Iran. Journal of the British Institute of Persian Studies. Vol. 32 issue 32

IRAN Journal of the British Institute of Persian Studies 1994

VOLUME XXXII

CONTENTS Page . . . . . . . . . . . . . . . . . . . . . ii Governing Council iii Report of the Council . . . . . . . . . .. . . .. .. .. . Professor Louis Vanden v ............. Obituary: Berghe Preliminary Report on the 1993 Excavations at Horom, Armenia, by Ruben S. Badaljan, Philip L. Kohl, David Stronach and Armen V. 1 Tonikjan ........................ Elamites and Other Peoples from Iran and the Persian Gulf Region in 31 EarlyMesopotamian Sources, by Ran Zadok .......... The International Merv Project. Preliminary Report on the Second Season (1993), by Georgina Herrmann, K. Kurbansakhatovet al. 53 The Fall of al-Maddiin: Some Literary References Concerning Sasanian Spoils of War in Mediaeval Islamic Treasuries, by Avinoam Shalem ..................... 77 The Chinese Uighur Animal Calendar in Persian Historiography of the Mongol Period, by Charles Melville ............ 83 Anecdotes of a Provincial Sufi of the Dehli Sultanate, KhwaijaGurg of 99 Kara,by Simon Digby ...... .............. The Accession of Iskandar Khan, by Audrey Burton ......... 111 The Ijdzafrom Yidsuf (d. 1186/1772) to Sayyid Muhammad Al-Bahrmni Mahdi Bahr al-cUlfim (d. 1212/1797-8), by Robert Gleave . . . 115 James Baillie Fraser: Traveller, Writer and Artist 1787-1856, by Denis 125 Wright ......................... British Indian Views (Nineteenth and Early Twentieth Centuries) of the Later Followers of the Raukaniyya Sect in Afghanistan and Northern India, by Sergei Andreyev ............ Production of Sugar in Iran in the Nineteenth Century, by A. Seyf . Shorter Notices

Hijji cAbbas,byJ.W.Allan Diz, by Mary Burkett

................

..................

135 139

145 149

THE BRITISH INSTITUTE OF PERSIAN STUDIES A Registered

Charity

c/o The British Academy, 20-21 Cornwall Terrace, London ISSN 0578-6967

NW1 4QP

STATEMENTOF AIMS AND ACTIVITIES 1. The Institute has an establishment in Tehran at which British scholars, men and women of learning versed in the arts, friends of Iran, may reside and meet their Iranian colleagues in order to discuss with them subjects of common interest; the arts, archaeology, history, literature, linguistics, religion, philosophy and cognate subjects. 2. The Institute provides accommodation for senior scholars and for teachers from British Universities in order that they may refresh themselves at the source of knowledge from which their teaching derives, the same service is being rendered to younger students who show promise of developing interests in Persian studies. 3. The Institute, whilst concerned with Persian culture in the widest sense, is particularly concerned with the development of archaeological techniques, and seeks the co-operation of Iranian scholars and students in applying current methods to the resolution of archaeological and historical problems. 4. Archaeological excavation using modern scientific techniques as ancillary aids is one of the Institute's primary tasks. These activities, which entail a fresh appraisal of previous discoveries, have already yielded new historical, architectural, and archaeological evidence which is adding to our knowledge of the past and of its bearing on the modern world. 5. In pursuit of all the activities mentioned in the preceding paragraphs the Institute is gradually adding to its library, is collecting learned periodicals, and is publishing ajournal, Iran, which appears annually. 6. The Institute arranges occasional seminars, lectures and conferences and enlists the help of distinguished scholars for this purpose. It will also aim at arranging small exhibitions with the object of demonstrating the importance of Persian culture and its attraction for the world of scholarship. 7. The Institute endeavours to collaborate with universities and educational institutions in Iran by all the means at its disposal and, when consulted, assists Iranian scholars with technical advice for directing them towards the appropriate channels in British universities.

MEMBERSHIP OF THE INSTITUTE Anyone wishing to join the Institute should write to the membership Secretary, Miss Mary Totman, 63 Old Street, London EC1V 9HX. The annual subscription rates (lstJanuary-31 December) are as follows: ?25 Full membership (U.K. only) ?8.00 Member not receiving journal ?30 or $60 Full membership (Overseas) Student membership ?7.50 COPIES OF IRAN Full members of the Institute receive a post free copy of the current issue of the journal Iran each year. Copies of Iran may be obtained from the Publications Secretary, Miss Mary Totman (address as above) at the following prices: UK addresses ?30 or $60 each-to Current issue-single copies purchased by non-members post free -to addresses overseas plus postage & packing see publications list inside back cover Back numbers-please Those ordering from overseas may pay in US dollars or by sterling draft drawn on London, by international money order or by Eurocheque.

IRAN Volume XXXII 1994

CONTENTS

Governing Council ............................. of the Council ..... ... . Report .......................... .. .... ... Obituary: Professor Louis Vanden Berghe. ........... on the 1993 Excavations at Horom, Armenia, by Ruben S. Preliminary Report Badaljan, Philip L. Kohl, David Stronach and Armen V. Tonikjan. ........ Elamites and Other Peoples from Iran and the Persian Gulf Region in Early ... Mesopotamian Sources, by Ran Zadok ................ The International Merv Project. Preliminary Report on the Second Season (1993), . by Georgina Herrmann, K. Kurbansakhatov et al. ............... The Fall of al-Mada•in: Some Literary References Concerning Sasanian Spoils of War in Mediaeval Islamic Treasuries, by Avinoam Shalem ... ....... The Chinese Uighur Animal Calendar in Persian Historiography of the Mongol Period, by Charles Melville ................... ..... Anecdotes of a Provincial Sufi of the Dehli Sultanate, Khwaja Gurg of Kara, by Simon .. ............. .. ...99 ... Digby ............ The Accession of Iskandar Khan, by Audrey Burton ........... .... The Ijdza from Yusuf Al-Bahrani(d. 1186/1772) to Sayyid Muhammad Mahdi Bahr ......... by Robert Gleave . . ... al-cUlfim (d. 1212/1797-8), Baillie Fraser: Writer and Artist Denis Traveller, 1787-1856, by James Wright . . . . British Indian Views (Nineteenth and Early Twentieth Centuries) of the Later Followers of the Raukianiyya Sect in Afghanistan and Northern India, by ........135 Sergei Andreyev ................... Production of Sugar in Iran in the Nineteenth Century, by A. Seyf .. .. . ..... Shorter Notices .

HaIt-jicAbbas, byJ.W.Allan.......................

Diz, by Mary Burkett

Page ii iii v 1 31 53 77 83

111 115 125

139 145 149

.........................

ISSN 0578-6967

THE

BRITISH

INSTITUTE

OF PERSIAN

STUDIES

(A Registered Charity)

c/o The British Academy, 20-21 Cornwall Terrace, London

NW1 4QP

BRITISH INSTITUTE OF PERSIAN STUDIES (A Registered Charity) GOVERNING COUNCIL President

*tProfessor MICHAEL ROGERS, MA, D Phil, FBA, FSA Vice-President

tProfessor ROBERT HILLENBRAND, MA, D Phil, FRSE Honorary Vice-Presidents

*Professor D. B. STRONACH, OBE, MA, FSA SIR DENIS WRIGHT, GCMG MA Members

MrsJANET AIDIN, MA tDrJ W ALLAN, MA, D Phil Mr DONALD ALLEN Mr C A BURNEY, MA, FSA MrJOHN C CLOAKE,CMG, MA MrJOHN COOPER, MA DrJ E CURTIS, BA, PhD, FSA Dr R W FERRIER,MA, PhD MrJOHN R GRUNDON tDrJ D GURNEY, MA, D Phil Mr DESMOND HARNEY,OBE, BSc Professor A K S LAMBTON, OBE, D Lit, PhD, FBA Dr PAUL LUFT, MA, PhD tProfessor K S McLACHLAN, MA, PhD Dr CHARLES MELVILLE,MA, PhD Dr DAVID O MORGAN, BA, PhD Mr A H MORTON, MA Mr CJ S RUNDLE, OBE, MA Professor A REZA SHEIKHOLESLAMI,MA, PhD Mr PETERJ W TAYLOR,OBE, MA Honorary Treasurer

tMrJOHN S PHILLIPS, TD, MA, FCA Honorary Secretary

tDr VESTA CURTIS, MA, PhD Honorary Joint Editors

tProfessor C E BOSWORTH, MA, PhD, FBA tDr VESTA CURTIS, MA, PhD Assistant Secretary

Miss DIANA MORGAN Membership and Publications Secretary

Miss MARY TOTMAN Special Adviser

Mrs M E GUERITZ, MBE Auditors

PRIDIE BREWSTER,29-39 London Road, Twickenham, Middlesex TW1 3SZ. c/o The British Academy, 20-21 Cornwall Terrace, LONDON NW1 4QP

tMember of Executive Committee *Editorial Adviser

P.O. Box 11365-6844 Tehran, IRAN

REPORT OF THE COUNCIL to 31st March 1993 The most significant event of the year was the appointment of new officers and staff of the Institute. Professor Keith McLachlan retired after a period of three years as President and Professor Michael Rogers was elected in his place. Professor McLachlan took over in 1989 at a crucial time when the future of the Institute had once again reached an uncertain and critical point. With his enthusiasm, positive attitude and immense knowledge and dedication to Iran, he refused to accept defeat and created a period of stability and optimism which influenced all those who worked with him. At the same time Mrs. Mary Gueritz, our hardworking and enthusiastic Assistant Secretary handed in her resignation after more than thirty years of complete devotion to the Institute, its officers, fellows and members. Her expertise and encouragement and at the same time her love of Iran, its people and culture were unique. It was for these efforts that she was awarded an M.B.E. in 1982. Although she has officially retired as Assistant Secretary, Mrs. Gueritz has taken up a new appointment as Special Adviser and we look forward to her continuing association with the Institute. Another sad retirement was that of the Hon. Membership Secretary, Mrs. Molli Cloake, who for many years, on a voluntary basis, gave much able to assistance to Mrs. Gueritz. Finally, thanks are also due to Professor Robert Hillenbrand whose term as Hon. Secretary came to an end but who has now taken a new position as Vice-President of the Institute. The new Hon. Secretary is Dr. Vesta Curtis. For the post of Assistant Secretary the Institute has been most fortunate to find Miss Diana Morgan. She and the Membership and Publications Secretary, Miss Mary Totman, have set up the London office and are undertaking the process of computerising accounts, publication orders and membership records. Mr. Peter Davies was appointed Hon. Covenant Secretary. The extensive work on the Institute building in Tehran has continued under the supervision of Dr. Georgina Herrmann and the Treasurer, Mr. John Phillips. Their task has been made even more difficult by the need to direct the work from London. Nevertheless, the roof has now been replaced, the Assistant Director's flat has been made habitable and work is well advanced on the plumbing and rewiring. In Tehran, Mr. Manouchehr Bayat and Mr. E. J. Andrews, Technical Supervisor of the British Embassy, gave them invaluable advice and support. The year also witnessed visits to Iran by a number of members and fellows, including Dr. James Allan, Keeper of Eastern Art at the Ashmolean Museum, who attended a conference, on carpets, in Tehran. Mr. Allan was able to travel in Iran and also look at the reserve collections of the Iran Bastan Museum in Tehran. Awards for study visits and projects in Iran were also made to Dr. John Gurney and Mr. Kamran Safamanesh to undertake jointly "A Survey of Qajar Architecture of Public Buildings in Tehran" and to Professor Keith McLachlan to continue a project on qanats and traditional water provision, begun in 1991. Grants for travel to Iran were given to Dr. Vanessa Martin to study "Aspects of Modern Shi'ism, Particularly the Relationship Between the 'Ulema and the State in the Mid-Twentieth Century", to Dr. Richard Tapper and Dr. Susan Wright to attend an international conference in Isfahan on "Nomadism and Development" and to Miss Susan Bull to study "Fatima as a Role Model for Women in Iranian Shi'ism". Mr. Paul Mitchell, Ms. Sophie Godrick and Ms. Mariam Imani received grants to clear the library and sort out the many boxes of sherds, thus enabling the building contractors to continue with their work. Dr. Javad Golmohammadi and Dr. Michael Harverson took up the grants that had been awarded to them during the previous financial year and visited Iran. A travel grant for work outside Iran was given to Ms. Elaine Wright to study in the Topkapy Saray in Istanbul "The role of the Pir Buda Qara Qoyunlu and the Evolution of Fifteenth Century Manuscript Illumination". Also a special grant was given to Miss Jennifer Scarce towards the cost of a Day Conference at SOAS in November on "Language of Dress in the Middle East". Grants from the Institute's own funds aimed at promoting work in Central Asia were given to Dr. Georgina Herrmann (for the International Merv Project) and to Dr. Vesta Curtis (to attend the iii

Third International Merv Conference in Turkmenistan and for travel to Bukhara in Uzbekistan). The summer lecture entitled "Fops, Floozies and Farangis; the Late Work of Rizayi cAbbasi" was delivered by Dr. Sheila Canby, Assistant Keeper of Oriental Antiquities of the British Museum. The Annual Lecture was given by Dr. James Allan on "'Alams and Artisans; the Tradition of Steelworking in Iran". A large audience enjoyed the lecture and particularly appreciated the slides taken by Dr. Allan during his recent visit to Iran. This followed the Thirtieth General Meeting at the British Academy on Wednesday 18th November 1992. The resignations of Professor Ilya Gershevitch, Dr. Clare Hill and Mr.John Hanson were accepted with regret. Other members due to retire in rotation were re-appointed. These were Professor Michael Rogers, Professor Robert Hillenbrand, Mr. John Cloake and Dr. Vesta Curtis. Dr. David Morgan, a former member was welcomed back to the Governing Council and Mr. John Cooper was invited to join as a new member. We learned with great sadness of the death of Michael Browne, Q.C., on 1st April 1992. He was a member of our Council from 1966 until 1988 and gave valuable advice to the Institute on legal matters. Once again, a successful series of monthly lectures was held at SOAS on Thursday evenings in cooperation with the Centre of Near and Middle East Studies. Under the general title of "Iranian Art and Archaeology", talks were given by Dr. Barry Flood, Dr. Eleanor Sims, Mr. Marcus Frazer and Miss Teresa Fitzherbert. For help of various kinds the Governing Council wishes to express particular thanks to the staff of the Middle East Department of the Foreign and Commonwealth Office, to Mr. David Reddaway, the former Charge d'Affaires, to Mr. E. J. Andrews, Technical Supervisor in Tehran, for his advice and help with the building and to the British Academy for its support and financial assistance. Thanks are also due to Mr. Manouchehr Bayat, BIPS' financial supervisor in Tehran and Mr. Houman Kordmahini, caretaker of the Tehran Institute. We are also grateful to Mrs.Janet Aidin for her legal advice. VESTA CURTIS Hon. Secretary

MICHAEL ROGERS President

iv

OBITUARY LOUIS VANDEN BERGHE (1923-1993)

It is sometimes said of. great scholars that they do not just make contributions to their subject, they change its shape. In such a league was Prof. Dr. Louis Vanden Berghe, doyen of Iranian archaeologists, who died in Gent on 17th September 1993 aged 69 years. He was Professor of the Archaeology and History of Ancient Near Eastern Civilisations at the University of Gent, and also Professor of the Archaeology and History of Ancient Iran at the Universite Libre de Bruxelles, as well as being Director of the Iranian Section of the National Museum in Brussels. He was born on 24th December 1923 in Oostnieuwerke in West Flanders, between Bruges and the French border, and studied Near Eastern archaeology and the history of art at the University of Gent and Oriental languages at the Universities of Brussels, Amsterdam and Leiden. This linguistic training later stood him in good stead, and amongst his many achievements was a remarkable fluency in modern Persian. Having obtained his doctorate in 1950 with a thesis about prehistoric painted pottery, he became an assistant in the University of Gent in 1951, rising to professor in 1957 and being given a chair in 1965. During an extremely productive career he produced about 150 books, catalogues and papers, with the emphasis throughout being on Iranian archaeology and history. His first major task was to give some structure to a subject which was still in the 1950s in a state of some disorder. To this end, his Archeologiede l'IranAncienappeared in 1959. This is a masterly survey of Iranian archaeology, first by region and then by period, going from prehistoric to Sasanian times. It has well stood the test of time and the numerous discoveries made in the 1960s and 1970s, and containing as it does a wealth of information it is still widely consulted. Shortly afterwards he was instrumental in establishing, together with R. Ghirshman, a newjournal devoted to Iranian archaeology. The first volume of Iranica Antiqua appeared in 1961, and largely as a result of Vanden Berghe's energy and drive it still continues. Further evidence of his determination to introduce order and discipline into the subject can be found with his Bibliographie analytique de l'archeologie de l'Iran Ancien (Leiden 1979), supplements

to which appeared

in 1981 and 1987 covering the years 1978-1985. Vanden Berghe is perhaps best known for his archaeological work in Luristan. Between 1965 and 1979 v

he led an expedition sponsored jointly by the University of Gent and the Musees Royaux d'Art et d'Histoire in Brussels that for fifteen seasons worked in the Pusht-i Kuh, the western part of Luristan. More than 30 cemetery sites were investigated, varying in date between the Chalcolithic period and Iron III. Particularly productive sites were Parchineh, Bani Surmah, Kalleh Nisar, Kutal-i Gulgul, Bard-i Bal, Tattulban, CamziMuimah and War Kabud. A monograph about War Kabud appeared already in 1968, and preliminary reports about the other sites appeared with commendable speed, usually in the French journal Archedologiaor in Iranica Antiqua. The contribution that Vanden Berghe has made through his work in Luristan has been immense. This remote and mountainous region is poorly known archaeologically, and there has been only a small amount of scientific research here. Apart from Vanden Berghe's work, the few scientific projects include the Holmes Expedition to Luristan, that resulted in the 1938 excavation of E. F. Schmidt at Surkh Dum-i Luri in the Pish-i Kuh, and Clare Goff's excavations at BabaJan. It is known that the many "Luristan bronzes" in collections around the world come from this area, but practically nothing is known about the culture which produced them. Vanden Berghe's work has gone some way towards remedying this deficiency with the discovery of about a dozen canonical Luristan bronzes. It would be wrong, however, to think that Vanden Berghe's work in the field was restricted to Luristan. Among many projects we might cite, for example, his survey of the Mary Dasht Plain in the 1950s and his brief excavations at the necropolis of Khurvin which resulted in a book published in 1964 (La Nicropole de Khuirvin). He also discovered an ancient road connecting Firuzabad and Siraf, an Achaemenid tomb at Buzpar in Fars and many "chahar taqs" (fire temples). His versatility is demonstrated by the description of an early Islamic castle at Puiskafn in Fars (La Decouverte d'un chdateau-fortdu dcbut de l'dpoque islamique a Puskan (Irdn), Iranica Antiqua Supplement IV, Gent 1990). He also recognised the importance of promoting the subject, and to this end he was an indefatigable organiser of exhibitions. For example there were major exhibitions about Luristan in Munich (1981), Gent (1983) and St. Petersburg (1992), and an exhibition about Urartu in Gent (1983). These were all accompanied by attractive and informative catalogues. Mention should also be made of an exhibition of photographs (by Erik Smekens) of Iranian rock reliefs organised in Brussels in 1984 (Reliefs Rupestres de l'Iran Ancien, Brussels 1984). This was followed in the next year by ajoint publication with Klaus Schippmann on Les Reliefs rupestres d'Elymaide (Iran) de l'dpoque parthe (Gent 1985). Several of these reliefs were discovered by Vanden Berghe himself. During his lifetime many honours came to Vanden Berghe and he was much respected. In his own country he was a corresponding member (1962) and a full member (1968) of the Academie Royale des Sciences, Lettres et Arts de Belgique, becoming President in 1973. Abroad, honours included being a full member of Institut (1973), an honorary fellow of the Society of Antiquaries of London the Deutsches Archfologisches (1980) and a fellow of the Explorers Club of New York (1978). Much prized was his honorary doctorate from the University of Tehran, awarded as early as 1964. In some ways Louis Vanden Berghe lived life to the full. His consumption of alcohol was legendary and he was seldom without a large cigar. Yet at the same time he was a lonely person, and for much of his life lived in a hotel near the main railway station in Gent. From there he journeyed sometimes to his home village in West Flanders. The drawbacks to this solitary lifestyle were, however, largely offset by the devoted help and support of a number of friends and former students including Ernie Haerinck, the late Christiane Langeraert-Seeuws and latterly Alexandre Tourovets. It was one of these colleagues, Dr. Haerinck, who together with Prof. L. de Meyer edited a two-volume Festschrift in his honour in 1989 (Archaeologia Iranica et Orientalis: Miscellanea in honorem Louis Vanden Berghe) which included an informative memoir about Vanden Berghe together with a full bibliography. Happily, Haerinck has now succeeded his old teacher as lecturer in Near Eastern archaeology at Gent. In spite of failing eyesight, Vanden Berghe continued to work right up until the end of his life. Even in his last year he organised an enormously impressive exhibition of Sasanian art in Brussels accompanied by a sumptuous catalogue (Splendeur des Sassanides, Brussels 1993). And it is gratifying for friends and admirers in Britain that he was able to attend the 1993 Lukonin Memorial Seminar in the British Museum on "Later Mesopotamia and Iran, c. 1600-539 B.C." To this event he contributed a lecture on "Excavations in Luristan and Relations with Mesopotamia", and although he could not deliver it in person he was able to introduce it, which was much appreciated. The passing of Louis Vanden Berghe leaves a major gap in the ranks of Iranian archaeologists, but his many contributions have pointed the way to others who will surely follow in his footsteps. Above all, his commitment to the subject never faltered, and in spite of the political changes which have dominated the last 15 years he never doubted the value of studying Ancient Iran and remained optimistic about future prospects. JOHN CURTIS

vi

PRELIMINARYREPORT ON THE 1993 EXCAVATIONSAT HOROM, ARMENIA and Armen V. Tonikjan By Ruben S. Badaljan, Philip L. Kohl, David Stronach and Yerevan,Wellesley,Massachussetts,

Berkeley

NORTH HILL EXCAVATIONS

Collaborative American-Armenian and Americanon the investigations archaeological Georgian Shirak plain of northwestern Armenia and the Djavakheti plateau of southern Georgia (see below: Appendix A) began during the summer of 1990 as field component the principal, regionally-focused and structured of the more broadly conceived for International Anthropological Program Research in the Caucasus (or IPARC). The 1990 and 1992 excavations at the site of Horom, Armenia have been reported in previous volumes of Iran (Badaljan et al. 1992 and 1993), while the 1990 and 1991 field seasons in Georgia, particularly at the Early Bronze Age site of Satkhe, have been briefly described in the Bulletin of the Asia Institute the (Kohl et al. 1992). This report presents excavations the 1993 of results preliminary at Horom1 and includes the Palaeolithic investigations and later Bronze and Iron Age prein in southern historic excavations Georgia Appendix A.2 On the basis primarily of surface reconnaissances and topographic mapping undertaken in 1990, the c. 45-50 ha. Horom settlement, which is centered around two dominant hills set along a north-south axis (the North and South Hills) and characterised by standing cyclopean stone architecture, was distinguished from a c. 400 ha. mapped area termed the Horom site, which consists of undulating, rocky stone structures of uncertain terrain containing date and clusters of circular stone-ringed tombs or cromlechs dating to the late second or early first millennium B.C. (Badaljan et al. 1992: fig. 3). The Horom settlement was located within the northwestern quadrant of the Horom site. This view may need to be modified in as much as additional stone structures, including possible fortifications, were recently noted to the east and southeast of the Horom and future excavations within such settlement, structures may relate them to the Iron Age outlying Urartian fortifications on the North Hill. Thus, the overall size of the fortified settlement remains unclear, and the 45-50 ha. figure could well represent a minimal estimate. At the end of the taken from a heli1993 season, photographs copter provided a series of aerial views of the Horom settlement and its cyclopean architecture (Pl. Ia-b).3

Constructions of Urartian Date While no materials of early first millennium B.C. date chanced to come to light within the restricted sounding of 1990, the results of the 1992 and 1993 excavations have made it clear that the major visible fortifications on the North Hill are to be ascribed to the Urartian period. In more precise terms the period in question is likely to have lasted for more than a century and a half, between the time that Argishti I (c. 785-760 B.C.) annexed the presentday region of the Shirak plain (cf. Badaljan et al. 1993: 15) and the moment when the kingdom of Urartu came to an abrupt end, somewhere close to 600 B.C. The association of wheelmade, buff to red pottery of an Urartian or Urartian-related type with grey wares which clearly continue a local Late Bronze/Early Iron Age ceramic tradition is currently from Operations Bi, B2, C2, firmly documented and D1 (Fig. 5). The association was first acknowledged during the 1992 season (Badaljan et al. 1993: 12-23), and the work of the past summer, in Areas B2 and D1 in particular, has continued to demonstrate the strength of this connection. Above all, the recent discovery of a small cosmetic jar of a known Urartian type (Fig. 5: 19, P1. Ic), such as occurs elsewhere at Bastam, for example (Kleiss et al. 1979: 208, fig. 3, 5) in the deep fill of one of the better preserved rooms of the B2 area (Room 3 of B2 complex see Fig. 3, P1. IIa) has underlined the likeliof the hood that the more intact magazines century B.C. "border settlement" at eighth/seventh Horom will one day serve to document the extent to which the local Iron Age wares of the Shirak by elements of the wellplain were complemented known ceramic corpus of the Urartian core area. In the balance of this report all Horom's architectural remains of the eighth/seventh century B.C. are referred to as "Urartian". It should be understood, however, that this label deserves to be read in several separate ways. On the one hand, the label is used with the above-mentioned chronological connotation and, on the other hand, it is intended to be indicative of a period of intense interaction with Urartu or of the citadel's physical incorporation-at 1

2

JOURNAL

OF PERSIAN

least at certain times, within the accepted boundaries of the Urartian state. In other words, the present writers are acutely aware that, while the presently exposed architecture on the North Hill offers its fair share of Urartian traits, the site as a whole has so far produced remarkably few incontrovertible Urartian artefacts. But it may also be noted before we proceed to review the work of the past season which took place in Operations A2/3, B2, and D1 that the small piece of a bronze quiver that was found on the heavily burnt roadway of the BI North-West Gate at the end of the 1992 season (Badaljan et al. 1993: 12-15 and pl. IId) was decorated in a distinctive fashion; i.e., it was marked by groups of transverse ridges in repousse (P1. Id) in a design that otherwise occurs on a group of five bronze quivers from the Urartian citadel at Kayalidere (Burney 1966: pl. 18c). The 1992 Operations B1 and C2 were not continued in 1993, but several architectural features including a stone-covered drain extending down the central axis of the B1 North-West Gate, which were uncovered at the very end of the 1992 season, were accurately mapped. We hope to continue work in these two areas in future seasons. Operation A2/3 (Figs. 1 and 2) Early in the 1993 season a small, exploratory trench, only 3 x 3.5 m. in size, was opened on the flat summit of the North Hill near the presumed northeastern corner of the A terrace. The precise location of this probe, called Operation A2, was determined by the presence of stone remains that were visible inside an animal burrow. Although the exposure was limited, stone architecture was encountered, and a thick Early Mediaeval deposit was explored to a depth of 1.5 m. In the course of the work, a large, complete storage vessel of Mediaeval date was removed, and an adjoining burnt surface yielded a rich cache of well-preserved carbonised seeds. It was subsequently decided that it would be of advantage to extend the excavation for a distance of 9 m. to the northeast down the slope of the A terrace. It was hoped that the extension, termed A3, would throw useful light on the still little examined uppermost line of Urartian fortifications and that it would also serve to reveal any subsidiary architectural features that might have been built up against the inner face of the A Wall. To begin with, work in the A3 extension proved to be a considerable challenge: a nearly unbroken sequence of parallel stone faces was found to extend throughout the eastern end of the narrow trench. It was only when the outer face of the A Wall was at last exposed, in fact, that this apparent sequence of "walls"could be read

STUDIES

as a series of monumental steps that advance, for a distance of 6 m., up the steep slope to the line of the flat summit. While much more of the A Wall needs to be exposed in order to understand its construction in full, it does now appear that we have uncovered the point at which the A Wall turned to the north-west, in accord with the natural contours of the North Hill (Fig. 1), and that the wide series of steps very probably functioned as a socle for the now missing stone superstructure that presumably once defended the flat summit at the apex of the whole Urartian fortification system. The materials from the eroded, steeply sloping terrain of Operation A3 included, hardly against expectation, examples of Urartian red ware mixed with sherds of known (and presumed) Mediaeval date. Operation B2: Architecture located on the B Terrace inside the Inner Face of the B Fortification Wall (Fig. 3, P1.IIa) From the prior account of the work that was carried out in Operation B2 during the 1992 season (Badaljan et al. 1993: 15-18, fig. 13), it will be recalled that this part of the site (Fig. 3 and P1. IIa) had already revealed well-preserved architecture that was located just inside the inner face of the B Fortification Wall. Accordingly, excavations were at once resumed in Room 2, one of two adjoining rooms (Rooms 1 and 2) that had been at least partially defined in 1992. Since the floor of Room 2 had only been reached within the limits of a small test trench in the course of the previous campaign, one of our first concerns was to expose the remaining, available floor surface. The only feature of note to come to light in the course of this endeavor consisted of the remains of a clay walled oven (or tanur), the floor of which was still covered by a layer of ash. The flue which had been used to draw air into the oven was also detected; it could be seen to have been sunk into the top of the earth floor, near the middle portion of the room, close to Wall 2 (P1. IIb). It ought to be added, however, that such a distinctly domestic feature offers a curious contrast to the monumental walls which adjoin it; for this reason, therefore, the oven could well represent a later feature which is not to be connected with the otherwise undetermined, orginal function of Room 2. As far as the 1993 extensions to Operation B2 are concerned, these took the form of the B2.3 trench to the south, where Rooms 3 and 4 were exposed; the B2.4 trench to the west, where a well-preserved room or corridor, Room 5, was uncovered; and the B2.5 trench to the north, where the work disclosed a number of barely sub-surface, one-to-two course

REPORT

PRELIMINARY

f

ON THE

1993

AT HOROM,

EXCAVATIONS

oo1620

ARMENIA

3

++

2B

C2

,

C40

E2

El

//

.

???..

.

.

.

?

5M5

~cM4

......~rr\~~~r~,+

m

mM .

20

10

0

.

.

. . .

.

.

.

0

Fig. 1. Generaltopographicmap of Northand South Hills of Horom,showingexcavationunits of 1992 and 1993. Thesummitof the NorthHill is also markedbythe remainsof a previouslyexposed,EarlyMediaevalbuilding.

4

JOURNAL

OF PERSIAN

wide stone walls (Features 12-16), all of which undoubtedly post-date the more deeply set rooms of the main B2 architectural complex. Indeed, the narrow walls in B2.5 (Fig. 4) are almost certainly contemporary with two other virtually sub-surface, insubstantial walls (Features 1 and 2) which were found to run above the massive, two metre and more high remains of Wall 6. In sum, then, this single part of the Horom settlement may be said to offer persuasive evidence for the existence of a later, less monumental period of occupation that was either Late Urartian or post-Urartian in date (see Fig. 7).4 It is tempting, of course, to link the thin stone walls which eventually came to be erected over the abandoned B2 complex with other examples of inferior, very often clearly secondary construction that have been noted in other sections of the North Hill. One of the more striking instances of late, secondary building activity may be said to come from the BI North-West Gateway, where the socalled guardroom represents an obvious late addition to the original fabric of the Gate (Badaljan et al. 1993: 15 and pl. IIc). In addition, it would not be at all surprising to find that the whole of the D1 complex (Fig. 8), which has for long been seen to stand outside the limits of the otherwise formidable fortification system on the North Hill (see below and Badaljan et al. 1992: 44-5), was itself representative of such a widespread, late building phase. While the ceramics from the North-West Gate, B2.5, and D1 are clearly similar, with each containing well-fired Urartian red wares together with grey wares of the still persisting local Late Bronze/Early Iron Age tradition, Operations B2.5 and D1 are alone in also producing sherds of still another kind with either wavy-combed or pattern-burnished decoration (Fig. 6: 13, 14 and Fig. 9). Eventually, therefore, pottery of this latter type may come to serve as a valuable, separate indication of late occupation. As for the relationship of such a late phase of occupation to the date of the fire which ravaged the North-West Gate, it is a matter of decided interest that the traces of extensive burning on the BI roadway were found to run up against the exterior of the already extant east wall of the above-mentioned "guardroom". In other words, we can be sure that at least some of the late and very largely inferior stone walling on the North Hill pre-dates rather than post-dates the time of the Gate's destruction. For the present, therefore, it may be justified to categorise the more or less massive Urartian architecture on the North Hill as "Early Urartian" and those examples of clearly related, but assuredly later, local building activity as "Late Urartian."

STUDIES

HOROM1993 AREA A2

0o2m

W2

1669.72

1669.76 1669.38

1670.18

W

16 9.

7

52

1670.86

W5

S1670.14

W4

W3

1669.43

1669.14

1669.54

W6

UNEXCAVATED 1668.75

1668.23 1668.15 1668.43

"\,A.os Fig. 2. OperationA2/3, summitof NorthHill, illustratingstepped fortificationsystemon summitas it turns to thenorth-west.

