là mumsika lahà wa-ma yumsik fa-la mursila lahu min ba'dihi (Koran 35:2, "Whatsoever mercy God opens to men, none can withold and whatsoever He withholds, none can loose after Him"). He should muffle his reading of such verses among them, for they do not understand the meaning, but consider that the sounds of the words mean what they understand by them in their own language; so they commit a sin by laughing at them. Likewise, he should muffle his reading of the following verse: in hàdà ilia xtilàq (Koran 38:7, "this is surely an invention"), because in their language tilàq is "vulva" (farj almar'a). Also, he should muffle among the Oguz all phrases containing the interrogative am — as for example: a'antum anzaltumuhum min almuzni am nahnu l-munzilun (Koran 56:69, "Did you send it down from the clouds or did We send it?") — since am is "vulva" in Oguz dialect. But there is no objection with regard to one who understands the meaning. This discourse on Koran recitation is isolated in the Dlwan.

Elsewhere

w e do find an example of genuine taboo. The verb qil- (277) means "to do something," as in: ar is qildi, "The man did the job"; dr yukunc

qildi,

"The

man performed the prayer." But the phrase ar qizig qildi m e a n s "The man copulated with the girl" (cf. Eng. "to make"). On this Kàsgari remarks: The word is used as an indirect expression for copulation. For this reason the Oguz avoid it. Instead of saying qildi for "He did something" they say étti which means "He did it well." For example, they say: àr yukunc ètti to mean "He performed (aslaha) the prayer", the Turks say qildi. It is interesting that in modern Turkish the verb kil- is largely frozen in the phrase namaz

kilmak,

m e a n i n g "to perform the prayer." Perhaps the

substitution of ét- for qil- among the Oguz began in this religious context, as Kàsgari himself intimates. The technical term that Kàsgari uses for the semantic transfer involved in the verb qil- is kinaya,

which I have translated as "indirect expression." In

the remainder of this paper I wish to record all the words of this sort relating to parts of the body. The only example in which this Arabic technical term is used again is the word qudruq

(237-8), meaning "tail". Kàsgari says that this

means "asshole" by "indirect expression" (yukna bihi 'an al-faqha),

and he

MIDDLE TURKIC

VULGARISMS

129

gives the phrase qudruqi dtgan kisi which literally means "a person whose tail sings," and which Kasgari translates "a person who farts a lot" (nabbaj al-faqha). The remaining examples relate to the sex organs. Of the four words attested for the female organ, two were already mentioned above. Tilaq (207, mata' al-mar'a, Cigil dialect) is obviously the diminutive of til meaning "tongue". Am (31, jahaz al-mar'a, Oguz and Qifcaq dialect) is probably cognate with Mongolian ama meaning "mouth" — but this point is irrelevant here, since we are concerned with synchronic, not diachronic, derivations. The etymology of sdkruk (correct vocalization? 240, mata' al-mar'a) is obscure. Finally bitrik, meaning "clitoris", (239, bazr al-mar'a) is the same word as bitrik in Argu dialect, which means "pistachio". The oral provenance of this group of words is apparent. The common term for the male organ, already mentioned above, is sik (168, 'ard ar-rajul). This is semantically opaque, except that it is homonymous with the verbal root sik-, which is the common word meaning "to c o p u l a t e " . 1 yigac (448, dakar ar-rajul), on the other hand, is semantically transparent: it is the ordinary word for "piece of wood" or "tree". Similarly qamci, meaning "whip", is used for the "rod" of a horse, bull, or camel (210, qadlb). Finally there are two words specifically relating to the organ of a small boy, and both of them appear to derive from words that are names for birds: ctipak (193,faysala as-sabi) — cipak qarguy is a "sparrowhawk"; and cakik (409, 'ard as-sabi fi hal sigarihi) — otherwise a "lark". 2 Except for the remarks in connection with Koran recitation cited above, there is no outright indication that these words are vulgar. One other word that is labeled "vulgar" belongs in the list, but is treated separately here since it has not been recognized hitherto. I refer to the word salindi, literally, "the hanging one". This belongs to the notorious category of "ghost words," and this in two senses. First it is misspelled in the ms. (226) with U instead of A. This was recognized by Brockelmann 3 and by Clauson. 4 Thus solundy (Brockelmann) and sulundi (Atalay) 5 are ghost words. Second — and this is the point that has not been recognized — the Arabic definition given in the text, 'arf ar-rajul, is also a "ghost". 'Arf means "smell" (thus Brockelmann: "Geruch [?]"), which can hardly be thought of as "hanging" from a man — for Kasgari states quite clearly after giving the definition: 1 ( 2 7 5 J a m a ' a , and derived forms). Other words used to express this are has - (269, tasannama), lit. "to press"; qil- (277, jäma'a), lit. "to make" (see supra); and qawus- (320, nakaha), lit. "to join". See ED, 415, and read there gekik instead of $ekig. 3 C . Brockelmann, Mitteltürkischer Wortschatz nach Mahmud al-Käsgaris Diwan Lugät at-Turk (Budapest, 1928), p. 183. 4 E D , 828. ^B. Atalay, Divanü Lügat-it-türk Tercümesi Vol. I (Ankara, 1939), p. 449.

130

FROM

MAHMUD

KA § G A R i TO

EVLiYA

gELEBI

T h i s c o m e s from the e x p r e s s i o n salindi nan, m e a n i n g "The thing hung d o w n ( t a d a l l a ) , " since it is a l w a y s hanging. 1 The word is vulgar ( l u g a gayr faslha).

Atalay, apparently reading '¿/r/'instead of 'arf, translated it "the hair which a m a n t h r o w s in b a c k o f h i m " ( E r k e g i n arkaya

dogru

saliverdlgi

sag);

and

u

Clauson went this way too, reading the word 'urf( a man's plait of hair"). But this too is unsatisfactory. 'Urf can mean the "mane" of a lion or the "crest" of a cock — i.e., something that sticks up (from the root 'arafa, "to know, recognize"); there is no reason to think it can mean the "lock" of a man's hair, which hangs down. Also, why should Kasgarl label this as "a vulgar word" — for this is what luga gayr faslha means, and not "incorrect" as Clauson took it to mean. Now what is always hanging down from a man, and would be referred to by a vulgar word? Only one thing. We have merely to amend 'arf to 'awf "pudendum" — cf. ms. 451.10, and see Dozy II, 190. This solution to the problem not only disposses of a "ghost" definition; it also recovers for us a Middle Turkic vulgarism that we did not know before, and one that compares nicely with the English use of the word "hung" to describe the male organ.

the entry salin- (344), where the example sentence is: "The pearl hung from the ear."

10. YUSUF KHASS HAJIB WISDOM OF ROYAL GLORY ( K U T A D G U BILIG) A TURKO-ISLAMIC MIRROR FOR PRINCES

Introduction Characterization

of the Work

Kutadgu Bilig is the oldest monument of Islamic Turkish literature. A long didactic poem in the mirror-for-princes tradition, it consists mainly of dialogues set within a frame story. The language of the work is technically referred to as Karakhanid, or Middle Turkish. It is substantially the same language as that of the Turkish inscriptions on stone, mainly found in the Orkhon region of Mongolia and dating from the eighth century; of the huge translation literature in Uighur Turkish, mainly of Budhist content; and of the later efflorescence of eastern Turkish Islamic literature known as Chaghatay. As the first major literary work in the 900-year-long history of Turkish Islam, Kutadgu Bilig has a significance analogous to that of the Divine Comedy or, closer to our theme, to that of the Shah-nameh. In the latter work, completed in 1010, Firdawsi had managed to fit the Iranian tradition into a New Persian dress that suited Islamic sensibilities. Yusuf, the author of Kutadgu Bilig, was undoubtedly influenced by Firdawsi in his choice of verse pattern (the Arabo-Persian mutaqarib meter — see below, under "Text and Translation") and in his tendency to the epic style. More than this, Yusuf attempted, with some measure of success, to do for the Turkish tradition what Firdawsi had done for the Iranian: that is, to establish its place alongside the Arab tradition within the larger frame of Islamic culture. The method and result were different, however. Firdawsi translated the Iranian epic into an Islamic Persian form. Yusuf took as his starting point, not the Turkish traditions of legend and saga (which also existed), but rather the Irano-Islamic ideals of statecraft which he found in Arabic and Persian literature. His aim was to make a new home for these ideals as part of an Inner Asian Turkish literary heritage. Stated otherwise, he aimed to show that the Turkish tradition of royalty and wisdom were comparable or superior to their Arab and Iranian counterparts, and were equally compatible with Islam. One way he accomplished this was to cite, as authorities for the wisdom sayings scattered

132

FROM

MAHMUD

K A § G A R i TO E V L I Y A

£ELEBi

through the text, various Turkish princes (Khan of the Turks, Beg of Otiiken, Beg of the Ili Valley, etc.) and various Turkish poets (all unnamed). The Turkish word for such wisdom sayings was bilig; and the other word in the title embodies kut, the word for "fortune" or "the charisma of rule." Thus the title means "The Wisdom that Conduces to Royal Glory or Fortune." Yusuf's intent was to provide his patron a mirror of court life, in order to guide his conduct in an age full of opportunity and uncertainty, and thus to ensure the success of his rule. In doing this, Yusuf provides for us a mirror of the political and cultural life of the Central Asian Turks at the outset of their Islamic career.

The Karakhanids

and Turkish

Culture

The Karakhanid Turks converted to Islam in the Middle of the tenth century. They ruled in Central Asia and, after 999, in Transoxania. Their two chief cities in Central Asia were Kashghar (in the present Sinkiang province of China) and Balasaghun (now in the Kirgiz SSR, precise location uncertain). In 1069 or 1070, Yusuf, originally from Balasaghun, presented Kutadgu Bilig to the reigning prince of Kashghar, Tavghach Bughra Khan. 1 As a reward he was made Privy Chamberlain (Khass Hajib), and he is also known by this title. By the eleventh century, three ethnically Turkish dynasties — the Karakhanids, the Ghaznavids, and the Seljuks — had assumed the dominant position in the political life of Islam, having wrested authority away from the Iranian dynasties of the Samanids and the Buyids. Politically, it was the Seljuks who were eventually to triumph in the central Islamic lands, leaving the Ghaznavids their empire in Afghanistan and India, and reducing the Karakhanids to a weak and divided Central Asian state, to be engulfed in the twelfth century by the heathen Karakhitay. Culturally, it was the Karakhanids who played a role for the Muslim Turks similar to that of the Samanids for the Muslim Iranians a century before. The Ghaznavids and the Seldjuks, on the other hand, both patronized Persian and not Turkish culture. It was under the Ghaznavids, for example, that Firdavvsi completed the Shah-nameh originally begun under the Samanids.

' w . Barthold, "The Bughra Khan mentioned in the Qutadqu Bilik," BSOS 3 (1923-25), 151-58.

A TURKO-ISLAMIC

M I R R O R FOR

PRINCES

133

The Karakhanids were a confederation of tribes, including especially the Karluk and the Chigil, led by their Khaqans, who preserved the Inner Asian traditions of tribal aristocracy and nobility of rule. 1 In this they differed from the Seljuks, who began their career as a band of freebooters, and the Ghaznavids, who started out as slaves. The Karakhanids traced their ancestry to the legendary hero Alp Er Tonga, whom they identified with the archenemy of Iran, Afrasiyab (see KB 276-80). They cultivated the Turkish language, and also continued to employ what they called the "Turkish" script — what we now call the Uighur script. Mahmud al-Kashghari, the great lexicographer of the Turkish dialects, records many of the distinctive tribal, folk, and court traditions of the Karakhanids in his work, which is contemporary with the Kutadgu Bilig? Taken together, the works of Mahmud al-Kashghari and of Yusuf of Balasaghun represent the establishment of a Turko-Islamic literary culture, the one in the areas of linguistic scholarship and the recording of national lore, the other in the areas of royal wisdom and the synthesis of Inner Asian tradition with Islamic religious values. But, perhaps because of the declining political fortunes and waning prestige of the Karakhanids after the eleventh century, these works did not serve as the basis for the elaboration of that culture. No Turkish Firdawsi came along to celebrate the pre-Islamic exploits of Alp Er Tonga. It is true, judging by the three extant ms. copies, that Kutadgu Bilig did enjoy a certain vogue as late as the Timurid period. But it never served as the basis for an elaboration, or even an imitation; in contrast, say, to the Persian Qabus-nameh, which was translated into Ottoman Turkish no less than five times in the fourteenth and fifteenth centuries. 3 When the Ottomans and the Timurids did create an imposing Turko-Islamic culture, it was on different foundations than those provided by these Karakhanid monuments.

The Major Structure of the Work Yusuf informs us at the beginning of the work (lines 353-58) that it is based on four abstract principles, each of which is represented by one of the four major characters. These are as follows: 1 9

1 ElA, "Ilek-khä ns," (Bosworth); O. Pritsak, "Von den Karluk zu den Karachaniden," ZDMG 101 (1951), 270-300. Compendium of the Turkic Dialects (Diwän Lughat al-Turk). An English edition of Käshghari's work is in the course of publication (Part I: Duxbury, Mass., 1982). -*See E. Birnbaum, "A Lifemanship Manual: The Earliest Turkish Version of the KäbüsnämeT' Journal of Turkish Studies 1 (1977), 3f.; The Book of Advice, ed. E. Birnbaum (Duxbury, Mass., 1981).

134

FROM

MAHMUD

Name Rising Sun (Tk. Kiin Togdi, lit. "the sun has risen") Full M o o n (Tk. A y Toldi, lit. "the m o o n is full") Highly Praised (Tk. Ogdiilmi§, lit. "praised") Wide A w a k e (Tk. Odgurrm§, lit. "awakened")

KAÇGARÎ

TO

EVLÌYA

ÇELEBI

Occupation king

Justice

vizier

Fortune

sage

Intellect (or W i s d o m )

ascetic

Man's Last End

Principle

The first half of the work presents the relations among the first three of these characters, and deals with the traditional mirror-for-princes themes, mainly of Iranian inspiration. The second half centers on the opposing character of Wide Awake and introduces religious themes relating to Sufism, or Islamic mysticism. The following sections of this Introduction provide background and source analysis to the more important ideas in Kutadgu Bilig. These are followed by a short literary analysis and interpretation, where the reader will find a more extended outline of the work's contents.

Mirrors for Princes: A Survey up to the Eleventh

Century

Given the delicate political situation of the Karakhanids in the second half of the eleventh century, it is appropriate that the major literary monument of the day should be a mirror for princes. "Mirror for princes" is the name given in Western European tradition to a kind of literature that is in fact a universal genre. It can be defined most broadly as "advice on how to rule." It is to be distinguished from other genres which, however, it overlaps: on the one hand, instruction in etiquette; on the other hand, political philosophy. It belongs to the broad category of oral and written "Wisdom"; and Kutadgu Bilig is, as indicated above, an attempt to combine two different wisdom traditions: the Turkish Inner Asian, and the Irano-Islamic. 1 W e find mirrors for princes wherever there is a tradition of autocratic state organization. There are examples from the ancient Near East, f r o m China, India, the Islamic Near East, from the ancient Greek tyrannies, and f r o m medieval Europe. The prevalence of the genre is due to the function of such a literature in an autocratic society: it embodies a conservative tradition; 'R. Dankoff, "Inner Asian Wisdom Traditions in the Pre-Mongol Period," JAOS 101/1 (1981), 87-95.

A TURKO

ISLAMIC

MIRROR

FOR

PRINCHS

135

it ensures smooth transition of rulership; it is especially useful during times of crisis in the state — for example, a change of dynasty, an outside threat, or inner disintegration. A remarkable thing about mirrors for princes is their uniformity, irrespective of time and place. The common form is one of direct address, either by a vizier or other adviser to his ruler, or by a ruler himself to his heir, and the major tendency in this literature is always this-vvorldly. Other-worldly concerns take second place to the temporal goals of stability, prosperity, and prestige. The theme of realpolitik is present from the very beginning, well before Machiavelli brought it to its logical conclusion. As a typical example, consider the following: Do not be evil, for patience is good; make your lasting monument in the love of you... Make your magnates great, that they may execute your laws.... Great is the great one whose great ones are great; valiant is a king who owns an entourage; and august is he who is rich in magnates. Speak truth in your house so that the magnates who are on earth may respect you, for a sovereign's renown lies in straight-forwardness; it is the front room of a house that inspires the back room with respect. Do justice, that you may live long upon earth... Beware of punishing wrongfully; do not kill, for it will not profit you, but punish with beatings and with imprisonment, for thus this land will be set in order...

This is actually one of the earliest known examples of the genre; it is the instruction of an Egyptian pharaoh to his son, and dates from the twentysecond century BC. 1 In the ancient Near East, advice to rulers covered many topics: how to organize the court; the relationship with other courts and the treatment of ambassadors; the dispensing of justice; the need for wise and loyal counselors; controlling the military; relations with the various classes and ranks in society; cultivating a reputation for generosity; good speech and table manners; advice on choosing a wife and organizing the household. This literature always keeps in view an awareness of the inconstancy of fortune; and therefore it includes advice on how to act in a subordinate rank, and on the various ways of making a living. All these themes are common to later mirror, both in the near and farther East, and in the West. Although the practical, temporal goals are in the forefront, spiritual, other-worldly motives are also present. For one of the major thrusts of this literature is to promote what may be called the Royal Ideology; and this ideology has a religious foundation. A Babylonian proverb states; "Man is the ^W. K. Simpson, ed., The Literature "The Teaching for Merikare."

of Ancient Egypt (Yale University Press, 1973), 182-83,

136

FROM

MAHMUD

K A § G A R I T O EV LI Y A

gELEBI

shadow of a god, a slave is the shadow of a m a n ; but the king is like the very image of a god." 1 T h e imagery of sun and shadow is universal. T h e sovereign is both the s h a d o w of G o d on earth and the s h a d e (or protection — cf. KB 4 5 1 ) of his subjects. A n d he is the sun as well, d i s p e n s i n g light, w h i c h means that he creates order f r o m chaos and establishes rule over the earth by m e a n s of j u s t i c e . 2 In Babylonia, the gods c h o s e H a m m u r a b i , as he tells us in his Code, " t o m a k e legislation appear in the land, to destroy the evil and the wicked, so that the strong should not harm the w e a k , so that I should appear like the sun to the black-headed people and m a k e light the land, and create well-being f o r m a n k i n d . " 3 In E g y p t too, kingly rule is likened to the sunrise; and K i n g D a v i d ' s last words, as recorded in the book of Samuel (II, 23:3-4) are: " H e that ruleth over men m u s t be just, ruling in the f e a r of G o d ; and he shall be as the light of morning, when the sun riseth, even a m o r n i n g without clouds." 4 T h e Old T e s t a m e n t also provides the locus classicus f o r the opposite tendency: a demotion of the k i n g ' s position, since sovereignty belongs to God alone. (A c o n c e p t i o n w h i c h w a s later influential in Islam, as well as the medieval church.) Samuel (I, 8) warns the elders of Israel w h o c l a m o r for a king, saying: a king will take your sons for his armies and your daughters for his servants; he will appropriate your fields and crops, tax you to the hilt, and impose forced labor on you. This is a type of mirror for princes in reverse: it is what a king should not do! T h e classical Near Eastern ideology of kingship 5 is f o u n d with the first world empire, the A c h a e m e n i d or Persian empire, established by Cyrus in the sixth century B C . (There is a reflecton of this in the bible also; w e find in the b o o k of E z r a | 1 : 2 ] the text of t h e d e c r e e a l l o w i n g the J e w s to return, beginning: " T h u s says Cyrus, king of Persia: All the k i n g d o m s of earth has the Lord God of heaven given unto me.") A c c o r d i n g to the classical ideology, the k i n g ' s person is holy and inviolable. His divine descent is reflected in his birth legend and in the i m a g e r y surrounding him: he is " b r o t h e r of Sun and M o o n , his n a m e is a m o n g the stars." His d i v i n e status on earth is reflected in court ceremonial and the insignia of royalty. T h e king is the c o s m i c ruler,

'.I. B. Pritchard, ed., Ancient Near Eastern Texts Relating to the Old Testament (Princeton, 1955), 426; dated to the reign of Esarhaddon (680-669 B.C.). 2 O n king = sun, cf. A. M. Hocart, Kings and Councillors (Cairo, 1926; new ed. Chicago, 1970), 8 5 , 9 6 , 139f„ 151,207,279. 3 1.32-48; cited in H. Frankfort, Kingship and the Gods (Chicago, 1948), 308. ^Frankfort, 57, 150-51, 344. Widengren, "The Sacral Kingship of Iran," in The Sacral Kingship, Studies in the History of Religions . . . Numen, Suppl. 4 (Leiden, 1959), 242-57. Cf. R. Frye, The Heritage of Persia (New York: Mentor, 1966), 120; W. Knauth, Das altiranische Fürstenideal (Wiesbaden, 1975), passim.

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

137

lord of the seven climes, and sovereign of the age. He is the peaceful Sunking or the warlike Moon-king. He is the protector of peace and prosperity, and the bringer of happiness and wealth. His element is Fire; and his symbol is the fire-nimbus or aura, the aureole of Glory, which is identified with his Luck or Charisma, the Royal Fortune. As long as the aura is present, the realm prospers; if it is absent, the realm suffers. 1 Some of the ancient Greeks were favorably impressed by the Persian imperial system, most notably Xenophon, in the fourth century BC, who wrote an entire book in which he tried to explain to the Greeks the success of the Persian polity under Cyrus (Cyropaedia, "The Education of Cyrus"). Earlier, there is an unusual example of the mirror genre in Herodotus (I, 96 ff.) in connection with the description of the rise of the Medes to world power as prelude to the Persians under Cyrus. Herodotus explains the king's power in rational and sociological terms. The Median king Deioces first got a reputation for justice among his own people; then he maneuvered to get himself appointed king. He proceeded to build his palace and establish his private guard; instituted court ceremonial; and administered justice from his new capital. He also sent out spies and secured a base of power for military operations, by which he expanded his kingdom through systematic conquest. As for the native Greek school, the most influential political thinkers were Isocrates, also fourth-century, who in a treatise addressed to king Nicocles of Cyprus urged the tyrant to be just on rational grounds; and of course Plato, who developed the ideal conception of justice as social harmony. Alexander the Great tried to combine the Persian and the Greek traditions, and himself became a model for later rulers. Finally in Plutarch (first century AD) we find, in his "Discourse to an Unlearned Prince," what is perhaps the earliest reflection in Western literature of the classical Near Eastern conception: the good prince is the image of God; just as God placed the sun and the moon in the skies as symbols of His power and glory, so the prince, the vicar of God, is resplendent on earth. 2 Let us summarize the three major strands in ancient political thought that are taken up in later mirrors; and, in particular, how the king in each of these traditions relates to Fortune and to Justice.

Greek aura corresponds to Akkadian melammu, Old Persian xvarena (new Persian farr), Aramaic gadda de malka (see A. L. Oppenheim, Ancient Mesopotamia, Chicago, 1964, pp. 98, 181ff.; Frye, 64). Kut is the turkish equivalent: see A. Bombaci, "Qutlugh Bolzun!" UAJ 3.6 (1965), 284-91; 38 (1966), 13-43. o Paraphrase as cited in L. K. Born, The Education of a Christian Prince . . . Erasmus (New York, 1936), 72.

138

FROM

MAHMUD

K A§ G A R i TO E V L l Y A

gELEBI

First, in Persia, where the Near Eastern royal ideology crystallizes, the king is the very e m b o d i m e n t of both Justice and Fortune. H e dispenses justice as the sun dispenses light; and the f o r t u n e of his realm is symbolized by his aureole of glory. S e c o n d , in Greece, Justice is conceived of as social h a r m o n y , and is maintained by a j u s t king; while the king, like everyone else, is subject to the c a p r i c e s of Fortune, w h i c h is a f o r c e outside his control. T y p i c a l l y , these c o n c e p t i o n s are e m b o d i e d in allegorical i m a g e s , w h i c h are c o m m o n in Western literature and art. Justice is represented by the goddess Dike (or else T h e m i s ) w h o is shown holding a sword and sitting on a three-legged stool; while Fortune is represented by the goddess T y x e , w h o is represented as blind and sitting on a ball. 1 Finally, there is the theocratic tradition, represented by Judaism and its m o n o t h e i s t i c d e s c e n d a n t s , Christianity and I s l a m . T h e king is s u b j e c t to G o d ' s will, which in e f f e c t replaces F o r t u n e ; 2 and the king is j u s t if he rules according to G o d ' s will. All of these strands c o m e together in the Sasanian state and, f o l l o w i n g it, in the Islamic imperial tradition. T h e d o m i n a n t t h e m e in medieval Near E a s t e r n politics, both p r e - I s l a m i c and I s l a m i c , is the e f f o r t to c o m b i n e religious orthodoxy with social unity under a rightful dynasty. T h e Sasanians, ruling f r o m the third to the seventh centuries, saw themselves as heirs to the Achaemenids,

w h i l e at t h e

same

time

they

attempted

to

Z o r o a s t r i a n i s m as the state r e l i g i o n . T h e i r i d e o l o g y r e c e i v e d

establish classical

e x p r e s s i o n in a d o c u m e n t , p r e s e r v e d only in a m u c h later I s l a m i c w o r k , k n o w n as the " L e t t e r of T a n s a r . " 3 T a n s a r w a s chief priest or r e l i g i o u s counselor to Ardashir, the f o u n d e r of the Sasanian dynasty. His f a m o u s letter is addressed to a rebellious king and consists of a justification of A r d a s h i r ' s

See literature cited in A. Bombaci, "Kutadgu Bilig Hakkinda Bazi Mulahazalar," Fuad Koprulii Armagani (Istanbul, 1953), 72. Cf. Shakespeare's jocular description in Henry V, 3.6.29ff.: "Fortune is painted blind, with a muffler afore her eyes, to signify to you that Fortune is blind; and she is painted also with a wheel, to signify to you, which is the moral of it, that she is turning, and inconstant, and mutability, and variation: and her foot, look you, is fixed upon a spherical stone, which rolls, and rolls, and rolls: in good truth, the poet makes a most excellent description of it: fortune is an excellent moral." The additional element of the wheel, the commonest symbol of Fortune in the West, is apparently an invention of Boethius (6th cent.): it is not found as such in the Near East, though it is similar to the turning of the sphere of the universe, representing Fortune or Fate in Islam (cf. Ar. dawr, dahr, dawla, falak, dunyd; Pers. charx; etc.) On this point, see S. D. Goitein, Studies in Islamic History and Institutions (Leiden: Brill, 1968), 200 ff. A typical example is a tenth-century Fatimid mirror embodied in the "Pillars of Islam" by Qadi al-Nu'man (G. Salinger in Muslim World 46 [1956], 26-27): "or do you hope for some change of fortune which would remove you from His power to that of someone else? ... There is nothing that more easily causes the revocation of G o d ' s favors and provokes more quickly His opposition than obstinacy in oppression." ^M. Boyce, The Letter of Tansar, Rome Oriental Series, 38 (Rome, 1968).

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

139

rule: The king of kings has restored society to its just order; he has reestablished the tradition of the ancients, which is Equity or Justice, as opposed to the tradition of the moderns, which is Violence or Injustice; his counselors and informers are trustworthy, obedient, learned, and pious; he upholds right religion. For all these reasons, the princes of the realm should subordinate themselves to him; for "Church and State were born of one womb, joined together and never to be sundered." The most famous Sasanian ruler was Khusrau I Anushirvan, known as "the Just," who ruled in the sixth century (see KB 290-91 and note). For Muslims Khusrau became the model of a just ruler; and his vizier, known as Bozorgmihr, the model of a wise vizier. Bozorgmihr figures prominently in a work known as the Pandnamag ("Book of Counsel"), which is known from a ninth century Pahlavi version. 1 It was also translated into Arabic (included in the Javidan Khirad)2 and was versified by Muslim authors, notably Firdawsi in the Shah-nameh. The Pandnamag consists of moral and religious maxims, but in the form of a dialogue: the king proposes questions to Bozorgmihr, who responds at length. For example: Khusrau asked: " . . . does one become powerful through one's own efforts or by the effect of Fortune which delivers to kings their crown and throne?" He answered: "Fortune and innate ability form a pair; they act conjointly, like the body and soul. The body is visible and the soul hidden. The body is only a frame in which is based a man's actions, if Fortune which watches over him is on the rise; but effort alone will not make one powerful if Fortune does not help."^

It was shown by Noeldeke that Bozorgmihr is a legendary figure, modeled after the Babylonian sage Ahiqar, whose romance was immensely popular throughout the Near East; and it was Christensen's opinion that the name Bozorgmihr is a corruption of Burzoe, the historical vizier of Khusrau the Just. 4 This Burzoe is famous for his introduction into Persia of the Indian royal tradition, as represented by two items: the game of chess; and the book called Kallla wa Dimna, a translation of the Indian Panchatantra. This book is a popular reflection of the Indian tradition of practical politics, 5 and consists of dialogues among various animals set within a frame story. ' a . Christensen, "La légende du sage Buzurjmihr," Acta Orientalia ... Norvegica 8 (1930), 9798. ^Miskawaih (d. 1030), al-Hikma al-Khàlida [Jâvidàn Khirad], ed. 'A. Badawi (Cairo, 1952). ^Firdawsi, Shah-nameh, ed. Mohl, 4:364ff., according to Christensen, 120-21; Javidan Khirad, 33. ^Christensen, 104- f. 5 T h e most notable work in this tradition is the Arthashastra of Kautilya, written for the founder of the Maurya dynasty in 321 B.C. The Mauryas were the first centralized political and military power in India; and their power represents the stabilizing of society after the chaos resulting from Alexander's invasions and the establishment of Hellenic rule in the northwest. The Arthashastra, in turn, displays Hellenic influence, and also Persian influence, especially in the exaltation of royal power and in the elaboration of a complex bureaucracy.

140

FROM M A H M U D KA § G A R Ì TO EVLÌYA

gELEBt

In the middle of the eighth century Ibn al-Muqaffa', the famous vizier to the Abbasid caliph al-Mansur, translated Kallla wa Dimna as well as other Middle Persian works into Arabic. Ibn al-Muqaffa' also wrote original tracts of Adab (or "etiquette"), directed toward the bureaucratic class of the Abbasid state, and reflecting the Sasanian royal ideology. 1 Further Sasanian material was incorporated into such ninth-century Adab works as Ibn Qutayba's Kitab al-Sultan ("Book of Government," the first part of his 'Uyun al-Akhbar);2 3 and the Kitab al-Taj ("Book of the Crown"), attributed to al-Jahiz, and into such tenth-century translations of Iranian Andarz as the Javidan Khirad.4 At the same time, princes or high officials began to write epistles (risala) or testaments (wasiyya), recording their political experience and counsel. 5 In the eleventh century the mirror for princes emerged as an independent literary genre in the Muslim world, first in Arabic, 6 then in Persian. 7 Bilig fits into this movement. It draws extensively from this Kutadgu Perso-Islamic tradition, though there is an attempt, as indicated above, to harmonize this with a native Turkish one. For example, in listing the traits of various animals that should characterize the army commander (lines 2309-16), Yusuf perhaps draws on a living Turkish oral tradition; but he certainly draws as well on this tradition as it had become incorporated into Arabic Adab, as well as on an analogous tradition drawn f r o m Iranian Andarz. 8 Judging from the large number of Persian caiques in the language of Kutadgu Bilig (discussed below), it is probable that its immediate model was a Persian mirror for princes. If so, no trace of such a model has come down to us.

In a celebrated treatise, Ibn al-Muqaffa' urged the caliph to establish a uniform religious code in his realm and a uniform system of education for his army; this program was not really carried out until the eleventh century when the Seljuks attempted to force the sway of their dynasty over the entire Muslim world under the banner of Sunni orthodoxy. See Goitein, Studies, 14967. Translation: Ch. Pellat, "Conseilleur" du calife (Paris: Maisonneuve, 1976). ^Translation: J. Horovitz, Islamic Culture 4 (1930), 171-98, 33§-62, 487-530; 5 (1931), 1-27, 194-224. ^Translation: Ch. Pellat, Le livre de la couronne (Paris, 1944). ^ S e e note 18 above; also W. B. Henning, "Eine arabische Version mittelpersischer Weisheitsschriften," ZDMG 106 (1956), 73-77. Trans. M. S. Khan, "The Jawidan Khirad of Miskawaih," Islamic Culture 35 (1961), 238-43. ^Examples of this genre: C. E. Bosworth, "An Early Arabic Mirror for Princes: Tähir Dhü 1Yaminain's Epistle to his Son 'Abdallah (206/821)," JNES 29 (1970), 25-41; M. Nazim, "The Pand-Nämah of Subuktigìn," JRAS 1933, 605-28. 6 A1-Mäwardi, Nasihat al-Mulük, written 450/1085; unpublished (see GAL 1:483). Richter, Studien zur Geschichte der älteren arabischen Fürstenspiegel (Leipzig, 1932; repr. New York, 1968); A. H. Dawood, "A Comparative Study of Arabic and Persian Mirrors from the Second to the Sixth Century A.H." (unpublished Ph.D. thesis, London, 1965); E. I. J. Rosenthal, Political Thought in Medieval Islam (Cambridge, 1968), 62-83; H. Busse, "Fürstenspiegel und Fürstenethik im Islam." Bustan 9/1 (1968), 12-19; A. K. S. Lambton, "Islamic Mirrors for Princes," La Persia nel medioevo (Rome, 1971), 419-42 [Accademia Nazionale dei Lincei, quaderno 160]. 8 S e e R. Dankoff, "Animal Traits in the Army Commander," IIS 1 (1977), 95-112.

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

141

The earliest full-blown Persian mirror is the Qabus-nameh, written in 1082 by the prince of the Ziyarid dynasty of northern Iran, shortly before the Ziyarid realm was swallowed up by the Seljuks. In it the prince counsels his son on court etiquette; on such matters as chess, love, polo, purchasing slaves, buying horses, and rearing children; also on how to be a merchant, a physician, an astrologer, a poet, a courtier, a king, a farmer, and a sufi. It is largely secular in its orientation. 1 The second of the classical Persian mirrors is the Siyasat-nameh or "Book of Government", written about 1090 by Nizam al-Mulk, the famous vizier to the Seljuk sultans. It is in the nature of an administrative handbook, containing specific proposals on how the realm should be governed. The author's concern is to maintain the Seljuk empire, especially against internal dissension and weakness; and to promote orthodoxy and stability. 2 Finally to this group belongs the Nasihat al-Muluk ("Book of Counsel for Kings") by the great al-Ghazali, written shortly before his death in 1111. Al-Ghazali's concern is more theoretical: to establish the ethical and religious basis of the sultanate. The main theme is that just sovereignty and right religion are twins, born of the same womb, and cannot be separated. In confirmation, al-Ghazali quotes a panoply of authorities which is strikingly similar to those cited by the realpolitiker Nizam al-Mulk. In addition to the prophets and the Muslim rulers of old, both authors appeal to the pre-Islamic kings and sages of Greece and Iran, Alexander, Khusrau the Just, and others. 3 Kutadgu Bilig, although it predates these three Persian mirrors, and although it is arguably superior to them in its breadth of vision, its originality of form, and its literary quality, was practically forgotten, while the Persian mirrors continued to be read and to be imitated. While the late eleventh century was the highpoint in Islamic mirror writing, the genre remained popular, both in Arabic and in Persian, and later on in Turkish as well. 4

^Translation: R. Levy, A Mirror for Princes (New York, 1951). Translation: H. Darke, The Book of government (London, 1960; 2d ed. 1978). Translation: F. R. Bagley, Ghazah's Book of Counsel for Kings (Oxford, 1964). [See now Patricia Crone, "Did al-Ghazali write a Mirror for Princes? On the Authorship of Nasihat alMuluk" Jerusalem Studies in Arabic and Islam 10 (1987), 167-92; reprint in From Kavad to alGhazali . . . (Ashgate, 2005), # XII. Crone demonstrates that a l - G h a z a l i wrote only the first half of this book, not the second half which contains the mirror-for-princes themes.] 4 S e e p. 140, n. 7; also A. K. S. Lambton, "Justice in the Medieval Persian Theory of Kingship," Studia Islamica 17 (1962), 91-119. Two Ottoman examples: A. Tietze, Mustafa Ali's Counsel for Sultans of 1581, part 1 (Vienna, 1979); W. L. Wright, Ottoman Statecraft: The Book of Counsel for Vezirs and Governors [San Mehmed Pasha, Nesaih iil-viizera ve'l-umera, c. 1720] (Princeton, 1935).

142

F R O M M A H M U D K A § G A R I TO E V L i Y A Q ELEB 1 Appeals

to

Authority

As we just noted, the appeal to authority is characteristic of Islamic mirrors for princes. In the case of Kutadgu Bilig these authorities are almost invariably drawn from the Turkish Inner Asian background. We find the Khaqan (line 3126), the Khan of the Turks (3817), the Buyruq of the Turks (1163), the Buyruq of the counselor (? Oge Buyruqi; 2941), the Beg of Otiiken (1962, 2682), the Beg of Yaghma (1758, 4947), the Beg of the Three Camps (1594), the Khan of the Three Camps (2966, 3815, 5569), the Beg of the Ili (1779), the Chief of the Ili (841, 1629, 2319), the Irkin of the Ili (4752), the Sir Tengi of the Ili (2696), the Yavghu of Boke (? Boke Yavghusi; 5043, 5523), the Chigil (3491), the Kok Ayuq (2644), and Tonga Alp Er (5861). Only the last of these can be related to a historical or legendary tradition; the rest are apparently simply titles, sometimes attached to geographical or tribal names: the Ili river valley, the Otiiken (i.e., Tannu-ola?) mountain range, the tribes of Yaghma, Chigil, etc. Others are even more general, like the Beg of Begs (893), the Beg of the people (6138), the Beg of the Realm and the City (216), the Beg of the Great City (5354), etc. There is no discernible pattern to the citing of these authorities. The wisdom attributed to any one is usually of such a vague or general nature that it could just as well be attributed to any other, or to the standard Wise Sage who is in fact cited the most aften; although wisdom relating to trade is sometimes put in the mouth of the Chief of the Merchants (2745, 3002, 5754). There is also little if anything to distinguish the wisdom attrbuted to the above authorities from that embodied in what are called proverbs (110, 1623, 3514) or Turkish proverbs (273, 319, 880, 1798, 1826). The same holds true for the scattered quatrains attributed to various anonymous poets. Except for their gnomic character, and their peculiar rhyme scheme, there are no other characteristics which distinguish them from the surrounding text. The reason is clear. All of this material, quatrains, proverbs, and wisdom sayings alike, was composed by Yusuf himself, using the Perso-Islamic wisdom and ethics tradition as his model. A m o n g the appeals to authority we do find one citation of an Arabic saying (5809), also one pretending to quote the Iranian Sage (3265). In the introductory portions of the work, in fact, Yusuf forthrightly acknowledges his debt. At the beginning of the chapter "On the Virtue and Benefit of W i s d o m and Intellect," which is little more than a litany of proverbial wisdom, he mentions as model Nushirvan the Just (line 290). And in the preceding chapter he mixes an encomium on Turkish royal glory (exemplified by Alp Er Tonga = Afrasiyab) with praise for the Iranian royal tradition as it

A TURKO-1SLAMIC

MIRROR

FOR

PRINCES

143

is embodied in literature (276-82). The two key words in the title substantiate this debt: bilig, like Ar. hikma, means both Wisdom and wisdom saying, corresponding to Pers. khirad and andarz (bilig is found in hendiadys with hikmet in line 2698); while kut, corresponding to Pers. farr, is found throughout the work alongside its synonym devlet (Ar. dawla).

Persian

Caiques

The clearest indication of Yüsuf's debt to the Perso-Islamic tradition is his language. Here I refer, not to loan words (such as hikmet, devlet, etc.) which are actually very few in number, but to the semantic value attached to the Turkish vocabulary, pointing clearly to loan translations or caiques on Persian (occasionally with an Arabic antecedent). A large number of such caiques are listed in Appendix 4. These usages (for example, "long hand" indicating authority or tyranny) were probably no less unidiomatic in eleventh-century Turkish than they are in twentieth-century English; and for this reason I have preferred to leave the literal rendering stand in my translation, since this better conveys the effect the original must have had on its audience. Similarly, from our point of view, the language of a work like Kutadgu Bilig is overloaded with clichés. Thus death is a "horseman" (atgi: 6035, 6055-56; kógütgi: 3788,6112); sickness is "Death's vanguard"; hospitality is always indicated by "bread and salt"; food and clothing by "throat and shoulder"; a man who is content with his lot is "full-eyed" while one who is covetous and greedy is "hungry-eyed"; "red cheeks" mean health and prosperity, "yellow cheeks," illness; an "open brow" cheerfulness, a "knit brow," anxiety; etc. These clichés also depend on corresponding Persian expressions. For the most part I have refrained from paraphrasing, and left these too in their literal form. Very interesting is the value attached to the concept "world." On the one hand, World means Realm, the territory which any prince controls (e.g. line 2034: ajun = él). This accords with the Near Eastern royal ideology and, specifically, with Persian usage. Thus "world-conqueror" (281: ajun tutgu, 284 cihán tutgu, 445 ajun tutmi§ er, etc.) traslates Persian jahan-glr. On the other hand, World is a synonym of Fortune (e.g., 2759; dünyá = devlet; 6122-34: ajun = dünyá = devlet = kut). It thus shares in the ambiguous valuation of Fortune, now positive, now negative. "World" also has the negative connotations of Ar. dunyá as opposed to axira and to din (see below, under "Analysis and interpretation of the work").

144

FROM

M A H M U D

KA§ GARI

TO

E V L1 Y A Ç E L E B 1

A final group of terms clearly reflects the Iranian statecraft tradition. The terms may be headed by Gate, which means the ruler's court (cf. Pers. dar). Then the ruler's authority is envisioned as encompassing three pairs of opposed concepts. These are: 1. The inner court (Pers. andarun) and the outer court (Pers. birun). The former refers to the prince's private affairs, his harem, bodyguard, palace staff; while the latter refers to the public affairs of state, the military, and the bureaucracy. The corresponding Turkish terms are ig — ta§ (425, 1645, 2500, 2552, 5495); also ev igre — ta§tin (5848); ig elde — ta§irtin (5547; cf. 3115). 1 2. The subjects and the realm (Pers. ra'iyyet vs. valayet, mulk; Tk. bodun vs. el). The distinction here is between the populace and the territory. The two are often paired, indicating the sum total of the prince's responsibility. If he supplies justice and security, then the people grow rich, and the land thrives (e.g. 1039). 3. The common people and the military (Pers. ra'iyyet, mardom vs. laskar, sipah; Tk. bodun vs. er at). This is the basic division of society, the former including all those who pay taxes, the latter all those who do not (therefore courtiers and nobles of various sorts as well as the military proper). The subjects form the basis of the wealth of the kingdom, and the prince in return is obligated to provide them with justice; the troops, on the other hand, form the bulwark of defense, as well as the spearhead of conquest, and the prince must provide them with cash to assure their support (e.g. 2133 ff.). When Highly Praised instructs Wide Awake in how to deal with the various classes of society, he first discusses the nobles or "military" (chapter 48), then the commoners (chapter 49), before proceeding to the various occupations and social groups. 2

1 Related to this is the distinction between the "seat of honor" (lör) and the "courtyard" (el), respectively the noble and the ignoble parts of the chieftain's dwelling (262-63, 614, 844, 2253). See discussion in my "Inner Asian Wisdom Traditions in the pre-Mongol period," JAOS 101/1 (1981), 91-92. 2 F o r a fuller discussion of these topics, see H. Inalcik, "Kutadgu Bilig'de Türk ve Iran siyaset nazariye ve gelenekleri," in Resid Rahmeti Arat Için (Ankara, 1966), 259-71. According to Inalcik, in the Turkish tradition, the ruler does not simply dispense justice, which depends on his own magnanimity (the Iranian concept), but also is bound by customary law (törü), an objective set of rules. It is questionable whether this distinction exists in Kutadgu Bilig, where törü is a synonym of könilik and means simply "justice" (e.g. 355, 800, 2058 [the "circle of equity" discussed by Inalcik, p. 263], etc.). Also, the ruler has it in his power to lay down good laws or bad (1456-61). See further: M. Grignaschi, "La monarchie karakhanide de Kachgar et les relations de dépendance personnelle dans le ' K u t a d g u Bilig' ... ," Recueils de la Société Jean Bodin 20 (1970), 515-626.

A TURKO-ISLAMIC

Greek

MIRROR

FOR

PRINCES

145

Wisdom

While the wisdom imparted in Kutadgu Bilig has a predominantly Iranian source, there is a Greek component as well. It is sometimes hard to distinguish between these. Thus, the need for providing the military with cash and the commoners with justice is also found in Islamic literature as a counsel attributed to Aristotle. 1 Another idea ascribed to Aristotle by the Arabs, which is also echoed in Kutadgu Bilig (2452), is that "a man without intelligence is like a statue without spirit." 2 We have seen that the allegorical depictions of Justice and Fortune (620-835) have Greek prototypes. Bombaci demonstrated that the descriptions derive from Hellenistic depictions of the goddesses Dike and Tyxe (see above, "Mirrors for Prince," text at p. 138, n. 1), but he was unable to suggest what Yusufs immediate source might have been. With regard to Fortune, at any rate, it might have been an Arabic translation of the description of a wellknown painting attributed to Cebes, a disciple of Socrates. 3 More likely, Yusuf found the full-blown allegory in some Arabic or Persian treatise that is at present unknown to us. Greek wisdom had long before found a place (modest in comparison with Iranian andarz) in Arabic Arab or polite culture, and Yusuf was undoubtedly familiar with Greek philosophy in the form of Arab Falsafah. Otto Alberts made an exaggerated claim in this regard, viz. that he was a pupil of the great Ibn Sina (d. 1037). 4 Alberts's analysis is most unconvincing. The learning — ethical, political, psychological, and theological — that is reflected in Kutadgu Bilig is the standard unsophisticated kind typical of Adab and the mirror-for-princes tradition. The only doctrine that might be considered at all philosophical is the distinction between the faculties of wisdom (bilig) and intellect (uku§), wisdom being learned and intellect being inborn (1677 f., 1815 f.). According to Alberts 5 this corresponds to Ibn Sina's distinction between theoretical reason, which knows good from evil, and practical reason, which chooses the good. Even Alberts admitted, however, that we do not find ' He wrote to Alexander: "Make yourself beloved to your elite by means of generosity and to your commoners by means of justice." D. Gutas, Greek Wisdom Literature in Arabic Translation (New Haven, Conn., 1975), p. 183. F. Rosenthal, Four Essays on Art and Literature in ¡slam (Leiden, 1971), p. 10. Cf. p. 16: the notion that beauty of appearance and beauty of character go together, ascribed to Plato (cf. KB 2216). Note also: |Aristotle:| "he who has no family and no children has no name and no progeny" (Gutas, 187; cf. KB 3375). •'F. Rosenthal, "The Symbolism of the Tabula Cebetis according to Abu 1-Faraj ibn at-Tayyib," in Recherches d'lslamologie ... Anawati et... Gardet... (Louvain, 1977), 281. ^O. Alberts, "Der Dichter des in uigurisch-türkischem Dialect geschriebenen Kudatku bilik ... ein Schüler des Avicenna," Archiv für Geschichte der Philosophie, n.s. 7 (1901), 319-36. 5 Ibid, 328.

146

FROM

MAHMUD

KA§GARl

T O EV LI Y A

CELEBI

the distinction in this simple-minded form in Ibn Sina's own writings; and he failed to note that the philosophers did not consider the first to be learned, the second innate, which is the substance of Y u s u f ' s doctrine (Yusuf hardly distinguishes them from the point of view of function). A much closer analogue to Yusuf s teaching is to be found in the work of al-Mawardi (d. 1058). 1 According to al-Mawardi, intellect (al-'aqt) is of two types: innate (garizi) and acquired (muktasab). Innate intellect is what distinguishes man from the other animals (cf. KB 1843). As to acquired intellect, it grows through experience and learning, theoretically without limit (cf. KB 1827). If the Qur'an was revealed as a warning to those who are "alive" ( h a y y — Q. 36:70), this means those who are "intelligent" ('aqil — cf. KB 1839). According to a hadith, intellect is "a light in the heart which distinguishes the true from the false" (cf. KB 288). While some say intellect is located in the heart, others say it is in the brain (cf. KB 1836: the brain). I do not mean to suggest that Yusuf derived this particular psychological doctrine from al-Mawardi and not from Ibn Sina; only that the learning reflected in Kutadgu Bilig accords more with the standard, Sunni, Adab style literature than it does with Falsafah. 2

Buddhism? If the Hellenic influence on Kutadgu Bilig was quite limited, and certainly mediated through Arabic literature, the same can be said for influence from the East; although some exaggerated claims have been made in this regard as well. Early scholarship on the work suffered from the happenstance that the first manuscript to be discovered and printed was the so-called Vienna

1

Adab al-Dunyd wa-l-Dln, ed. M. al-Saqa (Cairo, 1393/1973), 20f.

have now abandoned the idea, which I argued formerly, that the figure of Highly Praised represents the law-giving wisdom of the philosopher-king, a wisdom that according to the Muslim neo-Platonists (notably al-Farabi and Ibn Sina) had found perfect embodiment in the Prophet himself; and that this explains the name Highly Praised (Ogdiilmi§), apparently a translation of the name Muhammad. (Og- is "to praise," but there must also be a play on words intended with og "mind.") Yusuf states only that Highly Praised stands for intellect (356). The idea that he "represents" Muhammad has been put forward by S. f a g a t a y ("Kutadgu Bilig'de Ogdulmi§," Turk Kulturu 98, Aralik 1970, 107). Franz Rosenthal writes (private letter dated 29 January 1978): "I do not doubt that 'highly praised' would be possible as a translation of Muhammad or Ahmad, but I have my doubts as to whether it would be possible to assign to the Prophet Muhammad a comparatively minor role in the scheme of things, even if he should be conceived of in the sense of philosophers as philosopher-king. I think it would require quite heterodox sectarians to put Muhammad in the place he seems to occupy in the KB."

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

147

ms. which is in Uighur script. 1 It is in fact the latest of the three extant mss.; and it reflects a revival of Uighur script among the Timurids in the fifteenth century. But the notion became current that Kutadgu Bilig represented "Uighur" literature — one ingenuous critic went so far as to call it "the national epic of the Uighurs." 2 And serious scholars have found reflected in it Confucian or Buddhist doctrines. In particular, Saadet Qigatay has proposed that the figure of Wide Awake, the ascetic and world-renouncer, represents Buddhism; and that the debate between Wide Awake and Highly Praised represents the conflict between Buddhism and Islam, a conflict that was going on at both the political level and the level of what may be termed "subversive asceticism," which somehow threatened the Karakhanid state. 3 The best argument in favor of this hypothesis is that the names of the two opposing figures — Wide Awake (lit. "Awakened," Tk. Odgurmi§) and Highly Praised (lit. "Praised," Tk. Ógdülmi§) are translations respectively of the names Buddha and Muhammad. From this the conclusion follows that Wide Awake "represents" Buddhism and Highly Praised "represents" Islam. 4 There are several fallacies in this line of argument: 1. Like the other proper names in Kutadgu Bilig, these are highly charged allegorical names, and are not meant to hint at historical figures (see below, "Analysis and Interpretation of the Work"). Yüsuf himself tells us what the figures represent: Highly Praised stands for intellect (no doubt a play on words is intended with og "mind"), and Wide Awake stands for the Last End — i.e. the consciousness wakened to religious responsibility and directed toward the hereafter (see discussion below). If the name of Wide Awake translates anything, it can only be Ar. Yaqzan, which had been used earlier as an allegorical name by Ibn Slna in his symbolic fable entitled "Alive son of Awake" (Hayy ibn Yaqzan).

H. Vambery, Uigurische Sprachmonumente und das Kudatku Bilik (Leipzig, 1870); W. Radloff, Kudatku Bilik, Facsimile der uigurischen handschrifl der K. K. Hofoibliothek in Wien (St. Petersburg, 1890); W. Radloff, Das Kudatku Bilik des Jusuf Chass-Hadshib aus Bälasagun. Teil I: Der Text in Transcription. Teil II: Text und Übersetzung nach den Handschriften von Wien und Kairo (St. Petersburg, 1891, 1910). 2

R . Devereux, "Yüsuf Khäss Häjib and the Kutadgu Bilig," Muslim World 51 (1961), 299. S . Cagatay, "Kutadgu Bilig'de O d g u r m i j ' m Kijiligi," TDAY-Belleten, 1967, 39-49. Cf. Devereux, 303; Bombaci, Köprülii Armagani, p. 75; I. Kafesoglu, "Kutadgu Bilig ve Kiiltür Tarihimizdeki Yeri," Tarih Enstitüsü Dergisi 1 (1970), 11. 4 I have tried to state the argument in its most cogent form, ('agatay herself states ("Ögdiilmis," 107) that the equation Ögdülmi§ = Muhammad was suggested to her by Professor Benzing, only after her article on Odgurmig was published; and although she accepts the suggestion she has reservations about it, since there is still the problem of Highly Praised's "pessimism" toward the end, when he seeks Wide Awake's advice with regard to his intentioin to "repent." ("agatay's argument is based on a view of Islam that is totally this-world oriented, seeing otherworldliness as alien to the true Islamic spirit. 3

148

FROM

MAHMIJD

K A § G A R I TO

EVLIYA

gELEBI

2. Highly Praised by no means comes off as the winner of the debate; quite the contrary, as we shall see (below, text at p. 169, n. 1). It is hardly possible to think that Yusuf should incorporate a Buddhist ethic so favorably into a work suffused with Islamic learning; or that a figure representing the Prophet of Islam should fare so poorly (see p. 146, n. 2). Indeed, the debate between the statesman Highly Praised and the ascetic Wide Awake is better understood in light of an ancient conflict within Islam itself (see discussion below, under "Solitude versus Society"). 3. While granting that Wide Awake has become a Muslim, £ a g a t a y holds that neverthless he represents those w h o are secretly partisans of their former religion, and that this shown that the ascetics of Islam at this period had not separated f r o m Buddhism. 1 This line of reasoning depends on the highly q u e s t i o n a b l e a s s u m p t i o n (favored by such earlier schoolars as Goldziher) that the ascetic and mystical movement within Islam was due to the alien influence of Buddhism. It also fails to separate the question of the possible influence of Buddhism on some of the early Sufi masters (notably Ibrahim ibn A d h a m ) and the question of an ongoing influence a m o n g the Turks of Central Asia in the eleventh century. Although the latter is not impossible, it is at best unproved; and it would be circular reasoning to argue for it on the basis of the figure of Wide Awake in Kutadgu

Bilig.2

There is nothing particularly Buddhist about the doctrine of world renunciation. It is prominent in Islam as well (see discussion below), and is even part of the stock wisdom attributed to both the Greeks and the Iranians in Arabic Adab literature. 3 If the extreme claim of direct Buddhist influence in the Kutadgu Bilig is to be rejected, the more modest claim of indirect influence may still be maintained in connecton with the literary theme of "ascetic versus prince," as mediated by Arabic versions of the Barlaam and Josephat r o m a n c e . 4 T h e aspect of the Buddha legend that appealed most to the popular imagination ^"Odgurmig," 46. ^What is perhaps the most far-fetched of fagatay's ideas is the suggestion (ibid., 45) that the name of Wide Awake's disciple Kumaru ("Testament") reflects Kun gah vo, the tibetan equivalent of Ananda, the disciple of the buddha. She fails to explain how Yusuf or his audience might have had knowledge of this fact; and fails to discern that this name has an allegorical meaning and must be understood in the context of the structure of the work as a whole (see below, under "Analysis and Interpretation ofr the Work"). 3 For example, it is counseled by Bozorgmihr according to the Javidan Khirad, 29. Note also: [Socrates:] "Your days are three" (Gutas, 115; cf. KB 3605f.). In Langland's Piers the Ploughman, Lady Holy Church admonishes Will in his first vision thus (trans. J. F. Goodrich, Penguin Books, 1959, 33): " Put no trust in your body, for its promptings come from the World, and the world is a liar out to betray you. And the Flesh and the Devil are in league to pursue your soul, and speak evil things to your heart. I give you good counsel that you may be watchful." (Cf. KB 3589-90) See EI2, article "Bilawhar wa-Yudasaf' (D. M. Lang).

A TURKO-ISLAMIC

MIRROR

FOR

PRINCHS

149

was that of a king's son, raised in royal luxury, giving up his princely station to seek his salvation as a religious hero. In its literary form, the legend was livened with dialogues and debates setting out the opposing ideals of kingship and religious quest, meanwhile touching on themes common to the mirrorfor-princes tradition. Thus, in the Buddha-carita (third century AD) we find the following: Lord of men, when Time binds and drags away the king, neither relations nor friends nor sovereignty will follow you; all will depart, afflicted and helpless. Your deeds alone will accompany you like a shadow.... Therefore guard your kingdom according to the Law, if you desire Paradise and a good reputation. For there is no kingdom at all in Paradise for the king who in his delusion misapplies the I ,aw.'

Based on Indian texts related to the Buddha-carita, a version of the Buddha legend (with specific Buddhist content omitted) was popular in a Manichaean version in Central Asia; and a related version, probably in a Pahlavi recension, was translated into Arabic in the eighth century, as the Book ofBalawhar and Budhasaf. This original Arabic recension is lost, as is the metrical version by Aban al-Lahiqi (d. 815); but several later versions are extant. 2 From Arabic, the legend passed into Georgian, Greek, and Hebrew and, via the Greek, into Latin, Armenian, and Old Slavonic, as the romance of Barlaam and Josephat (Josephat even became a saint of the Christian church). In the Muslim world, the book was popular in Manichaean (zindlq) and Ismaili circles, and perhaps for this reason was unpopular in Sunni circles (it is hardly referred to in books on zuhd "renunciation" and dhamm al-dunya "blame of the World")- 3 The theme, however, entered into the mainstream of Sufi literature, in the story of the prince of Balkh, Ibrahim ibn Adham; and is even found in the Arabian Nights. 4 The mirror-for-princes association is still evident in the Arabic versions, as in the following excerpt quoted by Ibn Babuya (d. 991).

^Quoted in D. M. Lang, The Wisdom of Balahvar: A Christian legend of the Buddha (New York, 1957), 19. 2 E . Rehatsek, trans., "The Book of the King's Son and the Ascetic," J RAS, 1890, 119-55; D. Gimaret, Le livre de Bilawhar et Büdäsf selon la version arabe Ismaélienne (Geneva-Paris, 1971). Gimaret, "Traces et paralleles du Kitâb Bilawhar wa Büdäsf dans la tradition arabe," BEO 24 (1971), 132. According to Gimaret, the esotericism lies not in the content but in the tone or atmosphere — the story of pious men who suffer persecution and martyrdom, a sage who in a hostile environment and disguised appears to communicate his doctrine to the prince whom he sees at night in great secrecy, etc. ^The story of Harun al-Rashïd and his ascetic son (Night 401; Burton trans. 5:111 f.); the story of the "Lord of the Cloud" and the devout king (Night 473; Burton trans. 5:274f.).

150

FROM

MAHMUD

K A § G A RI

TO

EVLlYA

£ELEBi

Oh king! Why do you leave your kingdom and neglect your subjects?... D o n ' t you k n o w that the highest rewards are for those w h o do g o o d to p e o p l e and the highest g o o d is the o b e d i e n c e to the c o m m u n i t y and to your p e o p l e ? . . . D o n ' t y o u k n o w n that the best w o r s h i p is g o o d d e e d s and the most important deed is to o c c u p y yourself with the government of the people? A n d you, oh king, are just with your subjects, a s s u m i n g their w e l f a r e by your g o v e r n m e n t , and y o u w i l l get y o u r reward according to the measure of the g o o d w h i c h you have d o n e . . . . A n d if y o u w i s h e d for their ruin, you have brought upon y o u more sin than the reward w h i c h y o u have achieved in l o o k i n g after y o u r s e l f . . . . B e w a r e , oh king, that y o u throw o f f your princely garments w h i c h are the m e a n s to the glory of this world and of the afterworld! 1

It can hardly be doubted that as a literary theme, the conflict between the statesman Highly Praised and the ascetic Wide Awake, in the second half of Kutadgu Bilig, is partially modeled after the theme of ascetic versus prince, popular in Arabic literature, and based originally on the Buddha legend.

Qur'an and Hadith Kutadgu Bilig is suffused with Islamic learning. Yet there is no direct reference to any Islamic source. This is in line with Yusuf s aim and methods in the work as a whole. We saw above (under "Characterization of the Work") that he was attempting to make a new home for the Irano-Islamic traditions and ideals as part of an Inner Asian Turkish literary heritage; and (under "Appeals to Authority") that in his appeals to authority, characteristic of Islamic mirrors for princes, he almost exclusively cited various Turkish princes as well as unnamed Turkish poets. Similarly, the display of Islamic learning in Turkish, without any direct reference to its source, tends to demonstrate that the Turks, as well as the Arabs and the Iranians, are in possession of the truths of the religious sciences, just as they are in possession of the truths of political wisdom. On the other hand, since the book was written for a Turkish audience, it must be assumed that the reader would immediately recognize the Islamic basis of these truths. In this source analysis it is our task to sift these references and assign them to their likely sources. Quotations from or allusions to the Qur'an are identified in the text (see notes 1-5, 29, 40, 42, 44, 53, 77, 91, 92, 103, 114, 133, 149, 162, 175, 179, 199, to the translation).

^S. M. Stern and Sofie Walzer, Three Unknown (University of South Carolina Press, 1971), 22-23.

Buddhist

Stories

in an Arabic

Version

A T U R K O - I S LAM IC M I R R O R

FOR

PRINCES

151

T h e identification of hadiths, or sayings which in Islamic sources are generally attributed to the Prophet, is a trickier business, since m a n y such sayings m a y also be attributed to other authorities ( ' A l i , f o r m e r prophets, etc.) or else are of the nature of proverbs. Nevertheless, most M u s l i m s would probably recognize most of the f o l l o w i n g lines in Kutadgu

Bilig as referring

to hadiths or other Islamic traditions given here: 3240, 3666: on seeking one's livelihood. "Seekings one's livelihood (.talab al-halal)

is a duty incumbent on every M u s l i m . " '

3975f.: on the positive value of wealth. "How good is goodly ( s a l i h ) wealth to the goodly man." 2 3344: the carnal soul is an enemy. "The worst enemy is your carnal soul (nafs) which is between your sides."-' 3928f.: on helping others. "The best man is he who is the most useful to other people." 4 The believer gets along with others; there is no good in the man who does not get along with others, and the best man is the one most useful to others."^ 4201-6, 4224-28: "Love for the sake of God; hate for the sake of God." 6 4486f.: on marrying. "A woman may be married for four things: her money, her birth, her beauty, and her religion. Get thou a religious woman (otherwise) may thy hands be rubbed in d i r t ! " W h e n a man marries a woman for her money or her beauty, he is deprived of both her beauty and her money, but when a man marries a woman for her religion, God provides him with her money and beauty as well." 8 4548: on neighbors. "First the neighbor then the house, first the water then the l a n d . " ( a l - j a r tumma

al-dary

4991, 5490: on friendship. "The Muslim is a brother to the Muslim."'® 4733f.: the world is a field. "This world is the seedbed for the other world" (al-dunya

mazra 'at

al-axira).''

' ß a d l al-zamän Furüzänfarr, Ahädith-i Mathnawi (Tehran, 1334/1955), 7, 168; M. Shaybänl (d. 804), Kitäb al-Kasb, cited in Goitein, Studies, 222. ^Furüzänfarr, l t . Furuzänfarr, 9; al-Sulami, Kitäb ädab al-suhba (ed. M. J. Kister, Jerusalem, 1954), 45; alSarräj, Kitäb al-luma' fi l-tasawwuf (cd. R. A. Nicholson, London, 1914, repr. 1963), 12; cited in A. Schimmel, Mystical Dimensions of Islam (University of North Carolina Press, 1975), 112. ^al-Makki, Qüt al-Qulüb (Cairo, 1351/1932), 3:38; Kaikävus ibn Iskandar, Qäbüs-nämeh (Tehran, n.d.), ed. Sa'id Nafisi, 146 = Levy, A Mirror for Princes, 194. Furüzänfarr, 179. 6 I. Goldziher, Muslim Studies, ed. S. M. Stern, 2 (London, 1971), 355-56. ^A. Guillaume, The Traditions of Islam (Oxford, 1924), 124; al-Hujwiri, Kashf al-Mahjüb (trans. R. A. Nicholson, London, 1911, 1936, repr. 1959), 361. °A1-Ghäzäli, Kitäb ädäb al-nikäh (book 16 of Ihyä' 'Ulüm al-Din; Cairo, n.d.), 2:38. 9 y A well-known proverb, sometimes quoted as a hadith. Al-Ghazali on the Duties of Brotherhood (trans. M. Holland, London, 1975), 46, 51. (Also cf. Qur'än, 49:9.) 11 Furüzänfarr, 112, 203-5; Schimmel, 107.