PRELIMINARY

REPORT

1993

ON THE

EXCAVATIONS

AT HOROM,

5

ARMENIA

HOROM1992-93 A

AREAB2/CI

B2 0

1

2M LL 1645.00 FLOOR 1645.20

S164,

WW9

1646.10 164 3.48

7

1

LL

6.1644.92/

,0

LL 1643.27

LL 1644.03

OVEN

aQ

1644.85

~~~1643.16 1643.97

L639

LL164369

W4 P

3 ROOM

1644.16 FLOOR ROOM

2

W2ROOM

1

1.1642.40

0L

1643.47 LL LIMIT OF EXCAVATION

o

0

AT UPPER LEVELS

D

o

o~ 08

wl

d0~

TOMB E.B. 1640.96 Cl

w

1642.71

E

UNEXCAVATED

A F8

1642.87

Fig. 3. Plan of the substantial "Early Urartian" architecture in Area B2. Note also, in Area C1, certain traces of "Late Urartian" walling and the location of the Early Bronze tomb outside (and well below the level of the B Fortification Wall.

6

JOURNAL

OF PERSIAN

If Rooms 2-5 in the B2 area may be taken to exemplify certain of the main characteristics of "Early Urartian" construction at Horom, such features would appear to include consistent orientation, symmetrical planning, and the use of relatively heavy, skillfully built dry-stone walls (Pls. IIa-c and Fig. 3).5 With respect to two of the rooms (Rooms 2 and 3), certain of the walls were also found to retain traces of a plaster surface. Doorways are, of course, well represented. One doorway linked the seemingly paired square rooms, Rooms 3 and 4 (Pl. IIc); another connected Room 3 with a still unexcavated room to the south (P1. IIc); and yet another provided a link between Rooms 5 and 2, at least until such time as the west wall of the latter room (Wall 4) came to be widened over the greater part of its length (Fig. 3).6 Of special interest, of course, is the perfectly preserved doorway between Rooms 3 and 4; rectangular in shape, and capped by flat stones, the doorway was given a height of almost 2 m. and a width of 89 cm. Needless to say, the doorway's all-stone frame was by the exclusive use of stone in predetermined these ground floor walls, many of which still stand to a height of 3 m. Last but not least, a still not fully defined opening at the north end of Room 5 is likely to have provided access to yet other associated rooms, including Room 1. The number of small objects from Rooms 2-5 was not large, but the discovery in Room 3 of the cosmetic jar (P1. Ic, Fig. 5: 19) above-mentioned and a fragment of a large storage jar with a seal impression (P1. Ie) could perhaps indicate that this part of the site was occupied, at one time, by administrators or, at all events, by residents of some status. Other finds of note include a winged bronze arrowhead from the north end of Room 5 which directly accords with a standard Caucasian type that was current for a period of several centuries both before and after 1000 B.C.,7 and a more perplexing object: namely, an iron door-key (Pi. IIe), also from Room 5, which looks strangely out of place in an century B.C. context. That is to say eighth/seventh that, while the key was indeed recovered from the floor of Room 5 near the later blocked doorway that led into Room 2, it has an oddly modern appearance; and, in view of the local presence of many animal holes, its actual find spot could be fortuitous. Finally, it should be noted that Early Bronze Kura-Araxes sherds chanced to be found beneath the floors of both Rooms 3 and 5. In other words, the whole of the B2 complex would appear to have been built over a much earlier Early Bronze finding which is also consistent Age occupation-a with the discovery, in 1992, of an Early Bronze tomb in the C1 Operation (Fig. 3 and 7), beyond the outer or eastern face of the B Fortification Wall

STUDIES

(Badaljan et al. 1993: 4 and pl. Ia). Architecture Operation D1-Domestic Terrace (Fig. 8)

on the D

Prior to excavations in the D 1 Area, it was speculated that the whole of the unfortified D Terrace, which is so very different in appearance from the various fortified sectors of the North Hill (Fig. 1), could have represented the remains of a separate "administrative complex" (Badaljan et al. 1992: 44-5). In the aftermath of the 1992 campaign this characterisation remained unchallenged, if only because it seemed premature to rush to conclusions on the basis of two strictly limited, local probes. Indeed, the one clear message that emerged from the 1992 excavations in this part of the site was the fact that all construction on the D Terrace could have been restricted to a "relatively late date" within the time-frame of Horom's eighth/seventh century B.C. occupation (Badaljan et al. 1993: 21). On the one hand, the work of the 1993 season has clearly strengthened the case for such a probable late date (see page 4 above), and, on the other hand, there is now a sound case to be made for the domestic, rather than the representational-or administrative of those D1 structures (Rooms 1-3) -character that have been examined to-date. Of the rooms in question-all of which run along the same north-south axis-Room 1 was excavated in its entirety in 1992. It is a sub-rectangular unit, 6.7 x 3.8 m. in area, with a flagstone floor and with a broad stone platform located in front of its slightly curved western wall (Fig. 8). In accord, moreover, with the less than regular character of the walls of Room 1, the corridor which leads northwards to the next room, Room 2, consists of no more than a narrow, bent passage of irregular width. Because of other calls on our time and resources, only about one third of Room 2's total floor area of 8 x 9.5 m. came to be exposed in 1992. The main features to emerge in the course of the excavation consisted of a large, squarish basin of pink tuff set close to the line of the west wall (Badaljan et al. 1993: pl. IIIe) and a low "double partition" (composed of two parallel rows of slim stones set on edge) which stood perpendicular to the west wall. The presence of this last fixture, and the considerable size of Room 2, suggested at the time that the room might have been subdivided into several separate "cubicles" and that, in addition, the whole space might have remained unroofed. Following the 1993 season, however, new assessments are in order. In the first place, the discovery of four symmetrically placed stone column bases (Fig. 8, P1. IId) would seem to prove that Room 2 was roofed; and, secondly, if due account is taken of cer-

PRELIMINARY

REPORT

ON THE

1993

EXCAVATIONS

AT HOROM,

7

ARMENIA

HOROM1992-93 AREA B2 UPPER LEVELS

O

1

2M

Fig. 4. OperationB2.5. "LateUrartian"architectureexcavatedto the east of theB FortificationWall and to the northof theB2 "Early Urartian"complexof Fig. 3.

8

OF PERSIAN

JOURNAL

tain of the present-day fixtures that are associated with the stabling of animals in the local village, there can be little doubt that Room 2 came to be used, for a time at least, to water and shelter animals. Thus, while the tuff basin, which stands on a low stone platform, almost certainly represented a suitable water container, both the original east-west partition, and a newly exposed double partition, are likely to have served as feeding troughs-presumably if the width of the doorways is kept in mind, for animals of rather modest size, such as sheep and goat. Since the two long feeding troughs appear to have been sited in such a way as to take advantage of the presence of three of the four columns (Fig. 8), it is also possible that there was a time when Room 2 was roofed, but not yet used as an animal shelter. In this last regard it is in fact evident that Room 2 retained a rival identity right to the end. That is to say that parts of the room were supplied with a well-laid flagstone floor and that a variety of domestic activities continued to be pursued, especially towards the north end of the room. Many pestle-shaped ground stones were recovered, for

STUDIES

example, in an area of the room where both a circular stone-lined hearth (Feature 7) and an adjacent stone table (Feature 9) were located. In addition, a small stone-lined pit (Feature 8) may have been on the analogy of contemporary used-again collect glowing coals from the nearby practices-to large stone-lined hearth. An entry from Room 2 to the north led into the Room 3. This slightly smaller, irregularly-shaped room contained four large flat grinding stones set on a stone platform (Feature 1) in its north-eastern corner. The remains of a hearth (Feature 2), possibly for baking bread, lay immediately to the south, and to its west lay a shallow circular stone "milling" not necessarily funcbasin, similar in form-though tion-with those repeatedly uncovered on the top of the South Hill. If Room 2 housed animals, Room 3 may have functioned as a bakery. Room 3 also was connected to another room on its northern side, which will be excavated in the future. As one proceeds north in this interconnected complex, the in places depth of deposit increases, exceeding 1.5 m. Noteworthy also were the remains of charred 4

1

5

6

7

9

8

10

11

18 __19

12

13

14

15

16

i 17

li Fig. 5. Ceramicsfrom B2 Complex.Reddish-Buffand Red Wares,Urartianperiod.

PRELIMINARY

REPORT

ON THE

1993

section of wooden beams in the north-western Room 3, as well as considerable evidence for uncontrolled burning near the hearth of Room 2. The evidence suggests that this complex may have been devastated by fire and recalls the evidence for largescale burning from the Bi North-West Gate area. The D1 complex, which presently appears to have a substantial domestic character, seems to have been occupied during one period, presumably in Late Urartian times. The ceramics from the area consis(c. 20 per tently include a significant component to those similar red fine Urartian of wares, cent) found in the Ararat Valley to the south, as well as more numerous wares continuing the local Shirak ceramic traditions of the Late Bronze/Early Iron periods. As noted above, a few more elaborately decorated vessels (Fig. 9) from this trench together with those from B2.5 may help define the Late or immediately Post-Urartian phase at Horom. Three radiocarbon dates taken from the 1992

8

9

EXCAVATIONS

ARMENIA

9

excavations in the B1 and B2 areas were received from the AMS Facility of the University of Arizona: 1) sample no. AA-10194--B1 a-c, gateway: 2,520+/-55 (yr BP) or calibrated at 790-430 B.C. (1 sigma; or locus 4, 800-410 B.C. at 2 sigma); 2) AA-10189-B2, north of Feature 8: 2465+/-55 (yr BP) or calibrated at 760-410 B.C. (1 sigma; or 790-400 B.C. at 2 TT4, level 4 (beneath sigma); and 3) AA-11129-B2, or calibratthe earliest floor of Room 2): 2770+/-55 ed at 970-830 B.C. (1 sigma; or 1050-800 B.C. at 2 sigma). The dates are not enormously helpful for refining the Urartian chronology at Horom, since their calibrated ranges simply preclude any further subdivisions. The AA11129 date from the 1992 deep sounding beneath the eastern third of Room 2 is somewhat surprising in that it suggests that there may have been an immediately pre-Urartian occupation in this area of Horom which presumably had been totally destroyed by the construction of the B Fortification Wall and the high-standing architecture

10

13

AT HOROM,

11

14

12

greywarevessels Fig. 6. Examplesof black,brownand greywaresfrom theB2 Complex.The distinctivepattern-burnished (nos. 13 and 14) from OperationB2.5 appearto berestrictedto a "LateUrartian"horizon.

10

JOURNAL

OF PERSIAN

STUDIES

HOROM1992- 93 AREA CI/B2 B2

3 ROOM

4 ROOM

5 ROOM WALL 8

IWALL9

FLOOR

FORTIFICATION

CI

WALL B

FLOOR DOORWAYS . i'

0 SECTION A - A

Fig. 7. SchematicSectionof C1/B2 showingheightsof EarlyBronzetomband Rooms3-5 of "EarlyUrartian"period. built against it; by itself, the archaeological evidence in the B2 area only demonstrates a much earlier Early Bronze occupation beneath the Urartian architecture (see above). The Second Millennium Occupation and the Early Bronze Occupations at Horom and Anushavan Operations and 11)

C3b/5 and E2: the Gate 6 area (Figs. 10

The initial 1990 sounding at Horom was placed just within and west of Gate 6 in the C Fortification Wall. No distinctive Urartian materials happened to be recovered from this sounding, and for this reason the fortifications on the North Hill were initially dated to the pre-Urartian Early Iron period (Badaljan et al. 1992: 37-41), a conclusion subsequently corrected after the more extensive 1992 campaign. Additional excavations on both sides of the C Wall and to the east and west of Gate 6 were conducted in 1992 (Badaljan et al. 1993: 4-5, 19-21), and these again confirmed the absence of of significant Urartian remains in this area-save, course, for the C Wall itself. The recovery of several painted Middle Bronze sherds, and a single, surprisingly early calibrated C14 date of 1887-1693 B.C. no. AA-7766) from the bottom (1 sigma-sample of the "Early Iron" level in the 1990 sounding sugin this area could have gested that occupation extended back even into the Middle Bronze period or the first half of the second millennium B.C.

Additional work in 1993 reinforces this impression. Work began in C5 which extended 2.5 m. west of the 1992 C3b trench and 4 m. to the north. C5 with C3b in order to ultimately was connected expose a sufficient area to understand the stratigraphic relationship of the diagonally running wall, initially discovered in 1990,8 with the C Fortification Wall. Since this diagonal Wall lb ran beneath the C Fortification Wall, a small 2 x 3 m. trench E2, which was later extended into a 4 x 5 m. sounding, was placed outside the C Wall to trace Wall lb's extension farther to the south-west. Excavations in C3a in1992 did not reach the level at which this wall would have appeared (n.b., the north-eastern continuation of Wall lb into the 1990 sounding is not shown in Fig. 10); nevertheless, a calibrated C14 determination from a surface-C3a, area G, locus 31 - apparently above the wall yielded a date of B.C. [at 1 sigma; sample AA-101931290-1050 BP (uncalibrated); or 1370-1010 B.C. 2975+/-55 calibrated at 2 sigma]. As currently exposed, Wall lb stretches from the northern section of the 1990 sounding across the connected trenches C3a and C5; it then runs under the C Fortification Wall into the trench E2, continuing into its southern section (Figs. 10 and 11 and Badaljan et al. 1992: pl. XIIa). Thus, this massive wall of uncertain function, which extends for at least 10 m. in this area of the North of the C Hill, clearly predates the construction Fortification Wall and presumably is earlier than the calibrated date from C3a; its orientation is

AREA DI

2M

HOROM1992-93

o0

O I ROOM DJ O

00....

CDC

,,..,

~

1641.41f

18 -642

74

EO

1641

1641.28

BASIN

s

BASIN HEARTH 641 26B

1641474

O 1641.7,16A2.13

6-FLOOR

ROOM 2

81E

SCOL BASE 16AI.51

I /16A

COL

76BASIN 1641 1. 58 16A I03 )6A 18 1641 COLoo BASE 1 16AI BASE I0 16

Fig. 8. OperationDI, Rooms1-3.

ROOM 3

12

JOURNAL

OF PERSIAN

STUDIES

102

3

4

5

Fig. 9. Dl Ceramics, tannish-brown incised nos. I and 2from Room 2 have only beenfound in this operation, also presumably of the "Late Urartian" period; no. 3 is a black painted on greyfragment.

roughly perpendicular to the wall found in 1992 in trench El further to the east which also extended beneath the C Wall. Although additional work is needed to confirm the stratigraphic relationship of these walls, a picture of a substantial earlier settlement in this area of the site, if not even an earlier was subsequently which fortification system, seems to be the Urartian construction, destroyed by coming into focus. During the C5 excavation it became clear that the topmost portion of Wall lb had been cut away by the foundation trench for the C Fortification Wall. As was first noted in 1990, Wall lb appears to have been built in at least two phases (Badaljan et al. 1992: 36-7, fig. 4), the upper phase being partially removed by the construction of the C Wall. Most of the surfaces encountered in C3b/5 and E2 were ephemeral, although one decent plastered surface was uncovered on both sides of Wall lb in C3b/5; and, as far as could be ascertained, this surface separated the two building phases of this same Urartian wares wall. As before, no diagnostic from in area. Ceramics this C3b/5 were appeared mixed: dominantly "Early Iron" local consistently grey wares and occasional Middle Bronze painted and impressed wares were associated with the later rebuild of Wall lb; and roughly equal amounts of unmistakeable, black-burnished Early Bronze KuraAraxes wares and the same type of "Early Iron" wares were associated with the initial construction of Wall lb (P1. IIIa and Fig. 12). Unmixed Early Bronze sherds were only found beneath Wall lb in C3b/5 at the end of the season. Besides ceramics, a stone head and a few stylized anthropomorphic horse bones were recovered from these presumably unmixed Early Bronze levels; the latter have been submitted for radiocarbon analysis. Bedrock was not reached in this operation.

The precise dating for the different phases of Wall lb is unclear, and some of the difficulty relates to the uncertainty of the "Early Iron" attribution for all the local grey ware forms. The continuity of the Bronze to Iron Age ceramic tradition on the Shirak plain may be underestimated by recourse to this terminology. It is true that most of the surfaces here are fleeting or ephemeral, and that there was considerable disturbance in this area associated with the construction of the Urartian fortifications. the architecture, C14 dates, and Nevertheless, ceramics from the different trenches in this area may suggest not so much that the materials are mixed, but that a substantial settlement in this part of the North Hill dates back to the first half of the second millennium B.C., overlying, in turn, a late Early Bronze occupation, dating to the middle or late third millennium B.C, as well as an earlier Early Bronze occupation. A stone cist burial appeared just beneath the surface in the south-eastern corner of trench E2; the burial contained the remains of a young juvenile contracted on its right side with its head oriented to the north-west (P1. IIIb). Two undecorated pots or lugged (Fig. 13), including a double-handled grey ware vessel resting on the pelvis, several glass beads (including one with an "evil eye" design), an iron bracelet worn on the skeleton's arm, and a hollow-based brown obsidian arrowhead were found in the burial. Although this burial was not sealed, it clearly postdated Wall ib, which extended across E2 to the west (Fig. 11). The cist burial clearly differed in its form both from the numerous stone-ringed "Early Iron" tombs or cromlechs located east and south of the Horom settlement (see below and Badaljan et al. 1993: 8-12) and from the collective Early Bronze tomb accidently found in trench C1 of the at the end 1992 season (Badaljan

PRELIMINARY

REPORT

ON THE

1993

EXCAVATIONS

AT HOROM,

13

ARMENIA

HOROM1992- 93 AREA C3/E2

C3

Wlb E2

S0

1

2M

AJ

Fig. 10. Operation C3b/5-E2, showing diagonal Wall lb running beneath C Fortification Wall and stone cist burial to the south-east.

14

JOURNAL

OF PERSIAN

et al. 1993: 4, pl. Ia). While one grave does not make a cemetery, this stone cist burial may reflect the presence of additional mortuary remains in this area of the site, all of which may date to a period substantially later than the construction of Wall ib; i.e., to a time that was immediately pre-Urartian or Urartian. Operation

C1

Nearly the entire 1992 C1 trench was excavated down to sterile soil at the beginning of the 1993 season in the hopes of encountering additional collective Early Bronze tombs similar to that excavated in 1992. None were found, though materials recovered throughout the trench beneath the surface on which the B Fortification Wall was built were almost exclusively identified as Early Bronze Kura-Araxes wares. As mentioned above, Early Bronze ceramics also were uncovered beneath the Urartian architecture in the B2 complex immediately to the west, suggesting that this area of the Horom settlement, like the C3b/5 and E2 area to the south, was occupied during the Early Bronze period. Two discrepant C14 dates were obtained from the Early Bronze tomb in CI: sample no. AA-10191 BP or taken from burnt human bone-4505+/-50 calibrated at 3350-3050 B.C. (1 sigma; or 3360-2930 B.C. at 2 sigma); and sample no. AA-11130 which was charcoal from the burnt surface of the tomb--

HOROM

5150 +/-60 BP or calibrated at 4070-3810 B.C. (1 sigma; or 4220-3800 B.C. at 2 sigma). A late fourth millennium B.C. date for the typologically early forms of the three Kura-Araxes vessels from this tomb (P1. IIIc) is reasonable, particularly given the revised downdating for the beginnings of this "culture" (see Glumac and Anthony 1992: 203; Kavtaradze 1983); the discrepant, earlier date of the late fifth to early foutth millennium B.C. is problematic. Some of the Early Bronze ceramics recovered from the B2/C1 area, as well as many from the C3b/5 area (Fig. 14), exhibit incisions and surface decorations; they are considered typologically late and should indicate a much later mid to late third millennium Early Bronze occupation of the settlement as well. While more evidence of the Early Bronze period at Horom needs to be uncovered, particularly from undisturbed levels in different areas of the settlement, it is possible that the site may have been occupied continuously from the late fourth millennium B.C. onwards, the major difficulty, of course, being that the Iron Age levels either overlie and/or have strata containing the Early partially destroyed Bronze materials. It remains the case, however, that Early Bronze remains have been encountered beneath the Iron levels in all the operations at Horom, except for the excavations on the summit

E2/C3 AREA

1992-93

E2

STUDIES

--

-

C3

----

FORTIFICATION WALL C

REBUILD O

TOMB WALL lb

SECTION A -A

O Fig. 11. SchematicsectionthroughOperationC3b/5-E2.

1

2M

PRELIMINARY

REPORT

ON THE

1993

4

EXCAVATIONS

AT HOROM,

ARMENIA

6

5

9

7

8 10

11

12

16

17

20

14

13

18

15

19

21

22 23

24

25

2627

I I I I 12. "EarlyIron,"paintedMiddleBronze,and EarlyBronzeMixed GreyWareceramicsfrom OperationC3b/5 (number15Fig. wares). pattern burnished,numbers24-7 black-on-red

15

16

JOURNAL

1

OF PERSIAN

2_

STUDIES

-

_

C),

0

Fig. 13. Two lightgreypotsfrom E2 burial. of the South Hill, and this suggests that the occupation may have been fairly substantial, larger possibly than other known Kura-Araxes settlements, such as Karnut, on the Shirak Plain and, more generally, within highland Transcaucasia. The Early Bronze Settlement

at Anushavan

the Early Due to the difficulty of exposing Bronze settlement at Horom, a 6 x 6 m. sounding was opened near the end of the season at the neighboring site of Anushavan, which is located to the east of Horom immediately north of the town of Artik. Characteristic Kura-Araxes ceramics, serrated flint sickles, and andiron/hearth fragments picked surface reconnaissance initially in up during summer 1990 had indicated that an Early Bronze settlement once occupied a few hectares on the southern slope of a natural hill south-east of the modern Anushavan village and a later small Hellenistic site. More surface Early Bronze remains were retrieved by walking over the site in 1993, including a double spiral-headed copper or bronze toggle pin. The sounding was placed on the second of its three terraces. The cultural deposit was very shallow and filled with loose rocks. Nevertheless, a two course wide stone wall ran north-south across the trench, and a stone cist burial (Feature 2), reminiscent in form to the burial in E2, and containing the remains of an infant, had been placed immediately east of this wall (Fig. 15); the burial could not be dated since only one Early Bronze sherd was found beneath the skeleton, and it was unclear whether it related to the burial or to the underlying Early Bronze cultural level. Given the shallowness of the deposit, Anushavan may not prove to be the appropriate site to obtain more extensive exposure of Early Bronze remains in the Horom region, though a future sounding in a different area of the site is no doubt warranted.

THE SOUTH HILL EXCAVATIONS Two operations were conducted on the South Hill of the Horom settlement in 1993: 1) M7, which began as a 5 x 5 m. trench placed immediately northwest of the 1992 M1/6 "horizontal exposure" on the eastern summit of the South Hill (Badaljan et al. 1993: 4-6, fig. 3) and which was expanded several times to the north, south, and west as more architecture of this complex was uncovered, yielding by summer's end a total excavated area together with M1/6 of 26.5 m. (E/W) x 17 m. (N/S); and 2) M8, an 8 x 2.5 m. wide trench which was placed farther to the southwest near the bottom of the southern slope of the South Hill within what appeared to be a room in order to determine whether or not this occupation, which was beyond the fortification walls of the South Hill, was contemporaneous with the M1/7 complex on the summit or later in date. The M8 room was paved with flagstones, similar to those in the M1/7 area and in the D1 complex on the North Hill, but its walls were only preserved to a height of a single course; ceramics recovered from M8 were identical with those on the top of the South Hill, a circumstance which suggests that the occupations were, in fact, contemporaneous. Operation M1/7: The Architectural Complex on the Eastern Summit of the South Hill (Fig. 16) A large area of interconnected architecture x 17 m. has been 26.5 (c. m.) exposed on the eastern summit of the South Hill (Fig. 16). This complex now appears to continue even farther to the west, possibly extending all the way to what appears to be the main entrance through the southern fortification walls of the South Hill. The extensive flagstone flooring and lack of evidence for roofing uncovered during the 1992 M1 excavations implied that this complex may have been an open-air forum, and the presence of numerous hollowed-out

PRELIMINARY

REPORT

ON THE

1993

EXCAVATIONS

AT HOROM,

7

1

2

8

AsI

3

/k/

~t

4I

,oo

10

0000

0%"

-D.

11

Fig. 14. Incised Early Bronze Grey Waresfrom Operation C3b.

ARMENIA

17

18

JOURNAL

OF PERSIAN

stone pits or "basins" often in association with stone "tethering rings" chiselled out of adjacent rocks suggested that the complex may have been the scene of non-utilitarian, cultic activities, possibly including the sacrifice of animals (Badaljan et al. 1993: 5-6). The 1992 excavations uncovered four distinct areas within the part of the complex that was exposed; to and at least two more related areas-connected the area of the M1 excavations by a narrow corridor -were revealed in 1993. The cultural deposit on the South Hill is rela1 m. in depth. tively shallow, rarely exceeding Clearly only a single period of architecture is preserved on the top of the South Hill, though four separate loci were distinguished while clearing to the flagstone and packed clay floor of the architecture: 1) a c. 30-5 cm. deep topsoil level containing generally elaborate "Early Iron" grey wares with handles, spouts, and burnished designs; 2) a relasoil tively sterile deposit of loose homogeneous extending c. 30-70 cm. beneath the surface; 3) a culturally rich, dense mixture of clay, ash, and charcoal at a depth of c. 70-85 cm. directly above the flagstone floor; and 4) the flagstone and hard packed clay floor on which the walls and other features of the architectural complex were erected. This floor was occasionally cut through by pits which also seem to be associated with the architecture. Thus, even though it may prove possible to trace a development of the local grey wares from relatively undecorated to more elaborate forms on the South Hill, there is really evidence for only a single period of occupation. Two main walls were excavated in M7 during the 1993 season: a slightly curved eastern wall (Feature 1) which was 1 course thick and 4-5 courses high (max. height of 1.2 m.) (Pl. IIId); and a second wall (Feature 4) composed of large stones (up to 80 cm. wide) to its west, which was 2 courses wide but only 1 course high. It was initially thought that this more massive western wall might define the western limits of the entire interconnected complex, but a series of features, such as stone basins set directly into its western face, suggest that the complex continues still farther west beyond the limits of the 1993 excavations. Several features were found in the room defined by these two walls, including more pits and stone basins. One large basin, Feature 3 (P1. IIId), is particularly noteworthy. It was carved from a large triangular-shaped bedrock stone and has a diameter of c. 28 cm. Channels were cut into the bowl of the basin for drainage and a small lip or depressed spout also was carved into it presumably to facilitate the pouring of liquids. The channels drain into a large pit (Feature 5) set immediately to its west. A surprising number of finely-made bone artefacts (P1. IIIe) were recovered from this room, including

STUDIES

two points which were found within a small rectangular installation (Feature 13-165 cm. N/S x 70 cm. E/W) of vertically set stones south of the large basin together with a cache of forty-five ankle (astragali) and fifty-six metatarsal bones of cattle (c. 80 per cent) and red deer (c. 20 per cent), some of which had been perforated. The purpose of this installation is unclear, though the combination of features again suggests something other than simple domestic utilisation (the casting of lots, ritual divination, or the like?). Three additional basins were set into the western face of the large western wall (Feature 4) in association with a stone-lined hearth or area of burning (Feature 15).

ANUSHAVAN 1993 2M

0

_F3 F2

Fig. 15. Walland featuresfrom thesounding at theEarlyBronze site of Anushavan; notestone cist burial (F2).

BASINS

952

aa

16Q

O)108 on

1

41669 o

?)

67 0TETHERING

1669 43

00

A

HORO A

Fig. 16. M1/M7 Interconnected architecturalcomplex;stonerectangularinstallationin thesouth-westcontainedbonepoints and cacheof cattleand re metatarsalbones.

20

JOURNAL

OF PERSIAN

and associated features The architectural artefacts from the interconnected complex on the eastern summit of the South Hill are difficult to interpret. Present evidence neither confirms nor negates the hypothesis that animals may have been ritually sacrificed in this complex. Several of the bone artefacts found in Feature 13 and within the room defined by Features 1 and 4 show no evidence of use, possibly suggesting that they were fashioned within the complex. A large, worked red deer antler found south of Feature 13 also supports this interpretation. The proximity of several basins with one another, such as the three set in the western face of Feature 4, makes it difficult to conceive that all functioned as receptacles for animal sacrifice. For these, a more prosaic, industrial function may ultiThe complex mately seem more reasonable. remains enigmatic, and only additional clearing, possibly of the entire summit, may elucidate it. Tombs East of the Horom Settlement Numerous walls and surface features are spread across several hundred hectares to the east and south of the two citadel hills which define the Horom settlement. Most noteworthy and unequivocally prehistoric are circular rings of stones or cromlechs clustered in groups to the east and particularly south of the settlement. Two of these "Early Iron" tombs (labelled T1 and T2) were excavated in 1992 (Badaljan et al. 1993: 8-12), and four structures( T3-T6), thought to be tombs, were dug in 1993. For the purpose of defining chronological or social variety in the contents of these tombs, it was decided to sample structures from different clusters of tombs or different areas of the site. T4, which was defined by a circular ring of stones on the eastern ridge of the northern slope of the South Hill, proved to be natural and yielded no skeletal or material remains. T5, which appeared as a deliberately raised pile of stones and not a ringed circle, was located a few hundred meters to the northwest of T1 and T2. It was hoped that this structure might prove to be a kurgan, possibly earlier in date than the ubiquitous Iron Age cromlechs. Fragmentary human bones and potsherds were found in a pit beneath the pile of stones in T5, confirming indeed that it was a burial. These remains, however, were so in antisparse and badly disturbed-presumably quity, that they could not be dated. T3, a typically shaped cromlech defined by a ring of stones c. 6 m. in diameter, proved more informative, though its central burial pit had also been robbed in antiquity. T3 was located in a cluster of recognisable cromlechs west of T1 and T2 (i.e., closer to the South Hill) on the eastern slope of a small natural ridge. A large red basalt capstone had

STUDIES

been placed in the centre of the ring of the stones covering the central burial pit. After removing the surface vegetation and defining the ring of stones, two subsidiary rings of stones covered by smaller to the east (Feature A) and south capstones (Feature B) of the central burial pit became evident; Tomb 1, excavated in 1992, also had such an auxiliary ring of stones to its southeast (Badaljan et al. 1993: fig. 7) where it marked the presence of an additional pit containing ceramic vessels. The central capstone was split, and upon excavation it was immediately apparent that the central pit had been robbed; a few disarticulated human bones emerged, including part of the skull near the western edge of the pit. Parts of three vessels were recovered from the central pit along with nearlyl00 small beads, made primarily of carnelian, frit, and shell. Feature A contained six vessels, three of which were whole, and Feature B the partial remains of two additional vessels. Typologically, the vessels were similar to those found in T1 and T2 and presumably date to the late second or early first millennium B.C.; i.e., to the "Early Iron", pre-Urartian period (Fig. 17: 1-5). T6 was defined by a an oval ring of stones (c. 4.9 m. N/S x c. 5.5 m. E/W) clustered together with other stone-ringed cromlechs to the west of T3 on the eastern slope of the last ridge of hills separating this area of the Horom site from the South Hill. Three stones covered the central pit, two of which were removed. T6 was a cenotaph containing no skeletal remains but three vessels, two of which were complete (Fig. 17: 6-8), were recovered. The larger burnished dark grey complete vessel (Fig. 17:8) had a fragment of obsidian set in its base, a trait sometimes found on Middle to Late Bronze vessels in this vessel and the Transcaucasia; morphologically second complete vessel from T6 seemed earlier than the vessels recovered from other tombs at Horom, suggesting that this cenotaph (and others on this ridge of hills?) possibly date to the mid to late second millennium B.C. Faunal Remains from Horom Faunal remains from Horom are numerous and extremely well-preserved. Nearly 6000 animal bones from the 1992 and 1993 seasons were identified by Dr. Saindor Bokonyi. Eight domestic (cattle, sheep, goat, pig, horse, ass, dog, and chicken) and sixteen wild (red deer, roe deer, aurochs, Asiatic moufflon sheep, bezoar goat, wild boar, onager, badger, gazelle, red fox, hyena, brown bear, wolf, two wild birds, and one type of fish) species were represented in the Horom materials. 92.4 per cent of the identifiable remains were of domestic animals. The Shirak plain today is dominantly a cattle-breeding region, and this type of animal husbandry clearly

REPORT

PRELIMINARY

ON THE

1993

characterised the Bronze and Iron Ages as well; cattle constituted c. 63 per cent of the identifiable domestic forms. Caprovines were far less frequent (c. 26.5 per cent), and these were followed, somewhat surprisingly, by horse (6.7 per cent) and then pig (1.6 per cent). Red deer were by far the most numerous wild species, constituting at least 60 per cent of the hunted animals, followed by aurochs. The natural habitat for most of the wild animals is forest or forest-steppe. The large body and antler sizes of the red deer also suggest that these animals lived under favorable environmental conditions; which in dense woods grew probably specifically, throughout the plain at that time. It is noteworthy that the wild relatives of the domestic species whose 1

EXCAVATIONS

AT HOROM,

ARMENIA

21

origins go back to the Neolithic (namely, cattle, sheep, goat, pig, and dog) were indigenous to the region and still lived in the area during the Bronze and Iron Ages. Horse, however, would have to have been brought into highland Transcaucasia from elsewhere, most probably from the Eurasian steppes to the north, though it is not yet known when this momentous introduction first took place. As mentioned above, domesticated horse bones from "unmixed" Early Bronze levels in operation C3b/5 have been submitted for C14 determination, and their analysis will, hopefully, demonstrate the early presence of horses in the area. Dr. B6k6nyi also observed that some of the cattle horn cores showed a circular impression near the base, such as could

2

3

6

5

_8

Fig. 17. Ceramicvesselsfrom HoromTombsT3 (nos. 1-5) and T76(nos. 6-8); vesselno. 8 had an obsidianflake set in its base.

22

JOURNAL

OF PERSIAN

have been caused by the pressure of a yoke, a circumstance that would point to the use of draught oxen. The Horom Digital Mapping Project In summer 1993 a mapping project was initiated which began to incorporate detailed architectural drawings of the excavated areas and the visible surface architecture into a digitised map of the site compiled through the use of the AutoCAD pro-

STUDIES

gram. All the excavation units, including the extensive stepped fortifications uncovered in Operation A2/3 on the summit of the North Hill, were located on the base map and digitised as well. The aerial images will now be used to check and further refine the digitised model of the base map. We are now able to view the site and its architecture three dimensionally along several different, spatially significant scales of resolution. Work on this mapping project will continue in future seasons.

APPENDIX A: ARCHAEOLOGICAL INVESTIGATIONS IN SOUTHERN GEORGIA 1993 By Barbara Isaac, Zaal Kikodze, Philip L. Kohl, Giorgi Mindiashvili, Alexander Ordzhonikidze, and Greg White PALAEOLITHIC INVESTIGATIONS Three weeks were spent studying Palaeolithic and Plio-Pleistocene deposits in the field and State in the Georgian materials Palaeolithic localities Museum in Tbilisi. Four Plio-Pleistocene were surveyed: the Tsalka Basin; the Lake Paravani Plateau-Diliska Basin; the Paravani (Djavakheti) was established Plateau. It Persati and the Gorge; that sediments and volcanic rocks along the Diliska Gorge and under the Paravani Plateau offer ideal contexts for Plio-Pleistocene archaeological surveys. Mulitple extrusive rocks provide excellent dating potentials, while minor pyroclastic deposits should enhance correlation of fluvial and lacustrine facies. Sedimentary environments include floodplain and lacustrine littoral settings that would have been suitable for prehistoric occupations. Bone preservation is excellent, as may be that of plant macrofauna as well. Steep canyon and quarry exposures make exploration of these thick deposits feasible with minimal manual excavation of trenches. Abundant lithic artefacts were found on the Persati Plateau in and on fluvial-colluvial clays in this small elevated montane basin. Faunal materials probably are not preserved here, but this site of record an important to contain appears Acheulean technology, typology, and raw material transport. Geologically, areas with intact spatial patterning are quite possible. Trenching will be necessary to define the full character of the artefact assemblages and their geological context. Two caves were visited farther north in the area of Sachkere (Imeretia): Dzudzuana; and Ortvala. This area is strikingly similar to the Dordogne region of France and is characterised by extensive karstic activity within an uplifted Palaeozoic lime-

stone plateau. Incision of the plateau has exposed numerous caves and rockshelters. The Dzudzuana cave had only c.10 per cent of its fill removed by previous excavations. Sufficient exposures revealed the presence of Neolithic/Early Bronze materials that of the Palaeolithic. overlying Upper Interestingly, and for reasons which need to be clarified more extensive excavations, through no Middle Palaeolithic layers have yet been found. The Ortvala rockshelter is filled by sediments remarkably similar to the eboulis rich deposits which from French rockshelters. typify the sediments These are highly calcareous loams rich in angular limestone debris derived from the walls and roof. Most of the iboulis is intercalated with zones of fine, sediments dusty, and crumbly matrix-supported some hearths and ashes. The so-called containing "transitional" industry situated between the Upper and Middle Palaeolithic layers may be particularly important. Although it has been suggested that there is an unconformity between these industrial complexes, no evidence (e.g., soil or weathering) of such a break was observed in the field. Further exploration of both these caves should be highly informative, as should continued investigations of the earlier Plio-Pleistocene deposits further to the south. Both the Georgian State University and the Georgian State Museum have extended invitations for such collaborative field investigations, and these have been gladly accepted. BRONZE AND IRON AGE INVESTIGATIONS Short

field seasons were conducted on the of southern in 1990 Plateau, Djavakheti Georgia and 1991 (Kohl et al. 1992), a relatively level upland

PRELIMINARY

REPORT

ON THE

1993

stretching c. 40-50 km. west to east and km. north to south. The border between Georgia and Armenia represents the watershed between the tributaries of the Araxes river flowing south into Armenia and the Kura and its tributaries, particularly the deeply downcutting Paravani river, flowing north into Georgia. While the Plateau slopes northeast to southwest, its average elevation (c. 1800 m.a.s.l.) is higher than that of the Shirak Plain, and its climate accordingly is more severe and suitable for the pasturing of sheep and goats than for cereal growing and the herding of cattle, the dominant practises today and apparently during later prehistoric times on the Shirak Plain. The Djavakheti Plateau is surrounded by high mountain ranges, the most notable of which is the metalliferous Trialeti Range to its north. To date, the later prehistoric investigations in southern Georgia have concentrated on two sites: Satkhe, a site covering at least 7 ha. on the eastern edge of the Plateau; and Amagleba,9 a site near its northern boundary immediately south of the Trialeti Range. area 30-5

Early Bronze Excavations at Satkhe The site is located on a small hill above the contemporary village of Satkhe, c. 8 km. northeast of the regional center of Nino Tsminda (formerly Bogdanovka), near the confluence of the Paravani River and a small stream which runs through the village. The site was initially defined by a c. 4 m. thick wall formed by two rows of boulders (generally preserved to a height of only one course) separated by rubble infill. This wall follows the natural contour of the hill, securing a c. 6 ha. horshoe-shaped concave depression that forms the hill's summit. Early Bronze sherds were collected from within this fortified area and to the south and west beneath the contemporary village. In 1991 a late Early Bronze vaulted stone kurgan, which had been plundered in antiquity, was excavated in a field immediately to the north of the fortification wall, but two small exploratory soundings within the fortified area revealed the presence of Hellenistic and Medieval in addition to the dominant Early occupations The 1993 excavations were Bronze settlement: the extent and directed towards: 1) determining preservation of Early Bronze remains both within the fortified area (Operation Al) and on a sloping terrace at the foot of the hill beneath the fortifications to the southeast where a villager had unof decorated Early covered a dense concentration Bronze pots and storage jars at the end of the 1991 season (Operation Bi); 2) dating the wall by digging against its inner face and sectioning it A2 and A3); and excavating what (Operations appeared to be another small kurgan located on

EXCAVATIONS

AT HOROM,

ARMENIA

23

the remains of an early agricultural terrace immediately east of the fortified hill (Operation C1). This last excavation determined that this raised mound of stones had simply been collected by villagers in the course of agricultural work and was not a prehistoric kurgan that would help to date the extensive terracing east of the settlement. A 10 x 10 m. exposure (Al) was opened immediately to the northeast of where a large Kura-Araxes vessel, set on a plastered surface,10 had been recovered in 1991. Only Kura-Araxes materials were recovered from Al, although the cultural deposit which lay almost immediately beneath the surface was not well-preserved. Single courses of dry stone were structures walls, delineating rectangular revealed, and one such structure had been erected over an earlier structure, suggesting the presence of at least two distinct Early Bronze building phases in this area of the settlement. The black and brown burnished ceramics included a few incised decorated fragments, similar to those from Operations C1 anc C3b/5-E2 at Horom and other well-known Kura-Araxes sites, such as Shengavit, Mokhra Blur, and Kvatskhelebi, suggesting a date in the first half of the third millennium B.C. Unfortunately, the shallowness of the deposit and later cultural activities in this area, including its contemporary cultivation by villagers, meant that the Early Bronze remains here were badly disturbed and unlikely to yield additional significant information. Operations A2 and A3 were both small soundings set against and through the fortification wall for the purpose of determining the wall's date (Fig. 18). A2, which was located near the northeastern corner of the wall next to an internal feature or tower, contained a shallow Early Bronze deposit directly beneath the surface which continued partially underneath the wall, suggesting that at least in this part of the site an Early Bronze occupation predated the construction of the wall. A3 sectioned a 1 m. wide cut through the wall at a place where excavations in 1991 had revealed a dense concentration of Early Bronze artefacts underlying Medieval constructions against its inner face. A few unidentifiable, presumably Medieval sherds were recovered from the rubble fill of this trench, and the wall itself rested on bedrock (i.e., the Early Bronze deposit did not continue here beneath the wall). Thus, the results of these operations were It is possible that the wall initially inconclusive. could have been constructed during Early Bronze times, but it is equally, if not more, likely that it dates to a much later period, an interpretation its primitive form and which leaves unexplained of state preservation. poor Operation Bi, which began as a 10 x 10 m. trench, yielded a complex of Early Bronze domes-

24

JOURNAL.

OF PERSIAN

STUDIES

SATKHE 1993 TRENCHES A2 AND A3

A3

PLAN 0

1

VIRGIN SOIL A3 SECTION E-W

A3 SECTION N-S

BEDROCK

EBA 7/71/71

,r

,7/7-1,7;-77/

-/77"/

VIRGIN SOIL

SECTION E- W

A2

2I

I A2

PLAN

Fig. 18. Satkhe Soundings A2 and A3 against and through fortification wall of uncertain date.

2M

PRELIMINARY

REPORT

ON THE

1993

EXCAVATIONS

AT HOROM,

25

ARMENIA

ROOM D

B

MROOM

o

E

ROOM C

2ID

ROOM D

E

B.

PITHOS ROOMA

A-A PIT E.B.

B- B

ROOM D

SATKHE 1993

ROOM B

AREA BI 0

Fig. 19. Satkhe, Uperation •l-Early

Bronze domestic archztectural complex.

1

2M

26

JOURNAL

OF PERSIAN

tic, rectangular or sub-rectangular structures (Fig. 19). Although found directly beneath the surface, these rooms had been covered by a deposit of wash eroding from the hill and were less disturbed than those from Al. The rooms had well-defined floors preserved to a depth of nearly 2 m. A large, nearly complete black-burnished cooking vessel or storage jar was found on the floor near the northwestern corner of Room D; behind this several Kura-Araxes vessels and deer antlers had been placed (P1. IlIf). This evidence and other Early Bronze materials recovered from this excavation suggests that the rooms of this complex had been left while still containing usable items and that the site may have been suddenly abandoned, a feature also suggested for other Kura-Araxes settlements, such as Karnut on the Shirak Plain. There was also evidence of a Mediaeval midden deposit above Room C to the east, and today this terraced area of Early Bronze structures is being encroached upon and further disturbed by the expansion of the village. Some of the faunal materials from Satkhe were analysed by Dr. B6k6nyi, and these suggested that the Early Bronze inhabitants of Djavakheti practiced a different form of animal husbandry from to the south on the that of their contemporaries Shirak Plain. Specifically, ovicaprine remains slightthose of cattle, implying a far ly outnumbered greater reliance on sheep and goats on the higher Djavakheti Plateau. Horse bones (Equus caballus) also were recovered, as were remains of aurochs and red deer, although the wild animals were not as relatively numerous as at Horom, possibly suggesting that Djavakheti was not as densely wooded as Shirak. The Sounding at Amagleba The site of Amagleba sits on the northern edge of the Djavakheti plateau c. 25 km. north of Akhalkalaki, immediately to the southwest of Lake Tabatskuri, at the point where the Plateau first widens after descent from the Trialeti Range to the north. The latter contains rich metal deposits which are known to have been worked in late prehistoric times, and it is possible that Amagleba could have functioned as a transit station and/or production center for a trade in metals to the south. The site itself consists of a thin cultural deposit on the summit of a cone-shaped hill that rises c. 70 m. above the plain at a height of nearly 1900 m.a.s.1. and a thicker Bronze and Iron Age deposit with visible walling and possible fortifications spreading out from the southern base of this hill. A 3 x 3.5 m sounding was opened next to some of the visible architecture at the hill's base at the end of the 1993 season (Fig. 20). This sounding un-

STUDIES

covered the following stratigraphic sequence. The first 15 cm. were mixed, containing Medieval, Early Iron, and Early Bronze materials, and this mixture continued for another 55 cm., containing the foundation for a wall of undetermined Iron Age or Medieval date. Below this, a white surface was exposed containing Early Iron materials followed by several levels of undisturbed Early Bronze KuraAraxes deposit until bedrock was reached at 2.1 m. beneath the surface. The bottom of an Early Bronze stone wall was encountered at a depth of 1.3 m. and two additional Kura-Araxes living surfaces were exposed at 1.6 and 1.8 m. respectively. The Early Bronze materials included brown and black burnished wares, two of which were decorated with spiral volutes and incised triangles, a portable hearth or andiron fragments, a tanged brown obsidian arrowhead, and characteristic serrated red chert blades. A large and nearly 10 cm. long tanged stone spearhead was collected from the site's surface. Ceramic parallels again can be made to Mokhra Blur and Kvatskhelebi, as well as to the neighboring Early Bronze site of Amiranis Gora near Akhaltsikhe to the west. Faunal remains from the sounding were well-preserved. The single limited sounding at Amagleba essenfrom tially confirms what could be reconstructed the site's surface materials and reflects the basic known settlement pattern documented in highland Transcaucasia: a relatively thick and presumably long-lived Early Bronze occupation followed by a period of uncertain abandonment, beginning probably in the late third or early second millennium B.C., and then a reoccupation in the Late Bronze or Early Iron periods towards the end of the second millennium. The results suggest that the preservation of Early Bronze materials from Amagleba may be better than at Satkhe and Anushavan and much easier to recover than at Horom. On the basis of its location, the upper "Early Iron" levels at Amagleba may ultimately prove important for documenting a late second/early first millennium B.C. metals trade to the south when metallurgy in the Caucasus was a spectacular period of florescence experiencing and included the mass production of "Colchidean" tools and weapons (or "Koban") tin-bronze 1992: more extenAdditional, 275-95). (Chernykh sive excavations at Amagleba are clearly warranted and are indeed planned for the future.

CONCLUSION Limited investigations are scheduled for 1994. Already excavated materials from Horom will be studied and prepared for publication, possibly in conjunction with minor work on the site, involving

V

A

IV -

IV-

VI

+

TOPSOIL

\

4

STERILE SOIL

LIGHTBROWN

BROWN OF

CARBON

SOFTCHALKY

BROWN

PACKED REDDISH

BROWN PACKED

BRWNIV-

PACKED LIGHT M BROWN

FTTFF

I

V TRACES

I - EB-A FLOOR II- EB-A FLOOR

B

III- CULTURALBREAK BAND OF SOFT ORANGE-BROWN SOIL

2M

A

V- BAND OF WHITE SOIL, POSSIBLE BREAK BETWEEN EB-A AND LB- A

1

DEPOSIT VI-ENDOFCULTURAL

0

from initial sounding at Amagleba,northof Akhalkalakion thenorthernperiphery Fig. 20. Sectionand architecture of theDjavakhetiPlateau.

28

JOURNAL OF PERSIAN STUDIES

the completion of already initiated operations, such as those in the B1 North-West Gate area. We also hope to extend the coverage of our project in the Caucasus far to the northeast by opening a small test trench at the site of Velikent 2 on the Caspian littoral plain north of Derbent in Daghestan, Russia;

this exploratory work will be conducted with archaeologists from the Daghestan Scientific Centre of the Russian Academy of Sciences. IPARC will then resume its various investigations on a more extensive scale in 1995.

Bibliography

Kress Foundation in New York and from the Centre for Field Research (Earthwatch) and the Azadoutioun Foundation (Carolyn G. Mugar) in the Boston area. Support also was received from the Stahl Fund of the University of California, Berkeley, the University of Arizona, and Wellesley College. 2 Barbara Isaac, Assistant Director, Peabody Museum, Harvard University was supported by a grant from the Mary Leakey Foundation to continue her Palaeolithic investigations in southern Georgia in collaboration with Dr. Zaal Kikodze of the Georgian State University in Tbilisi. In addition to Ms. Isaac, Prof. O. Bar-Yosef, Dept. of Anthropology, Harvard University, Dr. Paul Goldberg, Texas Archaeological Research Laboratory, Prof. Reid Ferring, University of North Texas, and Dr. Anna Belfer Cohen, Dept. of Archaeology, Hebrew University, Israel visited Palaeolithic sites and studied the Pleistocene geology of southern Georgia. Dr. Bejan Tutberidze of the Georgian State University, Dr. Mehrab Tvalcrelidze, Institute of Geology, and Dr. Tengis Mechvelia, Georgian State Museum also took part in these Palaeolithic investigations. Future Palaeolithic work in Georgia will be supported both by the Georgian State University and the Georgian State Museum. Mr. Greg White, Ms. Nahrein Daniel, and Ms. Bevin Etheridge participated in the later prehistoric IPARC excavations at the sites of Satkhe and Amagleba. This latter work was also sponsored by the Georgian State University and by the Centre for Archaeological Investigations of the Georgian Academy of Sciences in Tbilisi. Drs. (;iorgi Mindiashvili and A. Ordzhonikidze from the Centre led a Georgian team of students from the State University to excavate Satkhe and Amagleba, and Mr. Levan Kalandarishvili was the architect/draftsperson for these excavations. :'The Air Force of the Republic of Armenia kindly allowed us to fly in one of their helicopters around the North and South Hills of the Horom settlement. Indeed this same area may have also begun to provide evidence for an Intermediate Phase of construction. In this context it is still necessary to continue our excavation of a particular wall, Wall 8, which is not yet definitely associated with either the first, or the last, phase of local construction. 'The extent to which walls other than ground floor walls were also made of stone is not, of course, known. To judge from B1 NorthWest Gate area, the B Fortification Wall's four m. high stone socle was capped by mud-brick (Badaljan et al. 1993: pl. IIb). "Note, however, that the 'jog" in the north face of Wall 4 begins at a point that is 2.70 m., not 3.70 m. (for which see Badaljan et al. 1993: fig. 13), from the inner face of the B Fortification Wall. 7 Cf. C. F. A. Schaeffer, Stratigraphiecompareeet chronologiede l'Asie occidentale,III et II millenaires,London, 1948, figs. 273, 274, 282, and 283. Compare also the stone mold for producing such bronze arrowheads which was found in the Early Iron settlement at the site of Dvin on the Ararat plain to the south (K. Kh. Kushnareva, DrevneishiePamyatnikiDvina, Yerevan, 1977, p. 31,

Badaljan, R. S., Edens, C., Kohl, P. L. and Tonikjan, A. V. 1992 "Archaeological Investigations at Horom in the Shirak Plain of Northwestern Armenia, 1990," Iran, XXX, pp. 31-48. Badaljan, R. S., Edens, C., Gorny, R., Kohl, P. L., Stronach, D., Tonikjan, A. V., Hamayakjan, S., Mandrikjan, S., and Zardarjan, M. 1993. "Preliminary Report on the 1992 Excavations at Horom, Armenia," Iran, XXXI, pp. 1-24. Burney, C. A. 1966 "AFirst Season of Excavations at the Urartian Citadel of Kayalidere,"AS 16, pp. 55-111. Chernykh, E. N. 1992. Ancient Metallurgyin the USSR: The Early MetalAge. Cambridge. Glumac, P. and Anthony, D. 1992. "Culture and Environment in the Prehistoric Caucasus: The Neolithic through the Early Bronze Age," in RelativeChronologiesin Old WorldArchaeology,ed. R. W. Ehrich, 3rd ed., Chicago, pp. 196-206. Kavtaradze, G. 1983. K khronologiiepokhi eneolita i bronzi Gruzii. Tbilisi, Metsniereva. Kleiss, W. et al. 1979. Bastam I. Ausgrabungen in den Urartiischen Anlagen 1972-5, Teheraner Forschungen IV, Berlin. Kohl, P. L., Edens, C., Pearce, J. and Carson, E. A. 1992. "International Program for Anthropological Research in the Caucasus: Field Seasons 1990 and 1991," Bulletin of the Asia Institute,NS 6, pp. 143-50. 'Participants in the 1993 field season at Horom on the American side included the following students: Laura Tedesco from New York University; Kim Codella, Eleanor Barbanes, Sanjyot Mehendale, and Nahrain Daniel from the University of California, Berkeley; Adam Smith, Amy Tucker, and Doug Gann from the University of Arizona, Tucson; Armine Ishkanian from the University of California, San Diego; Greg White from Cambridge University in England; and Bevin Etheridge from Wellesley College. Mr. Gann, Mr. Smith, and Mr. Codella were also responsible for the digital mapping project described in this report. Dr. Karen Rubinson, Research Associate of the University Museum, University of Pennsylvania directed the 1993 excavations of the Late Bronze/Early Iron Age cemetery at Horom. Dr. Sdindor B6konyi, Director of the Institute of Archaeology in Budapest, Hungary, analysed the faunal remains, and Christopher Rasmussen from Berkeley, California was the excavation photographer. The help of all the above is gratefully acknowledged. Ms. Eleanor Barbanes was responsible for drafting the architectural plans and assembling all the line drawings for this article; her invaluable assistance must be especially mentioned. The original ceramic drawings were prepared in the field by Ms. Anna Dokhikjan. The Armenian participants included the following archaeologists from the Institute of Archaeology in Yerevan: Simone Hamayakjan, Sergei Mandrikjan, and Mrktich Zardarjan. We also wish to acknowledge the invaluable support of the Institute's Director, Dr. Aram Kalantarjan. The 1993 season could not have been conducted without the logistical and organisational assistance rendered by Vladmir Kostandjan and Ruben Tonikjan of the Shirak Foundation. Both the Institute of Archaeology in Yerevan and the private Shirak Foundation supported the excavations at Horom. The American work was sponsored by a matching grant from the National Endowment for the Humanities (RO-2252392), and additional support was received from the Wenner-Gren Foundation for Anthropological Research and the Samuel H.

fig. 45).

"This wall, here called Wall 1b, was referred to as Feature 5 in the report of the 1992 excavations (Badaljan et al. 1993: 20). "This site was incorrectly referred to as Amareleba in the 1992 report. "'A charcoal sample was obtained from beneath the vessel lying on this surface. It yielded a radiocarbon date of 4380 +/- 45 BP (AA-No. 7768) or 3090-2921 3293-2912 B.C. at 2 sigma).

B.C. calibrated

(at 1 sigma; or

29

JOURNAL OF PERSIAN STUDIES

Addendum The following 17 radiobarbon determinations of samples taken during the 1993 excavations were received from the AMS Facility at the University of Arizona after this article was prepared for publication: Provenience Calibrated age BC C14 age (BP) I.D number 1 sigma, 2 sigma Satkhe B1, loc. 10 AA-12853 3343-3043, 3365-2925 4500+/-60 Satkhe B1, loc. 2, pit sq, Room A Horom C3b/5 loc. 11 lev. II - horse bone

4445+/-60

3301-2926, 3345-2914

3035+/-55

1387-1204, 1413-1117

AA-12856

Horom C3b/5 loc. 6

3285+/-55

1620-1511, 1682-1425

AA-12857

Horom C3b/5 loc. 9

3190+/-55

1515-1407, 1591-1319

AA-12858

Horom M7, loc. 12, fea. 15, burned bone

3045+/-75

1401-1139, 1440-1038

AA-12859

Horom M7, C, loc. 6 fea. 6

3145+/-55

1444-1324, 1517-1265

AA-12860

Horom B2.4, loc. 10

2850+/-55

1110-918, 1190-847

AA-12861

Horom B2.4, loc. 16, surf. C

2485+/-50

770-423, 794-403

AA-12862

2480+/-55

770-416, 795-401

AA-12863

Horom B2.4, surf. A2 (so. baulk ext.) Horom B2.4, fea. 7

2540+/-55

796-543, 808-419

AA-12864

Horom B2.4, loc. 17

3285+/-55

1620-1511, 1682-1511

AA-12865

Horom B2.4, loc. 6 fea. 4

2410+/-55

752-398, 767-384

AA-12854 AA-12855

AA-12866

Horom B2.3, loc. 6

2570+/-55

802-605, 819-525

AA-12867

Horom B2.5, loc. 4 burnt bone

2490+/-60

779-420, 799-401

AA-12868

Horom D1, room 3 (northwest corner)

2520+/-55

792-529, 804-408

AA-12869

Horom D1, rm. 2, loc. 3 (west of fea. 4)

2485+/-55

773-420, 796-402

ELAMITES AND OTHER PEOPLES FROM IRAN AND THE PERSIAN GULF REGION IN EARLY MESOPOTAMIAN SOURCES* By Ran Zadok Tel-Aviv University

1. ELAMITES AND INDIVIDUALS BEARING ELAMITE NAMES IN MESOPOTAMIA

This section does not include a complete prosopography of the Elamites in Mesopotamia, but I apply here a "maximum" approach (cf. Iran XXV, 1987, p. 1). Every name is followed by at least one prosopographical number (for full documentation of homonyms see Table 2). The total number of the individuals who were Elamites and/or bore assured or hypothetical Elamite names in Sumer during the Ur III period is 340 (maximum; minimum: 269). Since all the persons are mentioned within 52 years (2072-2020 B.C.) it is possible that

most, if not all, the homonymous individuals who lived or were active in the same place were physically identical unless there is evidence to the contrary. Therefore, an estimate between the maximum and the minimum number, but closer to the latter, would seem more likely. Admittedly, only a comprehensive study of the various Ur III archives would give more accurate statistics. Nevertheless, I hardly expect a radical change of my results. The main purpose of this study is to establish criteria for identifying Elamites in the huge and ever-increasing Ur III documentation. Only 77 (maximum; minimum 72 as 1.3.1.17.27. 97-99 are anonymous [Elam(mu)] and 1.3.1.66; 1.3.2.9 are homonymous) individuals are defined as Elamites (75 as Elam or dumu and two as

*Abbreviations as in A. L. Oppenheim et al (eds.), The Assyrian Dictionary of the Oriental Institute of the University of Chicago (Chicago-Gliickstadt, 1956-), as well as R. Borger, Handbuchder Keilschriftliteratur1.2 (Berlin, 1967 and 1975 resp.), except for the following: ARET = Archivi reali de Ebia. Testi;AUCT = M. Sigrist, Neo-Sumerian Account Texts in the Horn Archaeological Museum (Berrien Springs [Mich.], 1985 f.); BNF = Beitrage zur Namenforschung,CSAU = D. C. Katz, A ComputerizedStudy of the Aga-tisof the Ur III period (Minneapolis, 1979); CTMMA 1 = I. Spar (ed.), Tablets,ConesandBricksof the Thirdand SecondMillennia B.C. CuneiformTextsin theMetropolitanMuseumofArt 1 (New York, 1988); DAS = B. Lafont, Documentsadministratifssumiriens (Paris, 1985); DCEP = J.-M. Durand, Documentscun?iformesde la IVe sectionde l'EcolePratiquedesHautes Etudes 1 (Geneva-Paris, 1982); Di Vito, Diss. = R. A. Di Vito, Studies in Third Millennium Sumerianand AkkadianOnomastics:TheDesignationand Conception of the Personal God (Harvard Dissertation; Cambridge [Mass.], 1986); Eb. 1975-85 = L. Cagni (ed.) Ebla 1975-1985: Dieci anni di studi linguistici efilologici. Atti del convegnointernazionale(Napoli, 9-11 ottobre1985; Naples 1987); ElWb = W. Hinz and H. Koch, ElamischesWorterbuch(Berlin, 1987); Englund, Fischerei = R. K. Englund, Organisationund Verwaltungder Ur-IIIFischerei(Berlin, 1990); EO = R. Zadok, the Elamite Onomasticon(Naples, 1984); FAOS = FreiburgerAltorientalische Studien;FordeandFlaugher = N. W. Forde and W. R. Flaugher, Neo-Sumerian Textsfrom South Dakota University,Luther and Union Colleges(Lawrence, Kansas, 1987); Fs. Lacheman = M. A. Morrison and D. I. Owen (eds.), Studies on the Civilization and Cultureof Nuzi and the Hurrians in Honorof E. R. Lachemanon his SeventyfifthBirthday,April 29, 1981 (Winona Lake, 1981); HE = E. Carter and M. W. Stolper, Elam: SurveysofPolitical HistoryandArchaeology(Berkeley, 1983); Huber, PNN = E. Huber, Die Personennamenin denKeilschrifturkunden aus der Zeit der Kinige von Ur und Nisin (Leipzig, 1907); ITT2 = B. Lafont and F. Yildiz, Tablettescuneiformesde Telloau Musie d'Istanbul datant de l'Ppoquede la HfP dynastie d'Ur (Leiden, 1989); Kutscher, Wadsworth= R. Kutscher, Neo-Sumerian Tablets in the Wadsworth Atheneum. WadsworthAtheneum Bulletin, 6th

series, vol. 6/2 (1970), pp. 41-64; Ladders to Heaven = O. W. Muscarella (ed.), Laddersto Heaven (Toronto, 1982); Mil. Steve = L. de Meyer, H. Gasche and F. Vallat, Fragmenta Historiae Elamicae.MelangesoffertsciM.J. Steve (Paris, 1986); MessT = R. C. McNeil, The "MessengerTexts"of the Third Ur Dynasty (Doctoral Dissertation, University of Pennsylvania, 1970; Ann Arbor, 1971); MTBM = M. Sigrist, Messenger Texts from the British Museum (Potomac, 1990); NATN = D. I. Owen, Neo-Sumerian Archival Texts Primarily from Nippur (Winona Lake, 1984); Sigrist, Princeton = M. Sigrist, Tablettesdu Princeton Theological Seminaryipoque d'Ur III (Philadelphia, 1990); Sigrist, Syracuse = M. Sigrist, Textes economiques neo-sumiriens de l'Universite de Syracuse(Paris, 1983); SNSA = T. Gomi and S. Sato, SelectedNeoSumerian Administrative Texts from the British Museum (Chiba, 1990); STTI = V. Donbaz and B. R. Foster Sargonic Textsfrom Telloh in the IstanbulArchaeologicalMuseum (Philadelphia, 1982); TAD = S. H. Langdon, Tabletsfrom the Archivesof Drehem(Paris, 1911); TE = Ch. Virolleaud and M. Lambert, Tablettesiconomiquesde Lagas'1 (Paris, 1968); TJAMC = E. Szlechter, Tablettes juridiques et administrativesde la 111H dynastied'Ur (Paris, 1963); TS = H. Limet, Textes sumriens de la 1f dynastie d'Ur (Brussels, 1976); und. = undated; UNL = G. Pettinato, Untersuchungen zur neusumerischenLandwirtschaft 1: die Felder (Naples, 1967); USP = B. R. Foster, Umma in the Sargonic Period (Hamden, 1982); Watson, Cat. = P. J. Watson, Catalogue of Cuneiform Tabletsin BirminghamCityMuseum, 1: Neo-Sumerian Texts from Drehem with Some copies by W. B. Horowitz (Warminster, aus der Zeit der 1986); WMAH = H. Sauren, Wirtschaftsurkunden III. Dynastie von Ur im Besitz des Musded'Art et d'Histoire in Genf (Naples, 1969); Yang, Adab = Zhi Yang, Sargonic Inscriptions from Adab (Changchun, 1989). The preparation of this study was supported by the Foundation of Basic Research of the Israel Academy of Sciences and Humanities. I should like to thank the Trustees of the British Museum for permission to quote from several unpublished tablets (courtesy Prof. M. Sigrist, cf. 1.3.1, 1.4) and Prof. D. I. Owen, who kindly showed me the MS of his MVN 15 ahead of publication.

1.0 Introduction

31

32

JOURNAL

STUDIES

OF PERSIAN

TABLE

1

Explicit Elamites and Elamites accordingto onomasticcriteria (with various degreesof plausibility; minimum amounts in brackets;t = total) 1

2

3

4

5

6

7

8

9

10

11

12

place

explicit

t

assured

t

probable

t

possible

t

dubious

t

total

Lagash (1.3.1)

2.6.10.17.18.22. 24.27.29-40.44. 50.59-79.82.83. 86.97-100.102. 121.122.125. 129.133

Umma (1.3.2)

9.17.25.31. 38-40

5.60.89.100

12.13.32.56.84. 85.88

12 1.11

49(44)

100(75)

6

11

7(6)

9(8) 13

1-3.7.8.10

4.5.9.11

13

6

69(52)

2-6.8.9

1.7 2

unknown 12 (1.3.6) Grand total

33(32)

2.4.6.7.9.10 1

1

2

1

28(9) 5

8

138(120)

1.3.4.9.11. 16-18.20.33.36. 39-43.45. 47-49.51.52.54. 58.59.61.63.65. 68.69.71-77.79. 80.83.86.87. 90-93.95.96.99

2.6-8.10.14.15. 19.21.33.34.37. 38.44.46.50.53. 55.62.64.66.67. 70.78.81.82.97. 98 4

7(6)

41(36) 2-4.7.10-13.15. 16.19.21.23.26. 27.36.37.43. 45-48.50.53.56. 58.60.61.63.65. 67

19(7)

6(5)

4(2)

7(6) Puzrish- 22-31.57.94 Dagan (1.3.3)

22(18) 5.14.20.24. 28-30.32.33.35. 41.44.49.57.62. 64.66.68.69

1.6.18.34.42.59

8.22.51.52

4.5.7.12.15.21. 23.26.42.45.49. 51.54.56.80.81. 84.85.87.91-96. 109-116.119. 124.126-128. 131.132.137

11.13.14.16.20. 43.46.52.53.55. 90.101. 103-108.117. 118.120.130

13(12)

6(2)

56(52)

Nippur 3 (1.3.4) Ur (1.3.5)

1.3.25.28.41.48. 57.58.88.89. 134.136.138

8.9.19.47.123. 135

1

1

1

4

6

77(72)

20(13)

25(23)

76(41)

142(130) 340(279)

NOTES: Percentage: Lagash: 40.58(40.89): Umma: 20.29(19.33); PD: 29.41(27.88); Nippur: 3.23(4.08); Ur: 2.64(2.97); unknown: 3.82(4.83). Explicit: 22.64(26.76); assured: 5.88(4.83); probable: 7.35(8.55); possible: 22.35(11.52); dubious: 41.76(48.32). concerning the minimum calculation (nos. refer to columns). Col. 3: Deduct 1.3.1.27.97-99; 1.3.2.9. Col. 5: Deduct 1.3.1.8.19.47.135; 1.3.2.52; 1.3.3.88. Col. 7: Deduct 1.3.1.25 (homonym of 1.3.2.1); 1.3.2.59.

1.3.2. Col. 9: Deduct 1.3.1.14.16.20.43.46.52.53.103-108.130; 1.3.3.2.7.14.21.33.34.37.38.44. 28-30.32.33.35.41.45.49.57.64.66; 46.55.62.66.67.70.78.81.82.98. Col. 11: Deduct 1.3.1.21.56.95.96.115; 1.3.2.44; 1.3.3.65.68; 1.3. 6.10.79.91.92.96; 1.3.5.8. Col. 12: In theory, the minimum may be further reduced to 236, by regarding all the homonymous individuals as physically identical disregarding their place (cf. Table 2), but this does not seem realistic.

Susians; Table no. 1, col. 2). But since Greater Elam was not in all probability ethnoliguistically homogeneous, only a minority of their names is explicable in Elamite terms (6 ? 8: 1.3.1.44; 1.3.2.25.40; 1.3.3.22.23.28 ? 1.3.1.2.6.30; 1.3.3.24-27; 1.3.4.3). The remainder are atypical or too short (17: 1.3. 1.10.18.24.31.32.60.78.79.86.125.133; 1.3.2.31.39; 1. 3. 3. 29-31; 1.3.6.12; many explicable in Elamite terms), Sumerian (20 ? 2: 1.3.1.29.35.36.50.62.66. 69,71-77.82.83.122; 1.3.2.9.17.38 and perhaps 1.3. 1.67.100), Akkadian (4: 1.3.1.22.34.68.121), hybrid (Sumero-Akkadian, 6: 1. 3. 1. 37. 59. 61. 64. 65. 70),

Hurrian (1: 1.3.1.129), unexplained (5: 1.3.1.33.63. 102; 1.3.3.57.94) and broken (3: 1.3.1.38-40). This is, of course, a very weak argument against their Elamite extraction. Several individuals bearing Akkadian names might have originated from Susiana which was heavily Semitised. The Sumero (-Akkadian) names may be due to cultural-political influence or considerations of prestige. All the remaining individuals listed in Table 1 are considered Elamite with various degrees of plausibility due to onomastic criteria (Table 1, cols. 4.6.8.10). More or less assured Elamite names which need

ELAMITES

AND

TABLE

OTHER

FROM

PEOPLES

Name

ProsopographicalNumbers Total

1. A-da-ra,Ad-da-ri 2 1.3.1.80; 1.3.2.27 2. Ad-da-na-pir/pir,/pi/pi-ir 1.3.1.8.9.19.47.135; 1.3.2.8 9 ? 1

Adda-na-(pi/pi)-ir A-da-na-pi-ir 3. Agu/ku-ni 4. D/Tan-kal-la

51.52 + 43; 1.3.6.6 1.3.1.81; 1.3.4.2; 1.3.6.3

3

1.3.1.25; 1.3.2.1

2

5. D/Tan-ne-ki. 1.3.1.91; 1.3.2.3; 1.3.5.2 D/Tan-ni-gi4 6. D/Tanzir-ri,D/Tan-nu-ri 1.3.1.93; 1.3.2.47

3 2

7. E-lak-ra-(at),E-la-kdrlqar 1.3.1.94-96

3

8. E-lam(-mu) 9. EI/I-puga-ru

1.3.1.17.27.97-99

5

1.3.1.4; 1.3.3.11; 1.3.6.1.8

4

10. E-ze-me-na/ni

1.3.2.54;1.3.5.5

2

11. Ga-an-za

1.3.1.101; 1.3.3.50.78

2

12. Hu-ba(-a),U/U1,-ba-al Hu-un-ba-a

1.3.1.11.14.16.20.43.46.52. 53.103-108.130; 1.3.2.20. 28-30.32.33.41.45.57.64.66; 1.3.3.2.7.14.21.34.38.44. 46.55.62.66.67.70.81.82;

1.3.6.9.11 13. Hu-ba-na/an

43

1.3.1.(1 la). 135; (1.3.2.4a);

2(!)

14. Hu-ba-ni/ni-il

1.3.1.57; 1.3.2.34.59

3

15. Hu-hu-ni

1.3.1.55; 1.3.2.24.35.49

4

16. Hu-nu-nu-ur/Hu(-un)-nu-ri 1.3.3.74; 1.3.4.10

2

17. (Hu-)un-du-du

1.3.2.16.44

2

18. Hu-un-ha-al-bi-it/da

1.3.3.8.19

2

19. Hu-un-ha-ap-ur

1.3.3.15.33.37.98

4

20. Hu-un-hu-up-ls

1.3.5.7; 1.3.6.4

2

21. Hu-(un-)na-zi

1.3.3.77; 1.3.6.10

2

22. Hu-undlul-gi

1.3.2.61; 1.3.3.47.68.91.92 1.3.1.18.56

5

24. Ig4-ri/ru, I-gi4-ru-um(?) 1.3.3.20 ? 41 1.3.1.15.21.115 25. I-lam-ma

2 1? 1 3

26. In-da(-a)

1.3.1.54; 1.3.3.4

27. Kur/Kitr-da-lu

1.3.2.21; 1.3.3.79; 1.3.4.7;

1.3.5.3.8

5

28. Mu-ir-ti/ti(n)-ga-ba, Muiir-duo-gab 29. Se-ilx-ha

1.3.3.51.65.96

3

1.3.3.32.88

2

30. dsulgidan/da-angada

1.3.1.26;1.3.3.59

2

31. PI/U-li, U-e-li

1.3.1.32.131; 1.3.2.19;

1.3.3.36

4

32. Ur-ni9-gar 33. Za-na

1.3.1.66; 1.3.2.9

2

1.3.1.138; 1.3.2.42; 1.3.4.8

3

Grand Total Compare 1.3.2.46 with 1.3.3.16.

THE

PERSIAN

GULF

33

REGION

final component is correct-Si-im-ti-"gusir" (1.3.4.5; = Simti-kusir, cf. Te-ep-ku-ti-ir, EO, nos. 119.246).1

Probable Elamite names are e.g., 1.3.2.18, 1.3.1.42 (cf. 1.3.2.69 and perhaps 1.3.3.40; definitely not Akkadian., cf., hesitantly, Gelb, MAD 3, p. 125, s.v. "IHBR?"), La-al-,

Zu-zu-gu-ni

and

Pu-ul-ma

(1.2;

1.3.3.10; 1.3.4.11; see SEL 8, 1991, pp. 231ff.). Ad-dana-nam (1.3.1.1), which begins with atta (EO, no. 18), ends in the equivalent of the Marhashite's name Nana-ma (PDT 529, r.x, 32). Is its resemblance to the second component

of Lii-na-nam/L-na-na-mu

(listed

without interpretation in Limit, Anthroponymie,p. 483) just incidental? The segmentation of many of the following names is open to doubt. Moreover, the apparent presence of elements with Elamite parallels does not necessarily assure the Elamite character of a name as in most cases it cannot be determined whether the assumed compound is syntactically (1.3.1.4; 1.3.3.11; 1.3.6.1.8) permissible. EV/IU-puga-ru is tentatively included here in view of the resemblance of E/I~-pu- to Sarg. E?4-pum/ball (cf. SEL 8, p.

1.3.5.6

23. Hu-ru

AND

further comment are Am-ma-za-za(1.3.5.1; fem. EO, nos. 7.290; if not Sum.) and-if the reading of the

2

Homonymouspersons in general No.

IRAN

2

137 ? 2

226). Kir-dis (1.3.2.62) is on the face of it a perfect forerunner of the name of the Achaemenid dynasts who lived almost a millennium and a half later. Their name is in F. C. Andreas's view originally Elamite, but some other Iranists are of the opinion that the Achaemenids' name is of Indo-Iranian origin.2 In the latter case, the perfect resemblance of this short name would be just a coincidence. Silhak is represented by Se-il-ha-ki-li-ba-as'(1.3.3.27; cf. SEL 8, p. 231) and Si-il-ha-gi (1.6); Se-ill(ILxKAR)-ha

(1.3.3.32.88; not Akkad.) is related. Pu-ra-an-ha-al-bi-it (1.3.3.27) may end with the same element(s) as Huun-ha-al-bi-it(1.3.3.8). For the first component cf. perhaps

Pu-ra-an-hu-ut-ir-ra-an

(1.3.3.99)

whose

second component reminds one of the DN Hutran (with anaptyxis as NA d(j-du-ra-an, EO, no. 58?). A-da-

me-na (1.3.2.14) may be interpreted as atta (with a defective spelling, cf. EO, no. 18) and -min(a)(EO, no. 144), although Gelb hesitantly derived the second component from Akkadian (MAD 3, p. 179; cf. 1.3.2.54; 1.3.5.5; to izi, EO no. 85?). Not every name which apparently contains a-ta- (e.g. 1.3.1.30) is Elamite. Do 1.3.3.7, 1.3.4.9 and 1.3.1.5 contain ta-s (EO, no. 236c)? Does Da-gu-du-sa6 (1.3.2.53) end in tus (EO, no. 252a)? Ba-ak-za-na(1.2), which is followed by gemethereby referring to a female, probably ends in zana (EO, no 287). For Mu-mu-gi (1.2) compare perhaps RAE Mu-me-ig (ElWb. p. 951: Fort. 6043,4). Ha-ap-hi-if (1.2) possibly consists of hap and hil (EO, nos. 30.42); Ba-ar-li-ni (1.2; par + lini, cf. EO, nos. 175.205?). Ku-ki-ni/nidu/dulo (1.3.4.5a) looks as if it

begins with Elam. kuk-, but the remainder of the name has no Elamite counterpart (cf. Ku-ku-ni-ti, BNF NF 18, 1983, p. 166: no. 2362?).

34

JOURNAL

OF PERSIAN

The same applies to Ku-ku-da-i_(1.3.5.4). Ku-ku-alum (1.3.1.119) may be purely Akkadian (SEL8, p. 234; in OB Zi-im-ri-ku-ku,Greengus, Ischali, 264 r. 4f., it is preceded by an Amorite predicate). Do Gu4-ka-ri (1.3.1.2) begin with kukand Ur-ku-ku(1.3.3.76) end in it? For 1.3.3.60 cf. NABU 1990/39. Hu-ba-na (1.3.1. 134) is hardly a Sumerian anthroponym (as apparently understood by H. Sauren, WMAH, p. 311a, s.v.), but is Elam. Humpan (DN used as PN; cf. perhaps 1.3.1.58.81 with EO, nos. 8.61). Is Hu-ba-la EO, p. (1.3.1.109) a variant thereof? (cf. -hu-um-ba-al, 13: no. 48). Hu-un-ba-a(1.3.2.32) is probably a hypocoristicon of Humpan. Hu-ba(-a,e.g., 1.3.1.11.46), Uba-a (1.3.2.29) and U18(GISGAL)-ba-a (1.3.2.20) may also belong here. However, the name is the commonest one in the documentation under consideration: 43 individuals out of over 250 persons bearing Elamite names.3 Therefore it is probable that not all the individuals bearing this name were actually Elamites, the more so since hup- is also thought to be hypothetically Hurrian (cf. NPN, p. 218) and homonymous names are recorded in early sources from regions where one can hardly expect any Elamites. Such homonyms are Hu-ba-ANfrom Ebla. It can be compared with, e.g., U-ba-AN/nmn/lum (ARET 3, index, s.vv. with references; possibly Sem.), and Hu-un-bafrom MB Emar (RPA 6, 217,21). Hu-ba-riga/gl (ARET 3, p. 275 with refs.) from Ebla may also belong to a non-Elamite milieu, the more so since a component -rikV is not recorded in the Elamite onomasticon. U-ma-ni(Anthroponymie,p. 534, s.v.) seems to be Mesopotamian rather than Elam. Humpan.The fact that a certain U/U18-ba-a(etc.) was an ensi of Adamdun (RGTC 2, p. 4) may be in favour of an Elamite derivation of the name in question, but although many of the provincial civilian governors originated there (cf. P. Michalowski, SAOC46, p. 58), it cannot be demonstrated that ensis of certain peripheral regions always bore names indigenous of those regions. Moreover, there is very slight evidence to the contrary, viz. at least one Mesopotamian local high functionary bearing a non-Mesopotamian name (information of Prof. D. Owen who is due to publish the evidence). It is doubtful whether Hum/Lum-ba/ma(FAOS 15/1, 20, v, 6; pre-Sarg.), Hum/Lum-ba/ma-a(Huber, PNN, p. 124a) and the initial component of Hum/Lum-ba-ru(1.3.1.12) have anything to do with Humpan. The same applies to that of Hu-up-ni-ki-te-er-ra (recte-up-(pa)-with omission due to the fact that -up- is written at the line's end? SA 7,3) whose segmentation is not certain (cf. RGTC 2, p. 76, s.v. Hubni). In Goetze's opinion (JNES 12, 1953, p. 123), 1.3.1.129, who is explicitly described as an Elamite, had a Hurrian name. Is it a coincidence that 1.3.1.6 (explicitly Elam.) looks identical with the second component of Kass. Karziapku (Kass. St., p. 60; hypothetical segmentation as yapku seems to be equally possible)?

STUDIES

For Ga-an-za(1.3.1.101; 1.3.3.50.78) cf. Nap-ga-anza (Iran XXV, p. 3). Hu-hu-ni (e.g., 1.3.1.55) may render Elam. "fortress" or sim. (differently Limet, Anthroponymie,p. 111); cf. Hu-uih-NE(1.3.2.67)? Hu-ihu-' (Sigrist, Syracuse, 479, 16), Hu-hu-hu (see A. Falkenstein, NSGU 2, p. 318 ad ITT 3/2, 6563), Kuku-e (MVN 7, 140), Hu-ba-a (AnOr 1, 115, 3) and La(a-)ni (Limet, Anthroponymie,p. 447) are rather atypical like Hum/Lum-ka-kawhich is listed after 1.3.1.118 (cf. Za-an-ka-ka,STA). The same may apply to Ba-ar, Da-an, Gu-ri,Hu-un (1.2), Gui-gui-ni, Ku-gu-nu (1.3.1.49.135? cf. EO, no. 111?), 1.3.1.18, 1.3.2.23, 1.3.3.75 (cf. EO, nos. 53f. 70a.198.200), 1.3.6.12, U-eli (1.2; 1.3.1.131; homonymous with PI-li, 1.3.1.32; cf. DCEP, p. 67, n. 41 ad U-li [1.3.2.19]; not necessarily Akkad., MAD 3, p. 31; cf. U-i-li, 1.2?) and Um-pu (1.3.1.51) which are too short for an unambiguous interpretation. There is a very slight justification for considering them as potentially Elamite (cf. EO, nos. 57.198.200). Most of them are included here as they occur together with Elamites (cf. SEL 8, p. 234f.). Ab-ba(1.1) and Si-si (1.3.2.31) are definitely atypical, but they were borne by individuals who are described as Elam. Ha-ne and Ra-i (1.1) have the same designation (cf. EO, no. 29.193?). Si-a-a (1.3.6.12) was originally from Susa. In-zu (1.3.1.117) can be short for Sumerian compound names (cf. Di Vito, Diss., p. 71). Za-na (e.g., 1.3.2.42) is a short name, but is definitely neither Sumerian nor Akkadian. Since it refers to females, it can render the Elamite word for "woman". On the other hand, the male's name Za-na (var. Za-ana(-a), AOS 32, p. 208, s.v.) is atypical. An-za-za-a (1.3.2.6) may be the forerunner of Av•~aTi (cf. SEL 8, p. 226). Are La-al-gu (1.3.3.58), La-an-ku (1.3.4.4) and Pu-un-du (1.3.2.63) short for La-al-gu-ni,La-anku-ku and Pu-un-du-du (cf. above and EO, nos. 110.189) respectively? Is Ur III La-an-ga, which is listed without interpretation in Limet, Anthroponymie, p. 447, related to La-an-ku? Hu-ne-re (1.3.1.28) can be a homonym of MB Hu-ne-er (cf. Scheil, MDP 22, p. 189). For 1.3.3.9 see Iran XXV, p. If. 1.3.2.56 is very doubtful (cf. SEL 8, p. 233). The same applies to 1.3.2.10-13 (cf. EO, nos. 72.211?) and Muuir-ti/ti(n)-ga-ba, Mu-ir-dulo-gab (1.3.3.51.65.97; cf. EO, no. 150c?); for -gaba cf. perhaps Ad-da-gaba(1.3.2.50, Akkad. acc. to Gelb, MAD 3, p. 116). For 1.3.6.7 cf. perhaps EO, no. 171a in fine and for 1.3.1.42 EO, no 142c (differently Lafont, DAS p. 226, n. 37). 1.3.1.54, 1.3.3.3-5 (cf. EO, nos. 52.71.232aC/7.1) and 1.2.2.26 (to EO, no. 287?) seem to be explicable in Elamite terms, but are too short for an unambiguous linguistic affiliation. Is I-tu-ni-a(1.3.2.58) the forerunner of NA I-tu-ni-i(EO, no. 256a)? Does 1.3.3.95 end in the same element? Hybrid names are Rib-si/gi-mu-ut(Akk. -Elam.; not -ip- as in Iran XXV, p. 5). 1.3.4.1 is not necessarily hybrid: Hu-b/pu-nulis not certainly Semitic

ELAMITES

AND

OTHER

PEOPLES

FROM

IRAN

(listed without explanation in Gelb, MAD 3, p. 124). Hu-un-dul-gi (1.3.3.47) is Elamite-Sumerian if it is to be differentiated from Am. hunn- (also written defectively, cf. A. Goetze,JSS 4, 1959, p. 197 with n. 4; E. Sollberger,JCS 19, 1965, p. 28 with n. 7) and its initial component is identical with hun- of other non-Semitic names like Hu-un-zu-lu (1.2) which is undoubtedly Elamite (EO, no. 300), Hu-un-da-ri (1.3.6.5, prob. not Hurr. -(a)tal in view of Hu-un-dara, 1.3.2.5), Hu-un-kap-ku (1.3.2.2, cf. Kap-ku-ku, 1.3.1.118 provided it does not end with -kuk), "Huun-ha-ap-ri"?(1.3.3.90), Hu-ungu-uir-bi(MVN 11, 206, 212), as well as, perhaps, Hu-un-sa6(-3a61.3.2.60; 1.3.3.49) is atypical like In-za-za (BNF NF 18, 1983, p. 101:42). For Hu-un-ga6-an(1.3.2.4) cf. perhaps Ursa6-an (AOS 32, O 35, 9) and Hu-un-hu (1.3.3.35) which are alternatively atypical, as well as Hu-un-kiip-ri (1.3.3.41) which may contain Sem. -ki-ib-ri.Hu(cf. un-si/ze-ri(1.3.3.43) looks like Akkad. to in Old MAD 3, p. 129) but ms > ns is not attested Hums.rum Akkadian. However, the case for a segmentation hun + zeri does not seem stronger as Zi-ri (TrD 83) does not necessarily refer to an Elamite and is too short for an unambiguous linguistic affiliation. Huun-za (1.3.1.112) may be either Elamite or Akkadian (hunzua"lame"). Is the resemblance of Hu-na-zi (1.3.3.77) to the 2nd component of I-ti-hu-na-zi (from Ebla, cf. Du-hu-na-sell/e/zi, M. Krebernik, Die Personennamen der Ebla-Texte: eine Zwischenbilanz, Berlin 1988, pp. 89.168.209, s.vv.: "wohlnicht sem.") just incidental? A certain Hu-un-d"ul-giwas perhaps an Amorite (mar-tu).His namesake is defined both as Amorite and lz-SU.Aki (cf. RGTC 2, p. 172 with refs.), but this is not a definite proof against an Elamite derivation: compare the case of I-ap-tu-um (mar-tu,MVN 11, 180, 5; Drehem, und.) with Ii-ap-ti (gen.) father of a lt-iSU.Aki(cf. SEL 8, p. 229; cf. Ur III GN Ia-ap-ti-um,Gomi, Orient 16, 1973, p. 10 and Bibl. Yepet?).1.3.2.16 may consist of hun and dudu, unless it is atypical. Is 1.3.2.45 a variant thereof?. 1.3.3.1 is homonymous with A-ku-ku-nifrom Susa (BNF NF 18, 1983, p. 107:121). At least two individuals bearing hypothetical Elamite names belonged to another ethnic group (Amorite), viz. 1.3.2.50; 1.3.3.69 (= iri-ta-h; both from Yamutbal or near it?). For hun- names cf. Goetze, JCS 17 (1963), p. 18, n. 22; Limet, Anthroponymie,p. 178; and Zadok, Iran XXV, p. 1. Do Hu-unda-ah-fe-er(1.2), Hu-ne-ydr-raand Hu-un-k[u(?)-li(?)] (1.3.1.23.110) belong here as well? (cf. 1.3.1.116.136; 1.3.3.5.24.31.39.63.73; EO, nos. 30C, 33a, 193.224 and Su-mu-un-da-ak-le-er, 1.5; cf. also EO, no. 208). 1.3.1.45; 1.3.4.10 (--1.3.3.74) are hardly Amorite as hesitantly suggested by B. Lewis and E. R.Jewell (ASJ 4, 1982, p. 58), but perhaps Elam. Hun-uri, i.e. "my image" (?EO, no 49). Hu-un-na-ru-um(Watson, Cat. 139, 5) looks Amorite. 1.3.3.53.83 end in tuk (EO, no. 282 and U-du-ku,

AND

THE

PERSIAN

GULF

REGION

35

1.3.1.13). Cf. 1.3.2.48. 61? Hu-un-gi-si(?)(1.3.1.111) may be compared with Pap-gi-si (1.2). Hu-un-ur5-ti was from Hu-ur-ti (cf. SEL 8, p. 229f.). On the whole, the segmentation is not always clear. Hu-un-nu-bi (MVN 3, 305, 9) is Akkadian (cf. AHw., p. 356a). DfFan-(Da-an-) is rendered as Elamite "act" (M. Lambert, RA 64, 1970, p. 72) or "Gehorsam"(EIWb., p. 280f.) without sufficient evidence. Does the fact that D/Tan- (Da-an-) is not recorded after the 13th century B.C. strengthen the case for Akk. dan(nu)"strong"? In NEIRAE it is found in final position only (cf. EO no. 241). Was a hypothetical Elamite form D/Tan(n)-misinterpreted by the Mesopotamian scribes and contaminated with the resembling Akkadian element? The pertinent anthroponyms are 1.3.3.6 (to tahra, EO, no. 237a), 1.3.3.80 (to hilu, EO, no. 40?), 1.3.1.91, 1.3.2.3; 1.3.5.2 (to iki, EO, no. 60), 1.3.1.25 (to Kalla, cf. OB Ku-ku-kal-la, ElWb. pp. 421.555), 1.3.1.90; 1.3.3.86 (to EO, nos. 18.85 resp.); 1.3.1.48; 1.3.1.92 (cf. 1.3.2.4.60), 1.3.3.64 (cf. EO no. 229; OB Ku-uk-duu-ku,1.5), 1.3.1.44.93 (Akkad. acc. to Gelb, MAD 3, pp. 55.113, but cf. EO, nos. 241.277?). The same initial component may be followed in the case of 1.3.3.45 (acc. to H. de Genouillac, RA 27, 1905, p. 100:17) by an element related to the Elam. DN El-ha-la-hu(cf. EO, nos. 64.241; cf. GN MDP 14, P1. after p. 10, 111, 14; OAkkad., Ha-ta-ahki, not far from Elam); cf. 1.3.3.71? D/Tan-bi-si(1.3.3.48) is very doubtful (cf. EO, no. 184?). 1.3.3.41 may consist of iki and rum (EO, nos. 60.201). 1.3.3.20 would be related to EO, no. 60b if mah is a Sumerian appellative meaning "I. (who is) outstanding in size" (cf. 1.1 and Ur-ki-mahwith -gula, FAOS 17, 78, 11; 82, 4). 1.3.2.21, 1.3.3.79; 1.3.5.3 may alternatively be Akkadian (Qtir/Qizr-,cf. SEL 8, p. 235; MessT, p. 240, s.v.). Qur-da-s[u(?)](J. N. Postgate, Sumer 32, 1976, p. 100: 14, 2.seal 1) seems-if the reading is correct-to strengthen the case for an Akkadian derivation. La-mu-sa is an Akkadian name according to Limet, DPA, p. 21 ad 45, 3 (mentioned together with Hu-ba and other foreign names; 1.2) who does not interpret the name. It is homonymous not only with persons occurring in Messenger Documents concerning Susa, but also with his namesake who was connected with Shimashki (HLS 3, pp. 102:185; MVN 11, 160). The female name Am-ma-ha-tum(1.5) contains-if the reading is correct-amma (EO, no. 7). Is the final component related to -ha-at(1.3.3.72) and Hattim (patronym of 1.3.2.62)? Regarding the Elamites' names listed in 1.2, Adda-sel-4a consists of atta and perhaps lip (EO, nos. 18.223). Si-nam-ki-ri-ir may end in kirir (EO, no. 103a). Are Si-anha(/kir?, 1.2) and Pi-li-ri-ni (1.3.4.6) based on siyan "temple" (cf. Si-ani, 1.2) and pilir (EO, no. 181a) respectively? Does Si4-ni-fi-ba (1.3.1.128) consist of EO, nos. 205.223? For 1.3.1.116 cf. perhaps E1Wb., pp. 755ff.; EO, no 71.

36

JOURNAL

OF PERSIAN

STUDIES

same applies to E-la-ak-nu-id(also at Gasur, HSS 10, p. xxxiia with refs.; no. 5 below occurs in a Sargonic text from the Diyala region); or E-la-ak-nu-id (MAD 5, p. 103 with refs.; Sargonic Kish). All the explicable compound names with E/Ilag/k(u) have Semitic predicative elements (mostly Akkad.). They were borne by at least 55 (18 + 37) individuals in Sumer (nos. 2-31; 39-45 below; practically in every city during the entire period of the Ur III documentation; one [no. 1 below] is late Sargonic) and by seven persons in OB Mari (nos. 32-38 be RAE Appi-za-lu-um (OnP 8.101; unexplained) from 24231 more convincing? B/Put-ri-za-lum(BM below). Hence both the linguistic character and the Ur III Lagash; unpubl., courtesy of Prof. Sigrist) geographical distribution of the documentation do apparently ends in the same element. Cf. BAR-ga- not support an Elamite origin (except for 39-45 and perhaps E-la-ga-ak, WMAH 231, 3 [ratherlum (MessT 467)? For Bu6-ka-ra (1.3.1.87) compare Icomlast Does the below. >?]):(1) E-la-ak (ITT 878; MVN 6, 309, i, 7); (2) perhaps Elam B/Pugdrki D la-ak from E-la-akof 32, (4) (3) 1,4). (AOS I-la-ak-ku-r[u-ub]; ponent Su-mu-gi-in-ti (1.3.5.9) originate

The predicative elements of dSulgi-en-u-ba-ak and dSul-gi-ur-u-ru-ukwhose bearers occur in MVN 12, 125 (concerning 1.3.3.12.13) do not seem to belong to any known dialect (cf. Steinkeller, JAOS 108, 1988, p. 202, n. 37). "n4Gul-su-ga(segmentation?) is strange (1.3.1.102; cf. Deimel, SL 2/2 p. 461: 229, 23 in fine and perhaps 1.3.1.100). Ab/p-za-lum (1.3.1.85) has a Semitic appearance (Pinches, Amherst, p. 186 ad loc.). but it does not render any recorded Semitic form. Would an association with

Elam. * tempt (EO, no. 246)? Gelb (MAD 3, p. 261,

s.v. "S'M?")hesitantly regards it as Akkadian. Un-daga looks as if it consists of un and ta-k (EO, nos 238.270). Yet, the relatively high frequency of the name especially at Nippur (at least eleven out of seventeen occurrences ), the most prestigious cultic centre of Sumer, as well as the spelling Un-daga-a (TuM 3, 55, 4.9) seem to weaken the case for a non-Mesopotamian derivation. Kur(u)b-ilak (cf. MAD 5, p. 110, s.v. Ilag; from Sar-

gonic Kish; at least 18 individuals from Lagash, Umma, Puzrish-Dagan and Nippur between 2055 and 2028 B.C.; AUCT 2, 18; 254; 3, 128; 188; 213; CST 178.198; FAOS 17, 112, 93; p. 330: Seal A; Hallo, HUCA 29, 1958, p. 107f.12; ITT 7076; MessT 290.386.436; MVN 3, 82; 219, 13, 911, 4; TE 54) may end with the same theophorous element as in Akk. I-la-ak-nu-id(cf. A. L. Oppenheim, AOS 32, p. 188, n. 34; M. Lambert, TE, p. 147: 248). W. von Soden (AHw., p. 374a) renders e/ilak as "your god" (cf. MB Ba-ah-la-ka, i.e. "your lord/Baal", from Tall Munbaqa, M. Dietrich, O. Loretz and W. Mayer, UF 21, 1989, p. 136f., provided it is not short for a compound anthroponym like Bac(a)lkabar). Sollberger (TCS 1, p. 114b:187) compared E-la-gu.The latter is described as name of Sarpanitu in Elam in a SB list (cf. EO, no 65), but this is a comparatively late source (after such anthroponyms seem to have been out of use). In addition, E-la-guwas in all probability a female deity like SarpTinitu whereas the E/Ilag/ku-containing anthroponyms are all mascu-

ku-ru-ba; (5) E-lagu-(ru)-ub (see Gelb, MAD 3, p. 149f. with refs.); (6) E-a-ga-DINGIR (MessT 338, 18); (7) Ela-ak-MES (? MessT 234, 15), (8.9) E-la-ak-nu-ib (MessT

228, 6; 402, 12; mistake for nu-id?); (10-14) E-la-aknu-id (BRM 3, 53, 2; DAS 51, ix, 19; NATN 926, 7: 927, 5; PDT 341, r. 5); (15-21)

I-a-ak-nu-id (MVN 3,

165 r. 3; 209, v, 7; 271 r. 2; 279 r. 2; 320, 3; 322, r. 4; 376 r. 10; 13, 724, 6 and 732, 9); (22) I-a-ak-ra-am (FAOS 16, 1120 r. 1); (23) E-la-ak-(ITT 2 878, i, 31; x, i, 4), (24) E-tak-'u-qir (ITT 2 902, iv, 18; 911, 28; 1013,

iv, 2); (25) E-la-ak-ha-al-li (ARMT 22, 328, iii, 5); (26-31) I-la-ak-u-qir (FAOS 16, 1327, 3.seal; 1355,

3.seal; 1365, 3.seal; 1375, 3.seal; MVN 15, 142, 41; PDT 556, 3; YOS4, 254, 70); (32-37): - (ARMT 21, 401, 7; 22, 12 r. i, 13'; 14, ii, 13; 25, 690 r. 4; 714 r. 7; 729, ii, 20); (38) I-la-ak-su-qi-ir (ARMT 24, 24, iii,

41). (39) E-la-ak-ra-at(1.3.1.94) is the only name whose predicative element is not Semitic, but seemingly Elamite (cf. EO, no. 196). Its bearer was indeed connected with Elam. It is not impossible that (40-44) E-la-ak-rain the same documentation group (DAS 124, 3; 155, 11; 165, 7; 200, 7; 204, 14) is a defective spelling and that E-la-kiar/qar (1.3.1.95), who is connected with Anshan, is the same name. (48) E-lak/la-ga-ra-at may be a variant thereof (unlikely Gelb, MAD 3, p. 35, s.v.: an il(a)- name with a predicate deriving from Q-R-D thereby leaving the first two forms unexplained). Do Sarg. E-la-ak-gi (F. Rasheed, Himrin 4, Baghdad [1981], 41, 20) and Gu-ni-la-ak(with an Akkad. patronym, 1.2) as well as Ur III Ku-e-la-ak (AUCT 2, 282, 9) belong here as (MessT well? line, including E4aggu (1.3.1.113)/E-a-ga Linguistically, Ur-daulinak (1.2) is impeccably 121; Umma, SS 5 = 203312 B.C.)/E-lag/kag/k (ITT 2, but its theophorous element is the god of the face which on Sumerian, 714, 3)I-la-k[u] (MVN 15, 142) it looks like a DN used as PN. E/Ilag/ku was not of Susa. The theophorous element of Urdgu-nirra necessarily Elamite, the more so since Kur(u)b-ilak, (1.3.1.7; the initial component is unmistakably Sumerian) is listed in A. Deimel, Pantheon Babyloniwhose initial (predicative) element is definitely cum includhad a (Rome 1950), p. 89:550. It is not clear whether distribution, Akkadian, widespread ing regions northwest of Sumer (e.g. Ebla, D. I. Gunirra is identical with the much more common Owen and R. Veenker, Eb. 1975-85, p. 278, and Sumerian deity Gunura (see D. O. Edzard, RIA 3, p. Mari, cf. below and perhaps PDT, 161 r. 1). The 701 f.). Is it a coincidence that Gu-nir-raand Gu-ru

ELAMITES AND OTHER PEOPLES FROM IRAN AND THE PERSIAN GULF REGION

(cf. 1.3.2.65) resemble the Elamite theophorous elements Kunir and Kur respectively (EO, nos. 115.117? The theophorous element of Ur-d[b]a/[m]ada-[ti?]-[na] (? 1.3.2.7; cf. Limet, Anthroponymie, pp. 158.551; same name-type as the preceding) resembles the Shimashkian's name Ba/ma-da-ti-na (TCL 2, 5508). Was B/madatina a (petty) ruler's name? For such examples (ruler's name > divine name) see Michalowski, SAOC 46, p. 66f.; cf. perhaps the cases of Kur-bi-ak-gu-um (ITT2 665, 16), Li be-Ii-la-gdl (1.3.1.65), Ur-la-mus'-a (0. R. Gurney,JRAS 1937, p. 472: 4, 2), Dan-uir-ma-an-si (DAS 79, 5) and Mi-il-ki-li-il ("Milki-il is god", cf. Iran XXV, p. 4; differently G. Buccellati, The Amorites of the Ur III Period, Naples 1966, p. 173). (1.3.1.26; dSulgidan/dan"/da-anga-da element a theophorous 1.3.3.59) may contain which was originally a ruler's name (Sulgi-dan, itself with a ruler's name > theophorous element; pace Limet, Anthroponymie, p. 258, who offers an unlikely in view of da-an-dananSumerian interpretation; hardly Sum. -kalag- as understood by Di Vito, Diss., p. 51). dSulgi-si-im-ti/tzm (Schneider, Or. 23, p. 57: 916f.; cf. AOS 32, p. 185 with refs.) hardly ends in Elam. *tempt, but contains Akkad. simtum (the Elam. element is contained in 1.5!).5 Is Ur-da-ka (Schneider, Or. 23, p. 76: 1210) the forerunner of NA < Elam. Urtaku? (provided the latter is not an originally Iranian name);6 or is it an audial mistake for Sum. Ur-daga (cf. Limet, Anthroponymie, p. 152)? Elam ("NIM"), which (mostly accompanies follows) the names of 75 (70) individuals in Ur III (table no. 1, col. 2 above), precedes the names of many regions which cannot be located in Elam proper (cf. SEL 8, p. 227f.) like Si-z/umki (sakkanakku: Hu-ba-a whose daughter married a royal prince, see Goetze,JCS 17, 1963, p. 13; Michalowski, SAOC 46, p. 58f. with n. 16), U-lumki (cf. "Hu-un-ki-ip-ri lzi-ulliki,, which may have Elamite onomastic parallels; perhaps identical with OB [Susa] GN U-li-me and Neo-Sum. PI(wi?)-ilki [with PN, Su-ni-ki-ib/p], U-ulki [with PN Gu-up, atypical; 2039 B.C.; PDT 548, 17]; and U-ir-riki/Ur-rik (poss. = U-ra-umki). Elam lhiB/Pu-garki is recorded in ITT2 875 r. viii, 6. In two cases Elam is followed by the ruler's name instead of a GN, namely U-ba-a from Adamdun (MessT 55, 21; cf. H. Waetzoldt, ZA 65, 1975, p. 272) and Hu-ul6, from li-bar (e.g. Sigrist, Syracuse 480, mentioned is Duduli Duduli/Tutuli). frequently together with Susa (e.g., ITT 638.683.772), as well as with Shimashki and Huhnur (e.g., WMAH 225), but Duduli seems to have had a special relationship to Sapum according to DAS 153. The location of the latter is disputed.7 Does the fact that Sapum was terms with the Neoon peaceful generally of Sumerian state reflect a relative proximity Sapum to the Mesopotamian alluvium?8 Duduli is thought to have been located somewhere between the Tigris and the Zagros (see Edzard and

37

Farber-Fliigge, RGTC 2, p. 33 with lit., cf. SEL 8, p. 227). Another region of Greater Elam may be B/Pugdrki (ITT2 875, viii, 6). Was Hagarki, which is recorded in Messenger Documents (cf. RGTC 2, p. 72; MTBM 154, 6), in Greater Elam as well? A field of Elam people is recorded at Lagash in = 2049 B.C. A village and granary of such 46 S in undated documents are mentioned people (MVN 6, 300; TuT 160, iii, 20 and UDT 55 resp.). For store(house)s named after Elam people (one of them situated on a canal of such people acc. to Lagash documents) see G. Pettinato, UNL 2, p. 247 f: 27.50; cf. MVN 12, 5,44; Limet, CRRAI 18, p. 132 f. on settlements named after various foreigners. Several Elamite individuals and most members of the numerous groups of the various Elam people are anonymous (1.4). It is clear that Elam defines regions which-as found in the far as they can be localised-were Iranian Plateau or its piedmont, but not in northern Iran (Gutium and Lullubum; if Elam Si-ma-na with [MTBM 170, 9.10; no year] is identical Simanum Greater Elam would have extended more to the NW), Upper Mesopotamia and Syria which also included mountainous regions. Therefore, it stands to reason that Elam denoted eastern highland[er(s)] from the Lower Mesopotamian point of view. Oppenheim's proposal ("most probably a certain profession") seems less likely. McNeil proposed that most of them were labourers of some important government projects9 The SU(A )-people were identical with Shimashki according to P. Steinkeller (JAOS 108, 1988, pp. 197 ff.; cf. G. Selz, NABU 1989/4, p. 67 f.:94). An orchard from which wood loads were taken was named after these people (gi?-kiri-SU.A, Umma, IS 1 = 2028 B.C.; CST 527).

1.1 Pre-Sargonic in archaic accounts Elamites are mentioned from Ur (UET 2, 274, ii, account or name-list; 279, ii, concerning bread), provided the forerunner of the NIM sign has the same reading as in the postarchaic texts (this problem exists in other archaic texts as well). However, there is not a single named it is impossible to decide Elamite. Therefore, whether the individuals in question were ethnic Elamites or people from Susiana which was heavily Semitised. The same problem exists in most of the from Mesopotamia where earliest documentation (includmany recorded Elamites are anonymous and perhaps 115; cf. 1.2). ing, e.g., 1.3.1.10.27.97-99 There is no assured relevant documentation from Jamdat-Nasr (cf. RGTC 1, p. 43 ad OECT 7, 88, iii). The deity Lugal-Elam(a), i.e. "Lord of Elam" (following Lugal-Aratta), is recorded in the Abli

38

JOURNAL

OF PERSIAN

Salabi-i god list 63 (OIP 99, 82, iii, 19; cf. W. G. Lambert RIA 7, p. 137b).10 Few named and anonymous individuals (mostly officials) are described as Elam in the Shuruppak (Fara) texts (c. 2550 B.C.; cf. F. Pomponio, La prosopografia dei testi praesargonici di Fara, Rome 1987, pp. xvi.195, s.v. NIM). None bore an Elamite name. Note lhi-URUxAki from Shuruppak (TSS 131; S.N. Kramer, JAOS 52, 1932, p. 111:f.600 [PN?]; URUxAki, TSS 40.453 [PN?]). These documents may refer to people originating from a town which is to be sought in or near Elam not far from the Mesopotamian border (presumably in northwest Susiana according to Steinkeller, RIA 7, p. 381). All the following persons, except for sixteen, are from late pre-Sargonic Lagash (first half of the twenty-fifth century B.C., cf. Westenholz, ARES 1, 1988, p. 117; refs. to Nik. 1 are to Selz's re-edition in FAOS 15/1; see idem, CRRAI 36, pp. 27 ff.): 1-7. Za-na (for the Elamite character of the name see 1.0) occurs in at least 17 documents (CTNMC 4, i, 6; v, 15; xi. 7; xiv, 9; Nik. 1, 1, iv, 22; 2, v, 5; 6, i, 9 [restored]; v, 15; xi, 2; xiv, 9; 9, iii, 8; x, 15; r. ii, 10; 19, vii, 10; STH 1, 15, 10; 16, 10; 17, v, 2; 18 r. ii, 8; 20-23; TSA 10-12.14-17). The name refers to at least seven females between Lugalanda's sixth year and Uruinimgina's eighth year (most-if not allof them of servile status, cf. FAOS 15/1, p. 107; TSA, p. 122). 8. Ab-ba (Elam) is recorded in the same section with Za-na during the first year of Uruinimgina's reign (Nik. 1, 9 r. 11, 10; STH 1, 15, 10; 16, 10). Were Ab-ba and at least certain Za-na's musicians? (cf. Selz, FAOS 15/1, p. 107; for the connection between designations of musical instruments and geographical terms see H. Hartmann, Die Musik der sumerischen Kultur, Frankfurt a/M. 1960, p. 197). 9-14 Ha-ba-ra-du-ne(with -tuni, EO, 256a?), Ka-a (atyp.), Nin-DU-il-il, Ugiur-DUlo.DUlo (both unexplained), dNingir-su-ur-muand PAP.PAPama-da-ri(both Sum.); see Selz, CRRAI 36, p. 39. 15. Pu-da-gi-ir(Nik. 1, 310, see Limet, CRRAI 18, p. 131; Selz, FAOS 15/1, p. 540; time of Lugalanda and Uruinimgina) and 16. Lugal-nimgir-gi(dumu Elam, MVN 3, 53, iii, 7) occur in undated documents. 15 was active in the trade between Lagash and Elam. 17. Ka-ka-ri-tah and 18. Kum-ku-`e's son were engaged in the same activity then. The latter's (cf. J. Bauer, Altpatronym is unexplained sumerische Wirtschaftstexteaus Lagasch, Rome 1972, p. 524 ad 176, ii, 1 with lit.; note Kum-KU,MVN 6, p. 343b, s.v.) and the former's makes a nonMesopotamian impression. This trade was also operated via the watercourses: boatmen such as Dada (DP 486, 5; 637, 8) and Nigin-mud (DP 423, 2; 486, 5; 687, 7; Nik. 1, 19, 4; A. Deimel, Or 26, 1927, p. 39 f.: Wengler 2) brought especially timber and furniture from Elam; an Elamite boat loaded with

STUDIES

spices and livestock is mentioned in RTC 21, 11, 3 (see Lambert, RA 47, 1953, p. 62 f.; ArOr 23, 1955, p. 570 f.; Selz, FAOS 15/1, p. 134 ad loc.). Slaves were also brought by traders from Elam and adjacent mountainous regions to Lagash. A list of 12 slaves (igi-nu-duh; belonging to four Lagashites) from Uru-az is dated to the fourth year of Uruinimgina (DP 339; cf. Lambert, RA 47, p. 69f.). At least half of them bore (or were given by their masters) Sumerian and Akkadian names. The others had "strange" and atypical names (Ab-umbzim,B/Pu-su-e, KAM-ba-eand Ku-ku, Si4-ma respectively). Nin-URUxA-NI-dulo is recorded as an anthroponym in Uruinimgina's fifth year (Nik. 1, 19 r. iii, 10; cf. ad (lii-) URU.Aki above). A forwarding agent (? hli-U5) from Uruaz is recorded at Lagash as early as the first or second year of Lugalanda (Nik. 1, 143). For trade with Elam (e.g., Sumerian grain for Elamite cattle and wool) cf. Selz, FAOS 15/1, p. 302; W. F. Leemans, Foreign Trade (Leiden, 1960), p. 175; Englund, Fischerei, p. 14 f. It is interesting that an anonymous Elamite (Elam) occurs in a record of land distribution (SAKF 3, ii, 6; und.). Ku-kugu-la, i.e. "Ku-kusenior", who is followed by Kuku-tur "Ku-kujunior" in three documents (cf. Mas-gu-la followed by Mag-tur in DP 137, ii, 1 f.), is not necessarily Elamite, but atypical (see Selz, FAOS 15/1, p. 236-5: 12, 14). One of these documents (DP 134) mentions Ni(g)-dub-(b)aNIM. Ku-ku is typologically, but not necessarily linguistically, comparable to Hu-hu-tur (Nik. 1, 3, 4), i.e. "Hu-hujunior". Does Ku-uk-ki-mah(cf. Iran XXV, p. 24, n. 7) mean "Ku-ukki who is outstanding in size"? It is not certain whether gi-NIM, who is recorded in at least 11 Lagash documents (dated between Lugalanda's fourth year and Uruinimgina's fifth year, RGTC 1, p. 43 with references; VS 25, 85, ii, 3; CTNMC 3, viii, 1: g[i-NIM]; cf. En-gi-NIM, RTC 44, 13 from Lugalanda's sixth year [hardly Gi Elam; cf. FAOS 15/1, 264, ii, 5] and 1.2) has anything to do with Elam (Limet, Anthroponymie, p. 422 lists Gi4-ni-mu [certainly not Elam.!] from Ur III Lagash without explanation). In Selz's opinion (FAOS 15/1, p. 139, Nik. 19 r. i, 10), Ki-NIM has probably no relation to Elam. DP 164 (from Lugalanda's fifth year) records a delivery of grain (-products) to Aizkimzid (of Ninmar) at Migime (Bashime) after it was conquered by Eannatum (see Gr6goire, Prov. mer, p. 17 f.). An undated list of Elamites (Nik. 1 = FAOS 15/1, 11) has Ha-ne, Ra-i (cf. 1.0), Al-la, Mi-du (atypical) and Ne-sag (Sum.). Note Bara-URUxGU-a (TSA 2 r. 5; perhaps -Guru, a place which is mentioned together with Elamite locales, cf. RGTC 1, p. 184; 2, p. 239). Inim-ma-ni-zi(Sum.) Elam is recorded in DP 374, 3. For a dubious occurrence of Elam ("NIM") in Pre-Sargonic or early Sargonic Adab see Edzard, SR, p. 164, OIP 14, 75.

ELAMITES

AND

OTHER

PEOPLES

FROM

IRAN

AND

THE

PERSIAN

GULF

REGION

39

ively. Uru-azki ("Bears' town" acc. to Selz, FAOS 15/1, p. 359) is also recorded as an anthroponym in a document Da-an-'-ri is mentioned (< gentilic) then (ITT 1195 r. 3). It is interesting to Lagash. recording delivery of baskets (STTI 37:L. 1277, 2). find a scribe named after U.URUxA-mes, a place An Elamite priest is recorded in a list of metal which is to be sought in Elam (eventually = objects and military equipment (STTI 7: L. 1125, 6). URUxA, see Edzard, Farber and Sollberger, RGTC STTI 63: L. 1469 records a purchasing trip to Elam. 1, p. 188 ad Nik 2, 14, ii, 14). Anonymous Elamite An anonymous Anshanite is recorded in RTC 247 ugula acted in Gudea's eleventh year (MVN 7, 493, r. 17. A Susian occurs in CT 50, 148 (concerning 13). Umma. For Susians and Elamites from Bashime, rations; cf. 146). Hu-ba (possibly Lagash, perhaps who time of Sarkaligarri, 2217-2193 B.C. or slightly see Foster, USP, pp. 15 f. 113. huiu-li-nimki, later) received rations (DPA 45). The ensis of Susa presumably originated from Susa, is recorded in and Elam, who might have been expelled from CT 50, 56, 21 (concerning beer). The very same their regions by Sarkaligarri, held land in Girsu spelling recurs in T. Donald, MCS 9 (1964), 241 r. according to RTC 143 (see B. R. Foster, Mesopotamia 11; 242 r. 7 (from Umma? possibly late Sargonic or [Copenhagen] 9, 1982, pp. 19. 36 f. with n. 31). A Gutian). An anonymous Elamite acted as a scribe certain Ha-ne is recorded in the Mesag archive (DV 5, 61, 3). Nippur. An anonymous Elamite is recorded in a (perhaps near Lagash, cf. S. J. Bridges, The Mesag beer account (OSP 1, 57, i, 7; cf. the anonymous Archive: A Study of Sargonic Society and Economy, Yale craftsmen in i, 2.4.6; ii, 4.5). Westenholz (OSP 1, p. Dissertation 1981, p. 32). The following individuals occur in the late Sargonic period (in MVN 6 unless 38), who suggests a probable date in Sargon's reign, otherwise indicated): Ba-ak-za-na (fem.) is listed is of the opinion that he presumably was an before Ba-ar-si-ni whose name is likewise non- "ambassador". Other anonymous Elamite individMesopotamian (381 r. 7; cf. Ur III Ba-ra-gi-in, uals received beer and clothes (*NI[M]- x I, TuM 5, 1.3.2.37). Hu-um-ba (-"MA";shepherd; 351 r. 5). 15 38, i, 6; 108 +, iv, 7; cf. Westenholz, OSP 1, p. 94, individuals (most probably slaves), who belonged s.v.) there at about the same period. Another docuto four groups (a-d), are listed in a contract (500). ment (OSP 1, 129, v, 2 f.; contents uncertain Some of them bore atypical names, but all are (records Elam-mu from Du6-LUL (Du6-ka5, i.e. explicable in Elamite terms: a. Gu-ri,Na-pis-ir (both "Foxhill"? or Dul-us? see Westenholz, p. 110, s.v. 1.2. Sargonic

fem.), Hu-ba, Da-an; b. Ka-ka, Gu-ri, Ba-ar (three females), Hu-un, U-e-li; c. Si-im-ta[n], A-bi, Am-ba/ma-ar, Hu-un-zu-tu, Za-na (fem.); d. Hu-hu-me (fem.); Ha-ap-hi-

with lit.; cf. 112, s.v.

Gdna-Giggi-dul4-aki). Adab. Elamite prisoners received rations (cf. Yang, Adab, A 672, 4). It is very doubtful whether Uis (late Sargonic? 100, 2). The following individuals i-Ui(ibid. 632, i, 3) was an Elamite. Gasur (all in HSS 10). Pu-ul-ma(185, iv, 15), Si-a-ni are listed in ITT (transcription/transliteration only): Gi-NIM (5839; cf. 1.1); In-tar-ra (3107; cf. (169, 9; 199, 2), Ti-ru-s'a-ki(129, 13; 156, 7; 197, 10; 1.3.3.3); Mu-mu-'i (4559; apparently cultivated a cf. EO, no. 211) and Za-na (fem., 188, ii, 18; iv, 21; cf. field), Zu-zu-i-lum (ITT 4518) and possibly Ur- Th. Meek, ibid., p. xxix n. 15 ad loc.). Other strange names such as Hi-ir-ha-ga(176, 13), Hu-hi-ir (26, 9; dusinak (2855; list of offerings) originated from Susa. An anonymous ensi of Susa is mentioned in 61, 5; 162, 10), Ki-ip-tu-ru(129, 11; 153. iv, 31) and U(ITT 4560. Susians are recorded in ITT 4700 and i-li (cf. Gelb, MAD 3, p. 31!), are not necessarily Elamite as Gasur was situated near regions where Elamites (lh-Elam) in ITT 2905 (after "Nim-hid") and 4514 (account of cattle). An anonymous Elamite Lulubian and other unclassifiable dialects were possessed a slave with a Sumerian name (ITT 5798). spoken. Eshnunna. Ku-ru-zason of 1, 85:TA Ma-ad-ga(RTC 253, 8 r. 5) is homonymous with a Se-ilx-ha (MAD if name of Elam not the Greater 1931, 9). region's region Other places. Susians (sg. li-sulinaki, e.g., Pinches, with omission of the determinative. People from the neighbouring region of Kimash (also sometime Amherst 7, 4'; cf. RGTC 1, p. 154 f.), as well as an within Greater Elam) are recorded in RTC 248, 7; Elamite (E-lam,MAD 4, 16 r. i, 7), are recorded. If a 251, 3'; 252, 5. Pu-ul-ma is mentioned among segmentation Elam ci-um is correct (a name Nim-iaiSusians (MVN 6, 90, 5 r. 9; late Sargonic?); Ha-al-ka um [M. van de Mieroop, RA 79, 1985, p. 20:3, 12] (late Sargonic and 11th year of Gudea, MVN 6, 377, has no parallels and is unexplained), then it may 1.15 and 504, 6 respectively). Anonymous Elamite refer to an Elamite (concerning delivery of guruJwomen received rations according to documents workmen). An Elamite woman (e-la-mi-tum) is from the late-Sargonic Suna archive (MVN 6, recorded in OAIC 9, 10. Hu-un-da-ah-le-eris men105.335.492). They are mentioned twice (105.492) tioned in MAD 3, p. 130: A. 3004 and [Ru(?)]-hupubefore people from E-la-nirki (in Elam?). People ni in BIN 8, 139, 4 (delivery of sheep). from Bashime and Uru-azki are recorded in Nik. 2, Foster (ASJ 4, 1982, p. 24; BJRL 64, 1982, 35, ii, 12 (from Lagash?) and RTC 113, 3' respect- pp. 459.472) notices that CST 19 (possibly post-

40

JOURNAL

OF PERSIAN

STUDIES

Sargonic) contains Elamite names, viz. Si-an-ha Hu-ba-a (S 48; TuT 152, 18; also TIM 6, 34, AS 2; (IKIR ? cf. Westenholz in Foster, loc. cit.), Si-nam-ki-ri- from Puzrish-Dagan, cf. CSAU, pp. 57.87); 17. An ir, Ad-da-sell-baand Pap-gi-si. However, they were Elamite (Elam) is mentioned (AS 1; CT 10, pl. from B&d-da-NIMki,i.e. "Fortress beside Elam". If 38:15296, i. 23) together with 18. Hu-ru (AS 1; HLC Foster's interpretation is correct, then these Ela- 2, pl. 83:73, i, 3.6; the latter also in Jean, RA 19, p. mites did not reside in Mesopotamia, but were 42: 82, 4; MVN 15, 167 [IS 2]; SET 206, 13); 19. Adson probably stationed near the Elamite border. da-na-pirx (AS 1; CT 7, 34:18409 r. 6); 20. Bu6-sQ(?) of Hu-ba (AS 1? TuT 159, i, 13; cf. vii, 14). 21. Li'dnaru-a son of I-lam-ma(AS 4; from Gu'abba; CT 10, pl. 1.3. Ur III cf. 115? 22. Se-li-bu-um(AS 5; STH 2, Abbreviations of royal names (chronology acc. to 16:12921, i, 24); 75 r. 4). An Elamite is mentioned then (AS 5; TuT J. A. Brinkman in A. L. Oppenheim, Ancient 37 r. 10); 23. Hu-un-k[u(?)-li(?)] (AS 7; SNSA 110 r. 1); Mesopotamia, rev. ed. completed by E. Reiner, 24. Ad-da (AS 8-9; CT 10, 46, 2; DAS 97, 18; 137, 6; Chicago and London 1977, p. 336): S = Sulgi 154, 14; 176, 22; 182, 19; 192, 7; 193, 8; 229, 7; 311, (2094-2047 B.C.); AS = Amar-Sin/Suen (20467; HLC 2, pl. 75: 57, iii, 6; ITT2 683, 23; 879, ii, 10; 2038 B.C.); SS = Su-Sin/Suen (2037-2029 B.C.); IS MTBM 21, 6; 24, 7; 49, 3; TS 99 r. 2; WMAH 227, 1 = Ibi-Sin/Suen (2028-2004 B.C.). The abbreviated f. occurs frequently together with 44 below: AAS RN is followed by the year number (e.g., S 47) 176 [cf. Gregoire, AAS, p. 209 ad 171]; DAS 84, 21; below. 85, 19; 155, 20; 157, 20; 159, 20; 177, 21; 183, 25; 195, 33). 25. D/Tan-kal-la(AS 9), who belonged to a 1.3.1. Lagash guardians' group (DAS 206, 12), was most probably tukul, AS (From Girsu = modern Tall al-Loh unless other- identical with D/Tan-kal-athe guard (l-zigfAS 9 acc. wise indicated.) 9; DAS 114, 4); 26. dSulgi-da-anga-da(etc.," 1. Ad-da-na-nam(S 32; MVN 7, 82); 2. Gu4-ka-ri(S to DAS 148, 7; mentioned frequently together with 33; MVN 7, 251 r. 2); 3. Dumu-dabs-ba-MUS.ERENki 44 below); 27. Elam-mu(AS 9; CT 10, pl. 26:14315, ii, (S 34; A UCT 1, 21); 28. Hu-ne-re(AS 9; WMAH 209, 9). (S 34; MVN 6, 66, 8); 4. Susa" (a-s' susinak, a-sa a20 Elamites are listed in TE 47 (AS 9), but only 12 303, 15); 4a. "The field ofIg-pu-ga-ru names is recorded a (29-40) are preserved: 29. Igi-ni-ta(2); 30. AtSusian(s)'s field?) gar -; originally between S 34 and IS 2 (see Pettinato, UNL 1, p. 64: ti-iz-ta (3); 31. Pd-hu (4); 32. Pi-li (5); 33. A-ad-da-hu (6); 37; 2, p. 194: 774; RGTC 2, p. 190; mentioned after 34. Ip-qui-sa(7; WMAH 233, 3); 35. Ma-ds'(8; WMAH the Elamites' village in TuT 159; cf. i-dub(-fi-) 214, 11); 36. dNanna-hi-li (9); 37. Be-li-a-zu (10); 38. susinaki,Waetzoldt, ZA 65, 1975, p. 276 ad ITT 7131, [...] -]i-im-[ ..] (r. 2); 39. Se-en6-[...] (r. 3); 40. [...] -pa19). Note dNin-usinaki (e.g., MVN 6, 588, 2). 5. ni-h[u] (r. 4); 24 above (r. 5). 41. Ad-da-bi-li-ir(SS 1; IHu(?)l -i-ni-da-a-se (S 37; TuT 226, 5); 6. Ap-ku (S DAS 152, 7); 42. Me-en-ra (SS 1; DAS 100, 3); 43. Hu44; HLC 2, Pl. 92:98, ii, 2); 7. Ur dgu-nir-ra(S 45; TuT ba-a (SS 1; WMAH 210, 7). For anonymous Elamites 258, 4); 8. A-da-napirx (PIRIG; S 45/AS 2; CT 10, 28: who are mentioned in the same year see 1.4. 44. Da14316, ii, 24); 9. Ad-da-na-pirx (S 46; WMAH 175, ii, an-u-pi, D/Tan-i-pi/pi/pi4, [D/Tan]-u-e12 (SS 1--IS 2; 27; cf. Steinkeller, ZA 77, 1987, p. 92 f.); 10. Lu-lu see 1.4). 45-48 are recorded in ITT 6787 from SS 2 (cf. de (wife of an Elamite nu-banda;S 46/AS 3; HLC 1, P1. ITT 5, p. 23; the other names occurring 11. HLC Hu-ba-a 46/AS 12: Genouillac, 3; 1, 4; ii, 4:5, i, (S 2); P1. 52, ii, 5; from Lagash?). 1la. The field of Hu-ba-na in this document are not Elamite; Su-ru is atypical): (cf. 134 below) is recorded in S 46-AS 8 and in 45. Hu-nu-ri; 46. Hu-ba; 47. Ad-da-na-pi-ir; 48. D/Tanseveral undated documents. It is often followed by na-hu-ti; 49. Chi-gi-ni (SS 3; ITT2 826, 15); 50. Ab-bavarious designations: a-sa H.-ambara (in a swampy sa6-ga (SS 4; ITT" 741, i, 13; MVN 6, 62, 7 f.); 51. Umregion), -agi-uis (of the gendarmes), i7-sug-ge-dar-ra pu (SS 4; TE 43, 15). 52. Hu-ba (SS 6; DAS 60, 16.29); (on the Suggedarra Canal). A plot of a-4aH. is fol- 53. Hu-ba and 54. In-da-a (SS 8; HLC 1, pl. 5:37, ii, 3 lowed by kzir being the only field which is desig- and iii, 4 resp.); In-da (went to Adamdun, Pinches, nated as such. For all the material on H. see Berens84 r. 10); 55. Hu-hu-ni (SS 9; ITT2 944); 56. HuPettinato, UNL 1, p. 278 f.: 421 f.; K. Maekawa, ru (IS 2; MVN 15, 167); 57. Hu-ba-ni-i (IS 3; IT7T 726, Zinbun 13 (1974), pp. 22.36. It was situated opposite iv, 24); 58. E-gi-il-ta (IS 3; ITT 906). 59-138 are recorded in documents whose year is the field of dEn-lil-l-[i]-ga6 (Enlil-ifa) according to either not known or not reported; 59-67 and 68-71 WMAH 42, B, ii, 2. For a-&-hu-ba-na igi-ug-uduki (perhaps opposite U'udu) see A. Falkenstein, AnOr belonged to two groups consisting of ten (WMAH 214, 2 ff.) and five (WMAH 233, 1 ff. where 44 above 30/1, 39 (cf. RGTC 2, p. 237). 12. Ur-dgulgi son of Hum/Lum-ba-ru(S 47; HLC 1, P1. 28: 248, ii 6); 13. U- is also mentioned) Elamites: 72-77 belong to a du-ku(S 47; Figulla, Cataloguep. 49: BM 12776, from group of six Elamites (BM 23157, pers. comm. Prof. Lagash?); 14. Ur-bdrson of Hu-ba (S 48; UDT 73, 22); M. Sigrist). 59. Su-d"ulgi; 60. Ba-a-a; 61. dSulgidan; 62. Ur-nin15. I-lam-ma (S 48; WMAH 176, xi, 23'); 16.