FROM

MAHMUD

K A § G AR I TO E V L I Y A

£ ELEB I

4809: the world is a prison. "This world is a prison for the believer." 1 5982: on sickness. "Do not curse it (fever), for it refines away sin as fire refines iron." 2 4373, 6001-3, 6021, 6038: on dreams. "The dream is fulfilled according t o i t s i n t e r p r e t a t i o n " ( a l - r u ' y a 'aid ma

tu'abbaru)?

5311f.: the world and religion are divergent paths. "This world and the other world are like the East and the West: the farther you move away from one, the closer you move toward the other." 4 "This world and the other world are enemies at variance, and roads that diverge. If you love this world, you hate the next. They are like the East and the West: the closer you are to one, the farther you are from the other."-' 6314f.: on death as an admonition. "Death is s u f f i c i e n t as an a d m o n i s h e r " ( k a f a bi-l-mawt

wa'izan

Sufism

Several of the ideas put forward in the work by the figure of Wide Awake are drawn directly from Sufi literature. Thus: 3386-87: marriage is like a ship. Ibrahim ibn Adham: "When a man marries he embarks on a ship, and when a child is born he suffers shipwreck." 7 3540f., etc.: the world is a bride/a hag. Yahya ibn Mu'ad: "The world is a bride..." 8 3600: the carnal soul is like a restive horse. 9 4754f. (attributed to the Irkin of Ila!): on serving a prince. Sufyan alThawri: "Do not keep company with the Sultan. If you are obedient, he will overburden you; and if you are disobedient, he will kill y o u . " 1 0 4786f.: mutual love of God and man. Bayazid al-Bistami: "At the beginning I was mistaken in four respects. I concerned myself to remember God, to know Him, to love Him, and to seek Him, and when I

'Furuzänfarr, 11; al-Sulami, 'Uyüb 1976), 91. 2 Furüzänfarr, 108.

al-nafs

wa-mudäwätuhä

(ed. E. Kohlberg, Jerusalem,

^M. J. Kister, "The interpretation of dreams: An unknown manuscript of Ibn Qutayba's ' 'Ibäret al-Ru'yä'," Israel Oriental Studies 4 (1974), 100. 4 A b ü Hayyän al-Tawhidi, Kitäb al-imtä' wa-l-mu'änasa (ed. A. A m i n and A. al-Zayn, Bedirut, 3 vols., n.d.), 1:15. 5 Furüzänfarr, 121 (attributed to 'Ali). ^Al-Ghäzäli, Kitäb dhikr al-mawt wa-mä ba'dahu (book 40 of Ihyä' 'Ulüm al-Din ; Cairo, n.d.), 4:450. ^Schimmel, 37, citing al-Sarräj, 199; cf. 'Attär, Muslim Saints and Mystics (trans. A. J. Arberry, London and Boston, 1966), 71. 8 al-Sarräj, 39, 1. 14. See Schimmel, 428; and the hadith (Furuzänfarr, 186): "The world on the day of resurrection appears as a hag." See also Ghazäli's Book of Counsel for Kings , 34. ^Schimmel, 112-13. (Also A. Schimmel, "Nur ein störres Pferd," Ex Orbe Religionum: Festschrift Geo Widengren, Leiden, 1972). Ghazäli's Book of Counsel for Kings , 87.

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

153

had c o m e to the end, I saw that He had r e m e m b e r e d m e b e f o r e I remembered Him, that His knowledge of me had preceded my knowledge of Him, His love towards me had existed before my love to Him, and He had sought me before I sought H i m . " 1

Even Highly Praised, in the course of his debate with Wide Awake, quotes standard Sufi doctrine: 3671: on fear and hope. "Fear and hope are the two w i n g s of action without which flying is impossible." 2

The figure of Wide Awake, living secluded on a mountain and pursuing religious devotions, is drawn on the model of the early Sufi masters. According to al-Sarraj, 3 the Sufis are those who for the sake of God have turned aside f r r o m all that might distract them f r o m Him and have preferred Him above all else.... For His sake they have deserted their native lands and forsaken their brethren and abandoned their kindred, and for Him have cut themselves off f r o m all ties and fled f r o m the f a c e of m a n k i n d , seeking intimacy with Him and alienation f r o m all e l s e . . . . T h e y are content with little in the way of worldly goods and are satisfied with the very m i n i m u m of f o o d to keep them alive, and limit themselves to the least that is necessary in the way of clothing and bedding. They choose poverty rather than riches, embracing want and avoiding plenty, preferring hunger to satisfaction and little to much. They renounce dignity and honour and rank, and show compassion upon mankind, and humility towards small and great.

And in the verse cited by al-Kalabadi: They do not quarrel over this w o r l d ' s pleasure — Honours, and children, rich and costly gowns, All greed and appetite! they do not treasure The life of ease and joy that dwells in towns. Facing the far and faint horizon yonder They seek the Infinite, with purpose strong: They ever tread where desert runnels wander, And high on towering mountain-tops they t h r o n g ! 4

' m . Smith, Studies in Early Mysticism in the Near and Middle East (London, 1931), 237-38. Cf. Schimmel, 39-40: "Rabi'a had meditated upon the Koranic statement that God's love precedes man's love: 'He loves them and they love him' (Sura 5:59)." 2 Schimmel, 127, citing al-Sarraj, 62. ^Al-Sarraj, 3, 11. The translation is that of Smith, Studies, 160. ^The Doctrine of the Sufis, trans. A. J. Arberry (Cambridge, 1935), 11.

154

FROM

MAHMUD

K A § G A R 1 TO

EVLIYA

Ç ELEB Î

N o t all the S u f i s a p p r o v e d o f this s e c l u s i o n . S o w h e n H i g h l y Praised tries to p e r s u a d e W i d e A w a k e to l i v e in s o c i e t y , he can again draw o n a standard S u f i teaching. T h u s , it w a s related o f Sari al-Saqati: One day a man came from Mount Lokam to visit him.... "Shaikh So-andso from Mount Lokam greets you," he said. "He d w e l l s in the mountain," commented Sari. "So his efforts amount to nothing. A man ought to be able to live in the midst of the market and be so preoccupied with God, that not for a single instant is he absent from God." 1 Wide Awake's

a l i e n a t i o n f r o m h i s "brother" H i g h l y

Praised

is

p r e f i g u r e d in another a n e c d o t e (cited again a c c o r d i n g to the later v e r s i o n o f 'Attar): Yahye-i Mo'adh [Yahya ibn Mu'ad] had a brother w h o went to Mecca and took up residence near the Kaaba. From there he wrote a letter to Yahya. "Three things I desired. T w o have been realized. N o w one remains.... I desired that I might pass my last years in the noblest place on earth. N o w I have come to the Sacred Territory, which is the noblest of all places. My second desire was to have a servant to wait on me and make ready my ablution water. God has given me a seemly servant-girl. My third desire is to see you before I die. Pray to God that he may vouchsafe this desire." Yahya replied to his brother as follows. "As for your saying that you desired the best place on earth, be yourself the best of men, then live in whatever place you wish. A place is noble by reason of its inhabitants, not vice versa. "Then as for your saying that you desired a sevant and have now got one, if you were really a true and chivalrous man, you would never have made God's servant your own servant, detaining her from serving God and diverting her to serve yourself. Y o u should yourself be a servant. Y o u desire to be a master, but mastership is an attribute of God. Servanthood is an attribute of man.... When God's servant desires a station proper to God, he makes himself a Pharaoh. "Finally, as to your saying that you desire to see me, if you were truly aware of God, you would never remember me. So associate with God, that no memory of your brother ever comes into your mind. There one must be ready to sacrifice one's son; how much more a brother! If you have found Him, what am I to you? And if you have not found Him, what profit will you gain from me?" 2

1Muslim Saints and Mystics, 167. Ibid„ 181.

2

A TURKO

ISLAMIC

MIRROR

FOR

PRINCES

155

The issue of whether to serve God or the prince, which bulks so large in Wide Awake's arguments (3747f., 4745f.) was a lively one for some of the early Sufis as well. In general the Sufis enlarged on the common Islamic theme of the purity of devotion to God as apposed to the taint of involving oneself in worldly affairs. The scene in Kutadgu Bilig (ch. 70), when Wide Awake finally comes to the king's court, in the middle of the night, and gives his counsel, is reminiscent of the following famous incident recounted in Sufi literature. 1 Harun al-Rashid one night summons his favorite vizier and says, "Take me to a man this night who will reveal me to myself. My heart is grown weary of pomp and pride." The vizier takes him to Fudayl ibn 'Iyad. They knock at his door. When Fudayl learns it is the caliph, he says, "What business has he with me, and what have I to do with him?" The vizier counters, "Is it not a duty to obey those in authority?" "Do not disturb me," cries Fudayl; but they go in anyway. The caliph gropes in the dark, and when their hands meet, Fudayl remarks, "How smooth and soft this palm is, if only it could escape from Hell-fire!" Harun seeks good counsel, and Fudayl admonishes him, among other things, to fear God, and obey His command. And be watchful and prudent; for on the Resurrection Day God will question you c o n c e r n i n g every single Muslim and He will exact justice f r o m you in respect of everyone. If one night an old w o m a n has gone to sleep in a house without provisions, she will pluck your shirt on the Day and will give evidence against you.

Before leaving, Harun offers him a purse of a thousand dinars, which Fudayl spurns: "The counsels I have spoken to you have yielded no profit.... I call you to salvation, and you cast me into temptation." As he leaves the house, Harun remarks that Fudayl is "in truth a king of men. His arrogance is extreme, and the world is very contemptible in his eyes." Wide Awake, similarly, warns the king that if a single one of his subjects goes to bed hungry at night, God will hold him to account for it (5165); although along with such sermonizing he intersperses typical mirrorfor-princes themes (do justice and assure a lingthy rule, 5170-73; be generous or you will get a bad name, 5220; do not drink or fornicate, for these destroy a king's court, 5263). Wide Awake states his aim to be devotion to God. He uses the Arabic term ta'at "obedience" as well as the Turkish tapug "service" (e.g. 3339; Ar. xidma, also 'ibada, 'amal). He sounds the Sufi themes of trust in God and satisfaction with one's lot (cf. Ar. tawakkul, rida); remembrance (cf. Ar. ^bid., 57-60. Cf. al-Hujwi ri, Kashf, 98-100; Ghazdli's Book of Counsel for Kings, 20.

156

FROM

MAHMUD

K A § G A R I TO

E V LI Y A

gELEBt

and purification (Tk. suzul3522-24, 3632, 3752, 4795; cf. Ar. safwa). His aim is to join his heart to God (4778 f.). As a means to this end he has adopted a life of poverty (Ar. faqr), renunciation (Ar. zuhd), and solitude (Ar. xalwa, 'uzla). He is an ascetic (here the Arabic term zahid is used: 3441, 3563, 3915, 3929, 3929, 3976, 5106). At one point he counts himself among the saints (Ar. abdal, 3563). He is vigilant in his struggle with the lower self or carnal desire (here too we find the Arabic terms nafs, hava, along with the Turkish et or. 3344-46, 3635-44, etc.). He lives isolated on a mountain, eats only barley gruel or millet bran, and wears a cloak of wool (4767; cf. Ar. suf.). dikr);

In the language of renunciation, the object is usually specified as "the world" (Ar. dunya, less frequently Tk. ajun), sometimes as worldly goods (Tk. nerig "things") or as society (Tk. ulu§ kend "the city"). The commonest verb used is "to put aside" (kod-, reflecting Ar. taraka: 3327, 3347, 3441, 3563, 3592, 3652, 4732, 4737, 4798, 4805, 5115, 6050, 6154, 6334). W e also find "to turn o n e ' s face" (yiiz evtir-: 3148, 3203), "to be separate" bol-: 3348), "to give up" ( t i i n g i i l - : 4792), "to be cut o f f ' ( i i z u l - : (idnngin 4797), "to be distant" (yira-: 3327, 3350, 3552, 4806), and "to flee" (kag-: 3502, 3695, 4797, 6050, 6334). Opposed to all this is the concept of associating with other men, expressed by the term "to mix" ( k a t i l - , also karil-, reflecting Arabic xalata: 3340, 3363 f., 3405, 3502, 3922, 3994, 4686, 5750). W e also find the phrase "to enter society" (ki§iler ara kir-\ 3366); and Highly Praised speaks of "the religious merit to be found in society" (ki§iler araki muyan: 3499). Of Wide Awake it is repeatedly stressed that he lives absolutely alone (yalnnguz: 3204, 3328-30, 3349, 3359 f., 3423, 3437). The positive value in this is that it frees him from all worldly distractions. In fact the key to the character of Wide Awake is his constant vigilance against anything that might distract from his religious purpose. If devotion to God, leading to the heart's union, is Wide A w a k e ' s aim, and renunciation is the means, the existential symbol of his condition is waking or wakefulness. The Sufi background to this doctirne as well is clear. Sleeplessness w a s one of the boasts of the early Sufis generally. 1 One of them said, "Whenever I doze awhile, I hear a voice calling me and saying, 'Sleepest thou to Me? If thou sleepest to Me, verily, I will strike thee with a whip!" 2 And Shibli said, "God looked upon me and said, 'he who sleeps is heedless, and he who is heedless is veiled.' " 3 According to al-HujwIri, repentance (tawbat), ^See e.g. M. Smith, Studies, 132, 139, 162. The Doctrine of the Sufis, 157 ^Al-Hujwiri, Kashf, 351; Muslim Saints and Mystics, 281-82. 2

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

157

which is the first station of pilgrims on the way to the Truth, "has its origin in the stern prohibitions of God and in the heart's being aroused from the slumber of heedlessness." 1 And according to 'Abdallah al-Ansari (d. 1089), who systematized the Sufi path into one hundred stages, the very first of these stages is "awaking" (yaqaza).2 This does not mean waking up from physical sleep; it is a symbolic wakening from the slumber of heedlessness or neglect of one's religious duties (Ar. gafla). Wakefulness is a key element in the value structure of Kutadgu Bilig as a whole, and is not confined to the character of Wide Awake. When King Rising Sun is first introduced, for example, and his virtues are recounted — wisdom, justice, bravery etc. — among them we find wakefulness (odugluk) and alertness (saklik), which, beside being beneficial for princely rule and long life, "is also a matter of religion ( § e r i ' a t ) " \ (446). Here, typically for the mirror-for-princes genre, religion, far from supplying the basic values, plays a secondary role; it comes in almost as an afterthought, and merely to back up a value that is established on other, practical, grounds. It is as a worldly quality that vigilance is desirous in a prince; and as such it is stressed again and again (1452, 1956, 2016, 2024, etc.). A different tune, but using the same notes, is sounded by the vizier, Full Moon, while giving admonition on his deathbed to his son Highly Praised: "Awake (odungil) and do tomorrow's work, for heedlessness (osallik) destroyed me" (1208-9). Similarly, when Highly Praised conceives the intention of repenting and taking up the dervish path himself, he says that in serving men rather than God he has spent his life in vanity and in the "sleep of neglect" (5632, 5690, 5693). The figure of Wide Awake typifies this religious consciousness (e.g. 4908, 6075, 6159, etc.). According to a hadith, "men are asleep, and when they die, they awake." 3 And it is Qur'anic doctrine (Q. 29:64), echoed again and again in Sufi thought, 4 that true life is the afterlife, and belongs to those who are "dead" to this world. Again, according to the Qur'an (Q. 28:83): "That is the Last Abode (al-dxira); We appoint it for those who desire not exorbitance in the earth, nor corruption. The issue ultimate (al-'aqiba) is to the god-fearing" (Arberry's translation). Yusuf tells us (357) that Wide Awake personifies 'akibet, this ultimate issue, the Last End, the Hereafter. He knows what is ultimately best for a man's soul, and he counsels the king from this point of view. He is thus at the opposite pole from Highly Praised, the personification l

Kashf, 295; cf. Schimmel, 107. Kitab manazil al-sa'irin, ed. Beaurecueil (Cairo, 1962); cf. Arberry in The Cambridge History of Islam (Cambridge, 1970), 2:618. ^Schimmel, 382; Furuzanfarr, 81. 4 A. J. Arberry, Sufism (Harper Torchbook, 1970), 40-41. 2

158

FROM

MAHMUD

K A § G A R i TO E V L I Y A

gELEBl

of this-worldly wisdom, who knows what is best for the world's governance. Their debate, which takes up most of the second part of Kutadgu Bilig, reflects what a contemporary scholar has called the "unmistakable tension in Islam between the claims of history and those of eternity." 1 The immediate literary background of the debate, however, is to be sought once again in Sufi literature.

Solitude versus

Society

The unsociability of many of the early pietists provoked a long-lasting debate in Islam over the best way of living a full religious life, whether alone or in the company of one's fellow Muslims. We find repercussions of this debate within Sufi literature itself. The Sufis had a lively awareness of the dangers of the solitary life; 2 but most of them adopted it and approved of it, at least for a part of their career. The issue, as it was generally posed, was between withdrawal from the world (Ar. 'uzla; also xalwa, wahda, tark aldunya, etc.) and "mixing" (Ar. mwcalata). Here is chapter 8, "on withdrawal from the world" (Fi al-'uzla), from the "Testaments" of al-Muhasibi: And I warn you, my brothers, against mixing with other men {muxalatat al-nas). For every transgression, and all burdens and cares, come from mixing with them, and associating with them. But you do not perceive that; only those realize it who are abstemious and examine their own consciences (ahl al-wara' wa-l-muhasaba). We are not the sort who will escape with our religion safe and sound, if we keep company with the satans of men and of jinn — and we are like them! — goading one another with words of vanity and deception. Nay, associate with two men only: he who helps you to piety and the fear of God, and he who succors you in your worldly concerns. And if God combine for you religious help and worldly succor in the form of one man, then hold fast to him, and avoid anyone else. For all of them are harmful to religion, except he who helps to piety. Nay, the greatest safety is in avoiding others altogether ( m u j a n a b a t al-nas), and in this is the most abundant merit; it far outweights your fears. Thus we have been told: that worship is one part silence and nine parts the avoidance of other men; and thus do I advise you, if you can accept it. But those who can accept it are few. And the endurance of isolation is difficult. It is a grace of God, which He gives to w h o m He wills — may God grosper you and bring you to every good by His

^F. Peters, Allah's Commonwealth (New York, 1973), 413. See e.g. the warnings against it in al-Sarraj, 417ff.

A TURKO- ISLAMIC

MIRROR

FOR

PRINCES

159

mercy! Nay, depart from other men with your hearts, but join with them for well-being and for fulfilling the obligations of Muslims. 1 In t h e Risala

of a l - Q u s h a y r i t h e r e is a Bab

al-xalwa

wa-l-'uzla

in

which the sayings of the Sufi masters on the subject are gathered. At the end of the chapter al-Qushayri cites the following: When God desires to transport a servant (i.e. a man) from the ignominy of disobedience to the nobility of obedience He keeps him company when he is alone, and enriches him with the wealth of contentment, and causes him to see his own imperfections. Whoever has been granted this has been granted the good of this world and of the next. 2

It is in al-Ghazali's "Book of the Etiquette of Solitude" (Kitab Adab al-'Uzla, book 16 of the Ihya lUlum al-Dln), that we find a sustained and systematic treatment of the entire subject. 3 Al-Ghazali begins his discussion by listing the Sufi masters who preferred 'uzla (solitude, withdrawal from the world) over muxalata ("mixing"). And he says that most of the "followers" (including al-Shafi'i and Ahmad ibn Hanbal) approved of muxalata, the multiplying of acquaintances and companions, associating with other believers in mutual love, and helping them in matters of religion. He then discusses the authorities, especially hadiths, which the two sides appeal to. Among the sayings and anecdotes in favor of 'uzla is the following: "Malik ibn Anas used to attend funerals and visit the sick and pay his respects to his fellows; then he stopped doing these things one by one, until he abandoned them all, and he used to say 'It is impossible for a man to keep giving excuses.' " Al-Ghazali cites such hadiths as this: "Whoever departs from the community by a span puts off the lasso of Islam from his neck"; but he interprets this as referring to the community of belief and allegiance to the imam. And: "Whoever cuts off relations with his fellows for more than three days will enter the Fire"; which he interprets as referring to anger with people which a man persists in so that he does not talk to them or greet them or have anything to do with them. And the hadith concerinng a man who went to the mountains to practice his devotions, and was brought to the Prophet, who said, "None of you should do this; to be patient in one of the settlements (mawatin) of Islam is better for you than worshipping alone for forty years"; but al-Ghazali interprets this as referring to the man's abandoning of the jihad despite its urgency at the beginning of Islam. Also the hadith: "Satan is the wolf of mankind like the wolf of sheep l

Al-Wasaya, ed. 'Abd al-Qadir Ahmad 'Ata (Cairo, n.d.), 65. Al-Muhasibi died in 243/857. Al-Risdla al-Qushayriyya, 2 vols. (Cairo, 1385/1965), 1:276. 3 (Cairo, 1387/1967), 3:282-311. 2

160

FROM M A H M U D K A § G A R i TO E V L i Y A

CELEBi

who takes the one that strays or wanders off; avoid the mountain gulleys, and stick to the communities and the mosques"; according to al-Ghazali this refers to one w h o withdraws f r o m society before having complete religious knowledge. And the same for the traditions about Abraham, the Men of the Cave, and the Prophet himself, who broke relations (i'tizal, hijra) with their respective peoples; these all concern the breaking of relations with infidels after despairing of their conversion; none of these can be used as an argument against 'uzla. On the other side we have, for example, the hadith to the effect that 'Abdallah ibn 'Amir asked the Prophet, "What is salvation?" and the Prophet said, "Stick to your house and hold your tongue and weep for your sin"; but, al-Ghazali says, he did not command all his companions to do this, but only 'Abdallah ibn 'Amir, because he knew, through the light of prophecy, that for this man, sticking to his house was more suitable and safer than mixing with others. "The safety of many individuals," al-Ghazali argues, "lies in 'uzla not in muxalata, just as his safety may lie in his sitting at home rather than going on the jihad, but that does not mean it is better to quit the jihad altogether." ("Safety" means the safety of o n e ' s religion; cf. KB 3338.) Similarly, the hadith, "God loves the man whose piety is hidden" (cf. KB 3919-20) refers to the preference of obscurity over f a m e — "How many a hermit living in solitude is known to all the people, and how many an obscure ' m i x e r ' has no renown at all!" — but this is an objection to something that has no relation to 'uzla per se. The bulk of al-Ghazali's treatment consists in a full discussion of the pros and cons of 'uzla. First he deals with the advantages of 'uzla, under six headings (the first two are "religious" dim, the others "worldly" dunyawi): 1. Being free of distraction for devotion to worship and contemplation. "Only the power of prophecy allows one to combine the mixing with others openly and devoton to God in secret" — though some of the saints have nearly attained this rank. But the majority of men require the help of solitude to achieve it. 2. Escaping from those sins to which "mixing" conduces: slander and gossip, hypocrisy, failure to command the good and forbid the evil (because of the dangers this involves), and the bad habits arising from worldly pursuits and bad company. 3. Escaping from contention and strife. Hadith: "There will come a time when a man's religion will not be safe unless he flees with it from village to village and from mountain peak to mountain peak and from burrow to burrow, like a fox that is hunted." This will be when a man cannot make a living unless he disobeys God; and at that time celibacy will be permitted. According to al-Ghazali, celibacy, in this context, implies 'uzla.

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

161

4. Escaping from the evil of others. Al-Ghazali cites the saying of A b u ' l - d a r d a ' : "Men were once leaves without thorns, today they are thorns without leaves," and remarks that if this was so in his time (at the end of the first century of Islam) it cannot be doubted that later times are even worse. 5. Cutting off covetousness, of others for you and of you for others. 6. Escaping from the dull-witted and the foolish. T h e a d v a n t a g e s of muxalata,

on the other hand, are seven:

1. Instruction and learning. If someone already has religious knowledge, if he is an ' a l i m , then the situation of the present day, according to al-Ghazali, does indeed require him to withdraw from the world if he wishes to keep his religion sound. But "there is no good in the 'uzla of the illiterate and the ignorant." And the acquiring of learning is possible only for those who "mix." But scholars who seek worldly gain and prestige through their scholarship are advised to abandon the world. 2. Being of benefit to others and e n j o y i n g their benefit. He mentions earning money in order to give alms, and gaining merit by ministering to the needs of the Muslims. As long as these are o n e ' s aims, and as long as one lives within the limits of the Law, then "mixing" is better; but this is assuming that o n e ' s withdrawal would consist of nothing but supererogatory praying and bodily exercises. If, however, o n e ' s withdrawal means that his spiritual exercises open the way to continual remembrance and contemplation of God, then this is a superior state. 3. The disciplining of o n e ' s own character and that of others. However, "one who spends all his life disciplining himself is like one who trains a horse all its life but never rides it." 4. Companionship. He cites Ibn 'Abbas: "Were it not for the fear of the whispering of Satan I would not associate with others" (lawla maxafat al-waswas lam ujalis). 5. Gaining and giving religious merit ( t a w d b ) (cf. KB 3236 f.) through attending funerals and visiting the sick and being present at the rites of the two great festivals and at the Friday communal prayer, etc. Al-Ghazali says that one must weigh the merit of such "mixing" against the disadvantages that he has mentioned, and choose accordingly. There was a group of early Muslims, such as Malik (see above), who did abandon these things and w h o left their houses only for the Friday prayer and to visit graveyards; "and some of them even left inhabited towns ( a m s a r ) altogether and withdrew to the summits of mountains, in order to devote themselves to worship and to escape distractions." 6. Humility. "For there are those who choose 'uzla out of pride ... or out of fear that their faults will be apparent if they ' m i x ' and people will not believe in their austerities ( z u h d ) and their devotions.... The way to recognize this sort is that they love to be visited but not to visit, and they rejoice when the common folk or sultans gather at their door or in their path and kiss their hand to get the blessing...." 7. Experience. "There is no value to the 'uzla of one whose character has not been tried and tested by vicissitudes."

162

F R O M M A H M U D K A § G A RI T O E V L I Y A

gELEBI

Al-Ghazali's conclusion is that one cannot make an absolute judgment concerning 'uzla: the advantages or disadvantages differ according to individual temperament and circumstances. Although the weight of argument seems to favor the second position, al-Ghazali very fair-mindedly concludes that the choice of life styles has to be determined on an individual basis. At any rate, the solitary life, so full of dangers and drawbacks to the uninitiated, is definitely not suitable for the vast majority of men; although in its ideal form of real communion with God, it is much to be desired. It is no accident that this book of Ghazali's Ihya' is structured almost as a debate. If we look at the arguments on each side, we find that all of the six points in favor of solitude are negative, while the seven points in favor of mixing are all positive; but in each of the latter cases, Ghazali shows how the argument might be turned around in favor of the truly learned and pious man, one who has already acquired these benefits and who then chooses to embark on the life of solitary devotion. Certainly, Ghazali's treatment of the subject reflects actual discussions that took place among Muslim intellectuals, and bears witness to the wealth of earlier literature on the subject. It is the nearest parallel in Islamic literature to the debate between Wide Awake and Highly Praised in Kutadgu Bilig.1 The similarities between the arguments and conclusions are striking. It might also be noted that among the Persian mirrors for princes, it is Ghazali's Book of Counsel for Kings that most closely resembles Kutadgu Bilig in tone and in moral content.

Analysis and Interpretation

of the Work

Y u s u f ' s originality did not consist merely in adapting Turkish language and Turkish royal traditions to the genre of Islamic mirrors for princes. He also made a highly original contribution to that genre. He dramatized the issues in the form of dialogues set within a frame story; he gave the main characters significant allegorical names; and he attempted, quite successfully, to incorporate Sufi asceticism as an opposing, and ultimately complementary, ideal to the prevailing community and statecraft ethics. (In this last point he may be considered a precursor of al-Ghazali.) The work divides naturally into two parts, the fulcrum occurring at line 3120, toward the end of chapter 38.

^For a later reflection of the issue in Islamic culture, see D. Margoliouth, trans., "Devil's Delusion of Ibn al-Jauzi: Account of the way wherein the devil deludes ascetics and devotees" (section 9 of Talhis I bits), Islamic Culture 10 (1936), 339ff.

A TURKO

ISLAMIC

MIRROR

FOR

PRINCES

163

The first part is in the traditional mirror-for-princes mold, and centers on the characters of the king, Rising Sun, and the vizier, Full Moon. In the opening chapters, Yusuf sings the praises of God, the Prophet, and the Prophet's companions. Then (chap. 4) there is an ode to spring conjoined with an encomium of the poet's patron — a traditional feature of Islamic eulogistic poetry 1 — and a discourse on the heavenly bodies (chap. 5). Along with some short sections on the merits of wisdom, intellect, speech, and virtue, the author presents his apology (chap. 8), explains the title of the book (chap. 11), and introduces the king, Rising Sun (Chap. 12, line 405). The king longs for a good vizier, and his trusty chamberlain presents to him Full Moon, a young man of many virtues who has come to the court from far away in order to serve the king because of the king's reputation for justice. There now occur (chaps. 15-18) two scenes in the nature of tableaux vivants, in which Full Moon demonstrates that he is "Fortune", and the king demonstrates that he is "Justice". Some conversations follow concerning uprightness and virtue, and the benefits of speech. Finally the king appoints Full Moon as vizier (line 1036). Then (chap. 20) Full Moon himself suffers an unkind turn of furtune. He falls subject to an illness which cannot be cured, despite all the efforts of the physicians (lines 1057 f.). We are now (chap. 21) introduced to Highly Praised, Full Moon's young son. The father admonishes the boy, and also writes a testamentary letter to the king (chap. 23), in the course of which he commends his son to the king's care (lines 1484 f.). After Full Moon's death, the king adopts Highly Praised (line 1552). Some time later the king, missing Full Moon, recalls Highly Praised, and summons him into his presence. After sounding him out with tests of his wit, he decides to foster the lad; and Highly Praised soon becomes chief counselor to the king. He is especially eloquent on the subject of wisdom and intellect, and in his portrayal of Intellect (chap. 27) seems to be describing himself. The remainder of the first part is given over to the standard mirror-forprinces topics. In response to the king's queries, Highly Praised describes the qualities and duties of the various courtiers (chaps. 28-37): prince, vizier, commander, chamberlain, gatekeeper, envoy, secretary, treasurer, cook, and cupbearer. Finally (chap. 38) he explains the obligations of the prince toward his subjects. Since the king follows Highly Praised's counsels, and administers the realm with justice, the country prospers and the people are rich and happy (3102-8).

^See R. D a n k o f f , " O n nature in Karakhanid literature," ITS 4 (1980), 27-35.

164

F R O M M A H M U D K A§ G A R i TO E V L i Y A

gELEBI

The second part of the work introduces a new theme: the conflict between the political ideals of the community and the religious conscience of the individual. This conflict is dramatized in the form of the debate between Highly Praised and his "brother" Wide Awake. The king, while more than satisfied with Highly Praised, longs for one more excellent counselor, and asks Highly Praised to recommend one of his kinsmen. Highly Praised knows of such a one, "a hundred times more virtuous than h i m s e l f ' (line 3150), named Wide Awake. But the man is a recluse, living alone on a mountain, and Highly Praised is not sure he would come to serve at court. The king sends Highly Praised to fetch him, accompanied by a letter (chap. 39), and makes clear that it is a command (line 3277). Highly Praised carries out his mission. But Wide Awake is reluctant to abandon his contemplative life of religious devotion for the active life of state affairs. In the course of the debate, he rails against the world's faults (chap. 42) and insists upon the need for devoting one's life to God and seeking one's bliss in the next world. Highly Praised counters this (chap. 43) with the argument that the other world is won through this world, and that Wide Awake must devote himself to good deeds in order to gain religious merit. After considering the matter, Wide A w a k e eventually refuses to obey, explaining his position, at the other's insistence, in a letter to the king (chap. 44). Displeased that his will has been thwarted and still suffering an inordinate desire to avail himself of the holy m a n ' s services, the king sends off Highly Praised a second time, armed with a letter that is couched in stronger language than the first (chap. 45). This time Wide Awake pleads that he is ignorant in the ways of the world, and especially in royal custom and protocol. And so Highly Praised instructs him at some length (chaps. 47-61) on the proper manner of serving the prince and of conducting oneself with the various classes of society: courtiers and commoners, scholars, physicians, diviners, d r e a m interpreters, astrologers, poets, f a r m e r s , m e r c h a n t s , stockbreeders, craftsmen, and beggars. To this he adds other typically mirrorfor-princes topics (chaps. 62-66): choosing a wife and raising children, managing the household, and behaving as host and as guest. After this long digression on the proprieties of life in society, Wide Awake is still adamant (chap. 67), insisting that he cannot serve both God and the king. At this point (line 4874), Highly Praised has a change of heart; and when he returns to the court and reports the conversation, the king too realizes the superiority of Wide A w a k e ' s position and the injustice of his own. But though he has abandoned the plan to make Wide Awake serve him, he still

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

165

harbors the desire to see the holy man at least once. So Highly Praised goes to him a third time, this time without a letter (chap. 68). When Wide Awake hears the request, he agrees immediately; for "a Muslim is a brother to his Muslim brother, and visiting is proper" (line 4991). He comes to the court, but warily, at night; and assures his own safety by having the king greet him first before he returns the greeting (chap. 70). Though Wide Awake disclaims his own virtue and piety, the king insists that he give him his counsel. This he does (chap. 71), stressing religious matters (though not exclusively), and then returns to his mountain, leaving Highly Praised to reconcile the king's qualms about raising troops and to explain the duties of the prince and the mutual obligations of ruler and subjects (chap. 72). The king governs well, and the realm prospers. Highly Praised, however, is feeling his age, and conceives the intention to repent his past life and to purify his soul in preparation for death (chap. 73). Bui first he goes to Wide Awake to consult with him about it. Though lauding his good intentions, Wide Awake counsels Highly Praised to resume his station at court, on the grounds of the great benefit to the state that he is providing, and of his obligation to the king who adopted and raised him when his own father died (chaps. 74-75). Highly Praised cheerfully accepts this advice. Returning to court, he devotes himself to state affairs, while extracting from the king a promise not to heed any slanders against him (lines 5871 f.). In the final episode of the work, Wide Awake, on his deathbed, sends for his "brother" Highly Praised. The latter says that sickness is atonement for one's sins, and berates the former for expecting death. But Wide Awake has had a portentous dream, which he relates (chap. 78) after listening to the wise statesman's disquisition on dream interpretation. While Highly Praised pedantically tries to force an auspicious meaning onto the dream, Wide Awake demurs, insisting that it is a sign from God that he should prepare for death (chaps. 79-80). After returning home, Highly Praised learns of Wide Awake's death from the latter's disciple, named Testament (chap. 82). Wide Awake has left behind his worldly possessions, consisting of a staff and a bowl. 1 Highly Praised offers one of these to the king, as a memorial; and the king chooses the staff (chap. 85, line 6325). The book ends with some reflections on the corruption of time (lines 6451 f.) and a final ode in which the author sounds once again the themes of wisdom and of turning away from the world toward God (6605-45).

^The staff (Ar. 'asä) and leathern water-bowl (Ar. rakwa) were requisite for the wandering dervish; see Schimmel, 235.

166

F R O M M A H M U D K A § G AR I T O E V L I Y A

gELEBi

On the allegorical level, the "plot" of the work can be interpreted as follows. Justice, to be exercised properly, requires Fortune, Wisdom, and "the E n d " — what al-Ghazali calls "knowledge of the last breath" 1 or what we might call "religion." King Rising Sun, who stands for Justice, first enjoys the services of Full Moon, or Fortune. But it is Fortune's nature to be unreliable: Full Moon dies. Next the king finds Highly Praised, or Wisdom (more strictly: Intellect) who turns out to be his most steadfast servant. Still the king is unsatisfied: he also desires the counsel of Wide Awake, or "the End." But wordly and otherworldly concerns are not easily reconcilable. And at the end, Wide Awake too goes to his reward, leaving "Justice" with "Wisdom" alone left to help him administer the realm. Besides the four main characters, there are three minor ones, also provided with significant names. When Full Moon arrives at the capital, he is first helped by a man named Much Desired (Tk. Kiisemi§: line 502). He seems to reflect nothing more than Full Moon's desire to serve the king. Much Desired carries out this function by introducing him to the king's chamberlain, whose name is Manly (Tk. Ersig: line 505). The chamberlain performs his duty properly. Recognizing Full M o o n ' s excellent qualities, and knowing that the king is seeking such a man, he arranges an audience, and all goes well. Far more significant is the third minor character, Wide A w a k e ' s disciple, Testament (Tk. Kumaru). "Disciple" translates Ar. murid (line 6286), the specific term for the disciple of a shaikh on the Sufi path. Unfortunately, the text is defective at this point in the work, and there are indications that the author left it incomplete (see note 190 to the translation). But the significance of the name is clear: Wide Awake has left behind as a testament, someone to perpetuate the religious values which he represented. The Turkish word kumaru has a broad range of meaning, covering the English terms "legacy, inheritance, testament, memorial"; and it is the equivalent of Arabic wasiyya.2 When Full Moon dies he gives alms of gold and silver to the poor, and "many testamentary memorials (kumaru)" to his friends (or kinsmen; line 1150, also 1517). His dying discourse of admonition to his son he calls "a testament (kumaru)" (1341); and he also bequeaths a "testamentary letter" (kumaru bitig) to the king (chap. 23). Of course he also leaves behind his son Highly Praised, whom he calls "my life's blood" (lit.

1

Ghazali 's Book of Counsel for Kings , 38. N o t e hadith: "When a son of Adam dies, his activity ceases, except for three things; a permanent bequest, and knowledge by which men are benefited, and pious sons who invoke blessings on him after his death" (J. Stephenson, The First book of the Hadiqatu'l-Haqiqat ... Sand'i, Calcutta 1910, p. xxx). 2

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

167

"my liver's fire," bagrim oti, line 1482), and whom he commends to God (bayatka tutuzdum, 1483) and to the king's care. As for Wide Awake, he too leaves behind his words of admonition (kereklig tutuzdum sanga barqa soz, 6184); and his staff and bowl as a "memorial testament" (kumaru, 6322; vasiyyet, 6321), of which the king takes the staff "that it be a source of good fortune (kutadsu) to me"; as well as Testament, his disciple. What is prized here is the legacy of wisdom and virtue, whether embodied in words or in tangible objects (or, in the case of Highly Praised and of Testament, in persons). The author, smilarly, leaves behind Kutadgu Bilig as an eternal legacy (113, 6506). As a high court official, disillusioned in his old age with the world and the times (see the three odes at the end of the work), and leaning toward the solitary life (6570 f.), Yusuf suffered within himself the conflicts which he portrayed in his work, so that in the course of writing a mirror for princes he broke the traditional boundaries of that genre. Especially in the second part, in the confrontation and reconciliation of the statesman Highly Praised and the Sufi ascetic Wide Awake, he achieved a dramatic portrayal quite unique in Islamic literature. The Sufi and the statesman stand at opposite poles in classical Islam, and their confrontation is rich in dramatic possibilities. It has to be admitted that Kutadgu Bilig hardly realizes the category of drama — even less so than a Platonic dialogue. It is too static, too formal. On the other hand, the dramatic tension of the situation, despite the dry allegorical overlay, is clearly there beneath the surface: the imperious demand of the king to be served, against the pious ascetic's wish to serve only God; the brother's longing that his brother rejoin his kin and share his virtues with the community, against the friend-of-God's longing to share his life with God alone; the demands of society to fulfill communal and familial obligations, against the demands of the spiritual self to renounce worldly ties and to concentrate one's being on the afterlife. Built upon these opposing human types are opposing world views. To highly Praised, that God made the good things of this world permissible (Ar. halal) means that we ought to enjoy them; that God imposed laws regulating interpersonal affairs means that we ought to live a communal life in order to carry out these impositions fully; that God created heaven and hell means that He intended man to face the peril of sin and to face up to the duty of right living in the presence of temptation. Also — although this is not made explicit — that men need one another, that man is a political animal, means that living alone is unwise, unhealthy, and ultimately unhuman.

168

F R O M M A H M U D K A § G A R I TO E V L I Y A

CELEBl

To Wide Awake, on the other hand, God made this world the testingground for the next, a prison for the sons of Adam; but also the stage on which man has the opportunity to purify himself and thereby to earn a place of honor after he dies — and, possibly, to draw close to his Creator in this life. All else is vanity. In Kutadgu Bilig the conflict is resolved in a satisfactory way within the terms of the dialogue. The ascetic urges the statesman not to renounce his obligations to the king and to the people; while the statesman accepts that the ascetic is answering a higher calling. Wide Awake, far from being a danger to the world's governance (as some scholars have interpreted his role in the work), provides a leaven of otherworldly goals and ideals without which the ruler's life would be vain. The ultimate reconciliation of the "brothers," in the king's presence, demonstrates one of the deep-rooted themes of the IranoIslamic statecraft tradition: that just sovereignty and right religion are twins, born of the same womb, and cannot be separated (see above, "Mirrors for Princes"). In the typical Islamic division of human concerns between din ("religion") and dunya ("the world"), din replaces axira ("the otherworld, the hereafter") which in the Qur'an is opposed to dunya. With the demotion of dunya in Islam (Qur'an 2:86/80, 87:16), and with the identification of dunya with dawla ("temporal rale, royal authority and the imperial fortune") — an identification that was perhaps facilitated by the multivalency of the corresponding Persian word, jahan — there was a concomitant tendency to demote dawla. Historically this occurred when Muslim pietists turned away from the Umayyad dynasty in the seventh and eighth centuries as illegitimate rulers. A significant function, then, of the mirror-for-princes literature in Islam was to raise the moral stature of dawla to its original high position, as the "twin-brother" of din. In the eleventh century, when the caliph no longer had any political authority, and the sultan had no religious sanction for his rule, Sunni authors like al-Ghazali and Yusuf Khass Hajib found it important to reconcile din with dawla/dunya. The outstanding characteristic of Kutadgu Bilig is the interweaving of temporal, political, and social themes, drawn mainly from the mirror-forprinces tradition, with spiritual themes connected with the salvation of the individual soul, drawn mainly from the Sufi-ascetic tradition. Even W i d e Awake, who insists that religion and the world are irreconcilable (5311 f.; see above under "Qur'an and Hadith"), is not averse to mixing worldly wisdom along with his doomsday preaching (see above under "Sufism"). And even the king, the "world-conqueror," bemoans the emptiness of mundane existence, and has words of praise for the ascetic life (6379-86).

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

169

Although Wide Awake wins the debate, 1 Highly Praised is the hero of the story. The king, while he accepts Wide Awake's counsel, does not thereby abdicate; on the contrary, he continues to rule. And Highly Praised continues to serve him until the end of their lives, so that the realm prospers and the people enjoy good fortune (6424) — and the author breaks in to say that this sort of man is not a man at all but an angel (6426). The upshot is that, though there is debate and conflict, neither side disavows the other. Wide Awake recognizes worldly wisdom in its place, just as the king and Highly Praised recognize the religious values that Wide Awake represents. As Yusuf states at the end of the work (6498), "Here is the path of Religion and the path of the World." Religion and the World may be in conflict, but in the final analysis one needs them both, and Kutadgu Bilig is intended to prepare the prince to succeed in both endeavors ("both worlds"). Still, the lesson of Wide Awake's death, like that of Full Moon before him, is clear. The prince cannot depend on Religion to run his realm, any more than he could depend on Fortune. Justice and Wisdom are the pillars of the state.

Text and

Translation

The translation is based on Rahmeti Arat's critical edition of the text. 2 This in turn was based on the three extant manuscripts, which have been published in facsimile. 3 In Appendix 3 I list all the alternate readings I have adopted which make a difference in terms of the meaning of the text. 4 Quite a few verses, which Arat includes in his edition but which must be considered spurious, are translated in the notes to the text. The Cairo ms. (designated C) includes toward the end of the work (6521-6604) two odes not found in the other mss. These are undoubtedly the work of Yusuf, but they do not relate directly to Kutadgu Bilig, and so are consigned in the present translation to Appendix 1. A third ode, at the very

This is quite clear (see 5397f.); Bombaci is wrong then when he states (Kopriilu Armagani, 73) that "after a vehement debate there is no indication as to which side is right." Nor is Bombaci correct in saying that the opposition involved is between "faith" and "works" — these are theological categories which do not apply here. Rather, as we have seen, the opposition is between solitary devotion ('uzla) and social commitment (muxalata); or, on another level, between Religion and the World. 2

Kutadgu Bilig I: Metin (Istanbul, 1947). Note also Arat's translation into modern Turkish: Kutadgu Bilig II: Terciime (Ankara, 1959); second edition: Ceviri (Ankara, 1974). ^Kutadgu Bilig Tipkibasim: Viyana Niishasi (Istanbul, 1942), Fergana Niishasi (1942), Misir Niishasi (1943). 4 F o r further discussion, see R. Dankoff, "Textual problems in Kutadgu Bilig," ITS 3 (1979), 8999.

170

FROM

MAHMUD

KA§GARt

TO

EVLiYA

QELEBi

end (6605-45), is found in the Vienna ms. as well (designated A). [The Fergana ms. (designated B) unfortunately breaks off at line 6232, well before the end of the poem.] This third ode resumes the wisdom and antiworld themes of the work, as well as referring to its composition (6623 f.), and so is included in the present translation in the body of the text. The three odes considered together are distinguished from the rest of the text by several characteristics: 1. Rhyme. They follow the qasidah form of monorhyme rather than the masnavi form of rhymed couplets. 2. Meter. The first two odes, but not the third, employ the acatalectic variety of mutaqarib rather than the catalectic variety found in the main text. Thus: Main Text and Third Ode U U U U— First Two Odes U U U U 3. Morphology and lexicon. Certain features are not found in the main text: the cohortative in -ayi rather than -ayin, most notably in the rhyme position of Ode II: 1 the phrase keyik tagi (6570, 6617), here translated "wild mustangs" (but note tagi at 5375); the word kopek for "dog" (6601, translated "cur") rather than it (e.g. 5379, 6194, 6364-66, usually a hunting dog). On the other hand, several features are common to the odes and to the main text, including the bulk of the vocabulary and other linguistic features, and the sentiments expressed. A "lament on his old age" (the topic of Ode I) is also found toward the end of the introductory chapters of the work (lines 36375); another is put in the mouth of Full Moon on his deathbed (1098 f.); and another in the mouth of Highly Praised when he forms the intention to repent his past life (5637-43). A "complaint on the times" (the topic of Ode II) is found at the end of the main narrative just before the conclusion (6451 -94). Thus, there is no reason to doubt that these two odes are by Yusuf himself. It is impossible to state, however, whether he intended them to be included in the text of Kutadgu Bilig, or whether the copyist of the Cairo ms. (or of its prototype) was alone responsible for that decision. Two other additions to the text are definitely by later authors, and so are consigned to Appendix 2. These are the Verse Prologue (77 lines), written perhaps a century after Kutadgu Bilig, and the Prose Prologue, which is a summary of the Verse Prologue. (In the extant mss. the Prose Prologue comes first, then the Verse Prologue [omitted in ms. B]; then the list of chapter headings; and finally the text of the work.) The author of the Verse Prologue understood the work to be mirror for princes in the traditional sense. The principle that Wide Awake represents is said to be Contentment (line 66: kana'at ve 'afyet; 71; kana'at). This must depend on a misreading of line 357 in the text, where Yusuf specifies ^Ms. B occasionally has -ayi (e.g. 3759, 3996). Beside 1033, 2033[B and CI, 4172[B and CI).

-ayin

we also find

-ay

(e.g. 560[B],

A TURKO-ISLAMIC

MIRROR

FOR

PRINCES

171

that Wide Awake stands for Man's Last End ('akibet). Even here, one of the mss. (B) has Well-being ('afiyet). Similarly, the author of the Verse Prologue must have read ' a f i y e t , then expanded it to the hendiadys kana'at ve 'afyet (shortened for the meter). The Prose Prologue retains only kana'at (line 30). Another, less serious, alteration appears in the name of the author's patron, given in the text (line 88) as Tavgag Ulug Bugra Khan. The Verse Prologue (line 60) has Tavgag Kara Bugra Khanlar Khani; the Prose Prologue (line 25) simply Tavgag Bugra Khan. The author of the Verse Prologue also added the Arabic or Persian equivalents of Turkish abstract nouns for the three other principles represented by the other major characters (lines 64 f.) It might be noted, finally, that in two verses (70, 77) of the Verse Prologue there is a weakness in the meter. "A translator is to be like his author; it is not his business to excel him," said Dr. Johnson. I have mainly aimed at coherence and readability. The logic of prose is not the same as that of verse; and the virtues of modern English prose are even less those of medieval Turkish verse. Thus, while not omitting anything germane to the meaning that Yusuf was trying to convey, I have not hesitated to omit inconsequential line-fillers; to transpose words and phrases and, occasionally, entire lines; to abbreviate where Yusuf is wordy and to expand where Yusuf is obscure. In doing so, I have eschewed pedantic parentheses and brackets, which would only serve to distract the reader from the sense. It would serve no useful purpose, for example, if the translation of line 2715 looked like this: "When the sword drips (red) with blood, the prince conquers territory; and when the pen drips black (with ink), he acquires gold (lit. gold comes)." The text implies the additional elements, and my judgment was that the reader should have what is implied rather than possibly be left in the dark. Similarly, we read several times that man's final lot of the world is two pieces of cloth; at line 1238, where this comes up first, I add, without apology (i.e. parentheses), "for his shroud". In rendering the quatrains into verse I mainly aimed to distinguish them f r o m the surrounding text, although in the original they are distinguished only by the rhyme scheme (uniformly aaba). I have necessarily taken rather greater liberties in the translation of these quatrains, though striving always to remain true to the original. In cases where the meaning of Yusuf's verse depends on word play, this is indicated in the body of the translation. Otherwise, word plays are indicated in the notes to the translation. 1 Explanatory notes are kept to a minimum, consisting mainly of Qur'anic references and other points which would be familiar to every Muslim reader though possibly obscure to a Westerner.

' T w o frequent word plays are not indicated: on ilig "king," elig "hand" (e.g. 598, 1607, 2181, 2269, 2757, 2789, etc.), and koni "upright, just," kuni "his day, his sun" (2186, 2512, 2757, 2789, etc.).

11. TEXTUAL PROBLEMS IN

KUTADGU

BILIG

In the course of preparing an English translation of Kutadgu Bilig1 I have had to wrestle with a good number of philological problems. Some of these problems relate to the critical edition of the text, established by Re§id Rahmeti Arat. 2 I have naturally used this edition as the basis for my translation, and I would be remiss not to acknowledge my debt. Arat's herculean labors, combined with his scholarly exactitude and his sensitivity of judgment, provided scholars with a text they could use and rely upon. It has long been recognized, nevertheless, that Arat's methodology in establishing the text had certain shortcomings. Also, Middle Turkic philology has advanced somewhat in the thirty years since it was first published. 3 In light of the above, it is perhaps opportune to offer here some of my conclusions and suggestions, based on my study of the text over the past few years. Arat's major failing, in my opinion, was to consider the three extant manuscripts as equally valid witnesses in reconstructing the text. This was contrary to the classical principles of textual criticism. 4 Without establishing any criteria by which to prefer one reading over another, Arat's method was simply to choose the reading he preferred and to list the variants at the bottom of the page. Aside from the unmethodical nature of this procedure, it also made the apparatus criticus extremely cumbersome and redundant. These drawbacks were hardly offset by the advantage of having all the variants before one's eyes on the same page. The relation among the mss. is actually not at all obscure, and Arat recognized it and described it, at least in broad outline (Metin, XXXIIIXXXIX). The oldest ms. is B ("Fergana," 13th cent.?), then C ("Misir," ' I wish to acknowledge grant support for this translation from the National Endowment for the Humanities. 2 Kutadgu Bilig I: Metin (Istanbul: Milli Egitim Basimevi, 1947). 3 O n e important work is Gerard Clauson, An Etymological Dictionary of Pre-ThirteenthCentury-Turkish (Oxford, 1972), cited below as ED. 4 S e e , for example, Paul Maas, Textual Criticism, tr. B. Flower (Oxford, 1958); Ahmed Ate§, "Metin tenkidi hakkinda," Ttirkiyat Mecmuasi VII-VIII (1940-42), Ciiz 1, 253-267. Cf. Ateg's review of Arat's edition in Ttirk Tarih Kurumu Belleten 13 (1949), 157-163, p. 161, parag. 4. The trend of Soviet scholarship on Kutadgu Bilig has been to prefer the Fergana ms. (B = N [Namangan] according to the Soviets) as the basis for reconstruction. See A. Valitova, "Osnovnyje principy publikacii poemy Jusufa Balasagunskogo 'Kutadgu Bilig'," Issledovanija po ujgurskomu jazyku (1970), 144-166, p. 147. The edition by K. Karimov (Tashkent, 1971) is unavailable to me; see the review by E. N. Nadzip in Sovetskaja Tjurkologija (1972), no. 3, 108113.

174

FROM

MAHMUD

K A § G A R i TO EVLiYA

QELEBi

before 1367), then A ("Viyana," 1439). There are innumerable points in the text where A and C agree with each other but not with B (e.g. B omits the verse prologue). A and C, though they form a group, also diverge in many respects (e.g. A omits lines 6304-51 and 6521-6604). Thus, the relation among the mss. can be represented in the following scheme: 1.

Original (1069-70)

2.

Inclusion of verse prologue (12th cent.?)

3.

Inclusion of prose prologue: Archetype

4.

5

B (Omission of verse prologue; 13th cent.?) C (before 1367)

6.

In Arat's judgment, based on cultural considerations, the text at stage 1 was in Uighur script (Metin, XXXII). This notion was effectively refuted by Alessio Bombaci. 1 At any rate, Arat recognized that the archetype, at stage 3, must have been in Arabic script. The inclusion of the prose prologue comes at this stage, it being "a defective and poor summary of the verse prologue" (Metin, XXX). The hyparchetype x is on the same level as B in terms of reconstructing the text, and so is put at stage 4, although the evidence of its witnesses (AC) would support a substantially later date for it (14th cent.?). Everyone agrees that A, in Uighur script, was copied from an exemplar in Arabic script. But the copyist of A was too slavish and ignorant to have performed the various emendations, which evidence independent judgment and poetic sensitivity, that are traceable to A's prototype (these emendations are discussed later in this paper). For this reason we must posit the hyparchetype y at stage 5, on the same level as C. The clear conclusion to be drawn from the above is the following: Regard as spurious the peculiar readings in C or in A; and restore to the archetype whatever B and C, or B and A, agree on. Arat's failure to follow this principle has resulted, first of all, in a somewhat distorted picture of the vocalization and the consonantization of the text. For example, BC generally have -mas for the aorist negative, A has -maz. Arat's edition has -maz throughout, although -mas is clearly the form '"Kutadgu Bilig Hakkinda Bazi Miilahazalar," Fuad Kopriiltt Armagani (Istanbul, 1953), p. 68.

T E X T U A L P R O B L E M S IN KUTADGU

175

BILIG

that ought to be restored to the archetye. It might be objected that Arat was reconstructing not the archetype, stage 3, but the original, stage 1, and that we know the form was -maz in the eleventh century. But this argument has several flaws. Firstly, it is crcular, since much of what we known about eleventh century Turkic is based on Kutadgu Bilig. Secondly, in this case the evidence of the other major eleventh century T urkic text, the Diwan Lughat at-Turk, points in the other direction, since there too the aorist negative is uniformly -mas (Ka§garl even spells this out at fol. 301!). Thirdly, the editor's primary task is to establish the archetype. The task of reconstructing the original is a highly speculative one unless there are clear instances where the archetype was in error. Many of Arat's phonological reconstructions are based on the highly dubious assumption that the original was more archaic than the archetype. And some of the reconstructions go against the evidence not only of BC or BA, but of BCA together. Thus at line 703 we find BA kotki, C koski (similar distribution at 1696-1705, 2073, 2229-2232, etc.). The obvious form to restore is kotki, and this also agrees with the form in Ka§gari (fol. 215). (Note the play on words in the rhyme position in line 3527: kotki - kotki.) Arat's form, kodki, is a hypothetical one, nowhere attested; it is presumably based on the etymon of this word, kodiki, which is found in Uighur texts (see ED, 599). Like considerations would argue that the first forms in the following list are the correct ones (Arat follows the second unless otherwise stated): 794 662 1115 1138, 2989, 4657, etc. 2585 4088, 4090, etc. 4063, 4175, etc. 5562a 6219 721 1180 3626 4496 1948, 5252, 5481

BC korkiit-, berkiitBC yuvhnur A B C umun^siz ABC munung A B C seviigli BC seviinse BC tapunsa BC orta BC yaldradi A B C bek ABC yumki BC ohsar BC epgi A B iisteg

5261, 5306

BC tamug

A korkit-, berkitA boglunur

A A A A

sevinse tapinsa ortu yalrudi

A ok§ar A evci C usteng (but cf. 1796, 2269, 2357, 3150 A B C usteng) A tamu (cf. 3654, 4530 ABC tamug; 292, 917 ABC tamu)

Arat: Arat: Arat: Arat:

yuvlunur umuujsiz muning sevigli

Arat: yildradi Arat: berk Arat: yumgi

176

FROM

MAHMUD

K A § G A R I TO E V L l Y A

5556

BC mezgit

4551

ABC -ga

3428 982 1050 2728 3518 4215 6149 971 994 1018 1022 4701 6180 1093 4730 1232 1337 1335 2734

BC tilide BC yasi BC tiikeli BC asgi BA 'aybi BC umdusi BC köti A B C tilka BC tilka A B C tilig BC tangrig BA lsinmak BC yene ök ABC diinyeke BC diinyake ABC ecelka ABC zinake BC kilikdin BC yagukda

CELEBI

A mcscit (cf. 5486!) Arat: -gay (cf. 4541, 4545, 5595, etc.)

(A-)

A yasin A tiikelin A asgin C'aybin A umdizi A közin

Arat: tilinde

Arat: umdusin Arat: tilke

A tilke A tengri C isinmek A yana ok A tiinyaka

Arat: tilig Arat: tengrig

Arat: Arat: Arat: Arat:

diinyaka diinyaka ecelke zinaka

A kihktin A yagukta

The first two-thirds of the above list, which is by no means complete, illustrates the obscuring of interesting phonological and morphological details, due mainly to Arat's archaizing tendency. The last third reflects rather the influence of the phonology of modern Turkish. In some cases Arat's reconstructed from is impossible metrically. Thus me in lines 809 and 811 (ABC) cannot be read men (also 1260, 1561); 1 §erridin, hayridin at 5062 (BC; A: §erringe, hayringa) cannot be read §errindin, hayrindin; and ornunga at 5738 (BC) cannot be read ornungka (A) — the latter is more strictly correct grammatically, but the former is required for the sake of the meter. Similarly, hazz bezz at 5799 (B; haz sebz) fits the meter, as well as the sense, better than harir (C). In the most significant cases, Arat's reconstructed form is substantially different, and often with quite a different meaning, from the form justified on the basis of the mss. In all the following examples the forms listed first are to be restored to the text:

' See T. Tekin, "Determination of middle-Turkic long vowels through 'arud," Acta Orientalia Hung. 20 (1967), p. 157, n. 7. The vocative particle ay (or ey) must be read i (or e) in lines 89, 91, etc. (on this point see Ate§ in TDKB 13 (1949), pp. 162-3).

TEXTUAL 505 552 757 795 941 1058

1152 1715

PROBLEMS

A B C ati A B C 5züng B A bilse B C sózüng B A tiriglik B C iggil (not in apparatus criticus; see ED, 629) A B ani C teper B tutar

1774

B tatidi (cf. 4169, 5627) B C mengzetü emdi 1869 B C mengzetür (cf. 4673 B C turur 2018 B A tirengü 2285 2302 B C tu§ang A B C tile 2339b 2352 B C eger B C kóz (not in 2465 appar. crit.) 3154 B C birseng 3209, 3213, 3 3 5 0 B C kap 3348 B C giybet 3350 B C yüdürmedim B A ki§i telvesi 3526

3626 3736 3976 4016 4116b 4150 4157 4172 4229 4334 4350

B A tolmaz B 1 C tatig B A dünyani kodsa B C üzsemiz B C sózüng A B C tilenseng B C kilgu begke B C tapay tip B C kilgay B C kórü B C serme irme

4442

B C azar (see ED, 289)

I N KUTADGU

BILIG

177

Arat: ayi Arat: orting C kilsa A yusiim (C-) A ek§ek

C muni A ider

A tetigü (?) A mengzer emdi bu A mengzeyür) A kilur C tiringü A teseng A anggar A yüz (A-) (A-) (A-) (A-) C ki§ide ha§i

C A C A A

todmaz ya§ig 'ukbi kolsa ay tüzemiz sózüg

Arat: yuziing Arat: tirilgil

Arat: iter (but see Arat's translation!) C (Arat) yetildi

Arat: bo§ang Arat: bile Arat: angar

Arat: bir sen Arat: ka Arat: 'aybet Arat: todurmadim [problem of rhyme; perhaps read usol it esi in second hemistich]

Arat: ursamiz Arat: tilese

A begke kilgu A tapayin A kedgey A óyü (A -)

A erez

Arat: kitgey Arat: irme serme (note confusion in appar. crit.) Arat: erej

'See Kutadgu Bilig: Tipkibasim II: Fergana Niishasi (Istanbul, 1943), Yardimci §izgele, p. 17, 272:1.

178

FROM

MAHMUD

BC f a n g 1 BC §eksiz BC isetiir B ere§ (secondary form of erej!) B i§ BC ash ABC kut BA kiiçung BC nenging BC to§ekin tiiriip BC muni ... am BC i§ sening BA yakin (allows play on words in rhyme position; cf. 605) BC halkka BA sezig ABC karim (see ED, 660) BC ahir B erentiz (cf. 139,

4505 4616 4623 4907 5186 5190 5286 5410 5452 5545 5595b 5620

5768 6025 6063 6126 6221

5676) Prose

KA § G A R Í T O E V L Í Y A

prologue

BC -

A A A A

arig terkin isitür eri§ C erij

C iz A orni Arat: kod C kiiçi A nengig A tö§ektin turup A am ... mum A i§lering C b/y-?

ÇELEBÎ

Arat: karig [error for ii§iitiir?] Arat: erej

Arat: bakin

A ilke C kezig Arat: kirim A bir kiin A erenyez C eren kir-

Arat: erentir

A 'asiz

Arat: 'aziz

29

All of these restorations follow the principle established above, giving preference to BC over A, or to BA over C. On the same basis, those complete verses found in either A or C but not in the other two mss. must be considered spurious. This applies to lines 726-7 and 5771, found in C only but not in BA; and to the following lines in A only and not in BC: 478, 1460, 1875, 2673, 3534, 3647, 3806, 3921, 4964, 4968, 5051, 5061, 5281, 5342, 5466, 5503, 5507-8, 5644, 5672, 5712. By the same token, the headings at 670-1, 672-3, etc. ought to be omitted, since they occur in A only. Arat added a heading for every change of speaker, even when there was no basis for this in the mss. (e.g. at 1581-2, 1916-7, etc.); it seems to me most unlikely that this was the author's intention. Finally, the heading after 4354 ought to go after 4355, and the one after 5631 ought to go after 5633, following BC as against A. Despite the basic principle established above, which we have followed to this point, it cannot be excluded that the anomalous ms. (A or C against the other two) might occasionally have a better reading. This can occur in one of two ways: ÏCf. Atabetü '1-Hakayik, ed. R. R. Arat (Istanbul, 1951), line 515.