ELAMITES

AND

OTHER

PEOPLES

FROM

IRAN

AND

THE

PERSIAN

GULF

REGION

41

giz-zi-da;63. Ga-[ma]h-DU;64. Nu-ir-dsul-gi;65. Lii-be- 11); 131. U-eli (HLC 3, pl. 132:335, 5; IT7T 668, 28); son of Ku-gu-nu(TuT 210 r. i, 5); 133. li-lagdl; 66. Ur-niggar;67. In-ell; (35 above); 68. A-bu- 132. Ur-dba-ba6 um; 69. Luzi-sulgi-zi;(34 above); 70. Idi-~ul-gi; 71. Za-la-a(BM 27572, courtesy Sigrist). 134. [... . son of Li-gui-na; 72. Li'-dingir-sard; 73. Ur- dgegtin-an-ka, Hu-ba-na (grandson of S2a-ba-na-[gar],WMAH 282, 74. Lugal-zage-si; 75. Lti-dingir-ra;76. I-NElbi, 77. iii, 7; cf. MVN 2, p. 18 ad loc. and Ila above); 135. Inim- sara. Ad-da-na-pirx is mentioned in a provision list of six 78. Ba-la-a and 79. Se-ti-ti are mentioned before persons going to and returning from Susa (HSS 4, Elamites from Shimashki and Sapum in a Mes- 65, 5; see T. Fish, MCS 4, 1954, p. 89; Steinkeller, senger Document (ITT 2 893, i, 6; iii, 25); 80. A-da-ra WZKM 77, 1987, p. 192, n. 19; Lagash?); 136. Se-er(DAS 79, 13); 81. Agu-ni (TuT 56 r. 2); 82. A-ha-ma-ti napi-ir (from Lagash? cf. FAOS 16, p. 61 ad 900 r. 2; (HLC 1, pl. 88:92, v, 2) is mentioned in the same A. Uchitel, BSOAS 53, 1990, p. 126); 137. Ur-sag document as 83. Na-ba-s'd(vi, 11); 84. A-ma-an-ne-en (AAS 176, 12, cf. Gregoire, ibid, p. 214 ad loc.; men(connected with Adamdun, ITT 4097); 85. Ap-za-lum tioned in the same document as 24.44 above) was (Pinches, Amherst 111 r. 2); 86. Ba-ad-da-a(DAS 204, perhaps an Elamite; 138. Za-na (fem.; HSS 4, 47, 4). 11); 87. Bu6-ka-ra(TuT 194 r. i, 9); 88. Da-agu-nir An anonymous Elamite functionary (agd-iis) is (MVN 7, 377 r. 6); 89. Da-angu-ni (Jean, RA 19, 1922, recorded in WMAH 139, 2 (no year). An anonyp. 30:8, 12.16); 90. D/Tan-ad-da (MTBM 25, 2); 91. mous Elamite from Susa is mentioned in IS3 (SNSA D/Tan-nigi4 (Jean, RA 19, p. 34: 43, 9); 92. D/Tan-sa- 200, 2 f.). Anshan is recorded in another document sa (DAS 79, 19); 93. D/Tan-ir-ri (Jean, RA 19, 1922, p. (SNSA 252). An unpublished document in Istanbul (L-30315; 42:86, 3); 94. E-lag/k-ra-at (DAS 74), E-laga-ra-at (Pinches, Berens 80, 21; connected with Adamdun); cited by Gelb, Glossa 2/1, 1968, p. 93 f.) lists men 95. E-la-kar/qar(connected with Anshan, AOS 32, X (gurus) and women (gemer)who were probably taken 8, 8; HSS 4, 86); 96. E-lak-ra(ITT2 638, 12; 640, 19; as booty from Susa. Does lzi-Hu-a-a-a (MVN 11, K, 875, ix, 7'); the last document mentions an unloca- 42, und.) refer to an Elamite group?-For a field, 97. Elam- village, canal and granaries of Elam people see 1.0 lised Elamite group (Elamli-B/Pu-gdrki-me); mu (HLC 3, 384, ii, 14); 98. - (HSS 4, 149, 4; TuT 94, and Gregoire, Prov. mer., pp. 1.6.11.17.32.41. Cf. iii, 25; iv, 15); 99. - (STA 9, ii, 23; 10, viii, 8; ix, 16); 1.3.4.5a. 99a. E-lam-ma(an orchard on the Girsu Canal, ITT 6967, 15); 100. na4Elam-iadga (MTBM 154, 4, but 1.3.2. Umma 1. D/Tan-kal-la (S 34; Sigrist, Princeton 263, 2); 2. index, s.v. has: -"'a6-";no copy is published; mentioned together with U-ba-a, poss. the ensi of Hu-un-kap-ku(S 36; SET 243, i, 26); 3. D/Tan-ni-gi4(S Adamdun); 101. Ga-an-za(DAS 199, 3); 102. na4Gul- 40; A UCT 2, 173, 4); 4. Hu-un-ga6-an(S 40; AOS 32, H su-ga (TuT 201, 15); 103. Hu-ba (WMAH 228, 5). 1, 3); 4a. The field of Hu(?)-ba-nais mentioned in S 104-107, all named Hu-ba, are mentioned in Jean, 44/42 (Sigrist, Syracuse 149, 2; cf. CST 883); 5. HuRA 19 (1922): 104. (p. 32:91, 15), 105. (p. 39:1, 3), un-dar-a(S 45/AS 2; Schneider, Or. 47-49, 249, 59); 106. (p. 43:95, 8), 107. (p. 43:98 r. 9). 6. An-za-za-a(fem., c. S 45; FAOS 17, 94, 2; cf. p. 284); 108. Hu-ba-a (SET 209, 8); 109. Hu-ba-la(MVN 11, 7. Ur-dba/ma-[d]a-[ti?]-[nal (?) (AS 1; Gomi, MVN 12, P. 22); 110. Hu-ne-gdr-ra(TuT 204 r. i, 7); 111. Hu-un- p. 8 ad CST 554, 5); 8. Ad-da-napi-ir(AS 1;BIN 5, 106, gi-si (?E. Szlechter, RA 59, 1965, p. 112, 7); 112. Hu- 3); 9. Ur-nig9gar(AS 4; from Umma? AOS 32, C 16, un-za servant of 113. E4ag-gu (see Sollberger, TCS 1, 57); 10-15. A-bi-in-za,Ba-mu-ga-ak, Sa-ki-in-zu,Sa-ku-bap. 133:326 ad 339); 114. Hu-si-in (TuT 192 r. i, 7, li, A-da-me-naand Hu-b/pu-ul(AS 5; TCL 5, 6039); 16. provided it is not an error for Ri-); 115. I-lam-ma Hu-un-du-du(AS 5; BIN 5, 109, 7); 17. A-al-li/li-muson of x-x-e-ni (AS 8; C. A. Peters and D. R. Frayne, (TuT 160 r. iv, 9; the Elamites' village is mentioned in the same document; the same name is recorded ARRIM 8, 1990, p. 56:10); 18. Ha-ap-ru(AS 9;'AUCT in de Genouillac, Fs. Hiiprecht, p. 141:6, i, 6); cf. 21? 2, 168, 3; BIN 3, 382, 10; YOS4, 289, 8); 19. U-li (AS 116. "in-tah-gi-ir"(ITT 9966, most prob. = "in-tah-si- 9-SS 4; DCEP = HEU 34, 7; Schneider, Or. 47-49, ni", ITT 9972); 117. In-zu (Jean, RA 19, 1922, p. 462, 2; SNSA 453 r. ii, 9'; 500, ii, 10; YOS4, 260, 59); 44:118, 14); 118. Kap-ku-ku(ITT 3161); 119. Ku-ku-a- 20. U18-ba-a(SS 3; MessT 264, 6); 21. Kur-da-gu(SS 3; lum (connected with Nippur, MVN 7, 419, 7); 120. MessT 280, 4); 22. Napi-ir-da-szi(SS3; TS 66 r. 2); 23. Kur-in-tak(Jean, RA 19, 1922, p. 39:6 r. 3); 121. Li-ti- Hu-ti-ni (SS 3; CST 822). 24. Hu-ku-ni (SS3; AnOr 7, um (ITT 679); 122. L2-dnin-Jubur(ITT2 879, iii, 31); 306, 40; MessT 135, 8) 25. Hal-ti (SS 3; MessT 130, 123. Me-ri-ig(connected with Adamdun, MVN 7, 9 r. 30); 26. Za-an-na-za(SS 3; wife of dNu-mul-da-an-dul, 3; 386 r. 11); 124. Nagi-kal/tan is mentioned in con- AUCT 3, 373, 5. seal); 27. Ad-da-ri(SS 3; MessT 278, nection with Elamites (MVN 7, 260 r. 8); 125. PI-PI 4); 28. Hu-ba-a(SS 4; CST 841; MessT 517, 6; Owen, (or WA-WAetc., ITT 952); 126. Ri-ge(TE 55, 26); 127. Mesopotamia8-9, 1973-74, p. 158: 14, 12); 29. Ui-ba-a Si4-ni-gi-ba(MTBM 119, 4); 128. Tab1gu-ri (con- (SS 4; MessT 55, 20; 162, 30). 30. Hu-ba (SS 4; MessT nected with Anshan, MVN 7, 296, 3); 129. Tup-ki-na- 185, 2); 31. Si-si (SS 4-6; see McNeil, MessT, p. 78:19 mu (TE 55, 18); 130. U-ba-a(Jean, RA 19, p. 41:56, ad 1, 22); 32. Hu-un-ba-a(SS 5; MessT 450, 3); 33.

42

JOURNAL

OF PERSIAN

Hu-ba (SS 5; CST 835; SET 231, 19); 34. Hu-a-ni-is' (SS 5; MessT 51, 17); 35. Hu-hu-ni(SS 5; CST 835); 36. Ha-an-ur-a(SS 5; Owen, Mesopotamia8-9, 1973-74, p. 156:12, 8); 37. Ba-ra-li-in (SS 5; MVN 14, 425 r. 1; PN?); 38.39. Gime-dSara, Gu-a-a(SS 5; SNSA 503, 15 f.); 40. Su-ti (SS 5.6; MessT 59, 10; 93, 25); 41. Hu-ba (SS 6; DCEP 253, 14; MessT 476, 3.10; 477, 2); 42. Zana (wife of Ur-ni9-gar;SS 6; SNSA 509, ii, 4); 43. Adda-na-x(x)(SS 6; TS 102, 3; needs collation); 44. Undu-du(SS 9; SA 74); 45. Hu-ba (IS 1; MessT 505, 3); 46. Hu-un-gi-[la-ak(?)](IS 1; SA 71, ii, 8; cf. MAD 3, p. 130); 47. D/Tan-nu-ri, 48. Hu-nu-ha-ra (both IS 1; AUCT 3, 259, 13.41); 49. Hu-hu-ni (IS 2.3; CST 864; DV 5, 365, 11). 50-65 are recorded in documents which do not mention a year: 50. Ad-da-gaba(CST 728); 51. Ad-na(pi/pi)-ir (MessT 319, 9; cf. p. 200, n. 29); 52. A-da-napi-ir (Schneider, Or. 23, 494, 6; are both names homonymous with 43?); 53. Dagu-du-sa6 (AUCT 2, 127, 11; 276, 7 [both-(du)-], 3, 245, 5; 246, 7 [-(du)-]); 54. E-ze-me-na(YOS4, 289, 7); 55. E-zu-na-pi-ir(MessT 178, 25); 56. Gu-ga-at-
STUDIES

MVN 15, 325); 11. Ipu-ga-ru (S 46; A UCT 1, 261, 8); 12. Ha-nagu-ni-ir,

13.

(S 46; MVN 12,

Za-na-i-li-ir 108, p. 202, n. 39); 14. 125, 4; see Steinkeller,JAOS 46/AS 3; Hallo, HUCA 29, 1958, p. 107 f.: (S U18-ba-a 1); 15. Hu-un-ha-ap-ur(S 46.47; PDT 168; SET 51, 18; cf. CSAU, p. 49; -Ur, Nesbit, SRD 9); 16. Hu-un-gi44a(S 47; a[k] (S 47; CTNMC 13, 8); 17. Su-gi-ri-ni In-ri MVN 18. 15, 309); Kutscher, Wadsworth); (S 47; 19. Hu-un-ha-al-bi-[d]a(S 47; S. Lieberman, JCS 22, 1968, p. 57, 6); 20. Igi/gi4-ru (-?) mah (S 47-AS 9; A UCT 2, 384, 9; Forde and Flaugher 58, ii, 21; UDT 112, 8); I-gi4-ri(-?)mah (SS 1; PDT 336, 7); 21. Hu-ba-a (S 48; FAOS 16, 1243, i, 5; r. iv, 10; Forde and Flaugher, 3, 8; undated); 22. Hu-un-dapi (Shulgi's time? MVN 13, 695 26; PDT 529, x, 23). In the latter source he is described as Elam. The same designation is borne by 23. Da-hu-un-ba-an(MVN 13, 695, 24; PDT 529, x, 30), 24. Se-er-ra-ge-er (MVN 13, 695, 27; PDT 529, x, 26), 25. Za-mi-du-uk(MVN 13, 695, 28; PDT 529, x, 28), 26. Ha-ar-sa-ti-ip,27. Pu-ra-an-haal-bi-it(ibid. 33.34) and 28. Se-il-ha-ki-4i-ba-as (ibid., 35). The names of the other individuals with the same designation are too short for an unambiguous affiliation and interpretation (29. Ga-ti, 30. Ki-ip-ti, 31. Se-er-tiMVN 13, 695, 30-32; PDT 529, x, 31.34); 32. (PDT 529, vi, 28); 33. Hu-un-ha-ap-ur(AS Se-ilx-ha 1; S. Langdon, Babyloniaca 7, 1923, pl. 19 after p. 243: 3 r. 3); 34. Hu-ba-a(AS 1; Watson, Cat. 40, 7); 35. Hu-un-hu (AS 1; PDT 422 r. 1); 36. Pi-li (AS 1-prob. AS 8; FAOS 16, 959 r. r. viii, 20); 37. Hu-un-ha-ap-ur (husband of Arbitu; AS 2; BIN 3, 47, 3; 48, 9; FAOS 16, 958, ii, 17); 38. Hu-ba-a (AS 2; FAOS 16, 958, ii, 41); 39. Ha-ap-ru-se-er(AS 2-7; A UCT 2, 73, 2; CT 32, 26, i, 12; FAOS 16, 958, ii, 41; D. C. Snell, ASJ 9, 1987, p. 266: 70, 7; TrD 29, 19; UDT 91, 329); 40. Ha(AS 3; ap-rupu (AS 3; A UCT 1, 415, 2); 41. 1-gi4-ru-um A UCT 2, 99, 11); 42. Hu-un-ki-ip-ri(AS 4; CTMMA 1, 17, 71; TCL 2, 5508, ii). 43. Hu-un-z-ri (AS 4; TCL 2, 5508, ii); 44. Hu-ba-a (AS 4; D. C. Snell, ASJ 9, 1987, p. 267: 73, 3); 45. D/Tan-ha-a-ah (AS 4; CTMMA 1, 17, 70); 46. Hu-ba (AS 4; TCL 2, 5563 r.; 5584 r.); 47. Hu-un-dul-gi (AS 4-SS 9; AUCT 3, 198, 33; 211, 5; Schneider, Or. 47-49, 136, 14; SET 103, 21; Sigrist, Princeton 81, 3; TCL 2, 5504, ii; YOS 4, 107, 3); 48. D/Tan-bi-si (AS 5; TS 3, 8.r. 3); 49. Hu-un-ga6(AS 5; 1.3.3. Puzrish-Dagan BIN 3, 351, 1); 50. Ga-an-za(AS 5; Archi and Pomponio 1. A-gu-gu-ni(S 32; TrD 59, 10); 2. Hu-a-a (S 33-SS 60 r. 7); 51. Muiir-dulo-gab(AS 6; SET 63, 44); 52. Li8; CST 152.448; PDT 408, 7; TCL 2, 5504. 5584 r; R. ni-si-in (AS 7; FAOS 16, 1249, 3); 53. Hu-un-nu-du-uk D. Biggs and R. L. Zettler, ASJ 12, 1990, pp. (AS 8; A UCT 3, 318, 4. seal); 54. Nu a-an (FAOS 16, 25.41:10, 10; A. R. George, Iraq 41, 1979, p. 125:3, 7; 959 r. viii, 20); 55. Hu-ba-a (both prob. AS 8 acc. to T. Gomi, BJRL 1981, p. 111:68, 4); 3. In-tar-ri(S 34; Owen, WO 22, 1991, 182 f.; FAOS 16, 959, iii, 24. r. AnOr 7, 147, 18); 4. Inda-a (S 40; AOS 32, 22, 6); 5. Si- vii, 42); 56. dsulgi-un-ha-ni-ig(AS 9; 2038 B.C.; A UCT (AS 9; MVN im-ti-ip-ha-le-er (S 43; E. Dhorme, RA 9, 1912, p. 57: 3, 200, 4; BIN 3, 546, 21); 57. Ra(?)-l4u-ut SA 3, 4; cf. MVN 15, 9); 6. D/Tanda-h[ar] (S 431AS 4; 3, 243, 2); 58. La-algu (time of AS; YOS4, 246, 104); BIN 3, 491, 68); 7. Hu-ba (S 44; AUCT 1, 696, 4); 8. 59. dSulgi-dan/da-anga-da (AS 9-SS 2; Pinches, HuJun-haal-bi-it (S 44; CST 127); 9. Naahpi-ir-ti (S Amherst 119 r. 2; 120, 4; TAD 9, 6; went to Kimash 44-poss. S 47; STA 36, 4; TJAMC-FM 46, 18; cf. acc. to the second document which is not dated); CSAU, p. 53); 10. La-algu-ni's anonymous son (S 46; 60. Hu-ba-nudu (SS 1; Watson, Cat. 57, 5); 61. Su-gu-la

ELAMITES

AND

OTHER

PEOPLES

FROM

(SS 1; PDT 336 r. 4); 62. U,$-ba-a(SS 1; FAOS 16, 1202, 3); 63. Si-im-lu-ha-se-er (SS 1; MVN 3, 338, 28); 64. D/Tan-su-gu(SS 2; SETDA 183, 30); 65. Mu-ir-tiga-ba (SS 3; Nik. 1, 476, i; di(n)-GABA,CT 32, 36: 103403, ii, 9); 66. U-ba-a(SS 4; TCSD 285 r. 2); 67. Hu-ba (SS 4; found at Nippur; NATN 914, 4); 68. Huun-dulgi (SS 5; FAOS 16, 1073 r. 5); 69. l-ri-dah(SS 5; Sigrist, Princeton, 75, 2); 70. Hu-ba (SS 6; MVN 15, 142); 71. D/Tan-la-ah(SS 6-IS 3; SET 290, 10; TrD 62 r. 3); 72. Hu-un-ha-at(SS 7; CT 32, 34, i, 10). 73. Ra-bi-se-er(SS 8; MVN 3, 283, 6); 74. Hu-nu-nu-ur(SS 9-IS 2; A UCT 3, 292, 1; MVN 15, 281; PDT 261, 5; his seal, BIN 3, 602, 25; also in: FAOS 16, 1043, 2; TCL 2, 5536. 5553; see P. Michalowski,JCS 26, 1978, p, 162 ad 166:2); 75. Ru-hu (IS 1; CT 32, 15 r. 3); 76. Ur-ku-ku(IS 1; MVN 15, 267); 77. Hu-na-zi (IS 2; in broken context; Schneider, Or. 47-49, 126, 4; TRU 342, 5). 78-100 are recorded in documents whose years are either not known or not reported (78.79 are contained in UDT 91; 84-88 are possibly from Umma according to Pomponio, ZA 79, 1989, p. 10): 78. Ga-an-za (186); 79. Kisr-da-'u (186.199); 80. DITan-hi-tum (AUCT 2, 381, 12); 81. Hu-ba (AUCT 1, 321, 15); 82. - (SA 63, 4); 83. Hu-lu-du-uk(TRU 8, 12); 84. Hu-ba-dsulgi-da-as is mentioned in an account of 85. Hu-ba-i-ha--ne-es(BIN 3, 315, 2 f.); 86. D/Tan-mi-zi,87. Si-in-si-da-ak,88. Se-ilx-ha (ibid. 624, 11 f.); 89. Sim-dagu-ni(ibid.623, 8; cf. Sim-da-ku-nu,MVN 15, 142, 41); 90. "Hu-un-ha-ap-ri"(? CST, p. 11 ad 507); 91. Hu-un-z ulgi (FAOS 16, 915, 5); 92. - (Forde and Flaugher 3, 18); 93. Hu-un-[...] (mentioned together with other non-Mesopotamians; FAOS 16, 804, E. 7); 94. La-la-ha(Susian; FAOS 16, 1313, 1); 95. Li-ik-du-ni(FAOS 16, 913, i, 9.19; ii, 13.32; r. iii, 5; iv, 1: -ni)); 96. Mu-ir-ti(n)ga-ba (TRU 256 r. i); 97. Su-ganu-um (TAD 59, 11); 98. Hu-un-ha-ap-ur(TJADBIES 337, 9 f.; 337A, 4; from Puzrish-Dagan?); 99. Pu-raan-hu-ut-ir-ra-an(AnOr 1, 29, 2; PN?); 100. Si-im-ti-napi-ir (Puzrish-Dagan or Umma? YOS 4, 227, 4 f.). 1.3.4. Nippur 1. Hu-b/pu-ul-si-mu-ut (S 25126/32/44/45/IS 3; from Nippur? TuM NF 1/2, 236, 2 f.); 2. A-ku-ni(son of Nagi-bil-tum, S 38; from Nippur? AOS 32, TT 4, 13); 3. Da-an-zi-pu(AS 1; from Nippur? FAOS 17, S. 1, 21). 4. La-an-ku(SS 2; Sigrist, Princeton571, 3). 5. The son of Si-im-ti-"gusir"came to Nippur in IS 1 (TAD 28, cf. p. 20). Grain from the field of 5a. Ku-ki-nidulo was destined to Nippur sometime in Shu-Sin's reign (DAS 23). Lafont (p. 30 ad loc.) compares it with Ku-kur6-i-dulo (UNL 112, p. 44:503). The field of Ku-ki-ni-duis mentioned after Gidaha in DAS 303 from IS 1. 6. Pi-li-ri-ni (son of Su-e-'4tir; SS 7; AOS 32, TT 11); 7. Kurda-gu (IS 1; br./f. of Lli-Nanna and Su-Ea, see Steinkeller, FAOS 17, p. 212 ad 37, 5; Fish, Journal of the ManchesterEgyptian and Oriental

IRAN

AND

THE

PERSIAN

GULF

REGION

43

Society 15, 1930, p. 1072: 2, 4); 8. Za-na (fem.; IS 2; NATN 145, 1). Was the village of Su-sa-nuimk' (IS 3; NATN 320, 1) named after people from Sus(ian)a? 9-11 are recorded in documents which do not indicate the year: 9. A-na-in-da-s'i(NATN 983, 9); 10. Hu-un-nu-ri(B. Lewis and E. R.Jewell, ASJ 4, 1982, p. 58: 18, 2); 11. Zu-zu-gu-ni(T. Fish, Iraq 5, 1938, p. 176: 26, 5). Cf. 1.3.1.119; 1.3.3.78.79 and 1.4. 1.3.5. Ur All the material is from UET 3 unless otherwise indicated; 5-9 occur in documents which do not indicate the year: 1. Am-ma-za-za,a female slave of Aadugga (Shu-Sin's reign? 51; cf. H. Neumann, AOF 16, 1989, p. 229); 2. D/Tan-ne-ki(SS 9; AOS 32, P 4, 4); 3. Kitr-da_-u(IS 6; UET 9, 248, 2; 322, 4; cf. 303, 3?); 4. Ku-ku-da-i_(IS 9; 1048); 5. E-ze-me-ni(1452); 6. Huba-an(? UET 9, 135, i, 9). 7. Hu-un-hu-up-se(SNSA 541 r. 12); 8. S/Za-numson of Kiir-da-su(1414); 9. Su-mu-siin-ti (1040). 1.3.6. Place Unknownor not Indicated 1. (S 35; DCEP 214, 4); 2. Hu-hu-um-ti(S I-pu-ga-ru 558, 2); 3. A-ku-ni (S 40; SAKF 103 [tablet], 35; PDT 13); 4. Iddin-Sin son of Hu-un-hu-up-se(AS 2; A. Goetze,JCS 17, 1963, p. 18; n. 82: YBC 3635); 5. Huun-da-ri (AS 3; AUCT 1, 383, 6). 6-13 are recorded in documents which do not indicate the year: 6. "Ad-da-na-pi"(ir)(? Huber, PNN, p. 44b: Cohn 2, i, 3); 7. Ba-al-li-ri son of Ra-[(x?)]blpu-um (Delaporte, Cat. Bibl. Nat. 94); 8. E-pu-ga-ru (AUCT 1, 442, 20); 9. Hu-ba (DCEP 35, 20); 10. Hu-unna-zi (A 5477, unpubl., cited by Gelb, MAD 3, p. 130); 11. Lu'-tinanna(son of Hu-ba,DCEP 497, vi, 9); 12. Si-a-a(originally from Susa, R. J. Lau, Old Babylonian TempleRecords,New York 1906, 22, 7); 13. Tarna-pi-ir (Delaporte, Cat. Bibl. Nat. 68). 1.4. Summary The relatively scanty pre-Sargonic documentation concerning Elamites contains information not only on Elamite slaves, but also on such freemen who resided in Sumer and participated in the commerce between Lagash and Elam. Very few of them who acted in such a capacity might have attained official status (compare perhaps the situation at Shuruppak). The Elamite slaves were not necessarily prisoners of war, but were bought by dealers from certain Elamite locales which might have been situated not far from Sumer. No direct connection between Mesopotamian anti-Elamite wars and the appearance of Elamite slaves in Sumer can be established. On the whole, it seems that the preSargonic Sumerians had relations mostly-if not exclusively-with certain western regions of Elam,

44

JOURNAL

OF PERSIAN

with the hypothetical exception of Bashime which might have been situated somewhat more to the east on the coast of the Persian Gulf. The latter region was fairly easily accessible by boats from the Gu'abba. There is no evidence from nonmythological sources that Sumer had relations with more remote regions of the Iranian Plateau such as Aratta.13 Items of material culture which ultimately originated there could have reached Mesopotamia through Elamite transit trade. The Sargonic military actions and conquests in the Iranian Plateau inevitably widened the geographical horizon of the Sumero-Akkadians and caused an unprecedented intensification of the relations between Mesopotamia and Elam. There is more documentation of Elamites in several Mesopotamian urban centres. In addition to people from Western Iran, individuals from more eastern regions such as Anshan and Marhashi are recorded. The Elamites belonged to various strata of the population. However, in several cases it is difficult to determine whether the Elamites in question actually resided in Mesopotamia. It stands to reason that many-if not most-of the Elampeople, who bore Mesopotamian names, originated in western Iran which was influenced by the Mesopotamian culture. It is difficult to establish any correlation between the information on the 340 individuals (maximum; minimum: 269, cf. Tables nos. 1 and 2) presented above (1.3; for geographical distribution see Table 1) and what is known about the developments in the relations between the Neo-Sumerian state and the Elamite political entities (the former's vassals, see T. Maeda, ASJ 14, 1992, pp. 143 ff.). There were wars between Sumer and various political entities of Greater Elam in S 27, 34, 46, 48; AS 6, 7, 8; SS 7; and IS 9, but the intermediate periods were presumably relatively peaceful (cf. HE, pp. 16-22). The following list presents the chronological distribution of the explicit, assured, probable and possible Elamites (no. of individuals in brackets): S 25 (1), S 32 (1), S 33 (2), S34 (2), S 38 (1), S 40 (1), S 43 (2), S 44 (3), S 45 (3), S 46 (7), S 47 (2), S 48 (2-13), AS 1 (8), AS 2 (4), AS 3 (1), AS 4 (3), AS 5 (2), AS 8 (4), AS 9 (18), SS 1 (5), SS 2 (4), SS 3 (4), SS 4 (6), SS 5 (8), SS 6 (4), SS 8 (1), SS 9 (1), IS 1 (1), IS 2 (2), IS 3 (2). Only a comprehensive study of the Messenger Documents may establish such a correlation. The Elamites who bore non-Elamite names, as well as the anonymous ones, were not necessarily ethnic ones, but originated from Susiana and adjacent regions where the Sumero-Akkadian culture was always discernible. The occupations andlor status of many of the 3401/ 269 individuals under consideration here are mentioned in the documents. Only a minority of them were certainly Elamites. The remaining Elamites,

STUDIES

who were probably or doubtfully such (cf. 1.0), are preceded by ? in the following statistics. 2 ? 5 individuals (mostly females) with Elamite names (1.3.2.14 ? 10-13.15) are described as prisoners of war. Except for three individuals (males: 1.3.2.9; females: 1.3.2.6; 1.3.5.1) and 1.3.3.12.13, who were probably donated by the Shimashkian ruler Guriname/Kirname presumably to the Ur III state, none was clearly a slave. Two individuals (? 1.3.2.47.48) belonged to a large gurus-group. The group, which received oil rations, was-if to judge from its members' names-ethnically heterogeneous. 1.3.1.10 + 12.98; ? 1.3.2.50 possibly received rations as well. A group of no less than 454 gurus-workmen which is recorded in a document from AS 5 (from Umma?) was presumably Elamite according to Jones and Snyder (SET, p. 155 ad 264, 14). Three anonymous Elamite prisoners of war were employed as gurus-workmen according to a document from Lagash (ITT 6175, see Oppenheim, AOS 32, p. 19). ? 1.3.5.5 is mentioned in a roll of workers. Anonymous Anshanite workers are recorded in MVN 15, 162 (Umma, S 33) and Lutz, UCP 9/2, 77, 8 (SS 4). Most individuals who received rations acted as messengers (cf. below) ? 1.3.1.112 was an unspecified servant/dependent and + 1.3.3.64 was a kizui-servant.1.3.3.60 was a musician (NAR). 1.3.2.5 is recorded in a document dealing with shipbuilding and building material, but it is not known whether he was somehow involved with the river transportation between Sumer and Susiana (cf. 1.1 and ad 1.3.1.97 below). 1.3.2.53 and 1.3.3.9 (? 2) were bakers; 1.3.4.3 was a miller. Several Elamites were shepherds (1.3.2.4; cf. HLC 2, pl. 83, 73, 6, 3; TuT 37) or are recorded in documents concerning sheep and cattle (1.3.1. 6.21.53.54.81.99.118). 1.3.2.8 is mentioned in a document concerning shipment of various leather goods and vessels. 1.3.3.71 was probably also a leather worker. One individual (1.3.1.57) was an agricultural overseer/cultivator (engar, cf. presently) in charge of the gana-gu4 ("land worked by the oxen"; belonging to a great household acc. to Foster, Mesopotamia [Copenhagen] 9, p. 44) of Lui-dba-ba6(the nu-bandagu4-official, cf. Pettinato, AnOr 45, p. 39 f. ad 59, iv, 24). In J. M. Sharashenidze's opinion (VDI 1984/4, pp. 98 ff.), the engar (also 1.3.1.136 "ploughman") was an independent cultivator possessing landed property. At least three fields (1.3.1.4a.1 la; 1.3.2.4a), an orchard (1.3.1.99a), a granary and a store(house, cf. 1.3.1.4a) seem to have been named after Elamites (cf. also 1.0). It is not without interest that such strangers (if not merely their descendants) seem to have been allowed to possess landed property in the heart of Sumer. Yet it must be remembered that the evidence is not explicit and unambiguous. It is clear that Elamites resided in

ELAMITES

AND

OTHER

PEOPLES

FROM

the centre of Sumer (note especially 1.3.1.50 who explicitly acted at Girsu). Some of them assimilated to the local population if to judge from the Sumero-Akkadian names of their family members 1.3.5.8; 1.3.4.2.6; 1.3.3.37; 1.3.2.26.42; (cf. 1.3.6.4.11). 1.3.1.104 ? 1.3.3.50 served as dub-sar ("scribe", actually a middle level bureaucrat acc. to Michalowski, SAOC 46, p. 62), ? 1.3.5.7 acted as a judge and 1.3.5.2 was a comptroller. However, it should not be forgotten that Akkadian names were not uncommon in Susiana and some other western regions of Greater Elam. Note that Si-im-ti-ip-ha-seer (1.3.3.5), whose name is certainly Elamite, is recorded in a document concerning offerings to Mt. Bishri (hur-sagBa-far) which might have been the deified original region of the Amorites far away from Elam. The following individuals acted as functionaries (most in Messenger documents; cf. Sigrist, Mel. Steve, pp. 51 ff.): gabra(1:1.3.3.4);14sukkal (1:1.3.1.18; ? 11: 1.3.1.94.96 [also lh-i~'tukulgu-la].119.127.132; 1.3.2.30; 1.3.3.56.57.59 and poss. 53; - GAB: 1.3.2. 32; cf. M. Yoshikawa, ASJ 10, 1988, p. 231 f.); aga•-is (15 ? 2:1.3.1.24.26.27.29-40 ? 109; 1.3.3.48; 1.3.1. 16 had his own agai-is), aga'-isgalgal (2 ? 1: 1.3.1.1 ? 4.124), tukul (3 ? 4: 1.3.1.25 ? 26.51.54.55. I•-i-' 105 ? 131), lti-~' tukul-gu-la(1 ? 5: 1.3.1.41 ? 85. 87.96 [also sukkal].110.125); magkim(2: 1.3.2.18 [also rdagabalike 1.3.1.107.126]; 1.3.3.70); ugula (7: 1.3.1. 11.115; 1.3.2.69; 1.3.3.11.74.89.92), ugula of the millers (? 1: 1.3.3.11); nu-binda (1 ? 5: 1.3.1.101; 1.3.3.35.36.47.54.75; cf. 1.3.1.95); nam-10 (? 1: 1.3.1.52). 1.3.2.61 (not necessarily Elam.) was the sakanakkuof Umma (see Steinkeller, SAOC46, p. 25, n. 17). 1.3.1.44, who received rations in SS 1 (DAS 243, 23), was (has just become?) a magkimin the following year and a royal agaids gal-gal in IS 1.2 (DAS 7, seal; 11, seal). In the last two years he acted as the fourth judge of a collegium (ITT 6032 = NSGU 224, 7; see A. Falkenstein, NSGU 1, pp. 35.48:9; 3, p. 42). He is recorded without indicating his function in many Messenger Documents (cf. n. 12). At least 2 ? 2 individuals possessed seals (1.3.1.125; 1.3. 3.16; ? 1.3.2.19; 1.3.4.6). ? 1.3.1.113 was master of ? 1.3.1.112. 1.3.3.78.79 are mentioned in documents recording offerings for the Enlil temple at Nippur, and 1.3.5.3 in a document concerning evening offerings from the king in E-Nannar. The title ui-kulborne by 1.3.6.12 is of unknown denotation (+ 1, see Goetze,JCS 17, p. 25, cf. e.g. Da-da itkul, Arnold, 10, 2). Is it just a coincidence that it sounds like the element listed in EO, no. 150c? It is likely that 1.3.1.22 was identical with the homonymous ensi of Sapum (see T. Fish, MCS 4, 1954, p. 90). Was 1.3.1.67 identical with a homonymous individual from Uru-az who bore the same title (RTC 359, 3)? It is doubtful whether those who delivered livestock and offerings at Puzrish-Dagan

IRAN

AND

THE

PERSIAN

GULF

REGION

45

(most of them did.)15 actually resided in Mesopotamia (29.41% of the 340 individuals are recorded at Puzrish-Dagan). After all, Puzrish-Dagan was not a normal settlement, but a collecting point for cattle. The fact that very few Elamites are recorded in nearby Nippur (where most of the relevant documentation belongs to family archives, see M. Civil, SAOC46, p. 44) compared with their frequent mention in Puzrish-Dagan supports the assumption that the Elamites of the PD texts were not resident in Sumer. A case in point may be the messenger (sukkal) Ra/ibhuti who always acted together with the dignitaries of Harshi. Did he belong to their entourage? (cf. Hallo, HUCA 29, 1958, p. 78, n. 15 ad 6, 19; Owen,JAOS 108, 1988 p. 115). In this case he did not live in Sumer. He might have acted on behalf of his fellow countrymen like the anonymous Elamite (1.3.1.97) did for his compatriots according to HLC 3, 384, ii, 16f. Another clear case is Pi-ti-na-RIElam An-ga-na(in the Messenger Document MTBM 66 from AS 9).1.3.2.22 might have been identical with his namesake (cf. Michalowski, SMS 2/3, 1978, p. 13 f: 8, 2) who resided in Elam. 1.3.6.12 and presumably 1.3.1.3 were Susians. 1.3.1.84.95.124.128; 1.3.3.56 were connected with regions of Greater Elam, like Hu-ba-a the gakkanakkuof Sium (1.0). Cf. Steinkeller, SAOC 46, p. 34. Groups of anonymous Elamites frequently visited Puzrish-Dagan (e.g., people from Harshi in SS 7 acc. to MVN 5, 127). Many messengers (cf. above), if not most of them, probably were not residents of Mesopotamia: for instance, MVN 15, 142 (Shulgi's time) records, inter alia, a Susian, lI-Si-ik-risumi (in Greater Elam), lu-dulo-iki (location unkn.), two SU(A)-people (Ga-ra-da-tu and Pi-ip-ra [the name is either Elam. or perhaps Sum. bibra, a kind of bird]) and other foreigners (very few individuals in this text have Mesopotamian names). The fact that most of the Elamites from the Lagash and Umma sources (60.87% of the 340 Elamites are mentioned in these sources) are recorded in the socalled Messenger Documents (cf. M. Lambert, TE, p. 9f.; a misnomer retained here just for convenience sake) would considerably reduce the number of Elamites residing in Sumer. The abovementioned named Elamites were only a very small percentage of their recorded compatriots. Numerous anonymous Elamites are recorded in the abundant corpus of the Messenger Documents, e.g., (number of individuals in brackets) from Shimashki (25; DAS 91), Duduli (30; DAS 122) and Itnigi (40; DAS 93) as well as unspecified ones (36; DAS 90). These four documents are dated to AS 9. There are also such reports from other years (e.g. S 34 [MVN 7, 3], AS 1 and SS 9). Yet, most of the Elamite groups are recorded in Messenger Documents without indication of their number (e.g.,

46

JOURNAL

OF PERSIAN

STUDIES

and the latter an atypical name. I-la-mi-im("Elamite"; gen.) and A-at-ta-a(s. of U-dapa) acted as witnesses at Kutalla in Rim-Sin's time (Strassmaier, Warka 5: B 36a, 18; 17: B. 44, 29). Is U-ta-a-hu(f. of Ipus-Enki, ibid.,index, s.v.) an Elamite name (cf. EO, nos. 20.56)? Ahi-'agis and his father Samag-tiiram (s. of Warad-Sin), all bearing Akkadian names, were influenced by the Elamite cult: they worshipped Suginak at Kutalla (see Charpin in O. Tunca [ed.], De la Babyloniea'la Syrie,en passant par Mari. Melanges offertsa MonsieurJ.-R. Kupper a l'occasionde son 70e anniversaire, Liege 1990, pp. 64, 77). Si-ir-ma-pu[uk(?)] (f. of B/Pur-ri-ia, see C. B. F. Walker in D. Collon, Catalogue of the WesternAsiatic Seals in the British Museum, CylinderSeals III: Isin-Larsa and OB Periods, London 1986, p. 108 ad 179; cf. NABU 1990139) was servant of the Mesopotamian god Sin. On architectural connections between early Sukkalmah Elam and early OB Babylonia cf. E. Carter, HE, pp. 146, 149. 163 f. (material culture). It is noteworthy that Ellu(?)-Ningal and Ku-uk-giga-at (Iran XXV, p. 9) are mentioned in documents of the same provenience (locus) in Ur. Ku-uk-gu-ku (not -Ilu-Ju-ma!)son of At-ta-napi-ir acted as a scribe (1821 B.C.; also M. Anbar and M. Stol, RA 85, 1991, y 32 f.: 21, seal e). He is described as "servant of Sa-at-WA-ak"(prob. an Elamite deity). The appellative hu-uh-pu-umfrom OB Ur (UET 5, 795, i. 11; cf. AHw., p. 353a: "ein Bronzegefass") is originally Elamite. In von Soden's opinion (OLZ82, 1987, col. 495 f.), OB Di-r-Su-um-mu-uk-ri refer(-Si-im-mu-uk-ri) the Larsa a settlement in of to (cf. kingdom ring WO 16, 1985, p. 45, n. 110), is an Elamite name. For an Elamite at Sippar-Amnanum see K. van Lerberghe, MRl.Steve, pp. 151 ff. Si-mu-'zis recorded at Uruk in Irdanene's reign (1816-1810 B.C.; Sh. Sanati-Miiller, BaM 21, 1990, p. 179:125, 4). Am-ma-ha-tum (SA 170; the first two signs need col1.5. Further Evidencefor Elamites in Old-Babylonian a list of females receiving rations) and A-kulation; Sources (cf. Iran XXV, pp. 1 ff.) un (f. of B/Pu-nu-ma, AUCT 4, seal 68) are from The designation Elam is recorded in Ishbi-Erra's Larsa(-region). The provenience of the following three reign (BIN 10, 45, 8; 124, 3). Ri-ib-gi-mu-utis mentioned in the same reign (BIN 10, 157, 7; G. Th. Fer- individuals is unknown (except for the third one, werda, TLB/SLB 5, 18, 21). D/Tan-nu-ri (BIN 10, 47, they bear atypical names): Hu-pi son of U-tu-na 4) and Hu-un-te-[bi] (BIN 10, 46, 7, a hun-name?), (Elam.?) was servant of Adad (late OB, Hallo in B. who lived in Isin during the same reign, were Buchanan, Early Near Eastern Seals in the YaleBabyloperhaps Elamites as well. Sim-ge-il-haand Si/Se-il-ha nian Collection,New Haven and London 1981, pp. (the former is described as Elam) are recorded at 366.374.461: 1058). In-da-da-ason of Ra-bi-bi(cf. EO, OB Eshnunna (see R. M. Whiting, AS 22, p. 29, n. no. 193) worshipped Enki (OB seal, Moortgat, VAR 90). On cultural-artistic ties between Elam and 120). Su-mu-un-da-ak-ge-erhad relatives bearing southern Mesopotamia at the end of the Isin- Akkadian names (AbB 12, 195; see W. van Soldt, Larsa period cf. Ladders to Heaven, p. 195 (note ibid. p. 151 n. b ad loc.). For tubqumsee J. M. Durand, ARMT 21, p. 523 f. also p. 119 f.). La-ma(?)-ahfrom Lagash (DAS 354, undated) is (possibly near Sippar). For Huba(kO (prob. near described as Elam. Ku-ru-um-miand Gi-ingugu (time Sippar) cf. Abr-Nahrain 27 (1989), p. 155 f. The from Mari was noticed of Kudur-Mapuk dynasty, de Genouillac, FT 2, pl. Elamite PN Ku-uk-ri-ih-tu-uh 58 f.: AO 13004.13007.13009.13010) were possibly by von Soden (OrNS 56, 1987, p. 99; OB; a kukshepherds. The former had a hybrid (Elam.-Akkad.) name). Ku-uk-ku-wa-riis now read Ku-uk-gi-ia-riby women from Sapum, AS 3; Pinches, Amherst,68; Danim-ma-an-zi-at was in charge of people from Sapum acc. to BM 27944, courtesy Prof. Sigrist; Elamites from Sapum and "Elamites of the palace", Umma, AS 7; MVN 15, 191) or the year in which the document was issued. In such cases their approximate number can be calculated according to their allowances. They received rations (barley, flour, bread, oil, fat, meat and beer) while on their way from Ur to their homeland or vice versa (Elamites from Kimash, Marhashi, and "(lti-)Ki-la-tu"received rations at Ur in S 37-AS 7 acc. to UET 3, 90.1401.1633). They necessarily passed the Lagash region (notably at Kinunir south of Girsu, cf. DAS 91 from AS 9) as Lagash was the easternmost Sumerian city state thereby being the closest one to Elam. Irll-nanna the sukkalmah who actually controlled vast territories in the east on behalf of the king (notably Shu-Sin) was based at Girsu (see Steinkeller, SAOC 46, p. 26 with n. 21; cf. Jones and Snyder, SET, p. 123-206, 13). 1.3.125 was his servant according to his own seal. It may be that 1.3.1.125 did not reside in Sumer. Does HLC 3, 384, a record of various boats and their uses, reflect communication between Lagash and Elam via the watercourses? (trade in bulk was possible mainly on water: therefore Lagash tried to control Bashime already during the early dynastic period, cf. 1.1 and Steinkeller, SAOC 46, p. 40 with n. 68). Umma, from which a smaller, yet sizable corpus of Messenger Documents has survived, was next to Lagash regarding proximity to Elam. A definitive evaluation of the relevant evidence of the Messenger Documents cannot be undertaken here, as one has to wait for the final publication of this huge corpus.

ELAMITES

AND

OTHER

PEOPLES

FROM

IRAN

AND

THE

PERSIAN

GULF

REGION

47

Finkelstein, JCS 7, 1953, p. 171 f.: 76, 11; 82, 3) is merely a homonym (alternatively a toponym deriving from elammakku-tree?). N/LB has Sd-ta-ri-in-di(ancestor of Nergal-iddina; the latter was father of Nadin, scribe, Sippar, 562/1 B.C.; R. B. Dillard, Neo-BabylonianTextsfrom theJohn FrederickLewis Collectionof the Free Libraryof Philadelphia (Dissertation, Dropsie University, Philadelphia 1975, 1556, 14); cf. EO, nos. 199.215; Si-mu-d(ancestor of Gimillu; the latter was father of Madanu-ahhe-iddina, Sippar, 545/4 B.C.; H. G. 1.6. FurtherEvidencefor Elamites in Kassite Babylonia Stigers, JCS 28, 1976, p. 41: 30, 12 f.; poss. a (cf. Iran XXV, pp. 13 ff.) hypocoristicon of Simut). Ha-al-ti-kuis not Elamite (F. Joann&s in F. Bruschweiler, RA 83, 1989, p. 158 Unpublished early MB documents from the ad 55, 3.7.10.17 f.), but possibly Aram. "sparing, Diyala region16 contain the following names which for (lt) = /g/see F. M. Fales, can be Elamite: Ki-la-hi, Si-il-hagi and U-da-`a-[x-x] withholding" (to H-S-K; OrNS 47, 1978, pp. 91 ff.). It is not known whether (cf. EO, nos. 56.97a.220b; last name Kass.?) For 4-iraKu-uk-mu-ra-te-from Seleucid Uruk (L. T. Doty in E. as-se (of Kuk--) cf. K. Deller andJ. N. Postgate, AfO Leichty and P. Gerardi [eds.], A Scientific Humanist: 32 (1985), p. 75 f. ad GN Urage.For Elamite-Kassite Studies in Memory of Abraham Sachs, Philadelphia relations cf. J. van Dijk, OrNS 55 (1986), pp. 159 ff. 102 with n. 28: YOS 20, 72, 5 f.) had an A seal-impression from Malyan resembles the 1988, p. Elamite name (with kuk-). Nippur ones (from 1316 and 1297-1283 B.C.) according to M. W. Stolper, Textsfrom Tall-i Malyan (Philadelphia 1984), p. 16 (cf. Brinkman in Stolper, 2. FURTHER EVIDENCE FOR KASSITES IN ibid., p. 50 ad 19, 3 f.: FLP 1313, iv, 6: MB di-mutakMESOPOTAMIA sir in a list of Elamites from Nippur). Mesopotamian cultural influence diminished in Susiana 2.1. The Old-BabylonianPeriod during the latter half of the second millennium B.C. (HE, p. 144), perhaps because Babylonia itself (cf. Iran XXV, p. 16 f. and 1.0 ad 1.3.1.6) was occupied by a non-Semitic element. Is MB/NB For Ku-ru-us-[x(tu?)]cf. Ku-ru-ug-ti(KUB 48, 106, cultural influence echoed by the toponym Temen21; cf. E. Laroche, Hethitica 4, 1981, p. 23: 653a; Marduk-garrani which refers to a settlement in Kassite?). But Kurugtu is explicable in Akkadian Huzistan (cf. WO 16, 1985, p. 46; HE, p. 162 f.)? terms. Expenditures of barley, valued in beer or pihu, for offerings for (deceased?) Kassite kings and personnel are recorded in 1720 B.C. (OLA 21, 20; 1.7. More Elamites in First Millennium Mesopotamia cf. 19, und.) and probably from 1677 B.C. (ibid. 22). (cf. Iran XXV, p. 25 f., n. 54) According to K. van Lerberghe (p. 3 ad loc.) these SB sources mention Elamite figs, wool, algar-harp documents "give us more information on the and a chariot, whereas Elamite carnelian is problem of the integration and the role of the Kasrecorded in NA; note Istar of Elam (for refs. cf. sites in OB society". The names of the Kassite kings AHw., p. 196 f.). CTN 3, 145 (784 B.C., found at are Ti-iggi-il and I-gi-ik-ni-id/t.They are mentioned Sa/Ta-ha-am-di-it-ta. Calah) shows that the Assyrian army used and together with the tu-ur-gu-ma-num manufactured the Elamite bow (iv, 13) and that an OLA 21, 20 also lists Mar-ha-an-duand Mar-ma-adElamite ambassador (I6MAHkurElam-ma--a,iv, 26) [...]. Kassites (ERIN Ka-a'-'i-i) are also recorded in was received at Calah in that year (see S. M. Dalley 1634, 1630 (OLA 21, 61 and 67 respectively) and and J. N. Postgate, p. 256 ad loc.). This information 1600/1599(?) B.C. (VS 22, 77: -Ju-di),as well as in is from a period when there is a gap in the undated documents of the same period (OLA 21, M. Vandocumentation from Elam itself. An anonymous 19; VS 22, 74; Ilu-gu-ib-ni ERIN Ka-as-u-'ud, Elamite received rations (B. Parker, Iraq 23, 1961, dendriessche, OLoP 18, 1987, p. 63 f.: 2, 6; W. A. J. p. 32 f. and pl. 16: ND 2489 r. i, 10; undated from de Smet, Akkadica 68, 1990, pp. 1 ff.). They were the NW Palace, Calah). ADD 930 contains some active in the Sippar region. Does dKada--na-sir information on the Babylonian-Elamite border (A. K. Mohammed, Studies in Unpublished Cuneiform region and mentions an anonymous king of Elam Textsfrom Diyala Region-Himrin Basin-Tall of Rim(undated, cf. Postgate, Taxation, p. 146). ADD 857, M. A. Thesis, Baghdad 1985, 518, 3.14; son H.addid, iii, 11 f. mentions two Elamite prefects (Jaknu) Adad from Tall-Haddfid = Mr Turran on the (names broken: [...]-gi and [...]-gir; the middle of Diyala) contain a form (surely a theophorous elethe seventh century B.C.). E-lam-maki(or-ki?) (J.J. ment!) related to Kass. Kadagman?The form Kadal, Durand, Mil. Steve,p. 124 f. (var. Ku-uk-ku-)who also lists another two Elamites from OB Mari, viz. Ku-ia(a-)ia/Ka-ia-a-iaand In-ne-ri/In-ne-[ri]."Simti-hallurig" is also Elamite (D. Charpin, Mil. Steve, p. 136; cf. NABU 1990/39). Charpin (ibid.) lists a fifth Elamite Cf. from OB Mari, viz. Ku-(un-)na-am/Ku-un-na-ma-an. ARMT 23, p. 311 f. ad 407, 4. Note Ha-ag-tu-tu (ARMT 26, 397, 7). A group of Elamite messengers is perhaps mentioned in ARMT 26, 368.

48

JOURNAL

OF PERSIAN

if it is related at all to Kadasman, may favour a segmentation Kadas + man (cf. Pinches,JRAS 1917, p. 108; Kass. St., p. 156 f., but Kadasman is a predicative element). Are Ba-gi-ni, whose sons belonged to the garrison of Duir-AmmiSaduqa (AbB 7, 114, 12), and his namesake Ba-gi-nu (AbB 11, 103, 7) related to Kass. pak- (Kass. St., pp. 76.171 f. 229: 92; 238, n. 42)? Is Su-mu-un- (of - -a-bi-ia, Kh.A. al-A'dami, Sumer 23, 1967, p. 153 and pl. 1 on p. 165, 9, from Tall ad-Der) a Kassite theophorous element? Lu-tuuk-in-dafrom OB Alalah (Wiseman, 412, 7) was a Kassite (see E. GaMil,Ac. Ant. Acad. Sc. Hung. 30, 1982-84, p. 22).

STUDIES

Na-zi-i was borne by several individuals in Mesopotamia during the first millennium (e.g., B. Parker, Iran 25, 1963, p. 91: 106, 6; from Balawat; 784 B.C.). He is mentioned together with Ka-gu-di (ibid., 8), but the latter, as well as its namesake KasARU su-du (ARU 177, 20; 117a, 17, var. Kag-Su-ud/t'i, 117, 9.10; after 648 B.C.; found at Nineveh; owned land), may be Assyrian (Kasgudu,to kaliadum).Ma-leza-za (ARU 257, 9; 673/2 B.C.) is possibly the same person as Ma-e-e-za-zi(ARU 619, 11; date lost; both found at Nineveh) in which case a Kassite derivation is excluded. Ku-ri-gal-zu(f. of Tabniya, CTN 3, 62, 15, after 648 B.C., from Kar-As'surin NE Babylonia; found at Calah) is a traditional royal name without any ethnolinguistic value. It is very doubtful whether Ha-am-ba-nu(KAV 121, 11), Ha-an-ba-a-ni (CTN 3, 58 r. 2') and Ha-an-bu-nu(CTN 2, 91, 12) are Kassite names (cp. Bit-Hamban,AOAT 6, p. 147?). A'di-a-s'i(CTN 3, 145, iv, 18; in a broken context; GN? 784 B.C.; found at Calah) apparently ends in -ya'. It

2.2. The Middle-BabylonianPeriod Unpublished early MB documents from the Diyala region contain"16several hitherto unknown Kassite anthroponyms (references are to Kass. St.), viz A-ak-ti(-e, cf. -ak-ti, 179), Ak-ri-ia-a (hitherto a not identical-with uU'Al-ditoponym and a horse name, 90.124), Al-ma-ia-a-as' is homonymous-if (with yal, 155 f.), B/Pur-na-ha-ia(with burna, 48 f.), (ia-)a', a stronghold of Mannai (AOAT 6, p. 41), but Bu-ur-ra-za-ah(with sah, 114), Ka-si-ia-a`(with ya , cf. the initial component is not recorded in Kassite (with -Sugab, 115), Nun-na- onomastics and there is no evidence for Kassites in above), Ne-me-en-su-ga-ab 75: nun + a + k Mannai. The same document (CTN 3, 145) has (iv. Nu-na-ak-te/ti, ak-kiNum-ma-ak-ki (cp. + te/i. Ia-e-ia-ag(with -ya'?), Su-ri-i-i (< Suriya'?) and 20) A-de-e-li(?)-~ii.A N/LB name worthy of mention Bir-ri-in-da-ak-ki/Bur-ri-in-da-ak-ku (to burra, 175 f.?) here is Lu-pu-sa-ah (poss. f. of Nabii-zfra-lifir; G. are possibly Kassite. Well-known Kassite names Barton, AJSL 16, 1899/1900, p. 80: 30, 7; sometime which recur in these texts are, e.g., B/Pu-ur-na-za-ah, between 485 and 482/1 B.C.) which on the face of it B/Pu-ur-ni,Bu-ri-ia and Bu-ri-ia-'u.The element burni ends in -sah. is apparently contained in Bur-ni-u-i-gu (if the reading is correct). The anthroponym B/Pur-na-ak-ki 3. OTHER PRE-IRANIAN PEOPLES FROM THE which occurs in the same group of documentation IRANIAN PLATEAU is identical with the second component of the -Pu-na-ak-ku hybrid toponym Bit-B/Pu-ur-nak-ka, from the 1st millennium B.C. (cf. Zadok, WO 16, Lugal-baira (Sumerian name) from Marhashi is 1985, p. 46, bottom). It can now be analysed as recorded in early Sargonic KA-sa-bum (the receipt CST 8 concerning copper probably originates from Kass. b/purna + ka. Documents from Sippar, which may be dated in Umma). Dragomans were attached to Marhashites the same period, mention a certain A-gum-ma-ri(cf. in the Ur III period (see Gelb, Glossa 2/1, p. 96 f.). Kass. St., p. 45 f.). Were Ar-_i-du-i and Hu-ru-PI-mas' Tuga-as (SS 3; 2035 B.C.; SET 228, 3) may be the Kassites as well? (see R. A. A. al-Qaradagi, Unpub- same person as the homonymous Marhashite lished Textsfrom Sippar, M.A. Thesis, University of (MVN 13, 695, 21). A Marhashite is mentioned in the undated document AOS 32, C1 (r. ii, 20). His Baghdad 1970 [Arab.], pp. 168 f. 172). For Kassites at Nuzi (c. 1600-1450 B.C.) see K. name, dSarai-kam, is typical of Umma. Cf. P. Deller and G. Dosch, Fs. Lacheman,pp. 91 ff.-For Steinkeller, RIA 7, p. 381 f. The musical instrument MB Du-un-nad[za]-ah (fem.) see van Dijk, OrNS 55 parahs~tumat OB Mari originated from Marhashi (1986), p. 161 ad VS 24, 91, 30 (cf. Kass. St., pp. 52. (see P. Villard, ARMT 23, p. 546 ad 579, 16). A Lulubian functionary (ma.kim) with a 84 f. 114). Sumerian name is recorded at late-Sargonic Lagash (MVN 6, 86, 1). For Lulubians at Girsu during the 2.3. The First Millennium B.C. time of the second Lagash dynasty see Steinkeller, The documentation is partly irrelevant as it JCS 40 (1988), p. 53. In A. Falkenstein's opinion contains onomastic "fossils": The gentilic Kal-ga(?)- (AnOr 30/1, p. 34; cf. NSGU 1, p. 101 ad 162, 17) Luaa appears in a document which was found at lu-bu(na)kiwas a Lulubian colony in the city-state of Calah (CTN 3, 68, 13; 749/8 B.C.). Bur-naza/sas/zahis Lagash (Ur III; cf. RGTC 2, p. 112, but note the mentioned at Nineveh sometime between 648 and reservations of Edzard in S. Parpola, A. Parpola 612 B.C. (ADD 1165-StPohl. Ser. Maior 14, 231, 41). and R. H. Brunswig, Jr. JESHO 20, 1977, p. 148).

ELAMITES

AND

OTHER

PEOPLES

FROM

IRAN

AND

THE PERSIAN

GULF

REGION

49

Lu-lu-bi-tum acted as a messenger according to indeed indigenous to Kakmi (ensis' names were not ITT2 828, 6 from Lagash (AS 9). Lulubian(s) (Lu- necessarily indigenous to their regions!) it may be lu-bi-[.. .]) are recorded in S 44-46 (AUCT 1, 378, 15; surmised that the homonymous commoners occurplace unknown). See H. Klengel RIA 7, p. ring in other Puzrish-Dagan documents (mostly from AS's reign) originated from Kakmi and/or 164 f. A-3a Sa-ra-hu-um(-ma), i.e. "field of Sarahum" (near adjacent regions (e.g. FAOS 16, 959 r. viii, 35; ITT2 Umma; AS 4-SS 6) has nothing to do with a Trans- 922, 34; 1007, 27; TCL 2, 5504 r. ii; 6401, i, 8; TRU Tigridian

region (Sarithum/Suruthum). Cf. UNL 2, p.

185: 751; Owen, JCS 33, p. 262. An anonymous Gutian dragoman is mentioned in the Sargonic period (see Gelb, Glossa 2/1, p. 94 f.). Anonymous Gutians are reported then in Pinches, Amherst 9, 2. Such females are mentioned at Sargonic Adab (OIP 14, 141, 12; note the strange name E-ki-ku-ki in the same document). Cf. Steinkeller, RA 74 (1980), p. 7 f.; Englund, Fischerei, p. 54 f. Gutians are recorded at Sargonic Umma (see Foster, USP, p. 113). Note that the Gutian royal name Sii-uum (for variants and homonyms see Hallo, RIA 3, p. 712b, inter alia the pre-Sargonic female weaver Si-um-me from Lagash) is apparently homonymous with a region of Greater Elam (cf. RGTC 1, p. 65 f. and 1.0). Was this potentate identical with the master of I-lu-lu (A. Moortgat, VAR 186)? Cf. Hallo, RIA 3, pp. 709 ff. A Gutian named GIR-nun-ne'(son of Lugal-an-natum) is recorded either in the Late-Sargonic or in the Gutian period probably near Lagash (see B. R. Foster, in H. Behrens, D. Loding and M. T. Roth

[eds.], Dumu-e2-dub-ba-a, Studies in Honor of Ake W. Sj6berg, Philadelphia 1989, p. 161: 3). Du6-Gu5-ti-um(-

ma), -Ku-ti(-ma)was situated not far from Lagash. It might have originally been a Gutian colony according to Falkenstein (AnOr 30/1, p. 24; cf. Waetzoldt, ZA 65, p. 273; Ur III). A certain Gu-ti-ais mentioned there in the same period (ITT2 950, i, 8). For a Gutian (named A-qur-[qa?]-ba) at OB Alalah

133, 2; UDT 91, 149; Watson, Cat. 84, 3). All the occurrences may, however, refer to the abovementioned ruler (without his title), as the chronological distribution suggests. The wife of Du-uk-ra

(e.g. CST 480; TrD 19, 2) was possibly that of the ruler. Lu-dnanna, son of Du-uk-ra is recorded at Puzrish-Dagan in SS 5(2033 B.C.; Nakahara, Sum. Tabl. 36, 7 f.). Note Du-uk-ra of Nu-ga-ark' (AS 5 =

2042 B.C.; MVN 11, 140). Du-uk-raresembles Tu-ukservant of Ammiri, patronym of Ibni-Marduk VAR OB 495; seal; found at saduqa (Moortgat,

Babylon). It is hardly Semitic (cf. Gelb, MAD 3, p. 296, s.v. "TKR?"). The Turukian I-ni-ig-ki-ba-al (ARMT 26, 404, 91)

might have had a Hurrian name like Iniz-ulme

(ARMT 25, 104, tr. 2; cf. NPN, p. 271b); for kipal cf. NPN, p. 228a. Ar-ba/ma-da-al from Ashjali (UCP 10/1, 27, 1; OB,

und.; poss. a worker) probably ends in Hurr. -atal (cf. E. I. Gordon,JCS 21, 1967, p. 84; S. Greengus, BiMes 19, p. 112 ad loc.). For the rest of the name cf. perhaps A-ru-ma-a-tal and arm- (NPN, p. 205a) on the

one hand and Ur III Ar-ba-tal(AUCT 3, 294, 9.11) =

A-ar-ba-tal (MVN 15, 216, 16) = A-ri-ba-a-tal (MVN

13, 710, 2) from Shimanum on the other. An anonymous Subarian slave is mentioned in an undated OB letter, perhaps from the Ur region (greeting by Samas', Sin and Ningal, R. D. Freedman, The Cuneiform Tablets in St. Louis, Dissertation,

Columbia Univ. 1975, 194, 13). An-di-a-a (ARU 292, 9; 682/1 B.C.; found at (Wiseman, 54, 23) see E. Gaail, Ac. Ant. Acad. Sc. Hung. 30 (1982-84), p. 16. For MA/SB quti2 "Gutian Nineveh) originated from Andiya. Ni-nu-a-ri (665/4 B.C., found at Calah; Wiseman, Iraq 15, 1954, (garment)" see Deller, OrNS 58 (1989), p. 266. A certain Du-uk-rawas active at Puzrish-Dagan in p. 140: ND 3424, 37) reminds one of Zamuan/LulluAS 4 = 2043 B.C. (BIN 3, 101, 5). His namesake bean names ending in -r, but this is not a criterion acted as an ugula according to a document from Ur and the relevant documentation is very scanty.17 Several names of well-established Babylonian (AS 9; UET 3, 1499, obv. i) in AS 9 = 2038 B.C. Duuk-ra, who bore the same title, is recorded at families in the first millennium B.C. cannot be Puzrish-Dagan (SS 2-IS 2 = 2036-2027 B.C.; PDT associated with any known dialect, e.g., Sda-DIS-luh 114, 4; 346 r. 14). They were homonymous with Du- who was the ancestor of seven individuals (includFAOS 16, 807, ii, 6; TAD 67, uk-ra ing three scribes) from Borsippa (541-486 B.C.; lih-kak-mik' (e.g., 7). Kakmi (of Ur III and later sources) was situated B6hl, MLVS 3, p. 49:863; Joannes, Archives de Bornear-if not inside-Gutium. Du-uk-ra (also Tu-uk- sippa, Geneva-Paris 1989, p. 419a with refs.; TBER ra) acted near Cutha in AS 2 = 2045 B.C. (see pl. 75-TEBR 75; VS 3, 162; 4, 139; 6, 160), two scribes from Babylon (164 B.C.; SBH 1 r. = AOAT 2, Goetze,JCS 17, p. 16 with n. 54, 55). A connection with Tukrig does not seem likely if the latter is an 141) and another scribe from an unknown place Indo-European -i stem. Du-uk-rais once described presumably in Babylonia (date lost; ibid. 476; as li-Ni-lum-[maki] (MVN 13, 529, 20 f.; AS 1 = Hunger reads the last sign always as -i). The 2046 B.C.). The location of Nilum (an ensidom, cf. reading ld Sukkal Bdr-sipa (suggested by von Soden, Owen, JCS 33, p. 259) is not known (or is it ZA 73, 1983, p. 294-410:75, 19) is unlikely in view of AOAT 2, 141.476. simply a mistake for Kakmeki?). If Du-uk-ra were

50

JOURNAL

OF PERSIAN

STUDIES

Nippur, Walker, AfO 24, 1973, p. 123 and pl. 19, 3); perhaps not later than Shulgi's reign (-gan, concernat is recorded early Sargonic ing sheep from the field of Da-se-hu-um, TS 20, 10); Sargonic. Lzi-ma-gan Lagash (ITT 3106). A courier from Makan received undated (-gan,SA 157, ii, 6). It is not certain whether beer according to a late-Sargonic document prob- the anthroponym Ma-ga-nu-um(also -nuim as a field ably from Umma (T. Donald, MCS 9, 1964, 245; cf. name, UNL 2, p. 77:560) has anything to do with Gregoire, AAS, p. 18, n. 4). The provenance of Lui- Makan (cf. Gelb, MAD 3, p. 170). Puzrish-Dagan. S 44 (gan, concerning cattle; TS [ma?]gan (MVN 3, 28, viii, 9; an account of equids and barley) is not known. 80, 5); S 48 (FAOS 16, 1243 r. iv, 8, cf. 1230, 1, und.). Ur III (for the chronology of the regnal years see Nippur. IS 1 (-ganki, Fish, Iraq 5, 1938, p. 169: 1 r. 1.3). Lu•-mdganis not a rare anthroponym in this 4'). Ur. IS 2 (ganki, concerning jars; UET 3, 883, 15). period, but is totally absent as such in later periods The A concrete W. RIA 7, p. 199). (see village of the Makanites included a field, part of physical Heimpel, connection between most-if not all-the individ- which was royal property (see Pettinato, UNL 1, p. uals bearing this name and Makan cannot be demon- 173:253 ad UET 3, 1364, 4; perhaps time of Ibi-Sin). Is Ur III Su-ir-hu-umki (A UCT 3, 484, 11) identical strated (see Edzard in Parpola et al.,JESHO 20, 1977, with Sar. Se/irihum (possibly in the Persian Gulf Makanite of 166 the occurrence f.). However, p. cr. RGTC 1, p. 151)? acc. rations to in IS 1 at (received region, groups Lagash DAS 270.271) makes such a connection likely, the more so since most of the individuals in question 5. MELAHHITES received rations. A noteworthy exception is Lu-mSaThe evidence for Melahhites in early Sumer was ganki who had the Sumerian patronym Lil-dingir-ra and acted as a scribe at Girsu as early as S 47 (MVN thoroughly discussed by Parpola et al., JESHO 20, 12, 241:HS 5, ST 5). He is not a unique case of a 1977, pp. 129 ff.; cf. MVN6, 88, 5; Yang, Adab, A 712, scribe of foreign extraction in Sumer (cf. Limet, 10). They argue that at least in the case of the village 4. MAKANITES

CRRAI 18, p. 134 ad Unapsen and Nanna-ma(n)ba).

The following homonymous individuals (only variants are registered) lived also in Lagash (presumably at Girsu) and received rations (unless otherwise stated): S 25 (-ganki,TS 82 r. 4; Gomi, Bull. Anc. Or. Museum 2, 1980, pp. 35 ff.: 88.89.92-102: SS 5-Is 2 [concerning sheep]); S 25/IS 3 (ganki, MVN 7, 97, v, 3); early reign of Shulgi (?-gan,M. van de Mieroop, RA 79, 1985, p. 32 ad 26, 6); AS 1 (gan, MVN 12, 337, 2; ganki, Pinches, Berens 30, 5.7); time of SS (-gan, CT

3, P1. 13:13888, 2); possibly same period and SS 2 (ganki, MVN 7, 431 and 225 respectively);

SS 5 (ganki,

CBT 1, 13824.13847); SS 6 (-gan-na, CT 7, pl. 23: 13944, 9); SS 8 (-gan,A UCT 2, 45, 3); IS 1 (-(gan) DAS 269, 2; 270, 3; 271, 2; ganki, CT 3, pl. 20:16367, 2); IS 2 (-ganki,MVN 7, 400, 1; -gan; TS 83, 3); undated (-ganki, TS 81, 9; TuM NF 1/2, 119, 4). Lui-mdga-na

acted as a li-0g tukul gu-la in the Messenger Document HLC 3, pl. 130:310 r. 3 (year unknown). Umma. S 25/IS 3 (-ganki, prob. Umma; sent to

Is OB Te-te(?)-ep-ma-da (V. Scheil, RA 24 [1927], p. 41:5) a related name? 2 Andreas, Verhandlungendes 13. internationalenOrientalistenkongresses,Hamburg,Sept. 1902 (Leiden, 1904), p. 93 f.; W. Eilers, BNF 15 (1964), pp. 180 ff. (esp. 194 with n. 47); idem, IF 79 (1974), pp. 53 ff. Cf. R. Zadok, Iran XIV (1976), p. 63 with n. 14 where more literature is listed. 3 It is very likely that many documents quoted below refer to homonymous individuals living in Greater Elam, such as the ensi of Adamdun or the lakkanakku of Sium. The former, spelled Ux(URU)-ba-a,is mentioned in PDT 27 after his colleague Za-ri-iq of Susa (both without title).

of Melahha (E-duru5sme-luh(-haki,2063-2046

B.C.; in

the Lagash region), whose store(house) is recorded (RGTC 2, p. 42 with refs.; see Pettinato, UNL 2, pp. 246:18; 249 f.: 53), it was originally inhabited by people from Melahha who might have cultivated the orchard

(kiri6) of Me-luh-ha (STA 19, ii, 14). See,

however, the reservations of Edzard in Parpola et al., pp. 148 f. 166 f. It is surprising that the relatively rich documentation from Ur concerning Melahha (e.g., items of furniture, RGTC 2, p. 132 f.) contains no information on Melahhites. The anthroponym Dar-me-luh-ha(-tur, ibid.with refs.) is not directly connected with Melahhites as it is possibly originally an appellative "Indian poultry" (made of ivory, see Waetzoldt, ZA 65, 1975, p. 275). Ur-DAR.A-meluh-haki is followed by Marhashi (UET 8, 37, 9 f.; cf. CT 44, 1 linking Melahha with Anshan).1s Makan and Melahha are recorded in the early OB period (BIN 10, 14, 143, 2 and 114, 3 respectively). AUCT 2, 164, 17; NATN 111, 5. seal; 344, 5; 893, 7; 912, 2; 955, ii, 9; TuM 3, 13, 10; 40; 43a, 7 = 43b, 6; 55, 4.9 are from Nippur; NATN 213, 3 (provenience: Umma, but has the Nippur calendar and was found at Nippur). TuM 3, 179, 22 is perhaps from Nippur. A UCT 1, 703, 2; 2, 54, 7; 201, 4 and perhaps TuM 3, 283, 4.5 are from Puzrish-Dagan (near Nippur). A UCT 3, 421, 6 is from Umma. 5 Limet, Anthroponymie,pp. 398-401, lists the names with Sulgi as a theophorous element. Sul("DUN")-gd-(a-)at(RGTC 2, pp. 147.244) is perhaps Akkadian according to Gelb (MAD 3, p. 314). If it is a Sulgi-name one may assume that it is contracted from * Sulgi-a-at. The only Akkadian form which would suit

4

ELAMITES

AND OTHER PEOPLES

FROM IRAN AND THE PERSIAN