TEXTUAL

PROBLEMS

IN

KUTADGU

B1LIG

179

(1) The archetype had the correct form, but at two independent stages in the transmission of the text two scribes made the same error. This appears to have happened with verse 1601, which can be reconstructed as follows: yir opti kor 6gdulmi§ aydi ilel ilig koz yarutsa manga buldum el "Ogdiilmi§ kissed the ground and said: Verily, when the king's eyes shine upon me, then have I gained success." The rhyme position contains two unusual forms which the scribes must have found troublesome. Ilel, according to Ka§gari (fol. 52) is "a particle meaning 'yes' used in the royal Khaqan dialect as a reply to emirs and kings." It is found in similar contexts in Kutadgu Bilig (but always at the beginning of the verse) at 3281, 3946, 5027 and 5668. It must have been in the archetype, since it is preserved in C, and was certainly an archaic word by the time C was copied. In the transmission of both B and A it was replaced by ilig, a word that is very common in this position (e.g. 1553!). As for el, I take this to be the rare (at this stage of Turkic) secondary form of elig, found again at verse 3125: elim "my hand." It was natural to change this to elig and then to adjust the meter by shortening manga to ma. This is what we find in A; whereas in B we find the incredible phrase: mangar boldi ig "then have I become ill." 1 (Incidentally, the phrase elig bul- meaning "to gain success, to prevail," a caique on Persian dast yaftan, is quite common in Kutadgu Bilig: cf. 890, 1489, 2330, 4261, 5081, 5758. Possibly at 1601 we ought rather to read boldum il "then I am at peace"?) In this case, we must follow C as against B and A (Arat follows B). (2) The archetype had the incorrect form, but in the course of the transmission of one of the anomalous mss., a scribe restored the correct form. This appears to have happened with verse 60, where in the first hemistich tadu was replaced by tediik by the time of the archetype (probably under the influence of this word in the rhyme position at 58), so that we find tediik in B and tesiik in A; while in C (followed by Arat) we have the correct tadu, restored on the basis of the same word in the second hemistich. In cases like 4139, where A has the obviously correct kilig and BC have the obviously incorrect kdinq\ or in 5527, where C has the correct form konitge as apposed to AB koninge; it is hard to know which of the two processes was at work, since both are plausible. At any rate, in both of these cases I agree with Arat in restoring the anomalous form. In the following cases, however, I am not so sure. The anomalous form, followed by Arat, does appear to give a better reading in each instance; but it is hard to believe ' c f . S. £agatay in Turk Kulttiru 98, p. 109. Her interpretation (ig - yig) is untenable.

180

FROM

MAHMUD

K A § G AR I TO E V L I Y A

gELEBi

that in every case it restores the original (there is little question that the form Arat rejects, although giving a poorer reading, must be restored to the archetype): 1028 2169 4134 5067 5230 5520 5893 6006 6182

BC yavuzi BC uku§lug BC erse sen/bolga (C erse; restore: bolsa?) sen BC bolmadim BC ki§i BC bilig BC tu§i (or, read tii§i?) BC bu tab'i BC sabir kil

A uguzi (parallels agin!) A uvutlug A erse oz/kilgu (Arat: kilga) koz A butmedim A koni A silig A tengi A tadusi A tida tut1

With these examples, and especially the last two, we are confronted with evidence of independent judgment and poetic sensitivity that, as I mentioned above, are characteristic of A ' s prototype. Besides the substitution of words, which as we see above often results in an improvement, we find as further evidence of these qualities: (1) a liberal rearranging of the order of the verses, sometimes to advantage (e.g. reversal of 302 and 303; placing 987 after 984; etc.); (2) the omission of doublet verses (e.g. the one after 5836); and (3) the addition of verses. We saw above that Arat included many of the latter in his edition; and there is no doubt that some are of high quality (e.g. 4964; and 1780-1, replacing doublets in the archetype). It seems to me extremely unlikely that the originator of these happy emendations was the copyist of A himself, who otherwise appears as rather slavish and ignorant, merely copying his text out of Arabic script into Uighur script for his Timurid patron, and committing countless howlers in the process. Rather they must be due to an intelligent and sensitive scribe who, perhaps as early as the thirteenth century, redacted the hyparchetype common to his text and to C (labelled x in the scheme at the beginning of this paper) and produced a new hyparchetype (y) which the scribe of A subsequently mangled in his own way. To the preceding list I would now add the following, where A again appears to have a better reading, but where Arat (correctly!) follows BC: 757 3616 3667

BC eviirmez yiizin BC belgiiliig BC boguz

4892

BC tunerdi

A kagumaz turur (restores a quatrain) A ulgiiliig (cf. 1238, 1420) A yavuz (? - meaning uncertain; cf. 5095) A yarudi

^In the last two instances, A restores a Turkic form as against Arabic (cf. 4616, 5768, 6126 in above list).

TEXTUAL

PROBLEMS

I N KUTADGU

BIL1G

181

It might be said that what A ' s prototype has done in all these cases is to emend the text, just as a modern editor would do. But this is not so. When a medieval scribe substituted one form for another, he was not trying to restore the original text; he was simply replacing what was before his eyes with something more satisfying. 1 In the last cited example above (4892), the context is a poetic description of dawn: 'abir biirkirer teg tiinerdi kalik "The sky darkened (or, was dark) as though sprayed with perfume." The hemistichs immediately preceding and immediately following both speak unequivocally of the sky brightening; and so the substitution of yarudi "brightened" for tiinerdi "darkened" is satisfying, however we may visualize the crepuscular perfume-spray. The modern editor, in this case Arat, judiciously chose to interpret tiiner- here as meaning not "to darken" but "to take on a twilight character" (in his translation: alacalandi).2 Of course there are occasions when emending the text is justified, even required. At 4139, if all three mss. had had kiling, it would still be necessary to emend this to kilig, though in this case, as we saw, A points us to this conclusion. At 3753b all mss. repeat the rhyme word, tapug; Arat's emendation to butug is brilliant. I would recommend the following additional emendations: 94 447 1038 1544 2311 2354 2497 4068 5412

esir (i.e., Ar. atir) B esir A eser C arsikar kör (or, er tiker köz? - cf. 4487) A arsak-a ol C - B er teger kör Arat: er teger köz süzdi (cf. 421, etc.) A B C (Arat) sordi (but see translation!) iring (?) BA erij C - Arat: erej azgin (see ED, 283) C (Arat) azgir B argir A agsun süsi A B C (Arat) sözi (but see Arat's translation!) ugrinda (cf. 2537) BC (Arat) agzmda A ogr-i bar kiil irkin (see ED, 225) A C kök irken B kür tegin tigin beg burnung (?) BA yurnung C (Arat) yurtung

Arat's

Arat:

Finally, there are numerous instances where the weight of textual evidence could support a reading different from Arat's. In what follows, unless otherwise stated, the first form listed is in my opinion preferable, the second form is the one preferred by Arat. (Here X stands for the undotted Arabic letter that could be read as b ~ p, t, n, or y.)

1 S o m e t i m e s he did this because he did not understand the f o r m he saw, it being archaic, or the like. Note 4132: B C te§i "glutton"; A tecik "Persian"! 2

S e e Yusuf Has Hacib, Kutadgu

Bilig II: Tercume,

tr. R. R. Arat (Ankara, 1959).

182

FROM

MAHMUD

KA§GARI

TO

EVLÌYA

gELEBÌ

Verse prologue 57 A munung ki C muXXki Arat : bu tengi Read: munungki 50 A C yigi B iki 86b B tigi A tugi C- 1 121 B telinsüni A tökülsüü C Arat : yalinsuni 171 B Xulmadim A körmedim CArat: tulmadim Read: bulmadim (see ED, 491) 369 öq kek B ögke A ö?lük (cf. 2312: sürdüng?) 554 A tudug C (Arat) tapug B taXug Read: tutug (cf. 1211) 618 B bari A turu CArat: yari 662 B tezginür C yiikülünür A guvrunur Arat: yolunur 911 A yerip B yorip C1286 B §ük CAbük 1390 C tü§ tün (cf. 3605!) A tü§-tin B tü§kün Arat: tü§ün 1408 C engl (cf. 3625) A tingi B (Arat) tili 1424 B C udiXga A udunz-a Arat: uditgay Read: udinga (see ED, 62) 1500 B A koXinga C komnga Read: koyinga (see ED, 631) B XrXu C brbu A birnii Arat: bürnü ED, 367: burnu Read: barnu (from barin- "to gush [of blood, etc.]" in Ka§gari, fol. 338) 1 5 3 2 b A bilgüsüz C B belgüsüz 1541 A öliim-tin C B ölümke Read: ölümdin (also: buldagi, cf. 4837) 1642 C emitti berii A imid-ti naru B Arat) yumitti törü 1646 A bolugi (cf. 1628 boluglug!) C yoriki B (Arat) yüleki 1669 Ayoli C ögi B (Arat) tili 1723 B bu kilga C (Arat, but see translation!) kayir-ga A kiyik-ga 1819 BC Xengediir (C n-?) A yüke-tür Arat: tenegediir Read: nengedür (*nenged-, otherwise unattested, would mean "to gain in wealth" and fits the commercial conceit of the context) 2807 C yiiz agig (so read; play on words; and see A r a t ' s translation!) A B yiizde 51g 3160 B sani C sinl3283 B karardi A C kizardi 3357 B ula bolsa (?, cf. I l l ) C ulayulsa A (Arat) ulanulsa 3552 A C tegürme B tegürge 3603 C A tiler öz tiki (tegi?; cf. 4294) B tilegü am Arat: tiler öz tengi (4487 C A tiler öz tikip B tiler üstikip) (cf. Kä§gari: östik- derived from öz tik-!)

'('f. A. Bombaci, La letteratura turca (Milan: Sansoni-Accademia, 1969), p. 109.

TEXTUAL 3656 3736 3789 3973 4017 4139 4206 4389 4585 6151 6534

6555b 6557

PROBLEMS

I N KUTADGU

183

BILIG

CA kalgay (preserves sense and meter!) B kalmagay C tilemez B bulmayin A (Arat) toplamaz B yaragay A yarugay C (Arat) barugay B isiz kamug CA aggay kapug BXtar AC bolur Arat: batar Read: yutar (?) C yazgu A yasgu B Xargu Arat: targu B se si§ C se11 §i§ A saki§ Arat: se§i§ A kö(?)nik-se CB kinilse Read: kinikse (see ED, 731) A yasim C B yanm Read: yazim (implies an error in protocol; cf. 2497) BC Yüz A yüz Arat: yiiz Read: tüz (C only) galpiladim (?) Arat: cjeliledim Cf. (¿alpak, etc. (ED, 418); also see Brockelmann in Oriens (1949), p. 146. yigitlikte (not in apparatus criticus) Arat: tiriglikte XgXtim Arat: tagittim Read: yagittim

2

In several of the above cases (e.g. 3973) the d i f f e r e n c e between the two main textual traditions (B and C A ) is so great that it is not merely a matter of misreading, but of wholesale substitution, similar to what w e saw with regard to A ' s prototype. Here again, a scribe has replaced w h a t he saw before his e y e s with s o m e t h i n g m o r e to his liking; but it is hard to tell w h i c h is the original and w h i c h the r e p l a c e m e n t . A t 366, f o r e x a m p l e , the a u t h o r , c o m p l a i n i n g of old age, says that " s i x t y " (altmi§: A, Arat [C-]) — or " f i f t y " (elig:

B) — is calling him. And at 3883, 6 g d i i l m i § s a y s that if the k i n g

w a n t s Odgurmi§ once, he wants him "a t h o u s a n d " (ming\ B, Arat) — or " t e n " (on: C,A) — times. C h o o s i n g between the alternatives is often j u s t a matter of taste, and f o r the m o s t part I would agree with A r a t ' s choices. A diligent f u t u r e inquirer, h o w e v e r , m i g h t be a b l e to d e v e l o p s o m e m o r e stringent criteria f o r c h o o s i n g between them. Here are s o m e f u r t h e r e x a m p l e s of this interesting p h e n o m e n o n : 387 1663

A (Arat; C-) B B (Arat) CA

1723 3541

B (Arat) CA CA (Arat) B

yoritti tilig bitidi elig taki bir siliglik kilingi köni ikigü ajunda kutadur küni taki bir mengilig kilingi bütiin ikigii ajunda kutadur kutun küdezdi nengin ked özin sakladi küdezdi özin nengke suklanmadi (A sözlenmet[i]) köngül birse kurtga bolur bir (MSS yir) turi (A toz-i) yaka ying tutar künde kesmez un (A öz-i) köngül birding erse bu kurtga yüzi yakangni tutar künde kitmez sözi

184 4116

FROM

MAHMUD

CA (Arat) B

5659 6167

CA (Arat) B CA (Arat) B

K A § G AR I TO E V L Î Y A

ÇELEBÎ

seningde ulugrak sözin sözlese sözüng (A sözüg) barça kodgu am tinglasa özüngde ulugnung sözin sözleme sözüng yah§i tinglap am kizleme kinge§mes ki§ike katilma yira§ kinge§mes ki§ini ada§ tutma i§ ölüm yetgelir terk oti kil emi ölüm buzgalir kör tiriglik tami

One highly interesting case deserves closer attention. Line 2381 is the second part of a quatrain, of which the first part has the rhyming words tiinleri and kiinleri. Both B and A have sacrificed the quatrain scheme for two couplets; thus: B A

kür alp er sikirtip çerig boym sir yagini bulunlar alur hem esir bu alp er sekirdip çerig targam laçin ku§ kovar teg kalin kargam

In C we find the quatrain scheme preserved; and we also find some alternate readings below the line: C

küvez alp sekirtip gerig sürse bat yaXgay (?) bulun-lar alur hem töker kanlari lagin ku§ kovar tig

(Arat's reconstruction follows C, with the alternate reading in the second hemistich). It appears that C comes closest to preserving the original form, though there may have been confusion already at the stage of the archetype. A possible explanation is that the archetype had bulunlar alur hem toker kanlari; this was preserved in hyparchetype x, which also included the replacement phrase lagin ku§ kovar teg, reflected in A as well as C; and B altered the hemistich to accord with its own couplet rhyme. There is large scope here for speculation. Further study is needed. In conclusion, I append two final lists of emendations to Arat's edition. The first contains those cases where I believe a word should be read differently (Arat's reading is given in parentheses):

TEXTUAL 456 464, 2112 795, 1610, 4611 3160 795 885 1125 1474 1538 2115, 6454 2316 3469 3544 3679 4417 4506 4560 5081 5180 5763 5971 6424a 6615 6625

PROBLEMS

I N KUTADGU

B1L1G

185

birdi (bardi) tuz (tüz) (cf. Kasgari, fol. 42.3) saril- (seril-) saril- (siril-) töküldi (tügüldi) (cf. 195) yol-a (yula) odlur (udlur) yiri (yari) turi (tori) (turi "bitter" means "hated"; cf. 4553) kür (kör) (cf. 57) ögi (1151)1 beklig (beglik) (cf. 436, 484, 554) miini (muni) tü§ (tu§) ongi (öngi) (Also: 4080 [cf. 5288]) tip (tap) umi§ [*um-mi§?] (ömi§) buldi (boldi) gür (kör; but see Arat's translation!) utru (ötrii) (cf. 1188, 1311, 5618) öngdürdiye (öngdürdi ya) (the same word as "east" in the Uighur texts; ED, 181) ed kü (edgü) (cf. 618) büre (bura) (see ED, 355) tiire (tura) (cf. 3811)

öngdürti

The second contains misprints not noted in Metin, pp. 655-6, "Diizeltmeler": 1816 2506 3235 3272b 3390 3433 6569

ögrenirmü (ögrenipmü) almasa (olmasa) kizleyü (kizlegii) ba§i (ki§i) k acari (kagar) angar (anga) vefalig (cefalig)

ADDENDUM After completing the above article in June 1979, I sent a copy of the typescript to Osman Sertkaya for his comments. In a letter dated 24 November, 1979, Dr. Sertkaya generously provided me with the following list, along with this explanation: "Bu listeyi Arat'in orijinal niishasindan gikardim. Arat kendi transkripsiyonlarindan bazilarinin, kar§isindaki §ekiller §eklinde olmasindan §iiphelenmektedir." I give here Dr. Sertkaya's list, followed by my own comments.

^See R. Dankoff in Journal of Turkish Studies/TUrklük

Bilgisi Ara§tirmalari

I (1977), p. 96. n. 3.

186 B 11 B 48 D 10 111 118 133 135 164 236 251 308 354 532 552 767 1073 1255 1276 1573 1652 1737 1747 1748 1750 1890 2115 2354 2402 2426 2496 2741 2785 2866 3226 3283 3628 3705 3709 4022 4044 4073 4080 4193 4299 4381 4386 4955 5237

FROM

MAHMUD KA§GARÍ

tüzi yakin açu saçu bolsu kurimu§ köksün urundi er sak burun öd kiin töri kiini topladi topginga oka ötrü kûvençlik bolur bu kiin tügüldi yetliimedi utru kunuk özüng öziing i§ buldi kör sözi Ol elig emi yitürse bolsa suglur bulumaz i§in tezig/kezig könglün kilma önür tiirçise ornasa öngi tuda§ yara§mazsa yatur kön bolga bol

TO EVLÌYA

töz-i ? yikin ? eçii ? seçii ? bulsu ? kurimi§ köskün/kiiskün? öründi ? arsik ? bk. A ve B törün ? edgün ? turi ? köni? topladi tapginga uk-a ! utru ? kiivençlig ? bulur ? biigiin ? toñuldi ? tidulmadi ? ötrü ? konuk ? uzung ? közüng ? e§ ? boldi ? kür ? süsi ? al ilig ? imi ? yetürse ? bulsa ? soglur bolumaz ? i§in ? tezik/kezik ? könglüfi ? kil ma ? unur ? turasa ? öznese ? ongi ? tüde§ ? yara§masa yetür ? kün bulga ? bul ?

ÇELEBÎ

TEXTUAL PROBLEMS 5482 5524 5596 5650 5675 5682 5706 5721 5942 6029 6307 6358 6359

tilek artatur bolgusi tildemi òngdiin tekip telekge tudas òziin beklig bolur aman òz sefaci

IN KUTADGU

BILIG

187

tileg ? Kafiye artanur ? Kafiye bulgusi ? tildami ? ongdin ? tegip ? tilekge ? tüde§ ? uzun ? beglik bulur ? eman uz ? cefaci Comments

Certain of these corrections refer to misreadings or misprints (118, 532, 2402, 4 2 9 9 , 4386, 5682, 5706, 6307, 6359). Of the other suggested emendations, I accept the following without reservation: D 10 (cf. ED, 20), 164 (cf. 2363), 1073 {ED, 698), 2 1 1 5 (see above), 2 3 5 4 (see above), 3226, 4 0 2 2 (rather onur,ED,

168-9; also 1680,

2451), 4 0 8 0 (see above), 4193 and 5721 (ED, 461), 5596, 5650, 6358. The following appear to me to be plausible: 552, 767 (not 1652), 1573 (rather, tònguldil),

1750 (note wholesale substitution in this verse),

2426, 4 3 8 1 (meaning?), 6029 (rather, beklik bulur?, cf. 2132). The remainder appear to me unacceptable.

12. ANIMAL TRAITS IN THE ARMY COMMANDER

A persistent motif in Turkic lore assigns qualities of specific animals to culture heroes and to warrior chieftains. Thus, Oghuz is said to have had the feet of a bull, the loins of a wolf, the shoulder-blades of a sable, and the breast of a bear. 1 Great warriors assume the names of strong animals or of birds of prey: Arslan (lion), Bori (wolf), Bughra (camel-stallion), Boka (dragon), 2 Chaghri (falcon), Toghril, etc. The practice of assigning animal names and animal qualities to individuals or to groups is a widespread one, and does not necessarily imply a totemistic or a shamanistic background. In this paper I shall examine the particular motif which assigns certain animal traits (the l i o n ' s courage, the f o x ' s cunning, etc.) to the a r m y commander in a general or ideal way. This is not the same as a folklore motif attaching to a specific culture figure, such as Oghuz. It belongs, rather, to that body of wisdom sayings which the Turks called Bilig. It appears in the Kutadgu

Bilig by Yusuf of Balasaghun (written 1069); and also, it is perhaps

the only piece of Turkic wisdom to have found a permanent place in Arabic literature, notably in the classical collections of Amthal or Proverbs. Indeed, the author of Kutadgu

Bilig probably knew this tradition in Arabic literature

and depended on it in this form more than in the form of oral tradition. There is also indication that he knew of a related Iranian tradition, again as preserved in Arabic literature. The passage in Kutadgu Bilig (lines 2309-2316) occurs in Chapter 30 in which Ogdiilmi§ explains to the king the qualifications of the army commander (sii ba§lar er)? kerek sii ba§inga bu bir kac kilik yagika yiiz ursa bu tiizse yorik

1 W . Bang and G. R. Rachmati |AratJ, "Die Legende von Oguz Q a g a n , " SPAW (1932) 683724, line 1 Iff. See R. Dankoff, "Baraq and Buraq," Central Asiatic Journal XV. 2 (1971) 102117, p. 111. 2 S e e R. Dankoff, "Kasgari on the Beliefs and Superstitions of the Turks," JAOS 95.1 (1975) 68-80, p. 79. R. Arat, ed., Kutadgu Bilig I: Metin (Istanbul: Milli Egitim Basimevi, 1947). In line 2311 I read azgin for azgir, following G. Clauson, An Etymological Dictionary of Pre-ThirteenthCedntury Turkish (Oxford, 1972), p. 283; and kotuz "yak" for kutuz "rabid dog." In line 2316 1 read d^i "his vengeance" for ugi "his battle-wing"; assuming the latter reading, one might translate something like "prepared to bring his wing into the fray"; cf. Arat's rendering: zafer kazanir (Kutadgu Bilig II: Terciime [Ankara: Turk Tarih Kurumu Basimevi, 1959)).

190

FROM

MAHMUD

K A § G AR Í TO

EVLÍYA

£ELEBÍ

yagida kor arslan yiireki kerek kari§tukta esri bileki kerek tonguz teg titimlig bóri teg kügi adiglayu azgin kotuz teg o§i yana algi bolsa kizil tilkii teg titir bugrasi teg kór og siirse keg sagizganda sakrak kerek tutsa o z kaya kuzguni teg yirak tutsa k ó z ulug tutsa hamyet kór arslanlayu iigi teg usuz bolsa tünle sayu bu yanglar bile er yagigi bolur yagiQi bolur hem yaraggi bolur yagigi yaraggi ki§i ol tugi yagi sanggugi hem yetürgen ogi

This may be paraphrased as follows: "When the Commander faces the enemy he should have these traits: the heart of a lion and the wrist of a leopard, the obstinacy of a boar and the strength of a wolf, the ferocity of a bear and the malice of a yak, the cunning of a fox and the vengefulness of a camel-stallion with his herd, the alertness of a magpie and the far-sightedness of a raven on the cliff, the fury of a lion and the sleeplessness of an owl at night. A man such as this is ever ready to smite the enemy and to exact his vengeance." Carl Brockelmann was the first to notice the relationship of this passage to the tradition in Arabic literature which I shall call "the Nasr ibn Sayyár tradition." 1 Brockelmann cited it only according to Abü Hilál al'Askari; and he failed to mention that it is attributed to Nasr ibn Sayyár, the great Umayyad general whose career in Khurasan and Central Asia spanned forty-five years, beginning in 86/705 and ending in the year of his death, 131/748. 2 Brockelmann mentioned only the attribution to the Arab historian al-Madá'inl (d. 228/842), whose Kitáb Futüh Khurasan, unfortunately lost, was a major source for al-Tabari. 3

^C. Brockelmann, "Altturkestanische Volksweisheit," Ostasiatische 73, pp. 50-51. 9 1 See art. "Nasr b. Saiyär" (Zettersteen) in EI1. O ' 1 See art. "al-Mada'ini" (Brockelmann) in El .

Zeitschrift VIII (1920) 50-

ANIMAL

TRAITS

IN

THE

ARMY

COMMANDER

191

Fritz Hommel criticized Brokelmann's characterization of this tradition as a "proverb." 1 Hommel himself suggested (most unconvincingly) that the tradition instead relates to the ten banners of the ten tribes of the Western Turks. E. E. Bertels was the next to discuss this tradition, in an article in Russian with the somewhat pretentious title: "On the Problem of Tradition in the Heroic Epos of the Turkic Peoples." 2 Bertels cited the Nasr ibn Sayyar tradition only according to al-Turtushi; he was apparently not aware of the articles by Brockelmann and Hommel. He then cited the Kutadgu Bilig passage, and suggested that both this and the Nasr ibn Sayyar tradition go back to a common ancient source in the Turkic military aristocratic tradition. Finally, he demonstrated the persistence of the motif in later Turkic literature, especially Turkman poetry. The Nasr ibn Sayyar tradition is not confined to Abu Hilal al-'Askari and al-Turtushi. It appears as early as al-Jahiz, and is also found in the works of al-Tawhidi, al-Tha'alibi, al-Zamakhshari, and Ibn al-Tiqtaqa. I shall here examine the texts of all of these in chronological sequence. 1) Jahiz (d. 255/868), Kitab al-Hayawan? His Isnad is: Abu 1-Hasan ['All ibn Muhammad al-Mada'ini] — Nasr ibn Sayyar al-Laythi. The tradition is attributed to "the chiefs of the Turks" ( ' u z a m a ' al-turk). The great C o m m a n d e r ( a l - q a ' i d al-'azlm al-qiyada) must have ten animal c h a r a c t e r i s t i c s {'ashar

khisal

Generosity of the Rooster Tenderness of the Hen Heart of the Lion Attack of the Boar Cunning of the F o x Trickery of the W o l f

min

akhlaq

al-hayawan):

(sakhä' al-dïk) (tahannun al-dajäja) (qalb al-asad) (hamlat al-khinzir) (rawaghän al-tha'lab) (khatl al-dhi'b)

Only these six are given. This defect probably reflects the manuscript tradition of Jahiz's work and not Jahiz's ignorance of the other four. (The editor somewhat arbitrarily supplies four others from later texts.) 2) Abu Hilal al-'Askari (d. 395/1005). Kitab Jamharat al-Amthal.4 The tradition is given in connection with the proverb: "More restless than the

^F. Hommel, "Zu den alttürkischen Sprichwörtern," Asia Major Probeband (Hirth Anniversary Volume) (1923) 182-93, pp. 190-191. 2 E . Bertel's, "K voprosu a traditsii v geroiceskom epose Tyurkskix narodov," Svoetskoe Vostokovedenie IV (1947) 73-79. ^Jäliiz, Al-Hayawän, ed. 'Abd al-Saläm Muhammad Härün, 7 vols., Vol. II, Cairo, 1356/1938 (second edition 1385/1965), pp. 353-354. ^'Askarl, Jamharat al-Amthäl, ed. M. Abü 1-Fadl Ibrahim and 'Abd al-Majid Qutärn'isii, 2 vols., Vol. I, pp. 330-331. Brockelmann cited the edition printed on the margin of Maydäni's Majma' al-Amthäl, 2 vols., Vol. I (Cairo, 1310/1893), p. 220, line 1 Iff. The more recent edition gives the mysterious tenth animal in the form y. '.rü.

192

FROM

MAHMUD

K A § G A RI TO

EVLiYA

gELEBI

Qutrub" (ajwal min qutrub). Isnad: Abu 1-Qasim — al-'Aqdi — Abu Ja'far — al-Mada'inl — Muhammad ibn Ibrahim ibn Nasr ibn Sayyar. Attribution to "the chiefs of the Turks." The great Commander ought to have ten animal characteristics ('asharat akhlaq min akhlaq al-baha'im): Bravery of the Rooster Wariness of the Hen Heart of the Lion Attack of the Boar Cunning of the Fox Endurance of the D o g against wounds Watchfulness of the Crane Caution of the Crow Aggression of the Wolf Fatness of the Y.gh.ru, which is an animal that fattens on trouble (or despite trouble) Rastlessness of the Qutrub

(shajä'at al-dik) (taharruz al-dajäja) [same as Jähiz] [same as Jähiz] [same as Jähiz] (sabr al-kalb 'alä 1-jiräh) (haräsat al-kurki) (hadhar al-ghuräb) (ghärat al-dhi'b) (siman y.gh.rü w a h w a däbba tasman 'alä 1-kadd) (jawalän qutrub)

Several points are to be noticed here. First of all, eleven animals are named instead of ten. Brockelmann cut off the last and translated these as follows: "die Tapferkeit des Hahnes, die Keuschheit der Henne, das Herz des Löwen, die Angriffslust des Ebers, die Schlauheit des Fuchses, die Ausdauer des hundes bei Verwundungen, die Wachsamkeit des Kranichs, die Vorsicht des Raben, die Kampflust des Wolfes, die Wohlbeleibtheit des Jagru, eines Tieres, das trotz aller Mühen und Anstrengungen fett bleibt." Hommel made things worse by confusing the last two, making "restlessness" (jawalän) a characteristic of the Y.gh.rü. Hommel also suggested (unconvincingly), on the basis of the Kutadgu Bilig passage, that "rooster" ( a l - d i k ) was a misunderstanding of Turkic adig "bear" and that "hen" was originally "magpie" but was changed to "hen" to go with "rooster." As for the Y.gh.rü, Brockelmann assumed that this was a Turkic word: "Der Name des letztgenannten Tieres ist sonst nirgends belegt, aber doch offenbar vom turk. Jag, 'Fett,' abzuleiten." (In a footnote he said this word probably had nothing to do with the animal named yargun which is found in the Irq Bitig.) Hommel, because of the confusion noted above, connected the word with Turkic iigi "owl," which is also restless (but only at night), and which in the Kutadgu Bilig passage also comes last. To begin with the Qutrub, 'Askari glosses the words as follows: "an animal that scurries about all night, and all day, without sleeping" (wahya däbba tajülu l-layl kullahu wa-l-nahär kullahu lä tanäm). The Arabic

ANIMAL

TRAITS

IN

THE

ARMY

C O M M A N D E R

193

lexicographers apply the word to several scurrying animals and especially to the insect known in English as firefly or lightning bug. It it also worth noting that al-Kashgharl uses the Qutrub to define the Turkic word kozkeni (vocalization uncertain) and glosses it as follows: "a type of beetle which flies about at night with a buzzing sound" (wahwa min jins al-ji'lan yatiru bi-llayl ma'a tanln).' Perhaps this latter word is to be connected with another animal named by Kahghari, kdzkuna (vocalization uncertain), and defined as "a bird which resembles the sacer falcon and the lizard and lives on wind" (ta'ir yushbihu. l-saqr wa-l-'azaya wa-yataballaghu bi-l-rili).- At any rate, the Qutrub is not an owl. Since none of the other Arab writers include the Qutrub in the list, it can be assumed that 'Askari added it himself in order to hang the Nasr ibn Sayyar tradition upon an Arabic proverbial expression. Discussion of the Y.gh.ru must be postponed until we can compare the form of the word found in the other texts. Finally, comparing the two lists given so far, w e can see that, although the order of the animals is the same, the traits associated with them are different in three cases. The first two (sakha' al-dlk, shaja'at al-dik; tahannun al-dajaja, taharruz al-dajaja) can be explained by scribal misreadings in one or the other of the manuscript traditions. The third (khatl al-dhi'b, gharat al-dhi'b) can only be due to a replacement within the oral transmission of the tradition and is on the same order as the replacement of words in the phrase meaning "ten animal characteristics" ('ashar khisal min akhlaq al-hayawan, 'asharat akhlaq min akhlaq al-baha'im). 3) Abu Hayyan al-Tawhidi (d. 414/1023). Kitdb al-Imta' wa-lMu'anasa? No Isnad. The attribution is simply to "the Turks." The great Commander (al-qa'id al-'azim) ought to have ten animal characteristics {'ashar khisal min durub al-hayawan):

Käshghari, Diwan Lugät al-Turk, facsimile edition published by the Türk Dil Kurumu (Divanü

Lügat-it-türk Tipkibasimi, Ankara: Aläeddin Kiral Basimevi, 1941), 247:2. The translation is that of Clauson, Dictionary, p. 760.

^Käshghari, Diwän, 263:17. Clauson (ibid.) is faulty. According to Käshghari, the wolf also maintains itself on wind for one week out of each month (Dlwan, 317:6). See DariVoiT, JAOS (1975), p. 78. a Tawhidi Al-Imtä' wa-l-Mu'änasa, ed. A. Amin and A. al-Zayn (Beirut, n.d.), 3 vols., Vol. I, p. 144'.

194

FROM

MAHMUD

K A § G A R 1 TO

Generosity of the Rooster Tenderness of the Hen Courage of the Lion Attack of the Boar Cunning of the Fox Endurance of the D o g W a t c h f u l n e s s of the Crane Caution of the Crow A g g r e s s i o n of the W o l f Fatness of the B . ' . r . w a , w h i c h is an animal in Khurasan that fattens on (or despite) toil and trouble

EVLÍYA

Q ELEB Í

(najdat al-asad)

(siman b'rw' wahya dabba bikhurasan tasman 'ala 1-ta'b wa-1saqa')

Tawhidi undoubtedly knew Jahiz's version and probably cites the tradition according to the text of Jahiz which he knew, the only differences (besides the additional animals) being the trait of the wolf ( g h a r a instead of khatl) and the word for "courage" of the lion ( n a j d a instead of qalb). 4) Mansub}

Tha'alibi

(d.

348/037).

Thimar

al-Qulub

fi

l-Mudaf

wa-l-

Isnad, attribution, and introduction are the same as in Jahiz.

Courage of the Lion Trickery of the W o l f Cunning of the Fox Attack of the Boar Endurance of the D o g against w o u n d s Tenderness of the Hen Generosity of the Rooster Caution of the Crow Watchfulness of the Crane Guidance of the D o v e

(jur'at al-asad)

(hidayat al-hamam)

The difference in order is perhaps to be explained by Tha'alibi's hanging the Nasr ibn Sayyar tradition on the Arabic proverbial saying: "the lion's courage" (jur'at al-asad). Five four-legged animals are given first and are balanced by five birds. 5) Turtushi (d. 520/1126 or 525/1131). Sirdj al-Muluk.2 "Nasr ibn Sayyar said: 'I was emir of Khurasan under Marwan al-Ja'di, last of the Umayyad kings, and the chiefs of the Turks used to s a y . . . ' " Introduction same as in 'Askari.

^Tha'alibi, Thimar al-Qulüb, ed. M. Abü 1-Fadl Ibrahim (Cairo, 1384/1965), p. 382. ^Turtüshi, Siräj al-Mulük (Cairo, 1319/1901), p. 151. Bertels cited the edition Cairo, 1306/1887, p. 140. Cf. the translation by M. Alarcón, Lámpara de los Príncipes, 2 vols., Vol. II (Madrid, 1931), p. 313. Alarcón read the mysterious last animal "naguir" and rendered the gloss: "pequeño reptil del Jorasán, que engorda con la fatiga." Bertels read it "nugayr" and rendered the gloss: "melkoe Zivotnoe v Xorasane, kotoroe Zireet, nestmotrya na trudnosti i bedstviya."

ANIMAL

TRAITS

IN T H E

Bravery of the Rooster Searching of the Hen Heart of the Lion Attack of the Boar Cunning of the Fox Endurance of the Dog against wounds Watchfulness of the Crane Aggression of the Wolf Fatness of the N.gh.y.r, which is a little animal found in Khurasan that fattens on toil and trouble

ARMY

COMMANDER

195

(bahth al-dajaja)

(siman n.gh.y.r w a h y a duwaybba takun bi-khurasah tasman 'ala 1-ta'b wa-l-shaqa')

Only these nine are given. The trait associated with the hen is an obvious scribal error (bahth for tahannun or taharruz). 6) Zamakhshari (d. 538/1144), Kitab al-Mustaqsa fi al-Amthal} Under the proverb: "fatter than the Yaghru" (variants: b.'.r, y.'.ru, y.gh.r) — glossed as "a little animal in Khurasan which fattens on trouble" (duwaybba bi-khurasan tasman 'ala l-kadd). "The chiefs of the Turks say: the great Commander must have the Bravery of the Rooster, the Cunning of the Fox, the Caution of the Crow, and the Fatness of the Y.gh.ru." 7) Ibn al-Tiqtaqa (wrote 701/1301), Al-Fakhri.2 Attribution to "certain sages of the Turks" (ba'd hukama' al-turk). The Army Commander (qa'id al-jaysh) ought to have ten animal characteristics {'ashar khisal min akhlaq al-hayawan): Courage of the Lion Attack of the Boar Cunning of the Fox Endurance of the D o g against wounds Aggression of the Wolf Watchfulness of the Crane Generosity of the Rooster Solicitude of the Hen for her chicks Caution of the Crow Fatness of the T.'.ru, which is an animal found in Khurasan that fattens on travel and trouble

(jur'at al-asad)

(shafaqat al-dajaja 'ala al-farârîj) (siman t.'.rü wahya däbba takün bikhuräsän tasman 'alä 1-safar wa-1kadd)

The trait associated with the hen is obviously a replacement for the original "tenderness" (tahannun). In the gloss on the T.'.rii, safar ("travel") must be due to a misreading of shaqa ("trouble").

' Zamakhshari, Al-Mustaqsà fi l-Amthàl, 2 vols., Vol. I (Hyderabad, 1962), p. 171. Ibn al-Tiqtaqà, Al-Fakhri (Beirut: Dar Sader, 1966), p. 58.

2

196

FROM

MAHMUD

K A § G A R I T O EV Li Y A

gELEBI

It was undoubtedly this Nasr ibn Sayyar tradition which gave rise to the Arabic proverb "fatter than the Yaghru." This is clear from the Zamakhshari example above (6). It is found in all of the classical collections of Amthal, including that of Hamza al-Isbahanl (d. c. 351/961), 1 Abu 1-Hilal al-'Askari, 2 and al-Maydani (d. 518/1124). 3 In the edition of Hamza al-Isbahani, the name is given as Ya'r and glossed: dabba takiin bi-khurasan tasmun 'aid l-kadd. Freytag, in his edition of Maydani, noted the following manuscript variants: y.'.ru, y.gh.zu, y.gh.r, y.'.rud, y.gh.ru, y.gh.rud. His translation of the gloss (same text as Hamza) is: "Animal illud in Chorasan vivens, dum molestiam fert, pinguescere dicitur." The word did not enter the classical Arabic lexicographic tradition, as far as I can determine. It is found in the Muhit al-Muhit of Butrus al-Bustani (first published 1867-70), where it is assimilated to al-ya'r, a totally different animal with a different and unrelated proverb associated with it. 4 Three different sources of the word can be contemplated: Turkic, Arabic, and Iranian. At first glance, an Iranian source would seem the most likely in view of the Khurasanian provenance attributed to the animal in the gloss. However, there is no obvious Iranian etymon. An Arabic source is the least likley, in view of the numerous variants and the glaring unfamiliarity with the word on the part of the finest Arabic philologians. 5 It was noted above that Brockelmann assumed a Turkic etymology, connecting the word with yagh. However, no animal name derivative from yagh is attested. Furthermore, the ancient meaning of yagh did not include "fat" but only "oil" and "butter" (in Kashghari: duhn and samn; only the Oghuz used it for "animal fat," shahm).6 "Fat" as opposed to "lean" is semiz. "Animal fat" is yaqri, a word more likely than yagh to be connected with our mysterious animal. 7 Yarghun, which Brockelmann rejected (it is found only in Irq Bitig in the compound yarghun keyik) is possibly the same animal as arghun, which Kashghari states is "a small animal of the rat genus (duwaybba min jins al-jirdan); its body is as long as half a cubit, and it is

' Isbahani, Al-Durra al-Fakhira, ed. 'Abd al-Majid Qutamish, 2 vols., Vol. I (Cairo, 1972), p. 234. 'Askari, Jamhara (1384/1963), vol. I, p. 536. The word is given as Ya'r in this edition. 'Askari has no gloss here, but refers the reader to the earlier mention of the animal (above, p. 191, n. 4). ^Maydani, Majma' al-Amthal, 2 vols., Vol. I (Beirut: Hayat, 1961), p. 496; G. W. Freytag, Arabum Proverbia, 2 vols., Vol. I (Bonn, 1838), p. 644. Ya'r is glossed by the Arabic lexicographers as "a kid or lamb which is tied at the bottom of a pitfall as a lure for a wolf or lion," from which comes the proverb: "lowlier than the Y a ' r " (iadhall min al-ya'r). It is surely only a coincidence, then, that yaghru, imperfect of ghara, is used of fat adhering to the heart; or that nughayr (cf. the form in TurtushT, 7 above) is a word meaning "little sparrow." ^Kashghari, Diwan, 515:2. 7 Kashghari, Diwan, 458:2. A plant name derived from this, yaqriqan, is attested at 468:12.

ANIMAL

TRAITS

IN

THE

ARMY

C O M M A N D E R

197

used to hunt out small birds in the crevices of walls; if it attacks a sheep its flesh turns yellow, and if it attacks a man in his sleep he suffers retention of urine." 1 The form in Tawhldi (3 above) suggests bughra, "camel stallion," which is found in Khurasan and could be said to stay fat despite travel and trouble (a possible rendering of the gloss in Ibn al-Tiqtaqa, 7). Considering the Arabic vocabulary for "camel," however, one would not expect bughra to be defined by such a colorless word as dabba. The ms. variants of Maydani with final dal which Freytag noted suggest a possible misreading of toghrul. The toghril (note vowel!) was a somewhat faboulous bird for the Turks 2 but not particularly known for its fatness, nor an animal that could be referred to by the Arabic dabba (or even duwaybba). Thus far we have approached the problem from the point of view of orthography or palaeography. A more fruitful approach will be to ask: Is there an animal in the Turkic tradition that was proverbial for fatness, as the Yaghru came to be in the Arabic Amthal collections because of the Nasr ibn Sayyar tradition? Here, again, Kashghari provides the clue, in his gloss on borsmuq: "a small animal resembling the Warak; it is proverbial for fatness (duwaybba mithl al-warak wa-bihi yudrabu l-mathal fi l-siman); the Oghuz drop the mlm and say borsuq."3 Warak is an error for Waral, which is the monitor lizard. 4 Borsuq (porsuq) is the common Turkic word for "badger." That Kashghari did not know an Arabic equivalent strengthens the connection between the badger and the Yaghru. The Sanglax also states of the porsuq: "whenever they kill it, it is fat"; 5 and there is a Kirgiz proverb: "The badger is fat but has no straight intestine (considered a delicacy in horses)" — qasqulaq semiz, qarta zoqP It should be noted that the spelling of borsuq agrees with Yaghru in the following points: the first letter (B/Y differing by one dot) ; the r; and the last letter (qaf and waw often are confused in Arabic script). The ghayn or 'ayn might have been a misreading of an original sin, and this would give us 1

Irq Bitig, LXII; Kashghari, Diwán, 72:15-73:1. Clauson, Dictionary, p. 216, suggests the arghun might be a weasel. 2 "It kills a thousand geese and eats one," Kashghari, Dlwan, 242:1. See Dankoff, J AOS (1975), p. 78. •^Kashghari, Diwán, 625:6. 4 C f . definition of sughur (Diwán, 183:5) "weasel (wabr)\ it is a small animal resembling the monitor lizard ( w a r a f ) \ its skin is used to make raincoats." 5 N o t e Persian rüdak which the Sanglax (132 v. 22) gives as equivalent to porsuq (Clauson, Dictionary, p. 369) The Arabic equivalent, washaq, is, as Clauson states, unidentified. The "Isfahani" equivalent, xükrá (Clauson, in error, xüxra) is probably the same as gürkan which the Persian dictionaries give as one of the words for "badger"; the usual word in Persian is shighár. ^K. K. Yudakhin, Kirgizsko-russkiy

Slovar' (Moscow, 1965), p. 364.

198

FROM

MAHMUD

K A § G AR i TO

EVLIYA

Q ELEB i

the otherwise unattested metathesized form: *bosruq. I would hesitate, however, to reconstruct Yaghru as Borsuq, or the like, without further ado; possibly it reflects an altogether different word. 1 The identification as "badger" does appear to be sound. Before attempting to reconstruct the original form of the Nasr ibn Sayyar tradition and comparing it with the Kutadgu Bilig passage, we must first introduce a related Iranian tradition, also found in Arabic literature. In the voluminous collection of wisdom lore compiled by Miskawaih (d. 421/1030), Jdvidan Khirad, we find the following: 2 The ancient Persians (qudama' alfurs) only appointed as lord of the marches (yuwallun al-thughur) one who combined in himself fourteen animal characteristics (arba' 'ashrat khasla min akhlaq al-hayawan); he must be: more more more more more more more more more more more more more more

Hearing than a Horse Sightful than an Eagle Guiding than a Sandgrouse Cautious than a Magpie Courageous than a Lion Leaping than a Cheetah Cunning than a Fox Impudent than a Wolf Generous than a pecking Rooster Forward than a Leopard Gathering than an Ant Watchful than a D o g Patient than an A s s Obedient than a Camel

(asma' min faras) (absar min 'uqab) (ahda min qatah) (ahdhar min 'aq'aq) (ajra' min asad) (awthab min fahd) (arwagh min tha'lab) (awqah min dhi'b) (askha min laqitat al-dik) (aqdam min namir) (ajma' min dhurra) (ahras min kalb) (asbar min himar) (atwa' min jamal)

A slightly different version of this tradition is given by Abu Hilal al'Askari. 3 In the course of explaining a proverb concerning haughtiness, he cites the reasons given by Qutayba ibn Muslim for not sending a certain general to the front. In the course of his explanation, Qutayba states that he never saw a warlord (sahib harb) who was haughty who was not also defeated in battle, even though, when he confronted the enemy, he was:

1 Perhaps the Mongol word for badger, dorgo, should be taken into account. Also Mg. targun "fat" (of animals). Altai yoroqon and Lebed toroqon are probably borrowed from the Mongol. 2 Miskawayh, Al-hikma al-Khalida (Jawidan Khirad), ed. 'Abd al-Rahman Badawi (Cairo, 1952), p. 64. 3 'Askari, Jamhara (1384/1963), Vol. I, pp. 166-167.

ANIMAL

TRAITS

IN T H E

more Hearing than a Horse more Sightful than an Eagle more Guiding than a Sandgrouse more Cautious than a Magpie more Forward than a Lion more Leaping than a Cheetah more V e n g e f u l than a Camel more Cunning than a Fox more Tricky than a W o l f more G e n e r o u s than a R o o s t e r w h o c h i c k s with his beak more Thrifty than a Gazelle more Gathering than an Ant more Watchful than a D o g more Patient than a Lizard

ARMY

COMMANDER

199

(asadd iqdaman min al-asad) (ahqad min jamal)

f e e d s his

(aghdar min dhi'b) (askha min lafiza) (ashahh min z a b y )

(asbar min dabb)

The Iranian provenance of this tradition again appears in the Fakhri of Ibn al-Tiqtaqa. 1 Among the qualities of a king (al-malik), according to Bozorgmihr, are that he be more Hearing than a Horse, more Sightful than an Eagle, more Guiding than a Sandgrouse, more Cautious than a Crow (asadd hadharan min ghurab), more Forward than a Lion ( a ' z a m iqdaman min alasad), and stronger and quicker in Leaping than a Cheetah ( a q w a wa-asra' wuthuban

min

al-fahd).

It must first be noted that the confusion in the epithet for "rooster" laqita) is a natural one in the Arabic script. Also, both Arabic roots are used of the rooster, since it pecks (laqata) at grain, and it also takes the grain in its beak and, instead of swallowing it, spits it ( l a f a z a ) into the mouths of its chicks. It is the latter habit which gave rise to the epithet lafiza being applied to the rooster — the feminine ending, according to the Arabic philologians, is for exaggeration (mubalagha). And it is this habit also which gave rise to the Arabic proverbial expression, "more generous than a rooster who feeds his chicks with his beak" ( a s m a h min al-lafiza).2 On this basis, we may correct the form given by Miskawaih. On this basis, also, we may confidently say that the quality of the rooster in the Nasr ibn Sayyar tradition was originally Generosity (sakha' — thus 1, 3, 4, and 7 above) and not Bravery (shaja'a — thus 2, 5, and 6). 3 Along with this, the quality of the hen must have been Tenderness (tahannun (lafiza,

— thus 1, 3, 4; replaced by shafaqa in 7) and not Wariness ( t a h a r r u z — thus 2) or Searching (bahth — thus 5).

1

Fakhri, p. 25.

o _ _ oThus Tha'alibI, Thimar, p. 473, s.v. samahat al-dik. Here laqita should be read lafiza. But note 'Askari, Jamhara, I, 538: ashja' min dik\ (line 10).

200

FROM

MAHMUD

K A § G A R I TO EVLIYA

Q ELEB 1

Furthermore, it is clear from a glance at the Kutadgu Bilig passage that the quality associated with the lion in the Nasr ibn Sayyar tradition was originally Heart (qalb — thus 1, 2, 5), and that this was simply replaced by its equivalent, Courage, in later forms of the tradition ( n a j d a — 3,jur'a — 4, 7), in order to make it conform with the Arabic proverb: "the lion's courage" (jur'at al-asad). It was presumably Tha'alibi (4 above), or his immediate informants, who changed the original order of the list, and for the same reason, namely in order to attach it more readily to an Arabic proverb. Again, it is only in Tha'alibi's list that we miss "fatness of the Yaghru" and find instead "Guidance of the Dove" — perhaps, as suggested above, in order to balance the five four-legged animals with five birds. We may also eliminate from consideration 'Askari's "Restlessness of the Qutrub." From the above considerations, the Nasr ibn Sayyar tradition can be reconstructed to conform with the following list: Rooster Hen Lion Boar Fox Dog Crane Crow Wolf Badger (? "Yaghru")

Generosity Tenderness Heart Attack Cunning Endurance (against wounds) Watchfulness Caution Trickery or Aggression Fatness

This order accords with the oldest sources (1, 2, 3; also 5 and 6). The wolf's "Trickery" (khatl — thus 1, 4) would accord with the quality associated with the wolf in the Iranian tradition according to 'Askari's version (aghdar min dhi'b). His "Aggression" (ghara — thus 2, 3, 5, 7), on the other hand, would accord with the Iranian tradition in the version of Miskawaih (awqah min dhi'b), and also with the Kutadgu Bilig passage, to which we finally return. The list of animals and their traits in the Kutadgu Bilig passage may be resumed as follows:

ANIMAL TRAITS Lion Leopard Boar Wolf Bear Yak Fox Camel-stallion Magpie Raven Lion Owl

IN T H E A R M Y

COMMANDER

201

Heart Wrist Obstinacy Strength Ferocity Malice Cunning Vengefulness Alertness Far-sightedness Fury Sleeplessness

Counting the lion once, there are eleven. Both Brockelmann (following Radloff) and Bertels counted only ten, failing to note the second (esri "leopard" which Radloff had translated "strong"). Hommel pointed out Brockelmann's error, but he too counted ten, since in addition to the lion doublet, the yak-camel pair is also a doublet, on the grounds that the same quality (dg, dg keg) is assigned to each. In Kutadgu Bilig we have a poetic rendering of the Na§r ibn Sayyar tradition. Yusuf of Balasaghun might have known it as a living oral tradition among the Turks or else (or in addition) he knew it in its literary form in the Arabic authors. He also certainly knew the related Iranian tradition cited above, and borrowed from it freely to serve his artistic ends. From it, for example, he took the Leaping of the Cheetah (or, perhaps, the Forwardness of the Leopard — the two animals were often confused); and he substituted for this quality Wrist (bilek), in order to provide a rhyming and semantic parallel to the lion's Heart (ytirek). From the same source, no doubt, he borrowed the Vengefulness of the Camel (above, only in 'Askari's version; Miskawaih has Obedience); the Alertness of the Magpie (Caution in the Iranian version, a quality that characterizes the Crow in the Nasr ibn Sayyar tradition); and the Far-sightedness of the Raven (of the Eagle in the Iranian version). The omission of Rooster and Hen at the beginning of the list perhaps follows the model of Tha'alibi (4 above) who also began the enumeration with Lion. The addition of Bear and Yak serves to fill out the verses. The presence of two "doublets" is a bit awkward, but is not distressing in Yusuf s verse, which tends to wordiness and repetition. The second quality associated with the lion, Fury (hamyet, from the Arabic hamiyya, more exactly "ardor, enthusiasm") looks suspiciously like Attack (hamla) of the boar in the Nasr ibn Sayyar tradition, and lends weight to the contention that Yusuf relied more on written (i.e., Arabic) sources than on oral (Turkic) ones. Owl, finally, is a happy substitute for the Yaghru (badger?), whose somewhat undignified Fatness is replaced by Sleeplessness, a trait that accords better with the dominant ethic of Kutadgu Bilig.

13. SOME NOTES ON THE MIDDLE TURKIC GLOSSES

In my review of MTG1 I suggested that several of Eckmann's readings were "ghost words" — i.e., words based on a copyist's error or on an editor's misreading. In the summer of 1978 I had the opportunity to examine briefly the ms. of the Rylands Inter-linear Koran Translation. As a result of this examination, and of other considerations, I have modified some of the views I expressed in my review. I no longer doubt the genuineness of äl "hand," ong "share," and tiikiin "barren, sterile." 2 At 30/23b: 1 there is no sign of Uyiik, only livä. Eckmann must have been confused by the other example of iiyük, in the same phrase, which he cites on p. 316-7. That yönlägli

{MTG, 349) must be read tilägli is confirmed by the

spelling of the word in the ms., thus: NYL'KLY. In two other cases as well, Eckmann's readings must be emended by resort to a close look at the ms. MTG, 349, yosun "message, commandment." Menges has interpreted this as a Chinese loanword. 3 (But the Chinese word suggested means "tradition, custom, ceremony," not "message.") In my review I wrote: 4 "Eckmann thought that the manuscript was executed in the late thirteenth or early fourteenth century; and this dating is supported by the presence of at least one Mongol loan, yosun 'message, commandment' (Ar. risäla) from Mg. yosun 'custom, rule; doctrine,' unless this is simply a copyist's error for yumus (DLT, 449)." The reading as yumus (cf. ED, 938) 5 is confirmed by the shape of the word in ms. (approximately: YVMVS). Thus, there are no Mongol loans, and the text is definitely pre-thirteenth-century. Eckmann, Middle Turkic Glosses of the Rylands Interlinear Koran Translation (Budapest, 1976). Review by R. Dankoff in JAOS 98.2 (1978), 135-137. 2 Aside from äl "hand" alongside älig in A.K. Borovkov, Leksika sredneaziatskogo Tefsira XIIIXV vv. (Moscow, 1963), it occurs in Qutadgu Bilig in the following lines: 1439, 1454b, 1601(7), 3125 (see JTS/TUBA 3, 1979, p. 93). On ong, see JTS/TUBA 3, 1979, p. 461 and n. 3. On tükün (better: tügün) see S. Tezcan in 'IDA YB 1978-9, p. 290. ^K. Menges, "Etymologische Notizen zu Eckmanns Mittel-Türkischem Tafsir," WZKM 70 (1978), 35-52, p. 50-1. Dankoff, p. 136. -*ED: G. Clauson, An Etymological Dictionary of Pre-Thirteenth-Century Turkish (Oxford, 1972).

204

FROM MTG,

Bodrogligeti

MAHMUD 7 7 8 . ark

2

K A § G A R 1 TO EVLÍYA

qural

"power

and

c o m p a r e d this with a word qural

ÇELEBÎ

authority."

Menges1

and

( W B ii, 923: Turn. Bar.) ~

qoral (Bashkir) meaning "instrument, w e a p o n . " (This is again unsatisfactory on semantic grounds.) In my review I rashly suggested t w o or three w o r d s of w h i c h qural

m i g h t be the c o p y i s t ' s m i s r e a d i n g . 3 E x a m i n a t i o n of the m s .

r e v e a l e d that the w o r d is s i m p l y tiirk

(i.e., T V R K not Q V R a L ) .

The

hendiadys ark tiirk is well attested in Middle Turkic (see ED, 220). In conclusion, I give a transcription of the two pages w h i c h gave rise to these t w o ghosts, published herewith in f a c s i m i l e by p e r m i s s i o n of the Rylands Library.

1.

30/25b (Koran 15:42-3) 'alayhim laysa laka bar ïsân nist mar tura sanga olarning iizâ yoq

sultànun bar-gumâstagï va hujjati àrk tiirk

2.

ilia magar magar

man ânki ol kim

ittaba'aka pasravi kunadat sanga udsa

min al-gàwîn. az bi-râhân. yolsuzlardïn.

3.

wa-inna va har âyina butunlukin kirn

jahannama düzax tamug

la-maw'iduhum va'dagâh-i ïsân-ast olarning urulmïs 4 ôdlâri

ajma'ïn hama ol yumqïlarïning

1.

2.

3.

33/8b (Koran 33:39-40) yuballigüna risâlâti 'llàhi mï rasânand paygâmhâ-yi xudây-râ tangrïnïng yumuslarïnï tâgriirlâr-'

wa-yaxsünahu va mi tarsand az vay qorqarlar anda

wa-lâ yaxsüna va na tarsand qorqmazlar

ahadan

illâ

'llâha

az hïô kas

magar

xudây

kim àrsadà

magar

tangrï

wa-kafâ va bas ast tap

bi-llâhi xudây tangrï

hasïban sumâr kunanda saqïslagan^

ma kâna na büda ârmàz

^Menges, p. 43. ^A.J.E. Bodrogligeti, "The Technique of the Glossist as a Key to Understanding the Lexical Material of Early Eastern Middle Turkic Interlinear Q u r ' â n Translations," UAJ 50 (1978), 1924, p. 23. 3 Dankoff, p. 136. 4 ms.: 'VDRLMYS; cf. MTG, 309. -*Tàgiïr- "to convey" is not registered in MTG, 282-3 (but cf. tagurmaklik, 6 C f . MTG, 247.

p. 283).

SOME

NOTES

^

ON T H E M I D D E E T U R K I C

^

¿¿^

GLOSSES

'é l

l

i

i

f

r

T

S

^

J

S

John Rylands Library, Arabic 30, fol. 25b

205

206

FROM

MAHMUD

KA§GARÌ

TO E V L Ì Y A

John Rylands Library, Arabic 33, fol. 8b

CELEBÌ

14. THE LYRIC IN THE ROMANCE: THE USE OF GHAZALS IN PERSIAN AND TURKISH MASNAVlS*

In Islamic poetry one generally distinguishes between the shorter, monorhyming poems, whether lyrical ("ghazal") or panegyric ("qasideh"), on the one hand, and the longer poems consisting of rhymed couplets ("masnavi"), whether of an epic, romantic, or didactic character, on the other. While a poet like Jalaladdin Rumi excelled at both forms, he refrained from mixing the two; and so we have his Divan of over 3,000 ghazals and his Masnavi of 26,000 couplets. Firdausi's Shah-Nameh, consisting of 60,000 couplets, is likewise uninterrupted by any lyric outburst. So much for the mystical/didactic and epic masnavis. When we come to the romances, the situation is more complicated. 'Ayyuqi, who initiated the romantic masnavi in Persian in the eleventh century, included ghazals in the course of his poem (see below). His Turkish imitator Yusuf-i Meddah in the fourteenth century followed suit, and the use of ghazals became a common feature in the subsequent tradition of Western Turkish narrative romance, so much so that it has been considered a hallmark of the Turkish masnavi (Gibb, vol. 2, p. 173; Ate§, p. 132). But Nizami, who put his imprint on the genre in the twelfth century, eschewed the lyric insertion; and in the subsequent Persian tradition (notably the works of Amir Khusrau 1 and Jami), as well as the Eastern Turkish tradition (Qutb and Nava'i, but also the Western Turkish Fahri), the romantic masnavi has the same austere and unrelieved quality that we find in the epic and religious masnavis. One exception to the rule regarding the Persian masnavi tradition is Salman Savaji's Jamshld u Khurshld (763/1362). This relatively short work (2,700 verses) is fairly overloaded with lyrical insertions: they total 85 short pieces, including 45 ghazals, 18 qit'ehs, 21 ruba'is, and 1 fard (single couplet, different meter from the surrounding masnavi). Salman was apparently following the unconventional dictate of his patron, who told him (p. 12, 1. 9):

* For references, see list of works cited at end of this article. 'i.e., his Khamseh: there are ghazals in some of his historical masnavis (Rypka, p. 259) and 9 ghazals, one at the end of each section, in his Nuh Sipihr, which is unconventional in other respects as well.

208

FROM

MAHMUD

K A § G A R Í TO E V L Í Y A

f ELEB Í

Zi-har jins-i hikáyat dar ham ámiz Zi-har naw'i gazalhá'i bar angiz Add a pinch from every kind of story, and stir in some of every sort of ghazal.

In some sections the work is simply a hodgepodge of lyrics strung together by a slender thread of narrative. It is hardly surprising that Salman's example was not followed in Persian, although there are at least two Turkish imitations: one by Ahmedi (806/1403), the other by Prince Cem (881/147677). The first of these, recently published, in a text of 4,800 verses, contains no fewer than 93 short pieces, including 85 ghazals, 5 qit'ehs (3 in Arabic), 2 musammats, and 1 masnavi in a different meter. Ahmedi pretends to follow the dictate of his own patron, who directed (line 401): Gazellerle miiretteb eyle aril K i§idenüñ sevine cism ii cam Compose it with ghazals, that the listener may rejoice in body and soul (Ahmedi includes his makhlas or "signature" in one of the ghazals only, p. 145).

Another apparent exception is Maktabi's Layli u Majnün (895/148990). In Rypka's judgment (p. 213) this "is a work that even achieves new effects by means of lyrical ghazal insertions. Somewhat later it was imitated by Fudüli...." Fuzüli did not need Maktabi's example, since he was simply writing in the Turkish tradition; nor was Maktabi a particularly important influence on Fuzüli's conception of the romance (see Levend, p. 267). As for the ghazal insertions in Maktabi's poem these consist of two, the first at the end of Layli's letter to Majnün (p. 116), the second at the end of Majnün's reply (p. 121). They are in the same meter as the surrounding text; they do not share a common rhyme, and there is no makhlas. These features distinguish these ghazals from those in similar contexts in the Turkish masnavi tradition culminating in Fuzüli's poem (see below). Rather, they are features that appear to derive from the Persian Dah-námeh genre, in which love-letters in masnavi form generally nclude a ghazal at the end (see Gandje'f). *

Zablhullah Safa, the editor of 'Ayyüqi's Varqa u Gulsháh (written before 1030?), writes on p. ix of his introduction: In this poem there occur several ghazals, in the original ( m u t a q á r i b ) meter, but with monorhyme according to the manner of ghazal writers. This is something novel in Persian literature. The original reason for it, in my opinion, is that in the original legend of Varqa and Gulsháh (i.e.,

THE

LYRIC

IN

THE

R O M A N C E

209

in the story of ' U r w a and ' A f r a ' , discussed below [p.x, citing Aghani, Beirut, vol. 20, pp. 366-79]), we see, according to the old A r a b manner, that the explanation of an event is accompanied by mention of loveverses and ghazal-laments attributed to the Arab poet ' U r w a ; and the author w h o transmitted this legend into Persian verse, wherever he came upon love-verses ("ghazals"), included a ghazal in the course of the masnavi. These ghazals have a charming ( l a t l f and dil-angiz) effect, and are among the oldest extant Persian ghazals that are joined to context....

The ten ghazals may be analyzed as follows: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

a love-poem in Rabi' ibn 'Adnan's message to Gulshah (p. 13). Varqa bewails his loss of Gulshah (p. 15) Varqa. riding to battle against Rabi', vows to rescue Gulshah (p. 17) Rabl's boast in battle (p. 20) Varqa's lament on the death of his father in battle (p. 27) Gulshah, betrothed to the king of Syria, recites a love poem for Varqa (p. 75). Varqa, believing Gulshah dead, recites a lament (pp. 81-82) Varqa begs the physician to cure him of his love-sickness (p. 108) Varqa pours out his heart to Gulshah before expiring (p. 110) Gulshah's lament on hearing of Varqa's death (p. 112)

It is fitting that the earliest Western Turkish romantic masnavi should be an imitation of the earliest Persian one. Yusuf-i Meddah's Varqa ve Gul§ah (743/1342), besides using a different meter (remel 1), departs in numerous details from its Persian model, both in the narrative line and in the placement and function of the ghazals. 1 Thus, in 'Ayyuqi's poem (pp. 54-56) there is an exchange of vows between the two lovers, followed by a tearful parting at which Varqa gives Gulshah a ring. In Yusuf's adaptation, Varqa first gives the ring (lines 326-27), then recites a ghazal, and after his departure Giil§ah recites another using the same rhyme. The joining of two ghazals in a unit bound by a common rhyme (indicated in the analysis below by a bracket) appears to be an innovation of Yusuf that was widely practiced by later Turkish poets. Otherwise, Yusuf's ghazals are as simple and straightforward as 'Ayyuqi's, and his handling of them displays little more artistry than his predecessor.

' For the date, see my review of Yusuf-i Meddah (ed. G. M. Smith), in Journal of Turkish Studies (TUBA) 2 (1978): 146. For the meters, see list at end of the article.

210 1.

FROM

MAHMUD

K A§ G A R 1 TO

EV L i Y A £ ELEB i

Gul§ah, bound by Beni 'Amr, cries out to God to save her (line 125)

| 2.

Varqa, departing for Y e m e n , recites a l o v e p o e m to Gul§ah (line 3 3 1 )

[ 3.

Giil§ah b e m o a n s Varqa's parting (line 3 4 6 )

4.

Varqa, in the hands of 'Anter's black slave, recites a l o v e p o e m for Gul§ah (line 6 6 7 )

5.

Gul§ah, betrothed to king Muhsin, g i v e s Varqa's ring to her slave-girl

6.

Varqa, b e l i e v i n g Gul§ah dead, is prevented from killing h i m s e l f , and

7.

Varqa laments to Giil§ah their sad plight (line 1 3 4 7 )

and cries out to God lamenting her fate (line 9 5 3 ) recites a

love p o e m (line 1 0 3 1 ) .

I 8.

Varqa, about to part from Giil§ah, recites a love p o e m (line 1402).

I. 9 .

Gul§ah's reply (line 1 4 1 0 )

10.

Varqa b e g s God to let him die since he cannot live without Gul§ah (line 1442)

II.

Giil§ah, at Varqa's grave, recites a lament before killing herself (line 1517)

Before embarking on a deeper analysis of these two works 1 it will be useful to give a brief summary of the story. Varqa and Gulshah, members of the same noble Arab tribe and inseparable companions since childhood, are betrothed to each other, but on their wedding night Gulshah is carried off by the chieftain of a rival tribe. Varqa rescues her, but his father is killed in the ensuing conflict. Now the king of Syria persuades Gulshah's parents to betroth her to him instead of Varqa. Though Gulshah accompanies the king to Syria, she forces him to accept a Platonic relationship, and she remains faithful to Varqa. Gulshah's family, meanwhile, buries a sheep and tells Varqa it is the grave of Gulshah. Varqa, after mourning deeply, learns the truth and goes to Syria. The king permits the lovers to meet, but Varqa feels he must leave Gulshah to her husband. In his wandering, he meets a physician who prescribes union with the beloved as the only cure for his sorrow. This being out of the question, Varqa finally dies of grief. Hearing of his death, Gulshah also expires (in Yusuf's version, she kills herself). The different circumstances in which we find ghazals used in the early part of the story (A.l-5, Y.l-4) reflect the different approaches of the respective authors. 'Ayyuqi's poem, written in the meter of the Shah-nameh, has a distinctly heroic flavor. One of the first ghazals is the "battle boast" of Varqa's rival and is strongly reminiscent of pre-Islamic Arabic poetry. Another is the poem Varqa recites eulogizing the manly qualities of his fallen father. Throughout, the love story is subordinate to the heroic aspects of the ' i he analysis which follows is mainly the work of Judith Wilks.

THE

LYRIC

IN T H E

ROMANCE

211

narrative. Gulshah recites only two of the ten ghazals in the work, and two others are recited by Varqa's rival. Yusuf's poem is written in the meter of the Masnavi of Rumi, whom Yusuf greatly admired (see line 1604), and it plays up the romantic and ethical aspects of the story. Varqa's rival is rather less noble than his counterpart in 'Ayyuqi: he recites no ghazals, and in his speeches, after praising Gulshah, he later makes vile threats against her (lines 97-117). Varqa's tribute to his father's heroism is reduced to a few scattered lines (185, 194-95). The role of Gulshah is greatly enlarged: she recites five of the eleven ghazals, including the first and the last. The differences in the later part of the story are less apparent. Even the circumstances for reciting ghazals coincide somewhat. In both versions, ghazals occur at the two most crucial events: Gulshah's betrothal (A.6, Y.5) which leads to Varqa's death; and Varqa's death (A.9, Y.10) which leads to Gulshah's death and the end of the story. The remaining ghazals, however, occur at slightly different points in the two narratives. Compare the poems recited upon the report of Gulshah's death to Varqa and his visit to the false grave (A.7, Y.6) and the ones recited by Gulshah when she hears of Varqa's death and visits his grave (A.10, Y.ll). In both instances, 'Ayyuqi inserts the ghazal at the point the character receives the news of the beloved's death and feels the first pang of grief, and gives the graveside speech a little further on in masnavi form. Yusuf, in contrast, places the ghazal at the most dramtic point in the action, the graveside scene itself. Yusuf's ghazals generally serve to highlight important plot developments as well as allowing the characters an impassioned outburst. While some of 'Ayyuqi's ghazals occur at crises (A.6,9), more typically they are placed a little apart from a major event (A.7,10) or are inserted at times of high emotion for the characters but of less than crucial importance in the plot of the romance (A. 1-5,8). 'Ayyuqi seems to have considered the insertion of a ghazal at a crisis to be optional. The ghazals serve to give the reader or listener an occasional glimpse into the character's inner state at various points in the story, whether critical or not; and the characters often seem to be reflecting aloud, rather like an "aside" in drama, not caring whether anyone else hears their laments. (Even Varqa's plea to the physician [A.8] is basically just another review of his troubles, only partially intended for the physician's information.) The ghazals in Yusuf's poem on the other hand, occurring invariably at crises in the action, often elicit a real response from another character. Thus, Gulshah cries out to God to send Varqa (Y.l), and He does. Y.2-3, 8-9 constitute pairs, Gulshah responding to Varqa's love poems. In Y.4 Varqa's former slave recognizes him as a result of the ghazal recital and so spares his life; and in Y.10 Varqa cries to God to let him die, and God obliges.

212

F R O M M A H M U D K A § G AR I T O E V L I Y A

gELEBi

The two poets appear to use a similar formal device to link some of the ghazals. In 'Ayyuql, the radif (refrain element after the rhyme) in the first ghazal recited by each of the three major characters ('Adnan, the rival chieftain — A.l; Varqa — A.2; Gulshah — A.6) is -e man, meaning "mine." The rivalry between 'Adnan and Varqa is thus established at the outset by using the same radif in their speeches addressed to Gulshah, and her using the same radif in her first ghazal relates it directly to the other two, even though it occurs quite a while after them. Another link is provided by the word zinhar "beware!," repeated twice in the last couplet of both Varqa's and Gulshah's ghazals. We also find the -e man line-ending at crucial points of the narrative masnavi, seemingly as echoes of its use in the ghazals mentioned above. Thus, Varqa's second ghazal (A.3) is preceded by a masnavi couplet with both hemistychs ending in -e man. After Gulshah's first ghazal she resumes speaking two lines later with an -e man rhyme in the first couplet. In fact, the first verse spoken by Gulshah in the entire poem has -e man ending each hemistych (p. 54, line 16). When Varqa mourns at the false grave his speech ends with an -e man couplet (p. 82, line 13), and his passionate speech before his final ghazal (A.9) begins with an -e man couplet (p. 109, line 4). When Gulshah's husband, the Syrian king, expresses concern for her after her ghazal of grief (A. 10) the first couplet of his speech also ends in -e man, thus linking him to the others involved in the romance, and poignantly indicating his own small claim. Gulshah's speech at the final parting scene includes two consecutive couplets with -e man (p. 100, lines 6-7), and several lines of her lengthy death speech (pp. 114-115) have the same feature. While it could be argued that the first person possessive form is a very simple and common way to end a line in love poetry, and that therefore all these "links" among the speeches and the ghazals are accidental, the recurrence of this feature throughout the poem rather suggests that 'Ayyuqi used it intentionally, perhaps to bring the romantic relationship among the characters into sharper focus against the background of his heroic story. Yusuf's partial imitation of this pattern confirms such an opinion. The first ghazal pair (Y. 2-3), besides sharing a common rhyme, also share a common radif — beniim, meaning "mine." And in Varqa's next lament (Y.4) the masnavi lines immediately preceding and following the gazal (lines 666, 676) and comprising part of Varqa's speech have beniim at the end of each hemistych; this device forms a frame for this ghazal and also echoes the two provious ones. In sum, while Yusuf's use of inserted ghazals was plainly modeled on 'Ayyuqi's, he assigned a much more speecific role to the inserted lyric than did his predecessor. Many later Turkish authors of romantic masnavis

THE

LYRIC

IN

THE

213

ROMANCE

followed Yusuf's example and used the inserted ghazals in this more clearly defined way: placing them at critical points in the narrative, and even assigning them a role in the plot; sometimes connecting them strongly to each other by a common rhyme (and later, a common meter) and using them for impassioned dialogue between the lovers. *

Between Yusuf-i Meddah and §eyhi we have a group of three Turkish romances whose Persian models, if they existed, are unknown. All have interspersed ghazals, and all show the further feature of varying the meter, an innovation that was not present in Varqa ve Giil§dh. (They are also all considerably longer than that work, which had only 1,600 verses; and they all have happy endings.) A fourth, the episode of Iskender and Giil§ah in Ahmedi's Iskender-name, must also be considered along with this group. The first, written just a few years after Yusuf's work, is Mes'ud ibn A h m e d ' s Stiheyl ii Nevbahar (751/1350). The masnavi is in miitekarib. Of the 14 ghazals, 6 are hezec 2 (8, 9, 11-14), 3 muzari' 1 (1, 4, 7), 3 miictess (2, 5, 6), 1 remel long (3), 1 muzari' 2 (10). Their occasions and distribution are as follows: 1.

Siiheyl

gets drunk

in the F a g f u r ' s court

and

recites

a ghazal,

accompanying himself on the kopuz (p. 79) 2.

Suheyl's second recitation in the court, which nearly betrays his love for the Fagfur's daughter, Nevbahar (p. 82)

3

Siiheyl tunes his kopuz to "Neva" and sings, hoping that Nevbahar will be drawn out by his voice to the palace roof (p. 91)

4.

Suheyl's second attempt to draw out Nevbahar (p. 95)

5.

Suheyl's third attempt to draw out Nevbahar (p. 98)

6.

Siiheyl recites a poem {§i'r)

describing Nevbahar's beauty when she

departs (p. 101) 7.

Suheyl's fourth attempt to draw out Nevbahar (p. 107)

8.

Siiheyl finally sings (yirla-)

9.

directly to Nevbahar (p. 108)

Siiheyl, while merrymaking with Nevbahar, takes her kopuz and recites a poem at her request (p. 119)

10.

Siiheyl's second song at Nevbahar's request (p. 120)

11.

Nevbahar, escaping from Sa'luk, recites a lament (sagu) (p. 165)

12.

Nevbahar's second lament (p. 189)

| 13.

Siiheyl appears to Nevbahar in her dream and recites a poem (p. 192)

| 14.

Nevbahar, still in her dream, replies (p. 193)

214

F R O M M A H M U D K A § G A R I T O EV LI Y A g E L E B l

AH 14 ghazals are compressed into the first half of the poem. The lovers' adventures continue without lyrical interruption for another 175 pages or so (the entire work is 5,568 verses). This imbalance may be considered a weakness. On the other hand, the ghazals in the first half fall into two clearly defined groups, each group closely wedded to its context. The first 10 lead up to the climax of the love-story, the union of the lovers, graphically described in the love-making scene immediately after ghazal no. 10 (p. 121). Siiheyl is portrayed as an expert minstrel, while his initial ineptitude at promoting his love-interest wins him the reader's sympathy. (His singing lures out Nevbahar after ghazal no. 4, but when she disappears, he falls into the water and has to be fished out.) The 10 ghazals, all uttered by Siiheyl within 40 pages of the text, not only express his passion but are intricately enmeshed in the story. In the other group, the last 4 ghazals, the focus of the story has shifted to Nevbahar and her escape from her abductor. The 14 taken together culminate in a pair with the same rhyme and meter. The second of this group is § e y h o g l u ' s Hur§ld-name (789/1387), which has been published recently. In 7,900 verses there are 22 1 ghazals and 1 terci'-i bend. The masnavi is in hezec 1, and nine different meters are used in the lyrics: hezec 1 (1, 8, 9-10, terci'-i bend), hezec 1 long (2), hezec 3 (1516), miictess (3, 13-14, 22), muzari' 1 (4-5), remel 1 long (6-7), remel 2 (1718), remel 3 (12), recez 2 (11), mutekarib (19), miinserih 20-21). In over half of these (8-10, 13-14, terci'-i bend, 15-18, 20-22) we find the innovative feature of a makhlas: the character reciting the poem includes his name in the final verse (in the terci', in each stanza before the bend). Ghazals are paired in 7 instances: in one pair (15-16) the rhyme is constant, the radif varies; in another (20-21) part of the radif is constant, the rhyme varies; in the remainder, there is no variation. 1. 2.

King Siyavu§ discovers the beauty of his daughter Hur§id (lines 961-75) Azad tells Ferah§ad of his love for Hur§id (lines 1824-30)

3.

Gulam-i Divane tells Ferah§ad of his love for Hur§id (lines 1939-47)

f 4. | 5

Ferah§ad pleads with God to end his separation from Hur§id (lines 2553-59) Hur§id's reply (lines 2518-24).

J

Ferah§ad sends Hur§id a love poem (lines 2740-46) Hur§id's reply (lines 2761-67)

I78.

I I

Hur§Id bemoans her separation from Ferah§ad (lines 2861-69) Ferah§ad proclaims his love to Hur§id (lines 3043-51) Hur§id's reply (lines 3055-63)

' t h e editor claims that there are 23, but one of these (lines 5402-6), rather than being a ghazal, is simply masnavi employing rhetorical takrir (cf. lines 3302-5, etc.).

THE LYRIC 1 1. 12. | 13. | 14.

Ferah§ad

IN T H E

ROMANCE

215

sings to Azad of his love for Hur§id (lines 3100-08)

Hur§id sings to her nurse of her love for Ferah§ad (lines 3143-49) Hur§id's mutribs (minstrels) sing her song (lines 3335-43) Ferah§ad's reply (lines 3357-63)

Terci'-i bend: Hur§id sends Ferah§ad a long love poem (lines 3749-88) f 15. Ferah§ad bids farewell to Hur§id (lines 6089-97) I 16.

Hur§id bids farewell to Ferah§ad (lines 6102-10)

f 17.

Hur§id despairs of Ferah§ad's coming (lines 6698-6704)

I 18.

Ferah§ad's reply which Hur§id sees in her dream (lines 6759-67)

19.

Ferah§ad's dying father gives advice to his son (lines 6920-32)

I 20. I1 21 ' 22

Hur§id sends Ferah§ad a love poem (lines 7220-36) Ferah§ad's reply (lines 7265-73) Ferah§ad bemoans his separation from Hur§id (lines 7405-14

Once again, a clearly demarked set of ghazals (nos. 4-14) leads up to a love scene which follows immediately upon the last one (lines 3472 ff.). The first of these (no. 4) plays a role in the plot, since Ferah§ad's servant Azad writes it down and has it delivered to Hur§id, whose reply (no. 5) initiates a series of impassioned lyric exchanges: a second one by letters (6-7), one face-to-face at the lovers' initial meeting (9-10), and one at the elaborate party or meclis just prior to their union (13-14). These are interrupted by two interludes (8, 11-12) in which the lovers complain of their separation. The eleven ghazals taken together form a rhythmic counterpoint to the action. The terci'-i bend which follows them is also situated at the mid-point of the narrative, and serves to round out the first section. The next 2,000 verses, describing wars and heroic adventures, have no lyrical interruptions. Only toward the end, when the lovers once again part so that Ferah§ad may fulfill the heroic quest set for him in order to obtain Hur§id in marriage, is there another set of ghazals. Here, Hur§id's lament (also called sagu, line 6697) and the reply which she reads in her dream (17-18) are reminiscent of the similar exchange which we saw in Siiheyl u Nevbahar. The third in this group of "potboilers" is Mehemmed's 'I§kname (800/1397), composed in hezec 1, and consisting of 8,700 verses with 2 gaps in the text. The meters used in the 30 ghazals are: hezec 1 (1, 4, 12-13, 20, 21-22, 30), muzari' 1 (10-11), munsarih (14-15), and miictess (the remainder). Nine times the ghazals are paired: in each pair the same meter is used and, in all but 3 cases, the same rhyme; in those 3 (10-11, 14-15, 25-26), the same radif (cf. line 2036: "Ferruh kept the radif but changed the rhyme"). All the ghazals include a makhlas, being the name of the character uttering the ghazal.

216

FROM

M A H M U D

K A § G A R Ì

TO

E V L Ì Y A

1.

Ferriih on arriving in China (line 985)

2.

Hiimâ on awakening f r o m a dream (line 1015)

3.

Ferruh on arriving at Hurrem-âbâd (line 1125)

4.

Hiimâ after a sleepless night (line 1174)

ÇELEBi

5.

Hiimâ longing to meet Ferriih (line 1381)

6.

Ferriih longing to meet Hiimâ (line 1671)

7.

Hiimâ in exceptation of Ferriih (line 1706)

8.

Ferriih on reaching H i i m â ' s garden (line 1729)

9.

Ferrûh on waking f r o m his fainting fit and finding Hiimâ gone (line 1858)

I 10. j ] i

Hûmâ rejecting her nurse's counsel (line 1996) Ferrûh rejecting his f r i e n d ' s counsel (line 2038)

I 12.

Ferruh writes a love poem to Humâ (line 2109)

I 13.

Humâ, taking it f r o m the fainting Ferriih, writes a reply (line 2129)

I 14.

Ferrûh on attaining union with Hiimâ: to the night o f pleasure (line 2281)

[ 15.

H u m â ' s reply (line 2293)

I 16.

Hûmâ thanks G o d f o r providing union with Ferrûh (line 2 4 0 4 )

I 17.

Ferrûh's reply (line 2413)

I 18.

Ferriih sends a message to Hûmâ to bear up in times of distress (line 2 6 6 2 )

j 19.

Hiimâ sends her reply (line 2692)

20. I 21.

Ferrûh in H u m à ' s garden sings and drinks himself to sleep (line 2721) Ferruh on being reconciled with Hûmâ (line 3376)

j 22.

H û m â ' s reply (line 3 3 8 6 )

j 23.

Humâ grieves at being g i v e n to ' A l e m § â h (line 3 5 5 1 )

j 24

Ferrûh's reply (line 3573)

f 25.

Ferriih grieves at being separated f r o m Hûmâ (line 3707)

j 26.

H û m â ' s reply (line 3724)

27.

Hûmâ on escaping f r o m her N e g r o captors bewails her plight (line 4265)

I 28.

Ferruh on reunion with Humâ after many adventures (line 6474)

j 29.

H û m â ' s reply (line 6484)

30.

Ferrûh on again being reunited with Hûmâ (line 8075)

Ahmedï's

ïskender-nàme

differs from all the other masnavis

considered here in that it is not primarily a romance at all but a didactic Bildungsroman.

iskender's wooing and winning of Giil§ah, daughter of king

Zeresb of Seyistân, occurs early on in the work, after his conquest of Iran and before his conquest of India and the more discursive sections of the book. Ahmedi apparently thought that iskender's education required a romantic adventure in the early part of his career; and this naturally included an exchange of ghazals. At the conclusion of the episode, Ahmedi claims that this section of his work reveals the secrets of love, compares it favorably with

T H E

L Y R I C

IN

T H E

R O M A N C E

217

Nizami's stories of Leyll and Mecnun and of Ferhad and §irin, states that there are many 'I§knames (a reference to Mehemmed's work?) but none like this one, says that anyone who reads it will forget Cem§id and Hur§id (a reference to his own work by that name?), and says, furthermore, that he wrote it in two days and that it contains 605 "jewels" or verses. 1 The six ghazals are in the same meter as the masnavi (remel 1) with no makhlas. 1. Gul§ah sends a love poem to iskender written on silk (line 1434) 2. iskender's reply (line 1463) 3. iskender's love poem in his letter of separation (firak-name) to Giil§ah (line 1600) 4. Giil§ah's love poem in her letter of separation to iskender (line 1659) 5. Ghazal of iskender (line 1813) 6. Ghazal of Giil§ah (line 1874) *

In his Husrev u §irin (ca. 828/1425), §eyhi insists (line 567) that although he is a master of the ghazal, "it is the masnavi which is the touchstone of the coin of eloquence." Still, he cannot keep himself from introducing 27 ghazals into this long poem, thus aligning it with the mainstream of Turkish romances, rather than with his predecessor as Nizami's adaptor in Western Turkish, Fahri (768/1367). §eyhi's ghazals are again without a makhlas, except for two of them (nos. 1, 20), where the name, however, is that of the author, not the character uttering the ghazal. §eyhl's makhlas is also found in the final stanza of the long terci'-i bend (lines 47814850), which is otherwise put in the mouth of Ferhad; and in the miinacat (lines 5453 f.), which is uttered by §irin. This practice, which was followed consistently by Fuzuli (see below), is the final innovation in the tradition of ghazal insertions into the romantic masnavi. §eyhi uses 10 different meters, as follows: remel 1 long (12, 14, 17, 19, 27), remel 2 (10-11), hezec 1 (8-9), hezec 1 long (6-7, 22, 26), hezec 2 (2), hezec 3 (terci'-i bend), muzari' 1 (1, 3, 13, 20, 21, miinacat), miictess (4-5, 16, 24, 25), hafif (18, 23), mutekarib (15).

In the recently pulished facsimile (ed. Ünver) the episode covers lines 1,331-1,933 in a text of 8,754 verses. Some manuscripts of the work omit the first 2 of the 6 ghazals, and in these same texts Ahmed! states that the episode contains 551 rather than 665 "jewels." This entire problem will be dealt with in a separate article.

218

FROM

MAHMUD

K A § G A R Ì TO EVLÌYA

Ç ELEB Ì

1.

§âvùr's description of §ïrïn (line 1049)

2.

Çïrïn's mutrib, from her tongue, when she falls in love wth Husrev's picture (line 1308)

3. I 4. I 5.

I 6. I 7.

Husrev's mutrib, from his tongue, when he drinks after arriving in Armenia and finding §irin gone (line 2021) The minstrel Barbed from the tongue of Husrev (line 2985) The minstrel Nigïsà from the tongue of §ïrïn (line 3006)

Barbed from the tongue of Husrev (line 3021) Nigïsà from the tongue of §irin (line 3032)

I 8. I 9-

Barbed from the tongue of Husrev (line 3124) Nigïsà from the tongue of §ïrin (line 3133)

I io. I J]

Bârbed from the tongue of Husrev (line 3214) Nigïsà from the tongue of §ïrïn (line 3224)

12.

Husrev bemoaning his separation from §ïrïn (line 3549)

13.

Barbed from the tongue of Husrev, longing for §ïrïn (line 3885)

14.

§ïrïn longing for Husrev (line 3924)

15.

Mehin Bânû dying (line 3970)

16.

§irin longing for Husrev (line 4047)

17.

§âvûr from the tongue of §irin to Husrev (line 4160)

18.

Barbed from the tongue of Husrev (line 4181)

19.

Bârbed from the tongue of Husrev (line 4189)

20.

Ferhâd to §irin (line 4365)

21.

Ferhâd mad for §irin (line 4418)

22.

Ferhâd mad for §ïrïn (line 4485)

Terci'-i bend: Ferhâd on §irin's visiting him at the mountain (line 4781) Miinâcât: §irin begs God for mercy (line 5453) 23.

§irin on Husrev's coming drunk to her castle (line 5574)

24.

Nigïsà from the tongue of §ïrïn (line 6125)

25.

Barbed from the tongue of Husrev (line 6170)

26.

Nigïsà from the tongue of §ïrïn (line 6215)

27.

Bârbed from the tongue of Husrev (line 6277)

A s is apparent, §eyhi takes every opportunity to introduce a ghazal into the narrative; he even doubles N i z â m ï ' s f a m o u s s c e n e in w h i c h the minstrels, Bârbed and Nigïsà, serve as mouthpieces for the two lovers. Hamdullâh Hamdl, b e s i d e s u s i n g ghazals in his Leyll

ve

Mecnùn

( 9 0 5 / 1 4 9 9 - 1 5 0 0 ) as had §âhidï in his version of the same romance entitled Giil§en-i

'U§§dk ( 8 8 3 / 1 4 7 8 - 7 9 ; see Levend, pp. 108-32, pp. 160-76),

also

THE

LYRIC

IN

THE

ROMANCE

219

introduced them into his romance based on the Qur'ànic story of Joseph.' Hiiseyin Ayan, who has analyzed the ghazals in H a m d ï ' s Yûsuf ve Zelïhâ (897/1491-92), states (p. 32) that they occur at the most poignant points in the narrative, allow a peek into the characters' spiritual state vis-à-vis the action of the poem, and give the reader a chance to pause for breath and return to the masnavï with renewed interest and commitment. None of the 15 ghazals is uttered by Yiisuf himself. The first 3 are plaints of longing of Y a ' k û b for his son; the fourth, the Egyptians' song of praise of Y u s u f ' s beauty; the fifth though the ninth, poems of longing of Zelïhâ who has seen Yiisuf in a dream, and who later bemoans her bad fortune; the tenth, a sigh of regret by the caravan leader for having sold Yusuf to Zelïhâ; the eleventh, Bâziga's cry of frustration at her failure to distract the bidders so that she could herself buy Yûsuf; and the twelfth through the fifteenth, again poems of Zelïhâ who, rejected by Yûsuf, pours out her love and expresses her suffering. Hamdï's artlessness recalls 'Ayyùqï. The meters used are hafïf (masnavï, no. 1), remel 1 (no. 3, the tuyug [quatrain] of wisdom-advice on love occurring between nos. 8 and 9), hezec 1 long (no. 9), remel 2 (no. 13), muctess (nos. 2, 4, 7, 14, 15), and muzâri' (nos. 5, 6, 8, 10, 11, 12). Gibb (vol. 2, p. 207 = vol. 6, p. 77; no. 112) provides an example of a ghazal (= Ayan's no. 5) in context. In contrast to §eyhï, Fuzûlï seems to have held the ghazal in higher regard than the masnavï. In the preface to his Turkish Divan he envisions (p. 4) a "paradise-betokened" and "heart-refining" school (somewhat like the school where Mecnùn and Leylï fall in love): A charming courtyard, with beauties seated row on row, Their faces radiant-blessed, like lines of scripture...

But since the adolescent denizens of this celestial academy are yet innocent of the fine points of the religious sciences, "nothing was recited in that paradisical assembly except love poems, and no writing was to be found upon the pages they were reading except heart-igniting ghazals." (p. 6) The ghazal gives delight to the people of insight; The ghazal is the rose of the garden of excellence. Hunting the gazelle of the ghazal is not easy; The people of mystical knowledge ( ' i r f d n ) do not deny the ghazal. The ghazal discloses the poet's powers; The ghazal increases the poet's reputation. Although the heart be disposed to many kinds of poems Choose from among them the kind that is the ghazal! For the ghazal is the adornment of every assembly; The ghazal is the craft (son'at) of the people of wisdom... ' T h e later version of this romance by Yahyâ Bey (d. 983/1575), recently published, lacks ghazals. The earlier one by Mustafa Darir (768/1366) contains them (Iz, pp. 591-605), all in the same meter as the masnavï (remel 1).

220

FROM

MAHMUD

K A § G A R I TO E V L Î Y A

ÇELEBI

Fuziili's Ley II ve Mecnùn (963/1556) represents the culmination of the Turkish romantic masnavi tradition. There are 22 ghazals, 2 murabba's, and 2 munâcâts. The use in all of the lyrics of the poet's own makhlas is consonant with Fuzull's tendency in the poem as a whole: to raise the personal and human love-tragedy to the plane of mystical longing and ethereal aspiration. While generalizing this aspect, which was begun by §eyhi, Fuzuli reduces the number of meters, using only remel 2 (1, 7, 9-10), hezec 1 long (3, 8, second murabba', 12, 18-19, 21), hezec 2 (4), muctess (11) muzâri' 1 (6, first and second munàcàt), and remel 1 long (the remainder). 1. Leyli bemoans her separation from Mecnûn (line 715) 2. Mecnûn bemoans his separation from Leyli (line 789) 3. Mecnun asks his friends to excuse him for his love-madness (line 889) 4. Mecnûn's father counsels him to be reasonable (line 972) 5. Mecnûn asks his father to excuse him (line 1016) 7. Mecnûn at the K a ' b a asks God not to cure his love-malady (line 1123) 6. Leyli addressing the cloud complains of her love (line 1398) 8. Mecnùn complains to Nevfel of the faithless world (line 1485) I 9. I 10.

Leyli on seeing Mecnûn led as a chained captive (line 1641) Mecnùn, on seeing Leyli, pleads for justice (line 1670)

jj

Leyli complains on her being married to Ibn-i Selâm (line 1750)

M u r a b b a ' : Mecnûn at the end of a letter to Leyli complaining faithlessness (line 1922) Murabba': Leyli at the end of her reply excusing herself (line 1996) Munâcât: Mecnûn begs God for mercy (line 2323) 12.

Zeyd transmits Leyli's love-song (line 2379)

13.

Mecnûn on hearing of Ibn-i Selâm's death (line 2475)

Munâcât: Leyli begs God for mercy (line 2536) [ 14. Mecnùn proves w h o he is to Leyli by recounting troubles (line 2608) '•15. Leyli's reply (line 2637)

his

(16.

Mecnûn asks Leyli who she is (line 2657)

I. 17.

Leyli's reply (line 2686)

j 18. I U9.

Mecnûn insists his love is perfect without the need for physical union (line 2727) Leyli's reply (line 2767)

of

her

THE 20.

LYRIC

IN

THE

ROMANCE

221

Mecnun's mystical song (line 2809)

21.

Fuzuli on Leyli's death (line 2904)

22.

Mecnun at Leyli's grave (line 2981)

The lyric pieces are all deftly placed to break up the narrative without hindering its flow. Fuzuli distributes them with a kind of self-conscious artistry that we have not observed hitherto, employing rhythmic patterning and framing devices, and building to a climax. Three of the four sets of paired ghazals come grouped together at the climax of the narrative, when Leyli finally holds out the opportunity of union to the distraught lover, and Mecnun rejects it on the grounds that he has reached a higher plane than physical union. The pair of miinacats, employing the same meter and rhyme (and incidentally constituting a nazire on §eyhi's, noted above), is itself broken up to serve as a frame for two other ghazals. The two longer lyrical flights occur in the lovers' letters to each other, at the mid-point of the narrative. The inclusion of ghazals put in the mouth of Mecnun's father at the beginning and in the mouth of the poet (or an anonymous "wise man," hiredmend, line 2903) at the end, preserves the rhythm of these insertions, which can be outimed thus: 1-3: Leyli —Mecnun—Mecnun 4-6: Mecnun's father—Mecnun —Mecnun 7-20: Leyli —Mecnun (9 times) 21-22: Fuzuli—Mecnun

To illustrate Fuzuli's use of ghazals in his masnavi, we may take the following episode. Lines 1602-40: Mecnun substitutes himself for a chained captive. Led by an old man, he wanders near Leyli's house. When Leyli sees him in this pitiable state, she breaks forth in joyful song: (Yar rahm etdi meger nale vii efganumuza) The beloved must have had compassion on our crying and wailing That he set foot today in the hut of our sorrows. The rain of tears must have helped, since all at once This new princely rose shot up in our garden. From the moment the fire of sighs begins to burn, it is known to us That a torch has brightened our night of separation. It would have been possible to term this union a dream-state If only sleep had ever visited our weeping eyes. What we have seen must be a phantom, or else an idol; It is unthinkable that he himself should come to our side. The beloved has become our guest. Come, heart and soul, Let us expend whatever we may own for the sake of our guest. The sweetheart has designs on our life, Fuzuli. Come, Let us give up life for the sweetheart and do ourselves a favor.

222

FROM

MAHMUD

K A § G A R 1 TO E V Li Y A £ E L E B I

Lines 1 6 4 8 - 6 9 : Mecnun, s e e i n g Leyli, pours out his l o v e in the form of a plea for justice from his Sultan, culminating in this ghazal: ( K u f r - i ziilfun salali

rahneler

imanumuza)

Ever since the blackness/infidelity (kiifr) of your lovelock made breaches in our faith The infidel himself weeps at our forlorn condition. To see you appears to be impossible, since when we look Tears fill our weeping eyes. Do not be too cruel, lest the store of cruelty be used up at once, But pour cruelties little by little upon our soul. We never want for grief, so much so that whoever comes takes some From us, and leaves grieving, though he came joyfully to our side. Every link of our chain has a mouth with which Continually it reveals our grief that is concealed. Fuzuli! the grief of the days has been unjust to us; Lacking power, we have come to seek justice of our Sultan. The episode concludes with four lines of masnavi (1676-79): For a moment, crying in this manner, He sought justice of his Sultan for Love's oppression. Then he broke his chain in pieces And once again went off seeking solitude. His limbs broken and his eyes wet with tears, Disgraced and ruined, intoxicated and fearless. With gangs of children at his back, Some laughing at his condition, some weeping. *

Since f e w of the later Turkish romantic masnavis have been published, it is hard to say h o w entrenched the tradition of lyric insertions was, or what developments took place after Fuzuli. W e may note only that § e y h Galib's f a m o u s Hiisn u 'A§k ( 1 1 9 7 / 1 7 8 2 ) , which can be considered the swan song of the Turkish masnavi tradition, includes four lyric pieces in the c o m p a s s of 2 , 1 0 0 verses. They are all in the same meter as the masnavi ( h e z e c 3) and in the unusual 5-line stanzaic form, or mukhammas. 1. Baby 'A§k's nurse sings him a lullaby (lines 319-33) 2. 'A§k in love with Hiisn (lines 1188-1202) 3. 'A§k to Suhan who rescues him from difficulty and reminds him of Hiisn (lines 1527-41) 4. 'A§k recalls the happy days of first love (lines 1717-31)

THE

LYRIC

IN

THE

ROMANCE

223

These insertions serve to highlight crises in the career of 'A§k, the allegorized figure of Love, in relation to Hiisn or Beauty. The poet links them formally by using the same rhyme and rhyme-word (in§äd) in the immediately preceding verse in each case; also in the penultimate verses in the first three cases (line 317: feläket/rähat; line 1185: gäläk/efläk, line 1186: rähat/feläket; line 1525: gäläk/gäk). At the hands of a consummate artist such as §eyh Gälib or Fuzüli, the lyric insertions, while integrated harmoniously into the narrative, at the same time take on a life of their own, and can stand independently. (They are often anthologized along with the ghazals and other shorter pieces of these poets.) It is usually claimed (see, for example, Rypka, pp. 177, 259; Gibb, vol. 1, p. 309; vol. 2, p. 173) that the purpose of the ghazals in the masnavi is to break the monotony of the coupled rhymes. Undoubtedly, many of the romances (despite occasional brillant flashes [cf. Tietze]) are dull and repetitive — "von ermüdender Weitschweifigkeit," as Mordtmann characterizes Süheyl ü Nevbahär (p. 8) — and in these cases the lyrics do serve primarily to enliven the narrative and add charm to a lackluster poem. But this cannot be the reason for the original use or for the persistence of the tradition. 'Ayyüqi, who initiated the practice, certainly took his cue from his Arabic source (see above quote of Z. Safä). Also, the inclusion of lyrics in ghazal form is a natural and expected thing, especially where the hero of the story is himself a poet or minstrel. Why Nizämi chose to do without them, even in his Layli u Majnün, is a question that has not been addressed, so far as I know, by literary scholars of Persian, and I shall but suggest an answer here. Fakhruddin Gurgäni, before Nizämi, had also avoided them in his Vis u Rämin (ca. 1050) although Rämin, like Varqa, is a minstrel who composes poems to his beloved. That Gurgäni and 'Ayyüqi followed different paths might be due to their different sources, the one drawing from pre-Islamic Persian romance, the other from Arabic stories of love-poets (cf. Melikian-Chirvani, p. 30). If 'Ayyüqi and Gurgäni were the two possible models for Nizämi, it is easy to see why he chose Gurgäni, the superior poet by far. Then the great prestige of Nizämi determined the subsequent course of the masnavi tradition in Persian. As for the Turkish poets, it must be considered whether they were not disposed to lyric insertions because of a similar practice in Turkish folk literature — i.e., the interspersing of lyrics in a prose narrative. This is the practice of the Anatolian and Azerbaijani minstrels or 'A§iks in their "hikäyes," as well as the Uzbek Bakhshis in their "dästäns" (see Boratav in PTF, vol. 2, p. 31). Unfortunately, this type of folk literature is not attested before the sixteenth century. The heroic tales of Dede Korkut (fifteenth century?) employ a mixture of prose and a type of free verse; but it is hard to

224

FROM

MAHMUD

K A § G A R Í TO

EVLÍYA

QELEBÍ

tell whether the Turkish bards or Ozans had always used prose, or whether the Oghuz epics were originally verse throughout, as is the case with their modern counterpart among the Kirgiz. With the Dani§mendname (761/1360) we are pushed back to a time and milieu very close to the earliest Turkish romantic masnavis. This legendary account of the Türkmen conquests of Anatolia is in prose with n u m e r o u s verse insertions of varying length and both masnavi and ghazal/kaside in form. The verses employ the 'arüz meters and so belong to the high literary tradition. The author, 'Árif 'All, informs us (p. 289, trans, p. 459) that his prototype text had the fault, among others, of lacking verses; the clear implication being that he added them himself in order to display his poetic gifts and also, perhaps, to enliven the narrative. Another work that ought to be considered in this connection is the Battal-name (also fourteenth century?), a prose epic which does contain lyric insertions (see Kocatiirk, p. 295); but since it is unpublished except in translation (Ethé) it is hard to judge their character. 1 However, the Arabic work on which it is modeled, the interminable Sirat Dát al-Himma, employs sets of paired lyrics with the same meter and rhyme (imitated by Lane, pp. 425-26) in a fashion remarkably similar to that of the Turkish romances. 2 Of course, the inclusion of scattered verses in a prose text is a wellnigh universal phenomenon. Its relevance to our present discussion can only be tangential. In Uighur Turkic literature it is found in such Buddhist texts as the translation of the Suvarnaprabhásasütra known as Altun Yaruq (in the tale of "The Hungry Tiger," ed. Gabain, pp. 294-307), and there is a famous example in the Oguz Name (see Gabain's analysis in PTF, vol. 2, p. 220). It is characteristic of Arabic and Persian adab literature, of the Thousand and one Nights, and of such Persian prose romances as Bighami's Firüz Shah Nama (see Hanaway, pp. 19-20). In Turkish, the early Anatolian religiousdidactic text Behcetü'l-Hadayik is prose enriched by verse, as is the Hamzanáme of Hamzavi (ca. 1400: Gibb, vol. 1, p. 255). With the abundance of such examples, it is not surprising to find the Turkish poets introducing ghazals into all kinds of masnavis, not only romantic ones. As early as the thirteenth century, in the short pious poem of Ahmed Fakih entitled Kitábu Evsáfi Mesácidi '§-§erife, we find two ghazallike compositions, both in the same meter (hezec 1) as the surrounding text

The Istanbul ms. Universite no. 6353 has frequent verse interruptions of the prose text in the early part of the narrative (the first 20 folios), infrequent thereafter (for example, none between fols. 37a and 111a). Ethe, in his adaptation based on 5 European mss. and 1 Kazan imprint, strangely makes no mention or indication of any verse in the text. ^The function of the poetry in the Arabic prose epic has been analyzed by Steinbach, pp. 12126.

THE

LYRIC

IN

THE

225

ROMANCE

(lines 30-37, with makhlas; lines 79-85). They occur in versified Arabiannights adventure tales, Kemaloglu's Ferahname (789/1387: Kocatiirk, pp. 131-33) and ' A b d i ' s Cdmesp-ndme (833/1430: Gibb 437). At the end of the ninth/fifteenth century Hamdi introduced them into his Mevlid-i Nebl (Gibb, vol. 2, p. 191), although they are absent from the well-known Mevlid of Siileyman gelebi (812/1409). In sum, the practice of inserting ghazals in the masnavi, with its modest beginning in Persian romance, quickly faded in the Persian tradition itself, but flowered at the hands of the Turkish poets. This paper is intended as an initial survey of the field and a first stab at some of the related literary and historical questions. THE M E T E R S remel 1 remel 1 long remel 2 remel 3 hezec 1 hezec 1 long hezec 2 hezec 3 muzäri' 1 muzäri' 2 müctess hafìf recez 2 münserih mutekärib

_u

(J

(J_

_ u

u

u

uu —u u u — u— u u

u

(J_

u u— u u — u— u u

u u u u uu u u u u —u— u u —u— u —u—u u u —u u —u u — u— u u u—u— u u — —u u — u—u u — u u — u — u— u u — u —u —uu u uu u— u u u u— u

u

W O R K S CITED PERSIAN 'Ayyuqi. Varqa u Gulshah. ed. Z. Safa. Tehran, 1964. Maktabi. Layli u Majnun. Ed. I. Ashraf. Shiraz, n.d. Salman Savaji. JamshlcL u Khurshid. Ed. F. Vahman and J. P. Asmussen. Tehran, 1969.

226

FROM

MAHMUD

K A § G A R i TO E V L i Y A

gELEBi

TURKISH Ahmed Fakih. Kitäbu Evsäfi Mesäcidi' §-§erife. Ed. H. Mazioglu. Ankara, 1974. Ahmedi. Cem§id ü Hur§id. Ed. M. Akalin. Ankara, 1975. Ahmedi. Iskender-näme.

Ed. i. Ünver. Ankara, 1983.

'Ärif 'Ali. Däni^mendnäme. Fahri. Husrev

1 vols. Ed. I. Melikoff. Paris, 1960.

u Slrln. Ed. B. Flemming. Wiesbaden, 1974.

Fuzüli. Leylä He Mecnun.

Ed. N. H. Onan. Istanbul, 1956.

Fuzüli. Tiirkge Divan. Ed. K. Akyüz et al. Ankara, 1958. Mehmed (Mehemmed). I^k-näme.

Ed. S. Yiiksel. Ankara, 1965.

Mes'üd ibn Ahmed. Süheyl ü Nevbahär.

Ed. J. H. Mordtmann. Hanover, 1925.

§eyh Gälib. Hiisn ü A§k. Ed. A. Gölpinarli. Istanbul, 1968. §eyhi. Husrev ii §irin. Ed. F. K. Timurta§. Istanbul, 1963 and 1980. §eyhoglu Mustafa. Hur§id-näme. Yahyä Bey. Yüsuf ve Zelihä.

Ed. H. Ayan. Erzurum, 1979.

Ed. M. Qavu§oglu. Istanbul, 1979.

Yüsuf-i Meddäh. Varqa ve Gül§äh. ed. G. M. Smith. Leiden, 1976. SECONDARY WORKS Ate§, A. "Mesnevi," Islam

Ansiklopedisi.

Fasc. 80, pp. 127-133.

Ayan, H. "Hamdulläh HamdT'nin Yüsuf u Züleyhä Mesnevisindeki Gazeller." Edebiyat

Fakiiltesi

Ara^tirma

Dergisi

5 (1972): 31-49 (Erzurum, 1974).

Banarli, N. S. "Ahmedi ve Däsitan-i tevärih-i mülük-i äl-i osman." Mecmuasi Ethe. H. Die

Türkiyat

6 (1936-39): 49-135. Fahrten

Sittenroman.

des

Sajjid

Batthäl:

ein

alttürkischer

Volks-und

2 vols. Leipzig, 1871.

Gabain, A. Alttürkische

Grammatik.

3. Auflage. Wiesbaden, 1974.

Gandje'i, T. "The Genesis and Definition of a Literary Composition: The Dahnäma ('Ten love-letters')." Der Islam 47 (1971): 59-66. Gibb, E. J. W. A History

of Ottoman

Poetry.

6 vols. London, 1900-9.

Hanaway, W. L. Jr. Love and War: Adventures Sheikh iz. F. Eski

Bighami.

Türk Edebiyatinda

Kocatürk, V. M. Türk Edebiyati Lane, E. W. Manners Levend, A. S. Arap,

from

the Firuz Shäh Näma

of

New York, 1974. Nazim.

2 vols. Istanbul, 1967.

Tarihi. Ankara, 1964.

and Customs

of the Modern Egyptians.

Fars ve Türk Edebiyatlarinda

Leylä

New York, 1908. ve Mecnun

Hikäyesi.

Ankara, 1959. Melikian-Chirvani, A. S. Le roman de Varqe et Golsäh [1970]).

... (Arts

asiatiques

22

THE

LYRIC

IN

THE

Mordtman, J. H. Introduction to Mes'üd, Süheyl ü Nevbahär PTF, vol. 2: Philologiae Rypka, J. History

Turcicae

Fundamenta.

of Iranian Literature.

arabischen

Volksroman.

(above).

Vol. 2, Wiesbaden, 1964.

Dordrecht, 1968.

Safä, Z. Introduction to 'Ayyüqi, Varqa u Gulshäh Steinbach, U. Dät al-Himma:

227

ROMANCE

(above).

Kulturgeschichtliche

Untersuchungen

zu

einem

Wiesbaden, 1972.

Tietze, A. "Mehemmeds 'Buch von der Liebe': ein alt-osmanisches romantisches Gedicht" in W. Hoenerbach, ed. Der Orient in der Forschung: für Otto Spies. Wiesbaden, 1974, pp. 660-85. Yüksel, S. Introduction to Mehmed, l^k-name

(above).

Festschrift

15. THE ROMANCE OF ÏSKENDER AND GUL§ÂH

Among the episodes in Ahmedi's Iskender-name that are not to be found in the Persian prototypes of that work, and can therefore be termed "original", is a romantic interlude in which iskender falls in love with Giil§ah, the daughter of Zeresb, king of Sisitan (Seyistan). The episode occupies roughly 600 verses out of a total of 8750, or about 7% of the whole. 1 It will be useful to begin my analysis of this romance by quoting in full the judgment of E.J.W. Gibb: 2 This interlude I strongly suspect to be, if not altogether apochryphal, at any rate an after-thought and no part of the original scheme. My reasons f o r this opinion are these: 1, This dastan, and this alone, is omitted f r o m many mss. 2, Nothing in any way corresponding to it occurs in the S h a h - N a m e version of the story, w h i c h A h m e d i e l s e w h e r e f o l l o w s generally. 3, The literary style is different from that of the rest of the work; here alone are ghazels interspersed through the mesnevi, in the fashion of S h e y k h i and later poets. 4, It is a w k w a r d l y interpolated, h a v i n g no c o n n e c t i o n w i t h the story proper, and b e i n g e v i d e n t l y introduced merely for the sake of the love element which is otherwise quite unrepresented in the Iskender-Name.

Although in my opinion Gibb was wrong on each of these points and in his judgment as a whole, it must be admitted that the romance does at first glance seem out of character with the rest of the work, which is wholly didactic in nature. 3 Though the suggestion that is is "apocryphal" — i.e., not the work of Ahmedi — is to be rejected out of hand, it is possible that it was "an afterthought and no part of the original scheme." Even so, it does comprise an important part of Ahmedi's maturest version of his work, and therefore deserves to be studied within the context of the work as a whole. Although the Iskender-name has not yet been edited, we are nevertheless in a position to dispose of Gibb's arguments: 4

' Ahmedi, iskender-name : inceleme-Tipkibasim, ed. Ismail Ûnver (Ankara, 1983), fols. 12b17b (lines 1331-1933). Unver comments on the episode in his introduction, pp. 12, 23. A History of Ottoman Poetry, vol. 1 (London, 1900), p. 272 n. 3 See Unver, p. 13, on the didactic character of the work. 4 W e are in this position mainly due to Unver's efforts. Unver promises a critical edition, forthcoming. 2

230

F R O M M A H M U D K A § G AR 1 T O E V L l Y A

CELEBI

1) That this episode is not found in some mss. 1 merely indicates that it did not occur in the earliest redaction or redactions of the work. It is one of the additions to the text that Ahmedi made after 792/1390, the date of the first redaction. 2 In fact, based on Gibb's account and on my own examination of six of the Istanbul mss., I hypothesize four stages in Ahmedi's treatment: A, The episode does not occur. B, The episode occurs, with no ghazals and no concluding remarks. 3 C, The episode occurs with four ghazals, and in the concluding remarks the author states that it occupies 551 verses. 4 D. The episode occurs with six ghazals, and the author states that it occupies 605 verses. 5 This can only be a provisional typology. Only a thorough study of all the mss. and their interrelationship will allow us to determine categorically the progression in Ahmedi's conception of the work and his attitude toward his hero's love-adventure. 2) The romantic interlude does in fact correspond to something in the versions of Firdawsi and Nizami, viz. Iskandar's marriage of Rushanak (Roxana), the daughter of Dara (Darius). 6 This occurs after the conquest of Iran and before the conquest of India — i.e., in the same position that Iskender's conquest of Sisitan and marriage of Giil§ah occur in A h m e d i ' s work. While in the Persian prototypes Iskandar's marriage serves the function of a political reconciliation, in A h m e d i ' s version it serves as illustration for the themes of love and mystical union. 3) The literary style is in most respects identical with that of the rest of the work. 7 As elsewhere, the poet begins the dastan with an exhordium addressed to a bird; intersperses the narration of events with long passages of a didactic character; and ends with an epilogue which draws the moral. As for the inclusion of ghazals, this was standard procedure for romances in the Turkish tradition, and began well before §eyhi. 8

Gibb says "many" mss., but does not say how he arrived at this judgement. Unver pp. 24-27 lists over 75 mss. of which 18 are in Istanbul, 14 elsewhere in Turkey, most of the remainder in libraries in Europe. "Unver p. 14 lists some of these additions, but does not include the episode discussed here. 3 m s . £elebi Abdullah 240, fol. 57b ff. (but note that one folio is missing between 58 and 59!). 4 m s . Universite TY 166, fol. 50b-72a (actually contains 553 verses); TY 409, fol. 32b-51b (519 verses). Dr. Gtinay Kut informs me that another Istanbul ms., Muallim Cevdet 0.102, belongs to this group (fol. 78-115, actually contains 550 verses, plus 49 added later in the margin). 5 m s . Universite TY 921 (published in facsimile by Unver); Universite TY 848, fol. 31a-48b (note that there is a gap in the text between 34 and 35, and thus the second ghazal is missing and there is total of only 529 verses); Laleli 1995, fol. 48a-68a (599 verses). 6 F o r convenient summaries and references, see M. S. Southgate, t r J s k a n d a r n a m a h : A Persian Medieval Alexander-Romance (New York, 1978), pp. 170, 175. Unver pp. 17-21 lists all the oints where Ahmedi departs from his Persian sources. Unver pp. 21-24 analyzes the "dastan" structure of the work. ^See R. Dankoff, "The Lyric in the Romance: The Use of Ghazals in Persian and Turkish Masnavls," Journal of Near Eastern Studies 43.1 (January, 1984).

THE

ROMANCE

OF

ÍSKENDER

AND

G Ü L § AH

231

4) The place of interpolation, and the function of the episode, were touched on under point 1) above. It is true that the "love element" is otherwise absent from the work; but this only indicates that Ahmedi felt it belonged here, not elsewhere. If Ahmedi's purpose was "to educate the perfect man" 1 using the world-conqueror as the human model, then it stands to reason that he would include a love affair in the early part of Iskender's career. As is the case with other dástáns in the work Ahmedi conceived of this one as a self-contained unit. We may note that in one other case — the episode of the Prophet's birth 2 — he again gives a verse-count at the end; and also that at the very end of the book he gives a final verse-count for the entire work, using virtually the same language in each case: (1933) derc dur bu dürcde éy pür-hüner alti yüz ü dahi bé§ dañe güher Contained in this casket, O skilled one, are 605 jewels. 3 (6615) nige dáne dur déseñ bu 'akd-i dür alti yüz ü dahi on bé§ dáne dur If you ask how many pearls are on this string, there are 615. (8753- bil ki i§bu 'ikd dürcindeki

dür

8754) 'akd-i dest ile sekiz biñ dáne dur hem iki yüz elli anunla bile rahmet aña kim güher kadrin bile Know that the pearls in the casket of this necklace are, by a hand-count, 8250. Mercy on him who knows the value of jewels. The following summary of the romance includes a translation of the opening exhordium and closing epilogue, as well as the text and translation of the six ghazals. [fol. 12b, lines 1331-38] "Sing, O love-mad bird, sing! Make all the lovers astonished. The air is mild on [the day of] Nevrüz: you too make a pleasant air on [the musical mode of] Nevrüz. Tune your instrument to the pitch of love, and attune your lament to that of the phoenix. Sing the words of love so sweetly that people will forget the suffering of Husrev [for §irin]. Play your instrument with the lament of the heart-burning one. Make the ^So Ünver p. 24. ^Cf. f. Ünver, "Ahmedfnin ískender-námesindeki Mevlid Boliimli," TDAY-Belleten, 1977, pp. 355-411. The earlier redaction (see p. 230, n. 4) has instead "551 jewels" (bé§ yüz elli dahi hir dane güher).

232

F R O M M A H M U D K A § G AR I T O E V L i Y A g E L E B i

virgin's (or: 'Azra's) mind enamored (or: Vamik) with your words. When you begin to enchant with your song, and make the Leyli of your work love-mad (or: Mecnun), write a soul-reviving Hiimayun-name, that the phoenix (or: Humay) may be with it well-omened (hiimayun-fdl). Offer your song, whatever it be, to the sovereign (or: Cim§Id) of earth, nay, to the Venus and the Sun (or: Ziihre and Hur§id) of heaven." Following these references to famous lovers, Ahmedi begins the dastan with a comparison of love and fire, and makes clear that he is talking about love in the mystical sense: [1350-53] "When love puts an end to your existence, at that moment it conveys you to your desire. It is you who have become the veil to that beloved. Can the Beloved be seen if the veil is not removed? Let me give you some news of love, that haply it make an impression on your soul, and from that impression your soul and your heart will thrive and be full of light with Manifestation of Truth." The story proper begins at this point: [1354f.] iskender had subjugated both Iran and Turan, but Seyistan (or, as it is called here, Zabilistan) had not yet submitted. It was ruled by the great king Zeresb, scion of the hero Riistem who had ruled it in ages gone by. The Shah (iskender), longing to conquer it, disguised himself as his own envoy and made his way to the city of Sisitan. In the royal audience Zeresb was astonished at his beauty and eloquence, and doubted that he was merely an envoy: [1373] "He claims to be a slave, yet a hundred sultans would serve as vanguard to his army." Zeresb caroused with iskender for three weeks. At night, in a drunken state, he would retire to the harem and sing the envoy's praises to his beautiful daughter Giil§ah (whose charms are described at length): [1395-97] "He spoke so much of him, unbidden, that his daughter fell in love with him, unseen. The fire of love worked on her soul and overturned the kingdom of her heart. By trickery, and by effort, she managed to get a glimpse of the Shah through the skylight." Giil§ah's nurse, finding the girl sighing and weeping, brought her in from the roof top, and tried to bring her to her senses through wise counsel, but: [1415] "No fire was ever extinguished with counsel." So she turned to scheming. She found a painter and had him draw Giil§ah's portrait; [1432] "It was Giil§ah, her color and charm made visible; the only thing missing was her soul." On the back of the portrait, the princess wrote this ghazal: [143440] valihem yare 'aceb yarum kani 'a§ikam dildare dildarum kani yanaram 'i§k odina pervane-var bilmezem ol §em '-ruhsarum kani

THE R O M A N C E OF Ì S K E N D E R AND GÙ L § ÀH

233

kanumi içdi fusunile gôziih bilsem ol sermest-i hùn-hvàrum kam 'i§k bïmâr étti béni rahm édiib démedi bir gun ki bimàrum kam oh urur gézler kemânin gamzesi bilemezem ol cefâ-kàrum kam tutiyem ol §ekkeristân kanda dur bulbulem ol tàze gulzârum kam gitdi yolunda dil u cân u hired varmi varlikdan dahi âsârum kam I am distraught for the beloved, I wonder where he is. I am in love with the sweetheart, where is he? I burn in love's fire like the moth. I know not where my candle-cheeked one is. Your eye enchanted me and drank my blood. If only I knew where that blood-thirsty drunken one is. Love made me ill. It had no mercy on me, nor asked, one day, 'Where is my ill one?' His glance notches the bow and shoots arrows. I know not where that cruel one is. I am a parrot; where is that sugarland? I am a nightingale; where is that fresh rose-garden? Gone on your road are heart and soul and mind. Are there any traces of my existence? Where? They wrapped the portrait in silk, sealed the package, and had it delivered to iskender who, as soon as he saw it, fell in love with Giïl§âh. He had a eunuch brought from the harem and, plying him with silver and gold, charged him to bring the reply, which consisted of this ghazal: [1463-69] 'igkile ytizune hayrânuh benem zulfune zàr u perîçânun benem étdi cân derdiinile peymàn ebed saklayan 'ahdin bu peymânuh benem gamze okin atar kemân-ke§ ka§larun râst umaci bu peykânun benem 'i§kuhuh gencine vïrân yér gerek bende ko am ki vlrànuh benem cdna vérurlerse ayaguh tozi mù§terlsi canile anun benem gôrmedin ziilf u zenehdânun çahin 'â§iki ol bend ti zindànuh benem sen gulistânsin açilmi§ bï hazàn bulbul-i §éyzâ vu nàlânuh benem

234

FROM

MAHMUD

KA§GARÍ

TO

E V LÍ Y A

gELEBÍ

I am he who marvels with love at your face, and wails distraught at your locks. My soul made an eternal pact with love-pain for you: I am he who keeps his pacts. Your bow-like brow shoots arrow glances, I am the target of your darts. A ruined place is needed for the treasure of your love: leave it with me, for I am one ruined for you. If they gave the dust of your foot for the price of a life, I am he who would purchase it with my soul. Before ever seeing your locks and the dimple of your chin, I was the lover of your chains and your dungeon. You are the rose-garden which blossomed autumnless, I am your distraught and wailing nightingale. The nurse informed the Shah that the women would be going the next day for a pleasure-outing in the rose-garden. [1486-88] "They brought Gül§äh to the rose-garden and passed by the Shah's tent. As he watched from a hidden place she raised her veil, the sun displayed her face, and that beauty eradicated the Shah like dew " Now iskender was captivated completely: [1513] "although the Shah had conquered the world, he was made a captive by a single hair." Suffering a great deal, he wrote a "letter of separation" (firäknäme) to Gül§äh, including this ghazal: [1600-07] diinyi vii 'ukbida saña kalmi§am kapum cänuma kibla kilmi§am täc u taht u miilk ii bahti terk édüb cänile 'i§kuñi satun almisam 'i§vesine gamzeñüñ meftün olub zülfüñe zär u peri§än olmi§am baña dérler sakla näm u neñi, ben §i§e-i nämüsi ta§a galmi§am mäh u pervin kar§usina géceler yüzüñe agzuña taña kalmi§am cän safä bulmaga sidk u mihr ile subh bigi e§iküñe gelmi§em u§ kapuñda cän u dil yitiirüben yérine gussayle mihnet bulmi§am yüzüñüñ §em'ine kar§u cänumi §öyle kim perväne oda salmi§am [baldan a§di mevc-i bahr-i furkatuñ kil meded vasluñla kim ben gelmi§em

'This verse, missing in Üniversite TY 921, is supplied from Üniversite TY 848, fol. 38a.

THE R O M A N C E

OF

ÌSKENDER

AND

GÛL§ÀH

235

In this world and the next I remain for you. I have made your door the kibla for my soul. Abandoning crown and throne, empire and fortune, I have purchased your love with my life. Enchanted by the cocquetry of your glance, I am wailing and distraught at your locks. They tell me, "Keep your honor and good name," but I have broken the bottle of honor on the rock. Nights, confronting the moon and the Pleiades, I am wonderstruck at your face and mouth. For my soul to find wellbeing, sincerity and love, I have come to your threshold like the dawn. There at your door I lost heart and soul, and found in their place suffering and grief. Confronting the candle of your face I flung my soul as would a moth. [The waves of the sea of separation have gone over my head. Save me by union with you, for I have come.] After sending the letter, sealed with musk: [1616-24] "He cried and suffered so much that an ailment affected his body. The pustules so filled with heat from anguish that he required to be bled in order to be relieved of the melancholic humor. Just as the phlebotomist set to work, and opened the Shah's cephalic vein — just then, they say, Giil§àh too required to be phlebotomized. As the Shah's blood poured on the ground it spelled out the name of Gül§ah, and as the blood from her vein poured on the ground it spelled out the name of iskender. And just as the lancet touched the Shah, simultaneously Gül§ah was lanced. There was no twoness left, but only oneness. What happened here happened there as well. When there is love, it must be so. That union should result, that is the desire." Giil§ah lay incommunicado for three days after receiving the Shah's letter. Finally she wrote a reply, in the course of which "that gazelle extemporized this ghazal": [1659-65] éy nigâr-i giil-'izâr u serv-kad véy biit-i gonca-dehàn u làle-had 'i§kuñuñ derdin dilerem §erh édem nideyem dükenmez étdûgiimçe 'ad çevkumi yüzüñ cemâlin gòrmege nice édeyim vasf çiin yok aña had érte géce el-hamdu lillâh vasfum okiram çiin kul hiive 'llâhu ahad furkatuñuñ hurkatinda géceler hirzi dur cânuma allâhu 's-samad hicrüñ [u§] kahrile alur cânumi vasluñuñ mihrinden érmezse meded baht bigi kapuña yüz siiriyi gelmiçem ümlzle étme béni red

236

F R O M M A H M U D K A § G A R i TO E V L i Y A Q ELEB i O beauty with rose-cheeks and cypress-stature, O idol with bud-mouth and tulip-cheek. I would recount my love-pain for you, but how can I? there is no end when I count. How can I delimit my longing to see the beauty of your face, for there is no limit to it! Morning and night, praise be to God, I recite your quality thus: "Say He is God, One." In the burning of separation f r o m you, at night, my soul's refrain is: "God, Eternal." Separation from you will take my soul captive by force, unless aid arrive from the love of union with you. Prostrating myself at your gate, like fortune, full of hope I have come. Do not reject me.