the spelling of the final component is the hypocoristic suffix -yat-um(MVAeG44, p. 113), but such an interpretation is difficult due to the lack of mimation. Are we to deal here with a non-Mesopotamian predicative element attached to Sulgi, like -atal (cf. Limet, loc. cit.) and lipani (Owen,JCS 33, 1981, p. 258)? 6 Cf. Iran XXV, p. 26, n. 54 in fine. 7 Sapum is localised by B. R. Foster east of Der, but P. Steinkeller (WZKM 77 [1987], pp. 185 ff.) demonstrated that Foster's localisation is not sufficiently based on textual evidence. Cf. H. Neumann, OLZ 84 (1969), col. 523 f. with nn. 34-37. F. Vallat (Paliorient 11 [1985], p. 53) localises Sapum in the region of AghaJari on the coast of the Persian Gulf south of Huhnur, southwest of Shiraz and northwest of Liyan. The tribute (GUN) of Sapum is recorded in S 48-AS 1 (AUCT 1, 743, 5; 2, 179, 11). Cf. P. Michalowski, ZA 68 (1978), p. 48 f.; HE, p. 18. 9 See AOS 32, p. 12 f. and MessT, pp. 64 ff. resp.; cf. D. O. Edzard (following a suggestion of Gelb), AfO 19 (1959-60), pp. 17.21 with n. 54 and in S. Parpola et al.,JESHO 20 (1977), A. Goetze (JNES 12 [1955], p. p. 166 f. ("Berufsbezeichnung"). 116) considered them as mercenaries. He is followed by Sauren (WMAH, p. 236). Cf. Falkenstein, JAOS 72 (1952), p. 43a; Jones and Snyder, SET, pp. 299 ff.; Gelb, Glossa 2/1 (1968), p. 96, nn. 7.8; Gregoire, AAS, pp. xiv.178, n. 128; Sollberger, TCS 1, p. 114 f.; Lambert, TE, p. 181; Limet, CRRAI 18 (1972), p. 132 ("soldiers and workers"); Carter and Stolper, HE, p. 18 with n. 110. 10 For Elamite deities in later Sumer and Akkad, cf., e.g., dNinguginak (Sulgi 6, 5-7, 12, 1 ff.); dNab (Er?emma10, 6).

11

GULF REGION

51

dgulgida(an)gada (DAS 192, 10); dsulgiidanga-da (DAS 153, 9; 155, 8; RTC 392 r.); dSulgidana"ga)-da (DAS 156, 8; 159, 14; 167, 17; 177, 12). 12 D/Tan-d-pi (DAS 7, seal; 11, seal; Delaporte, CCL 1, T 220 +; ITT 6032, 7); D/Tan-zi-pi(AAS 167, 6; 176, 2; DAS 84, 5; 85, 5; 86, 3: [-pi]; 153, 5; 155, 5; 157, 5; 159, 5; 166, 5; 167, 5; 171, 20; 177, 5; 183, 5; 186, 5; 193, 5; 195, 5; 204, 5; 242, 31; 243, 23; ITT2 1021, vi', 4; Jean, RA 19, 1922, p. 33:47 r. 5; 40:25 r. 5; MTBM 49, 1; TE 48, 5; WMAH 235, 6); D/Tan--pi4 (WMAH 233, 16); [D/Tan]Ji-e(ITT6843, 23). 13 On the relations between Sumer and Iran as reflected in Sumerian myths, see recently S. N. Kramer, Bulletin of the Asia Institute 1 (1987), pp. 9 ff. 14 I prefer not to insert English renderings of the various titles since in most cases it is difficult to distinguish between their functions (e.g., nu-btnda "supervisor" and ugula "foreman"). '5 Note "Elamite meat" in AUCT 1, 337 (AS 1). 16 I. J. M. al-cUbayd, UnpublishedCuneiformTexts of the OB Period from the Diyala Region (M.A. Thesis, Baghdad 1983 [Arabic]; from Tall Muhammad within Greater Baghdad). These texts contain also Akkadian and Hurrian names. To the latter group may belong, e.g. (pp. refer to NPN), Ha-at-ra-ak-ki(cf. Ha-at-rage, 215a), Ha-az-zagi-ia-Iu(226b, cf. Gi-ia-a-Ium?),Ha-ziip-dzi-la-kum(278a), Hu-bi-di-ia(218a), Ki-ir-ri(228a if not atypical), and Si-migi (257). Gu-ub-buand perhaps Mu-lagi(ki?)-di may be Amorite. Note the patronym Gu-ti-i ("Gutian"). 17 E.g., Ki-ir-te/ti-a-ra,Me-eg-di-a-ra(cf. I. M. D'yakonov, Istoriya Midii, Moscow 1956, pp. 102, n. 4; 156 f.). Is For Melahha in the great poem "Enki and the World Order", see S. N. Kramer and I. Bernhardt, WZJ9/1-2 (1959/60), pp. 231 ff.

THE INTERNATIONAL MERV PROJECT PRELIMINARYREPORT ON THE SECOND SEASON (1993) By Georgina Herrmann, K. Kurbansakhatov et al. LondonandAshgabat

Our expedition was sponsored by the National Geographic Society, the British Academy, the British Museum, the British Institute of Persian Studies, the Stein Arnold Fund, the Society of Antiquaries of London, the Royal Society, the University of London (Central Research Fund), and University College London, and we are extremely grateful for their support. Our plans for mapping Erk and Gyaur Kala were made feasible by three magnificent loans: the University of East London lent us their superb Leica 200 Global Positioning System, and its skilled operator Ian Peet, for ten days, while Leica International continued their invaluable support by lending us a "Total Station" and their latest Penmap logging system. Once again, "Optimal Solutions" provided us with a "Husky" field computer. We thank them all for their generosity. We would also like to thank the Ordnance Survey and the Institute of Archaeology for the loan of additional survey and photographic equipment, and ECHO (Supply of Equipment to Charity Hospitals Overseas) for cost-price medical equipment. We were glad to be visited by Professor V. M. Masson, Director of the Institute for the History of Material Culture, St. Petersburg, and Scientific Coordinator of the Project. We are always grateful for his advice and support.

The International Merv Project is a three year Collaboration (1992-4) between University College London, YuTAKE (the South Turkmenistan MultiTurkmen Expedition), disciplinary Archaeological Academy of Sciences, and the Institute for the History of Material Culture, St. Petersburg, and operates under Permits granted to Dr. K. Kurbansakhatov and Dr. S. D. Loginov of YuTAKE. The second season took place from 25 August to 16 October. The team consisted of up to thirty personnel,1 three from YuTAKE, twenty from London and seven from St. Petersburg and Moscow. However, not all personnel were present at any one time: eight members of the British team came only for the first month, others came for the second half. Some of our Russian members equally were not at Merv for the full season. There was a pleasing spirit of friendship and cooperation between the three teams, Turkmen, Russian and British, and the results achieved are a collaborative effort.

ACKNOWLEDGMENTS Work is only possible at Merv because of the constant support of the Hakim of the Vilayat of Mary, Kurban Muradovich Orazov, and of his Deputy, Mohammad Chari. Not only do they give us permission to work in the Archaeological Park "Ancient Merv" but they accommodate us in their fine expedition house. They also give regular assistance with the inevitable problems of supply in keeping such a large expedition in the field at this difficult time. As if that is not enough, this year the Hakim lent us his personal helicopter both to fly round the ancient cities of Merv to take photographs and also to fly down the River Murghab to the Afghan border, which gave us an unparalleled opportunity to view the sites within their landscape. We are deeply grateful to the Hakim, his Deputy, and his Director and Deputy Park "Ancient Directors of the Archaeological Azim Akhmedov and Nursakhad Mamedov, Merv", Annaev is Annaev. Akmohammad Akmohammad always quietly ready to solve every problem with a smile, whether it is building a shower or a flotation tank, negotiating for bulldozers, or informing us about a variety of topics, including the local flora and fauna, building materials and traditions.

AIMS AND RESEARCH PROGRAMME The historic urban centre of the Merv oasis was selected for study by the International Merv Project not only because of its intrinsic importance as a capital city and major trading centre but also because of the unparalleled opportunities the site offers as a model for testing investigative archaeological techniques. Ancient Merv consists of a series of adjacent walled cities, lacking occupational overburden, unoccupied today and relatively free of vegetation. Our decision to work at Merv came at an opportune time when major advances were being made in the techniques of ground survey and in the power and affordability of Personal Computers. Two factors underlay our research strategy,2 one the need for an up-to-date plan of the site, together with a reassessment of earlier work, the other that large scale excavation is currently impossible, both

53

54

JOURNAL

OF PERSIAN

because of rising standards and costs and because of the absence of local workmen. From the outset, therefore, a mapping facility powerful enough to allow the rapid recording and storage of a large quantity of survey information was built into the project and was designed so that all past, current and future surface and sub-surface archaeological fieldflesh out the bare work can be plotted. Equally-to

STUDIES

tinued to flourish through the following Parthian and Sasanian periods, before moving west to the new site of Sultan Kala, which itself was assumed to have been abandoned after the devastation wrought by the Mongols. In what way has our work either agreed with or amended the above picture? First, our Turkmen colleagues, Terkesh Khodjaniasov, Akmohammad Annaev and Alaguli Berdiev,

TABLE

1

MAPPING Site Work

Records I

Old maps Archive reports Museum study

I

Aerial Photography

I

I

Geoarchaeology Borehole Magnetometer

Geophyisical Survey

I

Topo. survey Total Station G.P.S.

I

Surface Artefact Survey

EXCAVATION I

Stratigraphy Architecture Soil Samples

I

Artefact Assemblages Coins and Inscriptions Ceramics, Glass, etc.

wide range of bones of the mapping programme-a non-destructive, investigative techniques, see Table 1 above, have been undertaken, together with an The research limited excavation. essential, although section lockas a each is whole, programme planned ing into the overall design. While the focus of the first two seasons in 1992 and 1993 has been the acquisition of survey data, the third season, scheduled for 1994, will be more archaeological in emphasis. Not only will work on the Erk Kala trench continue, but it will be reinforced by a new programme of area excavation on a long low mound in the north east quadrant of Gyaur Kala (P1. IVa). This mound was selected in 1993 for an intensive survey programme, see reports below, because of the architectural remains visible on air photographs. Our new excavation programme, tested in 1993, see below, will involve the surface scraping of up to 8,000 sq. m. and is designed, firstly, to verify the nondestructive techniques employed, secondly, to provide a larger architectural plan than heretofore and, the stratified chronological finally, to extend sequence of material culture and environmental remains through the Sasanian period. PRELIMINARY RESULTS Archaeological orthodoxy suggested that Erk Kala was the Achaemenid foundation, while Antiochus I (281-261 B.C.) was credited with building the city of Antiochia Margiana, modern Gyaur Kala, and the outer wall, Gilyakub Chil'Burdezh. The city con-

I

I

Environmental Seeds and Charcoal Bones, Shell

I

Technological Ceramic, Glass Metal, Stone

have clearly demonstrated that occupation at Sultan Kala did not cease as a result of the Mongol destruction in 1221. There is considerable depth of postMongol and Timurid occupation. The team has also succeeded in collecting material for a gazetteer of the standing monuments in the Archaeological Park. These were photographed in 1992 and were drawn in 1993 by the talented Russian architect, Alexei Konenenko, under the guidance of Professor J. M. Rogers. We should like to record our thanks here to Akmohammad Annaev who made available plans of monuments prior to restoration, which had been prepared for the archive of the Archaeological Park. The team's principal effort has been the study of Erk and Gyaur Kala, and we have already established a considerably greater Islamic use of much of the city than previously thought. This is demonstrated in the maps illustrated as Fig. 1 A-D, which record the preliminary findings of David Tucker's Surface Artefact survey. They illustrate the extent of occupation of three distinct phases, Late Sasanian, and Earlier and Later Islamic. It is also made clear by the Erk Kala excavations of an extensive mudbrick house, phases of which are dated by associated coins, ceramics and glassware to the sixth-seventh centuries A.D., that is to the transition from the Late Sasanian to the postSasanian periods. Despite a reduction in house size in the final phase, there is no indication of any break in sequence. This evidence of continuity between Sasanian and post-Sasanian levels adds to the gradually emerging picture, established elsewhere on historical (Morony 1984), architectural and art historical grounds (Thompson 1976 and Kroger 1982).

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

The final phase of our Erk Kala excavations is postSasanian (Arab-Sasanian) in date and may be constructure which temporary with the monumental once occupied the citadel mound within Erk Kala. this structure was bulldozed away Unfortunately to Nevertheless, this prior proper documentation. with extensive occupation major building, coupled both within Erk and Gyaur Kala, demonstrate the continued importance of the city in the early Islamic period. Small, low-value copper coins are common at Merv, see below. They are found lying on the surface in large numbers and are also recovered from the Erk Kala excavation, both within deposits and accidentally incorporated into mudbrick construction. Most of them were minted at Merv, and their study makes a major contribution to our understanding of the history of the site. It is, for instance, well known that Ardashir I (A.D. 224-241) claimed sovereignty of the city. However, his earliest coins from Merv belong to one of his final issues (P1. VI: 5), corresponding to G6bl Type III.2 (Gobl 1971), by which time his crown had developed into the familiar cap surmounted by a korymbos. As Nikitin and Loginov state below, this demonstrates that Merv only became a Sasanian province in the later years of Ardashir's reign. The work of our first two seasons has already challenged some of the orthodox views held about the cities. While we hope that a number of problems will at least have tentative solutions at the end of the 1994 field season, many key questions will remain, some of which will undoubtedly be too heavily buried to answer at Merv itself.

THE MAPPING PROGRAMME The Topographic Survey, by Glynn Barratt The 1993 topographical survey programme followed on from the promising start made in the mapping of Erk and Gyaur Kala in 1992 (Herrmann et al. 1993: 44-8). The contemporary survey techniques employed in 1992, using a Total Station with a dedicated logging facility to collect map information in digital format were continued, while a further advanced survey technique, a Global Positioning System (GPS), was also employed for a limited period. The aims of the 1993 mapping programme were to move out from the early city of Erk Kala, investigated in 1992, into the more extensive conjoined city of Gyaur Kala. The investigative methodology between the remained one of close cooperation topographic survey team, surface artefact survey and the geophysical and geoarchaeological investigation

REPORT

ON THE

SECOND

SEASON

(1993)

55

teams. Due to logistic considerations, the survey season was divided in two. The initial two weeks concentrated exclusively on the use of the GPS equipment, the remainder of the season worked with the Total Station. The GPS Survey: GPS, which uses orbiting satellites to obtain a fixed global position, was chosen for use at Merv for several reasons. 1. Merv is free of standing buildings and of high vegetation cover, allowing good satellite visibility. 2. GPS allows rapid fixing of control stations over a large area. 3. Recent developments have provided the facility to track surface topography in a continuous kinematic fixing mode. 4. Fixing a GPS grid to the mapped site area so as to interface data from Remote subsequently to the GPS Sensing platforms (SPOT/LANDSAT) coordinates. Alhough GPS equipment has been used for a considerable time as a navigation aid and as a survey control tool, it has only comparatively recently been adopted as a tool for accurate detailed survey. The equipment used (Leica System 200) was introduced only within the last year. Using phased signal collection and differential fixing, it is capable of giving a WGS 84 position to 4 mm. accuracy in a horizontal plane and to 5 cms. in height above ellipsoid. The equipment was used in several areas of the mapping programme, some of which are regarded, at least in part as experimental in terms of ground survey technique. A network of post '92 season air photography control points was fixed for the future photogrammetric plotting of Gyaur Kala as well as possibly for the extensive remains of Sultan Kala. visible by Similarly surface features potentially remote sensing from satellite were fixed to allow correlation between the land surface and sensed data. A further attribute of the equipment allows continuous tracking once satellite lock is initiated, so that in a kinematic mode, the satellite receiver can be moved across a surface, either by foot or in a vehicle, and will provide continuous mapping of the feature or ground surface. This potential, for real-time survey with a vehicle-mounted kinematic GPS, has been used for the first time on an archaeological site to provide height information. Two transects across Gyaur Kala north to south and east to west were tracked using the GPS mounted on a Gaz 66 truck, while a contour survey of a mound (MGK: 5), chosen for detailed investigation, was accomplished with the receiver mounted on a Lada Niva. In both cases the speed of data collection was impressive. If this experimental use of kinematic GPS proves to be successful once the field data is processed then rapid contour survey of complex sites such as Gyaur Kala will become increasingly possible

56

JOURNAL

OF PERSIAN

within the limited available time of a field survey season. TotalStationSurvey:The second and larger part of the season, following the departure of the GPS equipment, was concentrated on the collection of data using a Total Station Theodolite (Leica TC 500) and two complementary logging systems (Leica Penmap and Optimal Autograd). The middle portion of the season concentrated on the detailed recording of a portion of the lowlying area in the north-eastern corner of Gyaur Kala (P1. IVa). Such low areas are a feature of the corners of Gyaur Kala: those in the north-west and south-west are largely free of any mounded areas, whereas those in the north-east and south-east each have a varied topography. The mound, designated MGK: 5, was chosen for intensive investigation after examination of aerial photographs collected in 1992. Soil marks on these indicated that a range of substantial buildings was present on the mound. The area thus formed a focus for geomorphologic and geophysical investigations, and was accordingly surveyed in association with these teams. A detailed contour survey was carried out of the mound and its environs. All surface features were recorded, all blocks of vegetation were surveyed, all geophysical grids positioned and all boreholes located. Recording was in three dimensions to facilitate later modelling. The (Penmap) logging system used comprises a 386 pen-driven (Tandy Gridpad) computer running a programme designed to record survey data in a graphical form. The ability to see in map form the data recorded as it was collected greatly facilitated this intense and detailed recording, the ability to layer data and to view contour plots in the field proving particularly useful. There is, however, at present a limitation to this type of system. Battery life of the computer was limited to two hours, making access to a charging facility essential if this type of system is used. This logger was, accordingly, backed up by a less power demanding computer (Hewlett Packard HP 100LX). This palmtop computer, while not having the power of the Gridpad, runs on simple dry cell batteries which in this case lasted the full duration of the season without replacement. Using feature coding logging software (Optimal Autograd) the system was used extensively throughout the season as a general purpose logger both for survey data and for attribute information. With the completion of the micro-contouring of MGK: 5 area the final part of the 1993 season was taken up with the location and fixing of the positions of the former excavations within Gyaur Kala. This was accomplished with a highly mobile, two man team using the team vehicle, radio communications, the TST, HP 100 and Autograd logging system. Previous excavation sites and surface features, such as indus-

STUDIES

trial areas, were tied into the control framework established in 1992. The two seasons of survey at Merv have so far drawn on a wide and varied range of data collection techniques with a view to maximising the amount of information that can be collected within a limited period of time. It now remains to compile that information into a useful format. Post field processing may suggest the necessity for further ground work in selected areas. Without the help and cooperation of all the team members, both topographical survey and other investigative disciplines, it would not have been possible to coordinate the programme. Similarly without the generous loan of expensive equipment made by Leica U.K. and the University of East London, it would not have been possible to approach the quantity or quality of information which has been recorded so far. The Geoarchaeological Survey, by A.J. Barham and S. Mellalieu Three

key questions have led the objectives of research in the last two seasons of geoarchaeological work. These were: 1. What was the topography and nature of the landscape environment prior to and during occupation at Erk Kala and Gyaur Kala? 2. What processes cause erosion and modification of the mud brick structures when abandoned, and how far does the modern topography reflect the shape of underlying building architecture? 3. Can sediment studies be used to define former activity areas, e.g. smelting, kiln sites, or produce information on the type of environment that existed adjacent to the buildings (e.g. in low lying areas) in the period c. 500 B.C. to A.D. 1200? The questions have been addressed in three related ways. Firstly, by examining the type of sediments that exist in stratified contexts of the excavation areas. Second, by understanding the sediments and processes that move across the site and downslope into the low-lying areas, and third, by examining sediments derived from the mounds now redeposited in the seasonal swamp areas with salt pans. Stratigraphic investigations during the 1992 season were concentrated on the depression within Erk Kala. This located a buried land-surface pre-dating occupation of the site. The objectives of the 1993 season were to determine whether a similar sequence could be identified within a depression of Gyaur Kala. This information will help determine the mode of formation of the area, whether by excavation of material for mud brick production or unoccupied gardens. Investigations used the same methods and techniques employed in the 1992 season. Boreholes were

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

drilled (2-5 m. deep) using manual augering equipment along a transectacross the depression so that the sedimentarysequence could be described and sampled.Interpretationswere based upon characteristics of sediment particle size and sorting, colour, compaction, sharpness of contacts between layers, and inclusions, e.g. salt crystals, charcoal, bone, ceramics.All boreholes were mapped and levelled into the main project mapping grid using EDMand GPS techniques. Similar stratigraphies to those found in Erk Kalahave been identified in this area. The preliminaryfield results indicate gradual infilling of the low-lyingarea by up to 3 m. of sediments during the last 3,000 years. Sediments eroded from the archaeologicalhorizonson the adjacentmounds have been identified lying 1-2 m. beneath layersof saltsand anoxic black muds. The salt-richsediments lying nearer to the surface preserve evidence for recent salinisationand local changes to the vegetation, probablydue to changes in the level of groundwater produced by construction of the Kara Kum canal. Two narrowtest trenches dug by a back-hoewere opened in the lastfew daysof the team'spresence in Merv,one in the slopewashof the Trench 5 Mound and one in the adjacent salt pan. These were areas alreadyexamined by borehole and were designed to test the findings based on such augering. A wellstratifiedlevel with numerous molluscsmayindicate that the palaeo-landsurfacewas reached. Sediment samplesfrom bore holes and trencheswere collected and taken to London and St. Petersburg.Field data will be supplemented by detailed laboratoryanalysis of sediment samplesat the Museumof Soil Science, St. Petersburg, and the Institute of Archaeology, London. Surfacesedimentshavealso been analysedusing a portable magnetic susceptibilitymeter (Bartington InstrumentsMS2).This method allowsdepositsto be characterisedby their magnetic mineral content, a propertywhich varieswith sediment source, potential firing history, e.g. hearths/kilns, ceramic and organic content. Field analysishas concentrated on mounds and saltpansin GyaurKalaand also on the excavationwithin Erk Kala,where the method has been used to characterise and identify contexts. However, the usefulness of this technique at Merv requiresfurther calibrationusing chemical analysis in the laboratory. The preliminary results of the magnetic susceptibility field surveys indicated that it is possible to discriminate between the main surface sediment types in the area. In particular, salinised sediments marginal to, and within, the low-lying depressions yield much lower values (2-4 x 10-8 SI) than the leached and windblown surficial sediments of the mounds (typically 25-35 x 10-8SI). Additionally, surface con-

REPORT

ON THE

SECOND

SEASON

(1993)

57

centrations of artefacts (particularly pot and brick fragments) enhanced the magnetic signal (to 35-60 x 10-8 SI)-so that along transects where artefact denmeasured as frequency sity was independently counts/m.2, the susceptibility measure shows a direct correlation with artefact concentration. Surface magnetic susceptibility survey can therefore be used as a direct (digital) measure of sherd frequency, and also then be applied to vertical sections to estimate density of artefacts within discrete stratigraphic contexts. In the machine-excavated trench at the southwestern edge of MGK: 5 a 2-4 m. deep sequence of predominantly conformable colluvial wash deposits was exposed. This sequence provided valuable corroboratory data on: 1. the processes responsible for modification of the archaeological mound surface; 2. the value of core/augering methods; and 3. the long-term history of the swamp margins. Processes Responsiblefor Modification ofArchaeological Mounds The stratigraphy exposed within the trench showed discrete bedding, with variable ratios of inclusions within the dipping beds. bone/ceramic Significantly, the angle at which beds dipped varied with clast content, and individual thin beds of fine to medium size gravel clasts were present. This strongly suggests episodic winnowing by wind and/or wash processes during formation. Also the zone of clast deposition shifts laterally towards the swamp-edge up-sequence, with clear truncation of earlier deposits at upper contacts. This suggests that earlier steeplyinclined colluvial mound-edge deposits are truncated by later lowering of the mound surface. Several phases of re-working are therefore involved in the modification of mound architectures. Locally, steeply dipping beds of large (> 100 mm.) clasts of mudbrick indicate rolling downslope (the angle of dip of the bed reflects slope angle at time of deposition), which suggests "palaeo-slope" modelling of the mound edge since inception can be undertaken using stratigraphic methods. This could, by extrapolation, ultimately permit modelling of critical angles A key observation was that colluof cohesion/repose. vial stratigraphy has migrated out over underlying a model of net swamp stratigraphy-confirming of the areas over time. infilling low-lying The Value of Core/Augering Methods The machine-excavated exposures confirmed the interpretations made from hand augered boreholes along the transect. Although boreholes fail to detect stratigraphic structures (e.g. thin-bedding), the basic stratigraphic units, and estimates of archaeological

58

JOURNAL

OF PERSIAN

content had all been correctly interpreted from the logged boreholes. This confirmed: 1. the validity of data generated from boreholes in other areas; 2. the approach adopted of modelling sub-surface architecture from borehole data archaeological linked to digital terrain mapping (DTM) from the Global Positioning System (GPS) and Total Station surveys. It is therefore concluded that sub-surface DTM approaches can be successfully adopted and extended to other areas at Merv (or other comparable sites in Western or Central Asia) using the techin the last two niques designed and implemented field seasons. Long-term History of the Swamp Margins It is clear from the trench stratigraphy that several key aspects of the site history can be approached by stratigraphy. At the site study of mound-marginal m. of lateral encroachment by collu>25 investigated vium over swamp deposits had taken place. The chronology of this is not yet established, but given the abundance of allochthonous pottery and bone, situ with in remains, dating of the coupled organic seems possible. This phasing of the encroachment should provide indirect estimates of the degradation rates and losses of archaeological stratigraphy from adjacent mound surfaces since site abandonment, and possibly during later occupation phases. In addition, the cores from the swamps indicate high humification, but preservation of some resistant seeds, and probably pollen. Stratification of authigenic salts at lower levels of the swamp, also indicated that episodes of salinisation have occurred prior to the twentieth century, and possibly during periods of The dating of the site occupation/abandonment. swamp cores and assessments of pollen and invertebrate micro-fossil content will therefore form a key objective of future laboratory research. In summary, our preliminary results indicate that both wind erosion, rainsplash and weathering by salt are important processes in forming stratigraphy rapidly when buildings are abandoned. These sediments erode and redeposit as new slopes, and locally infill depression areas nearby. The pre-archaeological landsurface can be identified, beneath both buildings and redeposited mud brick sediments in low-lying areas. As almost all deposits at Merv are derived from natural deltaic sandysilts, chemical and physical analysis of high precision and accuracy offer the best prospects for resolving former human activity areas and industrial activity at Merv. For field survey, augering procedures (locally tested by manual or machine excavation) and controlled by DTM linked to Total Station survey, offer great potential for sub-

STUDIES

surface mapping of the architectural served subsurface at sites like Merv.

structures pre-

The Geophysical Survey, by P. Strange and R. Falkner Geophysical survey has been widely advocated as a useful, rapid and non-destructive means of providing information on sub-surface archaeological features over more extensive areas than could normally be covered through conventional excavation. Its use in the Near East has been largely experimental and its effectiveness variable, particularly when applied within a mudbrick site environment owing to weak distinction in signals received from in situ and destroyed mudbrick (Lanza, Mancini and Ratti 1972; McGovern 1983; Yassi and Majeed 1989/90). However, greater success has been achieved by the authors using resistivity (rather than magnetic) survey at Nicopolis ad Istrum in northern Bulgaria over an area of 5.7 hectares. This experience, together with the inevitable constraints of equipment transport and the need to apply, within three weeks of fieldwork, one method that would cover as wide a range of features as possible with a minimum of computational data prothat this initial should be cessing suggested survey carried out using a resistivity method. The need to investigate relatively large areas, initially at one metre intervals, and as speedily as possible, effectively limited the choice to the twin-electrode (or splitdipole) method. Not only does this electrode configuration permit the highest measurement rates (up to 600 points per hour) but derived images from the measured data are less ambiguous over linear features than those of other configurations. Its principal disadvantage is the close dependence of the "background" readings on near-surface geology. Five areas were selected for survey, two within Erk Kala and three within Gyaur Kala. Each was laid out as a grid of 20 x 20 m. squares, surveyed into the main site map using EDM and GPS techniques. Practical difficulties in making measurements arose immediately when many areas were found to have localised patches of highly compacted wash only 1-2 cm. below the surface. This not only caused problems in inserting the probes but also resulted in many cases in unacceptably resistance high probe-contact values. The effect of the latter is usually minimised (Geoscan Research by operating the equipment RM4) in its high contact resistance mode which effectively increases the compliance of the constant current driver source by a factor of three but reduces the effective sensitivity of the instrument by the same factor. Operating firstly in this mode it became obvious that, although the probe-contact resistance problem could be reduced, the background and resultant anomaly relative resistances being recorded were

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

REPORT

ON THE

SECOND

SEASON

(1993)

59

much too low for sensible interpretation. Only by mudbrick loosely compacted decayed losing moisture more readily than more compact in situ using the instrument in its normal current driver mode was it therefore possible to take sensible mudbricks. measurements and then only in the range typically 1.1-3.0 ohms. The Systematic Surface Artefact Survey of Erk and Gyaur A total of some 18,000 measurements were taken, Kala, byD.J. Tucker and B. Stoll Tucker most of which seem to be reliable. The most unexThe only published survey of Merv is the pionpected result was that mudbrick structures exhibited a lower (rather than higher) value than the sureering work of the YuTAKE VIII expedition rounding deposits: the results thus were best dis(Al'Khamova 1953). That survey was extensive in its played to show negative anomalies. The results from coverage, taking in the entire area within the each of the five survey areas are as follows: Gilyakub Chil'Burdezh wall. Forty years on new quesArea A (1,060 sq. m.). Erk Kala: along the north tions are arising concerning the nature of the settleand east sides of Trench 1. The depth and nature of ment at Merv which can only be answered with more the deposits and some wall alignments could be conintensive sampling and re-evaluation of the material culture in the light of refinements in chronology and fidently surmised from the excavation. The east wall of Room A was shown to continue to the north before functional interpretation. The aims were to gather and assemble a coherent body of data on the nature being obscured by the remains of former excavation was demonstrated further architecture outand character of the surface material culture in Erk dumps; side the trench to the east. and Gyaur Kala; to provide a spatial context for the Area B (580 sq. m.). Erk Kala: along the crest of material culture found in previous excavations and the Erk Kala defensive wall. The chosen area almost integrate this data into a uniform map; to define the extended from the outer location and spatial extent of discrete functional certainly beyond original face of the wall and for 20 m. back into Erk Kala areas; and to establish procedures for the mapping of across the suspected wall. all surface finds in the future and to further a general Area C (4,400 sq. m.). Gyaur Kala: on the west respect for the spatial intra-site context of archof the Trench 5 mound. Aerial slope photographs aeological material. In 1992 the extensive 170 hectare mound within taken in 1992 revealed a spread of rectilinear archi- a tectural remains here. The purpose of this survey was Gyaur Kala was sampled with a systematic grid of 80 to verify and map these walls. A generally satisfactory quadrats (measuring 20 x 20 m.), spaced at 150 m. match was obtained, although the raw geophysical intervals, resulting in 3.2 hectares of the mound data do not reveal walls to the same extent as the being intensively sampled. In this manner quantitative and qualitative statements can be made concernphotographs. Area D (600 sq. m.). Gyaur Kala: an L-shaped area ing the spatial nature of the distribution of material down and across the SW corner of the MGK: 5 culture. This combined with excavated material will mound prior to the excavation of MGK Trench 1. be used to define the nature of Merv's intra-site strucThe raw geophysical data suggest the presence of ture through time. Settlement extent for the Late structures on the summit but not on the lower slopes. and post-Sasanian periods are already well defined. Confirmation of the latter was obtained as a result of The maps (Fig. 1 A-D) outline the occupational Trench 1, see above. extent for the Late Sasanian period (Fig. 1 A) and two Area E (11,400 sq. m.). Gyaur Kala: along both successive phases of the Islamic period, the earlier sides of the main EW "hollow way". The track itself dating to c. ninth-tenth century A.D. (Fig. 1 B) and the later to c. eleventh-twelth century A.D. (Fig. 1 C). either produced unacceptably high probe-contact The fourth map (Fig. 1 D) shows the distribution of resistance values or proved too compacted to get the coins dating to the eighth-ninth centuries A.D. (see probes into the ground. However, possible structures seem to respect this alignment on both sides of the following report by L. Treadwell). These maps are track. based on survey material acquired during the 1992 the inherent problems of very season, which was restricted to the area of the Notwithstanding low "background" and high contact resistances asso170 hectare tell enclosed by the walls of Gyaur Kala. The 1993 survey extended to all corners of Gyaur and ciated with this technique at Merv, the application of Erk Kala. However, the 1993 data awaits processing. resistivity survey to those areas of known occupation but minimal disturbance seems highly promising. The results presented here must therefore be regardof There remains, too, the likely enhancement ed as preliminary. Whilst the precise dating of the apparent contrast under different conditions, such pottery remains insecure, the broad patterns of occuas in the early spring. Resistivity contrast enhancepational presence and absence illustrated by the in ment under the dry conditions experienced maps have not been significantly contradicted by work conducted during the 1993 season. is almost certainly due to more August-September

60

OF PERSIAN

JOURNAL

A

STUDIES

B

O0O

O ..

. . .... . . . ... .

.

O .

.

.

.

..........

.. ... .~..........

....o. N

.N..o..

...........

... .. . . . . . . .

..............?.?..: ...................?.

?

?

OO O

.

............. ............... .............

C

D

O

'"'' '''''''''''''''o 0o-1kBST

N

O

oO

o

N

a

o0 BST

km ast

as

Fig. 1. Surfacesurveydistributionmapsfor Late Sasanian-Islamicperiodsof occupationin Merv:Gyaur-Kala. A. Late Sasanian ceramicdistribution,c. 6th-7th centuryA.D. B. EarlyIslamicceramicdistribution,c. 9th-10th centuryA.D. C. Islamicceramicdistribution,c. 1lth-12th centuryA.D. D. EarlyIslamiccoppercoin distribution,8th-9th centuryA.D. Opencircle= one coin; solid circle= two or morecoins.

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

To illustrate schematically occupational extent for three distinct phases in the occupation of Gyaur Kala a single ceramic type for each period has been selected. Each type is well defined, easily identifiable and relatively common. The extent of the deeply buried Achaemenid, Parthian and Partho-Sasanian occupations was unclear in 1992 and requires further processing. The general occupational

scheme is as follows

Late Sasanian Occupation, 6th-7th century A.D. (Fig. IA: 72 Specimens; Pl. Va). Rippled surface ware (Type 17). Vessel: medium sized jar. Shoulder bodysherd. Fabric: pale brown buff, pale pink-orange. Surface: pale brown, pale yellow buff, lightly impressed with subtle, shallow, vertical ripples (Herrmann et al. 1993: fig. 9, nos. 1, 2 and 5). Light grey tone on Fig. 1 A represents one specimen, while a dark grey tone represents two or more specimens. The rippled bodysherd type is concentrated in a north-south strip between Erk Kala and the South Gate and also appears in the vicinity of the West and East Gates. Sherds bearing a rippled surface are scattered across the entire area of the tell, the sherds are more visible where subsequent Islamic overburden is thinnest-such as on the southern part of the tell and in the area close to Erk Kala. Where the Islamic occuthe Sasanian pation is densest, correspondingly sherd distribution is present but much decreased.

REPORT

ON THE

SECOND

SEASON

(1993)

61

flat based open bowl. Fabric, often red, red-brown ware. Surface: white slipped, incised with curvilinear motifs, splashed with yellow and green glaze. Grey tone = two or more specimens. This sherd distribution is most visible as it represents the last major occupational phase within Gyaur Kala and thus has little overburden. The scatter is concentrated between the West Gate and the site of the Mosque. The scatter falls off rapidly towards the East Gate. The green glazed sgraffiato red ware may be partly contemporaneous with the late phases of the earlier underglaze black painted yellow ware. Copper Coins, 8th-9th century AD. (Fig. JD: 37 specimens) The distribution of the copper coins (Fig. 1Dopen circle = one coin, solid circle = two or more coins) closely mirrors the scattered black underglaze painted ware distribution (Fig. 1B), as opposed to the more restricted distribution of the later sgraffiato ware (Fig. 1 C). On the western side of the main tell, where the black painted ware (Fig. 1 B) is most prevalent, the coins tend to derive from locations on the the very edge of perimeter of this distribution-at the tell and not in the middle of the mound where the later sgraffiato ware (Fig. IC) cultural deposits are thickest. Significantly, no coins have been found to correlate with the later glazed sgraffiato ceramic phase, see report by L. Treadwell below. The 1993 Season

Earlier Islamic, 9th-10th centuryA D., Samanid (Fig. IB: 176 specimens, Pl. Vb). Underglaze black painted ware (Types 2, 11 and 29). Vessel: flat based open bowl. Fabric: pale yellow, pale brown. pale yellow-buff, pale pink-orange, Surface: black painted linear motifs and spots, overlain by a transparent glaze which is often decayed and absent. Grey tone = two or more specimens. This occupation is scattered thinly across much of the mound, but is concentrated along the old hollow way route between the West and East Gates, with another cluster on the tell adjacent to the South Gate. The anomalous empty swathe between the western and eastern sherd clusters across the middle of the site is surprising and may represent an area not associated with mass cultural material discard-such as a cemetery? A mosque and covered cistern (sardoba) complex lie within the west edge of the eastern cluster. Islamic Occupation, 11th-12th century A.D. (and earlier?)--Seljuk? (Fig. 1 C: 184 specimens, P1. Vc). Glazed sgraffiato ware (Types 3, 22 and 47). Vessel:

An arbitrary grid was superimposed on a plan of Erk and Gyaur Kala to facilitate the location of all surface finds, thus enabling all future casual finds to be plotted by the finder to an area of 75 x 75 m. Isolated mounds scattered within vegetated depressions in the four corners of Gyaur Kala were systematically searched and a further 24 quadrats sampled. In an attempt to define the extent of the buried pre-Sasanian occupation in Gyaur Kala systematic searches were conducted in deep gullies and on actively eroding, steep, west-facing slopes of the mounds. A combination of quadrat sampling and systematic searching of arbitrarily defined areas was conducted at Erk Kala. This involved extensive coverage of the citadel, the interior walls and the steep exterior west and south facing slopes. A row of small mounds just beyond the north-west wall of Gyaur Kala were also searched.

THE NUMISMATIC SURVEY Three hundred and fourteen coins and coinfragments were found in 1992 and 1993 at the sites of

62

JOURNAL

OF PERSIAN

Merv, Erk Kala, Gyaur Kala and Sultan Kala. Seventy were recovered from the excavations, MEK Trench 1, the rest from the Surface Artefact Survey or from chance finds. Most identifiable coins belong to one of three main groups, Parthian, Sasanian and Islamic, Sasanian coins predominating. Most of them were minted in Merv, imported coins are rare, although a few Roman coins were found. All the coins mentioned here are bronze (Cu alloy); no silver coins have been found. No less than six experts worked on their conservation and identification. These were S. D. Loginov and A. B. Nikitin (Parthian and Sasanian coins), N. Smirnova (Hellenistic coins), R. Hobbs (Roman coins), L. Treadwell (Islamic coins) and M. Kotchoubeev (conservator). Reports on the Parthian, Sasanian, post-Sasanian, Islamic and Roman coins appear below: Parthian, Sasanian and Post-Sasanian Coins, by A. B. Nikitin and S. D. Loginov. Many coins described here are of unusual types not previously published: a brief description of the principle coin-types most frequently found in Merv is given below. The following numismatic publications Parthian coins: have been used for reference: Sellwood 1980; Parthian "bronze drachmas" of local mintage: Pilipko 1980; Sasanian coins: Gobl 1971; Kushano-Sasanian coins: Gobl 1984. Early Parthian Bronze Coins Mithradates II (123-87 B.C.) * Tetrachalcos. bust to left, diadem Obv.-king's (like Sellwood, types 24-26); rev.-human figure to right, right arm raised, traces of Greek legend arranged in square. * Dichalcos. Obv.-like the former; rev.-tripod. ? Mint-Nisa * Dichalcos. Obv.-king's bust to left, diadem; rev.-Nike walking to right, holding wreath, N in front below, legend: BAXIAE0Z BA2IAEQN APXAKOY EMI4ANOYY QIAEAAHNOX. MintNisa. * Chalcos. bust to left, diadem Obv.-king's traces of (Sellwood, type 26); rev.-caduceus, Greek legend arranged in square. Unknown king (first quarter of the 1st century B.C., Sellwood, type 31) * Dichalcos. Obv.-king's portrait to left, tiara decoto right, Greek legend rated with star; rev.-horse arranged in square. * Dichalcos or chalcos. Obv.- like former; rev.Nike walking to right, holding wreath, traces of Greek legend.

STUDIES

Unknown king (first quarter of the 1st century B.C., Sellwood, type 30) bust to left, diadem; * Dichalcos. Obv.-king's rev.-horse walking to right. Orodes II (57-38 B.C., Sellwood, types 45-47) * Chalcos. portrait to left, diadem. Obv.-king's prancing to right, P in front, traces Rev.-Pegasus of Greek legend. * Rev.-horse standing to right, P in front, traces of Greek legend. * Rev.-horse head to right, P in front, blundered Greek legend. * Rev.-eagle to right, P in front, traces of Greek legend. * Rev.-star, P at right, traces of Greek legend. Phraates IV (38-2 B. C., Sellwood, types 51-54) * Chalcos. Obv.-king's portrait to left, diadem, behind rev.-Nike head; eagle walking to right, P traces of Greek legend. wreath, below, holding * Rev.-archer seated to right. Mintmark ? Traces of Greek legend * Rev.-head of Helios facing, P at right, traces of Greek legend * Rev.-bow, below P, traces of Greek legend. * Rev.-fire-altar, P at right, Greek legend. * Rev.-crescent, P at right, Greek legend. * of the Arsacidae, P at right, traces of Rev.-sign Greek legend. * Rev.-P, traces of Greek legend. Late Parthian : "BronzeDrachmas" of Margiana Mintage, after V. N. Pilipko A group of bronze coins following the types of Arsacid drachmas, minted in Margiana in the firstbeginning of the third century A.D. 1.1. following the type of Phraates IV. Obv.king's bust to left, diadem, eagle behind seated to right, traces of head; rev.-archer legend arranged in square. 1.2. following the type of Phraataces. Obv.king's bust to left, diadem, above-two winged figures holding wreaths; rev.-archer seated to right, P below bow, blundered Greek legend arranged in square. like the former, but different in style. II.1. 11.2. king's bust to left, diadem, star and crescent in front; rev.-archer seated to right, P below Greek blundered bow, legend. like the former, but different in style. II.3. bust to right, diadem, dotted III.1. obv.-king's seated to right, P below border; rev.-archer bow, Greek legend in circle: BAXIAEQZ BALIAEQN. III.2. like the former, but the Arsacid device behind archer on the rev.

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

111.3. obv.- like III.1, but crescent in front of king's portrait; rev.- like III.1, but Greek legend BAXIAEYXE ANABAPHX. 111.4. obv.-star and crescent in front of king's portrait, diadem with triangular knot; rev.-the same as 111.3, but Greek legend blundered. 111.5. obv.-like III.2, but diadem with triangular to right, monogram knot; rev.-archer formed by letters A and P below bow, blundered Greek legend. bust to left, diadem, crescent in 111.6. obv.-king's seated to right, P below front; rev.-archer bow, encircled by blundered Greek legend. bust to left, diadem; rev.111.7. obv.-king's archer seated to right, P below bow, sign of Gondophares behind. bust to right, diadem; rev.-IV. 1. obv.-king's archer seated to right, monogram formed by letters P and T below bow, blundered Greek legend. IV.2. like IV.1 but different style. V.1-3. like III.1 but schematised representations, legends blundered. V.4-6. like V.1-3, but device of the Arsacidae on rev.-behind archer, legends blundered. like V.1-3 but different style. V.7. the type of Vologases IV obv.-following VI.I. (Sellwood, type 84), king's portrait to left, high tiara; rev.-archer seated to right, sign of the Arsacidae behind, traces of Pathian legend above bow. seated to left, VI.2. obv.-like VI.1, rev.-archer sign of the Arsacidae behind, Parthian legend. the type of Vologases V VI.3. obv.-following (Sellwood, type 86), king's portrait frontally, diadem, bunches of hair on top and behind seated to left, sign of the ears; rev.-archer Arsacidae behind, Parthian legend. VI.4. close to V.4 but more schematised. to VI.1-2, to VI.4; rev.-close VI.5. obv.-close traces of Parthian legend of beardless king to left, diadem; VI.6. obv.-bust seated to right, P below bow, rev.-archer of Gondophares above, traces of legend sign behind. Sasanian All bronze coins of Ardashir I (c. 220-240). Ardashir I coming from Merv belong to one of his last issues corresponding to type III.2 after Gobl: obv.king's portrait to right, segmented cap surmounted with huge ball of cloth covering king's locks, traces of altar with flames. Earlier legend; rev.-lion-footed types are not represented among finds-obviously Merv became a Sasanian province only in the later

REPORT

ON THE

SECOND

SEASON

(1993)

63

years of Ardashir's reign. Shapur I (240-273). There are two main types of his coins. Type 1: obv.-king's portrait to right, mural crown with no ear-flaps surmounted with globe, imitation of inscription; rev.-fire-altar, two attendants standing by the sides, their backs towards the altar, holding long staffs. This type corresponds to G6bl's type II.1. Many coins of this type bear traces of being struck over bronze coins of Ardashir I of the type described above. same, but mural crown with earType 2: obv.-the like flaps, type 1.1 after Gobl. The name of Shapur in Middle Persian script is legible on some well-preserved specimens. Rev.fire-altar with two attendants. Coins of type 2 were usually struck on smaller and thinner flans, they can be dated to the second half of Shapur's reign. The King of Merv under Shapur I (c. 240-260). Coins of the same size and fabric as those of Ardashir I and type 1 of Shapur I. Three types are distinguishable: Type 1. obv.-ruler's portrait to right, diadem, Middle Persian legend-mlwy MLKI = the of rev.-horseman to Merv-in front; King or wreath diadem in diadem, right, wearing his right hand, long sword by the side. same, but ruler's head to right; Type 2. obv.-the rev.-the same, but horseman wearing mural crown. same, but ruler's head to left; Type 3. obv.-the rev.- the same, but horseman wearing mural crown. The ruler can be identified as Ardashir, the King of Merv, mentioned in the SKZ inscription of Shapur I among the vassal-kings under Ardashir I. Some of his coins were struck over bronze coins of Ardashir I minted in Merv. Varahran II (276-293). Only one type of his bronze coins is represented among finds, correto after IX/2 type sponding G6bl: obv.-busts of king and queen to right, king in his winged crown surmounted with globe, queen-in high kulah with aniIn front of of prince to them-bust mal's protome. left, wearing high kulah with bird's protome. Rev.fire-altar, two attendants by the sides standing, their backs towards the altar. Shapur II (309-379). The obverse representations of the five types of bronze coins of Shapur II found in Merv differ only in style: king's portrait to right, mural crown, sometimes the name of the king-

64

JOURNAL

OF PERSIAN

shpwhry--is traceable. The five reverse types are the following: Type 1: high fire-altar flanked by two attendants holding barsoms (the first years of Shapur's reign). Type 2: the same, but the stem of the altar tied with ribbon, its ends suspending by the sides (c. 325-350). Type 3: altar with bust of a deity surrounded with flames (c. 350-379). 4: lion-footed altar, like that of Ardashir I, no Type attendants (c. 340-350). Type 5: altar with a waist-figure of a deity holding wreath and staff (or trident), similar to the obverse type of several Kushano-Sasanian bronze issues (the second half of Shapur's reign). This last type is predominating among the finds of coins of Shapur II in Merv. be distinThe mintmark of Merv-mlwy---can guished on the reverse, above the altar, on wellpreserved specimens of types 1, 2 and 4. Shapur III (383-388). Small bronze coins minted on thin flans. Obv.-king's portrait to right, flatwith globe. Rev.-topped crown surmounted schematised diadem or wreath within circular dotted border. Varahran IV (388-399). Obv.-king's portrait to right, winged crown with globe. Rev.-almondshaped leaf, two ribbons rising from its stem by the sides. Obv.-king's portrait to Yazdgird I (399-420). rev.surmounted by-globe; right, segmented cap cross surmounted with crescent, two ribbons rising by the sides from its foundation. Obv.-king's portrait to Yazdgird II (438-457). and his name mural crown, title-kdy yzdkrt'right, wreath. inscribed in front; rev.-schematised 5th century. Obv.-fork-shaped Anonymous. Minted on small thin flans. tamga; rev.-blank. Peroz (457-484). Type 1: obv.-king's portrait to right, wingless crown (crown 2 after Gobl); rev.-schematised wreath, dotted border. Type 2: obv.-king's portrait to right, winged crown, xwarr-behind head; rev.legend-GDH= fire-altar, two attendants by the sides facing This last type often it, mintmark-'mw. found in Merv comes most probably from Tabaristan, from the mint of Amul.

STUDIES

Kavad I (484, 488-97, 499-531). Most of Kavad's bronze coins from Merv follow his regular types: I.1, II.1 and V.2 after G6bl. On some of these coins the mintmark of Merv is visible-ml, or mlw. Khusrau I (531-579). Small bronze coins of types 1.1 and 11.2 after G6bl are present, bearing the mintmark of Merv-ml. At the same time there are coins of type 11.2 with no mintmark or date at all. A series of small coins (type 11.3) have on the rev. a schematised wreath instead of a fire-altar with attendants. Hormizd IV (579-590). His bronze coins found in Merv are few in number, they are of his regular type. We cannot be sure that all those coins come from the mint of Merv. Khusrau II (590-628). Very few of his bronze coins are preserved well enough to distinguish minor details. They were minted on rather thin and broad flans. Obv.-king's portrait to right, winged crown, GDH-behind head, hwslwb-in front, legends: double rim; rev.-fire-altar, two attendants standing by the sides frontally, triple border. Several coins of the same fabric displaying winged crowns, which are conventionally ascribed to Khusrau II, could be as well minted by other Late-Sasanian or even PostSasanian rulers. Post-Sasanian Peroz (late 7th century). Obv.-portrait of a bearded ruler en face; rev.-fire-altar, legends: leftMLKX) or governor mlw m. = king (if m.-for (marzban) of Merv, right pylwc= Peroz. under Halid Varahran, (690-91). governor Obv.- portrait of a beardless ruler to right, crown surmounted with crescent, in front: hlyd = Halid, three star-and-crescent devices in the field; rev.monogram, circular legend: dhwpt ZY wlhl'n = dahibed Varahran. "Military governor" late 7th century. Obv.ruler's portrait to right, winged crown; rev. fire-altar, to left-plmwt, to right -s(yn)pt = "ordered (by) the military govemor". Islamic Coins, by Luke Treadwell The total number of Islamic copper coins retrieved from the sites of Erk Kala, Gyaur Kala and Sultan Kala (MEK, MGK and MSK respectively) is forty-three, forty-two of which came from the surface survey (three from MEK, thirty-eight from MGK and one from MSK) and one from the excavation of Trench I on MEK. The majority of these coins are

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