The Shah, having returned to his realm and resumed the throne, charged Aristo with the mission of seeking Giil§ah's hand. Aristo travelled to Sisitan with rich gifts, secured an audience with the king, and delivered the message: [1708-15] "When Zeresb heard it his heart was oppressed with anguish. For anciently, when Riistem fought with isfendiyar, he had tricked him and killed him fearlessly by an arrow wound. Now Zeresb was of that [i.e., Riistem's] lineage. Even a dragon would not do battle with him. But isfendiyar was the father of Behmen who, when he became king of Iran, sought vengeance for his father's blood, and utterly devastated Sisitan, when he also killed Riistem's son Feramiirz. Since iskender was of that [i.e., isfendiyar's] lineage, the ancient grudge was now renewed." Zeresb refused to give his daughter's hand, and Aristo returned with the message to iskender: [1720-22] "Who, when he heard it, his blood boiled with rage. For there were sovereignty, and love, and youth; carousal, and victory, and the wine-cup. How then could there not be war for the sake of the beloved; battle, and strife, and killing?" iskender marched on Sisitan, routed Zeresb's troops, and laid seige to the city. Suddenly it occurred to him that this might frighten Giil§ah. He explained to his army that if he wounded her he would be wounding himself, since they were one, not two: [1761-62] "What I am explaining to you now is what Mecnun tried to tell the people. He went and struck Leyli with a knife, but himself was struck with the wound." After confirming the point with an anecdote about Ayaz and Mahmud, he concluded: [1779-82] "Let us be patient for a day or two and see what fortune brings. Perhaps we can have victory without battle, and settle the matter without bloodshed. If the goal can be achieved without fighting, why commit evil?" At the Shah's words they sat down there and raised the siege. Seated in a grand tent, iskender wrote a letter to Giil§ah including this ghazal: [1813-19]

THE R O M A N C E OF Í S K E N D E R AND

GÛLÇÀH

237

ey yiizi taze gülistán kandasin gôzi nergis ziilfi reyhân kandasin di§leri lu'lü-yi la le kam sin tutagi la'l-i bezehsân kandasin éy kamu zahmuma merhem kam sin vey kamu derdiime dermân kandasin furkatuñuñ hasretinden te§ne-dil yanaram ey âb-i hayvân kandasin senden ayru dü§di cânum mihnete kanda sin ey râhat-i càn kandasin zülfüñe kayd oldum [u/ bir gün baña démedüñ kim éy peri§ân kandasin bunca müddet dur gülistándan irak àhir ey murg-i seher-hvân kandasin You whose face is a fresh rose-garden, where are you? Whose eyes are narcissus and locks are sweet basil, where are you? Whose teeth are tulip-pearls, and whose lips are Badakhshan rubies, where are you? You who are balm to all my wounds, and remedy to all my pains, where are you? With heart athirst in the longing of separation from you, I burn; O water of life, where are you? Separate from you my soul fell into grief; where are you, O soul's comfort, where are you ? I am captive to your tress, and not a single day did you say, "O distraught one, where are you?" For so long far from the rose-garden, at last, O bird singing at dawn, where are you? When Giil§ah read the letter, shot over by arrow from iskender's camp, it affected her so that she lay unconscious for twelve days. Finally she wrote her reply, including this gazal: [1874-80] 'á§ik oldum hálum añlamaz habïb derdimendem derdiimi bilmez tablb gôrse ol ziilfi ki ben 'akdindayam ola mecnün her haminda bin lebib sénüñ içiin béni bïgâne étdi hvi§ kaldum u§ kavmum arasinda garlb bu 'aceb kim baña cüz derd ü ta 'ab olmadi hüsnüñ nisâbindan nasïb gâh hàcib cevr éder cânuma gàh cevr éder baña 'ayn-i rakïb

238

F R O M M A H M U D K A § G A R 1 TO E V L Í Y A

gELEBÍ

kar§uha efgan ederem ruz u §eb nite kim gill kar§usina 'andelib hacetiim yuzun saqun dur subh u §am kim gorem hayrile vallahu l-mucib I am in love; the beloved does not realize my condition. I am in pain; the physician does not know my pain. If they saw the tress in which I am entangled, a thousand sensible ones would go mad in each curl. Because of you my kinsfolk treat me as a stranger, so I am left a foreigner in the midst of my people. The wonder is that aside from pain and misfortune I have received no portion of your beauty's estate. Now the doorkeeper torments my soul, now I am tormented by the rival's eye. Morning and night I address my lament to you, just as the nightingale addresses the rose. I need your face and hair, that morning and evening I may see them auspiciously. "And God answers prayers." Fearing lest Gul§ah die of grief and longing, the nurse and the eunuch fashioned a ladder out of silk brocade, "as long as the tower was tall" (1903), and all three scrambled down at night and went to the Shah's camp, iskender prepared a secluded palace for Gul§ah: [1911] "But he did not go to Giil§ah, nor lift her veil and see her beauty." Hearing of this, Zeresb at last was reconciled with iskender, and the two lovers were united. In the epilogue to the dastan there is reference to earlier romances, particularly the Persian ones of Nizami, and apparently two Turkish one: 'I§kname of the poet Mehemmed (composed 800/1397) and Cem§id u Hur§id of Ahmedi himself (composed 806/1403): [1924-33] "Ahmedi it is who told you this discourse. Love's secret it is that he revealed. He who listens to this discourse with his soul will comprehend love's mysteries in their true nature. Nizami in his Penc Gene did not utter two words in this fashion, or even one. 1 Love's secret is all in this defter — whatever defter there is aside from this one, roll it up! Were Leyli and Mecnun alive they would make this defter a temple, O pasha! He who reads one page of this defter will take a lesson in love from Ferhad. Though there are many 'I§k-names there is none like this one. He who remembers Cem§id and Hur§id will forget them when this discourse reaches him. In the two days that I wrote my work I embroidered it thus. Contained in this casket, O skilled one, are 605 jewels."

1

Reading

démedifor déme, with

the other mss.

16. "INNER" AND "OUTER" OGUZ IN DEDE KORKUT

Although the notion that the Oguz consisted of twenty-four tribes is attested as early as the eleventh century, 1 it is only in the late twelfth/early thirteen century that we find the first reference to a moiety division, the Ug-ok (or Uguk) and Boz-ok (or Bozuk). 2 Rashid al-Din provides the first scheme in which the twenty-four tribes are distributed between the two moieties, 3 and this scheme persists in Central Asian literary tradition at least until Abu '1Ghazi in the seventeenth century. In western Turkish literary tradition Ug-ok and Boz-ok are replaced by ig ("Inner") Oguz and Ta§ ("Outer") Oguz respectively; and the moieties are thus designated in The Book of Dede Korkut (= DK), although the older designation Ug-ok—Boz-ok is mentioned three times in episode XII. 4 The Oguz genealogists, beginning with Rashid al-Din, constructed elaborate schemes to show the relations among the twenty-four tribes and their descent from Oguz Khan. 5 These genealogies served a political purpose in justifying the promotion or demotion of a dynasty depending on the precedence of the dynastic eponym. There was also a concern for which tribes shared a common badge (ongun), what place each occupied in a courtly setting (iorun), and what portion of the animal each deserved at feasts (iilu§). 6 The formal listing of these requisites and prerogratives, such as we find in Rashid al-Din, we may call the Oguz royal tradition. DK, as we shall see, has traces of the royal tradition, but the concerns of the story-teller are quite different from those of the genealogist. The questions "Who rescues whom?", "Who fights with whom?", and "Who marries whom?" lie behind the elaboration of most of the twelve episodes. The tales and legends that, in part, answer these questions we may call the Oguz popular tradition.

' M a h m u d al-Käsgari, Diwan Lugät at-Turk (facsimile edition, Ankara, 1941), p. 624; R. Islamica I Dankoff, "The Alexander romance in the Diwan Lughat at-Turk," Humaniora (1973), p. 236. Ibn al-Athi r, al Kürnilfi 'l-Ta'rikh, ed. C. J. Tornberg, vol. XI (Uppsala, 1851), p. 54; cited in M. Houtsma, "Die Ghuzenstämme," WZKM II (1888), p. 220. ^K. Jahn, Die Geschichte der Oguzen des Rasid ad-Din (Wien, 1969), pp. 45-6; A. Z. V. Togan, Oguz Destam (Istanbul, 1972), pp. 50-2. ^The twelve episodes are indicated here, following Zhirmunsky, by roman numerals. Page references are to the translation by G. Lewis, The Book of Dede Korkut (Penguin Books, 1974). E. Woods, The Aqquyunlu (Minneapolis and Chicago, 1976), pp. 186-96. ^F. Siimer, Oguzlar (Ankara, 1972), charts after p. 210; A. inan, " ' O r u n ' v e 'Ülü§' Meselesi," Türk Hukuk ve Iktisat Tarihi Mecmuasi I (1931), repr. in Makaleler ve Incelemeler (Ankara, 1968), pp. 241-54.

240

FROM

MAHMUD

KA§GARÌ

TO EVLÌYA

£ELEBÌ

Before examining the tales themselves, it will be useful to list the warriors according to their kin and moiety affiliations, judging by indications in DK, and supplemented by the list of Bayburtlu Osman. 1

igOGUZ

Bayindir Khan

TA§OGUZ

Deli Dundar Kazilik Koca ========= daughter

I

Biigdiiz Emen

Yigenek

Aside from the prophecy in the Introduction of DK (no doubt composed under Ottoman auspices) that sovereignty will "return" to the Kayi, DK contains little trace of the twenty-four tribal system that so preoccupied the genealogists. Only some of the Oguz warriors' names are drawn from the tribal eponyms, most notably that of Salur, chief of the Inner Oguz. Bayindir, assigned by the genealogists to the Ug-ok (or Inner Oguz), seems to occupy a position outside, or above, the moiety system in the tradition common to DK IfBayburtlu] Osman, levarih-i Cedid-i Mir'at-i Cihan, ed. Atsiz (Istanbul, 1961), p. 24/25; quoted in M. Ergin, Dede Korkut Kitabi I (Ankara, 1958), p. 40. The key passage reads: ... They were called the Oguz tribe. The one called Kazan Khan was the vezir of Bayandur Khan and also his son-in-law. Among them mention is made of the Outer and the Inner Oguz. Belonging to the Inner Oguz were Kazan Khan, Budak Beg, Yigen [usually identified with Yigenek in DK], §emseddin Beg, Beyrek Beg, and Kara Konuk Beg. Belonging to the Outer Oguz were Dundar Beg, Emen Beg, Avgar Beg, Doger Beg, and Riistem Beg. They were ninety thousand warriors in all. Dede Korkud was their sheikh. ...

I N N E R "

A N D

" O U T E R "

O G U Z

241

and to Bayburtlu Osman — no doubt reflecting the political precedence of Bayindir/Ak-koyunlu at the time this tradition was fixed. Another Ug-ok/Inner Oguz eponym that appears is Biigdiiz — but, curiously, both DK and Bayburtlu Osman assign "Biigdiiz" Emen to the Outer Oguz. The only other Ug/Inner eponym mentioned (if the vocalization is correct) is Kinik, in DK in the name Kirk-Kimk son of Kara Cogur. K 11 ^ Kinik's moiety affiliation is not indicated in DK, but he is perhaps to be identified with Bayburtlu Osman's Kara Konuk (Inner), unless the latter is rather to be identified with DK's Kara Gone (Inner). 1 The only Boz-ok/Outer Oguz eponym in the body of the text of DK is Doger (or Dogiir). Bayburtlu Osman adds Av§ar as an Outer Oguz warrior. The Introduction of DK informs us that Dede Korkut himself belonged to the (Boz/Outer) tribe of Bayat, which the genealogists closely associate with Kayi. 2 Of greater interest to the story-teller was the ceremonial arrangement of the warriors at the "court" or camp of Bayindir Khan. At the beginning of III (Lewis tr., p. 59) 3 we are told: B e f o r e B a y i n d i r K h a n t h e r e s t o o d K a r a B u d a k son of K a r a G o n e , l e a n i n g o n h i s b o w . O n t h e K h a n ' s r i g h t s t o o d U r u z s o n of K a z a n , o n his l e f t s t o o d P r i n c e Y i g e n e k son of K a z i l i k K o j a .

Elsewhere (VII, 133) Kazilik Koca is called "the minister \vezir \ of Bayindir Khan." In this tradition, all three ceremonial places are occupied by warriors of the Inner Oguz. The parallel text at the beginning of IV (p. 88), where, however, the scene is the "court" of Salur Kazan, reads: H i s s o n U r u z s t o o d f a c i n g h i m , l e a n i n g o n h i s b o w . O n h i s r i g h t sat h i s brother Kara G o n e , on his left his maternal uncle Uruz.

Elsewhere (XII, 184) we learn that Beyrek is "Kazan's minister." In this tradition the formal arrangement is thus: Right I

Salur Center I

Left O

^O. §. Gokyay, Dedem Korkudun Kitabi (Istanbul, 1973), p. CLXV, argues strongly for the latter identification. The traditions on Korkut have been gathered and analyzed by V. Zhirmunsky in Kniga moego deda Korkuta (Moscow-Leningrad, 1962), pp. 145-74; and Gokyay, op. cit., pp. C X W - C X U . Both these scholars overlooked Ibn al-Athir's report (see p. 239, n. 2), according to which the emir of the Boz-ok was Qurgut ibn ' Abd al-Hamid. •^See p. 239, n. 4 for references.

242

FROM

MAHMUD

K A § G A R i TO EVLiYA

£ELEBi

Since the scheme of these two texts is the same, it seems reasonable to suppose that one served as the model for the other; and it is generally supposed that the Salur Kazan tradition was prior. For our purposes, it is enough to note that in the Oguz royal tradition as reflected in DK, the Inner Oguz definitely take precedence over the Outer Oguz. It is the slighting of the Outer Oguz in a ceremonial or ritual context — the pillaging of the Khan's tent — that provokes them to rebellion against the Inner Oguz in the final episode (XII). The Salur tradition, of course, contradicts the Central Asian genealogies, according to which Boz-ok took precedence over U?-ok; so in the "Uighur" legend of Oguz Kagan we read: "On his right sat the Boz-ok, on his left sat the U^-ok." 1 Turning to the popular tradition, the question "Who fights with whom?" is answered twice in DK in nearly identical words: (II, 57) Wild Dundar with the nobles of the Outer Oghuz attacked on the right, Kara Budak son of Kara Gone attacked on the left with his brave warriors. Kazan with the nobles of the Inner Oghuz attacked the centre. (IV, 105-6) Wild Dundar with the nobles of the Outer Oghuz attacked on the right, Kara Budak with his brave warriors attacked on the left, Kazan with the nobles of the Inner Oghuz attacked the centre.

The formal battle arrangement is thus: Right

Center

Left

O

I

I

And this agrees with the tradition in Rashid al-Din, according to which the Boz-ok constitute the right wing of the Oguz army; the U§-ok, the left wing. 2 We should note that in the catalogue of heroes given in the course of the battle description in two of the tales (II, 55-6; IV, 104-5), representatives of Inner Oguz (Kara Gone, Kara Budak, Beyrek, etc.) and of Outer Oguz (Deli Dundar, Uruz Koca, Bugdiiz Emen, etc.) are listed indiscriminately. Similarly, when Bayindir Khan orders "the lords of the twenty-four provinces" (a reminiscence of the twenty-four tribal organization?) to accompany Yigenek to rescue his father (VII, 134-5), of the five listed the two whose moiety affiliation is clear belong to the Outer Oguz (Deli Dundar, Dolek Evren), although Yigenek himself belongs to the Inner. And when the Oguz warriors go to rescue Salur Kazan in the company of the latter's son Uruz (XI, ' W. Bang and G. R. Rachmati |Arat], "Die Legende von Oguz Qagan," SPAW (1932), p. 704 (lines 367-8). 2 Jahn, op. cit., pp. 4 3 , 4 8 ; Togan, op. cit., pp. 48, 53.

I N N E R "

A N D

" O U T E R "

O G U Z

243

178-80), of the four mentioned, two are Inner (Kara Gone, Beyrek) and two Outer (Dolek Evren, Alp Riistem). Finally, Tepegoz is a menace to all the Oguz, Inner (Kazan, Kara Gone) and Outer (Alp Riistem, Biigdiiz Emen, Uruz Koca, Kiyan Selciik) (VIII, 142-3).1 As to the question "Who marries whom?" the answer, only implicit in DK, appears to be: men of Inner Oguz marry women of Outer Oguz. The three cases that illustrate this rule — Salur Kazan's father Ula§, Bamsi Beyrek, and Yigenek's father Kazilik Koca — have been taken as evidence that DK preserves traces of a system of "dual exogamy" between the Inner and Quter Oguz. 2 There are, however, three details which seem to contradict this rule: (1) Salur Kazan is the son-in-law of Bayindir Khan, who, at least originally (as we have seen) belonged to the U?-ok/Inner-Oguz. This may be explained away, however, if we agree with Zhirmunsky that the marriage tie between Salur Kazan and Baymdir Khan is a late and artificial interpolation. 3 (2) One of the epithets attached to the name of Kara Budak is: "who came and won Kazan's daughter by his valour" (IV, 104-5). This too may be explained away if we assume that the preferred Islamic marriage between first cousins on the fathers' side was tending to replace the exogamous system of the Oguz tribes. (3) When Kan Turali wants a wife, he seeks one among the Inner Oguz; but when his father seeks a wife for him, he first looks among the Inner Oguz, then the Outer Oguz, then the infidels of Tribezond (VI, 117-8). This too may be explained away if we assume that Kan Turali did belong to the Outer (although there is no other indication of this), and so properly would seek a wife among the Inner Oguz; and that the story-teller wishes to say that no Oguz girl at all, but only a Greek princess, was worthy of him. There are some other details in DK that become clearer if we assume an exogamous system between the Oguz moieties. When Uruz (son of Salur Kazan) and when Deli Dumrul fear that they will die, and wish to make a disposition for their wives, they refer to them as yad kizi "a girl not of our tribe" (IV, 101: V, 113); this has been taken as evidence for an exogamous system among the Oguz. 4

The moiety affiliation of the others mentioned in the Tepegoz episode (U§un Koca, Aruk Candan, Demir Donlu Mamak, Yiinlu Koca, Yapagilu Koca) is unknown. There is no reason to assume that they were Outer Oguz, or that this episode basically concerns the Outer Oguz, as does Siimer, op. cit., pp. 389,422. 2 S o Zhirmunsky, op. cit., p. 212; also Siimer, op. cit., p. 404, who is more skeptical. •^Zhirmunsky, op. cit., p. 180. inan, "Turk Dugiinlerinde Exogamie Izleri," Turk Dili ve Tarihi Hakkinda Ara^tirmalar I (1950), repr. in Makaleler ve Incelemeler (Ankara, 1968), p. 342.

F R O M M A H M U D K A § G A R i TO E V L I Y A Q ELEB i When Beyrek is dying (XII, 186) he refers to his wife, Uruz Koca's daughter, thus: Before Uruz's son Basat comes and takes my white-skinned beauty ... Kazan must come to my aid ... He must marry my white-skinned beauty to his son ...

The fear seems to be that the woman would revert to her Outer Oguz father and brother, instead of retaining an Inner Oguz husband. Indeed, the final episode of DK (XII), in which the Outer Oguz rebel against the Inner Oguz, may be looked on as the tragedy of Bamsi Beyrek, who is caught between loyalty to his own moiety, the Inner Oguz, and loyalty to the moiety of his father-in-law, the Outer Oguz. The moiety conflict is reduced, in the manner of heroic tales, to the individual conflict of Uruz Koca against Salur Kazan. Uruz Koca summons Beyrek as the mediator, since "he is my son-in-law, but he is K a z a n ' s minister" (p. 184). When Beyrek refuses to rebel against Kazan, Uruz Koca, in anger, deals him a blow with the sword that proves fatal. It has been suggested that the epithet attached to the name of Deli Dundar — "who three times in one fight unhorsed so great a hero as Kazan" (IV, 104) — refers to the battle between the moieties. 1 But in the description of that battle (XII, 188) Uruz Koca (Deli Dundar's grandfather!) chooses Kazan as his adversary, and Kazan unhorses him and has him beheaded, at which point the Outer Oguz submit and peace is restored. It is hard to tell whether the Oguz moiety division was ever anything more than a fiction. In DK it is a system projected on a remote heroic past and serving to explan intra-Oguz rivalries as well as, perhaps, marriage connections. Like other myths, it no doubt reflects a social reality, but in the absence of external evidence it would be hazardous to treat it as anything other than a myth.

!

So Siimer, op. cit., pp. 396,400.

17. THE SEYAHATNAME LITERARY MONUMENT

OF EVLIYA CELEBI AS A

The study of the Seyahatname as a literary monument has hardly begun. This essay focuses on issues relating to genre, form, and style. With regard to both contents and structure, designating the work as travel literature is too simplistic; the most exact generic description would be: Ottoman geographical encyclopedia structured as travel account and personal memoir. The first-person account of Evliya's itinerary and adventures forms the armature within which he builds up his main structure consisting of description. Evliya's descriptive style is expansive and digressive, with frequent use of alliteration, jingles, wordplay, and similes, all of which add charm to a text that is otherwise admittedly quite tedious. Evliya's literary flair comes out in the narrative sections, which reveal him as a master prose stylist. His occasional sallies into verse, on the other hand, are perfunctory.

Evliya £elebi (1611-85 [?]) is one of those authors more quoted than read. The vast size of his work and its encyclopedic nature invite skimming and probing rather than concentrated and extensive perusal. Also, the lack of a reliable edition until now has discouraged serious analysis. 1 Thus, there has been little effort to try and characterize the work as a whole. In this essay, I leave aside the broader contextual questions relating to Ottoman cultural history 2 and focus on strictly literary issues relating to genre, form, and style.

The Question of Genre The Seyahatname recounts journeys over a forty-year period (roughly 1640-80). To that extent it falls in the category of travel literature. But there are two kinds of difficulties with this designation, one based on internal, the other on external considerations. ' T h e text published by Yapi Kredi Yayinlari can be considered a critical edition beginning with Book V; there is a plan to re-edit Books I-IV. 2 F o r evaluations of these questions, see the following works: "Evliya £ e l e b i " by M. Cavid Baysun, "Evliya £elebi" by Mucteba Ilgiirel, "Evliya (Jelebi and the Seyahatname" by Robert Dankoff, and An Ottoman Mentality: The World of Evliya felebi by Robert Dankoff.

246

FROM M A H M U D K A § G A R I TO E V L I Y A Q ELEB I

1. The main point deriving from internal considerations relates to contents. The Seyahatname is by no means simply a straightforward travel narrative. Rather, the roughly chronological travel account merely provides the armature within which the author does many other things. These other things — mainly description, but also narratives of various sorts, such as historical and hagiographical, as well as elegies and other kinds of insertions in verse — make up the bulk of the work, and so cannot be considered simply as digressions to the travel account. Another point relates to structure. The division of the work into ten "books", while perhaps required because of its voluminous nature, is by no means a mechanical one based simply on a chronological division. Rather, there are organizing principles both within each book and among the books that are different from the chronological principle of the travel narrative which provides the armature of the work. The framing books, Books I and X, are devoted to Istanbul and Cairo respectively, the major metropolises of the Ottoman Empire, where Evliya spent the parts of his life before and after his forty-year period of travel. These two books are the only ones provided with a chapter organization, and appear to be modelled on each other in various respects. For example, the description of the shops and guilds in Cairo (Book X, ch. 49) is a reduced and more straightforward version of the corresponding sections concerning Istanbul (Book I, ch. 270). Further, the survey of quarters and villages up and down the Golden Horn and the Bosphorus (Book I, chs. 235-266) has its analogue in Evliya's trips up and down the Nile (Book X, chs. 65-74). While Books II-IX do not have the same tight structure as the frame books, there is clear evidence that the author intended to provide a kind of shape to each book. Although his itineraries over forty years repeat and crisscross, his accounts of them tend to be coherent and interrelated. It is not wrong to characterize Book II by the rubric "Anatolia and the CelalT Rebels", Books III and V by "The Career of Melek Ahmed Pasha", Book IV by "Safavid Borderlands", Book VI by "Hungary and the German Campaign", Book VII by "Habsburg Borderlands", Book VIII by "Greece and the Conquest of Crete", and Book IX by "Pilgrimage". It is also no accident that several of the books end with saga-like accounts like those of the lives and deaths of the Celali rebel and bandit Kara Haydaroglu in Book II and of the brave and tragic commander Seydi Ahmed Pasha in Book V, or the account of the author's own adventurous escape from Istanbul when threatened by the dangerous grand vizier Ip§ir Pasha in Book III.

A LITERARY

247

MONUMENT

Thus, with regard to both contents and structure, designating the work as travel literature would be too simplistic. Yet the author called it Seyahatname or "Book of Travels" and presented it in the first place as a literary record of his achievement as seyyah-i alem or world traveler. 2. This brings us to the external considerations, relating to Islamic and O t t o m a n c o n v e n t i o n s . Seyahat

(Ar. siyaha)

traditionally refers to the

j o u r n e y s undertaken by a Sufi adept seeking mystical g u i d a n c e and illumination, such as travel in search of a sheikh or at the behest of o n e ' s sheikh, also the visiting ( z i y a r e t ) of the tombs and shrines of evliya

(holy

men, saints, or friends of God). While this motivation is certainly present in the Seyahatname,

and also relates to the author's name (Evliya) and to his

self-characterization as dervish, it is by no means the primary driving force. Nor is the description of such journeys a well-founded literary genre; on the contrary, Evliya's designation of his work as Seyahatname

is unprecedented

(according to his own account, the title was suggested by his father; II 241b23). 1 T h e other traditional type of travel in Islam is rihla.

T h e term

designates both a journey motivated by the hajj (pilgrimage to Mecca) as well as a genre of literature devoted to recording such a journey. This motivation is also present in the Seyahatname,

especially in Book IX, but again it is

subordinate. Rather, E v l i y a ' s primary motivation was to provide a c o m p l e t e description of the Ottoman Empire and its neighboring regions. His model was less the rihla tradition of the A r a b travelers (such as Ibn Jubayr and Ibn Battuta) and more the masalik

wa mamalik

and khitat traditions of the Arab

geographers (such as al-Muqaddasi and al-Maqrizi). The rihla mode is personal or autobiographical: Evliya's junkets and adventures, while often following recognizable narrative patterns, tend to be quirky and anecdotal, sometimes sliding into satire or fantasy. T h e geographical mode, imperial in scope, embraces history, customs, folklore, and much else, all tending to fit into preestablished formulas and grids.

1 References to Books I-VIII are to the following autograph mss.: Bagdat 304 Books I and II Bagdat 305 Books III and IV Bagdat 307 Book V Revan 1457 Book VI Bagdat 308 Books VII and VIII References to Book IX are to Bagdat 306. References to Book X are to lUTY 5973. References include book number, folio number, and sometimes line number. All mss. are preserved in the Topkapi Palace Library, with the exception of Book X, which is to be found in the manuscript collection of Istanbul University.

248

F R O M M A H M U D K A § GAR I TO E V L i Y A

gELEBI

In the Ottoman context, with the partial exception of Mehmed A§ik of Trabzon (d. ca. 1600), Evliya would have found little precedence for what he undertook. Thus, the most exact generic description of the Seyahatname is: Ottoman geographical encyclopedia structured as travel account and personal memoir. It is a genre without precedent and without imitation.

The Poetics of Description If you open the Seyahatname at random, your eye is likely to light on a section heading (set off in the original mss. by indentation, large letters, or red ink) beginning with the word evsaf (description) — or, less commonly, sitayi§ (praise) — and including the name of a town. These town descriptions are the most characteristic literary unit of the work. They generally follow the same pattern, beginning with the history and administrative organization of the town, its names in various languages and their etymologies, and its geographic position. They then continue with a description of the town's topography, with particular attention to fortifications, and including descriptions of houses, mosques, madrasas, schools, inns, baths, and fountains. These sections are followed by comments on town quarters and religious affiliations; climate; the appearance, dress, manners, and customs of the populace; proper names and speech habits; the ulema, poets, physicians, and other notables; markets, shops, products, and comestibles; and parks, gardens, and picnic spots. Town descriptions are usually concluded with comments on graves and shrines, along with biographies or hagiographies of the dead. Some tendencies of Evliya's descriptive style, as also pointed out in Hanneke Lamers's article "On Evliya's Style" and Nuran Tezcan's article "Bir Uslup Ustasi Olarak Evliya £elebi" (Evliya £elebi a s a Master of Style) are the following: 1. The scope of the work being encyclopedic, the treatment of a topic, and also the transitions from one topic to another, are expansive, digressive, and leisurely. These tendencies extend to the narrative portions of the work. Evliya is in no hurry. Even word choice and phraseology are unconstrained by considerations of directness and concision. Thus, speaking of someone's brother or daughter, Evliya seldom uses karinda§ or kiz but rather biirader-i can-beraber or duhter-i pakize-ahter. A school is rarely simply mekteb, but rather mekteb-i sibyan-i ttflan-i ebced-han. If he mounts his horse he might say ata biniip but he might also say esb-i saba-reftarlarimiza siivar olup (IX 201 a20).

A

LITERARY

MONUMENT

249

2. Rhyme and Persianate phrasing are common throughout (as in the above examples) but particularly in headings and/or at the end of sections. E.g., IX 87a8: Ve iku yiiz diikkandir, amma bezzazistani iki aded handir, ve bir hammami onilnde bir gemenzar musalla-yi giilistandir, cdnib-i erba'asi kavak dirahtlari ile araste olmus ibddet-hane-i ehl-i irfandir, ve etrafi kdrgir bina-yi amaristandir. 3. There is frequent use of alliteration, jingles, wordplay, etc. E.g., VI 161al6: berrak ve ak-pak bir ab-i muravvakdir\ VII 95a31: daglari ciimle baglar ve baglarinda zaglar baglarin iiziimlerin yedikce baglar sahibleri zaglardan feryad ediip aglarlar; IX 148b2: gayet girdab sapa yoldur, sipasin ga'ib eden gelmez ve gelen giilmez ve sipasin bulmaz, boyle bir bt-eman yollardir. This tendency also extends to the narrative portions of his work. E.g., IV 2 6 3 a l l : Erzurum begleri ve behadir&run Rum yegleri; V 13b26: beyaz kar iizre kara ku§ gibi §ikar iizleyiip gelirler; V 144a25: anlar bizlere biz anlar a bizler sokup kirarken. 4. There is frequent use of similes and other images that add vividness to an otherwise standard account. Thus, the seeds of a certain fine pomegranate are "as large as partridge eggs" (IX 87a 18: abdar ve danedar ho§giivar ve la 'Igun Nif kirazi derler bir giilnari var kim her danesi keklik beyzasi kadar vardir); and, in a legend, the fourth caliph Ali slices a certain stone with his sword "like a cucumber" (IV 378a5: htn-i fetihde hazret-i Ali . . . dii^mana kar.ol ta§i Ziilfikar ile iki pare etmi^dir, bi-emri-huda guya manend-i hiyar dilmi^dir). 5. In his town descriptions Evliya typically uses ciimle to introduce the number of mosques, madrasas, baths, etc., but sometimes he gets tired of ciimle and substitutes other terms, and once he gets started doing this, he often runs out of standard terminology and has to create terms of his own, quasi-Turkish or quasi-Arabic. E.g., VIII 248a34-249a3: ciimlesi, kamusi, hepsi, umumisi, olancasi, barisi, olandasi, hemusi, hamusi, emetasi, yekunisi, cem'asi, olanisi, gotiiriisi, ammetasi, hemisi, hemetasi, herbarisi. 6. An analogous technique is to assume the narrative style of a different persona, thus alerting or awakening the reader. E.g., V 180a 12180bl7 (shrine of Memi Bey Sultan in Egribucak): Assuming a dervish style, he ends each period with an endearment, as though addressing the reader: imanim, canim, giiliim, omriim, canimin cam, ptrim, canamm, etc. While Evliya constantly employs such devices in order to add charm, it must be admitted that even with these techniques at play, most of the descriptive sections of the Seyahatname (and therefore most of the work) make for very tedious reading. There is little to distinguish the description of

250

FROM

MAHMUD

K A § G A R I TO

EVLÌYA

gELEBi

one mosque from that of another. Even (as often) when Evliya waxes ecstatic over the beauty of a building's structure or decoration, his language tends to cliché. A fine minaret is always "well-proportioned" (mevzun). A crevice or precipice is always likened to an abyss of hell (gayyà, derk-i esfel). If he mentions a garden we can expect a reference to the Koranic "ìrem of the pillars" (usually reduced to iretn-i zàt), and if he mentions fragrances we can be sure that they "perfume the brain" (demàg-i beni-àdem mu'attar eder). In soup-kitchens ( i m a r e t ) and hospices ( t e k k e ) we invariably find that hospitality is extended to all visitors (dyende vii revendeganlara ni'meti mebzùl), and musical performances, where they occur, always resemble those at the Timurid court of Herat around 1500 (Hiiseyin Baykara fasillari).

Evliya as Versifier As pointed out in Osman Horata's article " E v l i y a £elebi SeyahatndmesVndeki Manzum Kisimlar" (Verse Portions of Evliya £elebi's Seyahatname), Evliya frequently cites verses by others. These are usually Turkish, although Arabic and Persian are not infrequent, and he also cites poems in Greek, Kurdish, Syriac, and some African languages (including one that he calls "Hebrew" and that is perhaps invented). Aside from chronograms—whether noted down from the monuments observed or else relating to battles, births and deaths, etc.—he cites verses relating to places or events that have caught his fancy, and occasional favorite verses illustrating the working of fate or expressing his point of view. 1 None of these displays any marked literary sensibility. Usually he quotes from memory and therefore inaccurately (example of Yahya below). Occasionally he names his source, but as far as poets are concerned, he is clearly more interested in anecdotes about their lives than he is in their verse. 2 Turning now to his own compositions, aside from occasional short chronograms and throwaways, we find the following more ambitious examples:

'E.g., Gorelim aytne-i devran ne suret gosterir (I 38a3, II 261b9, 363b25, III 182bl4, V 73b25, VI 54a7, 183bl8, VII 165b24, X 183bl8—only here as the second misra of a beyt beginning: Idhaha varalim ol giinde dollab seyrine). The longest poetic text quoted by Evliya is Cemali's eulogy (sitayif) of Eregli (III 17a-18a). ^ M u s t a f a ¡sen, in his article "Edebiyat Tarihimizin Kaynaklarindan Evliya (Jelebi Seyahatnamesi" (The Seyahatname of Evliya (."elebi as a Source of Our Literary History) shows that in Book I Evliya depends on Mustafa Ali. In Manisa, where he names ten living poets, he seems to depend on his own judgment (IX 38a29-38bl8). This is pointed out on pages 120-23 of Nuran Tezcan's work Manisa nach Evliya Qetebi.

A LITERARY

MONUMENT

251

1.1 214b29-33: On spiritual intoxication as opposed to drinking wine; mesnevi, 7 beyts; meter: U U U . 2. IV 347b: Eulogy Qehrengiz) of Baghdad; kaside, 59 beyts; meter: 3. 12 beyts; 4. 14 beyts\

V 146b6-14: Chronogram on the conquest of Ribni^se; meter: U U U U . VII 182b30-183a2: Elegy (mersiye) for a dead slave boy; meter: — U U U —.

5. IX 79a25-79bl5: Eulogy (sitayi§) of Bal-binan in Aydin; 16 beyts; meter: U U U U .

mesnevi, mesnevi, kaside,

In general, Evliya's verse is uninspired. He mainly resorts to it in order to display his ability to eulogize a place (sitayi§, gehrengiz) in verse, in the manner of CemalT and other poets, just as he can in prose.

The Role of

Narration

While Evliya was an indifferent versifier, he was a master of prose, as has oeen pointed out in Pertev Nai'i Boratav's article "Evliya £ e l e b i ' n i n Hikayeciligi" (The Storytelling of Evliya £elebi) and Mine M e n g i ' s article "Evliya Qelebi SeyahatnamesV nin Birinci Cildinde Tahkiye" (Narration in the First Volume of Evliya £elebi's Seyahatnarne). His literary skills were based firstly on a gift for narrative, honed over the years in his capacity as entertainer and companion of sultans, viziers, and pashas; and secondly on a superb education in Ottoman and Islamic culture, including Koran and commentaries, biographies and hagiographies of the prophets and saints, chronicles, epics, travelogues, belles-lettres, etc. 2 Since he aimed as much to entertain as to inform, he had no compunction about inflating numbers and spicing his otherwise sober travel account with exaggerations, humorous anecdotes, tall tales, and other fictions or embroidered truths. Some of these are highly polished narratives, indicative of literary sensitivity and ambition, and appealing to a sophisticated Ottoman audience. The one instance where

Contrary to Jessica Lutz's assertion in her article "Evliya ("elcbi's Qasida on Baghdad" that the kaside is devoted to Ma'ruf Karkhi (63), only line 18 (verse 25) is devoted to him. Rather, the poem is a §ehrengiz devoted to Baghdad, recounting its praises. Line 22 (verse 32) should be rendered: "They chant its (i.e., Baghdad's; not his, i.e., Ma'ruf Karkhi's) praises." In the following lines Evliya is speaking (as he often does in his praises of various cities) of the lovely boys of the town and their lovers. 2 While some attention has been paid to Evliya's sources of information, there have been few studies concerned with literary influence. The epic style adopted in battle narratives seems to echo popular Turkish genres like Battalname rather than chronicles, and some expressions favored in the hagiographical accounts perhaps derive specifically from Menakib-i Ibrahim-i Gul^eni, but these suggestions need to be confirmed.

252

FROM M A H M U D K A § G A R i TO E V L I Y A

ÇELEBl

his narrative flair spills over into verse ("Hebrew" — i.e., an unidentified African language — poems in Books IX and X attributed to the "prophet" Kaffah) seems to be an elaborate juggling of outlandish linguistic and prophetic lore, calculated to appeal to a recherché Ottoman taste f o r whimsicality. 1 As mentioned above, the first-person account of Evliya's itinerary and adventures forms the armature or scaffolding within which he builds up his main structure which consists of description. This narrative core occupies roughly five percent of the whole. Most of the other narratives that we find in the Seyahatname — whether historical, legendary, or hagiographie — are subordinate to the descriptive sections. The main exceptions are the two sagalike accounts mentioned above, the heroic biographies at the end of Books II and V. In the case of some of the narratives, the autograph manuscripts (i.e., those prepared by Evliya himself, which we have for Books I-VIII) show evidence of very careful editing and revision. For example, Evliya recounts the last days of Sultan ibrahim in two different places (A: I 78al3-78b24; B: II 369b29-370al9, 370b6-373a3). Comparing the two accounts, it is plain that B is based on A, with many elaborations. Sometimes there is only a change in wording. Most of the changes are outright additions, making B nearly twice the length of A (not counting the lengthy interpolation of Evliya's visit to Cinci Hoca's palace). These changes lend color and detail to the narrative, and additional personal reference. Thus, to the list of items which Cinci and the other courtiers caused the sultan to waste money on, B adds a troupe of gypsy musicians. B also includes a list of bad consequences to the sultan's prodigality. When it comes to Cinci being put to death, B mentions how much of his wealth was confiscated by the state, and also the fact that Evliya continued to use the horse which he had received f r o m Cinci as a gift (mentioned in the interpolation). Finally, B spins out the sultan's macabre conversation with his own executioner, adding the remark that he too would join in the funeral prayer. But version A shows reworking as well: Additions made at the final fair copy stage are noted in the margin. Thus, Evliya specifies the date of the events leading up to the sultan's execution, adds the information that the conspirators tried to obtain a decree in their favor from the mad sultan but were unsuccessful, and also includes a rather long note concerning the sultan's funeral. All this demonstrates that Evliya had second thoughts about this narrative two different times, recorded separately and without any overlap. This instructive example, which can also be found in the ' More information on this literary phenomenon can be found in chapter 5 of An Mentality: The World of Evliya Çelebi by Robert Dankoff.

Ottoman

A LITERARY

MONUMENT

253

article "§u Rasadi Yikalim mi? Evliya £elebi ve Filoloji" (Shall We Demolish That Observatory? Evliya £elebi and Philology) by Robert Dankoff, offers a way of demonstrating Evliya's method of composition by comparing his two versions of the same account. One can also compare Evliya's recounting of certain events with the standard Ottoman histories. Thus, Evliya's account of Melek Ahmed Pasha's fall from the grand vizierate can be compared with that of the major historian of the period, Naima. When this is done, as on pages 12-15 of The Intimate Life of an Ottoman Statesman: Melek Ahmed Pa§a (1588-1662) by Robert Dankoff, we can see how Evliya inflates the human drama, colors the narrative with the use of dialect, resorts to the epic mode for battle descriptions, and lightens the whole with his personal touch. Evliya's account of his adventures in Bitlis, from the beginning of Book V, can serve as an example of Evliya's literary art on a somewhat larger scale (text and English translation in Dankoff, Evliya Qelebi in Bitlis). Far from being a mere string of personal anecdotes, this account is a wellstructured narrative with careful setting of scenes, clear transitions, and cumulative mounting of tension, interrupted by comic relief and issuing in a climax and denouement. It can be outlined as follows: 1. The stage is set in Book IV with Melek Pasha's defeat of Abdal Khan. Abdal Khan flees, and the Pasha appoints as puppet khan Abdal's lovable and compliant young son, Ziyaeddin. At the homage ceremony we are introduced to the latter's elder brother, Nureddehir, whom Evliya, in a kind of novel-like foreshadowing, characterizes as "bloodthirsty" (IV 273b36). At the beginning of Book V Melek Pasha learns he has been removed from office as governor of Van. He immediately sends Evliya to Bitlis to help his other agents collect some arrears. Shortly after Evliya arrives, news reaches Bitlis that Melek Pasha has been removed from office, and the following day Abdal Khan himself returns. The situation becomes critical for Evliya. 2. Prelude (V 9a36-9bl): In Evliya's dream, his father reassures him that he will escape. 3. Breakfast scene (9b7-10a22): The khan's wife sends him a warning to leave, he is tested by the khan's sons Bedir and Nureddehir, murder of the "treacherous" steward Haydar Kethuda. 4. Transition (10a22-35): He betrays his apprehensions, exercises his horses daily in the snow to prepare for his escape. 5. First conversation with Abdal Khan (10bl-31): He flatters the khan, who gives him gifts and promises more.

254

F R O M M A H M U D K A§ G A R i TO E V L I Y A

gELEBI

6. Transition (10b32-35): "When I came out, I displayed these gifts to all the courtiers and pretended to be deliriously happy at this good fortune. But the truth was that all joy had vanished from my heart ever since Haydar Agha was cut up at my side. I could only think about mounting my horse, girding my sword, and running away. My joy had turned to sorrow, my wine to poison, as in the verse: Not everyone knows what pain I suffer / Only I know, and God knows |Derdin nice dii§var idigin her ki§i bilmez / Bir ben bilirim gekdicegim bir de bir Allahj.11 remained in Bitlis two more months in this state, unable to go anywhere, since we were completely snowbound". 7. Second conversation with Abdal Khan (10b36-l l a l 6 ) : The khan curses Melek Pasha, Evliya tries to mollify him. 8. Transition ( l l a l 6 - 2 2 ) : The situation daily grows more dangerous, the khan having everyone killed whom he suspects of collusion with Melek Pasha. 9. Skiing scene (1 Ia23-12a4): Murder of the "treacherous" Molla Mehemmed. 10. Transition (12a5-27): Elegiac remarks on Molla M e h e m m e d , another month of agitation and planning escape while entertaining the khan. lines Bath scene (12a27-12b20): Nureddehir becomes riotous and makes threatening overtures to his brother, the "new khan" Ziyaeddin. 12. Transition (12b20-13al): Passing the time playing cards and eating candies and fruit, Evliya excuses himself to urinate, checks his horses and gives instructions to his retinue. 13. Scene in Z i y a e d d i n ' s sleeping quarters (13a 1-15): That night Ziyaeddin has trouble sleeping and asks Evliya to keep him company. In the course of conversation, Ziyaeddin says that he would flee to Melek Ahmed Pasha in Van if only it were not the dead of winter. Finally he falls asleep. 14. Transition (13al5-18): Evliya lies down with his clothes on, as usual, but cannot sleep himself. 15. Climax ( 1 3 a l 8 - 1 3 b 2 ) : A t two hours past midnight he sees Nureddehir enter the chamber, with a dagger in his hand. Evliya, pretending to be asleep, watches as Nureddehir approaches first himself, then two other sleepers, pausing to address each one in monologue fashion, until he comes before his brother Ziyaeddin. Suddenly he flares up and stabs his brother to death. 16. Evliya's escape (13b2-14a35).

' By §eyhiilislam Yahya (not named). On page 2 4 4 of Divan-i Yahya, w e find the following lines: Allan nice du§var idiigin her ki§i bilmez / Bir ben biliirim gekdiigimi bir de bir Allah.

A

LITERARY

MONUMENT

255

A s i d e f r o m its v a l u e as a s h e e r a d v e n t u r e story, the m o s t m e m o r a b l e aspects of this narrative are the f i n e characterizations of the key players. A b d a l K h a n , d e p i c t e d earlier as genial host a n d R e n a i s s a n c e m a n , is n o w revealed as a c a n t a n k e r o u s and suspicious tyrant, r e v e n g e f u l a n d bitter over the loss of his w e a l t h a n d t h e great b l o o d s h e d a m o n g his f o l l o w e r s . E v l i y a m a n a g e s to stay on his g o o d side t h r o u g h j o k e s a n d f l a t t e r y , w h i l e r e v e a l i n g to the reader his o w n f e a r s a n d a p p r e h e n s i o n s . T h e n there is the blustery H a y d a r K e t h u d a , t h e a n c i e n t f u n c t i o n a r y w h o m M e l e k P a s h a h a d a p p o i n t e d as t h e child k h a n ' s s t e w a r d , w h o spits at B e d i r a n d N u r e d d e h i r w h e n they rebuff h i m and s u f f e r s the c o n s e q u e n c e . E s p e c i a l l y p o i g n a n t is t h e c o n t r a s t b e t w e e n N u r e d d e h i r , t h e violent elder b r o t h e r w h o has taken his f a t h e r ' s side, a n d Z i y a e d d i n , the m e e k a n d c o m p l i a n t y o u n g e r b r o t h e r w h o c o n f i d e s in E v l i y a . N u r e d d e h i r ' s n e r v o u s tension and violent thoughts, and Evliya's o w n Falstaffian character, are w o n d e r f u l l y sketched in the climactic scene ( 1 3 a l 8 - 1 3 b 2 ) : It was two hours past midnight. . . . I was lying in bed, and my eyes were on the door. Suddenly, who should appear there but Nureddehir Beg, with a waistband round his middle and his sleeves rolled up. Softly he crept in, stretched and yawned, cracked his knuckles and snapped his lower vertebrae, then fingered his dagger and straightened his belt. As I lay there watching, he first came over to me, saying, "Evliya the dervish. What are you doing in this land?" My heart was in my throat. Taking refuge in God, I gave a snort as if in my sleep, and began to snore like a pig, just as the others were. I was really trumpeting in a comical fashion. Seeing me thus, Nureddehir said, "This dervish's beautysleep is also like the sleep of swine," and he passed me by. Moving on to the story-reciter Molla Dilaver at the foot of the closet, he said: "You pimp from Isfahan, get to work, or go to your house down in the city and sleep there!" He passed him also and went over to the coffeeserver Rustem, lying next to the khan, and looked him over too. Then his eye fell on his own "dear" brother the khan, who was lying on the couch and slumbering peacefully, "bedded on roses and covered with hyacinths." He glanced about furtively, then returned to gaze at the khan. I was peering out from under my coverlet like a dog from under the skirting-board of a privy, wondering if they had some private matter to discuss. Suddenly he drew the dagger from his belt, flung off the khan's gold-embroidered quilt, and gave him a kick, shouting, "Get up, you catamite!" Groggy with sleep, the khan opened his eyes and saw that it was his own "dear" brother. As he cried out, Nureddehir plunged the sharp dagger in his breast, then again at the midriff, twisting it into his belly. Grasping in

256

FROM

MAHMUD

KA § G A R I TO

EVLÌYA

gELEBi

terror and futility for his o w n dagger, the khan fell to the floor, where his brother stabbed him once again. 1

While there are many other sustained narratives in the Seyahatname

—

e.g., sagas of the Celali rebels at the end of Book II, and a Grimmelshausenlike account of adventures during the battle on the Raab at the beginning of B o o k VII — none of these approaches the Bitlis narrative f o r novel-like richness and coherence. One notes the fine pacing of exposition and narration, dialogue, and transitional elements; the density of setting, character, and plot; and the interaction of various social types on the same plane, each retaining his own voice or dialect (heteroglossia). Another direction in which Evliya carries his narrative flair is satire. An example is "The Girl W h o Gave Birth to an Elephant" which comes at the end of the very full description of Sivas (III 79a6; text and German translation in K o r k u t B u g d a y , Evliya

felebis

Anatolienreise

translation in Robert D a n k o f f , An Ottoman

Mentality

204-09;

English

. . . 173-74). Evliya

grounds the narrative in a historical moment, characteristically plays up his m e d i a t i n g and p e a c e - m a k i n g roles, uses dialect and other kinds of verisimilitude to portray the hapless and hilarious plight of the Anatolian villagers, disavows fantasy by appealing to popular theology, and directs his satirical barbs against the greed and corruption of the Ottoman officials on the scene. Other examples of satirical fiction are his account of Genghis K h a n ' s conversion to Islam, directed against the fanaticism of the religious authorities (VII O l a ; English translation in D a n k o f f , An Ottoman

Mentality

... 73-75),

and his description of the Kadizadeli ^ e l e b i s w h o come out to view the battle against Celali Gurci Nebi in Uskudar in 1648 (III 31a25).

Nisfli'l-leyli iki sà'at gegmi$di. . . . Yatdigim yerde iki gozlerim kapuda idi. Ani gòrdiim, Nùre'ddehir Beg orta ku$ak ile bàzùlann sigami§ kapudan igeri yap yap giriip bir kerre gerindi ve stindi, ve parmaklann gatirdatdi ve belin kutiirdetdi, ve elin hangerine koyup kemerin onarup gòrdiim, ibtidà bana dogri gelirken eydir: "Ey faktr Evliyà, seti bu diyàrda ni^lersen " dedikde cànim didime alup "Ilàhì sana sigimram" deyti derunumdan bir tevecciih ediip uykuda htìzlar gibi mu§ mu§ uyur §ekilli olup hinztr-vàr òbirleri gibi horuldamaga ba$ladim, ammà ¡aka-gune nefìr-i hàb gekerdim. Bu hai Uzre Nùre'd-dehir hakiri gorup eydir: "Garibiin hàb-i nàzi dahi hàb-i hinzir gibidir" deyiip hele beni gegitp yiikliik dibinde kissa-hon Monlà Dilàvere varup "Ey gidi Isfahàni, i§ine yà a§agi ¡ehirde hànene gidiip yatsana" deyiip ani da gegdi ve hànun yaninda yatan kahveci Rusteme varup ana da bakup hànun serìri yanina varup gòrdi kim serir uzre hàn hàb-i nàzda nàz-i na'im ile gtil dò}enup stinbul òrtiinen biiràder-i c&n-beràberidir, bir kerre cànib-i erba'asinda bizlere bakup yine hàna bakdi. Hakìr yine ihràm altindan nazar ediip "Ayà bir tenhà mii^àvere edecek kelimàtleri mi var ki?" deyii kadem-gàhda etek tahtasi altindan av zagari bakar gibi hakir nigeràn ediip dururken hemàn Nùre'd-dehir belinden hangerin gikarup hànun uzerinden zer-ender-zer yorganin serpiip hàna bir depme urup "Kalh bir e hey htz gótlek" deyince hàn dahi nevm-àlud seràstme kalkup nigeràn etdikde gòrdi kim biiràder-i càn-beràberidir, hay derken hànun memesi listine bir hanger-i ser-ttz ve tìz bir dahi bagrina urup hangeri karmnda burdi, hàn càn havliyle ba§i altindan hangerin gikaram derken serìrden a$agi dii.siip hàna bir hanger dahi urunca (a transcription and English translation of this

A

LITERARY

MONUMENT

257

Conclusion T h e s t u d y of the Seyahatname

as a literary m o n u m e n t h a s h a r d l y

b e g u n . I h a v e s u g g e s t e d here a f e w of the possible directions in w h i c h such a study m a y go. O t h e r s m a y wish to a n a l y z e this protean text as an a n a t o m y , or a p i c a r e s q u e ; o r m a y w i s h to a p p l y m o r e novel a p p r o a c h e s of l i t e r a r y c r i t i c i s m . N o w that a reliable edition is in the p r o c e s s of c o m p l e t i o n , o n e m a y at l e a s t h o p e t h a t t h e Seyahatname

will in t h e f u t u r e b e r e a d , e v e n

e n j o y e d , as m u c h as it is quoted.

W o r k s Cited Baysun, M. Cavid. "Evliya £elebi". isläm Ansiklopedisi. Vol. 4. Istanbul: MillT Egitim Basimevi, 1948. 400-12. Boratav, Pertev Naili. "Evliya f e l e b i ' n i n Hikäyeciligi". Folklor ve Edebiyat (1982). Vol. 1. Istanbul: Adam Yayincihk, 1982. 297-303. B r u i n e s s e n , Martin van and Hendrik Boeschoten, eds. Evliya Qelebi Diyarbekir'de. Trans. Tansel Güney. Istanbul: ileti§im Yayinlari, 2003. — Evliya felebi in Diyarbekir. Leiden: E. J. Brill, 1988. Bugday, Korkut, ed. Evliya ('elebis Anatolienreise. Leiden: E. J. Brill, 1996. Dagli, Yücel, et al., eds. Evliya Qelebi Seyahatnämesi. Vols. 1-10. Istanbul: Yapi Kredi Yayinlari, 1999-2007. Dankoff, Robert. An Ottoman Mentality: The World of Evliya Qelebi. Leiden: Brill, 2004. "Evliya C^elebi and the Seyahatname". The Turks. Eds. Hasan Celäl Güzel et — al. Vol. 3: Ottomans. Ankara: Yeni Türkiye Publications, 2002. 605-26. — "Evliya £elebi ve Seyahatnämesi I§iginda O s m a n h Toplum Hayati". Türkler. Eds. Hasan Celäl Güzel et al. Trans. Nasuh Uslu. Vol. 10. Ankara: Yeni Türkiye Yayinlari, 2002. 268-91. — "§u Rasadi Yikalim mi? Evliya £elebi ve Eiloloji". Tezcan and Atlansoy 99-118. — The Intimate Life of an Ottoman Statesman: Melek Ahmed Pa§a (15881662). Albany, New York: State University of New York Press, 1991. Dankoff, Robert, ed. Evliya Qelebi in Bitlis. Leiden: E. J. Brill, 1990. Horata, Osman. "Evliya Qelebi Seyahatnämesi'ndeki Manzum Kisimlar". Tezcan and Atlansoy 155-67. Ilgürel, Mücteba. "Evliya ( x l e b r ' . islam Ansiklopedisi. Eds. Tayyar Altikula£ et al. Vol. lines Istanbul: Tiirkiye Diyanet Vakfi, 1995. 529-33. Isen, M u s t a f a . " E d e b i y a t T a r i h i m i z i n K a y n a k l a r i n d a n E v l i y a £ e l e b i Seyahatnamesi". Türkliik Aragtirmalari Dergisi 4 (1988): 229-33. Lamers, Hanneke. "On Evliya's Style". Evliya Qelebi in Diyarbekir. Eds. Martin van Bruinessen and Hendrik Boeschoten. Leiden: E. J. Brill, 1988. 64-70. — "Evliya (Telebi'nin Uslubu". Evliya Celebi Diyarbekir'de. Eds. Martin van Bruinessen and Hendrik Boeschoten. Trans. Tansel Güney. Istanbul: ileti§im Yayinlari, 2003. 117-25.

258

FROM

MAHMUD

K A§ G A R Î TO E V L Î Y A

ÇELEBi

Lutz, Jessica. "Evliya Çelebi's Qasîda on Baghdad". De Turcicis Aliisque Rebus Commentarli Henry Hofman dedicati. Ed. Marc Vandamme. Utrecht Turcological Sériés, vol. 3. Utrecht: Instituut voor Oosterse Talen en Culturen, 1992. 59-77. Mengi, Mine. "Evliya Çelebi Seyahatnamesi'nin Birinci Cildinde Tahkiye". Tezcan and Atlansoy 197-208. Çeyhiilislâm Yahya. Divan-i Yahya. istanbul: Matbaa-i Amire, H. 1334. Tezcan, Nuran. "Bir Ùslup Ustasi Olarak Evliya Çelebi". Tezcan and Atlansoy 23143. Tezcan, Nuran, ed. Manisa nach Evliyâ Çelebi. Leiden: Brill, 1999. Tezcan, Nuran and Kadir Atlansoy, eds. Evliya Çelebi ve Seyahatname. Gazimagusa: Dogu Akdeniz Universitesi Yayinlan, 2002.

18. TURKIC LANGUAGES AND TURKISH DIALECTS ACCORDING TO EVLlYA £ELEBI

In addition to the thirty non-Turkic languages which Evliya £elebi noted down in the course of his travels, 1 the Seyahat-name contains rich materials on Turkish as well. Evliya was trained in the science of Koran recitation, and he also exploited the resources of Ottoman-Turkish orthography to the full. 2 He was thus able to note phonetic nuances of the kind that Turcologists generally despair of finding in Arabic-script texts. Evliya's standard, of course, is the Turkish of Istanbul. While growing up, Evliya heard all around him the speech peculiarities of the various groups settled in the capital. 3 In book I — his vast panorama of Istanbul — he records samples of the street-cries, the cant and banter of different traders: Albanian sweetmeats' cooks, Greek tripe-cooks, Laz coppervvare merchants, Armenian sewer-cleaners, etc. (See below) He was also sensitive to dialect diferences according to social class. Thus, at the top of the hierarchy, the Ottoman ruling house apparently had some peculiar speech-habits of their own, such as referring to "wine" by the archaic term gaqirAt the bottom of the ladder, the tanners and rebellious artisan types used such expressions as bizi begnemedin mi ("Don't you like us?" = begenmedin), and oldi ("Kill!" = oldir ). 5 Evliya even gives a sample of the fractured Turkish spoken by someone raised in Europe ("Fireng lehcesi"): xunkarci (= xiinkar), Muhammed efendici (for the Prophet, = efendi).6 ^See Robert Dankoff, "The Languages of the World according to Evliya (Jelebi," Journal of Turkish Studies/TUBA 13 (1989), 23-32. 2 S e e Robert Dankoff, "Evliya Qelebi'de Imla Hususiyetleri," paper read at 6th Intl. Turcology Congress, Istanbul, 1988; M. van Bruinessen and H. Boeschotten, Evliya Celebi in Diyarbekir . . . (Leiden: Brill, 1988), 71-99. 3 N o t e I 1 4 4 a l 0 where Evliya states that most of the populace of Uskiidar are settlers from Anatolia and therefore speak rustic Turkish (Etrak lisdni). References to books I-VIII of the Seyahat-name are to the autograph ms. as follows: Bagdat 304 books I and II Revan 1457 book VI Bagdat 305 books III and IV Bagdat 308 books VII and VIII Bagdat 307 book V References to books IX-X are to: Bagdat 306 book IX ¡UTY 5973 book X Other page numbers, without a or b, refer to the Istanbul printed text, 10 vols., 18861938. (In these footnotes, references to the printed text are not included for the linguistic material; see Appendix E.) 4 I 69b20 (247: omitted). 5 I I I 103b7,20, V 6 a l 4 . 6 I 132b2f.

260

FROM M A H M U D K A § G A R I TO EVLiYA Q ELEB I

In the course of his travels, Evliya employs different methods to convey the varieties of Turkish speech: (1) Where the speech is sufficiently close to the standard to be readily understood, he is content to give a few samples, without translation, just to indicate phonetic or morphological peculiarities. Thus, of Uskiip dialect (Macedonia), he remarks: "It lies between the Anatolian and Albanian dialects" (Rum ve Arnavud lisani arasinda), e.g.: Ya biz size demi§miz, gelepi 1 efendimize varamiz, ve anlar pizim efentimizdir. Of the Turkish dialect spoken in Buda (Hungary) Evliya reports that it is basically Bosnian (Bo§naq), since the (Turkish-speaking) populace are all from Bosnia; e.g.: Muhassal-i gelam, zahira, demi§um, §ayed gelmeyesun, sinqo cibi sevmi§um, ya dahi sagmisun.2 In Amasya (Anatolia) he says that the speech of the upper classes (igelebiler) is very elegant (fesahat u belagat iizre) but the lower classes (sayir re'ayalari) use such expressions as: §una goqu§raq i§ qayralim, §u i§i alatlayi gorelim, onatca er imi§And for various places in Greece he just signals the dialect by two or three words: varmisik, gelmisik (Drama), varmisiq, gelmisiq (Siroz), varmi§ik, gelmi§ik, almi§ik (Vodina), etc. 4 (2) Sometimes he does not bother with a formal section on the local dialect, but rather introduces samples of local speech in his description of manners and customs. Thus, discussing hamsi fish in Trabzon — xapsi palugi — he remarks that when the fishmongers cry their first wares of the seaon, even the imams in the mosques break the ritual worship and run out in the streets with their h a n d k e r c h i e f s crying: Babah camygtin olsun, axgacugumla bir maqrama xapsi ver ("By your father's life, give me a handkerchief of hamsi for my little asper"). People joke that they are sorry to see the brine dripping from the wrapping, and say: Bire, paligih suyin yana aqidirsin, suyina bir plavciq salsana ("Hey, you're letting the fish juice drip out, why don't you toss in a little pilav"). Evliya also records a tekerleme with which the Trabzon sophisticates taunt the simple townsfolk ( Q i g u re'ayalari): Trabuzandir yerimiz Axga dutmaz elimiz xapsi paluq olmasa nige olurti xalumuz (Our place is Trabzon, our hand can't keep an asper; if not for hamsi fish, what would be our state?). 5 171b23. V I 88a9. 3 I I 281b33. 4 V I I I 218al4, 220a30, 232a5. 2

5

II 252b.end.

TURKIC

LANGUAGES

AND TURKISH

DIALECTS

By the way, the same phonetic peculiarities revealed in Trabzon also characterize the street-cry of the Laz copperware merchants in Istanbul: Tavalar tengereler paqur avadanliqlar xurtavatlar ahrim ("I buy pots and pans, copper vessels and utensils"). 1 Evliya records another taunt-song in Toqat — here the town wits making fun of the Armenians' Turkish dialect: Erzingandan inq xabar Toxatun qa§malari axar mi Axar amma §ir-mir axar (What news from Erzincan? Do the fountains of Toqat flow? They flow but they flow §ir-mir).2 Similar phonetic features mark the street-cries of the Armenian whitewashers in Istanbul: Seksen axqahi alup agirdirim, yatmi§ axqa disan olmaz ("I'll take your eighty aspers and whitewash, but if you say seventy I won't"); and of the Armenian sewer-cleaners: §urada bir kariz vardir, aqup ayirdlasax, qox kariz badami qvcardi, xarnimiz acdur, badam yesax ("There is a sewer, let's open it and clean it out; many sewers brought up the almonds, we are hungry, let's eat almonds"). 3 (3) Where the speech includes elements rather distant from the standard, Evliya may give a full-blown specimen, listing words and phrases with their standard Turkish counterparts — the same method he employs with the nonTurkish languages. In two cases, Diyarbekir and Bitlis, these lists are appended to poems composed by local poets in order to illustrate the dialectal peculiarities of their respective cities. 4 Elsewhere in Anatolia, Evliya gives extensive specimens of three dialects: that of the Bolu region, recorded in Gerede; that of Gordes in western Asia Minor; and that of the Turcomans of Mar'ash. 5 Because the Bolu and Gordes specimens resemble each other so strikingly, a comparison is in order (see Appendix A). Evliya's Bolu materials date from 1646 and consist of ninety-one terms and their standard Turkish equivalents. 6 The Gordes materials, collected (ostensibly) in 1671, consist of thirty-one items, exactly reproducing the first half of the Bolu list (with a few M 185b 15. ^V 20al5; cf. Robert Dankoff, "Evliya Chelebi on the Armenian Language of Sivas in 1650," Annual of Armenian Linguistics 4 (1983), 47-56, p. 55. 3 I 205a20, 205b23. 4 F o r Diyarbekir, see Evliya Qelebi in Diyarbekir (above, n. 2), 100-02, 172-77, 198200; for Bitlis, see Robert Dankoff, Evliya Qelebi in Bitlis (Leiden: Brill, 1990), 18-26, 84-89. 5 For BOLU (II 278a30f.) and GORDES (IX 28a7f. [57]) see Appendix A; for MAR'ASH (III 62a23f.) see Appendix B ^Eren's article mentions only nineteen of these. Hasan Eren, "Evliya (,'clebi ve Anadolu Agizlari," I. Turk dili bilimsel kurultayi .... 1972 (Ankara, 1975), 113-119.

262

F R O M M A H M U D KA § G A R Í T O E V L Í Y A

gELEBÍ

gaps and one word out of order). The conclusion is inescapable that Evliya borrowed from the first list in order to produce the second. It seems that Evliya was determined to find traces of archaic speech in this part of western Asia Minor, where his own forefathers had settled.1 For, immediately following the Górdes list, he states: "This is truly the speech of our ancestors, although there is no trace of it in Kiitahya and Demirci, where our own forebears settled." 2 At any rate, the spurious nature of the Górdes materials should serve as a warning not to treat anything in the Seyahatnáme in isolation, and not to take anything at face value. On the other hand, we should also not deny the real value of Evliya's materials for linguistic investigation. The Mar'ash specimen, gathered in 1649, in addition to single terms, contains entire phrases. (See Appendix B) For Rumelia, the only similar lists are those for the Qitaq dialect in Silistria and in the Dobruca 3 (See Appendix C.) All these materials deserve a separate study. (4) Finally, Evliya records a great deal of everyday conversation, and often, not without comic effect, notes the dialect of the speaker. For example, one of his close associates during the years 1650-55 was a certain Qudde, the major-domo (ketxüdá) of his patron and kinsman Melek Ahmed Pasha Qudde was from Diyarbekir, and in pressure situations his dialect would out. Thus, when threatened by a rival, he quotes a proverb: Her qansi ta§ qatidir ba§iñi aña urgilen ("Whichever stone is hard, strike your head on it"). And when, at the end of his life, he is being put to the torture, he cries out to Evliya: Evliyacix, menim halime pes, etc. 4 Some prominent Ottomans with whom Evliya came in contact were of Bosnian origin, and spoke in Bo§naq lehcesi, such as Varvar Ali Pasha 5 and Isma'il Pasha. Others were from the Caucasus, and spoke in Abazaca or Qerk.es lisáni, such as the Abkhazian Seydi Ahmed Pasha. Evliya is fond of quoting the latter's barbarisms, e.g., his exhortation to the troops before a raid on the cossacks at Giinye in 1647: Allaxim hem seni kerimdir hem beni kerimcidir; or his reply to Sultan Ibrahim who warned him not to strike any of his courtiers with the cerid javelin: Valaha padi§ahim, onlar beni vururdur, beni de onlari hap vururdur, cilide horta yoqidir, onlar beni

1 Evliya tells us that his forebears came from Mahan in Transoxania with Ertoghrul, the proto-Ottoman chieftain, and settled in Kiitahya and in Demirci. See IX l l a 9 (21), H a l (28), 24al3,18 (49), 3 9 a l 7 (80), 3 9 b l 8 (81). IX 28al5: Gerqi aba ve ecdddimizin lisanidir, amma bizim cedlerimiz sakin oldugvi

Kiitahiyye ve Demirci §ehrinde bu gune kelamlar yoqdur. 3

III 120al9f., 125a25f. Note also Qitaq of Megre and Qaraciq at VIII 207b26. III 98b32; 184b24. ^Examples of his speech at II 363b25: Qatlana bir qag ciin ("Wait a few days"); V 20a28.

4

riefs-i

TURKIC

LANGUAGES

AND TURKISH

DIALECTS

263

ba§ vurursa ben onlari di§ vurur.' Evliya even records a conversation between the Bosnian Isma'il Pasha and the Abkhazian Seydi Pasha, shortly before the latter's execution in 1661: Isma'il: Baqa canurn Seydiim Ahmed Pa§a, sen beniim diinya ve ahiret buraderümsün (Look here my dear Seydi Ahmed Pasha, you are my brother In this world and the next.) Seydi: Ben kopek ile qardas mi olur? Yoqdur valaha, olurum yoqdur (Can I be brother to a dog? No, by God, I can't.) 2 Isma'il Pasha's rounded suffixes recall those of the Bosnians in Buda, cited above. Other examples are the Bosnian and Croatian frontier warriors, who in the course of Evliya's narrative say such things as: Bire cele canum Hoco Qelu qazi yigit, bir manca yeyelüz ("Come on my dear ghazi, Hoca Qelebi, let's have something to eat"); Hernán bir i§ górelüz, dóneliiz, gideliiz, goq oturup osurup qoqmayaluz ("Let's go raiding right now and return, let's not sit around farting"); Ya sen safa celdun, ya sag misun, ya sen bu daglarda ni§lerisun ("Welcome! Are you well? What are you doing in these mountains?"). 3 The Ottoman armies had warriors from the Caucasus as well on the Austrian and Venetian frontiers, and Evliya recalls a conversation among some of Melek Pasha's Abkhazian night-guards in Livno in 1660, trying to dissuade a would-be assassin who demanded that he be allowed to present the Pasha with a petition: Baq baq 'adamdir, bu qarahq gecedir davaci 'arz-i haldir olur midir — which Evliya glosses: Baqa baqa adam, bu qaranliq gecede da'váci 'arz-i háli olur mi ("Look here man, is this dark of night any time to present a petition?"). 4 One of the most colorful dialogues which Evliya records is from Chorum in north-central Anatolia Some samples: Avanliqlarinda qurtulaq. ... Islambol veziri qalab Allah qilicina du§ gele, ráhatliginiñ ve qizañlarimñ ve kendiniñ devri done, bu yatan gerqek er demine hu deyelim hu. Imam saña bulamac ve yalamac ve gülemec a§i getirmi§, qapul edüp a§ami§sin, pa§añi da seni de Allah qaradan qadadan ve yavunculu dü§den ve a§up yorudduguñ yerde yaramaz a§dan sizi Alláh

329M2; 335a9. V 184bl8. 3 V 146a3; 161a22; X 124al8. Note also VI 185b20: 2

durmayaluz, varalum ve viläyetleri esirler alaruz.

4

V 143al9.

vuralum, yaqalum

Varmi¡uz ve gelmi¡üz, ve yiqalum, varaluz käfirleri

yüriye qiraluz,

264

FROM

MAHMUD

K A § G AR I TO

EVLiYA

saqlasun ... issi dama girüp §aräb hörpüldedirler ba§larina birer ganaq geküp ... etc. 1

Q ELEB i

ve ol qizil

kekremsi

*

The dialect group for which Evliya provides the most extensive materials is that of Eastern Anatolia and Azerbaijan 2 — although he does not use these names for it. He sometimes calls it "Kurdish" dialect, as in Diyarbekir and Bitlis. In Tebriz he specifies it as the language of the Turcomans, Avshars, and Gök-dolaq. Aside from the Bitlis and Diyarbekir materials, mentioned above, he gives a few samples f r o m Erzurum, Van, Mosul, Hamadan, and Tebriz. 3 Much richer materials can be gleaned from the lengthy conversations Evliya had with the Safavid governors ("khans") of Urmia, Tebriz, Shamakhi and Yerevan, and with other officials, in the course of his missions to Azerbaijan in 1646 and 1655. The following examples will give the flavor of these conversations: I 70bl7 Sen daxi o hurr-i gehidin qamti xäke salanlarun 'irqim coymadin mi II 292a27 Endek hälime merhzamet etgilen, bir dani§igim vardir deyem 301b32 Qansini dilersen saha cabalaram mu 303a22 Yax§i daniniq edersen 314a6 Hey Evliya Aqa, mine heze Revän xäm geledir IV 291b2 Evliyä Aqayi dükeli yedi top atmisix 296al3 Meni äzäd etgilen göyindürüpsen 299a6 Her dani§iqda meni xäna yax§i maxtayasin 302a4 Bunin elbette gözlerine milqe ve ba§ma ta§ga gegiriserdir ve ödün smixdirasardir Evliya uses the word kend ("village") as an isogloss marker for this dialect group. When he first enters the Caucasus region from Turkey in 1646 he remarks that this is the term used here for "small town." 4 Twenty years 352a9 - 10,16-25. [ S e e now Christiane Bulut, "Türkische Dialekte der Übergangsgebiete nach Evliya (Jelebi," in: Scholarly Depth and Accuracy: A Festschrift to Lars Johanson, ed. N. Demir, F. Turan (Ankara 2002), 83-100; Hayati Develi, "Evliya (Jelcbi Seyahatnamesine göre XVII. Yüzyil Azerbaycan-Türkmen Agzi," V. Uluslararasi Türk Dil Kurultayi Bildirileri, 1 (Ankara: Türk Dil Kurumu, 2004), 767-86.] Erzurum (II 288a21): haradaydin, bixeyriyi göyndür (ocagi yaq); Van (IV 252b31, 258a31): yemqix, vermemi§iz; pisix (kedi), bixeri (ocaq), haradaydin (nereye vardin), apargilen (getirsehe), etc.; Mosul (IV 4 0 3 a l 8 — "Kiird lisanina qarib"): varmi^ix, haradaydin ; Hamadan (IV 310b5): harada idih, heze telisemen, etc.; Tebriz (II 300a4): heze tanmamqam (henüz görmedim), darcixmi§am (darildim), yavu baxmi§am (dii§men oldum), apargilen qaqiri (getir §aräbi). 4 II 293bl3: kend lafzi bu diyärlarda qasabaya derler. 2

TURKIC

LANGUAGES

AND TURKISH

DIALECTS

265

later, when he is again travelling in the Caucasus and leaves Circassian country for Daghistan, he remarks that villages are now called kend and not qabaq.1 Such linguistic finesse, of course, does not prevent Evliya from using the word kend for variety's sake elsewhere in his travels, even as far afield as Translyvania and Egypt. 2 Evliya was a better linguist than a historian; but among his antiquarian interests, as we have seen, was following the traces of the early Turkish settlers in Anatolia. One of the stopping-points of the proto-Ottomans between Central Asia and Western Anatolia was Akhlat, on the western shore of Lake Van. And among the many ruins of Akhlat, Evliya remarks particularly on the mosque of a certain Emir Qay who covered the walls with inscriptions detailing the flourishing state of the city in his time. There was one inscription, up near the roof, in tall letters, which Evliya could only make out (so he tells us) using a spyglass; it was composed in an unusual dialect reminiscent of Chagatay and Mongolian. 3 Evliya reproduces it in the text, carefully vocalized and punctuated (punctuation, indicated here as commas, presumably setting off phrases that Evliya could pick out with his spyglass), and with some glosses: I§ven, gitmi§ven, gurumiza varmismiz, bizim Qay gibidir, bogaz eyitdi, dukeli geleserlerdir, meni tjuncamisdir, §ad [u] bay ki§idir, ayitdim §ol ki§i menfadinadir, savular sayladim (ya'nl aglaya qaldim), barimiz urus khislmiz (ya'nl cumlemiz ceng adamiyiz)

This is a remarkable amalgam of Turkish dialecticisms and archaicisms. It mirrors nothing so much as Evliya's heated imagination; and it is not, as we shall see, unique in the Seyahat-name. *

If the dialect spectrum from the Danube to the Caucasus south of the Black Sea can be subsumed under the rubric "Turkish," the language of the Crimean court and its hinterlands north of the Black Sea was known as "Tatar," and Evliya has a great deal of information about it. His Tatar materials are based on his contact with the Crimean Khan and with Tatar raiding parties in the Ottoman armies during the Polish campaign in 1657, and on his extensive travels in the Crimea and beyond in 1665-66. ' VII 161b4: bu vilayetde koylere ¡imdengerii Qerkezistan gibi qabaq ta'blr etmezler, cumle kend derler.

2 3

VI 27al3 (= qurd); X 360b27 (776) (= koy ).

IV 239b33: Cami'uii siituhina qarib olan 'all xadari dur-bin He qird'at edtip terqim etdim, amma lisan-i gagataya ve lisan-i mogola qarib bir lehce-i gayr-i mukerrerdir.

266

FROM

MAHMUD

KA § G A R Í TO E V L Í Y A

£ELEBÍ

Evliya g i v e s two lists of Crimean Tatar words and phrases, one ( T A T A R ) in the old capital of A q - m e s c i d near Baghche-saray, the other ( N O G A I ) in the fortress of Choban. 1 (See Appendix D . ) There is little to distinguish the language of the two lists. 2 Both illustrate the mixture of southern and northern forms characteristic of the Crimean dialects, 3 plus downright Ottomanisms, compounded by Evliya's "hypercorrections" (bursun = vursun,

diigger

= du§er,

bizqa

= boza).4

The same features mark the

several extended passages in Tatar that Evliya gives, of which the following is a sample (with Evliya's glosses noted in parentheses): V 4 3 a l 8 Bu qadar cildir derd-i bela gekermiz, §imdi maldan ve candan ulaq toyum bulmadan Qirima nice barup biyimlerimize bogeclerimize ve §ura ve mirzalarimiza ve qart kisilerimizih betine benzine ne ytiz He baqaq? Alaydan reva bolami? Imdi xanim biz bugun barimiz da qirilirmiz, taburdaki qavurni soqamiz, barimiz da qavuri talava urup doyum olup bay ki§i olganda Qirima gitemiz 43a22 (Crimean Khan) Qaracilarim qo$da§larim ve qarda§larim, agiqman (ya'ni 'acele etmeii) aqrin aqrin edin, alatlaman, salldh doyum bolursihiz 52a31 (Crimean Khan's letter to Melek Ahmed Pasha) Benim aqaqay vezirim ataligim, carligimiz ilten Dede§ §urahiz ile depter mucibine ¡ol qadar uru§ malindan doyum boldugumuzca hedayalar saldiq, "Essexi ma melek." Seper iizre bolungan §olqaydir. Alay bolursa vilayetimiz Qirimga bardiqda Qaya Sultan aytilan efendimize elli Qerkes ve elli Abazga ve elli Gtirci ve elli Urus devgeleri ve iilgalari ve adiqalarina husiindar er u qizlarindan gonderemiz, ve cenab-i §erifiniz xidmetine layiq bolmaq igiin yax§i sadaq ta§igan zor cigit koleler ve agirmaq yorgalar ve elli gift qargigalardan toyqunlar ve sonqurlar gdndermemiz muqarrerdir in§a'llah. Bende Mehemmed xan VI 19b5 5 Bir yarlig verseniz §u §ehre bir gapul civerup §u §ilga qizlardan ve devgelerden alup Qirimga doyum varsaq dltiirdigimizge gam yemezdik VII 105b6 Eger xanim qadirseh bizi Qirimga gotiir, seniiile uru§ ediip barimiz da qirilup senin leaker it'll ve seni soqamiz ... Mina (ya'ni i§te) Nogay ve yah Tatarlarinih bari da atlanup sadaqh ve savatlanup xanimin iistiine cav dii§mege hazzir kitdiler

Wll 133b, 147b ^Evliya calls the language of a letter by the Crimean Khan (see below) lisan-i Tatar (V 52a31) and also lisan-i Nogav (52b3). He claims to know several varieties of Nogai (VII 175b4 (845: lisán-i Nogayuñ envá'in bildigimden); also states that there are twelve Nogai dialects requiring interpreters for mutual comprehension, and that he himself has memorized many Tatar expressions (lugát) (VII 133b30-134a2). 3 Cf. Gerhard Doerfer, "Das Krimtatarische," in Philologiae Turcicae Fundamenta I (Wiesbaden, 1959), 369-90, p. 370. Cf. firgin = firm in a Badraq Tatar tekerleme (VII 128b34 [618]); and, in the texts below, Abazga = Abaza -'Some of the text from books VI and VII lack diacritics; several erroneous ones were inserted by a later hand; the printed text is not reliable.

TURKIC LANGUAGES AND TURKISH

DIALECTS

267

107b27 Ekmek necikdir qanda olur? ... Ekmek gopekde bitgendir, degirmeninde un yasadup hamur edtip bisgendiriip ate§de koytindtiriirler ekmek olur. ...Adam Atayi cennetden giqarmaga sepep §ol gopekde bitgen dtmek midir? Behey atah anari cavil bolayim, oztim soqulursa da ben oni a^amamdir (ya'ni ben oliirsem de ben ekmegi yemem) 113a32 Ctigiir kisi Alay coluq bersin (segird adam Allah yol versin) 115a2 (the Khan urges on his soldiers) Aman guralarim (ogullarim) aman batir eigitlerim (bahadir yigitlerim) ve qaraci (is erleri) qarda^larim, gun bugunki gtindtir, din-i Muhammed 'agqina ve ceddimiz cinqiz xan canin §ad yasadin (ya[p]dih), galap (Allah) Allah qorsa bizge (bize) pirsat (firsat) ve nusret berirse (verirse) doyum bolganda (zengin olanda) barinizih (cumlenizin) qollarinda olan suralar ve devgelerihizden carim savqa (yarim '6§ri) alayim ve size semmur ton ve semmur siipirtma qalbaqlar vereyim. Ve benim qaragilarim, saqimh doyum (zengin) olmaga a§iqmah ('acele), burun (on) qavurun arisin soqup (oldurun) coyun (yoq ¿din), soiiraten (sohrasi) balini a§ayih. Qop (goq) mdl tapmaga (bulmaq) tubmen (du§iinmen) ve kelepure (§ikara) derilup (cem') u§men, qop ulgen 'asker doymaga astqirken ('acele) bozulup qaytavul etiip (donek edtip) singanlardir (bozulmu^lardir) 144a5 Eger xamm sen 'Osmanliya ciigurup kitersen senin ve olanlarinin barisin da soqup oldtire, heman ctigiir 'Osmanli He bozulup uru§ edemiz VIII 196a31 (complaints to the Khan concerning the Shirin Begs who tried to exact a toll [tamga beha] on their cossack captives) Negik ola §olay lemegik ola, Qirim yurdinda alti kerre yuz bin casir qazaqm yazmi§lar, qana sunuh qopna ki§gene sural an ve qana sunuri marya defgeleri yazdir, ni§in yazarlar, zulum yasayup tamga alirlar, Alahung emri §olay midir, Qirimga alay zulm bolsa 'Osmanli kisisi erse yasadiqlari zulmi bilifrijzdir 200al2 (51) Esrik olganmiz ("We are drunk" — what the Tatars say when they smoke tobacco!) Evliya's Tatar materials deserve a separate study.

On several occasions Evliya gives the texts of what purport to be ancient Tatar or "Chagatay" grave inscriptions. Thus, in book I, while commenting on the patron saint of the parade sergeants of Istanbul, the Prophet's companion Malik Eshter, Evliya states that during his sojourn in the Crimea in the year 1076 H. (1665-66) Mehemmed Giray Khan excavated some ground in the royal cemetary at Eski Yurd in Baghche-saray in order to construct a tomb for himself, and uncovered a white marble slab with a Chagatay inscription as follows: 1

' i 153a35. Evliya neglects to record this incident in his description of Eski Qirim at VII 136b23f. (657).

268

FROM

MAHMUD

KA§GARi

TO EVLiYA

gELEBI

Ey sii'al etgen §olqay merqadm eyesin, Salsal qavurni soqqan Malik E§ter oludiir, §ol — [blank] sahabesi ki§ilerindendir, eilgasi iiqyuzki cildir (Oh you who inquire as to the owner of such a tomb: it is the corpse of Malik Eshter, killed by the infidel Salsal; he was one of the companions of the Prophet and died in the year 300) Evliya g o e s on to say that the Crimean ulema performed a calculation and concluded that 770 years had passed since Malik Eshter had fallen victim to Salsal's arrow. 1 Similarly, from the extensive graveyard of the Dadiyan kings near Tarkhu in Daghistan, are the following: 2 Ey sii'al etgen §ol merqadm sahibini, Boga Bay ki§i men, ikiyiiz ya§qa yetgen na-murad yigit men idim, ekmek yeytip yiiregime ekmek yapi§up olgenmen, sebeb-i mevtim §olqaydir (Oh you who inquire as to the owner of this tomb: I am the man Bogha Bay [or, the wealthy man Bogha], I was a warrior who reached the age of 200 but did not attain my desire, I died while eating bread and a piece stuck to my heart — that is the cause of my death!) Ey sii'al etgen bu merqadm eyesin, ikiyiiz qirq be§ ya§qa na-murad HZurma Bike xamm xatun idim, yiiz be§ ya§inda qiz olan iken qocaya barup bir u§aq dogururken su igtip oldiim, sebeb-i mevtim §udur (Oh you who inquire as to the owner of this tomb: I was Hurma Bike Khanim Khatun, who reached the age of 245 but did not attain my desire; I married when I was a virgin 105 years old, and died when I drank water while giving birth — that is the cause of my death!) Ey sii'al etgen §ol me§hetni sahibini, iXqyiiz yasqa yetgen Quba Alp zor batir yigitken Ejderxan savannda oq pa§agi po§ugalanda oq temreni doqunanda dziimi beni oq soqup olgenmen, sebeb-i olgenim §olaydir, oziim igin patiha fatiha (Oh you who inquire as to the owner of this tomb: I was a champion warrior [named] Quba Alp and had reached the age of 300 when I was struck by an arrowhead during the battle of Astrakhan and died — that is the cause of my death! [Recite] a Fatiha for me) Finally, along the Volga near Astrakhan Evliya "discovers" the tomb of the ancient Iranian king Tahmuras Khan son of Hushang Shah. It is seventy-seven paces tall and made of porphyry. The inscription, recounting his feats and inventions, ends: ve Turkistanda ve ciimle

Tiirkistam

bind etgendir,

rahmet-i

of such inscriptions is said to be f agataysi

1

Yesu §ehrini Izid bolgandir.

ve

§dpuristam

The language

lisan?

Malik al-Ashtar actually was killed by the logistarius of al-Qulzum in the year 37 H. (658); see EI2, 2 VII 160b33f. 3 VII 170b8,33.

TURKIC

LANGUAGES

AND

TURKISH

DIALECTS

269

Evliya places these "inscriptions" in their (quasi-)historical context while discussing another ancient text, this from the time of Alexander the Great, located in Strumica (Macedonia). He fancifully connects the inscription with the etymology of the Turkish name of the town — Ustirumca — by saying that of the six lines of text, the three on the bottom are Turkish and the three on the top are Greek ( a l t i üg satir türkqe ve iisti iig satir Rümca xatlar tahrir édüp iisti Rümca olmu§ olur).' He goes on to say that although two thousand years have passed since this inscription was composed, apparently Turkish was spoken in those days. As confirmation, he cites the ancient grave inscriptions in Daghistan. He also notes that the Prophet sent letters to several peoples in the Caucasus and beyond, including the Turks, who responded in their own languages and accepted Islam. Furthermore, the Koran commentators maintain that Husheng, and others who boarded the arc with Noah, spoke Turkish; when asked what language it was, Noah replied: "It is the language of the future world conquerors." 2 Elsewhere Evliya states that "Turkish, which is the language of the Tatars, became widespread with the prophet Esau." 3 We find more "information" about the Turkic languages in the sections before and after his specimen of Mar'ash Turcoman dialect (see above and Appendix B): In these mountains and v i l l a g e s dwell the Turcomans, w h o speak an ancient tongue deriving from that of the Tatars of Bukhara and of [blank]. The true language of the Turks (or, the language of the true Turks [lisan-i Etrak-i sahlh]) has t w e l v e branches. Originally they c a m e from Mahan in Transoxania They set f o o t in these regions of Anatolia a l o n g w i t h the D a n i s h m e n d i d s , the A q - q o y u n l u , and the Seljuks. Each group o c c u p i e d a g i v e n territory, and each has its o w n peculiar dialect. The ancient Turkish language is the language of the Tatars, w h i c h also has t w e l v e dialects, and this Turcoman dialect is derived from them. ... T h e s e T u r c o m a n s have hundreds o f d i a l e c t s w h i c h are m u t u a l l y i n c o m p r e h e n s i b l e without interpreters. But the most elegant is that of the Chagatay Turks. Many p e o p l e s — M o g h o l , B o g h o l , Etrak-i Quzaq, H e s h d e k , Daghistan, L e z g i , Qumuq, Tatar of Bukhara, N o g a i Urumbet, Ulu N o g a i , Kichi N o g a i , Shidaq N o g a i , Haydaq, Badraq of Crimea, etc. — s p e a k the o l d Turkish or Tatar l a n g u a g e . T h e T u r c o m a n s and the O t t o m a n s derive f r o m these. But the Kalmuk Tatars, Chin, Faghfur, Khita, Khoten, and all the Kalmuks extending b e y o n d M u s c o v y as far as the Land of Darkness — these are a different kind of Tatars. 4 W i l l 374al3. 374a29: Bu lisan sahibi anlardir kim bu cihanuh edim-i arzi anlar ciimle qabza-yi teszlre alacaqdtr. 3 I I 256al6 (99): Hazret-i 'Isdan lisan-i tiirki §ayi' oldi kim lisan-i Tatardir. For the larger context of this quote, see "Languages of the World . . . " (see p. 259, n. 1). In his treatise on the devil (IV 395al8) he gives the common Turkic (lisdn-i Etrak) name of the devil as Elbi§ — an obvious invention based on Iblis. 4 III 62al7-21, 62bl3-20.

270

FROM

MAHMUD

K A § G A R 1 TO E V L i Y A

£ ELEB 1

This murky passage is Evliya's longest statement on the subject of Turkish origins. What is clear is that he regarded the language of the Kalmuks — the only Mongolian that he had wide acquaintance with — as belonging to a very different branch of the Tatar family, more akin to Chinese than to Turkish. 1 As is the case with all information in the Seyahat-name — whether linguistic, historical, geographical, architectural, etc. — with regard to the Turcological data as well, we must develop criteria by which to separate the factual from the spurious, the bogus from the true. While recognizing that a good deal of the material is only a tribute to Evliya's creative literary powers, we should not lose sight of the residue that may be genuinely useful for scientific inquiry. Appendix A BOLU (II 278a30 - 278b20) aziq tahil dehle zibar gafili tin rähatliq yumi (?) gitmec (?) qizan ciliz cibar merd Safà yalavaç Çalab haqqi Yalavac haqi Ese yalavac çebu faqi heleci aydici bafilayan salä mezgit eyne dami savu yeri sin yeri zähirliq savu göblez tula

GORDES (IX 28a7-14 [57])

ekmek bugday gözet yat araba äväz 'avret

aziq tahil dehle zibar ganh tin rahatliq

qiz u§aq kiiçiik anq adam Ädern peygamber Allah haqqi peygamber haqqi 'ïsâ peygamber çelebe xväce söz §eyx mü'ezzin namäz mescid cum'a cämi'i mezärliq mezärliq hammam aglamaq yavru köpek zagar

gitmec qizan ciliz cibar

ekmek bugday gözet yat 'araba äväz 'avret (?)

gobiez tara§

' On Kalmuk (and Qaytaq), see "Languages of the World

qiz oglan kiiçiik ariq

köpek yavrusi zagar

TURKIC

LANGUAGES

DIALECTS

271

G O R D E S (IX 28a7-•14 [57])

B O L U (II 2 7 8 a 3 0 - 2 7 8 b 2 0 )

gomar masti yetegen ginaq saplica dik dük ki§ne balli dari kelem qizil agag yer sapi piirgiikli hinza sepüger gogag oglan babuci tüglice donbaq tülice yumru dombaq qaplica burqa beg a§i

AND TURKISH

qoyun kòpegi fino kopek tazi arslan kiràz kiràz qurusi vi§ne incir lahanà havug havuc havu^ kereviz turp güne§ armud qurusi dikenli k e s t à n e §eftalü kestàne yumurta yumurta

gomar masti yetegen ginaq saplica dik dük ki§ne balli dan kelem qizil agag

q o y u n kòpegi fino kopek tazi arslan qizd kiràz kiràz qurusi vi§ne incir lahanà Hzavug

hinza yelli saz

kereviz turp

*50ga§

oglan babuci tüglice donbaq tülüce yumru

güne§ armud qurusi dikenli k e s t à n e §eftalü

qaplica burqa beg a§i

yumurta yumurta

BOLU (continued) gullu (jiiciik tavuq me§atliq mezarliq gdynumi§ olmi§ meyve qaq quru qaqac pasdirma faqi xoca zili sini zirfe zagri qa§iq mavmav kedi gelezdi baqdi diziqdim darildim palatiq §ahni§in qiran da§ra hayad avlu yigma depe pilav sara§ erde qari§ qati§ a§ir (?) a§i qaqal maqal qalye burani siqi bulgur goluq dutmag toqmagi herse a§i sarigi burma baqlava gibidir havayi dizliq bork gibidir dongel mu§mula boduq agag budaq yilip (?) palude kekremsi §arab horpuldedir iqcr ziven mi gider misin zilib hall diqi qann agnsi soykenenek yasdiq soyken yasdan qiraq mum api§di sac ayaq tig basdi sac ayaq maqriq mundar musmul temiz qep go? toq toq havan sumeke faydasiz qaqaq egseri kirman ig gigliq sada Ise haqi isa xwace cavvaq lebbey ne yahn nedir halin onat gat eydir ey gizgindi dedim agndi ba§im efelim qarda§im setlek aferim Appendix B: MAR'ASH (III 62a23-25) gelep (galap?) Allah yalavag peygamber yiice gelep 'all Allah heykel tilsim eyne cami' mezgit mescid faqi imam unlen mii'ezzin qancaridaydih nerede idin acarli yeni arvana di§i deve darciqma danlma onat diiz eyi eyle qancari yiligan nereye gidersin be bire gomeg ekmek leva§e ekmek plfi ekmek * In the text this entry appears at the end of the list.

272

FROM

MAHMUD

K A § G A R í TO EVLÍYA

gELEBÍ

(III 62a35 - 62bl2) balli gara harnup kekremsi §aráb dutuq duvaq helat xil'at muhidi geyeñ mi ferráce geyer misin bargim yavincidi qarnim agridi qóngeldim qocadim bu gaz ancilayin düz necear bu kerre §óyle yap dülger qirañda §arli kekremsi hórpüldedir ta§rada §ehirli §aráb igerler irmaga gimen mi u§aq suya girer misin oglan na§i avanla hele§i mi ideñ he mi u§aq bilmedigiñ adamla sóyle§ir miydiñ oglan eyne daminiñ faqi muzallqidir §u ki§i cámi'iñ imám-i kebiridir o ki§i ham halet (jergi esbábi §ariqdi §ehirli oldi sindi maqás emcik meme dam dazlaq girgiplaq derimde baña oyañ mi diyeñ be u§aq gadirda baña beñzeye yi (?) mi dersin bire oglan gilli gi§li §arli gara hórpüldedirler kinli kibirli §ehirli qahve igerler alat samat tiz fevrice óyküneñ mi ki§i uyar (?) misin adam be nideñ §u qepeli bire neylersin §u murdari belki dehleye gen yerden belki gózede uzaq yerden el bizi de dañlayalar ki§i xalq bizi 'acebleye adam gózgü ayna tutuma geldi fikrime geldi gókcek güzel cilliqdirir bagrim oynadir yüregim tahil bugday dehle gózet birikdi alañ yére cum cumat bir yére geldi meydana cümle cemá'at qoqu§mañ qizañ ü§meñ oglanlar tara§ zagar góblez kügük zagar yitgen tazi tula zagar maftil bázü-bend avnadim yuvarlandim ha§al it uyuz kopek sanqi bengi Appendix C: CHITAQ (III 120a20-21) alafia tiz ol qavra getir ocaq getir áte§ getir hi§ir qavun hi§irqa qarpuz qogaq güne§ alaf saman qalayiq cáriye baytali aqit ati sola qopayin dibelik ne apayin ne i§leyim defge qiz qalma sanq meqikli tá'ün bórk qalpaq üg beni bekle beni aydi ver sóyleyi ver ocaq qafvjra áte§ getir keremiñ aydi vereyim eyligiñ soyleyeyim selam ayitdim selám 'aleyküm qorbaci ne apar ayd agañ ni§ler sóyle tahirla bekle qidi ver eyi ver §ikast oldum xasta oldum yundum qoluni qisragim tayi órp at ta§agi baytal caví at siki (III 125a29-35) hi§ir qavun hi§irqa qarpuz qogaq güne§ qalayiq cáriye qavra getir baytal at baytali aqit ati sola qopayin dibelik napayin ne i§leyim defge cáriye qizi qalma sanq meqikli tá'ün bórk qalpaq üg beni bekle beni aydi ver sóyleyi ver ocaq áte§ góyün ocaq ocaq qavra áte§ getir keremiñ aydi vereyim eyligiñ sóyleyi vereyim selam ayitdim selám 'aleyküm qorbaci ne apar ayd agañ ne i§ler sóyle tahirla te'xir eyle §ikast oldum xasta oldum yundum qolunudir qisragim tayidir órp at hayási baytal caví at zekeri qidi ver eyi ver

TURKIC

LANGUAGES

AND TURKISH

DIALBCTS

273

Appendix D: TATAR and NOGAI TATAR (VII 133Ò5-27) galap Allah ismi yalavac peygamber ismi kòmec ekmek potga lapa oyan gorba may yag kemertme ahlad armudi gom qizilciq salabci legen qumgan ibriq gòmlek bardaq ayaq ganaq bulduq me§rebe melenrek orta bardaq qo$ qulaq iki qulpli bardaq tupan (?) saman gepeg tavuq kóykenek kerkes qargi atmaca bükterek delice qu§ topalken gaylaq la g in §áhin qargiga dogan biidene bildircin torgay serge kilcügen oq yelegi qu§i bural ala§a qil at gapdar at quia at ceyren at al at boz at qir at agirmaq kiiheylàn at ala§a beygir at taylaq tay cügürgen segirdicidir ere§ (javdar aqay aga abi ana §ura bar mi kòle var mi devge qiz marya qan xalàyiq toxta dur yax§i eyi cax§i eyi y aman fenà caman fenà atmaca bursun ta'ün vursun batir cigit as bolsun yarar yigit a§ goq olsun qart ki§i qoca adam gòngelgen egrilmi§ neyleyek kòngiil dü§ger neyleyeyim goñlüm dü§er òziimge baña seni severmen seni severim kòp goq talqan qavrulmi§ dan uni qimiz qisraq südi yazma yogurd ayrani bizqa bar mi boza var mi qonaq yolda§ cügür git tileyim (?) bar aytiyim isteyecegim var sòyleyem bir sadaq bergi manga cigit bir tirke§ veriñiz baña yigit §idaqm bile bolsun qamgisi bile olsun kirman ^àVcxan padiíjáh carliq emr-i §erlf yarhq emr-i §erif salay kògler beceneyk (?) sarp ormanlar bar ki§i var adam batir ki§i coluq bolsun yarar adam yoluñ àsàn olsun alay bolsun eyle olsun qizilga altun aqina barup doyum bolganda bergenim bar sefere varup mài bulursam verecegim var galap tanigandir Allah gòriicidir cax§i devgeler tilerseñ qo§umuza kelci eyice qizlar isterseñ ordumuza geliñiz ózüñ bilir batir sen bilirsiñ yigit qayda barasin tentek nereye gidersin kopek ózüñi soqarim seni óldürüriim ózüñi atmaca tapsin bursun seni yumurcaq bulsun vursun ózüñi soqar miltiq cügürme cigit seni oldürür qur§um gitme yigit qutaq edebde ta§aq kótlek §ura pu§t kòle azingi burnuñi gom yemi§ yite o§atirim seniñ ózüñi agziñi burnuñi qizilciq yemis kòpege beñzedirim seniñ kendiñi NOGAI (VII 147b3-13) pir 1 ikiz (?) 2 üs 3 tórt 4 be§ 5 alti 6 cedi 1 sekiz 8 toquz 9 on 10 galap Allah caradan Allah yalavag peygamber coluqduq ràst geldik sadaqlap tirke§li savat pusat kübeli zirxli zor cigit quvvetli yigit batir cigit bahadir yigit gapul civer sefer qoyver qo§ qazan agirliq yüki §ura oglan muña civiqdir i§te yaqindir bir ba§ urtim bir ricà etdim tuvalar yasadim du'àlar yapdim aytim bar aytayim sozüm var sòyleyim cügür ki§i segird adam alay col bersin Allah yol versin §uralarim bar ogullanm var qara§ilarim i§ erlerim toyum bolganda doyum olanda a§iqmañ 'acele étmeñ potga a§i lapa ta'àmi tirek aga5

274

F R O M M A H M U D K A § G A R I T O E V L1 Y A Ç E L E B Î

Tekerleme: Osurgamn orunbolsun / qol sòykede barati bolsun / qo dilaga pindìrsin / qol mòykede sendirsin / cavul atqa pindirsin / xàn sòykede sendirsin / tamuq taqi§ qan çatlasin / tava içki aylasin Tekerleme: Ay Nogayli Nogayli / arpacigi fugali / kôturup soqmaga qolayli / valayli vay valayli

Appendix E: Conspectus of Turkic Languages and Turkish Dialect Materials* I 69bl7, 70b 17,20 (251: omitted), 110b27 (373) Azeri 132b2-10 (441: omitted) Fireng 153a35 (512) "Chagatay" 170a34 Albanian sweetmeats cooks: aa'à çelepi. . . 170b8 Greek tripe-cooks: Al-a iki axça Axmet Beçeden 180a9 Manavgat: Çuluna nehnii gerek nihnu 185bl5 Laz copperware merchants 205al7,20 (629-30), 205b23-25 (632) Armenian ass-drivers (andira xalasi), stone-draggers, whitewashers, sewer-cleaners II 231a24 (34) Bursa: Memet Çebu (= çelebi), etc. 249b21-31, 252b34f. (84, 92) Trabzon: xuta-verti Pe§e, etc. 278a30-278b20 (175) Bolu 281b33 (190) Amasya 288a21 (214) Erzurum 292a27f. (227), 301bl7f. (258), 314a6 (299) Azeri 300a4 (253) Tebriz (Azeri) 324b 16 Qars: . . . olaq (333: olalim) 329b 12, 335a9 (350, 365) Seydi Ahmed Pasha (Çerkes, Abaza) 352a9-10,16-25 (413-14: mostly omitted) Chorum 362b25 (447) Varvar Ali Pasha (Bo§naq) III 13b8 (22: omitted) Qonya 62a22f. (171-73) Mar'ash 65bl0 (182) Qayseri: Ebbe bire alatla bire gôtii beri 94b9 (251) Chefa: yiliq-, ivan-, etc. 98b32 (266) Qudde Ketxiida (Kurd = Diyarbekir) 99b29 a rebel soldier sings quatrains (Eastern Anatolian): kesipsen, etgilen, menirn (270: kesdin, eyle ki, benim) 103bl7 (285) Laz 120al9f. (338: omitted) Silistre (CHITAQ) 125a25f. (358) Dobruca (CHITAQ) * Includes a few examples not given in the text above.

TURKIC

LANGUAGES

AND TURKISH

DIALECTS

165a26 (468) Edirne 184b24 (530) Qudde Ketxuda (Diyarb) IV 194b27 (12) Malatya: helemet sogayi getir (elbette bardagi hemesi puzandadir (cumlesi bagdadir) 207a33, 207b4-26 (50-51) Diyarbekir 216b8 (71: omitted) Ali Faris (Diyarb) 221bf„ 226b9f. (82f., 96f.) Bitlis (Rojiki = Kurd) 239b34 (138) "ancient Turkish inscription" in Old Akhlat 252b31, 258a31 (172, 186) Van

275

getir)

289bl8,22, 290a34-293bl9, 296al3,35, 298al6f., 299a5-300a24, 301bl3-303bl5, 304bl0, 335a33f., 355b28, 365al4 (284f.) Azeri 310b5 (348) Hamadan 403al8 Mosul V 20al5 (56) Toqat: Armenian taunt 20a28 (57) Varvar Ali Pasha (Bosnaq) 43al8, 52a31f. 137, 169-70) Tatar 96a7 (319) Ece in Gallipoli: comerdim qaldim (baqa qaldim) ebbe (te'accub etmek) keremgeli (eylik edici) 116al (382) Belgrade 143a 19f. (477-78) Melek Ahmed Pasha's guards (Abaza) 146a3f. (486: omitted) frontier warriors (Bo§naq) 161a22 frontier warriors (Croatian): goreliiz, etc. (527: gorelim, etc.) 171b23 (559) Uskiip 178a26 (576) Gol-i Kesri 184b 18, 186a9 (593, 597) Isma'il Pasha (Bo§naq) 184bl9 (593) Seydi Ahmed Pasha (gerkes) VI 19b5 (64) Tatar 80a30 (231-32) Jews of Buda: yalayalum oleliim mi yoqsa qurtulalum mi 88a9 (247-48) Buda (Bo§naq) 90bl8,26 (255: omitted) Pest: dovletli vezir bozmaya Gazi Suleyman xanun qanunin . . . 185b20 frontier warriors (Bo§naq): durmayaluz, etc. (544-45: durmayalim, etc.) VII 105a25 (495), 105b5 (496), 107b27f. (509), 112b34, 113a20f., 115a2f. (545), 128bl2,34 (616,618), 133al7 (639), 133b (641), 144a (69395), 155a21f. (750), 158b4f. (766), 159a8 (769), 165b21f. (799), 172al6 (829), 196a27,31f. Tatar 147b (712), 152b30 (739) Nogai 160b33f. (779), 170b8 (822) "Chagatay"

276

F R O M M A H M U D K A § G A R i T O EV LI Y A Q E L E B I

VIII 196a33 (35), 2 0 0 a l 2 Tatar 207b26 Megre and Qaraciq (Axiryan Qitaq): hululama ('acele etme) gilimiz de bile misik (ciimlemiz bile idik) saryiga edelim acayur (mu§avere edelim nice olur) sulubayimiz negik buyurami§ eylece edesik (voyvadamiz her ne emr ederse eyle edelim) an qolayina (filler in speech) 218al4 (125) Dirama 220a30 (133) Siroz 231a7 173) Vardar Yenicesi 232a5 (179) Vodina 2 3 5 a l 7 (191) Alasonya 236b31 (197) Yeni§ehir 237b6 (199) Ermiye (= Urumga) 239b24 (207) Tirhala (= Urum§a) 259al (280) Corinth: gormifik, etc.; qaqomir (bed-baxt, siinepe (napak) qalimbuq (Misir bugdayi)

zavalli)

helefus

IX 28a7-14 (57) Gordes 274b20 (602) gahbe cirbagi bacaqli Arablar (lehce-yi Etrak) 368a2,12 (799) hay benim olda§im ve qarinda§im, men yax§i endazem, anlarih birinin canin apara bilemiz (Ozbekqe) X 4 2 b l 0 , 2 8 (96), 4 3 a l l - 43b6 Azeri 73b 19 - 74a12 (159-60) Egyptian Mamluk 124al8(271) Bo§naq

tir-

19. THE LANGUAGES OF THE WORLD ACCORDING TO EVLIYA CELEB I

To the master of Turkic and non-Turkic languages: al-mar'u maxbiiwun tahta

lisanihi

Yakist turki vu tazi darln mu 'amala Hafiz Hadith-i 'isq bayan kun bidan zaban ki tu danl

For roughly forty years, from 1640 to 1680, Evliya C^elebi, who was born and raised in Istanbul, travelled throughout the Ottoman domains, and into its peripheries — west as far as Vienna, north to Kiev and the land of the Kalmuks, east to Tabriz, south to Sinnar and Abyssinia. Wherever he went, he recorded local speech, and included the samples in his voluminous Seyahatname. Although he states repeatedly that the world-traveller must know 147 languages, 1 he gives specimens of some thirty non-Turkic languages, plus samples of at least the same number of Turkish dialects or other Turkic languages. 2 Evliya's mother was Abkhazian, and it is perhaps not wholly accidental that the first full-blown specimen of a foreign language in his work is that of his mother's tongue. 3 After all, Evliya already knew quite a bit about the Caucasus, and he took the first opportunity to travel there. In addition to Abkhazian, he recorded some Ubykh, Qaytaq, Georgian, and Mingrelian. 4 These have been known to Western scholarship since 1850 when Hammer published his English translation of book II of the Seyahatname.5 Subsequent studies brought these specimens to the attention of Caucasian ^ . g . , I 151a7 (503; tr. Hammer 118461, ii,99), IV 2 4 1 b l 4 (141 - omitted), VII 149b35 (724); but at III 29a22 the figure is 140; and at X 88b26 (190) he states that 140 languages are spoken in Egypt alone. References to books I-VIII are to the autograph ms. as follows: Bagdat 304 books I and II Revan 1457 book VI Bagdat 305 books III and IV Bagdat 308books VII and VIII Bagdat 307 book V Reference to book IX is to Bagdat 306 or to P (= Pertev Pa§a 462). Reference to book X is to IUTY 5973 or to Q (= Begir Aga 452). Other page numbers (without a or b) refer to the Istanbul printed text, 10 vols., 1886-1938. 2 S e e Dankoff (1990). 3 I I 258b (109). 4 I I 258b (109), 311b (291), 320a (319), 332b (359). ^Hammer (1850), 58: Abkhazian and Ubykh (= Abaza and Sadga Abaza), 173: Georgian, 197: Mingrelian. Hammer unaccountably omitted the Qaytaq on p. 157.

278

FROM

MAHMUD

K A § G A R I TO EVLlYA

Q ELEB i

specialists. 1 As for the Qaytaq Mongol, Bartold touched on it as early as 1910, and Pelliot, who studied it exhaustively, concluded that Evliya's specimen is based on a written source, not an oral informant. 2 Before leaving the Caucasus we will jump to book VII, since Evliya returned there in 1666. Now, however, the obscure Caucasian languages seem to have caused him some trouble. Although he mentions "Daghistani" in Tarkhu, he leaves a blank space for the specimen. 3 And of Circassian he states that it sounds like the cawing of magpies, and therefore he was unable to record it. 4 He does however give a sample of Kabardian. 5 And when he ventures into the land of the wild Kalmuks, he does learn their language. 6 The other foreign language specimen in book II is Persian, the dialect of Nakhshivan in the Caucasus. 7 Evliya records other Near Eastern languages in books III and IV: Arabic in Damascus, the curious Teymani in Jabal Naqura, Ladino in Safed, Armenian in Sivas, and Kurdish in Miyafarqin. 8 Hammer brought attention to the Kurdish specimen as early as 1814, and more recently Van Bruinessen has devoted to it a full-fledged study 9 . Armenian, one of the two languages entirely omitted from the Istanbul printed text (the other was Albanian) was published for the first time in 1983. 10 By contrast with the Near Eastern languages, the European languages which Evliya records in books V through VIII have received a good deal of scholarly attention — naturally, by European scholars. Although he gives

' Bleichsteiner (1934), 103-09: Abkhazian, 109-16: Ubykh, 89-96: Georgian, 96-102: Mingrelian; Dzikija (1936, 1954): Mingrelian; Dumezil (1978) and Provasi (1984): Ubykh. [And see now Gippert (1992). | 2 Pelliot (1927, 1930-32). 3 V I I 163a (789). He also leaves a blank space for Tat at VII 136b (656); the reference is to a people (the Crimean Goths?) living on the S. Crimean coast: cf. W B iii, 899; EI (1) "tat" (Minorsky). 4 V I I 149b35 (724; cf. 779); also at I 151al0-12 (503; tr. Hammer 118461, ii,99). Compare his remarks on New World languages at IV 241 b22: Qerkes lisani gibi bogazdan soylenir, qaleme gelmez elfdzlar var imi§. He also says of Abkhazian and Ubykh that they are like the speech of magpies (II 259a5). 5 V I I 157b (763). This was included in Bleichsteiner (1934), 117-24. 6 V I I 179b (866). The Kalmuk specimen is included in the Russian translation of this portion of the text (Kniga 2, 1979, p. 176), but unlike the Qaytaq has not yet been the special subject of study by a Mongolist. Note also VII 176b 7 I I 295b (238). It was omitted by Hammer (1850), p. 128, and has not been the object of scholarly interest, though it contains some unusual forms (e.g. sanzdeh for "thirteen") and some curious names of foods. 8 III 29a (69-71), 40a (104-05), 44b (119-21), 79b (210 - omitted); IV 219a (74-76). ^Hammer (1814), Bruinessen (1985, 1988). 10 Dankoff (1983). And see Weitenberg (1984).

THE

L A N G U A G E S

OF

T H E

W O R L D

279

nothing in Polish beyond the numerals, 1 and barely mentions Russian, 2 he does give full-blown samples of four other Slavic languages: UkrainianCossack dialect in Uman (a fortress in the Ukraine), Serbian in Belgrade, Bosnian in Sarajevo, and Croatian in Dirnish. 3 Aside from the Ladino or Judaeo-Spanish, mentioned above, Evliya records two other Romance languages: Italian, from an informant in the Venetian colony of Split on the Dalmatian coast, and Rumanian in Bucharest. 4 Turkova included the Italian as well as the Croatian in her study of Evliya's travels in Dalmatia. 5 The Rumanian was touched on by Guboglu, and recently was the object of a definitive study by Lazarescu-Zobian. 6 Evliya's Hungarian sampling, collected in Nagy-Banya, has long been the object of study by Hungarian scholars, most recently in Hungarian by Ligeti and in English by Halasi-Kun. 7 The Albanian (of Shkodra/Scutari), by contrast, has until now wholly escaped scholarly notice, for the simple reason that, like the Armenian, it was omitted from the Istanbul printed text. 8 The German specimen was studied intensively by Kissling and Turkova, and of course is included in Kreutel's translation of Evliya's account of Vienna. 9 During his sojourn in Greece, Evliya collected a sample of Gypsy in Giimulcine = Komotini, 1 0 and samples of three Greek dialects: in Athens, in Tsakonia, and in Maina in the southern Morea. 11 Evliya's data on the strange Tsakonian dialect have been studied by Pernot. 12 During his travels in the Sudan, Evliya recorded five different African languages: "Hebrew" of Sinnar, "Hebrew" of Boruste, "Abysinnian," "Syriac," and '"Imranl." 1 3 Habraszewski has identified the "Abysinnian," along with the numerals of the "Hebrew" of Sinnar, as Kanuri; the remainder have not been

W 45a (146); in Lvov. V 47al (151 — omitted); in Kiev. 3 V 49a-b (159-60), 116a (382), 131 b-132a (439), 147b (490-91). And note 46b (Kiev). For the Ukrainian-Cossack dialect, cf. Kniga 1 (1961); for the Serbian, Putopis (1979). Note too the blank space for Bulgarian at III 147a2, and the blank space for "Latin" at VI 153bitop (Dubrovnik). [And for Bosnian see now Boeschoten (1992).] 4 V 149b-150a (499-500), VII 102b-103a (482). 5 Turkova (1965), 77-79, 62-64. [And see now Boeschoten (1990).] 6 Guboglu (1967), 46-47; Lazarescu-Zobian (1983). 7 V I 9a (32-33); Ligeti (1971), Halasi-Kun (1979-80). 8 V I 34b-35a (111 - omitted). [See now Elsie (1998).] 9 V I I 72a (321-23); Kissling (1935), Turkova (1952); Kreutel (1957), pp. 198-201. At VII 4b (17-18) Evliya describes a marvelous tree and states that he left a graffito on it in German (!), but he does not give the text. Note too the blank space for Dutch at VI 127b 17. 10 VIII 210a-b (94-95). [And see now Friedman and Dankoff (1991).l 1 'VIII 256b (271-72), 276b (349), 336b (605). 12 Pernot (1934), 497-505. 13 X 415b20 (895), 423a23 (911-12), Q 3 3 9 b l 4 (962), Q339b43 (964), Q340b20 (967-68). 2

280

F R O M M A H M U D K A § G AR 1 T O E V L I Y A

gELEBI

identified.' Finally, Evliya gives some tantalizing specimens of Hindi songs and prayers, some of which he heard from entertainers in Funcistan, others of which he learned from the Indian "Banyan" communities settled in Suakin, Massawa, and elsewhere on the Red Sea coast.2 One prayer — Ized Alla[h] ki penah qalti hurt ("I desire the refuge of God") — became a favorite of his, and turns up several times without explanation. 3 *

Like many other aspects of the Seyahatname, the linguistic specimens follow a set pattern. There is often an introductory section, explaining that the language in question has seven or twelve mutually incomprehensible dialects. 4 The numerals, at least 1-10, come first. The lists always include staple food items, especially bread, water, and meat or cheese. Other terms of daily life generally include some animal names, but beyond that there is a wide variety of what we find in the specimens. Relational terms are not as common as we might expect. Simple verbal commands (Go! Come! Sit!) are common. In six cases (Armenian, Kurdish, Croatian, Hungarian, Albanian, Rumanian) we find a kind of dialogue that reflects a traveller's need to feed his horse: "Bring barley." "There isn't any barley." "Bring barley or I'll crack your head open." (In Hungarian, wheat substitutes for barley.) The lists generally end either with obscenities (twelve cases) or verses (seven cases). 5 Evliya often excuses the obscenities by the vade-mecum argument that a traveller has to recognize them in order to stay out of trouble. So we have variants of "I'll fuck your sister" from the Adriatic to the Caucasus. Qaytaq and Hebrew (or, "so-called Qaytaq" and "so-called Hebrew") stand out as exceptions to this pattern, and in both cases something funny is going on. In the case of Qaytaq, Pelliot showed that Evliya merely took over some of the Mongol animal names from Qazwini's Nuzhat al-Qulub, written three hundred years earlier. There is no reason to alter Pelliot's final

' Habraszewski (1967). Spaulding (1973) was unable to identify the language of the verses in the "Hebrew" of Sinnar. 2 X 422a 12 (908), 422b3 (909), Q339b27 (963). For the Banyan communities, see X 4 3 6 a l 5 (939), 438bl0f. (944), 442b3f. (953). -*IX l b 2 3 (1), X 272b25 (579), 4 5 0 b l 4 (IX, 841). 4 T h u s for Arabic (III 2 9 a l 8 : twelve), Armenian (III 7 9 b l 3 : seven), Kurdish (IV 219a4: twelve), Russian (V 46b22: twelve), Kalmuk (VII 179b9: twelve), and Gypsy (VIII 210bl2: twelve). ^Obscenities: Abkhazian, Georgian, Mingrelian, Armenian, Ukrainian, Italian, Hungarian, Albanian, Rumanian, Kabardian, Kalmuk, Gypsy. Verses: Persian, Teymani, Kurdish, Maina, "Hebrew" of Sinnar, "Hebrew" of Boruste, Imrani.

1"HE

L A N G U A G E S

OF

T H E

W O R L D

281

judgment: 1 "Dès à présent, je considère qu'Evliya-Çelebi a froidement pillé Kazwini et que le prétendu dialecte mongol des Kaitak est une mystification." As for "Hebrew", the second language thus labelled is also, to borrow Pelliot's term, a mystification. It is certainly not Hebrew; and it looks very much like gibberish. Aside from the pretended Sudanese specimen, we find verses in the same "language" elsewhere: Adam's prayer in Egypt, Hagar's prayer in Mecca, and the prophet Qaffâh's prayer, also in Mecca. 2 Similarly unidentified, the Abyssinian language labelled "Syriac" is not Syriac, although it does look like a genuine language. Elsewhere Evliya gives some (genuine?) Syriac verses. 3 The identification of Teymani is also a puzzle. *

Evliya had a good ear, and recorded what he heard in the most exact manner possible given the limitations of Arabic script. Scholars who have studied the specimens as they appear in the autograph ms. agree on Evliya's reliability as a witness to at least one mid-seventeenth century variant of the languages in question. 4 And as valuable as Evliya's information is for some non-Turkic languages, it is even more valuable — and much more extensive — for Turkic languages; but that is the subject for another study. While the thirty or so full-blown specimens naturally form the core of our study of Evliya's information on the languages of the world, investigators should not ignore the rest of the text, since we find references to specific languages scattered throughout, sometimes in places we would not expect. Thus, in addition to aya "saint" and pirgaz "castle" in the Greek vocabulary, Evliya more than once, while explaining proper names of Greek origin, explains that aya means "saint" and burgaz "castle". 5 Similarly, he explains that the ending -var in such Hungarian place names as Zitvar, Uyvar, etc., is the Hungarian word for "castle". 6 In my article on the Armenian specimen I listed several terms (e.g. vank "church") found elsewhere in the text. Similarly for Kurdish, in addition to the words in the specimen discussed by Bruinessen, we find av "water;" bapir "sultan;" a children's ditty; and five

1

Pelliot(1930-32), p. 580. X 2a24 (3), IX 347a20 (755-56), 354al2 (771). [See now Dankoff (2004), pp. 176-84.| 3 I V 399a. 4 H a l a s i - K u n on Hungarian, Dankoff on A r m e n i a n , Lazarescu-Zobian on Rumanian, Bruinessen on Kurdish, Provosi on Ubykh. (Note Bruinessen: "I should not hesitate to call him the first Kurdologist.") 5 e.g. VIII 234b 13 (189); VI 33b8 (108 - omitted). 6 V I 3b24 (13); not in the Hungarian sample. 2

282

FROM

MAHMUD

K A § G A RI

TO E V L I Y A

gELEBI

untranslated expressions (including one obscenity) in the Bitlis narrative. 1 The quotations from a German prayer book are astonishingly faithful. 2 Occasionally Evliya's linguistic identifications turn out to be quite fanciful. For example, when Evliya is in Bosnia, his patron, Melek Ahmed Pa§a, has a premonitory dream during which he sees a bear in the mosque. 3 Following the battle in which the dream is fulfilled, a certain rowdy captive whose name is Vasil is put to death. Upon learning that vasil means "bear," the Pa§a exclaims that his dream has come true. 4 Elsewhere we learn that vasil means "bear" in Greek. 5 The clue to this puzzle comes in Evliya's description of the Gypsy dancing-bear trainers ( e s n a f - i vasilciyan ya'nl ayuciyan). One of the popular names they give their bears is Vasil, as Evliya attests in the ditty they use to start the bears dancing, which begins: Qalq-a 6 beriya Vasil. This was apparently the origin in Evliya's mind of the notion that vasil means bear, a notion that he put to good literary use in his account of Melek Pa§a's dream. * *

*

Evliya had theories, or rather notions, on the origins and relations of the various languages. One statement of his linguistic mythology is found in his introduction to the specimen of Arabic: 7 First God commanded all the angels to speak Arabic. When the prophet Adam entered Paradise out of the earth, Gabriel taught Adam Arabic, and he spoke Arabic with Eve, with the angels, and with God Himself. . . . But after Adam fell from Paradise he forgot Arabic, out of grief at separation from God. Then Adam met with Eve on Mt. Arafat . . . and by G o d ' s command they spoke a language close to Arabic, namely Hebrew. In fact, the word for Hebrew, 'ibri, is merely a metathesis of the w o r d for Arabic, 'arebi; but it is an independent language, recorded below in volume [blank]. When the descendants of Adam spread, they spoke Hebrew, Syriac, and Imrani. This was so until Ishmael was sent on his prophetic mission, when for the first time Arabic appeared among them. A s for Syriac and

1 IV 286a20, 305a22, 3 0 9 b l 8 ; VI 48a9 (bapir, also at IV 213a21, 215b3, 216a31); IV 287b5; IV 224a20, 225a29, 263b9, 265a34, 36. 2 I 2 0 3 a l 6 (623 - omitted; tr. Hammer [ 1846], ii,227), VII 7 2 a l 6 , 21 (322; tr. Kreutel [1957], 199-200). 3 V 135a.end (452). 1 3 6 b l 0 - 2 7 (456 - omitted). 5 VIII 2 1 4 a l 9 (109). The actual word for bear in Greek is arkouda; in Gypsy it is ric. or vasil = "bear" see also VI 50b22; X 100a7-9 (215). 6 I 169al3f. 7 III 29a2 ff.

For vasil

THE

LANGUAGES

OF

THE

WORLD

283

Imrani, these and several (other) languages have survived since the time of the prophet and scribe Idris. A more elaborate account is the following: 1 In Paradise Adam the Pure spoke Arabic and Persian. But when he fell to the earth he forgot Arabic and they spoke Hebrew, Syriac, Dehqili and Deri. Even today in Funcistan and Berberistan and the other countries of the blacks there are various languages spoken. They got by with these languages until the flood. Then from the sons of Noah — Ham, Shem, and Japeth — there arose 72 nations and 72 languages. At the time of the prophet Ishmael, Arabic and Persian were again heard. And thereafter the various races spread over the earth and the various languages spread abroad, so that in each land a different language arose (peyda oldi). The first to invent (peyda eden) languages of various sorts was the prophet Idris, since he was the first whom God invested with the myriad sciences. He was a scribe, and he recorded the scriptures revealed to him and bound them in books. Before the flood he secreted all of these books in the pyramids across the Nile — these had been built by the sorcerer Surid, and are mistakenly known as Pharoah's mountains. After the flood these books were taken out and read by the ancient philosophers. Hence 147 languages spread abroad. Arabic and Persian appeared with the prophet Ishmael; Turkish, which is the language of the Tatars, became widespread with the prophet Esau; in short, from the speech of the Arabs and the Tatars God adorned this world with as many as [blank] various peoples. Those descending from Tatar are: Tatar, Hind, Sind, Mughani, Laristani, Multani, and Banyani; twelve peoples and languages of fireworshippers in India; (Jin, Khita, Khoten, Fagfur, Qozaq, Mogol, Bogol, Turk-Tatar, Ozbek, and 'Acem (Iran); twelve Qumuq in Dagistan; twelve tribes and dialects of Qtlmakh (Kalmuk); Nogay, He§dek, Lipqa, (Jagatay, Lezgi, Giirci, Migril, §u§ad, Dadyan, A?iq-ba§, Ermeni, Urum, Turkmen, Qababita, Israili or Yahudi, Moskiiv from Giirci [sic], Ya'qubi, and Qarayi; also Fireng - they are twelve tribes and dialects, Jews but of the so-called Christian sect (yahudldir amma meslha-mezheb derler) — I§panya, Firansa, Ciniviz, Portaqal, Venedik, Dudu§qa, Sirf, Latin, Bulqar, Khirvad, Luturyan, and Talyan. Of the infidels emerging from 'Acem, four sons of Menucjehr fled and settled in the region of Egre (Egri/Erlau in Hungary). When asked why they came, they replied: men gariz, meaning "We are four men." By corruption of menqar emerged the people of Macar, who are ten times seventy (on yetmi§) infidel peoples, including Orta Macar, Erdel Macar, Seykel, Saz, Cay, Du§aq, Leh, Qeh, Qorol, Tut, Qiraqov, Rus — twelve tribes, Eflaq, Bogdan, Sirce, Isvac, Felemenk, Dunqarqiz, Danimarqa, Ingilis, Nemse, Ingiltere, Di§-Firansa, Khirvad, Macar, Bo§naq. From the noble and eloquent tribes of the Arabs are descended, first of all, in the peninsula of Egypt (i.e., Africa), forty peoples of various stripe: Magribi, Fes, Merankii§i, Afnu, May-bornu, Cicil-qan, Isvani, Sudani, Funci, Qirmanqi, Baganiski, Munci, Berberi, Nubi, Zenci, Habe§i, Kelapisi, 'Alevi, Dumbi; and the Arabs of Yemen, 256a6-36 (99-100; abbreviated tr. Hammer 11850J, 53). Note similar statements at VII 107a26f. (506), X 22b7-14 (47).

284

FROM

MAHMUD

KA§G ARI

TO E V L i Y A

QELEBI

Baghdad, Meval, Mekke, Medina, Badiye (i.e., bedouins), and 'Umman. In toto the Arabs come to 3060 tribes — some say more.

The addition of Persian to Arabic as one of the primordial languages is related to the hadith which Evliya records on the authority of Tafsir alDaylami'} The Prophet once asked the archangel Michael, "Does God say anything in Persian?" "Yes," came the reply. "In the scriptures of Abraham 2 cuz anki peyam arenv' God says: gi kiinem ba in mu§t-i xak-i sitemkaran ("What shall I do with this handful of dirt of oppressors, except bring them a Message?"). Furthermore: 4 Arabic was first heard [after the fall of Adam] and became widespread with [blank], the son of Ishmael. Ishmael himself spoke Hebrew with his father Abraham. In fact, the scriptures that were revealed by God to Abraham were all in Hebrew, which is the language of truth (lisan-i haq). Some words ( e l f a z ) in it agree with Persian, and certain commentators assert that Persian is Hebrew. In fact ... [quote from Tefsir-i Deylemi]. And some commentators say that Persian is the language of paradise. In fact, the grand mufti Kemalpa§a-zade Ahmed Efendi cited this noble hadith: "The language of the people of paradise is Arabic and Persian and Dari." And the Prophet [Muhammed] spoke Persian. ... It is a refined and elegant language, and, since it has common features with Hebrew (lisan-i 'ibri ile mu.perek olundugundan), must be quite ancient.

According to a related legend, 5 the prophet Seth refused to marry his sister, and requested his father Adam to get him a houri instead. God obligingly provided the houri, straight from paradise; but because she spoke Arabic, and he only spoke Hebrew, they were unable to converse until they had a child, who served as interpreter. And that son of Seth is the patron saint of all interpreters. * *

*

' Untraced; Evliya mentions his grave in Alexandria, X 3 2 8 b l 4 (699): E$-§eyx Deylemi, tefsir sahibidir. ^suhuf Ibrahim — cf. Koran 87:19.

Mehemmed-i

^11 2 9 5 b l 8 (238 — garbled) has: pz anki piyam rem', IX 353b23 has: cuz anki peyamram (ms. Pertev Pa§a 462, fol. 1 4 6 b l 4 [770] peyam em); X 423b7 (912) has: giz anki peyam em. 4 I X 3 5 3 b l 5 f . (770). 5 I X 2 5 0 b l 3 f . (552), 259a26f. (569).