heavily worn fragments which do not allow precise dating. Furthermore, it proved impossible to clean all the coins before the end of the season, which added to the difficulty of making secure attributions. However, all the coins from MEK and MGK can be dated to either the late Umayyad or the early Abbasid periods (eighth to first quarter of the ninth centuries A.D.) on the basis of their fabric and the few inscriptions which survive. The large number of fragments of fulits found is puzzling. These fragments appear to have been has broken rather than cut. The same phenomenon recovered been observed among contemporaryfulus from the site of Rusafa in Northern Syria (see L. Ilisch, "The Coins", in the German Archaeological Institute's Rusaifa excavation report, forthcoming). At present our understanding of the mechanisms governing the circulation of early Islamic copper is very limited and further excavation material is required before we are able to offer a satisfactory explanation for this and related phenomena. The main concentrations of Islamic coins lie in three pockets across the large tell which has accumulated on either side of the hollow way which divides MGK on an East to West axis, see Fig. 1 D. It is interesting to note that these concentrations of coin finds

REPORT

ON THE

SECOND

SEASON

(1993)

65

coincide to some degree with concentrations of so-called black underglaze pottery from the site (Fig. 1 B). While the dating of this pottery remains uncertain, the evidence of the coinage suggests that these areas were among the more heavily populated quarters of the early Islamic city. The following are examples of coin types which proved to be identifiable with the aid of comparative material collected from the site over the past three years by a private collector, who was kind enough to lend his collection for this purpose. As yet the paucity of Islamic numismatic evidence from the site, does not permit us to say very much about the monetary system of early Islamic Merv or the dating of Muslim occupation of the site. This can only be achieved by the publication of the much greater number of Islamic coins held in the YuTAKE archive in Ashgabat. One important question which remains unanswered is whether the numismatic evidence will help to date the removal of the main Islamic urban centre from Gyaur Kala to a more westerly site. The latest coin struck in Merv which has been recorded in this report is a fals of 206 A.H. If this proves to be the latest copper coin found in MGK, then it might be tempting to interpret it as evidence for the subsequent abandonment of the site of and Remarks

Inv. no.

Location

Wt. (g)

Size

Description

27

MEK Tr. 1

2.54

15.5 cm.

Obv: al-mansur/al-amir qutayba ?/ibn muslim Rev: mi'a wa/cishrin/bi-dirham No marginal inscriptions. In the Inventory Catalogue of YuTAKE Coin Cabinet the name of the amir on other specimens of this type is given as Qasim b. Muslim. However, the die is usually poorly engraved and there is reason to believe that this fals was struck on the authority of Qutayba b. Muslim al-Bahili, Hajjdj's famous general.

2049

MGK (6.D.I)

0.91

15.5 cm.

Obv: (1ii ildh illd/* * l/allah wahdahu) Rev: muhammad/*(-o-) * 2/rasul alldh Illegible marginal inscription. 2/3 fragment. Other specimens of this type have been tentatively dated to 13 (9) A.H.

2332

MGK (11.G.III)

1.68

20.5 cm.

Obv: Id ildh/illTh alldh/wahdahu Rev: muhammad/rasud/alldh/star? Illegible marginal inscriptions. Probably struck in Merv in 156 A.H.

1041

MGK

0.46

11.0 cm.

Obv: (lidildh illa/allah wahdahu/la) shank (lahu) Rev: muhammad ra(sil/alldh) salld all(ah/'alayhi wa sallam) Illegible marginal inscriptions. 1/3 fragment. Style is identical to a fals struck in Merv in 206 A.H. by order of Tahir b. Husayn, governor of Khurasan.

66

JOURNAL

OF PERSIAN

MGK. However, one should also bear in mind that copper coinage ceased to be struck in the central Islamic lands from the mid-ninth century. Whether copper production ceased in Merv at the same time is not yet known. Roman Coins, by Richard Hobbs A small number of Roman coins have been discovered at Merv in both Erk Kala and Gyaur Kala. The first coin of Justinian I was discovered by Schmidt (1908: 196). Approximately fourteen further specimens were recovered during the period of the YuTAKE excavations. One example was recovered from the 1992 MEK Trench 1 excavation, while four more were found in the 1993 surface survey. A full list and discussion is in progress. The material is generally poorly preserved, but most coins are sufficiently legible to be dated to the late fourth and fifth centuries A.D. The majority are the distinctive of the smallest AE4 denomination; diademed right-facing imperial bust is visible pearl on two examples. Although only a fraction of the total number of coins from the site, the Roman coins do add an important element to the wider distribution of Roman coins from eastern Arabia (Howgego and Potts 1992) and Mesopotamia to Central Asia (Simpson, Hobbs and Loginov 1993) and India (Turner 1989). Future research will, it is hoped, clarify this picture, particularly with regard to the nature of contacts between the Roman/Byzantine and Sasanian empires. THE EXCAVATIONS During the 1993 season, six excavations were conducted within Erk and Gyaur Kala. The principal one was that initiated in 1992 in Erk Kala: Erk Kala, Trench 1, see below. The other five excavations took place within Gyaur Kala. Trenches 1 and 2 consisted assessments within the of two geoarchaeological North East Quadrant, close to and south of the extensive mound known as YuTAKE Trench 5. Trenches 3 and 4 were excavation probes by Merkel aimed at identifying in situ Islamic metallurgical and glassworking activity areas south of the major "hollow way" running past the excavated mosque, see report by Merkel below. The main operation within Gyaur Kala was focused on the Trench 5 mound, with the aim of verifying and if possible dating architectural anomalies recorded on the 1992 aerial reconnaissance, see report by StJohn Simpson et al. below. Merv: Erk Kala, Trench 1, by StJohn Simpson, S. Loginov and Richard Hobbs3 The focus of work was the continuation

of MEK

STUDIES

Trench 1 (YuTAKE MEK Trench 10), begun in 1992 and situated towards the eastern fortification wall. Unfortunately, workmen were not available, and a relative lack of excavators, following the death of Valeri Ossipov and a large amount of essential off-site coin processing, reduced the quantity of excavation that could be undertaken. The primary planned objectives were nevertheless fulfilled and comprised the following: 1. Extension of the excavation area towards the east. This was aimed at obtaining further archaeological and environmental data from external deposits contemporary with the eastern wall of Room A. Several Middle Persian ostraca were recovered from a single deposit here. 2. Removal of the walls and lowest floor of Room A prior to excavation of earlier deposits. Attention was paid to removing the walls, course by course, with articulation of the bricks, for the purpose of recording brick-laying techniques (Pl. IVb). All finds from the walls were retained. These included two ostraca, coins, a bone pin, pottery and animal bones, thus providing useful quantifiable data for comparison with other residual material in deposits interpreted as construction make-up layers. 3. Traces of earlier construction, floors and a pit filled with organic refuse, interpreted as a midden, were explored below Room A. Numismatic evidence from these contexts implies a seventh century date, at the earliest within the reign of Khusrau II and possibly as late as immediately after the Arab conquest. The proposed midden pit provided a 100% flotation sample rich in animal bone and carbonised seeds, the forthcoming analysis of which is expected to provide a detailed case-study prior to the eventual final publication of all excavated environmental evidence. 4. Removal of remaining room-fill-rich in wellthe preserved animal bone (including fish)-plus lowest floor in Room B, immediately west of Room A, followed by investigation of the relationcontexts and those ship between underlying beneath Room A. 5. Articulation and excavation of contexts south of Rooms B-C, an area within the trench not investigated in 1992. Two rooms (I, K) were defined, the latter containing a sufa bench inset into its north wall. Floors were reached in both rooms, a connecting doorway defined in plan, and later refuse deposits with posthole features removed. Clear differentiation of material between different contexts was noted in these Rooms: one context contained several further Middle Persian ostraca. Sealed beneath these refuse deposits were two earth ramps, leading north up to doorways into Rooms B-C. These therefore mirror the situation found during the 1992 season where a ramp was

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

traced leading from Room E up to a raised floorlevel in Room D. 6. Cleaning and removal of floor re-modellings in Rooms D-E, followed at the end of the season by a probe through the wall separating Rooms E-K the maximum that was aimed at determining with the visible of deposits contemporary depth walls. The 1993 season was successful in defining more of the plan and development of the extensive domestic mudbrick building initially explored in 1992. The same stratigraphic techniques and recording systems were used that had proven both popular and effective in the first season. Hobbs carried the responsibility of maintaining the paper record. Increased use was made of on-site dry sieving using 1 mm. and 50 mm. wire mesh, aimed at improving recovery rates for small mammals, birds, fish, eggshell and charcoal. Larger numbers of potentially diagnostic glassas a by-product. ware sherds were recovered Systematic sampling for flotation recovery by Nesbitt, see below, of carbonised plant remains was initiated in 1993, as were sediment and micromorphology samples for analysis by Barham. Selected contexts within the excavation area provided a useful means of testing Barham's magnetic susceptibility equipment against archaeological deposits of known character and depth. Finally, Strange and Falkner used the excavated evidence to help calibrate their resistivity data adjacent to the trench. The 1993 field season at MEK 1 was successful, therefore, in providing further detailed information of the partially on the stratigraphic development excavated house. New coins prove that the building existed later than previously expected. The stratified artefacts and environmental evidence should provide a clearer picture of developments in the local material culture and economy during the Late to Post Sasanian periods. Ceramics and Small Finds from MEK: 1, by StJohn Simpson The ceramic fabrics and forms largely complement those reported from the first season. 14,124 sherds (>152.2 kg.) from this season were processed with Munsell recording of diagnostics (8% of the total). Sherds were present in virtually all contexts; higher frequencies in the uppermost 10 cm. over the entire trench and in refuse deposits are indicative of deflation and sweeping-out practices. Among the unusual types was a single sherd with a heavily decayed glaze. Its context is seventh century and its fabric is suggestive of a central/southern Mesopotamian origin; it is worth remarking that Sasanian glazed wares are remarkably rare outside the latter region.

REPORT

ON THE

SECOND

SEASON

(1993)

67

Partly reconstructable forms included small flatbased, bowl-like lamps with incurved rims, a pinched lip wick-support and partially blackened interiors. This is the only type of lamp represented among the 1992-3 finds: given the importance of local cotton cultivation (see Nesbitt below), it is likely that cottonseed oil was the preferred fuel.4 This contrasts with preferences for sesame oil or pitch in other parts of the Sasanian empire (Newman 1932: 102-4; cf. also Hauser 1993: 390). It is not known whether variations in locally available fuel types affected regional typological developments in Sasanian lamps. Finally,joining fragments were found belonging to a large vessel with a high pedestal foot and trough base. This unusual type is known from seventh-ninth century in Iran and Mesopotamia contexts (Wilkinson [1974]: 312, 347, no. 66; Whitcomb 1985: 136-7, 142-3, fig. 53aa, w-z): the trough bases are believed to have been filled with water as a means of deterring insects. Thirteen Middle Persian ostracon fragments, including twojoining sherds, were recovered this season. Apart from two examples, found within the mud mortar of the west wall of Room A, these came from discrete refuse contexts located immediately east of Room A and in the upper fill of Room I. Stratigraphically, these contexts are approximately and date within the seventh cencontemporaneous tury, according to the coin evidence. A report on the 1992 ostraca is given below, those recovered in 1993 are still being studied. Twenty-two ceramic figurine fragments were recovered in 1993. Nineteen belonged to animals, mostly consisting of broken legs and/or hindquarters of quadrupeds, but including three recognizable horse figurines, two ovicaprids, a dog and a distinctive monkey-like creature. The last of these is paralleled by other examples known from earlier investi1962: 168, fig. 32) and is gations (Pugachenkova curiously reminiscent of figurines found by Stein (1907: vol. II, pls. XLVI-XLVII; 1928: vol. III, pl. II) at Yotkan in the Taklamakan. The three remaining figurines are again well-known local types, being variations of the so-called "Great Margiana Goddess" type that apparently date from the second century onwards (Pugachenkova op cit.: 140-141, figs. 16-18: Herrman, et al. 1993: 55, pl. XIVc). Excavated figurine fabrics were usually pale yellow (5Y8/3-4) with occasional light finger-tip or nail impressions on their undersides. Other finds categories include ground stone, worked bone and large fired clay tetrahedra similar to examples found on the 1992 survey. A single, eggshaped, hard-fired slingshot was also recovered (cf. Schmidt 1908: 200-01, 208-10). Supplementing the 1992-3 archaeobotanical evidence for local (cotton) textile industries and the earlier discovery of a

68

JOURNAL

OF PERSIAN

woollen textile fragment at Erk Kala (Usmanova ceramic spindle1963: 70-2), five hemispherical whorls were found this season. Rare evidence for Late to post-Sasanian board games comes in the form of chipped sherd discs, and rounded some with deliberately smoothed and a cubical and unbaked cones, edges, clay spheres ceramic die with opposing sides adding up to seven, as on examples from Pasargadae (Stronach 1978: 182-3, 214-15, fig. 92:8, pl. 169 e-g) and Qasr-i Abu Nasr (Whitcomb 1985: 191, 196-7, fig. 74d; cf. also Curtis 1984: 53-4). A further die was found in MGK YuTAKE Trench 6 (Varkhotova 1958: unnumbered plate), but unfortunately this rather interesting type of archaeological evidence is rarely published. Chess was played at this date from Central Asia to Iran and Mesopotamia, judging by sporadic finds of chessmen (e.g. Buriakov 1980) but the 1993 finds may instead belong to race-games, such as backgammon. A polished bone pin fragment with the head evidently in the stylised form of a clenched fist was found within the east wall of Room A. Bone pins are a characteristic minor element of Roman, Parthian, material culture Sasanian and Kushano-Sasanian and several have (Crummy 1981; Zavyalov 1993) been reported from earlier excavations at Merv (Schmidt 1908: 201, 209, pl. 56: 7; Usmanova 1955: 19, 48: Katsuris and Buriakov 1963: 153). A small number of beads (stone, glass, coral) complete this year's excavated inventory of items of personal adornment. Further pink or rosy coral beads were recovered from the 1992 surface survey and earlier excavations (Schmidt 1908: 199, 209, pl. 54, fig. 17: Varkhotova, 1958: nos. 149-62. A small number of glassware sherds were also found. Metal (copper alloy, iron) was again represented among the excavated material only in a fragmentary and highly corroded state. The impression is one of careful recycling in antiquity, followed by heavy deterioration owing to adverse ground conditions. Recognisable metal artefacts other than coins were rare but include a small trilobate copper alloy arrowhead and a fibula bow, both probably residual from earlier periods. However, evidence for the casting of small metal pendants was found this year in the form of a mould (1. 5.8 cm.), well stratified immediately above a floor in Room K. This has been made by shaving down the edges of a Sasanian ceramic jar strap handle and incising one surface to allow the simultaneous casting of two different objects (P1. Vd). One of these (1. 1.9 cm.) appears to have a central leaf shape with small, plain, equal-length arm crosses ("Greek crosses") at the terminals. The second was a cross (2.1 x 1.7 cm.) with equal-length, splayed arms, a pair of small blobs on the tip of each arm and a fur-

STUDIES

ther five blobs on the cross itself. This is the first such mould to have been found at Merv, although stone jewellery moulds of similar casting principle are known from other sites (e.g. Rteladze and Kato 1991: 321, no. Pugachenkova, in the Merv mould is uncertain, The metal used 303). but a bronze pendant in the form of a "Greek cross" with splayed arms but lacking the blobs was excavated at Qasr-i Abu Nasr (Whitcomb 1985: 176, 178-9, fig. 66q). The form of the cross is compatible with the Late Sasanian date suggested by its findspot: c. sixth century parallels can be found in the Byzantine world (e.g. BM.EA. 57298) and more elaborate types occur on eighth-ninth century stucco plaques in churches in lower Mesopotamia and the Persian Gulf (Okada 1990, 1992; cf. also Wilkinson [1974]: 335, 362, no. 200). Small plain "Greek crosses" and "Latin crosses" (i.e. crosses with an elongated lower arm) often resting on a base also occasionally feature on Sasanian glyptic, sometimes associated with Christian personal names (Lerner 1977: 3-8; Shaked 1977: 17-24). The discovery of such an obviously Christian mould at this site is not particularly surprising as, by the sixth century, Merv had become a metropolitanate within the Nestorian church and a major centre for the transmission of the faith deeper into Central Asia (Fiey 1973). However, firm archaeological evidence hitherto has been lacking from Merv. A ruined, vaulted, mudbrick building at Kharobakoshuk, c. 15 km. north-west of Erk Kala, has been interpreted as a church but its function and date are unconfirmed 1967: 86-7).5 The (Pugachenkova identification of a Late Sasanian "oval building" complex within the north east corner of Gyaur Kala (P1. IVa) as a Christian monastery, mainly on the basis of some finger-marks in the wall plaster of one room, is even more dubious (Dresvyanskaya 1974).

Epigraphic Finds, by A. B. Nikitin The epigraphic finds of 1992 include four fragments of Middle Persian ostraca and short Bactrian inscriptions on bone coming from MEK: 1, together with a handle of a large jar with five lines in Parthian script, engraved before firing. The latter was a chance find made near MGK YuTAKE Trench 6 on old dumps and is currently being studied by Professor V. A. Livshits. In 1993, thirteen more Middle Persian ostraca were recovered from MEK: 1, as well as one Parthian ostracon recovered from the new excavations at MGK: 5. These are being worked on. Presented below is a report on the 1992 excavated epigraphic

finds.

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

REPORT

ON THE

SECOND

SEASON

(1993)

69

Middle Persian Ostraca (P1. VII) 1. No. 23 1, from MEK: 1 (P1. VIIa), Context 114. Size 3 x 2.5 cm. Fragment of an ostracon. External surface, two incomplete lines in black ink: Transliteration: 1. ] Y lywk'. [ 2. ]n/w yzd't Y [ Internal surface, two incomplete Transliteration: 1. '] ZLWNt ' bn/w/r [ 2. ]... THNN'Il. [

Translation: ] (son of) Rewag. [ ]Yazdaid, (son of?) [

Transcription: ] i Rewag. [ ]n/w Yazddd i [ lines in black ink:

Translation: ](he) went.. [ ]... miller ... [

Transcription: ]sud b.[ ]... drddr ...[

Comments: Like most other Middle Persian ostraca from Merv, this fragment apparently came from some household document containing personal names and accounts of various products. or patronymic from rkv N.Pr. Revag-hypocoristic be an < av. *raivant "rich" N.Pr. Yazda-d-can abbreviated form of Yazd-dad or Yazdacn-&id-= Sudan "to "created by the god/gods". 'ZLWNtn' could be the the word second ideogram BR' = go", be. THNN'l "miller"-from THNNtn' = drdan "to grind, mill" 2. No. 238, from MEK: 1 (P1. VIIb), Context 114. Size 5.5 x 3.5 cm. Fragment of an ostracon. External surface, end of one line: ] (t)n mtr'n = ] .. (son of) Mihr. 3. No. 237, from MEK: 1 (P1. VIIc), Context 114. Size 9.5 x 8.5 cm. Fragment of an ostracon. External surface, six incomplete faded lines in black ink. The reading suggested here is tentative. Transliteration: 111 [ 1. ]IIII 100 YHTWMVyt 2. ] '/hn zhk (or 'YK ?) MINWltwnd Y[ 3. ]1L.. '/hmn/w[...]n(t)..[ 4. ]In/wcn plhwd't 'lmn(?) (B)RH.. [ 5. ] (...) .. plh 't 30 (or LK ?) Wlsn(?) .... [ 6. ](')k bwlcmtr 1[ Comments: The character of the document is not clear. It contains several Iranian personal names: Farroxdaid, Frahaid, Rain (?) Burzmihr, as well as numerical indications. 4. No. 134, from MEK: 1, Context 1. Size 5.5 x 3.0 cm. Fragment of an ostracon or of an inscribed vessel. External surface: three letters: ]'y 1 [. Bactrian Inscriptions 5. No. 274, from MEK: 1, Context 92. Fragment of a rib, size 4 x 1.4 cm. One line in Bactrian cursive script drawn in black ink: abbbo abbaba abbb[. The inscription makes no sense. Most probably it is a school exercise, with

someone b.

learning how to connect the letters a and

6. No. 313, from MEK: 1, Context 32. Fragment of a shoulder bone, size 5.5 x 3.0 cm. Three short lines drawn in black ink: 1. (Bactrian cursive) loixo 2. (Middle-Persian) . dh 3. (Unknown script, faded, three characters) Comments: Bactrian loixo "land, country"< OIr. dahyu-. Cf. loixobosari "the helper of the country" in the Surkh Kotal inscription (Gershevitch 1966: 101). Middle-Persian dh = deh presents the direct translation of the Bactrian word loixo. We have here obviously also an example of a school exercise left by some scribe learning the Bactrian language. Excavations at Merv Gyaur Kala: Trench 5, by StJohn Simpson, K. Kurbansakhatov and V. Zavyalov New exploratory excavations, lasting only a week,6 were begun on the low-lying mound in the north-east quadrant of Gyaur Kala known as the YuTAKE Trench 5 mound, because of an earlier, unpublished YuTAKE sondage there. Our more extensive archaeological operation took place on the west-facing slopes of the mound. Its purpose was to verify and, if possible, to date architectural anomalies recorded here during the 1992 aerial reconnaissance of the In these structures were the 1993, city. presumed focus of both a detailed GPS survey of the mound and a geophysical survey, see reports above. A temporary grid was established over the mound and four areas excavated (P1. IVc-d). Three measured 5 x 10 m., the fourth measured 10 x 20 m. Remains of mudbrick architecture was found in all areas, with a close match obtained between the aerial/geophysical surveys and the scraping exercise. In order to ascertain deposit depth within one of the identified structures, one room was partially excavated to floor-

70

JOURNAL

OF PERSIAN

level at a depth of 0.80 m. below the top of the walls. A hearth was sunk into this floor and the goodquality plaster found to curve up at the wall-junctions. The deposit within this room consisted of seven courses of deliberately laid mudbricks, at least one of which bore a fingermark impression on the underside.7

SPECIALIST REPORTS Analytical Investigation of Crucible Steel Production, Preliminary Results, byJ. F Merkel, A. Feuerbach and D. Griffithss to the early Islamic texts, three According methods are described for indirect production of steel (fiildd) as discussed by Allan (1979) and alHassan and Hill (1986). The most common, traditional method is solid state carburisation of wrought iron. There are many variations on this method. It is also known as "case hardening" or in other instances "cementation". This is a diffusion process in which wrought iron is packed in crucibles or a hearth with charcoal, then heated to promote diffusion of carbon into the iron to produce steel. Alternatively, another indirect method uses wrought iron and cast iron. Although there has been some uncertainty on the translation of the word (dus) in Islamic texts, the cast iron interpretation is generally accepted (Allan 1979). In this process, wrought iron and cast iron may be heated together in a crucible to produce steel by "fusion". This is also called a "visco-liquid diffusion process" (Needham 1958) and may operate below the melting point of true cast steel (Smith 1960). A third indirect method to produce steel is partial decarburization of cast iron or a high carbon steel bloom. Again, there are variations of this method, but generally it is considered very difficult to control (see Rostoker and Bronson 1990). Outside the Islamic textual evidence, inadvertent direct production of steel during bloomery iron smelting represents another possibility, but it is not considered here in the context of an indirect or multi-stage process to routine production of steel. A detailed account of the by Rostoker and many variations is presented Bronson (1990). It is against these three main methods for indirect steel production that this prelimievinary report evaluates the archaeometallurgical dence and its interpretations for early Islamic times at Merv. Two areas with surface concentrations of crucible fragments, green "glass" slag fragments and slagged furnace fragments have been located in the survey (P1. VIIIa) in MGK 7.F.II during the 1993 season. The scatter of pottery around and within the archaeois predominantly remains Early metallurgical

STUDIES

Islamic. A small-scale excavation was conducted, but the surface remains were unstratified and directly upon earlier, non-metallurgical layers which were not investigated. The remains are interpreted simply as superficial residue. dumps of metallurgical Charcoal recovered from the slag will be dated by radiocarbon. Fragments of crucibles, furnace wall and tuyeres as well as the glassy slag were collected from the metallurgical dumps by J. Merkel for technical investiScience gation in the Wolfson Archaeological Laboratory at the Institute of Archaeology, U.C.L. The analytical work is undertaken, in part, as third year B.Sc. research by A. Feuerbach. The metallurgical process has now been identified as crucible steel production. The identification of the metallurgical process is based foremost upon the presence of abundant carbon steel droplets in the glassy green slag adhering to the inner surface of crucible fragments collected from the two areas. The steel is identified using etched metallographic sections (Pl. VIIIb) and microhardness measurements (Hv 140-320). Against metallographic standards, the carbon concentrations of the steel droplets seem to range from <0.1-0.8%. The structures are characteristically variable proportions of pearlite and ferrite. Microanalysis also detected silicon and sulphur in the iron droplets. The droplets range in diameter up to approximately 0.3 mm., but most droplets are too small for microhardness and adequate etching for microstructure identification. Quantitative microanalysis of the steel droplets for phosphorous and other possible alloying elements will follow. The Merv oasis is alluvial and without local iron ore deposits, so iron smelting locally is very unlikely. Steel production at Merv was probably based upon recycled wrought iron scrap. Qualitative compositional analysis (SEM/EDS) of the associated green "slag" adhering to the interior surface of a crucible fragment identified silicon, aluminium, calcium, iron, manganese, magnesium and potassium. The "slag" does not have a crystalline structure: it is a glass with a variable composition. Concentrations for calcium, potassium and aluminium from the SEM/EDS, however, are quite different than those typical for glass. Fragments of the glassy green slag are observed to melt under reducing conditions in an electric furnace at a temperature of about 1250o C. This is a temperature below the melting point of true cast steel (Smith 1960). Several glassy slag fragments exhibit viscous flow patterns which will be investigated. Some crucible fragments indicate the upper level of the molten contents as a "fin" of glassy green slag (P1. VIIIc). It is this slag which contains abundant steel droplets. Above this "fin" of slag, the adhering pattern of corroded iron is interpreted as "splashes" onto the upper crucible

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

wall. Of course, these forms will be investigated in detail for relict structures in the corrosion products. Below the "fin" appears a honeycomb pattern in the glassy slag on the inner surface which appears similar to that on the inner surface of other crucible base fragments. Similarities in descriptions of a characteristic honeycomb pattern in the slag (see Percy 1864) as well as striations in some glassy green slag fragments suggest incomplete fusion (see Smith 1960). Crucible fragments are variable in thickness and condition but appear to represent a single type. Wall thickness ranges from 0.5-2.5 cm. Fragmented circular segments with a reconstructed outer diameter of 6-8 cm. and a thickness of about 0.5 cm. are interpreted as crucible lids. These fragments have a central perforation about 1 cm. in diameter and the outer edges exhibit evidence of a clay seal to a crucible wall. The use of a luted, perforated lid is a variation in accounts of traditional steel making in crucibles (see century during the nineteenth Bronson 1986; Rostoker and Bronson 1990). The thinner fragments and lids are interpreted as having been fired at high temperatures under reducing conditions. The thickest fragments have a light-coloured core and dark, reduction fired inner surfaces. Investigations of the refractory properties of the crucible fragments and furnace wall fragments will be undertaken in order to characterise the materials relative to their performance at high temperatures and to estimate the duration of firing. These would be important considerations for steel production. Fragments of furnace wall consist of the local mudbrick but lined with crucible fragments on the interior. The inner surface is covered with a black, adhering layer of slag arising from high furnace temperature and fuel ash. It has not yet been investigated. The furnace wall fragments were about 5-10 cm. in length and 5-10 cm. in height. Several are interpreted as representing corner fragments. There are examples of crucible base fragments slagged onto a lower ceramic support (P1. VIIId), which was placed on the furnace bottom. No exact joins have yet been made between the furnace wall fragments. Although none of the furnace wall fragments were in situ, it is possible to suggest rectangular furnace dimensions. This proposed reconstruction is influenced by the furnace descriptions included by al-Hassan and Hill (1986). Further assessment would require more extensive excavations One large fragment of a tuyere was found along with the other metallurgical remains. It is well-fired with a large diameter of about 4 cm., but, as the ends are missing, it shows no evidence of direct contact with the other furnace wall fragments. While these archaeometallurgical samples are attributed to crucible steel production, there remain many questions and lines of evidence to investigate.

REPORT

ON THE

SECOND

SEASON

(1993)

71

There are many samples still to section and analyse. Distinction between the three possible indirect methods (cementation, fusion or decarburisation) for crucible steel production at Merv would depend upon further discovery of minute fragments of raw materials in the glassy green slag or adhering to the crucible walls. Raw materials might include wrought iron with elongated slag inclusions, carbon impressions or components (such as rice husks in the later Indian examples presented by Lowe et al. 1991) or perhaps even cast iron coatings on wrought iron droplets. Final products such as ingots or fragments may not necessarily be conclusive to distinguish between fusion or cementation. For the interpretation of these archaeometallurgical specimens from Merv, the most salient observations relate to the abundant steel prills of variable carbon content in a relatively small amount of glassy green slag with a of approximately 12500 C melting temperature adhering to the inner wall of crucible fragments. Further excavations of the archaeometallurgical remains are anticipated. Archaeobotanical Research in the Merv Oasis, by Mark Nesbitt The presence of an archaeobotanist made posof a flotation machine, sible the construction equipped with a pump. This resulted in the processing of far larger quantities of soil than had been possible in 1992, when water had to be raised by bucket from the canal. This in turn made it possible to respond to a kind invitation from Professor Maurizio Tosi to collect samples from the excavations at Tahirbaj Tepe.9 The Merv flotation machine is an adaptation of the Siraf machine (Williams 1973), working on the same principle in which carbon is separated from the soil matrix in water. The floating plant remains flow into a 0.5 mm. sieve (the minimum mesh that could be used here without clogging), the heavy silt and sand falls to the bottom of the tank, and other heavy bone and small artefacts-are colitems-especially lected on an internal 2.5 mm. mesh. As sample sizes at Merv are, of necessity, small (typically 10-20 litres) and portability is a major factor for equipment, a smaller machine than the usual 40-gallon oil-drum type was built. Constructed by a neighbouring workshop, it is a cube 40 cm. square, and was operated using water from the newly operational shower tank. Flotation rates of 50 litres/hour for normal samples and 20-30 litres/hour for exceptionally small or difficult samples are, in line with the small size of the tank, about half of typical rates for the Siraf machine. In general, samples disaggregated and floated well, although in about ten samples from Merv persistent clods of earth floated into the flot sieve, and six other

72

JOURNAL

OF PERSIAN

samples from Merv had to be discarded where this problem was particularly bad. Merv: Sample size was limited by the necessity to carry soil samples, as well as other dig equipment, some 300 m. from the excavation trench to the nearest point accessible by vehicle. In practice, three samples each of 15-20 litres was the daily maximum. Samples were taken from as wide a range of contexts as possible, including all pits and hearths. In all 1,074 litres of soil from 100 samples, representing sixty-two discrete deposits, were recovered from the 1992 and 1993 seasons. Particular effort was devoted to floating the entire contents (323 litres) of the shallow pit consisting of units 299 (upper fill) and 333 (lower fill), which was rich in bone and carbon. Preliminary figures for carbon, bone and ceramics density, calculated from the heavy residues, are shown in Table 2 and Fig. 2. Despite the small number of samples from some deposit types, and incomplete data sets (some samples collected on the final day of excavation could not be fully processed), some interesting patterns are already apparent. In terms of carbon density, the samples clearly fall into three classes: lowand floors, make-up, and mud-brick; medium-refuse and burnt These results pits; high-hearths areas.l0 are consistent with those from other sites and are important for two reasons: firstly, they indicate that the excavator's assessment of deposit types is broadly correct (hence the consistent figures; secondly, the low amounts of carbon in the mudbrick-derived material (i.e. the three low density deposit types) suggests that old mudbrick is not an important source for the charred plant remains at the site, and that therefore most of the plant remains are essentially contemporary with the excavated architecture. There is much less of a distinction in pottery and bone densities between refuse areas and mudbrick derived floors and make-up, suggesting that residual of refuse pottery may form a high proportion and that be a more reliable source deposits, pits may of pottery for dating purposes. More detailed analysis depends on increasing the number of sampled

STUDIES

900 Pits

800-

o o

700 600

Hearths

E 50soo ,

400 Refuse + + Floors

S300-

a

+Make-up

200

100

S100+ 0

0

Burntarea 100

200 300 400 500 Carbondensity (grammes/litrex 100)

600

700

Fig. 2. Average carbon and bone densities in different context types. Mudbrick samples have been excluded as bone densities are not year available. Note that densities have been multiplied by one hundred to make more manageable numbers.

deposits, especially the low density types which are currently under-represented. All the carbon samples were quickly scanned: of the sixty-two deposits, thirty-seven contained charred seeds. Simple presence/absence for each deposit has been used to assess relative abundance of the different crop species (see Table 3). TABLE 3 Number of seed-containing % presence deposits (n=37) Cotton 28 76 Hulled barley 17 46 Bread wheat 14 38 Lentil 9 24 Pea 3 8 3 8 Grape 3 8 Hackberry 1 3 Melon/cucumber Broomcorn millet 1 3 Peach 1 3 Almond 1 3

TABLE 2

Number Carbon density sampled (g./litre x 100) n. n. productive density Hearths 17 18 17 689 Burn areas/ashes 5 5 5 342 Pits 6 6 6 137 14 Refuse deposits 14 15 93 7 7 7 34 Make-up Floor 6 7 6 30 4 4 Mud-brick 4 16 Deposit type

Bone density g./litre x 100) n. n. productive density 10 8 634 3 3 83 6 5 847 13 13 299 6 5 222 6 6 263 0

Ceramics density g./litre n. n. productive density 11 8 9 1 3 3 6 5 17 13 13 5 6 5 6 6 6 5 0

Densities of carbon, bone and ceramics in flotation heavy residues (n. = number of deposits for results were obtainable n. productive = number of those deposits with more than zero contents;

THE

INTERNATIONAL

MERV PROJECT

PRELIMINARY

Weed seeds are present in 65% of the deposits; wheat and barley chaff in 11% and dung in 2%. The high proportion of samples containing weed seeds means that ecological studies will be possible. The abundance of cotton strengthens the interpretation made in the 1992 preliminary report that most of the plant remains derive from burning dung and cotton sticks. The abundance of cotton in the samples is not, therefore, a direct reflection of its economic importance: at such a large urban site food plants such as wheat and barley must have been most important. It is also likely that wheat and barley are over-represented because of their presence in dung. The difference in abundance between lentil and pea is more likely to reflect a difference in economic importance because these are both pulses, and should therefore be subject to the same preservation biases. The fruits known from Merv now include peach (one charred stone), grape and a probable fruits were found almond. Several uncarbonised during excavation, but crumbled to dust prior to consolidation. Even taking into account the factors favouring preservation, cotton is still an important crop: its sheer ubiquity in the samples argues for a widespread presence on site. It is present in both Late Sasanian and Arab-Sasanian samples. Wheat, barley and lentil can be considered staple food crops; the relative importance of grape, pea and millet is uncertain. as a single grain-adds The presence of millet-only another summer-season crop to cotton and cucumber. Millet is common in the Near East and adjacent areas from at least the Iron Age (Nesbitt and Summers 1988), and it is surprising that it is not more common here. The main results of the seed analysis have been to confirm the conclusions drawn on the exceptionally rich single sample discussed in last year's report (Nesbitt 1993), in particular the importance of cotton. It is hoped that flotation in 1994 will result in a longer sequence of samples, especially from MGK: 5, which will enable us to look at the possibility of change through time. Further flotation may well also produce new evidence of those crops present at very low frequencies, such as millet and melon/ cucumber. An extensive vegetation survey of the Merv area was completed, focusing on the extensive shrubby growth, especially in areas affected by the rising water-table caused by the Karakum canal. The oasis vegetation was compared to that at the Repetek nature reserve, which contains relatively dense desert woodland. While the dominance of cotton monoculture has greatly reduced the potential for ethnographic or ecological studies of irrigation agriculture in the oasis, most village houses still use traditional cooking techniques and ovens, and there is a great deal of local knowledge of the fuel properties of

REPORT

ON THE SECOND

SEASON

(1993)

73

different woody species. Ethnographic work with Akmohammad Annaev carried out on this, and other medicinal and food properties of wild plants, is of obvious applicability to understanding our charcoal and dung assemblages. Tahirbaj Tepe: The recent Italian excavations at Tahirbaj Tepe, a prehistoric site in the north of the oasis, have aimed at obtaining a good stratigraphic sequence of pottery and radiocarbon dates for the Iron Age. Initially defined in the 1950s by Professor V. M. Masson's excavations at Yaz Tepe, the Yaz sequence runs from the later part of the Late Bronze Age through to the late Iron Age (mid-second to midfirst millennium B.C.). This period has been divided into three phases on the basis of pottery types, with the earliest, Yaz 1, dominated by handmade pottery. Ten flotation samples were taken on 27 September and floated at Merv. Two samples were particularly rich: sample 5 from a Yaz 2 burnt destruction level, contained dung, hulled barley and broomcorn millet. Sample 9 (Yaz 1), from a black lense deep within a structural platform, contained spherical dung pellets (from sheep or camel?), many weed seeds (some hulled barley grains and with dung attached), broomcorn millet. Although Tahirbaj Tepe today lies well north of evidence the Murghab river delta, archaeological indicates that in the Iron Age river waters would have reached the site. The plant remains presumably result from local cultivation. Comparative material from other sites is limited to Merv and the nearby Bronze Age (early second millennium B.C.) site of Gonur Tepe (Miller 1993). At Gonur Tepe plant remains were recovered by 2 mm. dry-sieving, and the absence there of millet seeds is therefore not necessarily significant. Crops that were present include much hulled barley, little free-threshing wheat and emmer wheat, many lentils, little chickpea, pea and grass pea, much grape and few plums. The range of plant remains from Tahirbaj is too limited for detailed comparison, but several interesting points emerge. Firstly, cotton seeds are absent from both Tahirbaj and Gonur. Introduction of cotton into the area must have taken place at some point between the late Iron Age and the late Sasanian period. Barley is the most frequent cereal at all three sites, while broomcorn millet is present in the area from at least the Iron Age onwards. Lentils are the most abundant pulse, both at Gonur and Merv. There is evidence for shortage of wood fuel in the abundance of dung in samples from all three sites. If the Parthian gap can be filled, either at Merv itself or elsewhere, the prospects for studying long-term agricultural change in the Merv oasis are encouraging. Clearly, despite modern (or ancient) salinisation, preservation of charred plant remains is not a problem in this region.

74

JOURNAL OF PERSIAN STUDIES

Bibliography Allan,J. 1979. PersianMetal Technology,London. Al-Hassan, D. Y. and Hill, D. R. 1986. Islamic Technology:An IllustratedGuide,Cambridge, 251-60. Al'Khamova, Z. A. 1953. "Polevoi otchet VIII otrayada po izucheniyu rabada gorodishcha Starogo Merva v 1947 g." [Field Report from Team VIII on the Study of the rabadin the City of Old Merv in 1947], Tr. YuTAKE2, 404-12. Bailey, F. M. 1992. Mission to Tashkent,Oxford. Bronson, B. 1986 "The Making and Selling of Wootz, A Crucible Steel of India", Archaeomaterials 1, 13-51. Buriakov, Ju.F 1980. ["On the Dating and Attribution of Some Chess Sets (in the Light of Finds of 1977 as Afrasiab], SovArk 1980: 3, 162-70. Crummy, N. 1981. "Bone-working at Colchester", Britannia 12, 277-85. Curtis,J. 1984. Nush-iJan III, TheSmallFinds, London. Dresvyanskaya,G. A. 1974. [Christian Community's "Oval-Shaped" House in Old Merv], Tr. YuTAKE15, 155-81. Fiey, J. M. 1973. "Chr(tientes syriaques du Horasan et du Segestain", Le Musion 86, 75-104 [reprinted: Communautessyriaques en Iran et Iraq des origines d 1552. London: Variorum Reprints, 1979] Filanovich, M. I. 1959. [Preliminary Stratigraphic Excavation at Erk-Kala]. Diploma work, Historical Faculty, University of Tashkent. Gershevitch, I. 1966. "The Well of Baghlan", Asia Major 12, 101 f. G6bl, R. 1971. Sasanian Numismatics,Braunschweig. 1984. System und Chronologie der Miinzprdgung des Wien. Kuschanreiches, Hauser, S. R. 1993. "Eine arsakidenzeitliche Nekropole in Ktesiphon", BaM 24, 325-420, pls. 125-37. Herrmann, G., Masson, V. M., Kurbansakhatov, K., et al. 1993. "The International Merv Project, Preliminary Report on the First Season (1992)", Iran XXXI, 39-62. Howgego, G. J. and Potts, D. T. 1992. "Greek and Roman Coins from Eastern Arabia", AAJ 3, 3 (October), 183-9. Katsuris, K. and Buriakov,J. 1963. ["Study of the Artisan's Quarter of Classical Merv by the North Gate of Gyaur Kala"], 7r. Yu7AKE12, 119-63. Baghdader Forschungen 5, Kr6ger,J. 1982. SasanidischerStuckdekor, Mainz. Lanza, R., Mancini, A., Ratti G. 1972. "Geophysical Surveys at Seleucia", Mesopotamia7, 27-41, figs. 12-13. Lerner,J. A. 1977. ChristianSealsof theSasanian Period,Nederlands Historisch-Archaeologisch Instituut te Istanbul, Leiden. Lowe, T. L., Mark, N. and Thomas, G. 1991. "An Historical Mullite Fibre-Reinforced Ceramic Composite: Characterization of the 'Wootz' Crucible Refractory", in Materials Issues in Art and II (P. A. Vandiver, et al. eds.), MaterialsReseachSociety Archaeology Vol. 185: 627-32. McGovern, P. E. 1983. "Test Soundings of Archaeological and Resistivity Survey Results at Rujm Al-Henu", ADAJ27, 105-41. Miller, N. F. 1993. "PreliminaryArchaeobotanical Results from the 1989 Excavation at the Central Asian Site of Gonur Depe, Turkmenistan, "InternationalAssociationfor the Study of Cultures of CentralAsia InformationBulletin 19, 149-63. Morony, M. G. 1984. IraqaftertheMuslim Conquest,Princeton. Needham, J. 1958. The Development of Iron and Steel in China, London. Remains", in Herrmann, G. Nesbitt, M. 1993. Archaeobotanical et al. 1993, 56-8, pl. XVc-d. Nesbitt, M. and Summers, G. D. 1988. "Some Recent Discoveries of Millet (Panicum Miliaceum L. and Setaria Italica (1.) Beauv.) at Excavations in Turkey and Iran", AS 38, 85-97. Newman,J. 1932. The Agricultural life ofthe Jews in Babylonia Between the Years 200 C.E. and 500 C.E., London. of Plaque-type Crosses from Ain Okada, Y. 1990. Reconsideration

Sha'ia near Najaf, Al-Rafidan 11, 103-12. 1992. "Ain Sha'ia and the Early Gulf Churches: an Architectural Analogy", Al-Rafidan 13, 87-93. Percy,J. 1864. MetallurgyII ron: Steel.London. Pilipko, V. N. 1980. "Parfyanskiebronzovyemonetyco znakom P pod lukom",VDI,No. 4, 105-24: Pugachenkova, G. 1962. "Koroplastic of Ancient Merv", Tr.YuTAKEII, 117-73. 1967. IskusstvoTurkmenistana,Moscow. Pugachenkova, G. A., Rtveladze, E. V. and Kato, K., eds. 1991. Antiquitiesof SouthernUzbekistan,Tashkent Soka University. Rostoker, W. and Bronson, B. 1990. Pre-lIndustrial Iron, its Technolgy and Ethnology, Archaeomaterials Monograph No. 1, Philadelphia, Pennsylvania. Schmidt, H. 1908. "Archaeological Excavations in Anau and Old Merv, or: The Excavations in Ghiaur Kala (Old Merv) ", in Explorationsin Turkestan,Expeditionof 1904 (Pumpelly, R. ed.), I, 187-210, pls. 49-56. Sellwood, D. 1980. An Introductionto the Coinageof Parthia, 2nd ed. London. Shaked, S. 1977. 'Jewish and Christian Seals of the Sasanian Period", in Studiesin Memoryof Gaston Wiet(Rosen-Ayalon, M., ed.), 17-31, Jerusalem. Simpson, StJ., Hobbs, R. and Loginov, S. D. 1993. [From Rome to Charjou and China: Byzantine Coin Finds in the East.] [Ancient Amul] Charjou October 1-3, 1993. Smith, C. S. 1960. A IHistoryof Metallography,Chicago. Stein, M. A. 1907. Ancient Khotan, Detailed Reportof Archaeological Explorationsin ChineseTurkestan,Oxford. 1928. InnermostAsia, Detailed Reportof Explorationsin Central Asia, Kan-suand EasternIran, Oxford. Stronach, D. 1978. Pasargadae.A Reporton theExcavationsConducted bytheBritishInstituteof PersianStudiesfrom 1961 to 1963, Oxford. Tashkhodjaev, S. 1963. ["Section Through the City Wall in GyaurKala"], Tr. YuTAKE12, 95-118. Thompson, D. 1976. Stuccofrom Chal Tarkhan-Eshqabadnear Rayy, Warminster. Turner, P. 1989. Roman Coinsfrom India. ILondon. Usmanova, Z. 1. 1955. [ TheParthian Craftsman's House on the Site of Gyaur Kala], Diploma work, Historical Faculty, University of Tashkent. 1963. ["Erk-Kalaand Excavation of the Craftman's Workshop of the Parthian Period in the City-site of Gyaur Qala"], Tr. Yu7AKE12, 20-94. Varkhotova, D. P. 1958. [ArchaeologicalExcavations1956-1957 in the Quarterof Millerson the Site of Gyaur-Kala],Diploma work, student of the Historical Faculty, University of Tashkent. Whitcomb, D. S. 1985. Beforethe Rosesand Nightingales,Excavations at Qasr-iAbu Nasr, Old Shiraz,New York. Wilkinson, C. K. [ 1974], Nishapur:Potteryof theEarlyIslamic Period, New York. Williams, D. 1973. Flotation at Siraf, Antiquity 47, 288-92. Yassi, A. I. A. and Majeed, B.S. 1989/90. "GravityInvestigation in Ctesiphon Archaeological Site", Sumer46, 114-19. Zavyalov,V. 1993. [Bone Toilet Articles from Central Asian Sites of the Kushan and Post-Kushan Period]. Kratkie soobshchenia Instituta arkheologii(shortreportsof theInstituteofArchaeology),209, 31-41. -

'The Yu,7AKE team,Ashgabat. Dr. Kakamurad Kurbansakhatov, Director, Dr. Sergei Loginov, Director of Excavations in Erk Kala, Alaguli Berdiev, Director of Excavations in Sultan Kala. The London team Dr. Georgina Herrmann, Director, Institute of Archaeology, UCL, 3 September to 16 October. A. J. Barham, Deputy Director and field director, GeoInstitute of Archaeology, UCL, 24 August to archaeology, 17 September Professor J. M. Rogers, Director, Islamic Buildings, School of

THE INTERNATIONAL MERV PROJECT PRELIMINARY REPORT ON THE SECOND SEASON

Oriental and African Studies, University of London, 3 to 15 September Glynn Barratt, Assistant Director and field director, Topographic Team, English Heritage, 24 August to 17 September Dr. St. John Simpson, Assistant Director and field director, Excavations, Erk Kala and Gyaur Kala, Institute of Archaeology, UCL, 24 August to 16 October David Tucker, field director, Surface Artefact Survey, Institute of Archaeology, UCL, 24 August to 16 October Dr. Patrick Strange, field director, Resistivity, University of Nottingham, 24 August to 17 September: TopographicTeam Ian Peet, G. P. S., University of East London, 24 August to 3 September Christopher Barratt Phillips, English Heritage, 24 August to 17 September Edwin Moth, University of Sheffield, 3 September to 16 October Mark Herrmann, 17 September to 16 October Simeon Mellalieu, Institute of Archaeology, UCL, Geoarchaeology: 24 August to 17 September Resistivity:Robin Falkner, University of Nottingham, 24 August to 27 September Excavation: Richard Hobbs, British Museum, 24 August to 16 October Illustrators: Andrew Fisher, Birmingham, 24 August to 16 October Jane Read, Herts. Archaeological Unit, 24 August to 16 October: Dr. John Merkel, Institute of Archaeology, Archaeometallurgist: UCL, 17 to 25 September Archaeobotanist:Mark Nesbitt, Institute of Archaeology, UCL, 17 September to 16 October Islamic coins, Dr. Luke Treadwell, University of Oxford, 29 September to 16 October Liaison Officer,Kathy Judelson, Southampton, 24 August to 25 September. TheRussian Teamfrom St. Petersburgand Moscow Dr. A. Nikitin, numismatist and archaeologist, and Maxim Kotchoubeev, conservator, Hermitage Museum, St. Petersburg, 3 September to 16 October Dr. V. Zavyalovand N. Savvonidi, archaeologists, Institute for the History of Material Culture, St Petersburg, 24 August to 16 October

(1993)

75

Dr. Natasha Smirnova, archaeologist and numismatist, Pushkin Museum, Moscow, 24 August to 10 October Alexei Kononenko, architect, Moscow Architectural Institute, 3 September to 15 October Valeri Ossipov (St Petersburg), deceased 27 August. The principal contributors to the Research Design were StJohn Simpson, David Tucker and Glynn Barratt. 'The work was carried out under the Excavation Permit No. 16/377 of Dr. S. D. Loginov and was directed jointly by S. D. Loginov and St John Simpson. The team consisted of R. Hobbs, V. Zavyalovand N. Smirnova,joined for part of the season by A. B. Nikitin and M. Kotchoubeev. A. Fisher, M. Nesbitt, N. Savvonidi assisted towards the end of the season. V. Ossipov tragically died on the third day from a heart attack. 4Bailey (1992: 38, 44-5) described how in Tashkent earlier this century cotton seed oil was used to fill saucer lamps that were lit with the help of twisted cotton wool wicks. 5 Early Islamic pottery was noted here during a visit in 1993. "The team consisted of St John Simpson, V. Zavyalov, N. Savvonidi and up to six workmen. 7Finger-marked mudbricks have been recovered from a number of earlier YuTAKEexcavations at Merv, notable MGK Trench 2 (Usmanova 1963: 176), the south-western fortifications (Tashkodjaev 1963: 104, fig. 6) and the so-called "castle"on the southern circuit of the MEK walls (Filanovich 1959: Usmanova 1963: 50-1). SWe would like to acknowledge the valuable advice from Dr. I. Freestone, T. Lowe, Dr. B. Gilmour, Professor H.-G. Bachmann and others on the investigation of the remains. 'Archaeobotanical work benefitted greatly from assistance by KathyJudelson and Akmohammad Annaev in building the flotation machine and from Merv excavation staff in carrying bags of soil over long distances with little complaint. At Tahirbaj Tepe the director of excavations, Dr. Cattani Maurizio, gave active assistance in sampling. "'Deposit types have been defined as follows: hearths-clearly defined, small areas of in situ burning; burn areas/ashes-large spreads of ashes, mostly thought to be in situ; pits-clearly defined; refuse deposits-areas with higher than normal densities of bone; make-up-relatively sterile soil used to level up surfaces; floor-floor surfaces and associated underlying layers; mud-brick: individual mudbricks, and one sample of bonding material which proved to be similar in character.

THE FALL OF AL-MADA'IN: SOME LITERARY REFERENCES CONCERNING SASANIAN SPOILS OF WAR IN MEDIAEVAL ISLAMIC TREASURIES By Avinoam Shalem University of Edinburgh

"We urged (our) horses to Mada'in, Its waters like its dry land fruitful. We cleansed the coffers of this Kisra, They fled that day, like he, despairing." Abui Bujayd Nafi' b. al-Aswad.1 It is curious that the better known Sasanian objets d'art in mediaeval treasuries are those which were kept in the Christian churches of the Latin West. Apart from many Sasanian textiles, which are scattered in church treasuries throughout Europe, the most celebrated objects are probably the Cup of Khusraw-the so-called "tasse de Salomon" in the Paris (no. 76 in the invenNationale, Bibliotheque of no less fame, the so-called tory of 1634)-and, "vase d'Alienor" in the Louvre (MR 340).2 Both of them were formerly kept in the treasury of St. Denis.3 But unfortunately, it is difficult to unravel that their long history before they reached treasury. Tradition tells us that the Cup of Khusraw was among the presents which were given by Haruin alRashid to Charlemagne, and that Charles the Bald presented it later to the treasury. But such accounts were usually invented in order to bestow an aura of importance and mystery upon an object.4 More interesting is the account of Ibn Zafar (1104-70), who described in his Sulwan al-muta• that in the Byzantine treasury was a drinking crystal bowl decorated with gold, silver and glass (al-zujdj almuhkam) on which the portrait of the Sasanian king Shapur was carved.s This literary source suggests that the Cup of Khusraw could have reached the if so, this treasury of St. Denis via Constantinople; was most probably after the sack of 1204.6 The history of the "vase d'Alienor" was lately brilliantly discussed by Beech, who succeeded in tracing part of its "pre-Christian" history.7 Beech has suggested that the name of the anonymous donor Mitadolus, which is inscribed in Latin on the mounting of the lower part of the vase, refers to 'Ima-d al-Dawla 'Abd al-Malik b. Hiid, the last Muslim king of Saragossa (1110-30).8 However, the question how the vessel reached al-Andalus is still unclear. It is probable that most of the Sasanian objects