T H E

L A N G U A G E S

OF

T H E

W O R L D

285

For Evliya, "Hebrew" is not (what we know as) Hebrew. Rather it is one of several languages once widespread in the world, and associated with such pre-Islamic prophets as Adam and Idris, displaced in the Arabian peninsula by Arabic since the time of Ishmael, but still found in remnant form in Africa. Evliya is drawing here on a wealth of Islamic lore relating to the pre-Islamic prophets, much of it connected with Idris and with Egypt and the Nile Valley. It was Idris who invented hieroglyphic writing, which Evliya fancifully connects with the difficult Siyaqat script used in Ottoman chancelleries. 1 As far as "Hebrew" is concerned, Evliya's information is a farrago of folklore, folk etymologies, legends derived from Arabic literary sources, one or two facts, and possibly some fabrication. Evliya did not know enough genuine Hebrew (or Syriac) to posit a linguistic connection with Arabic. He did sense that German and Persian were related; but the manner in which he states the relation does not warrant calling him a proto-Indoeuropeanist. What he says is: "Many of their words (i.e. in German) are Persian, because they too (i.e. like the Hungarians) came from Persia with the descendants of Menugehr." 2 As Kreutel explains, 3 Evliya's judgment must be based on the striking resemblance of such words as Persian duxter, est, nam, tu and German Tochter, ist, Name, du. The connection of the Hungarians with the Persians is based on a fanciful etymology of Macar, connected both with Menugehr and with Persian men gar "we are four." 4 If remnants of Persian are found in German and Hungarian, remnants of Arabic are found in Albanian as well as in the Caucasian languages — Abkhazian, Laz, Mingrelian, and Circassian — as Evliya recounts in a very improbable legend. 5 Part of this legend goes to show that Arnavud ("Albanian") derives from Persian 'ar-na-bud "there was no shame." This is in line with what may be called Evliya's "etymologitis." Thus Bukre§ ("Bucharest") is from Arabic Ebu Qurey§;6 Trabzon is invariably called Tarabefzun ("joy-increase"); etc. etc. Aside from these flights of fancy, Evliya occasionally gives more-or-less correct etymologies for place names, or genuine folk-etymologies. 7 ' i v 275 b3 ff. V I I 7 2 a l 4 (322; tr. Kreutel [19571, 199). Cf. I 2 0 3 a l 8 (623 - omitted; tr. Hammer [ 1846|, ii,227): "many of their expressions (elfaz) are Persian." 3 Kreutel (1957), p. 263. 4 Evliya repeats this at every opportunity, e.g. Ill 137a22 (392 — omitted), V 126bl7-22, 31-32 (422 — omitted; here he says that many Persian words are used in Hungarian), VI 9a36 (33 — omitted: "many words in this language [i.e. Hungarianl are Persian, because in origin they stem from the children of Menu?ehr . . ."), VII 67a26 (301; tr. Kreutel, 165), IX 373b2 (811), X 34a21 (76). 5 V I 34b22 ff. (111 - omitted), VIII 353a (671-72), 367b (729-30). 6 V I I 100al5 (467). 7 For Armenian examples, see Dankoff (1983), 54 (on f a p a q 9 u r ) and footnote 14 (on Sivas). 2

286

FROM

MAHMUD

KA§GARi

TO

EVLiYA

Q ELEB i

One language preserved apparently uncorrupt from antiquity is Greek, of which Evliya states that it "was heard from the wise men of King Goliath during the time of the prophet David, and became well-known during the reign of Jereboam the son of the prophet Solomon. During the time of the prophets Zachariah and John (the Baptist) and Jesus ... everyone spoke Greek, so God revealed the gospel to Jesus in that language. Afterward it was translated into Latin, Syriac, and Coptic." 1 Following this statement he gives an apocryphal quote from the Gospel concerning the Prophet Muhammad; the same Greek quote is cited twice elsewhere in his work. 2 "Latin" in Evliya's mind is confused with the Slavic languages. In his introduction to Polish he states: 3 "Their language is the ancient tongue of Latin (latince). There are twenty different people that speak Latin, including those of Dubrovnik, Bosnia, Herzegovina, the Bulgars, the Serbs, the Voyniks, the Slavonians, Qorol, Dudu§qa, the Croats, the Slovaks (Tot), the Ukrainians (Rus), Lipqa, the Russians (Masqov), Cracow, and the Bashkurts (He§dek)." Similarly in his introduction to Croatian: 4 "Croatian is Latin. But the Croats speak a special dialect, with terms and locutions all their own. In fact the Croats, the Bosnians, the Serbs, the Bulgars, the Voyniks, the Slavonians, Dudu§qa, Qorol, the Poles, the Czechs, Lipqa, the Ukrainians, the Russians, and the Bashkurt Muslims — all these peoples speak Latin . . . but each has its own peculiar expressions and dialect." The Romance languages, on the other hand, come under the rubric "Italian." In his introduction to Italian he states: 5 "Language of the war-like Franks (Fireng-i pur-ceng). Their language is called Italian (Talyan); but Spain, France, Genoa, Livorno (Alagorna-Giranduqa), Portugal, Antwerp (? Dunqarqiz), Denmark, Holland (Felemenk), and England — all of these kingdoms are Franks. ... Their language is Italian; but each one has its own special dialect and terminology, and they communicate with one another only with interpreters. The most eloquent is the language of the Frankish Venetians. ..." *

1

VIII 265b5 (271). III 39b27; X 235bl6-18 (502-3). The passage in book X is edited in Haarmann (1988). 4 5 a l l . Cf. V 146a34 (487 — omitted): lisan-i bo^naqda ya'ni latin lisaninca. 4 V 147bl0 (tr. Turkova, 62). Turkova (1965), 130, has notes on Voynik (a class of frontier soldiers), Qorol (? — she suggests Friul), Lipqa (she reads Liqa, it. Licca on the CroatianDalmatian border); elsewhere (1950) she identifies Dudu§qa with Korushka in Slovenia. 5 V 149b33-34, 150a2-3. 2

THE

LANGUAGES

OF

THE

WORLD

287

Like many travellers, Evliya enjoyed comparing p h e n o m e n a f r o m d i f f e r e n t areas. L i n g u i s t i c a l l y , this t e n d e n c y can b e o b s e r v e d in his a t t e m p t s to r e n d e r a g i v e n t e r m in all the l a n g u a g e s h e k n e w . A s i d e f r o m " I s t a n b u l " 1 h e d o e s this s e v e r a l t i m e s f o r " h o t s p r i n g " 2 a n d o n c e f o r " c r o c o d i l e . " 3 T h e m o s t a m b i t i o u s e f f o r t in this r e g a r d is f o u n d in his treatise on t h e d e v i l . S i n c e t h e f i n a l q u a r t e r of b o o k I V h a s n e v e r b e e n p u b l i s h e d , a n d s i n c e t h i s D e v i l ' s Dictionary illustrates E v l i y a ' s linguistic horizon, w e end this paper with the section on " d e v i l " in all t h e l a n g u a g e s of t h e w o r l d . 4 The devil (iblis) has seventy names in the Torah. It would take too long to list them all. First in the book of the torah one of his names is: Yeser-hera' Harithum.-' He is not named in the Psalter, since the Psalter does not contain p r o m i s e and threat, legend, c o m m a n d and prohibition, but consists entirely of prayers. [Cites Koran 4:163 = 17:55.] In the Gospel the devil is named seven times, together with the names of the jinn — a fact that has led some commentators to include the devil among the race of jinn. His names in the Gospel are: Yugula, Diyavlo, Isqirnut (i.e., Iscariot), Demuna, Sadana. In the scriptures of Abraham he is called: Emare, J u d a j . 6 In the scriptures of several of the prophets he is called Satan (§eytan). In others his name is simply The Carnal Soul ( n e f s ) , and God has warned his servants saying (Koran 12:53): "Oh my servants, beware of The Soul that Commands Evil ( n e f s - i emmdre)." This humble and sinful one, Evliya, in fifty-one years of travel, while sojourning in eighteen empires and kingdoms, (learned) the names of the devil — may he be damned! — in the various tongues and the 147 languages. I have deemed it fit to set them down here as far as possible. Arabic: §eytan, Iblis Syriac: M i m m - m i m m 7 Coptic: Qatirnas 'Imran: M i l f a k 8 Hebrew: Nihab, Huna§, Khajciz RabI': 'Azazil

h 14a.end (55), Hammer (1846), p. 11. II 345b20 (393), Hammer (1850), p. 211; III 139b35 (399); V 174a.mid (567 - omitted); IX 236a.mid (521). 3 X 160bl2-14 (346). 4 I V 395a5-30. ^Heb yeser ha-ra' ("evil inclination") and hartum ("sorcerer"). These are the only genuine Hebrew words in the Seyahat-name. ^Emare for emmdre ("that commands evil" — Koran 12:53). With Jiidaj cf. Adam's prayer in "Hebrew" at X 2a25 (3): Huj-i gijiriba glossed as "beware of Satan" (¡eytandan saqla). ^Appcrcntly derived from Syr. mamuna "riches," figuratively used to mean "devil" in Urmia dialect. I am grateful to Edward Y. Odisho for this information. 8 O r Milgak; included in the "Imrani" specimen at X 968 (= ms. Be§ir Aga 452/2, fol. 340b31). 2

288

FROM

MAHMUD KA§GARi

Jinn:' Qabrum Angels: Harith — his name (ism),

TO EVLIYA Q ELEB i

Ebu Murre — his nickname ( l a q a b )

As I recorded above, in the Koran he is called iblis, Vesvas, Khannas, Yuvasvis,^ and Cinne; in the Torah Yeser-hera'; in the Gospel Yugula, Diyavlo — when the Greeks curse one another they say bire diyavlo! Turkic: Elbi§ Persian: [blank] Hindi:: Chaghatay: Tatar: Kurdish: Khilverik^ Func: Berber: Latin: Georgian: Mingrelian: Abkhazian: Circassian: Russian (masqovca): Polish: Czech: Swedish: Ukrainian (urusca): Didqo 4 Dutch: German: Bosnian: Bulgarian: Albanian: French: Hungarian: Rumanian: Gypsy: Armenian: In sum, there are various names for the devil in the various languages. We have given a brief account. [Excursus on how the Sufis refer to the devil.] The upshot is that in this world, the devil has a name in the seven layers of earth and the seven layers of heaven. Even those who have no language — I mean, the tongueless and the mute — have a gesture to indicate the devil: if they place the index finger^ of their right hand next to their right eye, and crook the finger, it means "devil."

'For examples of "jinn-speech" see IV 3 9 6 a l 0 , 12: bire meded gitdim, bire qaf qaf bije bije vay, semif semif senindir\ and 395b8, where Evliya remarks that "devil-speech" contains many q's, f's and j's. For these three, see Koran 102. ^By "Kurdish" Evliya means here "Rojiki" or Bitlis Turkish dialect; cf. IV 2 2 7 a l 0 , where xilverik is glossed as "black bogey" (qara qoncoloz). < Arm. dial, xlvlik "ghost, phantom." 4 A i s o found in the Ukrainian specimen at V 49b5. 5 salavdt parmagi — also at VI 26a8-9 (= sebdbe), IX 230a3 (508).

THE

LANGUAGES

OF

THE

WORLD

289

BIBLIOGRAPHY Bleichsteiner, R. (1934). Die kaukasischen Sprachproben in Evliya Çelebi's Seyahetname. Caucasica 11, 84-126 Boeschoten, Hendrik (1990). A Short Sample of Seventeenth-Century Venetian Italian in Evliya Çelebi's Seyahat-name. Journal of Turkish Studies, 14, 139-145. — (1995). Bosnische Metrik. In: M. Erdal and S. Tezcan (ed.), Beläk Bitig: Sprachstudien für Gerhard Doerfer zum 75. Geburtstag (Wiesbaden), 3349. Bruinessen, M. van (1985). Onyedinci Yüzyilda Kürtler ve Dilleri: Kürt Lehçeleri Üzerine Evliya Çelebi'nin Notlari. Studia Kurdica 3, 13-37 (1988) Les Kurdes et leur langue au XVIIème siècle: notes d'Evliya Çelebi — sur les dialectes kurdes, Studia Kurdica 5, 13-34 Dankoff, R. (1983). Evliya Chelebi on the Armenian Language of Sivas in 1650. Annual of Armenian Linguistics 4, 47-56 — (1990). Turkic Languages and Turkish Dialects according to Evliya Çelebi. In: Altaica Osloensia: Proceedings from the 32nd Meeting of the Permanent International Altaistic Conference, ed. Bernt Brendemoen, Oslo, 89-102 (2004). An Ottoman Mentality: The World of Evliya Çelebi. Leiden: — Brill; 2 n d edition, 2006 Dumézil, G. (1978). L'Oubych d'Evliya Çelebi. Journal Asiatique 266, 57-66. Dzikija, S. S. (1936). Sovetskaja Jazykoznanie 2, 109-28 — (1954). Evlija Celebi o lazah i lazskom jazyke. Iberijsko-kavkazskoe jazykosnanie 6, 243-56 EI: Encyclopedia of Islam Elsie, Robert (1998). Das albanische Lexikon des Evliya Çelebi (1662), und was ein Derwisch auf der Durchreise alles wissen muss. SüdostForschungen 57, 95-102. Friedman, Victor A. and Robert Dankoff (1991). The Earliest Text in Balkan (Rumelian) Romani: A Passage from Evliya Çelebi's Seyähat-näme. Journal of the Gypsy Lore Society 1.1., 1-20 [Reprint: Victor A. Friedman, Turkish in Macedonia and Beyond: Studies in Contact, Typology and other Phenomena in the Balkans and the Caucasus (Wiesbaden: Harrassowitz, 2003), 156-68] Gippert, Jost (1992). The Caucasian language material in Evliya Çelebi's 'Travel book': A Revision. In: Caucasian Perspectives, ed. George Hewitt (Unterschleissheim/München : LINCOM EUROPA), 8-62 Guboglu, M. (1967). Evliya Çelebi ... de la situation sociale-économique des pays Roumains ... Part 2. Studia et Acta Orientalia 5-6, 3-48 Habraszewski, T. (1967). Kanuri - language and people - in the "Travel-Book" ... of Evliya Çelebi. Africana Bulletin 6, 599-66 Halasi-Kun, T. (1979-80). Evliya Çelebi as Linguist. In: Eucharisterion ... Omeljan Pritsak i, 376-82 [Harvard Ukrainian Studies III/IV] Hammer, J. von (1814). Ueber die kurdische Sprache und ihrer Mundarten, aus dem III. Bande der Reisebeschreibung Ewlia's. Fundgruben des Orients 4, 246-47 — (1846, 1850). Narrative of Travels in Europe, Asia, and Africa ... by Evliya Efendi, I (i, ii), II. London

290

FROM

MAHMUD

K A § G AR i T O E V L Î Y A

ÇELEBI

Haarmann, U. (1988). Heilszeichen im Heidentum -- Muhammad-Statuen aus vorislamischer Zeit. Die Welt des Islams 28 , 210-24 Kissling, H. J. (1935). Einige deutsche Sprachproben bei Evliya Celebi. Leipziger Vierteljahrschrift für Südosteuropa 11, 212-20 Kniga: Kniga Putesestvija 1, 1961; 2, 1979. Moscow Kreutel, F. R. (1957). Im Reiche des goldenen Apfels. Graz Lazarescu-Zobian, M. (1983). Evliya Çelebi and the language of the rebellious Eflâks. Archivum Ottomanicum 8, 308-30 Pelliot, P. (1927). Le prétendu vocabulaire mongol des Kaitak du Daghestan. Journal Asiatique 210, 279-94 (1930-32). Les f o r m e s turques et mongoles dans la nomenclature — zoologique du Nuzhatu'l-kulûb. BSOAS 6, 555-80 Pernot, H. (1934). Introduction à l'étude du dialecte tsakonien. Paris Provasi, E. (1984). Encore sur l'oubykh d'Evliya Celebi. Annali 44.2, 307-17 Putopis (1979). = Evlijâ Celebi, Putopis odlomci o jugoslovenskem zemljama. Tr. H. Sabanovic. Sarajevo Spaulding, J. (1973). A text in an unidentified language of seventeenth-century Sinnar. Meroitic Newsletter 12, 30-34 Turkovâ, H. (1950). Mutmassliche Erklärung des W o r t e s dudushqa im Seyähatnäme des Evlijâ Celebi. Archiv Orientalni 18.4, 317-20 — (1965). Die Reisen und Streifzüge Evliya Çelebis in Dalmatien und Bosnien in der Jahren 1659/61. Prague W B : W. Radioff. Versuch eines Wörterbuches der Türk-Dialecte, i-iv. St. Pétersbourg, 1905 Weitenberg, J. J. (1984). "Evliya Chelebi on the Armenian Language of Sivas in 1650. Some Remarks." Annual of Armenian Linguistics 5, 99-108

20. EVLIYA CHELEBI ON THE ARMENIAN LANGUAGE OF SIVAS IN 1650

T h e T u r k i s h t r a v e l l e r E v l i y a C h e l e b i , in his t e n - v o l u m e Seyahat-name,

i n c l u d e s s o m e t w e n t y - s i x s h o r t g l o s s a r i e s of

work

various

l a n g u a g e s and dialects. O n e of these is A r m e n i a n . T h i s f a c t has hitherto e s c a p e d the attention of A r m e n o l o g i s t s , the reason being that the printed edition of E v l i y a ' s w o r k omits it. 1 ( T h e relevant v o l u m e w a s published in 1896 under conditions of strict c e n s o r s h i p . 2 It is also omitted in the m o d e r n Turkish edition prepared by Z. Danism an. w h o otherwise has restored censored p a s s a g e s . 3 ) T h u s , there is n o mention of it in A . X. S a f r a s t y a n ' s A r m e n i a n translation of E v l i y a ' s work, since this is based on the printed edition. 4 In recent years, one of the Istanbul mss. of E v l i y a ' s w o r k has been identified as the autograph, and current scholarship is f o c u s s e d on this m s . 5 T h e A r m e n i a n glossary occurs in v o l u m e 3, in the course of a description of the city of Sivas (Sebastia) in Central Anatolia, which E v l i y a visited in the year 1650. 6 In his i n t r o d u c t i o n to the g l o s s a r y , E v l i y a a s s e r t s that A r m e n i a n c o m p r i s e s s e v e n " d i a l e c t s " ( l i s a n ) , and i m p l i e s that t h e s e d i a l e c t s are c o n g r u e n t with seven Christian "sects" (firka).

A s i d e f r o m the main dialect,

w h i c h he t e r m s " J a c o b i t e " ( Y a ' k u b i ) , he m e n t i o n s the n a m e of o n e other, " M i g d i s i " . H e a l s o characterizes f o u r others w h o s e n a m e s , h o w e v e r , he apparently forgot, or never knew, since he leaves blanks f o r t h e m in the text. Here is a transcription and a translation of the introductory section:

Lisan-i Ermeni, Can-i Yarmeni demi§ler, Millet-i

Kadimdir

Bu memleket-i Sivas beled-i kadim olmagila banlsi 'Amlak 'asrindan berii kavm-i Ermeni destinde olup 'Is ibni Ishak nebiye miintehi olur. 1 Evliya felebi, Seyahat-name, ed. Ahmed Cevdet, Istanbul, 1314/1896, vol. 3, p. 210. S e e A. Taeschner, "Die neue Stambuler Ausgabe von Evlija Tschelebis Reisewerk," Der Islam 18 (1928), pp. 299-310; "Evliya (xlebi," islam Ensiklopedisi (M. Cavid Baysun). 3 Vol. 5, Istanbul, 1970, p. 98. 4 T'urk'akan Albyurner G (Tureckie Istocniki III): Evliya Celebi, Erevan, 1967, p. 154. 5 R . F. Kreutel, "Neues zur Iivliya-Cclcbl-Forschung." Der Islam, 48 (1972), pp. 269-79. P. MacKay, "The Manuscripts of the Seyahatname of Evliya ('clcbi. Part I: the Archetype," Der Islam 52 (1975), pp. 278-98. ^Topkapi Sarayi, Ms. Bagdat Ko§ku 305, fol. 79b 2

292

FROM

MAHMUD

KA§GARÍ

TO

HVLÍYA

gELEBÍ

Kavm-i cebbarlar idi. Lisán-i Ermení anlariñ z u ' m i n c a 'Amlák peydá etdi derler. Bu lisániñ fesáhat [u] belágati §ehr-i Rüm-i Sivas'da tekelliim olunur. Millet-i Ya'kübidir. B a ' d e h u Hazret-i Mesihe Imán getiiriip kavm-i mesihádan olup c ü m l e Íncílídir. Bunlar dahi yedi firka ve yedi lisán üzerine tekelliim ederler. A m m a ciimlesi 'isevi olmagila incile 'amel ediip tilávet-i incil éderler. 'Arabistániñ Ermenilerine Y a ' k ü b i derler. §ám'da ve Haleb'de ve Irák'da cümle Y a ' k ü b i lisáni tekellüm ederler. Ve áyin-i bátili bi-mezhebleri gayri günedir. Lisán-i Y a ' k ü b i dahi mahalliyle tahrir olunur. Ve kavm-i Migdisi lisám dahi ba§ka lisándir kim b a ' z i elfázi ve nige lugatleri Fársiye yakindir. Gáyet fasih lisánlan odur. Ve lisán-i — -si lisán-i Ekráda yakindir. Ve lisán-i ' A r a b i y e yakindir. Ve lisán-i Gürciye yakindir; bunlar Enü§irvánidir. Ve lisán-i bir lisána mü§ábeheti yokdur; E r m e n i kavminiñ (¿ingenesidir. A m m a cümlesi íncili ve millet-i Mesihidir. Ancak mezáhib-i bátileleri Rüm gibi degildir. ' í d - i b e y z a - y i nesárá gecesi E r m e n i yagli yer ve savm-i b á t i l l a n n c a U r u m l a r sabáhisi yagli yerler. Bu lisán-i E r m e n i d e n dahi seyyáh-i 'álem ve nedim-i ádem olanlara elzem-i l e v á z i m i n d e n lisán olmagila bu kadarca tahrir olundikim b a ' z i menázillerde ve kurá ve kasabátlarda tahsil-i ma'i§et édüp her kesle hüsn-i ülfet etmegi?ün kifáyet mikdári tahrir olundi. Evvelá 'aded-i hisáb-i lisán-i E r m e n i . Yedi 'aded lisániñ h i s á b l a n birbirine mugáyirdir.

Language Community

of the Armenians,

called

Soul

of Yármenl,

an

Ancient

This land of Sivas is an ancient city which has been in the hands of the A r m e n i a n s ever since the time of its builder, A m a l e k . They trace their descent back to Esau, son of the Prophet Isaac. They were a people of giants (or tyrants). According to their own claim, it w a s Amalek who created the A r m e n i a n language. T h i s language is spoken in its (most) elegant and r e f i n e d f o r m in the city of Sivas. T h e y b e l o n g to the Jacobite religious c o m m u n i t y . Since they believe in the Messiah, they are all Christians and f o l l o w e r s of the Gospel. Furthermore, they are (divided into) seven sects and speak seven dialects. But being Christians, they all act a c c o r d i n g to the G o s p e l and recite the G o s p e l . The A r m e n i a n s of the A r a b lands are called J a c o b i t e s . All of them in Damascus, Aleppo, and Iraq speak the Jacobite dialect. And their false and unorthodox rituals are of a different sort. The Jacobite dialect is written down in its place (perhaps refers to the glossary below). The dialect of the Mighdisi people is a different dialect, of which some words and phrases are close to Persian. It is the most elegant of their dialects. T h e dialect of [blank] is close to the language of the Kurds. The dialect of is close to Arabic. The dialect of is close to G e o r g i a n ; these are A n ü s h i r v á n i s . T h e dialect of does not resemble any other language; they are the gypsies of the Armenians. But they all f o l l o w the G o s p e l and b e l o n g to the Christian r e l i g i o u s community. Only their false doctrines are not like (those o f ) R ü m (i.e., the Greek Orthodox). The A r m e n i a n s eat oily f o o d s on the eve of the Christian Festival of the E g g (i.e. Easter), while the U r u m s (i.e. Greek O r t h o d o x ) eat oily f o o d s on the following morning, according to their false fast.

THE ARMENIAN

LANGUAGE

OF S I V A S

IN

165 0

293

The world traveller and boon-companion of mankind greatly requires (to know) some of this Armenian language. Therefore, this small amount has been written down here so that he will have a sufficient amount to get sustenance in the stages of his journey and in the villages and towns, and so that he will be on good terms with everyone he meets. First are the numbers for counting in Armenian. The (forms o f ) the numbers are (slightly) different in the seven dialects.

In the following edition of the glossary, each entry contains four elements: - A transliteration of the Armenian, in caps. Note that' stands for alif, and a raised dot (•) stands for sukun (sign of quiescence or no vowel). - A transcription of the Armenian. This is an attempt to record as closely as possible the phonetic character of what Evliya heard from his informant. For a linguistic analysis, see the discussion below. - The Turkish gloss, in modern othography, in parentheses (except for the numbers 1-12, where we find figures in the text). - An English translation. Note that here parentheses ( ) indicate elements that are in the Turkish but not in the Armenian, brackets [ J, elements that are in the Armenian but not in the Turkish. 1. M A R meg 1 2. 'ARKUVK erguk 2 3. 'IRK erek 3 4. CUVRS cors 4 5. HINK hink 5 6. VYC vec 6 7. YVT yot 1 8. 'UVT ut 8 9. 'INIY ins 9 10. DA'S' das 10 11. M A R DAS meg das 11 12. ' A R R U V K D ' S erguk das 12 13. 'AS VA'S- asvas (allah) God 14. IIA'C hac (ekmek) bread 15. CUVR cur (su) water 16. MIS' mis (et) meat 17. GA'VUVX gavox (ya§ uziim) grapes 18. CMIYJ camic (kuru uziim) raisins 19. 'KUV eku (gel) come! 20. KIYNAH kma (git)

294

FROM

MAHMUD

K A § G A R Í TO EVLÍYA

gELEBÍ

21. NISTIY nosté (otur) sit! 22. 'AL el (kalk) get up! 23. KINIY teni (uyukla) sleep! 24. XINJUVR xancor (elma) apple 25. ZUV QINAH PIYR KA'RIY zu tena pér kari (bire yüri arpa getür) Hey, go bring barley! 26. CIYQAH caka (yokdur) There isn't any 27. QINA' PIN DIR DIY tena pandardi (var ara§dir) Go look for some 28. YUVR QIDIY NA'M' yur teclañam (nerede bulayim) Where shall I find some? 29. QINA' KIDIYNIY ZUV KINIY KID KUVNIYM tena kadani zu ganikad kuném (var getür yoksa 'avretiñi kilayin) Go (bring) [find] some or I'll fuck your wife! 30. ZUV QINA' zu tena (bire gel) Hey, come! 31. KUVRTA'NIK BUVSTA'NIH KINXMINK kürtank bostana kini xménk (gel gidelim bostana §arab iqelim) (Come) let's go to the garden and drink wine 32. 'ÁH 'IM HUVKIZ DIGA' ah im hokiz daga (ah camm oglan) O my dear boy! 33. 'ÁXBA'R 'AQUV 'IYNJ QUVZIS 'IMHUVKIZ axbar eku inc kuzés im hokiz (karda§ gel ne istersin camm) Brother, come, what do you want my dear? 34. BAHA'DIR XIS T KIZIY KISIYRIM bahadir xist kézi tesirém (bahadir pek seni severim) My hero, I love you very much 35. YASIZ KZIY KISIYRIM' yes iz kezi tesirém (ben de seni severim) I love you too 36. 'AQUV 'IRTANK MIR DUVNY eku értank mér duna (gel gidelim bizim eve hem) Come let's go to our house (together) 37. BA'K MIYDUVR 'IYNJ BA'KH' ÁH DIGA' bak ma dur inc bake ah daga (bir §eftalu ver baña áh cámm oglan) Give me a kiss O (my dear) boy 38. VA'GIY KY'KVM TUVN KUNIY vaga tekum tun kini [...] (yann gelirim hemán sen §aráb al) I'll come tomorrow, you (get) wine (right now) 39. 'AZ NIS- TINK XIMINK ez nasténk xaménk (oturalim igelim) [Here (?)] let's sit and drink

THE ARMENIAN

LANGUAGE

OF SIVAS

IN

1650

295

40. BH'DUVR 'IYNJ KIGIYNA' 'Z KASIYR KIGINA' ' AQUV KINYK bahadur inc kagana ez kiser kogona eku konik (bahadir ne olursa bu gece olur gel yatalim) My hero, whatever happens will happen tonight, come let's go to bed As T. Halasi-Kun has pointed out in connection with Evliya's glossary of Hungarian, there are certain orthographic constraints that come into play when representing any language in the Ottoman (Arabo-Turkish) script. 1 Thus, to begin with the vowels, a, e, and e (closed e) may be distinguished by writing A ' (or A = alif plus medde), A, and I (or IY or Y) respectively. But there is by no means consistency in this regard, since A may also represent a and I (or Y) may represent i or a. On this basis, we are justified in representing "brother" (cl. arm. [Classical Armenian] elbayr) as 33 axbar and not *exbar. And for "this" (cl. arm. ays) 40 (and 39?) ez is a good guess rather than *az. Similarly, we may assume that Evliya heard a closed e in 3 erek (cl. arm. erek'); cf. the vowel in 6, 21 naste (cl. arm. nste), 25 per, etc. Final -a may be represented by A' or AH, and this explains the spellings of 20, 25, 30 kona, 32, 37 daga, and 26 caka. Rounded vowels are generally indicated by using the mater lectionis V in addition to the vowel-sign U (or zamme). This explains the plene writing in the loan-words 31 bostan and 40 bahadur (note the alternate vocalization, 34 bahadir). The same consideration sheds doubt on the significance of the spelling KUNIY for 38 kini; but reinforces the interpretation of "I'll come" (cl. arm. ka-gam) as 38 kakum, and of "let's go" (cl. arm. kart'ank'), as 31 kurtank. Ottoman script distinguishes d and t, so we may be sure that for "ten" (cl. arm. tas) Evliya heard 10-12 das; for "boy, child" (cl. arm. tlay) he heard 32, 37 daga; for "to find" and "to look for" (cl. arm. gtn-, p'ntr) he heard 2729 kadan- and pandar-; and for "house, give, you" (cl. arm. tun, tur, dun) he heard 36-38 dun, dur, tun. On the other hand, he apparently heard the secondperson possessive pronoun (cl. arm.-d) as -d (in 29 ganikad), not -t. As for g and k, these are not generally distinguished in Ottoman script, except with back vowels when the former may be written G (ghayn) and the latter Q. Otherwise both are indicated by K. Thus for "go!" (cl. arm. gna) it is certain that Evliya heard 27, 29, 30 kana; the spelling with K (20) perhaps represents palatazation. (In the transcription, k is maintained since a spelling with Q perhaps represents velarization.) But in another form of what is presumably the same verb, the spelling with G indicates g 40 kagana! For "to ' T. Halasi-Kun, "Evliya £elebi as Linguist," Eucharisterion: Essays presented Pritsak ...{Harvard Ukrainian Studies, Vol. III/IV 1979-80) Part 1, pp. 376-82.

to

Omeljan

296

FROM

MAHMUD

KA§ G AR i TO

EVLIYA

gELEBi

find" (cl. arm. gin-) Evliya heard 28 kadan-; but also (palatalized?) 29 kadan-. On the other hand, for "there isn't any" (cl. arm. c'ka) he definitely heard 26 caka. And for "come!" (cl. arm. ekur) he heard 33, 36, eku and (palatalized?) 19 eku. Since the preponderance of the spellings indicates that the dialect Evliya is describing exhibits the West Armenian sound shift, the forms chosen for the above transcription are the forms illustrating this shift, thus: 1, 11 meg (cl. arm. mek), 2, 12 erguk (cl. arm. erkuk'), 5 hink (cl. arm. hing), 25 kari (gari), 29 gan- (kn "wife"), 31, 38 kini (gini), 37 bak, bake (pag, pagne), 40 kiser (giser), 32, 33 hoki (hogi), 38 kakum ( kagarri). On the other hand, the voiceless quality of the present-tense particle (cl. arm. ka-) is quite clear in 33 kuzes, and so by extension the forms chosen for the transcription are 31 ktirtank, 34, 35 kasirem, 38 kakum, 40 kagana. In Ottoman script, p may be distinguished from b, but this is by no means done with any consistency. So while it is clear that Evliya heard 25 per (cl. arm. ber); and it is a good guess that he heard 37 bak (cl. arm. pag)-, it is somewhat less certain that he heard 33 axbar rather than axpar. In the same way, c may be distinguished from j. Thus Evliya definitely heard 15 cur (cl. arm. dur). But there is no simple way to indicate such Armenian sounds as j (dz), c (ts), and c' (ts'), or to distinguish between c and cc. These might well be indicated by J or £ or even S in Ottoman script. So it is possible that instead of 6 vec we should read vec'; instead of 14 hac, hac'; instead of 24 xancor, xanjor (as cl. arm.) or xancor (as mod. west. arm. [Modern Western Armenian]); instead of 37 inc "me", inj (as cl. arm.) or inc (as mod. west, arm.); and finally, instead of 13 asvas, asvac (cl. arm. astuac). Other points relevant to the seventeenth-century Armenian dialect of Sivas are the following. Cl. arm. xalol was (apparently) pronounced 17 gavox. Cl. arm. ekur dropped final r in 19, 33, 36 eku. Final s was voiced in 32, 33, hokiz (cl. arm. hogis), 35 iz(k) (isk), and 40 (and 39?) ez (ays); but not in 33 kuzes, 35 yes, 10-12 das, 16 mis. The imperative forms 23 kani (cf. first person plural 40 kanik), 27 pandardi, 29 kadani are noteworthy. We must leave to others the task of relating the data given here to the larger scope of Armenian dialectological studies. We will only note that the standard work of Grigoryan does not discuss the Sivas dialect except en passant. 1 The older work of Adjarian devotes a paragraph to the subject, informing only that cl. arm. e, o are represented in Sivas dialect by the dipthongs, ie, uo. 2 In light of this information, we should perhaps reinterpret as ie what we have suggested above as closed e in words like 3 erek, 25 per, ' A. V. Grigoryan, Hay Barbaragitut'yan Dasmt-ac', Erevan, 1957. Adjarian, Classification des Dialectes Arméniens, Paris, 1909, p. 67.

THE

ARMENIAN

LANGUAGE

OF

SIVAS

IN

1650

297

etc.; as well as uo instead of o in words like 7 yot (cl. arm. eot), 32, 33 hoki, etc. Any other literature that might exist that would shed light on the seventeenth century dialect of Sivas is unavailable to this writer. 1 Finally, we may note that Evliya does mention Armenian language elsewhere in his work. Thus, in connection with a folk etymology of the place name Capaldur, he relates a legend about Alexander the Great, who when he saw the beneficial qualities of Murad River [i.e. Murad Su] said: capakcur. In t h e dialect of M i g d i s i capak jur m e a n s " w a t e r of p a r a d i s e " (cennet suyi). E v e n n o w in A r m e n i a n ( l i s a n - i ermenl-i yarmem) jur

means "water".^

He also assigns to Alexander the naming of Van: "When he conquered Van fortress he gave the name Vank to the church of Goliath (Deyr-i Jalut). In Armenian they call 'church' ( k e n l s e ) : vank [i.e. cl. arm. vank']. ... Even now the name of Van has remained Vank which, by a common misprounciation ( g a l a t - i me.shur), people call Van." 3 Later he states: "Since this region is Armenia, all the churches are called vankBack in Sivas, he lists the names of the Armenian and Greek Orthodox churches ( v a n k - i ermeni ve urum), among which is Tuz Hisar Vanki ("Salt Castle Church") which the Armenians call: SITAHQA 'BIYD (?); and the church of Surb Nsan (Vank-i SIP NS'N). 5 He also relates legends according to which a certain Sap Nasan was the ancestor of Kayumarth, the "Armenian emperor" (Ermen Kaysar) who built Sivas, and whom the Armenians call Sivas-Asvas or Siv-Asvas. 6 And in a discussion of Changly Kilise (in the text: Qanli Kenise-i YarmenI) he mentions that the Armenians call the tomb within the monastery: Sap Garabet (SYPGARAH BIYT). 7

'LSee now J. J. Weitenberg, "Evliya Chelebi on the Armenian Language of Sivas in 1650. Some Remarks." Annual of Armenian Linguistics 5 (1984), 99-108.] ^Bagdat Kö§kü 305, fol. 85b (cf. truncated version in the printed edition |p. 291, n. 1 ], vol. 3, p. 225; Safrastyan fp. 291, n. 4], p. 166). For the correct etymology see H. Htibschmann, "Die altarmenische Ortsnamen," Indogermanische Forschungen 16 (1904), pp. 447-48. See also R. Hartmann, "Zu Ewlija Tschelebi's Reisen im oberen Euphrat und Tigrisgebiet," Der Islam 9 (1919), p. 236. Ibid., fol. 249b (cf. printed edition, vol 4, p. 161; Safrastyan, p. 228). 4 Ibid„ fol. 285b (cf. printed edition, vol. 4, p. 271 — truncated; Safrastyan, p. 259). -'Ibid., fol. 76a (omitted in printed edition). 6 Ibid„ fol. 73b and fol. 78b (cf. printed edition, vol. 3, pp. 197, 207 — truncated; Safrastyan, 149). Behind the Armenian designation lies a play on words with asvas "God" (see glossary above, no. 13). The latter designation (spelled: SIYV ÄSVA'S') might stand for sév asvas "black god" (?). ^Ibid., fol. 87b (cf. printed edition, vol. 3, p. 230; Safrastyan, p. 172). on Surb Karapet, see H. F. B Lynch, Armenia: Travels and Studies, vol. 2 (London, 1901), pp. 174-80; J. Mécérian, Histoire et institutions de l'église arménienne, Beyrouth, n.d. (Preface: 1965), pp. 240-44.

298

FROM

MAHMUD

K A § G A R i TO E V L Í Y A

QELEBI

Near Erzurum, Evliya comes upon the Armenian village of C a ' f e r Efendi and remarks: "All the houses in this [region of] Erzerum have roofs built from ship-masts, called keran ( K A R A ' N ) [i.e., cl. arm. geran

'beam').

They are heavy posts of pine-wood." 1 Concerning the town of Tokat in central Anatolia, Evliya says: In Armenian they call it , but the Armenians speaking Turkish say Toxat. Some of the youthful town wits taunt the Armenians [i.e. make fun of their Turkish dialect] thus: Erzingan'dan ing xabar (What news from Erzincan?); Toxat'ih ga$malari axar mi (Do the fountains of Tokat flow?); Axar amma §ir mir axar (They flow but they flow §ir mir).-

A m o n g the churches of Tokat, he lists (Gatnabur G A T N A ' B U V R ) which, he explains, is Armenian for "milk porridge" (siidli a§ — i.e., cl. arm. kat'napur).3

Finally, in U

Sai«j>t« F I

¡t\-rt i\t>a 4AJ;

f;. >> y J -

I Sit/. i

- ¿ A W Ì - - ? - A : *V •

ii-

¿ i A»* K "

t»-•y

~ j ,v J > ' / A

AA-^'A^V-A Sample G

i

'

*

i'

- AV. *

/V?A

^ A»

-V

'J-*f .>'•

- À C« » V X -J-:—1-A • ^ Y . V X ^ /

IFcriov l';i..s 462: iiSb etidi

o •• i

»•

* '

'

1

^

• i »

^ l i / J - ^ a j^iIr CJ. j j

»^ ^»V

1

¥ * J fk

r* ^-r ttv ^ ^

£ 'sr? < S ) i *

-•'f ^r ^

1

* ''

J J ^A-" a iVV

^V/1

350

FROM

MAHMUD

K A § G A R I T O E V LI Y A Q E L E B I

Bibliography Adivar 1970 A. Adnan Adivar, Osmanh Türklerinde Ilim. Istanbul. Babinger 1930 Franz Babinger, "Ewlija Tschelebi's Reisevvege in Albanien." MSOS As., 33, 138-78 |repr.: Rumelische

Streifen

(Berlin, 1938), 1-40].

£elebis

Reise

Bulut 1997 Christiane Bulut, Evliya

von Bitlis

nach

Van.

Wiesbaden:

Harrassowitz. Dagli and Kahraman 2001 Yiicel Dagli, Seyit Ali Kahraman, ed., Evliya Qelebi Seyahatnamesi

4.

Kitap.

Istanbul: Yapi Kredi Yayinlari. Dankoff 1990a Robert Dankoff, ed., Evliya Celebi in Bitlis. Leiden: Brill. Dankoff 1990b Robert Dankoff, "Turkic Languages and Turkish Dialects according to Evliya f e l e b i , " in: Altaica Permanent

Osloensia:

International

Proceedings

Altaistic

Conference,

from

the 32nd Meeting

of the

ed. Bernt Brendemoen (Oslo,

1990), 89-102. Dankoff 1991 Robert Dankoff, An Evliya f e l e b i Glossary: Words

in the Seyahat-name.

Unusual, Dialectal

and

Foreign

Cambridge, Mass. [In: Sources of Oriental

Languages and Literatures, ed. §inasi Tekin & Göniil Alpay Tekin.] Dankoff 2000 Robert Dankoff, "Establishing the Text of Evliya £ elebi's Seyahatname: Critique of Recent Scholarship and Suggestions for the Future" Ottomanicum

A

Archivum

18, 139-44.

Dankoff and Elsie 2000 Robert Dankoff and Robert Elsie, Evliya Regions

(Kosovo,

Montenegro,

elebi

in Albania

and

Adjacent

Ohrid). Leiden: Brill.

Dankoff and Kreiser 1992 Robert Dankoff and Klaus Kreiser, Materialien A Guide to the Seyahat-name

zu Evliya f e l e b i II (incl.

of Evliya (¿elebi and Bibliographie

raisonnee).

Beihefte zum Tübinger Atlas des Vorderen Orients B 90/2. Wiesbaden: Harrassowitz. Evliya Celebi 1935 Evliya

Qelebi

Seyahatnamesi:

Istanbul: Devlet Matba'asi.

Anadolu,

Suriye,

Hicaz

(Dokuzuncu

Cilt).

SHALL

WE TEAR

DOWN THAT

OBSERVATORY?"

351

Gemici 1999 Nurettin Gemici, Evliyä

felebi

in Medina.

Dissertation,

Otto-Friedrich

Universität Bamberg. Gökyay 1995 O. §. Gökyay, ed., Evliya Qelebi Seyahatnamesi

1. Kitap. Istanbul: YKY.

Haarmann 1976 U Haarmann, "Evliyä Celebis Bericht über die Altertümer von Gize." 8.1,

Turcica

157-230.

Inalei k 1973 Halil Inalcik, The Ottoman Empire: The Classical Age 1300-1600. New York, Kahrman and Dagli 1999 Seyit Ali Kahraman, Yücel Dagli, ed., Evliya Qelebi Seyahatnamesi

3.

Kitap.

Istanbul: Yapi Kredi Yayinlari. K reute 1 1972 F. R. Kreutel, "Neues zur Evliyä-^elebi-Forschung." Der Islam 48, 269-79 Kilisli Rifat 1928 Kilisli Rifat, ed., Evliya Qelebi Seyahatnamesi,

Yedinci Cild. Istanbul: Devlet

Matba'asi. Kur§un et. al. 1999 Zekeriya Kur§un, Seyit Ali Kahraman, Yücel Dagli, ed., Evliya Seyahatnamesi

£elebi

2. Kitap. Istanbul: Yapi Kredi Yayinlari.

Laut 1989 Jens Peter Laut, Materialien Karte B IX 6 "Kleinasien

zu Evliya Qelebi I: Erläuterung

im ¡7. Jahrhundert

und Indices

nach Evliya Celebi".

zur

Beihefte zum

Tübinger Atlas des vorderen Orients B 90/1. Wiesbaden: Harrassovvitz. MacKay 1975 Pierre A. MacKay, "The Manuscripts of the Seyahatname

of Evliya £elebi, Part

I: the Archetype." Der Islam 52, 278-98. Prokosch

1988-89

Erich Prokosch, "Die Gedenkinschriften des Evliyä £elebi." Jahrbuch Osterreichischen

St. Georgskollegs

Istanbul,

des

320-336.

Taeschner 1929 F. Taeschner, "Die neue Stambuler Ausgabe von Evlijä l'schelebis Reisewerk." Der Islam 18, 299-310. Tezcan 1999 Nuran Tezcan, Manisa nach Evliyä f e l e b i ... Leiden: Brill.

26. SOME REFLECTIONS ON THE EDITING OF BOOK 9 OF THE SEYAHATNAME

I. Comparison of the three manuscripts In a recent article, giving an overview of textual scholarship on the Seyahatname, I wrote: For Books 9 and 10 Haarmann 1976 showed that ms. series Y [Bagdat 306 (book 9), iUTY 5973 (book 10); formerly YildizJ is the best exemplar and shloud serve as the basis for a critical edition. The example was followed by Tezcan 1999 and Gemici 1999... All serious researchers, beginning with Haarmann... agree that ms. series Y is superior to the others. Series P [Pertev Pa§a 458-462 (books 1-10]) and Q [Be§ir Aga 448452 (books 1-10)], which were copied from Evliya's original in 1742, do have the merit, as MacKay has pointed out (1975-280) of having tried to recreate their exemplar page by page, and even line by line. Hence we can judge that for Books 9 and 10 Evliya used a larger folio, with 48 lines to the page (whereas in Book 1-8 the folios vary from 34 to 36 lines to the page), since this is what we find in P and Q. Y on the other hand, which was copied from Evliya's original in 1751, used 29 lines to the page and did not try to preserve the foliation of the exemplar. It did however try to preserve all of E v l i y a ' s spellings and vocalizations, which must have been complete for these two books — i.e., like books 1-5, but unlike 6-8, 9 and 10 — must have reached the final fair-copy stage. Why then is the autograph of Books 9 and 10 not preserved, and why instead do we have a very faithful copy of only these two books (= series Y)? I have proposed the following hypothesis: In 1742, when Evliya's autograph was brought to Istanbul, the volume containing Books 9 and 10 was damaged and some portions of the text were illegible. In 1751 an Ottoman official, perhaps Mehmed Rasim, tried to remedy the situation. He must have realized several things: (1) the autograph of Books 9 and 10 was in bad condition and probably unsalvageable; (2) the 1742 copyists had simply leapt over the illegible portions, usually without even leaving gaps to indicate where they had been; (3) those copyist had also failed to render exactly all of E v l i y a ' s detailed pointings and vocalizations, with the result that much information in the original was in danger of being lost. And so he had one of the divan scribes with an excellent hand, a certain Haci Mehmed, recopy these two books with careful attention to all of these matters.... All such garbeled passages occur at the bottom of P folios, indicating that the original damage also occurred at the bottom of Evliya's autograph. Indeed, where evidence of the damage first appears, P

354

FROM

M A H M U D

K A§ G AR I TO

E V LI Y A

Q ELEB i

also leaves blank spaces ... always toward the outer margin of the folio, indicating that the original damage was also at the outer margins. 1

In this article I wish to modify these ideas, based on a closer examination of all three mss. of Book 9 as part of the Yapi Kredi Yayinlan edition project. I no longer believe that Y (ms. Bagdat Ko§kii 306) is superior to P (ms. Pertev Pa§a 462) and Q (ms. Be§ir Aga 452), or that it should serve as the basis for a critical edition. The reasons for this judgment are the following. 2 1) While all three mss. are carefully done copies, all display the usual faults of copies. Thus, all three have skipped lines. An egregious example in P is at 76b22-27 (corresponding to Y 171b, M374) where the copyist skipped three separate lines. Q skipped a line at 52b35 (corresponding to Y 103a, M 212). Y skipped a line at 193a24 (corresponding to P/Q 84a30-31/32-33, M 424). 2) We find many examples where one of the three mss. has a reading different from the other two. In most cases it is very easy to motivate the aberrant reading on the basis of the usual principles of textual criticism (some samples of this motivation are given in Group 3 below, in brackets). Examples: G r o u p 1: YP-j- Q Y 1 b 18 = P lb9 hayvnatlariy§un; Q hayvanat igiin Y 2a3 = P lb20 vacibii'l-viicuba; Q vacibvi'l-viicubda Y 2al 1 = P lb27 vacibii'l-i'zam; Q vacibii'l-iz'an Y 2 a l 7 = P lb32 fethi baki kalan; Q feth u teshiri baki kalan Y 2b3 = P lb46 savabin ruh-i §eriflerine hibe eyleytip; Q sevabini ruhi piir-nur-i §eriflerine ibda ve hibe eyledim Y 3b 19 = P 2 b l 0 hasil edup; Q hasil olup Y 23a5 = P 12bl4 fetalanmizin; Q yigitlerimizin Y 23a6 = P 12bl5 zahrlari fetamiz; Q arkalari yigitimiz Robert Dankoff, "§u Rasadi Yakalim mi? Evliya Çelebi ve Filoloji," in: Nuran Tezcan and Kadir Atlansoy, ed. Evliya Çelebi ve Seyahatname ( D o g a Akdeniz Üniversitesi, 2002), 99-118. The works referred to in this passage are: Nurettin G e m i c i , Evliyâ Çelebi in Medina (Dissertation, Otto-Friedrich Universität Bamberg, 1999): Ulrich Haarmann, "Evliyâ C e l e b i s Bericht über die Altertümer von Gize," Turcica 8.1 (1976), 157-230; Pierre A. MacKay, "The Manuscripts of the Seyahatname of Evliya Çelebi, Part I: the Archetype," Der Islam 52 (1975), 278-98; Nuran Tezcan, Manisa nach Evliyâ Çelebi... (Leiden: Brill, 1999). 2 N o t e that M (for matbu, printed edition) = Evliya Çelebi Seyahatnamesi: Anadolu, Suriye, Hicaz (Dokuzuncu Cilt) (Istanbul: Devlet Matba'asi, 1935). ( — ) = blank in text roughly the size of one word.

SOME

Y Y Y Y Y Y

REFLECTIONS

ON THE E D I T I N G

OF BOOK

9

355

29b 1 = P 15b44 goyner; Q koknár 29b3 = P 15b45 yine; Q bedene 29b8 = P 16a2 3 sá'at dahi; Q 3 sá'atde 30b 17 = P 16bl 1 §e§me-i cán-fezá vardir; Q §e§me-i cán-firávándir 31a9 = P 16b38 kur§umludur; Q kur§un órtülüdiir 31 a9 = P 16b42 fukarádirlar; Q fukarálardir

Since the 1935 edition (= M) is based on Q it includes all these aberrant readings; the correct solution is the opposite of this, i.e. to include in the edition the reading where Y and P are in agreement.

Group 2 : YQ -r P Y 2 b l 6 = Q 2a9 dest-i §eriflerin; P dest-i §erífleriñ Y 2b24 = Q 2 a l 5 can evi; P cásdi (?) Y 3a21 = Q 2a37 mizáb; P miza Y 3a23 = Q 2a39 pencáh; P penca Y 3a27 = Q 2a42 §edd-i rahli; P §edd-i rahl Y 4 a l 0 = Q 2b27 cisrin; P cisriñ Y 4a20 = Q 2b36 evsáñ; P evsáf Y 4a26 = Q 2b40 láciver; P laciverd Y 4a28 = Q 2b41 1077; P 1075 Y 4b22 = Q 3al 1 kal'a-i íznik; P karye-i íznik Y 4b26 = Q 3 a l 6 Sultan Ahmedi; P Sultan Ahmed Y 5a6 = Q 3a23 demir kapulu elli aded; P temiir kapulu aded Y 5a7 = Q 3a24 cámi' sáhibi ... rusas; P cámi'i sáhibi ... sas Y 5a21 = Q 3a36 Hasan; P. Hüseyin Y 6b 17,22 = Q 4a9,14 atli; P atlu Y 6b29 = Q 4 a l 9 kol tüfenglerin; P kol tüfengleriñ Y 7a6 = Q 4a25 eyle; P óyle Y 7a8 = Q 4a26 eliñe al; P eline al Y 7 a l 2 = Q 4a30 al §u mizragi; P alver §u mizragi Y 7a24 + Q 4b46 bir galik yiizlii ve kór gozlii; P bir gal ik yüzlü ve bir kor gozlii Y 8a27 = Q 4a26 güzer; P güz Y 9a23 = Q 5a45 cumie; P cümleden

Group 3: PQ -r Y PQ 3bl8 zevávílerdir; Y 5b28 záviyelerdir [substituted a simpler form) PQ 3b27 be§; Y 6al 1 5 PQ 4a5 te§el§ül; Y 6 b l 2 te§a§ür

356

F R O M M A H M U D K A § G A R Í TO E V L Í Y A

gELEBl

PQ 4b 15 dámeninde; Y 7b22 dámeninden PQ 5a 1 e§-§eyh Ómer; Y 8b 1 §eyh Ómer PQ 5 a l 6 viizeralarin; Y 8 b l 9 vezirlerin [substituted a simpler forml PQ 5a27 defter emtni ve $avu§lar kethüdási vardir; Y 9a2 defter emini vardir gavu§lar kethudási vardir PQ 6a20 ?akilir; Y 10bl9 ?ikarlar PQ 7a3 bir §e§me-i züláli Zeregen; Y 12a20 bir ge^me Zeregen [skipped a word] PQ 7a 17 áb-i hayáti dahi gáyet; Y 12b8 áb-i hayáti gáyet PQ 7 a l 8 güne; Y 12bl0 dürlii [substituted a different word for the one before his eyes| PQ 7a21 filcáni; Y 12bl2 fincani [substituted a more familiar form] PQ 7a30 kavimdirler; Y 12b23 kavimdir P Q 7a32 etmi§lerdir. Ger?i; Y 12b26 etmi§lerdir kim genji PQ 7b39 hala ol háne Zeregen; Y 14a2 hálá Zeregen PQ 7b46 ni^e kerre yíiz biñ; Y 14al 1 nice yüz biñ kerre [transposed two words) PQ 8 a l 7 kible; Y 14b7 canib-i kiblcye [added a word) PQ 8b43 temellerinden; Y 15b28 temelleri PQ 9a 15 zurefá; Y 16a20 zurefan PQ 9a36 bir mülük bóyle hayráta málik olmami§lar; Y 16b 13 bir miílük málik olmami§lar [skipped two words] PQ 9b44 karye-i Komardis; Y 17b24 karye-i kasaba-i irdis PQ 10a20 mahkemede; Y 18a27 mahkemesinde PQ 10a47 gerdir; Y 18b28 (—) [did not know the meaning of ger ("mange") and so left a blank] P Q 10b2 ahalT-i Banaz; Y 19a3 ahallsi PQ l l a l 6 bir kal'a-i Gediis'dür a m m á ne cirmde idigi m a ' l ü m u m degildir. Ve §ehr-i Gediis Germiyan ... ; Y 2 0 a l 4 bir kal'a-i Gediis Germiyan ... [skipped aline] PQ 11 b8 yaylahane; Y 21al yaylaghane PQ 14a 18 etmegi; Y 26a3 ekmegi (substituted a more familiar form] P Q 1 5 a l 4 bir ba§indan bir ba§ina; Y 27b28 bir ba§indan bir ba§ina gelince [added a word] P Q 15a36 ve náhiyesi yedi pare; Y 28a24 ve náhiyesi pare [skipped a word] P Q 16b5 altmda; Y 3 0 b l 0 elbetde [misread a word] P Q 16b6 biñer ádem sigar; Y 30b 11 bárgír ve ádem sigar

SOME

REFLECTIONS

ON T H E E D I T I N G

OF B O O K

9

357

3) There are instances where all three mss. differ. Thus: Example 1: Y 22b25 diigiinler ve gencler, P 12b6 diigiinler ve gencerler, Q diigiinler Here it is clear that the copyist of Q did not know the meaning of gencer ("festival") and so left it out; Y dropped the final letter; only P kept the correct form. Example 2: Y 153bl6 'araban, P 70bl8 'arazbar, Q 70bl9 'urban Here only P has the correct form, a rare musical term (cf. VIII 324a33); both Q and Y misread it, probably because they did not know it. 4) The damaged passages occcuring at the bottom of P and Q folios appear throughout the text, as indicated in the table below. The standard way to indicate a damaged passage was to leave a blank space in the text. Sometimes all three mss. have the same blank spaces. At other times one or two, or all three, try to fill these in, presumably out of their own heads. Example 1: be§ yiiz pare gemi temevvtic-i derya olup karaya dii§erler: Y 192b 1 -nin §iddetinden telattiim P 83b48 deryaya Q 83b48 birbirlerine

ura ura pare pare

Example 2: bir havlisi var amma ortiiludiir, kiremit degildir: Y 10a20 cirmi (—) kire? P 4b45 (—) (—) Q 5b45 ctimle toprak Here the beginning of the unreadable phrase must have been visible, the letter c; Y and Q filled it in, each with a plausible reading, P just left it blank. Example 3: at boynuna Domalic Y 7b2 dii§up ala'l-acele P 4a47 dost deyince Q 4a47 dii§iip andan Here the beginning of the unreadable phrase must have been visible, the letters dus, and each copyist filled it in with a plasubile reading.

358

FROM

MAHMUD

K A § G A R I TO E V L i Y A

CELEBi

C o n s i d e r i n g all of these matters, the only possible conclusion is that all three mss. are independent witnesses of the archetype and therefore are on an equal level as bases f o r an edition. T h e only p r e - e m i n e n c e of Y is its beautiful script and legibility; otherwise there is no reason to privilege it. A s a general rule, wherever t w o of the mss. agree as o p p o s e d to the third, the agreed reading should be restored to the archetype.

II. Table of Damaged Passages In all three m s s . these p a s a g e s are indicated by blanks l e f t by the copyist, unless otherwse noted. In Q and P (with 4 8 lines per p a g e ) these passages occur at the end of the folio, unless otherwise noted; in Y (with 2 9 lines per page) they occur randomly. (-) = blank not indicated * = in mid-folio (pertains to Q and P only) # = orignally in margin? (see part III below)

Q

P

Y

2b

2b (-)

4a

7

3a

3a

5b

10

M

3b

3b (-)

6b

12

4a(-)

4a(-)

7b

14

4b

4b

8a

16

5a (-)

5a (-)

9a

18

5b

5b

10a

20

6a

6a

1 la

22

6b

6b

12a

23

7a

7a (-)

13a

26

7b

7b (-)

14a

28

8a

8a

15a

30

8b

8b

16a

32

9a

9a

16b

34

9b

9b

17b-18a

35

10a

10a

18b-19a

37-38

10b

10b

19b

38

11a (-)

11a

20b

42

lib

1 lb

21b

44

S O M E R E F L E C T I O N S ON T H E E D I T I N G O F B O O K 9 Q 63a (-) 63b (-) 64a 64b

Y

M

63a (-)

133b

282

63b (-)

134b (-)

285

64a (-)

136a

288-289

64b (-)

137b

292

P

65a (-)

65a (-)

139a

295-296

65b (-)

65b (-)

140a

298

66a (-)

66a (-)

141b

302

66b (-)

66b

143a

305-306

67 a

67a

144a-b

310

67b

67b-68a (-)

145b

313-314

68a

68a (-)

68a-b

317

68b

68b

148b

320

69a

69a

150a

323

69b

69b

151b

326

70a

70a

153a

330

70b

70b

154b

334

71a

71a

156a

337

71b

71b

157b

341

72a

72a (-)

159a

344

72b

72b

160a-b

347-8

73a

73a (-)

162a

351

73b

73b (-)

163a-b

355

74a 75a 75b

74a-74b (-) 75a (-) 75b

164b 167b-168a 169a-b

358 365

76a

76a 76b

170b-171a

368-369 372

172a-172b 173b-174a

375-6 379

175b

76b 77a

77a 77b

77b 78a

78a

177a

383 387

78b

78b (-)

178a-178b

390

79a

79a (-)

180a

394

79b

79b (-)

181b

398

80a

80a (-)

183a

401

80b

80b (-)

184a-184b

404-5

81a

81a (-)

185b

408

81b 82a (-)

81b (-)

187a

82a (-)

188a-b

411 414

359

360

FROM

M A H M U D K A § G A R Ì TO EVLÍYA P

ÇELEBI

Y

M

82b (-)

189b

417

83b (-)

192b (-)

423

Q 82b (-) 83b (-) 93a (-)

93a

217 (-)

481

93b (-)

93b (-)

219a

484

95a (-)

95 a

223 a-b

493

#95b*

95b" (-)

223b-224a (-)

494-495

96a

96a (-)

226a

500

97a (-)

97a (-)

228b

506

97b

97b (-)

230a

508

98b (-)

98b (-)

232b

514

99b

99b (-)

235a (-)

519

100a (-)

100a (-)

236b

522

100b

100b (-)

237b

525

#101b* (-)

101b* (-)

239a

528

101b

102a*

240b

531

102a*

102a (-)

241b

533

102b*

102b (-)

242b

535-536

105b*

105b (-)

250b

551

#106b*

106b* (-)

252b

555

106b

107a* (-)

253b

557-558

107a

107b* (-)

255 b

561

#107b*

107b* (-)

255b

560

#108a*

108a* (-)

257a

565

108a (-)

108a* '-)

258b

568

#108b*

109a* (-)

259a

569

109a (-)

110a* (-)

261b

575

#109b*

110a* (-)

261b

575

109b (-)

110b* (-)

263a

578

110a (-)

I l l a * (-)

264b

580

110b

111b* (-)

265b

583

Illa

112a* (-)

267a

586

111b

112b* (-)

268a-b

589

112a (-)

113a" (-)

270a

592

112b (-)

113b" (-)

271a

595

113a (-)

114a* (-)

272b

598

114b-115a (-)

115b-116a (-)

276b

606

115a (-)

116-b (-)

278a

609

115b (-)

117a* (-)

279a

611

116a (-)

117b* (-)

280b

614

S O M E R E F L E C T I O N S ON T H E E D I T I N G O F B O O K 9 Q 116b (-) 117a (-) 117b (-) 118a (-) 118b (-) 119a (-) 119b 120a (-) 120b (-) #12 la 121a 121b (-) 122a (-) 122b (-) 123a-123b (-) 123b (-) 124a (-) 124b (-) 125a (-) 125b (-) 126a (-) 136b (-) 127a (-) 127b (-) 128a 5-) 128b (-) 129a (-) 130a (-) 131a 131b 132a 132b 133a 133b 134a 134b 135a 135b

P 118a* (-) 118b* (-) 119a* (-) 119b" (-) 120a* (-) 120b* (-) 121a* (-) 121b* (-) 122a* (-) 122a-b (-) 122b* (-) 123a* (-) 123b* (-) 124a (-) 124b-125a (-) 125a (-) 125b* (-) 126a* (-) 126b* (-) 127a* (-) 127b* (-) 128a* (-) 128b* (-) 129a* (-) 129b* (-) 130a (-) 131a* (-) 131b* (-) 132b* (-) 133a (-) 134a (-) 134b* (-) 135a* (-) 135b* (-) 136a* (-) 136b* (-) 137a* (-) 137b* (-)

Y 281b 283a 284a-b 285 b 287a 288a-b 289b 291a 292a-b 292b-293a 293 b 294b-295a 296a 297b 299a 300a 301b 302b 304a 305 a 306b 307b 309a 310a 31 lb 312b-313a 315a 316b 319a 320b 320b 323a 324a 325b 326b-327a 328a 329b 330b

M 617 620 622 625 628 630 633 636 639 640 642 645 648 650 653 655 658 661 663a 666 669 672 675 677 680 683 687-8 690 595-6 698 698 704 706-707 709 712 715 717-718 720

362

FROM

M A H M U D KA § G A R i T O E V L l Y A

ÇELEBÎ

Q 136a

P

Y

M

138a* (-)

332a

722-723

136b

138b* (-)

333a

725

137a

139a* (-)

334b

728

137b

139b* (-)

335b-336a

731

138a-b (-)

140a-b (-)

337a

734

138b (-)

141a* (-)

338b

737

138b (-)

141a* (-)

338b

737

139a

141b* (-)

339b

740

[139b:omit]

141b-142a

341a

743

140a

142a-b

342 b

745-746

140b

142b

343b-344a

748-749

141b* (-)

143b* (-)

345b

752

142a" (-)

143b (-)

346 b

754-755

142b* (-)

144a (-)

348a

758

143a (-)

145a (-)

350a (-)

762

144a* (-)

145b (-)

351b

765

145a* (-)

146b (-)

354b

772

145a (-)

147a (-)

357a

777

146b* (-)

148a (-)

358b

780

147* (-)

148b (-)

360a

783

147a* (-)

148b (-)

360a

783

147b-148a (-)

149b (-)

362b

788

148b* (-)

150a (-)

364a (-)

791

148b (-)

150b (-)

365a (-)

793

149a-149b (-)

[151a: omit]

366b

796

149b-150a (-)

151b (-)

368a-b (-)

799

150b* (-)

152a-b (-)

369b

803

150b (-)

152b (-)

370a

806

151a-b (-)

154a* (-)

375 (-)

815

152b (-)

143b* (-)

376b-377a

818

#154b

[156b: omit]

382b

830

155b

[157b*: omit]

386a

836

156a (-)

157b*

387a-b

839

#156b

157b-158a

387b

840

#156a (-)

157b*

387a-b

839

SOME

REFLECTIONS

ON T H E E D I T I N G

OF BOOK

9

363

Studying the Table of Damaged Passages in Book 9 we can draw the following conclusions: Q consistently tries to recreate the exemplar ("S") page by page. This is shown by the fact that with very few exceptions (102a, 102b, 105b, 141b150b) the damaged passages occur at the end of Q folios. From this point of view, Q is the best guide to the original; that is why I have listed Q first in the above table. P tries to recreate the exemplar for the first hundred folios or so, but even in these folios occasionally gets out of step. Beginning at about fol. 106, P gives up trying to remain true to the pagination of the exemplar. This is show by the fact that, with a few random exceptions, the damaged passages no longer occur at the very end of P folios. Also, from this point on, or even earlier, P generally neglects to indicate the damaged passages — or, indeed, even blanks deliberately left by the author — with blanks in the text. (Toward the end of Book 9 P becomes extremely slipshod in reproducing the original text at all, often skipping lines and passages, paraphrasing, etc.) Y, while forgoing any attempt to recreate the exemplar with regard to pagination, consistently indicates damaged passages with blanks in the text, although occasionally supplying plausible suggested readings as well. From this point of view, Y is the best guide to the original.