under discussion, excluding those which directly reached the Byzantine court with Sasanian delegations or in like manner,9 were initially kept in the treasuries of Muslim rulers. This paper focuses on those Islamic documents which inform us of Sasanian objects in Muslim possession, and is mainly centred on accounts referring to the fall of alMada'in in 637. Al- Mada'in, "the Cities", as the Arabs used to call the two capitals Seleucia and Ctesiphon on both banks of the Tigris, was the winter capital of the Sasanian kings.l -Its capture was not only an important military achievement during the conof but an event which left quest Iraq, concomitantly a great impression on the Arabs, who were confronted with the wealth of a highly artistic culture. Though a large part of precious Sasanian objects was already destroyed by the first century following this event, the legends and the colourful accounts of later Arabic authors probably inspired Muslim artists and their patrons, and both aspired to bestow upon their artefacts an aura of Sasanian workmanship. Arab writers inform us of various episodes during this capture, all of which illustrate the abundance of Sasanian precious and rare objects in contrast to the ignorance of the Arab soldiers. AlDinawarT (died 895), who relies on Mihnaf b. Sulaym, describes how the soldiers were not able to distinguish between silver or golden bowls." AlTabari (839-923) tells how camphor, which the soldiers mistook for salt, was put in dough for in his book almaking bread,12 and Ibn al-Tiqtaq5 Fakhri (written in 1302) added that one of the soldiers sold a big jacinth (probably a sapphire) for the modest sum of 1,000 dirhams, because he did not know that there was a number bigger than 1,000.13 Although, according to al-Tabari, the keepers of the Sasanian treasury carried as much as they could of the "most precious and portable commodities"

77

78

JOURNAL

OF PERSIAN

and departed with the last troops to join Yazdagird III (632-51) in Hulwan,14 still a large quantity of treasures was left behind. Vessels, precious stones, presents (altdf, delicate objects?) of inestimable value were left in the treasure chambers.15 Obviously big and heavy objects were also left; for example, the plaster statues which decorated the Great Hall of the White Palace remained in their original place, and even later, when Sa'd b. Abi Waqqas, the commander in this battle, decided to use the Great Hall as a prayer hall, the statues were left as they were.16 Sealed baskets, which were probably packed in order to be taken by the fleeing troops, were found in two Turkish tents. The baskets were opened, and gold and silver vessels were found within.17 A mule, which was lifted out of the water after it fell from a bridge during the flight of the Sasanians over the Nahrawain canal (on the eastern side of the Tigris), carried baggage with the king's finery, including his clothes, gems, swordbelt and coat of mail.s18

Two other mules were seized, while carrying four baskets in which the clothes and the crown of the king were found. "I began to unload them. On one of the mules there were two baskets containing the crown of the king, which could only be held aloft by two jewel-encrusted props. On the other mule there were (also) two baskets containing the king's garments, in which he used to dress up, brocaded with interwoven gold thread and adorned with gems, as well as other garments made of different fabrics similarly interwoven and adorned."19 The king's helmet, greaves and armplates, as well as the coats of mail of different kings like Heraclius, the last Lakhmid king, the king of India and the king of Makrfin (in southeastern Iran) were found in two leather bags. Eleven swords of different kings were found in two other bundles.20 The swords of the Sasanian king and the last and also the king's Lakhmid king al-Nu'm-n, finery, his crown and his garments were sent to the caliph 'Umar b. al-Khattaib "for the Muslims to see and for the nomadic tribesmen to hear about".21 The spoils were probably sent to be displayed in the Ka'ba and in other important shrines. This was probably the fate of the king's crown. The crown which was sent by Sa'd from al-Madai'in to the caliph 'Umar in Medina in order to be shown to the Muslims, was, according to Erdmann, the hanging crown of Khusraw II (591-628).22 If this crown was presented in public, it was most probably the one which years later was displayed in the Dome of the Rock. The earliest account (which is known to us so far), referring to a Sasanian crown in the Dome of the Rock, appears in the Fadd'il al-Bayt al-Muqaddas of al-Wlsiti, which was probably written around 1020. Al-Wisiti says that

STUDIES

"During the time of 'Abd al-Malik, there was hanging on a chain above the Rock under the dome the Yatimah pearl, the horns of Abraham's ram, and the crown of Kisra (Khusraw). When the Banii Ha-shim (sc. the Abbasids) took over the caliphate, they sent them to the Ka'ba."23 The fate of these objects has been recently discussed by Nasser Rabbat;24 among them, at least the crown of Khusraw was still seen in the early eleventh century by the anonymous writer of the Nihaya.25 Other sources mention that two large golden ornamental crescents with precious stones, which were part of the spoils of Made'in, a golden pail and cups studded with precious stones and of great value, which were most probably Sasanian vessels, were sent by 'Umar b. al-Khattaib to the treasury of the Ka'ba.2Z Some precious statues were found packed in baskets: "In one of these [baskets] there was a golden figure of a horse, saddled with a silver saddle; on its crupper and breast girth there were rubies and emeralds encased in silver. Its bridle was likewise embellished. There also was a figure of a horseman made of silver encrusted with gems. The other basket contained a figurine representing a silver she-camel; on it were a saddlecloth and strap of gold. It also had a halter or bridle made of gold, all of this studded with rubies. On the camel a man, made of gem-encrusted gold, was seated. The king used to attach these two figures to the props used for holding his crown aloft."27 The description which is given by al-Tabari is unclear, and thus it is difficult to detect how these statues supported the hanging crown. It is more that they were decorative probable pendants. However, the statue of the horseman was most probably a Chinese one. Al-Mas'iidi (died 956) describes a similar statue of a horseman, which was sent as a present from the Chinese Emperor to the Sasanian king Khusraw I (531-79).28 The statue of the horseman was studded with pearls (fdris min durr munaddad), and the eyes of the horse and rider were made of rubies.29 A similar type of precious statue is mentioned later among the treasures of the Fatimids; al-Qa-di al-Rashid (ca. 1050) describes a golden peacock, rooster and gazelle, all studded with pearls, precious stones and rubies (the rubies were inserted instead of eyes).30 Probably the most impressive booty was the qitf. This was a gigantic carpet measuring sixty by sixty cubits (ca. thirty metres long and broad).3 The qitf, which was also known by its Persian name as the BahZr-i Kisra ("The King's Spring" or "The Spring of Khusraw"),32 was a gold-coloured carpet. Its decoration consisted of "... pictures of roads and inlays like rivers; among them were pictures of houses. The edges looked like cultivated lands planted with spring vegetables, made of silk on

THE

FALL

OF AL-MADA

stalks of gold. Their blossoms were of gold and silver ... the fruits depicted on it were precious stones, its foliage silk and its waters golden."33 According to al-Tabari, the Sasanian kings used to sit on this carpet in winter time, during drinking parties, imagining they were sitting in gardens.34 Sa'd, who tried to avoid the barbaric action of dividing the carpets among the soldiers, succeeded in convincing them to send it to the caliph 'Umar in Medina. But disgusted by the human desire of the material world, 'Umar unforpossessing tunately cut the carpet into equal pieces and distributed them among the warriors of this battle.35 The memory of the carpet was kept alive for centuries by Arab historians, and it appeared also in material poetry as a symbol for the vanishing about a It is that thousand quite probable world.36 years later a group of seventeenth and eighteencentury Persian carpets, the so-called "Garden Carpets", reflect the memory of the qitf. Their decoration consists of trees with heavy blossoms, flowers, wild animals and water canals, all seen from a bird's eye view.37 Another object which was stripped of its precious stones and distributed by the caliph 'Umar among the Arab warriors, was the famous Sasanian war standard, the so-called "Banner of Kaveh". This banner, which has been discussed in detail by Christensen,38 fell into the hands of the Arabs after the battle of al-Qadisiyya, probably in 635 (shortly It was made of before the fall of al-Mad-'in). animal skin, studded with precious stones, gold and silver coins. In time of war, it was taken out of the treasury of the Sasanian kings and carried on a wooden rod by the general commander. A large quantity of gems was looted from the Sasanids and sent to the caliph 'Umar. It is related that 'Umar "burst into tears" when he saw the rubies, chrysalites, peridots and (other) gems.39 After the battle of Nihawand in the year 639, the royal Sasanian treasure, containing gems which the king used to divide among his deputies, was also sent to the caliph. Some of these precious stones were probably sent to decorate holy shrines. The tenth-century geographer al-Muqaddasi says that "both within the mihrdb [of the Great Mosque in Damascus] and around it are set cut agates and turquoises of the size of the finest stones used in these rings";40 according to Yiqfit (1179-1229), precious stones were put there by the Umayyad caliph 'Umar (II) b. 'Abd al-'Aziz.4 Arab historians inform us of the unique gems used by the Sasanian kings. According to alMas'iidi, Khusraw I had four signet rings: a carnelian one inscribed with the word "justice", a turquoise with the word "agriculture", a carbuncle (almandine?) with the word "patience" and a ruby with the word "faithfulness".4 The most celebrated

IN

79

Sasanian precious stone was the red ruby called alJabal "The Mountain", probably a dome-shaped ruby. This stone was mounted on a ring, which had been passed from one Sasanian king to another. The belief that the king who incised his name on it would be murdered was the reason why it was worn by them without having their names incised on it. Haruin al-Rashid, who bought it for 4,000 dindrs, had his first name, Ahmad, inscribed on it. The stone was mentioned for the last time during the but it is possible reign of al-Muqtadir (908-32),43 that the ring was in the possession of Abu 'I-Futi.h Yiisuf, a Fatimid ruler of Sicily.44 The flight of Yazdagird III from the invading Arab forces ended in Khurasan, where the king was murdered by a millstone cutter.45 Some of his treasures probably reached Khurasan or even his last residence in Marw. says that the golden palm tree of theAl-Q•.dTial-Rashid Sasanian kings was discovered by the governors of Khurfisan. The tree was decorated with precious stones and was reported as being valued at two million dindrs.46 Though not necessarily the Sasanian palm tree, a golden palm tree with fruits of different precious stones passed from the Fatimid treasury to the hands of the Nazir al-Juyiish.47 A Sasanian brazier (kanun) was sent from Khurasan to the treasury of the Ka'ba, but it was destroyed out of fear that it might have "turned" the Ka'ba into a fire temple.48 A little table, probably of Sasanian origin, passed from the treasury of the Umayyads to the Abbasids and later to the Fatimids.49 According to Ibn Jubayr, who visited Medina in 1184, two objects, allegedly regarded as Sasanian, were hung above the mihrdb of the Mosque of the Prophet. 5 The first one was a square stone of a bright and shining yellow surface, which measured one square span (shibr fi shibr). This stone was regarded as the mirror of Khusraw. The second item was a little receptacle (huqq saghir) which was considered to be the drinking-cup (ka's) of Khusraw. Since Ibn Jubayr did not describe the objects carefully, it is difficult to know whether these objects were indeed Sasanian. Indeed, Ibn the Jubayr himself had some doubts concerning of the objects.51 origin Though mentioned as a jam min zujadfir'awni "a drinking vessel (bowl) of Egyptian glass", this vessel-which was kept in the treasury of Marwin most probably Sasanian or at least II (744-50)-was in Sasanian style.52 al-Rashid describes it as one finger (isba') thick A1-Q.diand one-and-a-half spans (shibr) wide. It was engraved in relief with a typical Sasanian hunting scene: a crouching lion was depicted in the centre, and in front of him was a kneeling person who was aiming an arrow at the lion.53

80

JOURNAL

OF PERSIAN

CONCLUSION These literary sources suggest that, during the fall in 637 and until the death of of Ctesiphon Yazdagird III in 651, a large quantity of Sasanian objects fell into the hands of the Arabs. The objects were usually collected by the general commander and sent to the caliph, who was the one to decide what should be done with them. Some of the objects were sent by him to major Islamic shrines, where they were shown in public, while others, mainly those which could have been easily stripped of their precious stones and metals, were destroyed and scattered among the warriors, probably as salary. This practice was probably carried out in accordance with the concept introduced by the Qur'an in Stira 8:41 (Sirat al-Anfdl "The Spoils"). According to this sara, spoils of war essentially belong to God and His Messenger. They all should be deposited by the soldiers, and the ruler should distribute

STUDIES

them. One-fifth of the spoils is to be deposited in public treasuries for public welfare, while fourfifths are to be distributed among the soldiers who took part in the fighting.54 that some parts of the It is worth mentioning looted Sasanian objects were not Sasanian at all. These were either presents sent by different kings to the Sasanian court or spoils of war. The objects were displayed by the Muslims as Sasanian booty, and therefore it is quite probable that in the course of time they were "Sasanised" by the mediaeval Islamic popular mind. However, the numerous literary sources referring to the booty of the battle of Ctesiphon suggest that the impression that these precious objects left on the Arabs was strong. Thus some of them, like the crown and the cup of Khusraw, or the Spring Carpet, became legendary precious objects, and therefore were regarded as models of high artistic value for generations to come.

1 Cited by al-Tabari, Ta'rikh al-rusul wa 'l-muljtk,Eng. tr. The

SPONSA. DEDIT. ANOR REGI. LUDOVICO. MITADOL[us]. AVO MIHI REX. S[an]C[tis] Q[ue] SUGER[ius] (as a bride, History of al-Tabari (Bibliotheca Persica), ed. by Ehsan YarEleanor gave this vase to King Louis, Mitadolus to her grandShater (Albany, 1989), vol. XIII (tr. and annotated by Gautier H. A. Juynboll), p. 15 (p. 2434). father, the King to me, and Suger to the Saints"); the English translation is that of Panofsky, AbbotSuger, p. 79. 2 On the cup of Khusraw, see mainly C.J. Lamm, Mittelalterliche Glaser und Steinschnittarbeitenaus dem Nahen Osten (Berlin, 9 Al-Mas'iidi in his Meadows of Gold describes some precious Sasanian presents which were sent around 590 to the Byzan1929-30), vol. I, pp. 187-8, vol. II, pl. 64,3 (with extensive tine court. Among them was a table made of amber with literature); Blaise de Montesquiou-Fezensac and Danielle three golden legs in the form of paws of a lion, eagle and Gaborit-Chopin, Le Trisor de Saint-Denis (Paris, 1977), pp. 62-3, pl. 46; Le trisor de Saint-Denis,Mus6e du Louvre, exhibigoat; see Bis zu den GrenzenderErde,extracts from TheMeadows tion catalogue (Paris, 1991), pp. 80-2, cat. no. 10; on the vase of Gold,tr. and ed. Gernot Rotter (Tiibingen and Basel, 1978), of Eleanor, see mainly Lamm, Gliiser und Steinschnittarbeiten, p. 137. A "renaissance" in the relationship between the Sasanians and their Byzantine neighbours occurred during vol. I, p. 187, vol. II, pl. 64,1; Montesquiou-Fezensac and the reign of Khusraw II, mainly after he had married his Gaborit-Chopin, Saint-Denis, pp. 63-4, pls. 47-8; The Royal favourite wife Shirin, who was a Christian. The wedding took Abbeyof Saint-Denisin the Timeof AbbotSuger (1122-51), exhibition catalogue, The Metropolitan Museum of Art, New York place in 592 or 593. During the same years, some votive offer(New York, 1981), fig. 36; Le trisor de Saint-Denis, pp. 168-71, ings (a plate, cross and vessels, all made of gold) were sent by this emperor, probably to a church in Antioch. See M. J. cat. no. 27; George T. Beech, "The Eleanor of Aquitaine Vase, William IX of Aquitaine, and Muslim Spain," Gesta XXXII Higgins, "Chosroes II's votive offerings at Sergiopolis", (1993), pp. 3-11; idem, "The Eleanor of Aquitaine Vase: its ByzantinischeZeitschriftXLVIII (1955), pp. 89-102. 10 On the city of al-Mada'in, see G. Le Strange, The Lands of the Origin and History to the Early Twelfth Century," Ars OrienEastern Caliphate (repr. London, 1966), pp. 33-5; E12 art. "altalis XXII (1993), pp. 69-79. For the account by Abbot Suger of this vase, see E. Panofsky, AbbotSuger,on the AbbeyChurchof Mada'in"; Saleh Ahmad el Ali, "Al-Mada'in and Its SurroundSt.-Denisand its Art Treasures(Princeton, 1979), p. 79. ing Area in Arabic Literary Sources", Mesopotamia III-IV 3 The (1968-9), pp. 417-39. cup of Khusraw reached the Cabinet des Antiques of the " Al-Akhbaral- tiwdl, ed. V. Guirgass (Leiden, 1888), p. 134; see Bibliothbque Nationale in 1791. The Vase of Eleanor reached also al-Tabari, op. cit., vol. XIII, p. 24 (p. 2445); Ibn al-Tiqtaqa, the Louvre in 1793. 4 Unfortunately, the earliest source in which this object was al-Fakhri, tr. C. E.J. Whitting (London, 1947), p. 79 (Ibn almentioned is the Grandes Chroniquesde France of the fourTiqtaqa explained that the ignorant soldiers asked to teenth century: "hanap d'or pur et d'6meraudes fines et fins exchange gold against silver). 12 Al-Tabar-, loc. cit.; Ibn al-Tiqtaq-, loc. cit.; Ibn al-Athir, al-Kdmil grenez (grenats) si mervielleusement ouvr6 que en tos les roiaumes du monde ne fut ainques ovre si soutille". Cited in fi 'l-ta'rikh,ed. C. J. Tornberg (Leiden, 1851-76), vol. II, 401. 13 Loc. cit. Le trisor de Saint-Denis, p. 80. 14 Op. cit., vol. XIII, p. 20 (p. 2440). Cited by Lamm, Gldser und Steinschnittarbeiten,vol. I, p. 499

(no. 95).

According to Villehardouin (ca. 1150-1218), who wrote the history of the Fourth Crusade from first-hand knowledge, the booty of Constantinople was equally divided between the French and the Venetians. See his The Conquest of Constantinople, in Joinville and Villehadouin, Chronicles of the Crusades,tr. M. R. B. Shaw (London, 1963), p.94. 7 Beech, "The Eleanor of Aquitaine vase", pp. 3-10. 8 Ibid., pp. 5-7. The Latin inscription reads: HOC VAS 6

15 Ibid. 16 Ibid., p. 17

23 (p. 2443). Ibid., p. 24 (p. 2445); see also Ibn al-Athir, loc. cit. We may assume that Sasanian luxury vessels of precious metals reached the treasuries of the Umayyads and later passed to the Abbasids. The detailed descriptions by Abui Nuwas of the decoration of precious metal vessels which were used by the Abbasids during wine banquets suggest that Sasanian vessels, or at least early Islamic ones with typical Sasanian motifs,

THE

FALL

OF AL-MADA

were used in secular contexts. The English translations of some of the songs of Abil Nuwas are given by O. Grabar, "An introduction to the art of Sasanian silver", in Sasanian Silver, exhibition catalogue at the Museum of Art of the University of Michigan (Ann Arbor, 1967), pp. 34-6. is Al-Tabari, vol. XIII, p. 25 (p. 2445); Ibn al-Athir, loc. cit. 19 Al-Tabari, vol. XIII, p. 26 (p. 2446); Ibn al-Athir, 401-2; A similar account is related by al-Qidi al-Rashid: "Some Muslims seized two loaded mules ... on them were four baskets (asfilt); in two of them was the crown of Khusraw and in the other two the garmets which he used to wear". See G. H. Qaddumi, A Medieval Islamic Book of Gifts and Treasures: Translation,Annotation and Commentaryon the Kitdb al-Haddaya wa-al-Tuhaf(Ph.D thesis; Harvard University, 1990), 165. It is worth mentioning that two cloaks of Sasanian horsemen were excavated in 1968 at Antinoe. These are kept in the Musee Historique des Tissus in Lyons, Inv.968.III.1 (34.872) and 968.III.1bis (34.872bis); for illustrations, see Marielle Martiniani-Reber, Soieriessassanides,copteset byzantinesVe-XIe siecles, Lyon, Mus6e Historique des Tissus (Paris, 1986), cat. no. 22, 23. 20 Al-Tabari, vol. XIII, pp. 26-7 (p. 2447); Ibn al-Athir, vol. II, 402. 21 Al-Tabari, vol. XIII, p. 27 (p. 2447); Ibn al-Athir, loc. cit. 22 K. Erdmann, "Die Entwicklung der sas-nidischen Krone", Ars Islamica XV-XVI (1951), pp. 111-12 and 115. See also R. Ettinghausen, From Byzantium to Sasanian Iran and the Islamic World(Leiden, 1972, pp. 28-9; for a discussion on this type of hanging spoils see O. Grabar "The Umayyad Dome of the Rock in Jerusalem", Ars Orientalis III (1959), pp. 50-2. 23 For a discussion of this account of al-Waisiti, see N. Rabbat, "The Dome of the Rock Revisited: Some Remarks on alWasiti's Accounts", Muqarnas X (1993), pp. 71-3; for later accounts of this subject see ibid., p. 75, n. 30, and Le Strange, Palestine under the Moslems(London, 1890), p. 147 (Le Strange refers to a fourteenth-century source of the author of the Muthir al-ghirdm). 24 Ibid. According to al-Azraqi (died 834 or 837), the Horns of the Ram were destroyed around 690, namely during the capture of Mecca by 'Abd Allah b. al-Zubayr, see F. Wilstenfeld, Die Chronikender Stadt Mekka (Leipzig, 1858), vol. I, pp. 155-6; this account might have been invented to explain their disappearance from the treasury of the Ka'ba. Moreover, it might have concealed their probable transfer from the Ka'ba to the Dome of the Rock in 692 by 'Abd al-Malik, who recaptured Mecca at the same year. This speculation has been also suggested by Rabbat, see "The Dome of the Rock Revisited", p. 73. The long history of the pearl, called alYatima, is partially revealed by Rabbat, see ibid., pp. 71-2; it seems that there are some contradictory accounts of this unique pearl, which were sometimes mingled with accounts referring to another famous pearl, the so-called al-'Azima ("the Unpierced"?); see al-Qadi al-Rashid, Kitdb al-Hdddyawaal-Tuhaf, pp. 180, 348, n. 1. The author intends to discuss these accounts in a separate paper. 25 Cited by Erdmann, "Entwicklung", 115. 26 Cited by M. Aga-oglu, "Remarks on the Character of Islamic Art", The Art Bulletin XXXVI (1954), p. 182. 27 Al-Tabari, vol. XIII, p. 28 (p. 2448); Ibn al-Athir, vol. II, p. 402; al-Q' di al-Rashid, p. 168 (the statues were placed near the two pillars[?] supporting the crown). 28 Muriij al-dhahab, ed. Ch. Pellat (Beirut, 1965), vol. I, pp. 307-8. 29 Ibid.; for a German tr. of this part, see Rotter, op. cit., p. 130. 30 al-Rashid, p. 261. The golden peacock from the Al-Q.di treasury of the Fatimids is described again by al-Maqrizi

IN

81

(1364-1442), see P. Kahle "Die Schiitze der Fatimiden", ZDMG, N.S. XIV (1935), p. 358. 31 Al-Tabari, vol. XIII, p. 32 (p. 2452), n. 103. 32 Ibid., p. 33 (p. 2453); Ibn al-Athir, vol. II, p. 403; alRashid, p. 169. al-Qa.di n Al-Tabari, vol. XIII, pp. 32-3 (pp. 2452-53). 34 Ibid., p. 33 (p. 2453). 35 Ibid., p. 34 (p. 2454); Ibn al-Athir, vol. II, p. 404. 36 "Khusrau, der goldene Teppich, das gelbe Gold von ParwezSie sind verweht mit dem Winde, sind eins geworden mit Staub". Afdal al-Din Ibr-him al-Khaiqini (1126-99), tr. from Persian by Annemarie Schimmel, Stern und Blume (Wiesbaden, 1984), p. 119. 37 The better-known among them is in the Islamic Museum in Berlin-Dahlem (I. 41/69). A similar carpet is in the Arthur M. Sackler Museum, Harvard University (Gift of Joseph V. McMullan, 1957.137). The earliest example of this group (dated to the seventeenth century) is in the Burrell Collection in Glasgow. For a discussion of this group of carpets, see W. Bode and E. Kiihnel, VorderasiatischeKniipfteppicheaus alter Zeit (repr. Munich, 1985), pp. 133-7; Erdmann, Europaund der Orientteppich(Berlin and Mainz, 1962), p. 131 (the author stresses the association between the "Garden Carpets" and the qitf); G. Curatola, "Gardens and Garden Carpets: an Open Problem", EnvironmentalDesign III (1985), pp. 90-7. See also E. Blochet, "Note sur une tapisserie arabe du VIIIe siecle",JRAS (1923), pp. 613-17. 38 A. Christensen, L'Iran sous les Sassanides (Copenhagen, 1936), pp. 496-9. 9 Al-Tabari, vol. XIII, p. 46 (p. 2466). 40 Descriptionof Syria, IncludingPalestine, tr. G. Le Strange (Palestine Pilgrims' Text Society, London, 1892), p. 18. 41 Cited by Le Strange, Palestine under the Moslems, p. 264. 42 Al-Mas'iidi, vol. I, p. 309, German tr. p. 131. 43 This information is mainly taken from al-Mas'Uidi'sMeadows of Gold;for an Eng. tr. of this specific topic, see The Meadows of Gold; The Abbasids, tr. and ed. Paul Lunde and Caroline Stone (London and New York, 1989), p. 295. The stone is also mentioned by al-Qaidi al-Rashid, pp. 185-6. 44 Al-Q2di al-Rashid, p. 353, n. 3. 45 Al-Tabari, vol. XV, tr. and annotated by R. Stephen Humphreys (Albany, 1990), pp. 78-90 (pp. 2872-84). al-Rashid, p. 179. 46 pp. 162-3; see also the account of al-Maqrizi in Kahle, 47 Ibid., Al-Q-.di "Die Schaitze der Fatimiden", p. 359. al-Rashid, p. 189. 48 49 Ibid., p. 200. Al-Qd.di 0o Rihla, ed. William Wright and revised by M. J. de Goeje (Leyden, 1907), p. 194. For an English tr., see R. J. C. Broadhurst, The Travels of IbnJubayr (London, 1952), p. 202. 5s Ibid. 52 The term zujdjfir 'awni should not be necessarily explained as Egyptian glass; al-Ghuzuili (died 1412) in his chapter on diamonds (almas) explained this term as referring to a whiteyellowish transparent glass. See idem, Matdli' al-budiir fi manazil al-surir (Cairo, 1882-3), vol. II, p. 151 (cited by Lamm, Gliiserund Steinschnittarbeiten,vol. I, p. 510 (no. 141). 5 Al-QLdi al-Rashid, p. 46. The same description is repeated by al-Ghuzuili, op. cit., vol. II, p. 138; Lamm, who cited this account, stressed the Sasanian character of this vessel; see Gliiser und Steinschnittarbeiten, vol. I, p. 498 (no. 94). 54 For a discussion of this verse in Siira 8, see Sayyid Abu 'l-A'li Mawdiidi, TaJhim al-Qur'dn, tr. and ed. Zafar Ishaq Ansari, Towards Understanding the Qur'dn (Trowbridge, Wilts., 1990), vol. III, pp. 153-4.

THE CHINESE-UIGHUR ANIMAL CALENDAR IN PERSIAN HISTORIOGRAPHY OF THE MONGOL PERIOD By Charles Melville Universityof Cambridge

I

systemes chronologiques dans le monde turc ancien, rein-

forces their claim on our attention. His work ends The expansion of the Mongol empire through- by underlining the desirability of a systematic out Asia was accompanied by a breaking down of survey of the animal dates appearing in Islamic political, commercial and intellectual barriers documents, for the light this would throw on the from China to the Crimea. The prominence of history of the calendar in the Muslim context.5 Turks as the most influential group in the empire, Such an investigation also offers the more immeboth culturally and politically, has been noticed by diate possibility of examining the precise date of various scholars, and Professor Buell's account of certain events in the history of Mongol Iran. At the same time, because the inauguration of a the Turkicisation of international cuisine provides a case in point.' One of the least conspicuous but new calendar generally reflects political and somemost durable results of the Mongol conquest of times administrative changes, the introduction and Iran was the introduction of the "Turkish" twelve- use of the Chinese-Uighur calendar is a facet of animal calendar. Little attention has been paid to Mongol rule and symbolic of a new phase in this aspect of the Mongol legacy; indeed, a recent Persian history. The destruction of the Abbasid survey concludes that "There can be no doubt ... caliphate in 1258 and the establishment of a that the original Chinese-Uighur form of this Mongol dynasty under Hiilegii Khan marked the calendar was never used by Iranians, either in the temporary eclipse of Islamic rule and the advent of a new imperial power with its own imperial calenMongol period or later'. In fact, the ChineseUighur calendar was quite systematically if not dar and system of government.6 Knowledge of the extensively used in Iran for about a century and, as Chinese-Uighur calendar was not limited to the Abdollahy notes, the basic twelve-year animal cycle astronomers; it is also found in the works of court continued to be employed, with an important historians of the Mongol period, clearly reflecting modification, until it was abrogated in March 1925, current usage in certain quarters. The purpose of this paper is therefore to on the eve of the new Pahlavi era. Indeed, in the seventeenth century, the Safavid historian Iskan- examine the extent and accuracy of the use of the dar Munshi wrote that if he adopted the hijri year Chinese-Uighur calendar in Persian historical beginning in Muharram for his chronicle, "most of literature, and to note other examples in docuthe people of Iran would not understand". He ments issuing from the Mongol chancery. The therefore settled on the Turki (i.e. animal) year, manner and duration of its use allow some passing with which the general public were more familiar.3 comments on the Mongol presence in Iran. Lack of A knowledge of the calendars used in the docu- space prevents us pursuing this topic beyond the end of the fourteenth century, and the systematic ments of different periods is a basic requirement for the chronological reconstruction of their survey of dates is restricted to the Ilkhanid period. history, a task that is still far from complete in many points of detail. Twenty years ago, Louis II Bazin drew the attention of historians to the of the to both dates question given according It might be helpful first to clarify briefly what is Chinese-Uighur twelve-animal calendar and the Muslim hiqr calendar. Working on the basis of a meant by the Chinese-Uighur animal calendar in short list provided by Osman Turan,4 Bazin dis- this context. The Chinese civil calendar employs cussed a few examples of the parallel use of hijr- an abstract duodecennial cycle of twelve chih and animal dates ranging from 633 to 873 A.H., (branches) in conjunction with a decennial series including two events of particular relevance here, of ten kan (trunks) to give a sixty-year cycle, which namely the birth of Ghazan Khan, and the date of is used to classify years, months, days and hours. his accession to the throne. Bazin's analysis of these The year is luni-solar, i.e. containing twelve lunar dates raised several interesting points, and the months of 29 or 30 days, adjusted periodically by recent publication of his masterly study, Les the insertion of an additional lunar month to keep 83

84

JOURNAL

OF PERSIAN

the year in phase with the sun. The beginning of the year is taken to be the arrival of the sun at 150 Aquarius (mean date 27 January (Julian calendar) in the thirteenth-fourteenth centuries).7 The start of the Chinese lunar month is also calculated, and does not depend on the actual sighting of the new moon, unlike the Islamic lunar month, which thus normally begins one or two days later.8 The adoption of the Chinese calendar by the neighbouring steppe people was a measure of China's success in imposing its authority and the benefits of its civilisation upon the "barbarians". Professor Bazin has demonstrated with a wealth of detail how the eastern Turks adapted the Chinese civil calendar, the abstract, official replacing twelve-year cycle by its popular equivalent, namely, the Chinese astrological cycle of twelve animals, which was not used for dating by the Chinese themof the selves. The ten and sixty-year elements Chinese system were dropped but the "Turkish" animal cycle continued to correspond with the of civil the Chinese calendar, twelve-year cycle despite the disharmony between the astronomistart of the Chinese year and the cally-determined Turkish nomadic traditions of the year starting at the beginning of spring. As we shall see, the in Iran eventually strayed back to Turko-Mongols the spring equinox as the start of their year, partly no doubt as ties with China weakened and partly because they found a similar indigenous tradition in the Persian solar calendar. There is evidence of the use of the animal calendar among the Uighur Turks from the eighth century onwards, and particularly from their sedentary civilisation centred in Qocho in the Turfan depression from the end of the ninth century. On the eve of Mongol expansion, a Uighur almanac from Qocho, containing a calendar for the year with the 1202, shows a complete correspondence Chinese civil calendar, even down to the most complicated astrological details. The Mongols in turn this Turkish (animal) version of the adopted Chinese calendar from the Uighurs, who played an important role in the Mongols' administration and cultural formation, not least in providing them with the Uighur vertical script.9 The first precisely attested date in the Mongols' Secret History is the year of the Cock (A.D. 1201).10 The Chinese-Uighur referred to as the Uighur-Mongol) (sometimes calendar of the twelve animals was introduced throughout the Mongol empire and was observed as and the civil calendar the Mongol among Turko-Mongol ruling classes. At about the same time, from the establishment of the Yiian dynasty in northern China in 1215 (and the elimination of the Sung dynasty in 1279), the Mongols acquired their own official Chinese civil calendar, in conformity with their status as a Chinese imperial power.

STUDIES

The official Chinese calendar of the Mongol Yuiian dynasty and the Chinese-Uighur (UighurMongol) calendar of the twelve animals, therefore, share a common astronomical base. Despite the of the Uighur scholars' knowledge sophistication of the Chinese calendar, it was sufficient for general use to indicate dates by the name of the animal year, the number of the lunar month, and the day of the month. The lunar months are numbered by the Turkish ordinals, from first to twelfth.' It is in this skeletal and simplified form that the Chinese-Uighur calendar mainly appears in the narrative histories of the Ilkhanid period; the years are sometimes given their Turkish names, In practice, sometimes the Mongol equivalents. these dates conform to the official calendar of the Yuian dynasty of China.12 The scientific community in Iran, as represented particularly by Nasir al-Din to the new calendar system TusT, was introduced not only in its Turkish guise but also in its full astronomical complexity. The introduction of the twelve-animal calendar into Iran As well as marking a change of dynasty, the introduction of the Chinese calendar system into Iran can be located in the context of the impetus given to Islamic astronomy in the Turko-Mongol world of the thirteenth-fifteenth centuries. This impetus was driven by a deep interest in astrology on the part of royal patrons, who gave support to scientists in a field that was to some extent frowned on by orthodox Islamic opinion. The creation of vaqf funds for the maintenance of the (endowment) observatories at Maragha, Tabriz and Samarqand is evidence of the commitment shown by Mongol rulers, which permitted teams of astronomers to work over the long periods required for their observations to be completed.13 One of the most famous products of these observatories, namely the astronomical handbook (zij) of Nasir al-Din Tusl, contains a detailed account of the Chinese calendar, and work done at Maragha made contributions to astronomical calculations that transcend their applications to dating. This work may in turn have had an impact on Chinese science during the Yiian period.14 The story of the foundation of the Mardgha observatory, and of Nasir al-Din's instruction in Chinese astronomy by the Chinese scholar Filmanji (Fu Meng-chi) is too well known to need repeating here.15 Tuisi is said to have encouraged Hiilegfi's support by pointing out that by studying the stars he would be able to foretell the monarch's future, the length of his life, auspicious days for his journeys, and so on. Hiilegfi's belief in astrology and on the judgement of Nasir al-Din his dependence Tiisi are borne out by other sources.'6 Tuisi

THE

CHINESE-UIGHUR

ANIMAL

CALENDAR

what he had learnt of the Chinese incorporated calendar (tarikh-i qatd'iydn va turkdn), which he says is the one employed by the rulers of Iran (pddshdhan-i ma),17 into the first chapter of the Zij-i Ilkhanz. He concludes his description with a table covering the period of 100 years from the start of Chinggis Khan's reign in the year of the Pig, A.D. 1203 (599 A.H.), to the year of the Dragon, A.D. 1304 (703 A.H.). This was designed to aid conversion from the hijr1 to the Chinese calendar, because the hijrz calendar was the one best known to "our astronomers". 18 Nasir al-Din Tusi's Zij-i Ilkhani achieved considerable renown and, despite its revision and the work of other astronomers in the thirty years after his death (in 672/1274), probably remained the standard work on the subject. It is likely that any attempts to calculate conversions between the animal and hijrz calendars during the Ilkhanid period would have been based on Tuisi's tables or on others for which his was the ultimate source.19 The use of the animal calendar in Persian historiography Rashid al-Din was informed about the nature as well as the advent of the Chinese calendar, which he describes in his section on China in the Universal History.20 He also uses it himself, in its simplified form, in the parts of this work that deal with the history of Ching s Khan, his successors, and the It is with this latter portion, Mongols in Iran. the containing history of the Ilkhans down to the death of Ghazan in 703/1304, that we are chiefly concerned. There are two reasons for this. Firstly, events elsewhere in the whereas in chronicling Rashid al-Din never gives more Mongol empire, than the year according to the animal calendar (and sometimes the season),22 when dealing with the history of the Ilkhans he often gives the day, month (ay) and year (yfl) according to the Chinese-Uighur system, together with the hijrz equivalent.23 It is only from such precise information that we can see were used how accurately the two calendars together. Secondly, it is the use of the animal calendar to date events in Persian history that is at issue here, both as an aspect of the Mongol presence in Iran and because it is only in this field that I can claim an adequate knowledge of the sources to offer explanations and corrections for dates that appear to be in error. Dates are also given by their "Turkish" month and animal year in Abu 'l-Q~asim Kishini's History of Oljeitii (regn. 1304-16), the brother and successor of Ghazan Khan.24 So far as I am aware, this method of dating is restricted to these two court chronicles,25 which between them provide about eighty examples of pairs of dates in the hijri and animal

OF THE MONGOL

PERIOD

85

calendars. These dates are listed in Table 1, which gives rise to several points to be discussed in the second part of this paper. III With very few exceptions, the events recorded in the animal calendar all concern the activities of the Mongol ruling class: births, marriages, deaths and accessions to the throne are regularly dated this way, as well as visits of ambassadors from other Mongol states, and, more occasionally, the movements of the khan's ordu or military actions. The are the deaths in 707/1308, of two exceptions Muslim notables, namely Tij al-Din Mu'mini Qazvmin and Malik Fakhr al-Din IHasan, reported by Kashani (nos. 65, 66). The former was an agent of Shams al-Din Juvaini and the latter was governor of Rayy and Varamin. Both were therefore closely attached to the ruling elite; in addition, Fakhr alDin Hasan was a specialist in Mongol culture and chancery practice, and knew the Mongol (i.e. Uighur) script.26 From this, it is clear not only that the calendar was indeed actively used Chinese-Uighur by the Mongols in Iran but also that the historians probably relied on the oral evidence of sources at court for the information recorded in this way. Rashid al-Din mentions the Mongols' celebration of the New Year (Turkish: kuin yankildmishi,27 Persian: sar-i sdl) several times in passing, with or without hijri equivalents,28 and it was clearly a living tradition, not just an abstract feature of the calendar. As for the question of sources, it is well known that Rashid al-Din had the benefit of the knowledge of Piilad Aqa Chinksank (cheng-hsiang), representative of the Great Qa'an at the Ilkhanid court, and of Ghazan Khan himself for information on and as well as traditions, early Mongol history access to the imperial archives.29 For the period before Rashid al-Din's own time, such information be dated according would to the naturally own calendar and Mongols' presumably reported in the same form by the historian. This is illustrated by the fact that, with one exception, 30in all the pairs of dates between 631 and 669 A.H., the element comes first. There follows Chinese-Uighur a transitional period, between the accession of and the accession of Ahmad Abaqa (669A.H.) (681 A.H.), when the hijr date tends to come first, for but the animal date is given by preference certain events, such as the sack of Bukhara in of Abaqa's ordu, the 67111273, the movements death of a noyan and the quriltay that decided on the election of Ahmad Tegiuder (nos. 21, 22, 24, 27, 30).31 From the reign of Ahmad onwards, paired dates are all given with the Muslim date coming

86

JOURNAL

OF PERSIAN

TABLE

STUDIES

1

Dates reportedin the hijri and Chinese-Uighur calendars hijrI date No. year/month/date

A.D. equivalent

Chinese-Uighur date day/month year

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

beg. 2 Feb. 1234 beg. 2 Feb. 1253 beg. 22Jan. 1254 23 Oct. 1256 (M) 15 Jan. 1257 (M) beg. 19 Jan. 1257 16 Jan. 1258 22 Jan. 1258 beg. 18 Dec. 1259 beg. 13 Feb. 1262 20 Aug. 1262 beg. 21 Jan. 1265 17 Feb. 1265 9 Mar. 1265 17 June 1265 19 June 1265 19 July 1265 beg. 13 Oct. 1265 26 Nov. 1270 (W) 4 Nov. 1271 (W) 22 Jan. 1273 beg. 27 June 1275 15 Apr. 1277 (Th) beg. 4 June 1277 25 May 1278 Winter 1278/79 beg. 2June 1280 29 Oct. 1281 (W) 1 Apr. 1282 (W) 6 May 1282 18 Jan. 1284 10 Aug. 1284 (Th) 11 Aug. 1284 (F) 22 Mar. 1290 (W) 10 Mar. 1291 (S) 15 Mar. 1291 (Th) 23 July 1291 (M) 20 Feb. 1292 (W) 29June 1292 (Sn) 24 Sept. 1294 (F) 17 Mar. 1295 (Th) 24 Mar. 1295 (Th) 17 May 1295 (Tu) 3 Nov. 1295 (Th) 21 Jan. 1303 (M) 19 July 1304 (Sn) 9 Aug. 1304 (Sn) 19 Sept. 1304 (S) 9 Dec. 1304 (W) 2 June 1305 (W) 24 July 1305 beg. 13 July 1306 4 Sept. 1306 (Sn) beg. 13 July 1306 28 Nov. 1306 (M) 25 Dec. 1306 (Sn)

27 Feb. 1234 ends 20 Jan. 1254 autumn 1254 12 Nov. 1256 (Sn) 17 Jan. (1257) (W) Li ends 16 Jan. 1257 16 Jan. 1258 Mughai 22 Jan. 1258 Miughay beg. 13 Feb. 1260 Bijin 25 Feb. 1261 (F) 25 Aram [S] Daqiqui beg. 17 Aug. (1262) Sikisinj start of beg. 19 Jan. 1265 Gav 18 Feb. (1265) 1 Ikindi beg. 19Jan. 1265 Huikar 16 June 1265 2 Shuin Huikar 19 June 1265 (F) 5 Shun [F] Huikar 18 July 1265 4 Altinj Huikair G v beg. 19 Jan. 1265 spring of Muirin (...) beg. 23 Jan. 1270 4 Dec. 1271 (F) 1 Bir Yikimminj QiLyin beg. 21 Jan. 1273 Aram Daqiqu beg. 29 Jan. 1275 16 Apr. 1277 (F) 12 Utuinj Huikar .Tiinqiuz beg. 5 Feb 1277 Huikar Tavushqan beg. 13 Feb. 1279 Pars beg. 25Jan. 1278 Li beg. 2 Feb. 1280 30 Oct. 1281 (Th) 17 TUiqsuinj Muighay 31 Mar. (1282) (Tu) 21 Ikindi (...) 6 Apr. 1283 7 Ujunj Qimyin 19 Jan. 1284 New Year's Day Bijin 31 July 1285 (Tu) 28 Altinj Daqiqul 29 Altinj 1 Aug. 1285 (W) Daqiqu 14 Mar. 1290 (Tu) 2 Ikindi Bars 9 Mar. (1291) (F) 8 Ikindi (...) 14 Mar. (1291) (W) 13 Ikindi Taulf 22 July 1291 (Sn) 25_Altinj Lii 21 Feb. 1292 (Th) 2 Ikindi [Th] L 14 Altinj 29 June 1292 (Sn) beg. 21 Sept. (1294) (...) Tiiqsfini ends 16 Mar. (1295) end of Ikindi (...) 7 Ujinj 23 Mar. (1295) (W) (.. .) 17 May (1295) (Tu) 2 Shuin 1 Nov. 1295 (Tu) 23 Tiiqs-nj Qiiyin 19 Jan. (1303) (S) New Year's Day L 18 Altinj 21 July 1304 (Tu) 7 Yitinj 28July 1305 (W) Yilan 16 Dec. (1304) (W) 19 Bir Yinkizminj 10 Bir Yinkizminj Yilan 26 Nov. 1305 (F) 1 Ikindi 24 Feb. (1305) (W) Yind beg. 10 Aug. 1306 Yitinj Qis beg. 30 June 1307 Altinj 3 Sept. (1306) (S) 25 Yitinj Qims beg. 29 Aug. 1307 Sikisininj 26 Nov. (1306) (S) 20 Unuinch 14 Dec. 1307 (Th) 19 Bir Yinkuiminj Qiis

631 Jumada I 650 Dhu 'I-H. 651 Dhu 'I-H. 654 Shawwal 2 [Sn] 654 Dhu 'I-H. 27 [M] 655 656 Muharram 9 656 Muharram 15 658 660 Pisces 5 660 Shawwal 2 663 Rabi' II 663 Rabi' II 28 663Jumadai I 19 663 Ramadain 1 663 Ramadan 3 663 Shawwal 3 664 669 Rabi' II 10 [W] 670 Rabi' I 29 [F] 671 Rajab 1 674 675 Dhu 'lQ. 10 [F] 676 677 Muharram 1 677 Winter 679 Safar 680 Rajab 14 [Th] 680 Dhu 'I-H. 20 [W] 681 Muharram 26 682 Shawwal 27 683 Jumada I 26 [Th] 683 Jumada I 27 [F] 689 Rabi' I 9 [W] 690 Rabi' I 7 [S] 690 Rab' I 12 [Th] 690 Rajab 24 [Sn] 691 Safar 29 691 Rajab 12 [Sn] 693 Dhu 'l-Q. 2 [F] 694 Rabi' II 28 [Th] 694 Jumada I 6 [Th] 694 Rajab 1 [Tu] 694 Dhu '1-H. 23 [Sn] 702 Jumada II 1 [Sn] 703 Dhu 'I-H. 15 [M] 704 Muharram 6 [Sn] 704 Safar 17 [S] 704Jumfida I 10 [W] 704 Dhu 'l-Q. 8 [W] 705 Muharram 1 706 Muharram 706 Safar 24 [M] 706 706Jumada I 21 [Tu] 706 Jumaida II 18 [Sn]

28 Aram end of autumn of 24 Utfinj New Year's Day ( ...) 11 Jaqshaba-t 17 Jaqshaba-t

Yiind Hfikar Bars Li

A.D. equivalent

Source RD,95 RD, 24 RD, 24 RD, 13 RD, 36 RD, 44 RD, 54 RD, 55 RD,70 RD, 195n RD,87 RD, 93 RD,94 RD, 100 RD,94 RD,95, 101 RD, 103-4 B,427 RD, 139 RD, 248 RD, 141-2 RD, 151 RD, 144 RD, 146 RD,152 RD, 151 RD, 153 RD, 162 RD,95,164 RD, 169 RD, 177 RD, 194 RD, 198-9 RD,220 RD, 195n,226 RD,227 RD, 230, 233 RD, 7 RD,236 RD,241 RD,243 RD,244 RD, 288 RD, 302 RD, 352 K, 24 K,31 K,31 K,41-2 K,44 K,45 K,52 K, 52 K,53 K, 53 K,53

THE

CHINESE-UIGHUR

ANIMAL

hijri date No. year/month/date

A.D. equivalent

57. 706 Rajab 7 [F] 58. 706 Rajab 29 [Tu] 59. 706 Shawwail 5 [Sn] 60. 706 Dhu 'I-Q. 13 [Tu] 61. 706 Dhu 'I-H. 24 [Tu] 62. 706 Dhu'l-H. 17 [Tu] 63. 707 Muharram 1 64. 707 Rabi' I 8 [Th] 65. 707 Rajab 25 [Tu] 66. 707 Sha'ban 20 [W] 67. 708 Muharram 1 68. 709 Muharram 1 69. 710 Muharram 1 70. 711 Muharram 71. 712 Muharram 72. 712 Dhu 'I-H. 10 [Sn] 73. 713 Muharram 74. 714 Muharram 75. 715 Muharram 76. 716 Muharram 77. 716 Ramadaln 27 [W]

12 Jan. 1307 (Th) 3 Feb. 1307 (F) 9 Apr. 1307 (Sn) 16 May 1307 (Tu) 26June 1307 (M) 19June 1307 (M) 3 July 1307 7 Sept. 1307 (Th) 20 Jan. 1308 (S) 14 Feb. 1308 (W) 21 June 1308 11 June 1309 31 May 1310 beg. 20 May 1311 beg. 9 May 1312 8 Apr. 1313 (Sn) beg. 28 Apr. 1313 beg. 17 Apr. 1314 beg. 7 Apr, 1315 beg. 26 Mar. 1316 13 Dec. 1316 (M)

CALENDAR

OF THE MONGOL

Chinese-Uighur date day/month year 9 Chaqshabat 1 Aram 6 Uchiinch 14 Tfirtiinch 6 Bishinj 20 Bishinij 10 Sikisinj 26 Chaqsa-bat 22 Aram

Chaqsab-.t Notes: a. Hijr- and animal dates are given exactly as found in the texts, with their A.D. equivalents. For corrections, see Table 3. b. The names of months and years of the ChineseUighur dates are transliterated directly from the original Persian text, which allows misreadings and the range of spellings to be readily apparent. Standard spellings are provided in Tables 2 and 3. c. (...) indicates a lacuna in the text.

first, with only one exception (no. 38), concerning an event within the immediate orbit of the ruling class.32 There are also one or two instances when Kashani mentions events outside Ilkhanid territory, using the animal date only and with no precise hijr- equivalent.33 It is easy enough to see why the Chinese-Uighur animal date was reported in those cases where it is found; it is rather harder to explain why it was not given more often. There are numerous occasions when one might expect a date to be given in the animal calendar, but in vain.34 It is doubtless partly a question of the sources of information available to Rashid al-Din and Kashani; but they must also have exercised their own conscious decision to omit animal dates, for we may presume that, if the animal calendar was in use, it was used systematically and not at random. Animal dates must have been known or recoverable for most of the Ilkhans' dynastic history. To pursue the question of sources would lead us too far away from our subject; it seems clear, at least, that the normal use of the Muslim hijri calendar continued unaffected by the

87

PERIOD

A.D. equivalent

13 Jan. (1307) (F) 3 Feb. (1307) 8 Apr. (1307) (S) 16 May 1307 Qfiyi 6 June (1307) (Tu) 20June (1307) (Tu) Tavushqan beg. 5 Feb. 1315 7 Sept. (1307) 20 Jan. (1308) 14 Feb. (1308) beg. 11 Feb. 1309 Daqiqui It beg. 31 Jan. 1310 Tfinghuiz beg. 20 Jan. 1311 Qiiluqana beg. 8 Feb. 1312 Ut beg. 27Jan. 1313 Ut beg. 27Jan. 1313 Bars beg. 17 Jan. 1314 Tavushqan beg. 5 Feb. 1315 Li beg. 25Jan. 1316 beg. 14Jan. 1317 YThin beg. 3 Jan. 1318 Yfi-n

Source K, 54 K, 54 K,54 K,61 K, 62 K,66 K,72 K,73 K,74 K,75 K,82 K,87 K, 109 K,121 K, 136 K, 144 K,151 K, 165 K, 173 K, 199 K,222

d. Only days of the week that are mentioned in the Persian text are given [in square brackets]; when this follows the animal rather than the hijr- date, this indicates that the animal date comes first. The corresponding weekday in the Christian calendar is also shown (in round brackets). Sn = Sunday, S = Saturday, etc. e. Years in brackets indicate the correct year, when no animal year is mentioned. In fact, an incorrect animal year is generally implied by the sequence in Kashani. f. Sources:RD = Rashid al-Din, ed. 'Alizada; B = BanaikatT; K = Kashan-i.

introduction of the Mongols' own calendar and that the two existed side-by-side. This raises the further question of whether our historians converted dates from one calendar to the other, or merely repeated the information they received. The relative paucity of recorded animal dates suggests the latter. Once Rashid al-Din came to cover his own times, he generally put the hijr date first when paired with an animal one, and this was also the practice of Kashani, as we have seen. Had they been used to offering conversions as a matter of course, rather than on the few occasions when the Chinese-Uighur date was available, we would expect them to have done so far more systematically. When Kashani does come to provide a systematic equivalent, it is for the year only, as will be discussed below. On the other hand, for many of the events in the earlier period, the Mongol date must have been the only (Chinese-Uighur) date available, making it necessary to calculate the hijrI equivalent. Bazin, in his discussion of pairs of animal-hijri dates issuing from the Turko-Mongol milieu, comes

88

JOURNAL

OF PERSIAN

to the conclusion that, in cases of doubt, when the two dates do not correspond, it is generally best to rely on the date given according to the twelveanimal calendar rather than the hijri equivalent, reconstructed a posteriori. His view is based on evidence of faulty calculations, or "the imperfection of the tables consulted".3 Unfortunately, I am not at present in a position to test the latter opinion. As noted above, the set of tables most likely to have been consulted at this period is the one given in Nassir al-Din Tuisi's Zij-i Ilkhanz. This is certainly the only contemporary conversion table that has been identified. It has not been edited, let alone translated and subjected to critical examination.36 It is designed to convert hijri dates into the ChineseUighur calendar, not vice-versa. It would be desirable to use Tusi's table to check all the pairs of dates listed in Table 1. This might reveal a systematic source of error, but until it is tried, it seems more charitable to assume that Tuisi's tables are accurate.37 It is very likely that other almanacs concalendar taining the imperial Chinese(-Uighur) were available among the administrative and literate classes, but one would not expect these to have been inaccurate either.38 On the whole, faulty calculation rather than faulty tables seems the likelier explanation for incorrect conversions, but various other sources of error can be imagined. In fact, Table 1 reveals a generally close and often exact correspondence between the hijri and animal dates provided, whether as a result of calculation or not. We can pass over those that agree without further comment. Those that do not agree need to be investigated, partly because we are faced with a choice of which date to accept as the correct one, and partly because it is interesting to try to explain the discrepancy. In doing so, we cannot assume that the given dates have a sort of absolute sanctity, for otherwise elaborate theories could be of advanced to interpret the non-correspondence or two dates in terms of calendar irregularities tables, when all the time imperfect conversion either or both of the pair are simply mistakes arising from other causes. Factual errors are usually difficult to verify, because both authors mention many events that are not found or not precisely dated in other sources. Many mismatching pairs can be attributed to scribal errors or Other, more regularlymanuscripts. imperfect are probably due to the occurring disparities nature of the two calendar systems themselves, and of them on the perhaps also to a misunderstanding part of the authors or their sources. It is not possible in the space available here to discuss all these scribal, factual, or systematic errors, more than one of which might be compounded in any individual case. Table 3 provides a summary of all the dates given in Table 1,

STUDIES

corrected where necessary so that both the hijri and animal elements can be reconciled with each other. Some are examined examples representative which to below, explain the rest. Stanmay help dardised spellings of the names of the months and years have been adopted, as shown in Table 2. Hiilegii's departure for Iran. Rashid al-Din's dates for the start of Hiilegfi's expedition to Iran (nos. 2, 3) are irreconcilable. Having left M6ngke's court, Hiilegii returned to his own ordu at the beginning of Hiikar-yil (Year of the Ox), not the end. From this error stems the further mistake of putting Hiilegii's the following animal year (Tiger), departure whereas in fact it should have been the autumn of the same year. Rashid al-Din's date in Dhu 'l-Hijja coincides with late winter. A far more precise account is given by Juvaini, who unfortunately does not use the animal calendar.39 The birth, accession and death of Arghun. There is no compatibility between the two dates given for the birth of Arghun (no. 10). The animal date, which comes first, is most probably correct, although 25 February 1261 was a Friday, not a Saturday. The Islamic date, with an astrological motif, is imprecise and is probably a later reconstruction.40 Neither date conforms well with the tradition that date Arghun died aged 33, but the Uighur-Mongol is more nearly correct. However, the "33 years" can be best explained by suggesting that the year of Arghun's birth should actually be the Monkey (Bichin), see below; this would be equivalent to Pisces in 658 A.H. TABLE

2

The Turkish monthsand twelve-yearanimal cycle* Months (Turkish) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Aram Ikindi Ochiinch Tortuinch Besinch Altinch Yetinch Sekizinch Toksunch Onunch Bir Yegirminch Chaqshapat Shun**

Years (Mongol)

Sichgan Ud Bars/Pars Tavishgan Lu Yilan Yunt Qoy(un) Bichin/Pichin Takagu It Tonguz

Qulugana Hiikar Bars Taulai Lu Mogai Morin Qonin Bichin Takiya Noqai Gaqai

Animal (Persian) Rat Ox (Gav) Tiger Hare Dragon Snake Horse Sheep Monkey Cock (D-qiqiqi) Dog Pig

The spelling of these terms differs not only in the Persian sources (see Table 1) but also in all the secondary works dealing with this topic. See e.g. the tables in Abdollahy (p. 667) and E. Chavannes, "Le Cycle turc des douze animaux", T'oungPao, VII (1906), p. 52. The spelling adopted here is based on that of Bazin, Systemes,simplified slightly. The Persian equivalents are sometimes found, as indicated in brackets. ** Shun is the intercalary month, see Doerfer, III (1967), pp. 327-8. *

THE

CHINESE-UIGHUR

ANIMAL

CALENDAR

TABLE

OF THE MONGOL

PERIOD

89

3

Reconciledhijri-animal datesfrom Table 1 No.

Event

Muslim A.H. date

Turko-Mongol date

A.D. equivalent

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

Birth of Abaqa Khan Hiilegfi camps Hiilegfi leaves for Iran Birth of M6ngke Temfir New Year Caliph's embassy to Hiilegfi Mongols to Baghdad Hiilegui to E. Baghdad Various deaths Birth of Arghun Hiilegfi to Alatagh Hiilegfi ill Death of Arikan Khatun Abaqa moves camp Death of Doghuz Khatun Accession of Abaqa Expedition against Golden Horde Abaqa moves ordu Re-accession of Abaqa Birth of Ghazan Sack of Bukhara Winter quarters Battle of Albistan Abaqa to Alatagh Abaqa leaves for Khurasan Nikiidari raids Death of Abatai Noyan Battle of Hims Death of Abaqa Election of Ahmad New Year's Day Death of Ahmad Accession of Arghun Arghun marries Bulughan Khatun Death of Arghun Despatch of messengers Election of Gaikhatu Death of Toqtani Khatun Coronation of Gaikhatu Death of Anbarji Despatch of amirs to Baidu Death of Gaikhatu News of Ghazan's advance Coronation of Ghazan New Year's Coronation Day.. of Oljeitii Oljeitii moves camp Arrival of Mongol embassies Arrival of Holden Horde embassy Oljeitui marries Bulughan Khatun Oljeitfi moves camp Beginning of hijri year Oljeitui moves camp Death of Du'a Oljeitui at winter quarters Arrival of envoy Arrival of envoys Arrival of envoy