III. The incorporation of marginal annotations MacKay characterized the relation of P and Q to the original ms. (S) as follows: The copyists of these two sets attempted not merely to represent the contents of S, but even to imitate the pagination of S At the beginnings of each book, and for a hundred or more pages thereafter, the copyists have tried to recreate their exemplar page by page, and even line by line. The one change that is regularly made is to incorporate all the abundant marginal notes of S into a continuous text line When a great deal of marginal annotation has to be incorporated, a page of P or Q may overflow past the boundaries of a page of S, and it may take the copyists a few more pages to get back in step. On the other hand, they sometimes finish the text on a page of S before reaching the end of their own writing line.... I have found no exception to the rule that a blank space in P or Q, which cannot be found in S, will always be at the foot of a page of P or Q, left there in order to regain an absolutely parallel pagination with S.... Since the effort by P and Q to follow the pagination of S can be proved in books 1 through 8, we can also determine many of the characteristics of the lost fifth volume of S from the evidence of P and Q. When the work begins on an edition of books 9 and 10, one of the first requirements will be to determine as far as possible what was the appearance of S. 1 1 MacKay, op. cit., pp. 280-1, 294.

364

FROM

MAHMUD

K A § G A R i TO E V L I Y A

Q ELEB I

The following lists all marginal additions that we may adduce for Book 9: — middle of PQ 9b: on opium as specialty of Afyon-karahisar; preserved in right margin of P and Q; integrated into text in Y (17a24-17b3). — middle of P 17b: on tulip as speciality of Ulucakh; preserved in right margin of P; integrated into text in Q and in Y (33al3-17). — top of PQ 42a: note directing passage about Feri§teoglu to be removed from this point (discussion of his medrese) and put below (discussion of his tomb): miikerrer yazdmi§dir bu mahalle, a§agida ziyareti mahalline yazila\ preserved in left margin of Q; integrated into text in P (but awkwardly stuck in at the middle of a sentence); only in Y has Evliya's instruction been followed, the passage being removed from 81b29 (cf. PQ 42a3) and put at 84b22-85a3 (cf. PQ 43b3), and the original marginal note omitted. — top of PQ 43b, top of Y 85a: note on death and burial of Feri§teoglu. In all three mss. this occurs following the intervening mention of two other sheikhs. Either the note relating to Feri§teoglu was originally in the margin and was misplaced, or else the mention of the two other sheikhs was originally in the margin and was integrated into the text at the wrong point. — top of PQ 56b: additional chronograms on the conquest of Rhodes; preserved in right margin of P and Q, also of Y (113b). — bottom of PQ 59a: additional ziyarets on Rhodes; preserved in left margin of P and Q; integrated into text in Y (121bl-5). — Q 67b39 and Y 145b 15: Ve Musa Beg medresesi. In Q this occurs just before a chronogram relating to a different medrese; in Y it occurs just before the main medrese heading: omitted in P. It must have been a marginal note in the original. — top of Q 68a: additional ziyaret; preserved in left margin of Q; omitted in P; integrated into text in Y (146a22-25). — middle of Q 68b and Y 148a8: Henk beli ve Kara Henk Sultan tekyesi. In Q and Y this occurs in the middle of a sentence; omitted in P. It must have been a marginal note in the original. — middle of PQ 70a: alternate tradition about the founding of Tarsus; preserved in left margin of P and of Y (152a); integrated into text in Q, inserted awkwardly in the middle of a name (between "Me'mun" and "halife"!) rather than somewhat earlier where it belongs. — bottom of Q 75a: a proverb (?) about Azez; apparently preserved in left margin of Q and integrated into text (but only partially?) in Y (167b2627); omitted in P. The situation is muddied by the fact that it coincides with a damaged passage.

SOME

REFLECTIONS

ON T H E E D I T I N G

OF

BOOK

9

365

— middle of Q 76b and Y 171526: Mi'mar Sinan binasidir. In Y this occurs above the line, which is perhaps its original palace; integrated into text in Q; omitted in P. — top of PQ 77b: additional ziyarets in Aleppo; preserved in right margin in P; integrated into text in Q and in Y (174b 16-29). — Q 87a9 and Y 202b 14: Ve Ibn Yamin Misir'da medfundur. In Y this occurs below the line, which is perhaps its original place; integrated into text in Q; omitted in P. — top of PQ 95b: section on the Church of the Holy Sepulcher in Jerusalem; preserved in right margin in P (but minus sevral lines at the beginning); integrated into text in Q and in Y (223bl9-224al3). — top of PQ 98a: two notes on Joseph in Egypt. The first (Hazreti Yusuf babasi Hazret-i Ya'kub Nebtfevtinden sonra yigirmi sene mu'ammer oldu) is preserved in the left margin of Q; integreted into text at slightly different points in P and in Y (230b5-6). The second (Zira vasiyyetleri eyle idi) is integrated into text at the correct point in Y (230b8), a bit later and incorrectly in P and Q. — top of PQ 101b: additional han and associated ziyaret in Sa'sa'a near Damascus; integrated into text in P and Q, also in Y (239al7-20). The gaps and inconsistencies in the text suggest that this section was originally in the margin, extending to the bottom of the folio which was damaged. — middle of Q 106a: additional specialty of Damascus (ve kand-i nebati); integrated into text at incorrect point (before list of specialties, interrupting discussion of the medicinal use of quince) in Y (251 b 17-18) and at different incorrect point (after list of specialties, interrupting discussion of olives) in Q and in P. — top of Q 106b: additional ziyarets in Damascus; integrated into text in P and Q, also in Y (252b4-l 1). The gaps and inconsistencies in the text suggest that this section was originally in the margin and was damaged. — top of Q 107a: additional ziyaret in Damascus plus notes on Umayyad history; preserved in left margin of Q; omitted where expected in P (107al5); integrated into text in Y (253b23-254al0). — middle of Q 107a: additional note on Umayyad history; preserved in right margin of Q; integrated into text in P and in Y (254b5-10). The phrase Ve cenbinde Suleyman ibn 'Abdiilmelik belongs before Al-i Emeviyyun giiztdesidir; only Y has put it in the correct place. — top of Q 107b: additional ziyarets in Damascus; preserved in right margin of Q; integrated into text in P and in Y (255bl 1-29). — top of Q 108a: additional ziyaret in Damascus; preserved in left margin of Q; integrated into text in P and in Y (257a20-26).

366

FROM

MAHMUD

KA§GARÌ

TO E V L Ì Y A

Q ELEB i

— top of Q 108b: additional stages south of Damascus; preserved in right margin of Q; integrated into text in P (109a) and in Y (259a6-16). — bottom of Q 109a: two notes on the canonical status of the Haj. The first (hacc-i §erif farz degildir) in Y occurs above the line, which is perhaps its original place (261al7); integrated into text in Q and in P (109b). The second (¿.irci Cenàb-i Kibriyà hacc-i ¡ertfi Kur'àn-i aztminde sùre-i Al-i ìmràn'da men istetà'a ileyhi sebilà àyetin buyurmu§dur) occurs several lines later in Q and in P (1 lOal), evidently in the correct place; Y puts it just after the first note and adds farz degildir. — top of 109b: additional section on side-trip from Muzayrib to two nearby ziyarets; preserved in right margin of Q; integrated into text in P (110a) and in Y (261b9-28). — middle of Q 120a: two half-verses on the caliph 'Umar; in Y they occur after another half-verse on the same topic (290b 1), evidently the author's intention; in Q and in P (121a) they occur several lines above, just after 'Umar's genealogy. They must have originally been in the margin. — middle of Q 121a: sections on hamams and hans in the fortress of Medina; integrated into text in Q and in P (122a-b), also in Y (292b-293a). The gaps in the text suggest that these sections were originally in the left margin, extending to the bottom of the folio which was damaged. — Q 136b3 and Y 332al3: dahi added below line; integrated into text in P (middle of 138a). — top of Q 138b: date (sene 975) appended to chronogram. In Q this occurs above the line, evidently its original place; integrated into text in P (top of 140b); Y mistakenly puts it one line up, adding tàrihinde (337a4-5). — bottom of Q 138b: There is some evidence in Q of interlinear and marginal notes, but the situation is muddied by the fact that it coincides with a damaged passage. — middle of Q 140a: section heading on the Ka'ba cover; preserved in right margin of P; integrated into text in Q (142a) and in Y (341 b27). — middle of Q 141b: six-line note listing his patrons for whom Evliya prayed at Makam-i Hanefi in Mecca; integrated into text in Q and in Y (345M6-25); summarized by a one-line note in P (143bl3). The tenor of it, and the fact that P considered it suitable for summary, suggest that it was originally in the margin. — middle of Q 142a: Persian note on Hebrew prayer for Zemzem (du'à guy end lisàn-i 'ibrt berà-yi àb-i zemzem); inserted into the middle of the prayer in Q and in Y (347a22); omitted in P (144a).

SOME

REFLECTIONS

ON T H E

EDITING

OF

BOOK

9

367

— middle of Q 154b: additional note on the wilderness of Tih and section on the ruined city of Karan; preserved in right margin of Q; integrated into text in Y (381a29-381b21); omitted in P (156b). — top of Q 156b: section on the stages between Cairo and Mecca and their compass points; integrated into text in Q and P (157b-158a), also in Y (387b6-25). The gaps and inconsistencies in the text suggest that this section was originally in the margin, extending to the bottom of the folio which was damaged. S o m e tentative conclusions may be drawn f r o m the above list. Q is most conservative in preserving marginal material; this probably has to do with its strict attempt to maintain the pagination of the original, since incorporating especially longer notes into the text makes it difficult to squeeze everything in and not " o v e r f l o w past the boundaries of a page of S" as M a c K a y put it. P, more lax in this regard, incorporated more marginal material into the text, though we also find some preserved, at least in the first hundred folios. Y is the most consistent in integrating marginal material into the text, which is to be expected in view of its abandoning the original paginaton; but even here marginal notes are occasionally preserved, perhaps because the copyist was uncertain precisely where they belonged. Based on the above examples, Y shows slightly better judgment than P or Q in integrating marginal material at the correct point in the flow of the text. In this book of the Seyahatname marginal notes are especially to be found in the Damascus section. This is not unexpected, given that Evliya had stopped in D a m a s c u s twice in 1648, and must have written his original description then; but it was only when he visited the city on his way to the H a j in 1672 that he incorporated the description into his book. At the beginning of the Damascus section he states that on his earlier visits he did not stay long enough to see everything, in particular all the ziyarets, and that he had been too busy going back and forth to Istanbul to work on the description (Y 2 3 9 a 2 8 : karar-dade olmayup ziyaretlerinde ve ba'zi

ahvalatina vakif olmadan def'ateyn asitane-i sa'adete ... gidiip gelmeden evsafina ctir'et olunmadi). Richard Kreutel in a seminal article demonstrated the importance of studying the marginal notes in analyzing Evliya's compositional methods and in editing the text. 1 The interpretations offered here for the marginal notes in Book 9 confirm Kreutel's insight.

1

F. R. Kreutel, "Neues zur Evliya-felebi-Forschung," Der Islam 48 (1972), 269-79.

27. SOME REFLECTIONS ON THE EDITING OF BOOK 10 OF THE SEYAHATNAME.

In an earlier article I discussed problems associated with editing the Seyahatname in a general way. 1 More recently I discussed problems relating to Book 9 in particular. 2 With the forthcoming publication of Book 10 in the YKY edition, 3 the dream of a complete edition of the text will finally be realized, albeit not in the ideal form that Goniil A. Tekin and §inasi Tekin proposed. 4 Here I shall put forward my conclusions based on a close study of the mss. of Book 10.

I. Comparison of the three manuscripts The situation with regard to Book 10 is essentially the same as with Book 9. We have three mss. - Y (iUTY 5973), P (Pertev Pa§a 462), Q (Be§ir Aga 452) - that are independent witnesses of the lost original (S) and therefore are on an equal level as bases for an edition. As a general rule, wherever two of the mss. agree as opposed to the third, the agreed reading should be restored to the archetype. As with Book 9, there is evidence throughout Book 10 of damage at the bottom of the original folios. The problem is expecially acute at the beginning of Book 10, but the situation gradually improves, so the picture as a whole is not so dire as with Book 9. The YKY edition makes clear where these passages occur, noting the alternate readings proposed by the three copyists and otherwise indicating the blank spaces that attest to gaps in the original caused by damage. I have not taken the trouble here to tabulate and analyze all occurences of this phenomenon, but will concentrate instead on those issues that relate to Evliya's composition of the text.

1 "§u Rasadi Yikalim mi? Evliya Qelebi ve Filoloji," in: Evliya Qelebi ve Seyahatname (ed. Nuran Tezcan, Kadir Atlansoy; Dogu Akdeniz Üniversitesi 2002), 99-118. o "Some Reflections on the Editing of Book 9 of the Seyahatname," in: Izzet Gutidas Kayaoglu Hatira Kitabi: Makaleler (ed. Oktay Belli, Yücel Dagli, M. Sinan Genim; Istanbul 2005), 122-32 [Turkish translation as introduction to Evliya £elebi Seyahatnämesi IX. Kitap (ed. Yücel Dagli, Seyit Ali Kahraman, Robert Dankoff; Istanbul: Yapi Kredi Yayinlari, 2005)J. This article refines and corrects thoughts expressed in the earlier article with regard to Books 9 and 10. Yapi Kredi Yayinlari. The recent publication of Book 1 (2006, ed. Robert Dankoff, Seyit Ali Kahraman, Yücel Dagli) is intended to supercede the previous version (1995, ed. Orhan §aik Gökyay). 4 "Evliya £elebi bin Dervi§ Mehmed Zilli Seyahatnämesi: Orhan §aik Gökyay'in Hazirladigi Istanbul Cildi ve ideal Metin Negri Üzerine Düjünceler," Kitap-lik 19-20 (Ocak-Nisan 1996), 12-20, p, 18,

370

F R O M M A H M U D K A § G A R 1 TO E V L Í Y A Q ELEB Í II. Table of contents (fihris)

and e n v o i ( i h t i t a m )

The mss. contain two different sorts of fihris or table of contents. One, which I will call the Large Fihris, is found at the beginning of Q (fols. 157a-158b, incomplete) and at the end of Y (fols. 451a-457b, complete). This is a detailed summary intended to show the contents of every opening of the ms. It begins with Evliya's fanciful notion that one might use the text as a kind of falname or fortune-telling book. Another sort, which I will call the Small Fihris, is found in Q (two added folios after the Large Fihris) and in P (fol. 158a-b). It is very summary and incomplete, and the texts of P and Q do not agree. The entire problem is bound up with the very interesting ihtitam or envoi (literally, "seal" - Evliya's last word and farewell to the reader). In it Evliya says that he has retired into obscurity in Cairo after fifty-one years of travel; that he is unable to consult histories or other travellers and writers, but has added Koran commentaries and hadiths learned from his master §eyh Ali §ümürlisi; and that since his travel account has now come to an end, he begs the reader to overlook its faults and pray for its author. This short section is found just before the Large Fihris at Y 450b and Q 156b, but again is missing in P. How can we account for this puzzling situation? We can never know for sure, unless Books 9 and 10 of S are some day discovered. In the meantime I offer the following conjectures: •

In the original ms. (S) the Large Fihris was at the end, but with a note to the effect that it should be put at the beginning. (In the other books, where there is a similar large table of contents it is found at the beginning.) Y left it at the end; Q put it at the beginning; P ignored it altogether.

•

In P there is a Small Fihris instead, but it seems to have been penned not by Evliya but by the copyist of P, perhaps with the Large Fihris before his eyes. Q too has a Small Fihris, side by side with the Large Fihris, and differing from P's. Perhaps when Q and P were being copied in 1742 under the aegis of Be§ir Aga, the copyists were instructed to draw up a Small Fihris to substitute for the Large Fihris; both did this, but Q kept the Large Fihris as well. Since neither Small Fihris was penned by Evliya, we have not included either one in our edition.

SOME

•

REFLECTIONS

ON T H E

EDITING

OF BOOK

10

371

As for the ihtitam, originally it must have been on the same folio as the beginning of the Large Fihris. When the copyist of Q transferred the Large Fihris to the beginning of Book 10 he transferred the ihtitam along with it. The result is that in Q (and in the older printed edition, based on Q [Istanbul 1935, p. 831]) the ihtitam mistakenly appears as the "seal" or ending of Book 9, rather than as the envoi to the Seyahatname as a whole. In our edition we have restored it to its rightful place.

III. Lacunae. There is evidence in the mss. of three sizeable lacunae, all toward the end: 1. Y 442b 17 (mid-line) and P 334b/335a (between folios): a gap of three folios. This portion of the text is only found in Q (338a-340b). Conjecture: when Q and P were being copied in 1742 under the aegis of Be§ir Aga, these three folios were removed from the original ms. (S) after Q had finished but before P had begun. We cannot know the reason, but it may have something to do with the obscene nature of some of the material in these folios. Whatever the reason, these three folios were not copied by P or (in 1751)by Y. 2. In the middle of the Oases section, just before the subsection on the visiting of tombs (Ziyaretgah-i Elvah) the text of Y suddenly breaks off (Y 4 5 0 b l l ; cf. Q 344b27, P 338b26) and comes to an end with the ihtitam. The intervening thirteen folios of the original are only reflected in Q and P. Again, we cannot know the reason for this, but perhaps it is bound up with lacuna number 3. 3. Q 344b/345a and P 338b/339a (between folios): a gap of one or more folios. At the beginning of the new folio we are no longer among the tombs of the Oases but among the mosques of Behnisa. This means that Evliya's travels between the Oases and Behnisa is lost and it is not possible to read this portion of the Seyahatname. Again, we cannot know the reason for this gap.

IV. Scribal peculiarities. Every copyist has his own quirks. Of the three in question, the most eccentric is that of Q. He usually writes -in (possessive plus accusative) with

372

FROM

MAHMUD

K A § G A R t TO E V L Í Y A

CELEBÍ

käf instead of nun. Aside from this he sometimes confuses nun and käf; e.g. BNWN instead of BNWK for bunun (301a), KZK instead of KZN for gezen (301b). He writes -dirl-dir plene: DYR. If he does not understand something he changes it; thus he did not understand gigle and made it geke geke (327a39; cf. Q 324a40, Y 419b28). As for P the copyist prefers the spelling timur to demir; e.g. he writes Öztimur instead of Özdemir (P 334bl3; cf. Q 337bl5, Y 442a2). The most sophisticated of the copyists is that of Y. Even where P and Q have ekmek and bedästen or bezästen he prefers etmek (365a. end) and bezzäzistän. He restores ta'in and kägtr to ta'yin and kargtr. He dislikes some of Evliya's favorite forms - saväb (for seväb), a§kiyä (for e§kiyä) and äbinü§ (for äbnüs) - and substitutes the more "correct" form. He considers tagayyuz vulgar and substitutes tagawut (402b22; cf. Q 321 alO, P 318al0). Occasionally he changes the original text for the sake of clarity; thus at Q 323bl3, P 320bl3 we find su'äl ediip followed by direct quotation with a change of speaker; Y 411bl2 makes this su'äl etdim, eyitdiler. What this means is that Y is the most attentive of the copyists as a reader of the text. Despite this he sometimes errs. E.g. he reads sigin as sigir (twice, 405b.bottom; cf. Q 320a.bottom - 320bi top, P 317a.bottom 317b. top).

V. The incorporation of marginal

annotations

As I tried to show in my article on Book 9, following the suggestions of Richard Kreutel and Pierre MacKay, the reconstruction of marginal annotations is a very fruitful editorial task, since it clearly demonstrates both Evliya's compositional techniques and the way that the copyists dealt with the original ms. The following lists all marginal additions that we may adduce for Book

10: Q 163a25 - 163b22 and Y 10b - l i b : long quotation from the §erefname concerning Alexander the Great in Bitlis; omitted in P (161b23). Q 166a9-10 and P 164a8-10: note concerning the Abbasid dynasty in Egypt; omitted in Y (18b3). Q 166b.bottom: additional note on Aybeg et-Tiirkmani. Preserved in margin of Q; integrated into text in P and Y (P 164b44-47, wrong place; Y 21al4-49). Y 40b5: gurre-i Rebi'ülülä. Preserved below line in red ink in Y; integrated into text in Q and P (Q 174a29, mäh instead of gurre; P 172a30).

SOME REFLECTIONS

ON T H E E D I T I N G

OF B O O K

Q 175a47, P 173a47 and Y 4 3 b l 2 : ve Derne ve Derteng Put in wrong place in P.

ve

10

373

§ehriban.

Q 175bl, P 173bl and Y 4 3 b l 7 : additional heading. Put in wrong place in Q.; in P originally put in wrong place, then crossed out and written in correct place. Q 175b24, P 173b27 and Y 44a26: §dhbdz. Only read correctly in Y (gorse kim Suleyman bir §ahbaz yigit olmu§); wrong in Q (gorse kim bir §ah Suleyman baz yigit olrrius) and P {gorse kim bir §ah Suleyman ydr yigit olmu§). P 173b38: additional heading. Preserved in margin of P (with integrated into text in Q and Y (with oldu, Q 175b36, Y 44b 16).

oldugu);

P 174a20: Selim §ah. Preserved below line in P; integrated into text in Y (45b 16); omitted in Q (176a22). P 174bl3 and Y 46b26: additional heading. Omitted in Q (176bl2). Q 177a42 and Y 49a21: additional heading. Omitted in P (175a47). Q 178a2 and Y 50b27 additional heading. Omitted in P (176a7). Q 178a36 and P 176b35: additional heading. Omitted in Y (53al4). Q 181a38: additional begs. Preserved in margin of Q; integrated into text in P and Y (P 179a39-43, Y 59bl4-21). Q 181a40, P 179a45 and Y 59b23: (—) erdeb glial ve (—) kadar at yemi. In P below line and in wrong place; in Q only as erdeb glial and in wrong place. P 179b3: additional note concerning Evcile. Preserved in margin of P; integrated into text in Q and Y (Q 181a48 - 181b2, Y 60a7-10). Y 6 0 b l 2 , P 179b.mid: additional note concerning kos sahibi ka§ifler. In P put in the wrong place; omitted in Q (181b.mid). Q 182a41: additional note concerning yedi boliik katibleri. Preserved in margin of Q; integrated into text in P and Y (in P placed not where it belongs at 180a41 but five lines up; Y 62b8). P 180b35-38 and Y 63bl9-24: note concerning Muhyiddin-i Arabi and prognostication of the Nile flooding. In P put after repetition of the following words; omitted in Q (182b35). P 182b2 ve Y 6 8 a l 5 : additional courts. Preserved in margin of P; integrated into text in Y; omitted in Q (184b 1). P 185b26 and Y 76a21: note about Baghdad being considered a mahalle of Cairo. In Q placed not where it belongs at 1 8 7 b l 8 but eight lines up. P 1 8 6 b l 2 and Y 7 8 b l l : additional note concerning the Cairo citadel. In Q placed not where it belongs at 188b5 but one line down. Q 188b30 and Y 79a21: additional heading. Omitted in P (186b34).

374

F R O M M A H M U D K A § G A R i T O EV Li Y A ^ E L E B i

P 187a36: addition to list of talismans. Preserved in margin of P; integrated into text in Q and Y (Q 189a28-31, put in wrong place; Y 80bl217). Q 189a42 and Y 81a8: additional chronogram (in Q before deyii, in Y after deyii). Omitted in P (187b2). Q 192a6-19: section concerning Bab-i Zuveyle. In P and Y (correctly?) put above, at the end of the gates (P 189b21-31, Y 87bl9 - 88al4). Q 193b7-12 and Y 93al5-25: section concerning mosque of Hay Atabeg. In P placed not where it belongs at 191b5 but below, after mosque of Sultan Mii'eyyed (191b27-32). Y 46b28: date of tomb of Sultan Gavri. Omitted in Q and P (Q 194b27, P 192b24). P 193bl: section concerning mosque of Arslan Kay. Preserved in margin of P; integrated into text in Q and Y (Q 195a43-46, began to write the following section on mosque of Sultan Kalavan, then crossed it out; Y 99a510).

P 194a44: additional note concerning mosque of §eyh Nizami-i Isfahan! Preserved in margin of P; integrated into text at different points in Q and Y (Q 196a43-44, Y 101bl4-17). P 195a40: additional mosques. Preserved in margin of P; integrated into text in Q and Y (Q 197a39-42, put in wrong place; Y 104a23-28). Q 198a2, P 195b47 and Y 106a3: additional mescid. In Q and P after ve'l-hasil... etc.; in Y (correctly) before ve'l-hasil... etc. Q 199a44: additional tekkes. Preserved in margin of Q; integrated into text in P and Y (P 197a42 - 197bl, Y 109bl2-23). P 197bl3: additional heading. Preserved in margin of P; integrated into text in Q and Y (Q 199b9, Y 110al7). Q 201b9 and P 199bl: additional heading. In Y placed not where it belongs at 115b2 but ten lines below at end of section. Q 201bl6: section on Timurta§ tekke. Preserved in margin of Q; integrated into text in P and Y (P 199b8-10, Y 115bl3-16). Q 202b.mid, P 200b.mid, Y 118b29: Ebutakiyye ham. In Q and P before Mehemmed Aga ham, in Y after Mehemmed Aga ham. Q 204b.mid, P 202b.mid, Y 124all: lisan-i Bo§nakqa. In Q and P after ya 'nt, in Y after dedi. Q 205bl and P 203bl: note about the inventor of tiryak-i faruk. In Y placed not where it belongs at 126b6 but at end of section (127a3). P 204all: note concerning the source of a spring. Preserved in margin of P; integrated into text in Q and Y (Q 206al8, Y 128b9-ll).

SOME

REFLECTIONS

ON T H E E D I T I N G

OF B O O K

10

375

P 206a41: additional well. Preserved in margin of P; integrated into text in Q and Y (Q 206a44-46, Y 129a29 - 129b5). Q 207b46 + 2 0 8 a l and P 206a4: additional lakes. Preserved in margin of Q and P; integrated into text in Y (133b 18-26, put in wrong place). P 206al3: note concerning the mu'arrif of Bahreyn in Bulak. Preserved in margin of P; integrated into text in Q and Y (Q 208a8, first wrote the following words, then crossed them out and put them after the note; Y 134al3). Q 210a31: note concerning the descendants of Noah in Egypt. Preserved in margin of Q; integrated into text in P and Y (P 208al9, Y 139bl6-19). Y 1 4 0 a l l - 140b 19: long addendum on Old Cairo (Fustat, Misr-i Atik). Only in Y in the proper place, between the opening description and the section on magistrates (hakiman); in Q placed not where it belongs at 210a44 but on the previous page in the middle of the Adiliyye section (209b41 - 2 1 0 a l 0 + upper and right margins); in P put below, in the midst of the eighth magistrate (208a42-48 + upper and left margins). P 208bl2: addendum on historical traditions relating to the governors of Egypt. Preserved in margin of P along with the following note: Kenarda sene 358 tarihinde gegiip yazila: integrated into text in Q and Y (Q 210b2133, Y 141b5-25). Q 210b44: bin Sultan Kalavan. in Q first written as part of line below, then crossed out and rewritten in margin; in Y integrated into the text (142al3); in P omitted (208b20). P 209a40 and Y 144a29: additional heading. In Q put one line up (211b30). Q 212a9: Amma Canpuladzade Huseyin Pa§a. Preserved in margin of Q; integrated into text in P and Y (P 209bl7, Y 145al3). Q 214a22, P 2 1 1 b l 0 , Y 151al5: sheikhs of the tarikat (Sufi orders) appended to the procession for the "cutting of the Nile." In Q and P they come before the imams; in Y after the imams. P 214a24: additional heading. Preserved in margin of P; integrated into text in Q and Y (Q 217a25, blank space instead of heading; Y 158a27). P 214b38 and Y 160al4: additional heading. Omitted in Q (217b41). P 215a40: additional information on the crocodile. Preserved in margin of P; integrated into text in Q and Y (Q 218a42, Y 161b22-24). Q 218a.bottom and P 215a.bottom: in the section on the wonders of Egypt, a note concerning the §eyhu'l-bahr ("old man of the sea"). Preserved in margin of Q and P; omitted in Y (162al3).

376

F R O M M A H M U D K A § G A R 1 TO E V L l Y A

QELEBI

Q 218b26: section on the Nilometer and crocodiles. Preserved in margin of Q; integrated into text in P and Y (P 215bl6-22, put in wrong place; Y 162bl9 - 163a2). Q 219b9: ve kadiasker hazretlerine. Preserved in margin of Q; integrated into text in P and Y (P 216b6, Y 165b8). Q 221a38, P 218a37 and Y 170b8: Tokuzuncu. In the correct place in P and Y (before fast); in the wrong place in Q (after fast). Q 222bl0: section on junk dealers (eskiciyan). Preserved in margin of Q; integrated into text in P and Y (P 219bl0, Y 173bl9-21). P 220b34: niima (after ibret). Preserved above line in P; integrated into text in Y (177al0); omitted in Q (223b33). P 221a31 and Y 178al7: mashariyyat-i diger. Omitted in Q (224a26). Q 227a43: additional heading. Preserved in margin of Q; integrated into text in P and Y (P 224a44, Y 186bl4). Q 228a8, P 225a8 and Y 188a21: peri§ani geyerler. In correct place only in Y (after bdlukba§i); in Q one line below, in the middle of a sentence; in P between boliik and ba§i. The phrase must have been below the line in the original. Q 228b34 and P 225b35: additional heading. Preserved in margin of Q and P; integrated into text in Y (190a27). Q 229b3: additional heading. Preserved in margin of Q; integrated into text in P and Y (P 226b3, Y 192a6). Q 229b39, P 226b38 and Y 192b29: mustagrak. In correct place only in P (zer-ender-zere mustagrak koqekleri He). In Q after He; in Y after kdgekleri. Q 230a32: additional heading. Preserved in margin of Q; integrated into text in P and Y (P 227a30, Y 194a5). Q 230bl0: additional heading. Preserved in margin of Q; integrated into text in P and Y (P 227b9, Y 194bl6). Q 232a26 and P 229a27: added phrases concerning the ceremonial howdah (mahmel-i §erif). In Y placed above, interrupting the flow of the text. (199al2). Q 233a3: additional heading. Preserved in margin of Q; integrated into text in P and Y (P 230a4, Y 201a23). Q 233bl8: additional heading. Preserved in margin of Q; integrated into text in P and Y (P 230bl8, Y 203a3-5). Q 235b20 + 26-27 and P 232bl9 + 25-26: additional heading + added remark about providing water to the pilgrims in Kiipler. In Y these two separate items are not where they belong at 208b 1 but are mixed together and put in the wrong place (208al8-21).

SOME

REFLECTIONS

ON T H E

EDITING

OF BOOK

10

377

Q 236b6 and Y 210a29 - 210bl: ve elli gobek mii§k-i ablr. In P put in the wrong place (233b3). Q 237al2 and Y 212a3-5: additional guilds in the procession of ulema. Put at different points in Q and Y; omitted in P (234al4). P 234al6: section on the procession of the molla of Cairo. Preserved in margin of P; integrated into text in Q and Y (Q 237al6-23, Y 212a9-23). Q 238al2 and P 235al2: note on restrictions during day of gunpowder transport. Preserved in margin of Q and P; integrated into text in Y (214b2529). P 235al3: additional information on gunpowder transport. Preserved in margin of P; integrated into text in Q and Y (Q 238al3-17, Y 215a2-7). Q 238b4 and P 235b4: note on the fate of Ruznameci Abdurrahman Efendi. Preserved in margin of Q and P; integrated into text in Y (216a7-9). P 236a20: additional note on which processions (alay) are included in the official number. Preserved in margin of P; integrated into text in Q and Y (Q 239a23-26; Y 217b20-24, put a little above, after the number of treasures [hazine]). Q 239b34: note to put here the account on the opposite page of the mevlud of §eyh Bekri. Only preserved in the margin of Q. Only Y follows this directive (219a25 - 219b27). The edition follows Evliya's orijinal plan according to which the fourth of the twelve mevluds is that of §eyh Bekrizade (Q 240a25-45, P 237a25-45) and then comes the mevlud of §eyh Ibrahim Giil§eni. In Y, on the other hand, the fourth mevlud is that of Ibrahim Giil§eni (220b7). Y 224a24: ibtida. Correct only in Y (dest-i gerifin ibtida bus eden kimesne); in the wrong place in Q (241b29 dest-i §ertfin bus eden ibtida kimesne); apparently not readable by P (238b28 dest-i §erifin (—) bus eden kimesne). Y 226al8-20: chronograms. In P one line below, in the middle of a sentence (239a44-46); omitted in Q (242b3). Q 242bl3 and Y 226b5-7: additional chronogram. In the wrong place in Q and Y (in the middle of the sentence Sultan (—) hayratidir); omitted in P (239b5). Q 242b24: ya'ni cebel-i Becamim. Preserved below line in Q; integrated into text in P and Y (P 239b 16, one line below correct place, between silahina and dayanan; Y 226b23). Q 245al5: quotes from ibn Celal and §ihabi on the pyramids. Preserved in Q in margin and in empty lines at bottom of folio; integrated into text in P (242al3-17, up to yatirlar\ the reference to §ihabi is in the margin); integrated into text in Y (232a25 - 232b4).

378

FROM

MAHMUD

K A § G A RI T O E V L i Y A

ÇELEBÎ

P 242b45: quote from ibn Celal on the Sphinx. Preserved in margin of P; integrated into text in Q and Y (Q 245b43-48, Y 234bl2-20). Q 247al9 and P 244a21: additional trees. Preserved in margin of Q and P; integrated into text in Y (227bl4-20). Q 248b4 and P 245b4: at beginning of chapter on novelties (bid.'at) and virtues Qiiisniyyât): note on Jewish funerals. Preserved in margin of Q and P; integrated into text in Y but before the heading (241all-17). P 246a35 and Y 243a29: sûr. In the correct place in Y (Âyîn-i sûr-i fellâhîn-i Misir); preserved below line in P but in the wrong place (after Misir); omitted in Q (249a37). Q 249b46: addendum on circumcision of Hazari girls. Preserved in margin of Q; integrated into text in P and Y (P 246b39-42, five lines above the correct place, in the middle of a sentence; Y 245a4-8). P 247a35: note on Jewish incense-sellers during mevlud of imam Hiiseyin. Preserved in margin of P; integrated into text in Q and Y (Q 250a3638, Y 246a9-14). Q 250b9 and P 247b9: additional note concerning Melamis ("When they enter the bath they wear an apron, when they come out they go around naked"). In Y put one line down (246bl8). Q 251b47: mekteb. Preserved above line in Q, after sibyân; integrated into text in P and Y (P 248b46, before sibyân; Y 250al3, as mekteb sibyâni). Q 253b32 and Y 255a7-8: additional note on the sixth clime: Evliya travelled the entire length of the Danube five times. Omitted in P (250b32). Q 257a44 + 46 and P 254a42 + 46-47: two additional notes on imam §afi'i. Both preserved in margin of Q; the first preserved in margin of P, the second integrated into text; both integrated into text in Y but the first one line below the correct place (263b9-ll + 13-16). Y 270al2: Muhammed. In correct place in Y (e§-§eyh Muhammed Fereskûrî); one line below in Q and P, in the middle of a sentence (Q 260al8, P 257a20: lokmalari Muhammed neden hâsil olur, kimse vâkif-i esrâr olmamiçdir). It must have been below the line in the original. Q 265b3: additional ziyaret in Mît Gamr: ve e§-§eyh Ahmed ezZenkalî. In Y put one line down (281bl5); in P the text is defective at this point (262b3). Q 266b47 and Y 285al6: additional heading. Omitted in P (263b48). Q 268al4: additional mosque in Tanta. Preserved in margin of Q; integrated into text in P and Y (P 265al5, Y 287b23). Q 275b36 and P 272b37: additional Arab tribes. Preserved in margin of Q and P; integrated into text in Y (305al5-18).

SOME

REFLECTIONS

ON T H E E D I T I N G

OF B O O K

10

379

Q 2 7 7 a l 3 , P 2 7 4 a l 2 and Y 307b25: akge. In Y (correctly?) after helàl, in P before helàl; in Q one line down. It must have been below the line in the original. Q 278b7-16 and Y 311al4-29: two added sections to Hu§-i Isa. Omitted in P (275b9). P 276b39: additional report about more than one Alexander in Alexandria. Preserved in margin of P; integrated into text in Q and Y (Q 279b32-34, Y 314a27 - 314b2). P 2 7 7 a l : additional report about Dacian (Takyanus) in Alexandria. Preserved in margin of P (written in a different hand); integrated into text in Q and Y (Q 279b45 - 280a5, Y 314b 19 - 315a2). Q 280b6: additional heading. Preserved in margin of Q; integrated into text in P and Y (P 277b6, Y 316al6). P 279a40: section on Alexandria water channels. Preserved in margin of P (written in a different hand); integrated into text in Q and Y (Q 282a34-41, Y 320bll-22). Q 282b6-8 and Y 321al4-16: addendum on Bi'r-i Mubsit. Omitted in P (279b6). Q 282b42-45 and Y 322al0-16: addendum on the construction of the Attarin mosque. Omitted in P (279b43). Q 284b36-39 and Y 326M0-14: added notes on vegetables and other features of Alexandria. Omitted in P (281b40). Q 285a29-32 and Y 3 2 7 b l 0 - 1 4 : additional ziyarets in Alexandria. Omitted in P (282a32). Q 285b8-17 and Y 328a20 - 328b7: addendum on tomb of §eyh §atibi. Omitted in P (282b8). P 2 8 9 b l l : arz. Preserved below line in P; integrated into text in Y (342b28); omitted in Q ( 2 9 2 b l l ) . Y 3 4 6 a l : dahi. In correct place in Y (Jki kapusu dahi, biri...); one line down in Q (294a5); omitted in P (291a4). It must have been below the line in the original. P 294a38 and Y 353b4: gar§udadir. Put one line below in Q (297a39). Q 297b39 and P 294b42: some notes concerning Menzile. Preserved in margin of Q and P; integrated into text in Y (354bl7-21). Q 297b48: completion of the folio (original text). Preserved in margin of Q; integrated into text in Y (355a5-8); omitted in P (294b. bottom). The discrepencies between Q and Y suggest that this section was originally in the margin where the folio suffered damage. Q 2 9 9 b l 2 : additional ziyarets in Mahalle. Preserved in margin of Q; integrated into text in P and Y (P 296b7-13, Y 358b5-15).

380

FROM

MAHMUD

K A § G A R i T O EV L I Y A

CELEBl

Q 301b31: agac. Preserved above line in Q; integrated into text in Y (363a26); omitted in P (298b31). Y 369b5: addition to heading: cebel-i Tayltmun dahi derler. In Q first put in wrong place, then corrected (304bl9); omitted in P (301bl5). Q 310a34 and P 307a33: well and pavilion of Kusayr and Evliya's chronogram. Preserved in margin of Q and P; integrated into text in Y (381bl7-19). This marginal note is very confused in P and Y, relatively clear in Q. Q 310a39 and P 307a38: additional information on the history of Kus. Preserved in margin of Q and P; integrated into text in Y (381b26 - 382a2). Q 310b21: additional ziyarets in Kus. Preserved in margin of Q; integrated into text in Y (382bl3-18); omitted in P (307b21). Q 316a9: Evsaf-i tetimme-i Say etc., a rather long section. Preserved in Q on the opposite page (315b); integrated into text in Y (394b6 - 395a5); omitted in P (313a8). Q 316a47 and P 313a47: additional heading. Put three words up in Y (395b29). Q 319b31 and P 316b32: Cebri olduklari andan ma'lumdur. Preserved in margin of Q and P; integrated into text in Y (403a27). Y 405al7: Bozahaneleri ve kahve vardir. Put one line down in Q and P (Q 320bl5, P 317bl6). Q 324al7: melik. Preserved above line in Q; integrated into text in Y (412b23); omitted in P (321al7). P 322b36: arak gibi. Preserved below line in P; integrated into text in Q and Y (Q 325b35, Y 416bl). P 329a42: bu. Preserved below line in P; integrated into text in Q and Y (Q 332a42, Y 430b28, put after the following word). Y 433b8: additional heading. In correct place only in v (after misk kedisi mukarrerdir); in wrong place in Q and P (Q 3 3 3 b l 4 , before mukarrerdir; P 330bl2 before misk kedisi mukarrerdir). P 332b3: Misir ve. Preserved below line in P; integrated into text in Q (335b3); omitted in Y (437b9). Q 349a20: section: Tetimme-i evsaf-i Feyyum. Preserved in Q in the margin on the opposite page (348b); in P not where it belongs at 343al6 but put randomly on the opposite page (342b3-6). P 344a8: bu. Preserved below line in P; omitted in Q. Q 351al9: additional events in the year '84. Preserved in margin of Q; integrated into text in P (345al6-18).

SOME REFLECTIONS

ON T H E

EDITING

OF BOOK

10

381

Q 351bl7: addendum on the arrival of an imperial rescript (hatt-i §erif) regarding raising troops for the Oman campaign. Preserved in margin of Q; integrated into text in P (345bl4-16). Q 354al2: mollas of Cairo in the years 971-1020. Preserved in margin of Q; integrated into text in P (348all-18). The list continues at the bottom of the folio on the reverse (Q 354b44-46, P 348b45-48). Q 356al3-17, P 350al0-13: a note to add description of battle between two Arab tribes, Beni Harb and Beni Sakar.

28. TWO ARMENO TURKISH TEXTS: LAMENT FOR A DEAD DAUGHTER and GAME OF CHANCE

Recent publications by Sanjian and Tietze 1 and by Hetzer 2 have drawn attention once again to the riches of Armeno-Turkish literature (i.e., Turkish literature in Armenian script). Here I present a transcription and translation of two short Armeno-Turkish texts, one dating to the late seventeenth century, the other, apparently, to the eighteenth. I wish to thank Andreas Tietze for examining an early draft of this article and suggesting several improvements.

I. Eremya £elebi's Lament for His Daughter Sogome The noted author Eremya (,'elebi (1637-1695) composed this lament for his daughter on the occasion of her death in the year 1690. 3 Sanjian, in the work referred to in note 1, has given an account of Eremya's life and his literary activity. For the present purpose it is enough to say that Eremya gave expression to his grief at his daughter's death in two laments, one in Armenian, the other in Turkish. Both of them are couched in terms of a monologue addressed by the dead woman to her parents, urging them not to grieve overmuch, lamenting her lost youth, and expressing longing for her orphaned son Joseph (Yovsep'/Hovsep ~ Usep). 4 The Armenian one consists of seven long stanzas, the first letters of which spell out Eremya's name (in Armenian script) and is structured in a formal and traditional Armenian rhyming and metrical scheme. The Turkish one, by contrast, follows the scheme of Turkish folk poetry. It is in two parts (not distinguished in the original). Part A consists of seven stanzas of five verses each (except the first stanza has four and the fourth, six), the last verse in each stanza being a refrain. With a few

' Avcdis K. Sanjian and Andreas Tietze, eds., Eremya Chelebi Komurjian's Armeno-Turkish Poem "The Jewish Bride," Budapest: Akademiai Kiado, 1981. 2 Armin Hetzer, Dackeren-Texte, Wiesbaden: Harrasowitz, 1987. ^The text is found in Mesrop Nshanian, Oragrutiun Eremia Chelebi Keomurjiani (Jerusalem, 1939), pp. 632-634. 4 According to Nshanian's estimate (see p. cxviii) Sogome was age 19-21 when she died, and Joseph at the time was age 4-6.

384

F R O M M A H M U D K A § G AR i T O E V L I Y A

CELEBI

exceptions, all the verses break up into the syllabic count: 4-4-3. (The overlong count in the penultimate verses of stanzas six and seven adds to the poignancy). Part B consists of eleven quatrains; all the verses break up as: 44. Tietze, in the work referred to in note 1, has analyzed the phonology and the other linguistic characteristics of Eremya's Turkish. Since the text presented here does not offer any new or unusual peculiarities in this regard, it is enough simply to refer to Tietze's study. The transcription, however, follows the lead of Hetzer (see p. 383, n. 2), who has set a new standard for transcribing Armeno-Turkish texts. - ANene beni siz aglaman sizlaman, Ecel geldi emrullahdi siz anlan, Sogimeniz oldi deyi pek yanman, Hemen deyin ki Sogome qurtuldi. Mama, don't cry and moan for me. The hour came; it was God's command. Understand that. Don't grieve much saying your Sogome has died; Only say that Sogome has been saved. Ben dunyaden behaberdim bilirsiz, Bir ter qon§e ondordonde ben heniz, Goz iimubda tezce ere verdiniz, Firaqlenmen ki elimizden qu§ u§di Hemen deyin ki Sogome qurtuldi. I had no experience of the world, you know. I was yet a fresh bud, in my fourteenth year. In a twinkling you married me off. Don't grieve saying "The bird has flown from our hand;" Only say that Sogome has been saved. Bu onulmaz derde gafil sata§dim, Olum haddini bildim his etdim, Sizi anib goz ya§imi qan iderdim. Yanman ki gencecig qizimiz oldi, Hemen deyin ki Sogome qurtuldi. All unawares was I inflicted with this incurable pain. I experienced, I suffered, the sharp edge of death. Thinking of you, I was weeping bloody tears. Don't grieve saying "Our tender daughter has died;" Only say that Sogome has been saved.

TWO

A R M EN O - T U R K I S H

TEXTS

385

Usepcigim, ah, hatrime geldikce, Zehirlenub ter daneler tokdukce, Parxun olub iiregcigim dukende, Axir nefes tutulubda sundukde, Feleklerle ceng etmen ki hayf oldi, Hemen deyin ki Sogome qurtuldi. Ah, whenever I remembered my little Joseph, Embittered, I poured forth beads of sweat. My small heart grieved and wore itself out. 1 In the end my breathing stopped and the tension broke. Don't quarrel with fate saying "What a pity!" Only say that Sogome has been saved. Usepimin bulbul avazi qande, Hig i§itmez qulagim dur oldukde, Hasret qaldigimdi derdim ol hinde. Yanman ki yavrumiz §u cins kill oldi, Hemen deyin ki Sogome qurtuldi. Where is my Joseph's nightingale voice ? Being far away, I cannot hear it. Then my pain was being separated (from him). Don't grieve saying "Our child has turned to ashes like that;" Only say that Sogome has been saved. Yasitlarim qizlar daxi oynarler, Ana ata evlennde qvanirler, Nagyah ecel camin tezye sundular. Tutu§man ki nazli besli yere yeksan mi oldi, Hemen deyin ki Sogome qurtuldi. The girls, my equals in age, are yet playing, Happy in their parents' home. Suddenly, quickly, 2 Death's cup was offered. Don't grieve saying "Has the one whom we spoiled and fattened been levelled with the earth?" Only say that Sogome has been saved. Riza oldum ben eflakin hukmine, Boynum qldan ince Xuda emnne. Qkdik maiden hem evladden siz deyi, Game garq olman ki ba§imize bu musubet ne idi, Hemen deyin ki Sogome qurtuldi. 1

Dukende for duken[dik]de? See Tarama Sozliigti, p. 3810: tiziye.

386

F R O M M A H M U D K A § G AR I T O E V L I Y A

gELEBi

I consent to the judgement of fate; I acquiesce to the command of God. Don't drown in grief saying "We have lost both wealth and child," And saying "Why did this disaster befall us?" Only say that Sogome has been saved. -B Ben murade iremedim, Hig bir murvet goremedim, Ecel camin zehnn i§dim, Gene olume yaqalandim. I could not attain my desire, Nor live to see my children come of age, But drank Death's poison cup. Young, I was caught in death's snare. Gozlerimden qanli ya§ler, Ba§ime qaza u qader, Yazilmi§di bu muqader, Ecel §erbetle mest oldum. Bloody tears from my eyes, Blows of fate on my head: This was written, fated. I was intoxicated with Death's drink. Bu garxi felegin igi, Tesiri naqsidi naq§i, Qaflet telefi can qasdi, Genchgime hi§ doymadim. Within this sphere of heaven, Its design (sign? portent?) pointed to destruction, 1 Its aim was sudden death. I did not get my fill of youth. Gog dar olubdir ba§ime, §ekvam belki seyran ele, Garib halindan 9eken bile, Xakile ki yeksan oldum.

1

Reading naqs as < naqz; alternately "imperfection" (< nags).

TWO

ARMENO

TURKISH

TEXTS

387

The sky became too narrow for my head: Misery to me, but to others a holiday. One who has suffered homelessness knows. 1 I was levelled with the earth. Gencligime yazig oldi, Oksuzime hay if oldi, Helakim bu iizden, idi, §ukur ki selamet buldum. A pity for my youth! l o o bad for my orphan! Such was my perishing. Thank God I have found peace. Nene ben gelin oldugim, Xeyali mi idi yoxse du§um, Sevinecegim hem omrum, §u qederce sure geldim. Mama, that I was married, Was it a fantasy or a dream? That I enjoyed my happiness and my life For that brief span of time, Bir zeman ben salinirdim, Izunaz ile hayatim, Begoglunin viranelerin, §imdi bekci ola geldim. One time I sauntered, Proud of my life. Now I am haunting. The ruins of Beyoglu. Fani dyunya mihnetinden, Ol ma§aqati ejgalden, Ol vefasiz me§gulatden, Borcum eda ide geldim. The trouble I borrowed of the fleeting world, The hardships and travails, 2 The inconstant preoccupations: I came to pay my debt. 3

' I.e., only one who himself has suffered the state of being a homeless wanderer (garih) can know what it is like.

o 3

Reading ejgal as < e§gal. I.e., I have suffered them all.

388

FROM

MAHMUD

KA§ GARI

TO E V L i Y A

gELEBI

Bir huma ki £ka elden, A§ianin bulmak neden, Bu murgi can qafes tenden, Pervaz idiib azad oldum. Just as a bird-of-paradise, set free, Tries to find its nest, So the bird of my soul has taken wing And has flown free from my body cage. Dyunyaye bel baglayanlar Zefki ile tek doysunlar, Ben imtila oldum dostlar Baqii seyrane geldim. Let those who trust the perishing world But get their fill of pleasure. I have got my fill, my friends: Now I have come to enjoy the eternal world. Ah, bir dane Usepcigim, Ben zevalli yetimcigim, Miirvetine hasret qaldim, Ahizar zar ide geldim. Oh, my only little Joseph, My poor little orphan: How I long to see you come of age. I keep sighing and moaning.

II. A Game of Chance An eighteenth-century miscellany in the Matenadaran library in Yerevan, #1675, contains on fols. 174b-177b the text presented here, described in the catalogue as a "game of chance" (Arm. vichakaxag).' It is placed between two historical works and clearly bears no relation to the serious texts surrounding it. At the top is a table of five columns each listing sixteen items of daily life. The columns are headed by Armenian letters standing for numerals, presumably a point value for the items listed. Most items are on more than one list — one, iron, is on all five — and there is no discernible pattern or order to the listings. The number of different items in the lists comes to thirty-one.

' I am grateful to the officials of the Matenadaran for providing me with a microfilm.

TWO ARMENO

TURKISH

389

TEXTS

F o l l o w i n g the table is a series of t h i r t y - o n e q u a t r a i n s . T h e y are n u m b e r e d using A r a b i c - s c r i p t n u m e r a l s — perhaps an indication that the entire text w a s copied f r o m an Arabic-script prototype. T h e quatrains are in the m a n n e r of riddles (note #27: bilmece).

T h e first three verses s e e m to

provide clues — but o f t e n the only clue a p p e a r s to be the r h y m e — while the f o u r t h n a m e s one of the items in the table. (An exception is #11 w h e r e the item b e i n g n a m e d is in the third v e r s e ) . E x c e p t f o r s o m e a p p a r e n t groupings by category (e.g., ##4-7: cotton, thread, fabric, cloth), there is n o discernible pattern to these quatrains. It m a y be s i g n i f i c a n t that the first t w o items in the quatrains (earth, c a n d l e ) are a m o n g those w h i c h a p p e a r only o n c e in the table (others that appear only o n c e are # 4 cotton, #8 leather, and # 1 6 bone); and that the final item is iron which, as w e noted, above, is in all f i v e of the columns. A l t h o u g h it is possible to s p e c u l a t e how the g a m e m a y h a v e b e e n played, this d o e s not s e e m a f r u i t f u l c o u r s e lacking a d e f i n i t i v e a n s w e r . I publish it h e r e in the h o p e that others, either with greater ingenuity than myself or having access to m o r e information about such g a m e s , will be able to break the code. T h e e m p h a s i s given here to the riddle items is not f o u n d in the origina. A n English translation a c c o m p a n i e s each unit of the text. 1 qarenfil clove demir iron agac wood yemi§ fruit etmek bread qur§un bullet bakir copper altun gold otluk pasture gumii§ silver bez p a r a s i piece of cloth

2 bez parasi piece of cloth da§ stone kehat paper sa? hair qur§un bullet yemi§ fruit §i§e glass altun gold giimu§ silver bokluk dunghill saman straw

4 ipek silk agac wood kehat paper yemi§ fruit bakir copper boncuk bead §i§e glass altun gold xatem ring otluk pasture demir iron

8 gon leather da§ stoen bugday wheat ipek silk agac wood yumurta egg kehat paper demir iron yapagi wool yemi§ fruit qarenfil clove

16 demir iron kemik bone etmek bread sa? hair qur§un bullet boncuk bead bakir copper §i§e glass yapagi wool yumurta egg bokluk dunghill

390

FROM

MAHMUD

KA § G A R Ì TO

EVLÌYA

CELEBÌ

1 saman straw iplik thread

2 xatem ring bugday wheat

4 iplik thread quma§ parasi piece of fabric

8 bokluk dunghill giimu§ silver

16 altun gold otluk pasture

bugday wheat toprak earth yumurta egg

demir iron mum candle quma§ parasi piece of fabric

bez parasi piece of cloth pembe cotton sabin soap

sabin soap otluk pasture xatem ring

giimu§ silver sabin soap xatem ring

1. Agdim didem gordiim eflak Dedim elhamdiilillah hak Avucinda heyallandim Aq gosder simsiyah toprak I opened my eyes and saw the heavenly sphere I said "Praise be to God the True." I imagined what is in your hand. Open and show it: pitch-black earth. 2. Gongiil ozledigin bildim Eman gozledigin bildim Tezce (jikar gosder §ahim M u m d i r gizledigin bildim I know what the heart desires Mercy, I know what you are seeking. Take it out quickly and show it, my Shah: I know that you are hiding a candle. 3. Eman a 1 efendim eman Aklim zayi qadtim keman Avucinde hayvaniet Yeyecegi sari saman Mercy oh my efendi, mercy. My wits are gone, my stature is bent. In your hands is what animals Will eat: yellow straw. 4. A 2 dosdum sin nihanim Qadti servi levendanim (,'ikar ki seyran idelim Sende beyaz penbe canim

' Can also be ready ày. Can also be read Ay.

2

TWO

ARMENO-TURKISH

TEXTS

391

Oh my friend whose secret is hidden, My handsome one with the stature of a cypress, Take it out so we can see: You have white cotton, my dear. 5. Divane §ekli oldum ben Gezer aklim Hind u Yemen Sokiiklere lazim olur iynesi yok iplik hemen I have become crazed; My mind is touring India and Yemen. It is needed for rips and tears; It has no needle, it is just a thread. 6. Ne toprak ne agac ne da§ Ne yemi§ ne yeyecek as Terzilerden artmi§ bildim Sende bir par§acik quma§ Not earth, not wood, not stone, Not fruit, not anything to eat. I have guessed it: something left over from the tailors; You have a little piece of fabric. 7. Iy gongiil aklini gezdir Evetleme dahi tezdir Ondan bundan qirpindicik Sende bir panjacik bezdir Oh my heart, send your mind on tour. Don't hurry, 1 it is still early. A little clipping from here and there: You have a little piece of cloth. 8. A^dim bu niyet imlasin Okurum angla enfasin Vicude siklet iyleme £ikar §u goniin parasin I have deciphered this script of (good) intention. I can read - understand me! - its spells. Do not oppress a body. 2 Take out that piece of leather. 9. Banga inan ey ehli dil Muhal akildir ne qabil Qoqusi burnuma deydi Ya§ deyil quru qarenfil Believe me, oh people of the heart: It is absurd, impossible. Its scent touched my nose: Not a fresh but a dry clove. 'See TS 2134: ivetlemek (evetlemek), DS 1812-13: evetlemek (s.v. evmek). I.e., do not bore me.

392

FROM

MAHMUD

K A § G A RI T O

EVLiYA

gELHBI

10. Siimeye iyleme seva§ Am bulur akilli ba§ Ya ni9in gizlersin dosdum £ikar e§gyare goster da§ Do not battle in vain.' A wise head will discover it. So why are you hiding it, my friend? Take it out into the open and show it: a stone. 11. Bu soze inan dediler £un bilmezsen dan dediler Gizli deyil sende bugday Etmek kiirt5e nan dediler "Believe these words," they said "If you don't know, consult," 2 they said It is not hidden: you have wheat. "Bread in Kurdish is nan," they said. 12. Uzakdir ta ba§dan gelir Kimi quru ya§dan gelir Bahasi bek agircadir ipek qizilba§dan gelir It is far and comes all the way from the head. Some comes from dry (and some) from wet. Its value is very great. Silk comes from the Kizilba§ (i.e., from Iran). 13. Avucunin aracigi A§ilmi§dir oracigi Ne bilmek gerek ani Gordiim agac paracigi The crack of your closed hand, That little part has opened. What is there to guess? I see a little piece of wood. 14. Fikir bir xatdir yazihr Deme isnatdir yazihr Pembeden hasile gelir Sende kehatdir yazihr What I have in mind is a line that is written. Don't say it is an ascription that is written. It is produced from cotton. You have paper on which is written. 15. A§ikler boyle demi§dir Saidama soyle demi§dir U§aklara ilazimdir Bir §oyle boyle yemi§dir !see TS 3619, DS 3711: siimeye. See TS 1004-05: danmak.

2

TWO ARMENO-TURKISH

TEXTS

Lovers have said so: "Don't hide it, tell it," they said. It is needed by the little ones. Thus and so: a piece of fruit. 16. Xakden xalkten yolda§idir Bu soz sòzlerin ba§idir Bize hacet deyil ama Gemik kelplerin asidir From the soil, from people, it is his companion. This word is the chief of words. We have no need of it, but It is food for dogs: a bone. 17. Dinim imanim paresi Dòkerim qanim paresi Din diregidir demi§Ier Etmekdir canim paresi A piece of my religion and my faith; I'll shed a bit of my blood. "It is the pillar of religion," they say: It is bread, oh you piece of my soul. 18. Aceb qa£indan gizledin Dòrt be§ ii§iinden gizledin Ben ani evelden bildim Yoldun sa^indan gizledin I wonder how many of them you have hidden. Four or five or three of them you have hidden. I guessed it from the first: You pulled out and have hidden some of your hair. 19. Boyle nazli duru§un ne Sarmalayip burujun ne Avcilara lazim olur Avucinda quru§un ne Why are you acting so coyly? Why have you wrapped it up? It is needed by hunters: Why is the bullet in your hand? 20. Bir yam u?uk dediler Qfitlarda 50k dediler Firenq lehin yadigyari Var gosder boncuk dediler "One side of it is faded," they said. "Jews have many," they said. A memento of the Europeans, the Poles. Show it: "A bead," they said.

393

394

FROM

MAHMUD

K A § G A R i TO

EVLlYA

QELEBI

21. Asiller okur gozledim Biilbuller §akir gozledim Qazancilara sata gor Bir par$a bakir gozledim I observed noblemen reading. I observed nightingales singing. Sell it to the kettle-makers: I observed a piece of copper. 22. Gongiil qurbetlige dii§e Ne dersin boyle bilige Isdanbolun pe§ge§idir Ya billur ola ya §i§e The heart may suffer exile. What do you say to such an acquaintance? It is gift of Istanbul: It must be either crystal or glass. 23. E§g elinden bagrim xundir Aklim zayi serni giindir Ne giimu§ ne bakir ne da§ Elinde sari altundir Because of love my heart is torn.1 My wits are gone, turned upside-down.2 Not silver, not copper, not stone: In your hand is yellow gold. 24. Budur soztin sahi sagi Cemi sinemler yatagi Qoyunlardan hasil olur Sende bir par§a yapagi This is the correct word, The bed of every idol.3 It is produced from sheep: You have a piece of wool. 25. Gordum avucun ortasi Siler derunimin pasi Qahve altiya lazimdir Qara tavuk yumurtasi I see the middle of your closed hand. It wipes away the rust of my heart. It is needed for breakfast: A blackbird's egg.

^ Lit.: From the hand of love my liver is bloody. Serni giin for ser-nigun. •3 Sinemler for sanemler (?).

2

TWO

ARMENO

TURKISH

TEXTS

26. Aklim kesmez sahi (¿okluk Goze zerarimiz tokluk Ondan bundan dev§irilmi§ Otluk desem deyil bokluk I really can't guess it most of the time. Not being greedy is our loss.1 It has been collected from here and there. If I say "pasture" that is wrong: dunghill. 27. Bu bilmece qayet giicdir Ugi birdir biri iigdir Quyumcilar sermayesi Gosder bir par?a gumii§dir This riddle is very difficult: Three of them is one and one of them is three. The jewellers' stock in trade; Show it: a piece of silver. 28. A1 benden §afi cevabin Agardir nazik esbabin Aktja pak^a ider her gis Bazarda satilir sabin Take from me your satisfactory answer: It whitens your fine clothes; It always makes things pure and clean; It is sold in the market: soap. 29. (,'ikarip meydane birak Nare yak virane birak Bize lazim degil hacet Otlukdir hayvane birak Take it out and put it in the open. Burn it in fire and leave it in ruins. We have no need of it; Pasture: leave it to the animals. 30. Bu icadi iden femdir Yaq§i §eye deme kemdir Barmagindan manzul olmu§ Avucmdaki xatemdir That which made up this invention is the mouth. Don't call a good thing bad. It has been removed 2 from your finger. What is in your hand is a ring. 31. Sendekin bilen fikirdir Bulan bir penfeyi §irdir Keser bigak mayasidir. Sende bir par§a demirdir 1

9Taking

goze ... tokluk as equivalent to tok gdzluliik. Manzul for ma 'zul.

395

396

F R O M M A H M U D K A § G A R i TO E V L I Y A Q ELEB I

That which guesses what you have is the mind. That which discovers it is a lion's paw. It is the stock of a cutting knife: You have a piece of iron. The first phrase in # 16 is a corruption of Haqdan xalqdan "From God, from creatures." The abbreviations DS and TS in the footnotes refer to Derleme Sozlugii and Tarawa Sozliigii.

T W O ARMENO- T U R KI S H T E X T S FACSIMILE

mÍÉ m M m ^ m •

gl " SS

w

I»

MUi/ht

il I b i i ^ f c i f f i iiì 11111 il j t e

mi

"riC. I^F"? • i

1

afyi

2 yîi

398

FROM M A H M U D K A § G A R Í TO E V L Í Y A

ÇELEBI

r Ml

Ott-,, W ~

tryw LÌ

rvFX * -4tiii I4-i

'(""tf If®*11*! yn»t»«i mili

4 M "r !

-

--1 :

3

^ t; -i

^