631 Jumada I 650 Dhu 'I-H. 651 Shawwal 654 Shawwal 22 654 Dhu 'I-H. 29 655 656 Muharram 9 656 Muharram 15 658 658 Pisces 660 Shawwal 2 663 Rabi' II 663 Rabi' II 29 663 Jumada I 19 663 Ramadan 1 663 Ramadan 3 663 Shawwal 3 664 669 Rabi' II 10 670 Rabi' II 29 671 Rajab 1 674 675 Dhu 'l-Q. 10 676 678 Muharram 1 677 winter 679 Safar 680 Rajab 14 680 Dhu 'I-H. 20 681 Muharram 6 682 Shawwal 28 683 Jumada I 26 683 Jumada I 27 689 Rabi' I 9 690 Rabi' I 6 690 Rabi' I 11 690 Rajab 23 691 Safar 29 691 Rajab 12 693 Dhu 'l-Q. 2 694 Rabi' II 28 694 Jumada I 6 694 Rajab 1 694 Dhu 'I-H. 26 702 Jumada I 29 703 Dhu 'I-H. 15 704 Muharram 6 704 Safar 17 704 Jumada I 10 704 Dhu 'l-Q. 29 705 Muharram 1 706 706 Safar 24 Muh.arram 706 706 Jumadfi I 21 706 Jumadfi II 18 706 Rajab 8 706 Rajab 29

28 Aram

1234 Feb. 27 1253 Feb. 1253 autumn 1256 Nov. 12 1257 Jan. 17 1257 1258 Jan. 16 1258 Jan. 22 1260 1260 Mar. 8 1262 Aug. 20 1265 Jan. 1265 Feb. 18 1265 Mar. 9 1265 Jun. 16/17 1265 Jun. 19 1265Jul. 18/19 1266 spring 1270 Nov. 26 1271 Dec. 4 1273 Jan. 22 1275/76 winter 1277 Apr. 16 1277 summer 1279 May 14 1278/79 winter 1280 June 1281 Oct. 30 1282 Mar. 31 1282 Apr. 16 1284Jan. 19 1284 Aug. 10 1284 Aug. 11 1290 Mar. 22 1291 Mar. 9 1291 Mar. 14 1291 Jul. 22 1292 Feb. 21 1292 Jun. 29 1294 Sep. 24 1295 Mar. 16/17 1295 Mar. 24 1295 May 17 1295 Nov. 6 1303 Jan. 19 1304 Jul. 19 1304 Aug. 9 1304 Sep. 19 1304 Dec. 819 1305 Jun. 23 1305 Jul. 24 1306 July 1306 Sep. 3/4 1306 Sep. 1306 Nov. 28 1306 Dec. 25 1307 Jan. 13 1307 Feb. 3

Yunt beginning Hiikiir autumn Hiikiir Lu 24 Onunch (1 Aram Mogai) Mogai 11 Chaqshapat Mogai 17 Chaqshapat Mogai Bichin Bichin 25 Aram Sekizinch (It) start of Hfikiir 1 Ikindi (Hfikar) Hfikiir 2 Shun Hiikar 5 Shun Hfikiir 4 Altinch Hiikiir Bars spring Morin (Onunch) 1 Bir Yegirminch Qoy Aram Takagu Tonguz 12 Uchfinch Hiikiir Hfikiir Tavishgan Bars Lu 17 Toksunch Mogai 21 Ikindi (Yunt) 7 iUchiinch Yunt Bichin (1 Aram) Bichin 28 Altinch Bichin 29 Altinch 10 Ikindi Bars 8 Ikindi (Taulai) 13 Ikindi (Taulai) Taulai 25 Altinch Lu 2 Ikindi 14 Altinch Taulai Toksunch (Morin) end Ikindi (Qoy) 8 UJchiinch (Qoy) 2 Shun (Qoy) 28 Toksunch Qoyun (1 Aram Taulai) 16 Altinch Lu 8 Yetinch Lu 19 Sekizinch (Lu) 11 Bir Yegirminch Lu 1 Altinch (Yilan) Yetinch Yilan Altinch Yunt Yunt 25 Yetinch Sekizinch Yunt 22 Onunch (Yunt) 19 Bir Yegirminch Yunt 9 Chaqshapat (Yunt) 1 Aram (Qoy)

90

JOURNAL

OF PERSIAN

No.

Event

Muslim H. date

59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77.

Orduqiya to ordu Oljeitfi on Gilan expedition Oljeitui enters Gilan Raid in Gilan Start of hijr- year Oljeitii goes hunting Death of Taij al-Din Mu'mini Death of Malik Fakhr al-Din Start of hijri year Start of hijr- year Start of hijri year Start of hijri year Start of hijri year Feast of Sacrifice Start of hijri year Start of hijri year Start of hijri year Start of year Death ofhi.ri Oljeitii

706 706 706 706 707 707 707 707 708 709 710 711 712 712 713 714 715 716 716

Shawwal 5 Dhu 'I-Q. 13 Dhu 'I-H. 4 Dhu 'I-H. 18 Muharram 1 Rabi' I 8 Rajab 25 Sha'ban 20 Muharram 1 Muharram 1 Muharram 1 Muharram Muharram Dhu '1-H. 10 Muharram Muharram Muharram Muharram Ramadan 30

STUDIES

Turko-Mongol date 6 Uichiinch 14 Tortunch 6 Besinch 20 Besinch 10 Sekizinch 26 Chaqshapat 22 Aram

Chaqshapat

A.D. equivalent (Qoy) Qoy (Qoy) (Qoy) Qoy (Qoy) (Qoy) (Bichin) Bichin Takagu It Tonguz Qulugana Ud Ud Bars Tavishgan Lu Lu

1307 Apr. 8/9 1307 May 16 1307 Jun. 6 1307 Jun. 20 1307 Jul. 13 1307 Sep. 7 1308 Jan. 20 1308 Feb. 14 1308 Jun. 21 1309 Jun. 11 1310 May 31 1311 May 20 1312 May 9 1313 Apr. 8 1313 Apr. 28 1314 Apr. 17 1315 Apr. 7 1316 Mar. 26 1316 Dec. 16

Notes: a. HijrI dates are taken to refer to the standard civil calendar, beginning 16 July 622.

b. Animal years missing from Table 1 are supplied in brackets.

Arghun's accession to the throne (no. 33) is dated first by the Islamic date. The Uighur-Mongol date is in perfect agreement if the year is changed, from the Year of the Cock to the Year of the Monkey (1284). An error of one year is not uncommon in these combinations, as we shall see.41 In this case, therefore, the hijri date is preferred.42 The hijri date of Arghun's death (no. 35) is given but it first; the animal year is not mentioned, should be the Hare. The two dates are one day apart, as is often the case. Since the Islamic Saturday begins at sunset on Friday, and the ChineseUighur Friday ends at midnight, it is implied that Arghun died between sunset and midnight on Friday 9 March 1291. Both dates could therefore be correct, but there are grounds for proposing the Muslim date should be 6 Rabr' I, which would denote that Arghun died during the day.43 Arghun is said to have reigned seven years and to have died aged 33.44 Seven years is an acceptable (690-683 = 7, 1291-1284= 7), but approximation of reckonthe to way Chinese-Uighur according ing, which counts inclusively, his reign would be eight years. This has a bearing on the figure of thirty-three years for his life, which would only be to our reckoning (1291thirty-one according 1260 = 31), or thirty-two lunar years (690-658 = 32). dates provided, According to the Chinese-Uighur of the Cock and came born in the Year was Arghun to the throne in the Year of the Cock (though in reality perhaps both should be the Year of the Monkey). He was thus "25" (two full cycles of

twelve, plus one, inclusive). His reign was then "8" years (Year of the Monkey to Year of the Hare, inclusive). The sum of twenty-five and eight gives our "33", whereby the year of his accession is counted twice. This solution relies on the year of Arghun's birth being 658 A.H./the Monkey; it may be considered preferable to retain the Mongol date given for his birth, and ignore the reference to his death aged 33, an error which would then require an even more tortuous explanation.45 The decision to enthrone Ahmad Tegiider. The dates given for the decision to elect Ahmad (no. 30) are confused. Apart from the year, there is an obvious discrepancy in the day of the month, for the 26th of the "Muslim" moon cannot correspond with the 7th moon. The year is of the (same) "Chinese-Uighur" easily disposed of, and we should substitute the Year of the Horse (1282) for the Year of the Sheep (1283); this may be a scribal error or a systematic error (Rashid al-Din is also a year ahead in dating Arghun's coronation, as seen above). Either way, the Year of the Horse is correctly given by Banaikati.46 With the day of the month, we have a choice. 26 Muharram is equivalent to 27 Uchiinch(6 May 1282), and we could therefore propose a .y simple scribal error, restoring *bist-u-haftum. Alter7 Uchiinch-ay is equivalent to 6 natively, Muharram (16 April), and it would then be necessary to remove the bist from the hijri date. Banakati, unfortunately, merely increases the uncertainty, by giving 17 Uchfinch-ay (26 April), equivalent to 28 Muharram (8 May). This confirms the element of 7

THE

CHINESE-UIGHUR

ANIMAL

CALENDAR

in the Mongol date, but also the 20 in the Muslim one. In this case, since the animal date comes first and the quriltay is a "Mongol" event par excellence, the animal date seems preferable. Although a careless scribe is perhaps more likely to have omitted a bist from the Mongol date than to have added it to the hijri one (as in no. 50, see below), another example of the same mistake is found elsewhere (no. 61, see below). Arghun's marriage to Bulughan Khatun. The mismatch between the dates given here (no. 34) can be explained by a simple scribal error. The Muslim date, which comes first, is accurate, as is confirmed by the correct day of the week. 22 March corresponds to 10 Ikindi-ay, and we may therefore propose *dahum for duvvum.47 Thefirst enthronementof Gaikhatu Khan. The hijri date for the election of Gaikhatu (no. 37) comes first, but the Uighur-Mongol date is probably more reliable, for it corresponds with the correct day of the week (Sunday), which was considered the most auspicious day for coronation.48 Although other sources follow Rashid al-Din in giving 24 Rajab, the correct date is probably 23 Rajab, Sunday 22 July.49 ThecoronationofOljeitii and his departurefor Tabriz. These dates (nos. 46, 47) need to be considered in combination to understand the difficulties that are frequently posed by Kashani's careless chronology. In the first place, the coronation on 15 Dhu 'l-Hijja is said to have been a Monday, whereas 19 July was actually a Sunday. This might suggest 16 Dhu 'lHiija/Monday 20 July as the "correct" date, to be explained by a possible delay in the local sighting of the new moon of Dhu 'l-Hijja.50Such a solution could be applied to other dates in the month, though here Kashani is erratic.51 It is also consistent with Kashani's date for the departure of Oljeitii for Tabriz: if 16 Dhu 'l-Hijja 703 was a Monday then 6 Muharram 704 was indeed a Sunday.s2 The Mongol date for Oljeitii's departure is the wrong year: it should still be Lu (Dragon), but Kashani mechanically introduces a new animal year with the change of hijri year. The corrected date, Saturday 8 August, corresponds to the beginning of the Muslim Sunday, i.e. after sunset on the evening of Saturday. However, since it is unlikely that Oljeiti set off in the evening, a more accurate animal date would be 8 Yetinch-ay (Sunday 9 August). We might therefore propose *hashtumfor haftum Yetinch-ly, either as a scribal or a calculation error."3 The Mongol date for Oljeitii's coronation does not in itself resolve whether 15 or 16 Dhu 'l-Hijja is correct, because it is either one or two days out. 18 Altinch-ay is possibly a straight-forward mistake on part, but hijdahum is a plausible scribal K.shini's error for hifdahum, and 17 Altinch-ay corresponds to 20 July or 16 Dhu 'l-IHijja.54Monday 16 Dhu '1-

OF THE

MONGOL

PERIOD

91

Hijja/17 Altinch for the coronation is also consistent with Sunday 6 Muharram/8 Yetinch for Oljeitui's departure for Tabriz. This "solution", which attempts to reconcile both pairs of dates, requires the correction of elements of three out of the four, and might seem over-elaborate. Furthermore, there is considerable support for the date 15 Dhu 'l-Hijja for the coronation in other contemporary sources.55 One would expect the Muslim date to be the most reliable, and despite the arguments outlined above, it seems preferable to take 15 Dhu 'l-Hijja (19 July) as the correct date, altering Kashani's "Monday" to a Sunday. Sunday was a good day for coronations, and the 15th day of the month was auspicious too.56 This still requires the modification of both the Uighur-Mongol dates, and leads us to conclude that, right from the outset of Oljeitii's reign, there are great problems with the accuracy and internal consistency of Kashani's dating. Oljeitii's marriage to Bulughan Khatun Khurdsanz. Kaishani's dates for this ceremony are on the face of it irreconcilable (no. 50). However, as we have seen, he is very careless in the matter of chronology, and in this case again several corrections are needed. It is clear from the context that Oljeitii's wedding followed his arrival in Tabriz on 20 Dhu 'l-Qa'da, and a bist seems to have dropped out. 28 Dhu 'l-Qa'da is equivalent to Tuesday 22 June 1305. The Mongol year is not stated, but it should be the Year of the Snake. Rather than the second month, June falls in the sixth month, and we may therefore suggest 1 *Altinch-ay, which is equivalent to Wednesday 23 June 1305. The correspondence is still not exact, and we may further refine the hijri date, partly because, for an event like this, one might expect the (corrected) Mongol date to be accurate, partly because Kashani's internal inconsistencies are so marked. He notes that Oljeitii's wedding was on a Wednesday and that 26 Dhu 'lQa'da was a Sunday. The date of the wedding should therefore be Wednesday 29 Dhu 'l-Qa'da/23 June.57 Oljeitii'soperationsin Gilain.Similar problems arise with Kashani's dating of the events on the G-la-n campaign (nos. 61, 62).58 In both cases, the Uighur-Mongol date is preferable (the year, which is omitted, is the Year of the Sheep), reinforced by the fact that they are consistent with each other, which the hijri dates are not. Oljeitii entered Gilin on 6 June/4 Dhu 'l-Hijija, so that a bist has been added. This was a Tuesday. Kashani then reports a raid led by amir Sevinch on Tuesday 17 Dhu 'lHijja, but that Tuesday would have been the 18th, which is confirmed not only by the animal date but also by other dates given on the same page.9 These are some of the more striking examples of the need to take the Chinese-Uighur date into

92

JOURNAL

OF PERSIAN

and often to retain it in preference consideration, to the Muslim hijr- date.60 They reveal a mixture of scribal and factual errors, some of which can be resolved by reference to other accounts. They also show that the lack of direct correspondence dates can be due between hijr- and Chinese-Uighur to more systematic differences. In the first place, since the Chinese-Uighur day, whereas the like our own, starts at midnight, Muslim "day" starts at sunset the previous day, there is often an ostensible difference of one day between the two calendars that does not require correction (e.g. no. 15, 17, 29, 35, 36, 41, 42, 53, 59). These dates are usually left as pairs on Table 3. The implication is that events dated this way occurred between sunset and midnight at the end of the Mongol day and the beginning of the Muslim day, as with the death of Abaqa (no. 29). However, this is often demonstrably not the case, and the disparity might represent a mechanical application of conversion tables that were not sufficiently discriminating; in all these cases, the Mongol date should probably be advanced (or the hijr- date reduced) by one day. It is noteworthy that sometimes within the same year, exact diurnal corare found as well as the "nocturnal" respondences overlaps, so that there is no discernible standard at work. This is particumethod of computation for 706 A.H., the last true of annal larly Kashani's calenyear for which he used the Chinese-Uighur dar regularly. The difficulties which he was having might have encouraged him to abandon it. As already noted, the problem of the disparity of one day between the dates given also raises the possibility that the chronicles do not always relate to the standard hijri era, starting on 16 July 622. In some of the cases discussed above, the differences could be resolved with reference to the "astronomical" calendar, starting the previous day, 15 July 622. When the weekday is given, this can help to distinguish which calendar is being used. On the other hand, the uncertainty remains that some discrepancies are simply the product of errors or of delays in sighting the new moon; furthermore, admitting the possibility that the chroniclers were inconsistent in their use of the two hyijr calendars has unwelcome implications, even for those dates that appear to be accurate according to the civil calendar. It is easier, if less rigorous, to assume (as here) a certain consistency in the use of the civil calendar throughout, while remembering that this may in fact be yet another possible variable in some instances. Another type of purely calendrical problem, referred to above, is that the start of the Muslim lunar month might on occasions have been delayed by the late sighting of the new moon, which would lead to a difference of a day (or even two) between

STUDIES

the local date and the standard hijrs calendar. The days of the week, however, would naturally carry on regardless. This sort of problem could explain the discrepancies found in nos. 38 and 57, and also in the dates given for the battles of Albistan (no. 23) and Hims (no. 28), although in the last two cases, the hyijr dates are confirmed by numerous Arabic sources.61 As in the case of days and months, there is normally a period of overlap between the two calendar years, since the year starts at a different point in each. In his history of the Ilkhans,62 Rashid al-Din generally gives the correct animal year correspondence, but in four cases (nos. 25, 30, 32, 33) the animal year is anticipated. These might be simple mistakes, but the same problem recurs in a far more systematic and conspicuous way in the chronicle of Kashani, who appears to mention the correct animal year only by chance. Kashani's twelve-animal year coincides with the Islamic year in which it begins, but the reverse is not true. This is most evident in the last few annals of Kashani's chronicle, from 708/Takagu to 715/Lu, when the start of the Islamic year (mentioned first) is matched to the animal year beginning later in the same hijri year, and not to the animal year in which it actually began. This systematic error helps to explain why he misdates Oljeitii's death, for he mechanically associates 716 A.H. with Yilan-yil, the next in the animal cycle.63 Apart from blatant misin the Uighurtakes, this error of anticipation is the of error found in these only type Mongol year from to of dates, 677/Tavishgan pairs 716/Yilan. It suggests a mechanical one-to-one correlation between the hijri and the Mongol year, whereas, the one being lunar and the other solar, they cannot be kept in phase so easily. This raises the possibility of the existence of simplified tables, the use of which lead to greater and greater would gradually inaccuracy. While there is no other evidence for the existence of such simplified tables, it sould be noted that these difficulties are not peculiar to Kashani. It is beyond the scope of this paper to examine in detail the continuing use of the twelve-animal cycle historians, but a by the late fourteenth-century review shows that the same element of preliminary seemingly mechanical correlation does sometimes occur, when the hijri year is mentioned first and the Mongol year is apparently anticipated (and the error never occurs the other way round). However, between hijri and animal the correspondences years given by Nizam al-Din Shami and 'AIT Yazdi are generally accurate, and in most cases where more precise (hijr) dates are given, they do fall in the animal year mentioned. Since the twelve-animal cycle was the only feature of the Chinese-Uighur calendar used by

THE

CHINESE-UIGHUR

ANIMAL

CALENDAR

later Persian annalists, it is worth looking a little more closely at another aspect of this question. As noted above, the Chinese-Uighur year adopted by the Mongols began when the sun entered 150 six weeks before Aquarius, that is approximately the spring equinox. By the Safavid period, the animal year was taken to start at the spring equinox (Naurfiz: then falling on 10/11 March of the Julian calendar). It is not clear precisely when this change took place, and the official Chinese astronomical of the new year gave way to the calculation and native Persian tradition of the Turko-Mongol in year starting spring. and Bazin considers this a late development, rightly notes that, in astronomical circles at least, the Chinese system was preserved, as in Ulugh Beg's zij dating from about 841/1437.64 However, it is evident from the dating in Yazdi's Zafar-nama that the animal year is taken to start at the vernal equinox, at least by the end of the fourteenth century, and this seems also to be the case in the earlier Zafar-nama by Sham.65 The Timurid historian HaIfiz-i Abrui implies that the animal year started in spring, though he also refers to the celebration of the Mongol New Year.66 His dating in both the cases cited is inaccurate, which suggests, if nothing else, that some uncertainty lingered round the complex relationship between the two systems. The disuse of the imperial calendar in the western Mongol lands might have coincided with the expulsion of the Yuian dynasty from China in 1368, though this cannot be demonstrated. We may also conjecture that the introduction of the Khani era on 12 Rajab 701/13 March 1302 was the first stage in the process whereby the start of the animal year became popularly associated with Nauriiz during the fourteenth century. Following the conversion of Ghazan Khan in 1295, the new era marked a change in the outward style of the dynasty and at the same time attempted to solve fiscal difficulties created by the disparity between the hifri lunar year and the agricultural year of solar seasons.67 The Khani year was solar, starting at Naurfiz, and adopted the names of the Turkish "as was appropriate for a calendar months, instituted by Mongol kings".68 It is not used by Rashid al-Din, who is supposed to have introduced it, nor by Kishani, though it was current in Abui Sa'id's reign, as is attested by his coinage.69 References to the Khani year in the mid-fifteenth century show it used in parallel with the animal year, demonstrating that certainl' by this time they both started together at Nauruiz. We may draw some more general conclusions from these examples of the historians' use of the twelve-animal calendar. First, the period for which we have dates in the full Chinese-Uighur form to (day, month, year) extends from 631/1234

OF THE

MONGOL

PERIOD

93

716/1316. Rashid al-Din uses the Chinese-Uighur calendar sparingly, in order to date events in the early history of the Ilkhanid dynasty, from the birth of Abaqa to the accession of Ghazan Khan. In this period, the Mongol date is often authoritative, and probably reflects the form in which his information reached him. A conversion from Mongol to Islamic dates was needed for the early period, and these conversions, perhaps on the basis of Tiisi's tables, are generally accurate. Rashid al-Din stops using the animal calendar for the reign of Ghazan. This surely reflects the fact to Islam and a that, after Ghazan's conversion certain distancing of the Ilkhanate from the rest of the Mongol empire,71 there was less emphasis on the Chinese (-Uighur) civil calendar of the parent Yuian dynasty in China. Kashani's continuing use of the animal calendar for Oljeitii's reign, however, shows that the Islamisation of the Hiilegiiid ruling family remained superficial. For a brief period of about three years (704-7 A.H.) he uses it quite frequently, compared with Rashid al-Din, for events both in Iran and elsewhere in the Mongol world. This reflects the situation at the beginning of Oljeitii's reign, when there was something of a renewal of pan-Mongol solidarity. Envoys arrived from the Great Qa'an, the Chaghatay Khanate and the Golden Horde, one consequence of which was Oljeitu's Gilhn campaign of 706/1307. The lack of further Chinese-Uighur dates soon afterwards is suggestive. It is also well known that there was a revulsion against Islam and a desire to revert to traditional Mongol practices in the early years of the reign, resolved by Oljeitii's adoption of Shi'ism in 709/1310. is frequently Although Kashani's chronology faulty, his animal dates are sometimes more reliable than their hijr- equivalents. Kashani's inaccuracies might be due to the unrevised nature of his work, but they might also indicate a general decline in the use of the official calendar, and a greater ignorance of its characteristics, especially after the start of the new Khani era. Also, it may be just a coincidence, but Tuisi's conversion table ends with 703 A.H./Lu-yil (the Year of the Dragon, 1304), precisely the point from which Kashani's chronology develops serious inconsistences. Taken on its own, the use of the Chinese-Uighur calendar by the historians, and its rapid abandonment after the fall of the Ilkhanate, might be taken to indicate a steady erosion of its political importance. On the other hand, the Mongol New Year to be celebrated continued evidently by the Furthermore, the picture of Mongols themselves.72 the decline of the Chinese-Uighur calendar is not borne out by the evidence of material issuing from the Ilkhanid chancery. Various documents have survived, including

94

OF PERSIAN

JOURNAL

STUDIES

TABLE4

Dates from Mongol chancerydocuments Document Hiilegfi to Louis IX From M6ngke Temiir Abaqa to Clement IV Safe-conduct from Abaqa Arghun to Philip the Fair Arghun to Nicholas IV Ghazan to Boniface VIII Oljeitii to Philip the Fair Decree of Abui Sa'Id Decree of Abli Sa'id Decree of Yisun Temfir Golden Horde to Simon Ivanovich Golden Horde Decree of Shaikh Uvais

Animal Year

Other details (day/month)

Dog 1262 Hare 1267 Dragon 1268 Hare 1267 1279 Ox 1289 Tiger 1290 Tiger 1302 Snake 1305 Monkey 1320 Ox 1325

10 April, 10th year of reign (1262)

Hare 1339 Pig 1347 Horse 1354 Dog 1358

Equivalent (A.D.)

4th qaghuchid,1st autumn month 7th month 27th day 23rd day, Altinch-aiy 6th month 16th day, 1st winter month 10th month 6th qaghuchid,1st summer month 4th month 25th day 5th day, 1st summer month 4th month 14th day, 3rd spring month, 701 A.H. 3rd month 8th qaghuchid,1st summer month, 704 A.H. 4th month 23rd day 8th qaghuchid,2nd autumn month, 720 A.H. 8th month 23rd day 9th qaghuchid,1st autumn month 7th month 22nd day 21 Ramadan 725 24th day, 2nd spring month 2nd month 5th qaghuchid,8th month 26th day 2nd qaghuchid,Aram-afy 1st month 29th day 15th day, 2nd autumn month 8th month 13 Dhu 'l-Qa'da 759

Ref.

= 10 Apr. 1262

M

= 18 Aug. 1267

MC2

=3 Aug. 1268

T

= 3 Nov. 1267 =21 Nov. 1279

MCI

= 15 May 1289

MC2

= 14 May 1290

MCi

= 12 Apr. 1302

MCI

= 16 May 1305

MC2

= 25 Sept. 1320

C

= 30 Aug. 1325 =31 Aug. 1325

HD1

=3 Apr. 1339

LC

= 4 Oct. 1347

MC2

= 22 Feb. 1354

MC2

= 18 Sept. 1358 = 17 Oct. 1358

HD2

Sources: M = Meyvaert;

C = Cleaves;

Cleaves,

"Trois

Cleaves,

"Lettres";

documents";

MC1 = Mostaert

MC2 = Mostaert

LC = Clark; T = Tisserant;

Herrmann

and

and Doerfer, "Seyh Oveys"; see note 73.

HD1 =

letters addressed to the rulers of Europe by different Ilkhans; see Table 4.73 These documents are in the present context, for several interesting, reasons, which we should note briefly. In the first animal year is used conplace, the Chinese-Uighur sistently and without interruption, from the time of Hiilegfi to the time of Abui Sa'Td, in combination with the hijri year from the reign of Ghazan. Secondly, the chancery documents differ from the in the way they express the lunar chronicles and sometimes also the days. Whereas the months, chroniclers invariably employ the form of months numbered one to twelve, in the chancery documents we often find expressions such as "the first summer month", "the second spring month", "the

and Doerfer,

HD2 = Herrmann

and

"725/1325";

second autumn month", and so on.74 There is also evidence of a complicated system of retrogressive numbering for the last ten days of the month.75 While the dating method employed by the chancalendar, it cery conforms to the Chinese-Uighur would appear to preserve characteristics that are more peculiarly Turkish than Chinese.76 It is not clear to me why the administration employed a dating system different from the one apparently in more popular use: possibly it reflects of the scribal classes. But the ethnic composition evidently the same "four seasons" system continued in use in the chanceries of the western Mongol states, and particularly by the Golden Horde, for at least twenty years after the collapse of

THE CHINESE-UIGHUR

ANIMAL CALENDAR

the Ilkhanate,77 which might indicate a conservative attachment to the imperial administrative practices of the parent Yuian dynasty in China, at least until 1368. Thus, despite the broader political trends hinted at by the declining use of the Chinese-Uighur animal calendar in the chronicles, there is evidence of greater continuity in its use in the administration. Juvaini's complaint that a knowledge of the Uighur script was a passport to advancement was probably as true at the end of the Ilkhanate as at the beginning. This reinforces other indications that the traditional Persian bureaucracy was perhaps not quite so vital to the functioning of Mongol government in Iran as they would like us to suppose.78

1

Igor de Rachewiltz, "Turks in China under the Mongols: a Preliminary Investigation of Turco-Mongol Relations in the 13th and 14th Centuries", in China Among Equals, ed. M. Rossabi (Berkeley, Los Angeles and London, 1983), pp. 281-310; P. D. Buell, "Pleasing the Palate of the Qan: Changing Foodways of the Imperial Mongols", Mongolian Studies, XIII (1990), pp. 57-81. 2 R. Abdollahy, "Calendars, ii. Islamic Period", in EIr, IV (1990), p.671; idem ('Abdallahi), Tdrikh-i tarikh dar Irdn (Tehran, 1366/1987), pp. 326-7. See R. D. McChesney, "A Note on Iskandar Beg's Chronology",JNES, XXXIX/i (1980), pp. 53-63 [esp. 56-7]. A table of the hijri dates corresponding to the start of the Turki new year from 1495 to 1925 is given by 'Ali Akbar Khan Muhammadi, "Sal-shumar-i tatbiqi-yi salha-yi turki va hijri qamarf", Tdrikhi mu'd4ir-i Irdn, I (Tehran, 1368/1989), pp. 95-119. I am grateful to Jean Calmard for this reference; the table, however, is not entirely reliable. 4 0. Turan, Oniki hayvanlzTiirk takvimi (Istanbul, 1941). Thanks to Metin Kunt for going through this with me. 5 Bibliotheca Orientalis Hungarica, XXXIV (Budapest and Paris, 1991), a revised version of Bazin's doctoral thesis (Lille, 1972, publ. 1974); see esp. pp. 405-10, 557. Certainly it is only in the area of Islamic sources that Bazin's work is less than exhaustive. 6 According to Saif-i Munajjim, Zij-i Ashrafi, Bibliotheque Nationale, Paris Ms. Pers. Suppl. 1488, fol. 3a, a new "era of Hillegfi" started on 1,592,087 + 322 days after the Flood, = 11 January 1258 (Julian), i.e. during the siege of Baghdad. This era does not seem to be attested in the chronicles, and Huilegfi's reign is generally taken to start in 652/1254 (cf. n.15). Rashid al-Din, Jdmi' al-tavarikh, ed. E. Quatremhre (Paris, 1836), pp. 60-3, is also aware of the start of a new era, tdrikh,which he dates back to Chinggis Khan. For fuller details of the calendar see, P. Hoang, A Notice of the Chinese Calendarand a Concordancewith the European Calendar (Zi-Ka-Wei, 1885);J. Needham, Scienceand Civilisation in China, III (Cambridge, 1959), esp. pp. 396-406. 8 It should be borne in mind that there is a difference between the civil or popular hijri calendar, beginning on Friday 16 July 622, and the astronomical reckoning, based on the true conjunction of the new moon the previous day, 15 July 622. Throughout this paper, the hijri date is taken to refer to the civil calendar, but the discrepancy between the two could explain some of the inconsistencies discussed below. Conversions are taken from G.S.P. Freeman-Grenville, The Muslim and Christian Calendars(London, 1963).

OF THE MONGOL PERIOD

95

ACKNOWLEDGEMENTS I would like to thank Professor Louis Bazin for responding very generously and in great detail to my numerous enquiries, and Professor Edward S. Kennedy for his unstinting assistance and encouragement. Both have also made helpful comments on the final draft, and are largely responsible for anything remotely numerate about this paper (except the errors). An early version of this paper was presented at a conference at the School of Oriental and African Studies, University of London, on "The Mongol Empire and its Legacy" in March 1991; I am grateful to David Morgan for allowing me to withdraw it from the projected conference proceedings for which it was originally prepared.

9 De Rachewiltz, "Turks in China", pp. 283-7; see also the paper by Thomas T. Allsen, "The Yuian Dynasty and the Uighurs of Turfan in the 13th Century", esp. pp. 266-7, in the same volume. The vital role of the Khitans has also been stressed by P. D. Buell, "Sino-Khitan Administration in Mongol Bukhara",JAH, XIII/ii (1979), esp. p. 124. o0 Bazin, Systemes,pp. 390-403. 1" Ibid., pp. 294, 354-5, 556. The first and the twelfth months derive from Buddhist/Iranian and Soghdian respectively. See Table 2. 12 All conversions in this paper are taken from P. Hoang, Concordance des chronologies niominiques chinoise et europeenne (Chang-hai, 1910), partial dynasty Yiuan, pp. 473-81, principal dynasty Yuian, pp. 267-78. 13 A. Saylli, The Observatoryin Islam (Ankara, 1960), esp. pp. 224-5, 233-6. 14 E. S. Kennedy, "The Chinese-Uighur Calendar as Described in the Islamic Sources", Isis, LV (1964), pp. 435-43. On the debated question of the influence of Islamic astronomy on China, see Needham, pp. 49-50, 372-82; Sayili, p. 207; and J. A. Boyle, "The Longer Introduction to the 'Zij-i-Ilkhani' of Nasir-ad-Din Tuisi",Journal of Semitic Studies, VIII (1963), pp. 244-54 (p. 254). 15 Saylll, pp. 189-91. The main Persian source is Rashid al-Din, ed. Quatremere, pp. 324-7; ed. A. A. 'Alizada (Baku, 1957), pp. 66-7; cf. tr. K. Jahn, Die Chinageschichtedes Rasid al-Din (Vienna, 1971), pp. 21-3. See also Rashid al-Din, Tanksuiqnamaya tibb-iahl-i khata, ed. M. Minovi (Tehran, 1350/1972), p. 16. Rashid al-Din's account is followed by Banakati, Tarikh, ed.J. Shi'ar (Tehran, 1348/1969), p. 338. 'Abdall-hi, Tdrikh,p. 324, seems to regard Banikati as the superior source. Banfikati, p. 420, puts the founding of the Maraighaobservatory in the 57th year of the era of Chinggis Khan (starting in 599 A.H.). Rashid al-Din says work was completed in the seventh year of Hiilegfi's reign, which was taken to begin in 652/1254, i.e. in 659/1261. Vassaf, Tajziyat al-amsairva tazjiyat al-a'sar (lith. Bombay, 126911852), pp. 51-2, puts the foundation of the observatory in 657/1259, and says it remained unfinished on Hiulegfi's death in 663/1265. 16 Vassfif, loc. cit., echoes Tiisi's own views on the value of observing the stars, as put forward in the Zij-illkhani, B.L. (India Office) Ms. Or. 7464, fols. 2b-3a. Although giving a detailed account of the work of the observatory and the contents of Tiisi's zij, does not specifically refer to the Vasss•f study of the Chinese calendar, nor does he use it in his chronicle. For Hiilegii's reliance on astrologers, see e.g. Sayili, pp. 192, 202.

96

JOURNAL

OF PERSIAN

STUDIES

one day early, perhaps to be explained by a delay in the sightZij-i lkhdni, Or. 7464, fol. 17a (misplaced in the Ms.); cf. Paris Ms. Ancien Fonds 163, fol. 5a, Cambridge Ms. Browne 0.2 (7), ing of the new moon of Safar (or of Rabi'I). fol. 3b; 'Abdallahi, p. 323. 3 Kshani, p. 38, for the arrival of an Ilkhanid embassy in Peking, and the death of Temfir Qa'an; also p. 54 for the 18 Or. 7464, fol. 13a, followed by the table on fols. 13b-16a. death of the Chaghatayid ruler Kunjak son of Du'a. 'Abdallahi seems at least partly to underrate the use of the Chinese-Uighur calendar on the grounds that it was little 34 Occasionally there are blanks in the Mss. of Rashid al-Din, known when it first appeared in Iran. indicating the intention to supply the Mongol date, e.g. ed. 19 According to the Timurid prince Babur, Ti-si's zij remained 'Alizada, pp. 93 (death of Huilegii), 169 (enthronement of in general use until it was replaced by Ulugh Beg's Zij-i Ahmad). Other such lacunae are indicated by (...) on Table 1. Gurgani, see Saylli, p. 264 and ibid., pp. 211-18, on later work at Mar-gha. A comprehensive list of zijes is given by E. S. 35 Bazin, Systemes,pp. 409-10, with particular reference to dates for Ghazan's birth and accession (nos. 20, 44). It should be Kennedy, "A Survey of Islamic Astronomical Tables", Trans. Amer. Philosoph. Soc., XLVI (1956), pp. 123-77. Of these, the noted, however, that the correct month of Ghazan's birth (Rabl' II) is found in one Ms.: Tdrikh-iGhdzani, ed. K. Jahn Zij-i khdqaniof Jamshid al-Kashi (1413) has tables for conversion from hijrzto Chinese dates for the period 801-901 A.H., (London, 1940), p. 3. This highlights the more likely source of India Office Lib. Persian Ms. Eth6 2232, fols. 17a-18a. error, namely scribal inaccuracy. The correction to the date of Ghazan's accession (see Table 3) is based on the view that 20 Tr. Jahn, Chinageschichte,pp. 22-3, followed in essence by it took place on a Sunday, as mentioned in the sources. Banakati, pp. 25, 339-40. Rashid al-Din's dates for the events of Dhu 'l-Hijja 694 lack 21 These sections of the Jdmi' al-tavarikh are conveniently internal consistency, and there is disagreement too among brought together in the "edition" of B. Karimi (Tehran, other sources. Banakati, p. 456, puts Ghazan's coronation on 1338/1959), I, pp. 165-679, from the editions by I. N. Berezin and E. Blochet, and vol. II (from Quatremere and K. Jahn). Sunday 28 Dhu 'l-Hijja; Mustaufi, p. 602, has the end of the month. 22 'Abdallahi, p. 326, refers only to this limited use of the animal calendar, citing the translation by J. A. Boyle, The Suc- 36 A paper in preparation by Mustafa K. Saiyid, E. S. Kennedy cessorsof GenghisKhan (New York and London, 1971). and Benno van Dalen will largely rectify this. tables for conversions between the Chinese-Uighur 23 For the terms dy and yil, see G. Doerfer, Die tiirkischenElemente 37 T•si's im Neupersischen,II (Wiesbaden, 1965), pp. 169-70, IV (1975), and Yazdagird calendars are entirely consistent, which 243-51, and Bazin, Systemes,esp. pp. 44-9, 65-71. inspires confidence. Furthermore, some trial conversions of 24 pp. Kashani, Tdrikh-iUljditii, ed. M. Hambly (Tehran, 1348/1969). hijri dates from the first folio of Tiisf's table into the 25 Bandkati also occasionally uses the animal calendar, genercalendar agree with modern tables Chinese-Uighur ally following Rashid al-Din, but sometimes able to provide (Kennedy, pers comm.). Further progress awaits an accurate corrections or alternatives. The edition should be used with transcription of the rest of Tusi's table. Cf. previous note. caution: for example, on p. 427, Banakati says Abaqa '8 See above, note 19, for subsequent work in the Ilkhanid returned to Tabriz from Mazandaran in the spring of the period. Amuli, Nafd'is alfunun, ed. Abu 'l-Hasan Sha'rni, III Year of the Ox (bahar-igdv) 664 A.H., which does not corre(Tehran, 1379/1960), pp. 459-61, describes the art of constructing almanacs, in which the "Turkish" years and months spond; Rashid al-Din, ed. Baku, p. 103, has spring (bahdrgdh) are featured. The author died at Shiraz in 753/1352. 663, which is inaccurate: it should be spring 664/1266 (Year of the Tiger); cf. Table 1, no. 18. ed. M. M. III (London, 39 Juvaini, Tdrikh-iJahdn-gushdy, Qazvini, 1937), pp. 95-6; tr. J. A. Boyle, The History of the World Con26 K-sh-ni, pp. 74, 75. For Mu'mini, see also Mustaufi, Tdrikh-i guzida, ed. 'A. Nava'i (Tehran, 1362/1983), p. 813. Fahkr alqueror (Manchester, 1958), II, pp. 610-11. Elsewhere, Rashid Din is also commemorated in a poem that appears to give his al-Din himself puts Huilegfi's departure west in the Year of death as 25 Sha'ban. These are the last complete dates the Ox, see ed. Karimi, I, p. 600, tr. Boyle, Successors,p. 223. 40 The Saturdays in Pisces 660 were 18 and 25 February, 4 and Kash-ni gives by the Chinese-Uighur calendar; it is possible that Fakhr al-Din was one of his sources of information. 11 March 1262. The only concrete information is that the sun was 50 over the horizon, i.e. he was born about 20 minutes According to Ibn al-Fuwati, Talkhis majma'al-dddbfi mu'jamalafter sunrise. alqdb,IV/3, ed. M.Jawad (Damascus, 1965), pp. 149-50, Fakhr al-Din died in 709 A.H., but here Kash-ni's dating is 41 There is an overlap, for the last month of 683 A.H. does fall in the first month of the Year of the Cock. Professor Bazin preferable. (pers. comm.) has suggested that the animal year of both 27 See Doerfer, Elemente,III (1967), pp. 657-60. 28 In addition to nos. 5, 31 and 45 on Table 1, see Rashid al-Din, Arghun's birth and accession might have been falsified delied. 'Alizada, p. 283 for New Year's day in 694 A.H., i.e. 17 berately, the Cock being a far more illustrious sign than the Monkey. If we take the Monkey (1260) as the year of Arghun's January 1295. Boyle's reference (CHIr,V, Cambridge, 1968, p. birth, there is still poor agreement with the day of the week: 396) to Ghazan's celebration of the New Year on 1 Jum-da II, 703/10 January 1304, however, is an error: the celebrations 25 Aram/8 March 1260 was a Monday. were to mark Abui Yazid's birthday, cf. Rashid al-Din, ed. 42 Bandkati, p. 441, follows Rashid al-Din's hijri and animal 'Alizada, p. 363; the Mongol New Year was on 6 February in dates; Vassa-f, p. 137 has 7 Jumada I (22 July), as does Ahari, 1304. Tdrikh-iShaikh Uvais, ed. and tr. J. B. van Loon (The Hague, 29 E.g. Rashid al-Din, ed. Quatrembre, pp. 74-9, ed. 'Alizada, p. 1954), p. 138/tr. p. 40. K. Jahn's edition of Rashid al-Din (The 379. Hague, 1957), p. 60, has 9 Altinch-ay (22 July). July 22 was a Saturday, and this would be a plausible alternative, except 30 The birth of Mongke Temfir in 654 A.H. (no. 4). The Mongol 17

date however, is preferable, see Table 3. The prince is said to have lived 26 years and 2 months, which is nearly correct according to the lunar calendar. 31 The death of Abaqa (no. 29) is given twice, once with the animal date first, once with the hijri. 32 The death of Toqtani (or Toqiti) Khatun, one of Hiilegfi's concubines; the text ed. Quatremire, p. 94, puts her death on a Monday. Thursday is correct, as confirmed by the Mongol date, which is to be preferred (see Table 3); the hijri date is

that it does not fit with the death of Ahmad the previous day. See also B. Spuler, Die Mongolen in Iran (Berlin, 1955), pp. 81-2; he does not refer to the animal date. 43 Vassaf, p. 245, has 6 Rabi' I. The day of the month is missing from Rashid al-Din's text, ed. 'Alizada, p. 226, but the time is given as mid-morning (jashtgadh). For internal consistency, the following hijri date (no. 36) should also be altered in order to allow a true diurnal correspondence with the ChineseUighur date. 44 Banakati, p. 446, says he ruled 7 years, 9 months and 20 days,

THE

CHINESE-UIGHUR

ANIMAL

CALENDAR

OF THE

MONGOL

PERIOD

97

62

60 The Chinese-Uighur dates are studiously ignored by historians of the Ilkhanid period, hence the lack of previous

But not in earlier parts of the work, see the constant corrections provided in Boyle's Successors. 63 The Mongol month is probably accurate, so the Muslim date also needs correcting; cf. above, note 58. 64 Bazin, pp. 410-12, is unaware of the Safavid use of the animal calendar (cf. above, note 3). He discusses an example (p. 406) that appears to confirm a continuing adherence to the Chinese calendar in 831/1428; we may also recall Babur's reference to the continuing popularity of Nasir al-Din Tiisf's tables (see note 19), though this does not specifically concern the animal calendar. 65 'AliYazdi, Zafar-ndma,ed. M. 'Abbaisi (Tehran, 1336/1957), II, pp. 170, 249, is unambiguous for 1400 and 1401, and a spring start to the year is also implied e.g. by the dates given for Timur's departure for Moghulistain in 776/1375 (I, 189). Yazdi's dating is generally more precise than Shimi's Zafarnama, ed. F. Tauer (Prague, 1937, 1956). Tauer wrongly takes Nauriiz to be 21 March at this period. I intend to review the chronology of these two sources elsewhere. 66 Hafiz-i Abrui, in Shami, ed. Tauer, II, pp. 35 (concerning 1374) and 119 (1396): the Mongol (Chinese) New Year was on 10 February in 1396, nowhere near the beginning of 798 A.H. (which started in October 1395). 67 Bandikati, p. 26, incorrectly puts the start of the Kh-ni era at the end of 699 A.H. See also S. H. Taqizadeh, "Various Eras and Calendars Used in the Countries of Islam", BSOAS, IX (1937-9), pp._903-22 (pp. 118-20); idem (Taqiz-da), Gdhshumari dar Irdn-i qadim (Tehran, 1316/1938), pp. 161-4, 296-7; 'Abdallfihi, pp. 328-30. The correct date for the start of the Khaini era, given by Mustaufi, p. 606, is confirmed by al-Kaishi, Zij-i Khdqadn,fol. 22a. I am most grateful to Prof. Kennedy for performing Kaishi's calculations for me. Kaishi says the start of the new era coincided with Ghazan's succession, which is incorrect. See also Shams-i Munajjim Wiibkanwi, Zij al-muhaqqaqal-sultani, Aya Sofya, Istanbul, Ms. 2694, fol. 28a. Despite a detailed account of the new calendar, which he helped to create, Waibkanwi does not specify the hijri date on which the Ghazani era began. 68 Wibkanwi, fol. 29a, also Sayili, p. 229. The Turkish months were associated with the zodiacal signs, thus Aram-aiy was equivalent to Hamal (Aries), and 1 Aram was now identical to of the 1 Farvardin, rather than the beginning Chinese-Uighur year. See also al-Kishi, loc. cit.; Taqizaida, p. 164. 69 Wibkanwi, fol. 36, says it remained little known, cf. Sayili, p. 230; see also G. C. Miles, "The Inscriptions of the Masjid-i Jiimi' at Ashtarjin", Iran, XII (1974), pp. 96-7, for Abil Sa'id's coinage, which includes issues from 31 to 35 Khini (1332-6), some of which are dated by both Khdini and hijri dates. This type is also marked by a revival of Uighur script on the coinage. The Khaini era is frequently used by Vassfif, who was a revenue official in Fars, and by Mustaufi, Dhail-i tarikh-i guzida, ed. and (Russian) tr. M. D. Kyazimova and V. Z. Pirieva (Baku, 1986). 70 Tij al-Din Hasan Yazdi, Jdmi' al-tavarikh-iHasani, ed. H. M. Tabaitabi'i and Iraj Afshair (Karachi, 1987), e.g. pp. 46, 47, 48. The use of the Khaini era is not always accurate, see pp. 56, 57, 95, and it is dropped from the reckoning towards the end of the work. 71 Spuler, p. 91; for the evidence of Ghazan's coinage, see Sheila

discussion of the subject. The question is briefly aired by R. Stephen Humphreys, Islamic History. A Framework for Inquiry (London, 1991), p. 130. 61 For the death of Toqtani Khatun, see above, note 32, and for the two battles, Abu 'l-Fidfi, al-Mukhtasar ft trirkh earlier al-bashar (Cairo, 1907), IV, pp. 9, 15, following sources. In Jahn's edition of Rashid al-Din (1957, p. 31) the battle of Albistin is put in Onunch-fiy (10th month), but is (3rd month) in the Baku ed., correctly OtunchlUchfinch p. 144.

S. Blair, "The Coins of the Later Ilkhanids: A Typological Analysis",JESHO, XXVIliii (1983), esp. p. 297. Ghazan's coins to use the Uighur script (696-703 nevertheless continued A.H.). For the revival of links under Oljeitui, see e.g. Boyle, "Dynastic and Political History of the Ilkhans", CHIr, V (Cambridge, 1968), pp. 398-9: for the date of the letter referred to there, see below (note 75 and Table 4). Oljeitii revived the use of Uighur script on his coins from 709 until 713 A.H. 72 Cf. above, note 66. Such references are disappointingly few. Letter of Hulagu, Il-Khan of 73 P. Meyvaert, "An Unknown

which is one year too many. According to the lunar calendar, he ruled 6 years, 9 months and 9 days. 45 I am indebted for this ingenious explanation of the 33 years to Professor Bazin. We may also note that according to correct Mongol reckoning, Arghun would have been 33 lunar years, inclusive. 46 Bandkati, p. 437. 47 In my article "Bologfin Kaitfin"in EIr, IV (1989), pp. 338-9, I wrongly said 22 March was Nauruiz; Nauriiz in 1290 was 13 March. 48 Cf. Bazin, p. 409; see also M.J. L. Young, "An Arabic Almanac of Favourable and Unfavourable Days",JSS, XXVII/ii (1982), pp. 261-79 (p. 273). 49 Ibid., p. 278, confirms that the 23rd of the month is "excellent for every matter", while the 24th is "continuously inauspicious". Bandikati, p. 447 has Sunday 24 Rajab,_but the year is given as Qoyi (sic, Sheep) for Taulai (Hare); Ahari, p. 140/tr. p. 43 also has 24 Rajab. Vassif, p. 260, merely has Rajab 690, and appears to think it was spring! The same author puts Gaikhatu's second coronation (no. 39), in Jumidfi II, 691, which he also says was springtime. Concerning this later event, it should be noted that in Jahn's edition (1957, p. 85) 12 Rajab is equivalent to 4 Altinch-fay (19 June); the correct correspondence, 14 Altinch-aiy, is given in the Baku edition (p. 236). 50 I.e. so that "Monday 15th" was really Monday 16th, according to the standard civil calendar. 51 On p. 30, he says that 17 Dhu 'l-Hijja was a Wednesday, which is consistent with his Monday 15th. On the other hand, on p. 23, he says 2 Dhu 'l-Hijja was a Monday, which would make the 15th a Sunday. s2 Otherwise, Kaish-ni is again inconsistent here, unless he took Dhu 'I-Hijja to be 29 days; but in 703, which was a leap-year, it had 30 days. 5 A difference of two days between the start of Muharram and the start of Yetinch-fiy could be expected because the previous months were of unequal length: Altinch-afy had 29 days, and Dhu 'l-Hijja 30. 54 Whereas it is an unlikely scribal error for shanzdahum(= 16) Altinch-fly (equivalent to 15 Dhu 'l-Hijja). 5s Bandikati,p. 474, Mustauff, p. 606 and Vassfif, p. 467. Kaishaini and Banikati both have munta4af (middle of) Dhu 'l-Hijja, which would normally be the 15th, or sometimes 14th, of a lunar month. The Muslim "Monday" begins at sunset on Sunday, but it is stated by Bandikati that the coronation occurred in the morning, which is what one would expect. Ahari, p. 147/tr. p. 49 has 2 Dhu 'l-Hijja, clearly in error. 56 Cf. note 48. The 15th is a good day, but the 16th is "continually inauspicious", Young, pp. 276-7. The sources confirm that the day was specially selected by the astrologers. '7 A further correction is therefore needed to Melville, "Bolog•in Kaituin",p. 339. s8 Resolved without explanation in idem, "The Itineraries of Sultan Oljeitii, 1304-16", Iran, XXVIII (1990), pp. 55-70 (p. 65). The same paper (p. 66 and note 137) discusses the correct date for Oljeitui's death (no. 77). 59 Kaishaini,p. 66, refers to Wednesday 19 and Thursday 20 Dhu

'l-Hija.

98

JOURNAL

OF PERSIAN

Persia, to King Louis IX of France", Viator, XI (1980), pp. 245-59; E. Tisserant, "Une lettre de l'Ilkhan de Perse Abaga, adress6e en 1268 au Pape Cl6ment IV", Le Museon, LIX (1946), pp. 547-56; A. Mostaert and F. W. Cleaves, "Trois documents mongols des Archives secr&tes vaticanes", HJAS, XV (1952), pp..419-506; idem, Les Lettres de 1289 et 1305 des ilkhanAryun et Oljeitiiti Philippe le Bel (Harvard, 1962); Cleaves, "The Mongolian Documents in the Mus6e de T6hhran", HJAS, XVI (1953), pp. 1-107; G. Herrmann and G. Doerfer, "Ein persisch-mongolischer Erlass aus dem Jahr 725/1325", ZDMG, CXXV (1975), pp. 317-46; eidem, "Ein persischmongolischer Erlass des Galayeriden Seyh Oveys", CAJ,XIX (1975), pp. 1-84; Larry V. Clark, "On a Mongol Decree of Yisiin Temiir (1339)", CAJ,XIX (1975), pp. 194-8. 74 For other examples, see Mostaert and Cleaves, "Lettres", pp. 52-3. Herrmann and Doerfer, "Seyb Oveys", pp. 57-8, also draw attention to this difference between the documents and the chronicles, but do not offer an explanation. Neither do they appear to notice that the dates in the decree of 759/1358 are not equivalent: the Mongol date should be the third autumn month, not the second (yielding 18 October 1358), and even then there is a discrepancy of one day. see Mostaert s5 For a discussion of the term qaghuchin/qaghuchid, and Cleaves, "Lettres", pp. 49-54. Only the bilingual decree of 725/1325 edited by Herrmann and Doerfer allows an exact concordance between the two calendars, and confirms that the qaghuchidare the last ten days of the month, numbered backwards, i.e. 21st = 10 qaghuchid,22nd = 9, etc. In fact, the correspondence is only exact if the Muslim date on the decree is taken to be according to the astronomical era; otherwise, there is a "nocturnal" rather than diurnal agreement. However, the agreement is worse using any of the alternative theories about the term qaghuchid,such as that it refers to the second half of the month, or that it is not numbered retrogressively. The other qaghuchiddates are therefore converted according to this method in Table 4. This allows a definitive solution to the date of Oljeitii's letter to Philip the Fair, written on 16 May 1305 at Barzand, which exactly fits his movements, see Melville, "Itineraries", p. 64 and note 93. It will also be

STUDIES

noted, however, that there are substantial differences between the calculations given in Table 4, and those found in H. F. Schurmann, "Mongolian Tributary Practices of the Thirteenth Century", HJAS, XIX (1956), pp. 304-89 (esp. 341-3), citing Priselkov; cf. Mostaert and Cleaves, "Lettres", p. 53. 76 Eidem, "Trois documents", p. 445, following Pelliot, suggest this seasonal dating was a Uighur characteristic and not necessarily exactly in line with the Chinese standard calendar. Bazin's researches confirm the Turkish character of the calendar based on the four seasons (Systemes,esp. pp. 430-51, concerning this usage among Comans (Qipchaq) of northern Crimea), but here it is used together with the twelve-animal calendar, which aligns it with the Chinese calendrical system. The first spring month is equivalent to the first Chinese month, etc., as confirmed by the concordance of the dates on the decree of 725/1325 (see above). 77 For the Golden Horde, see Schurmann, pp. 340-8, Mostaert and Cleaves, "Lettres", p. 53; for the Chaghatay, Clark, pp. 194-5; and for the Jalayirids, Herrmann and Doerfer, "Seyb Oveys". In laterJalayirid decrees, there is only a hijrf date: see H. Masse, "Ordonnance rendue par le prince ilkanien AhmadJalair en faveur du Cheikh Sadr-od-Din (1302-1392)", JA, CCXXX (1938), pp. 465-8 (concerning a decree dated 773/1372), and G. Herrmann, "Ein Erlass des Galayeriden Solttan Hoseyn aus dem Jahr 78011378", in Erkenntnisseund Meinungen, I, ed. G. Wiessner, Gottinger Orientforschungen,I (Syriaca), 3 (Wiesbaden, 1973), pp. 135-63 (pp. 136-8). 78 Juvaini, tr. Boyle, I, pp. 6-8. On the element of continuity, see A. K. S. Lambton, Continuity and Change in Medieval Persia (Albany, 1988), pp. 50-68, 297-8, 305-9. For some doubts, see D. O. Morgan, "Prelude: The Problems of Writing Mongolian History", in Mongolia Today,ed. S. Akiner (London, 1991), pp. 1-8 [esp. pp. 5-6]. For the continuing importance of Mongolian in diplomatic relations, see D. P. Little, "Notes on Aitamig, a Mongol Mamluik",History and Historiographyof the Mamluks (Variorum Reprints, London, 1986), no. VI, esp. pp. 392-6. The continuing use of Uighur on late Ilkhanid coinage also points in the same direction, though Blair, p. 305, considers Abui Sa'id's last issue a "throwback".

ANECDOTES OF A PROVINCIAL SUFI OF THE DEHLI SULTANATE, KHWAJA GURG OF KARA By Simon Digby Rozel,JerseyC.I.

The contemporary and near-contemporary literature of Sufis of the Dehli Sultanate sheds welcome light on the religious and social life of north India in the thirteenth and fourteenth centuries, and upon the ideas and concepts which prevailed in Muslim and Sufi circles in this environment. But the chances of manuscript transmission have left us with but a selection out of all that was originally composed in this period, and this selection is heavily weighted in favour of the more highly literate sections of the diverse and divergent body of Sufi men of religion. Leaving aside the "spurious Malfilzdt" attributed to the early Chishti Shaykhsin India, first recognised and denounced in Professor Mohammed Habib's pioneering article,1 we are still exceptionally well-informed on the attitudes and activities of the later great Chishti Shaykhs from the clearly, often precisely dateable records of the conversations of the conversations of Nizam al-Din "Awliya'", Nasir al-Din "Chiragh-i Dehli", Burhain al-Din "Gharib" and Sayyid Muhammad Gesuidaraiz. In addition, there are two posthumous but valuable Chishti biographical tadhkiras of the fourteenth century, the Surir alsuduirand the Siyar al-awliyd'.2The literature of the Firdawsi Shaykhs of eastern India has survived equally well, including letters and recorded conversations of Sharaf al-Din Maneri and a useful tadhkira, Manaqib al-asfiyyd'.3By contrast, unless more authentic manuscript sources are found, the Suhrawardi Shaykhs of Multan have fared very badly.4 From the late fourteenth century we have two collections of malffiz [discourses] of Shaykh Jaldl al-Din "Makhduim-ijahaniyan" of Ucch, both compiled in Dehli, though one of these has been characterised as inauthentic by its editor.5 This, then, is almost the entire corpus of Sufi biographical literature from the Dehli sultanate up to and shortly after the disaster of Timur's invasion of A.D. 1398.6 This paper examines and describes another work that may be added with confidence to the corpus surviving from before A.D. 1400. It is of interest in that it describes a different and lower social level of religious behaviour. We have, however, first to establish four basic items of information: the correct title of the work, the name of its author, the correct name of the holy man whose activities it describes and the date of composition. 99

The work in question has not escaped the vigilant eye of the veteran historian of North Indian Sufism, Khaliq Ahmad Nizami, who provides a brief description: "The Asrar-u'l-Makhduminis the malfuz of Khwaja Karak and contains interesting anecdotes concerning the saint. It was compiled by Karim Yar. Nothing is known about the compiler or the date of compilation."7 The work has previously been known only from the rare lithographic printing of the Nasim-i Hind Press, Fathpur-Haswa, 1893. As was common in Urdu and Persian printings of this period, the name of the patron who commissioned and doubtless subsidised the edition appears prominently upon the ornamental title page. It was Karim Yar, Ra'is of Yuhan, who was alive a hundred years ago, and not an author of earlier centuries. The erroneous identification of this "Ra'is" as the author first appears in Edwards's Catalogue of Persian Printed Books in the British Museum;and it is repeated in C. A. Storey's Persian Literature: a Bio-Bibliographical Survey (No. 1411, Appendix, No. [38]).8 We may note that the work is loosely described as a malfizi (record of conversations) rather than a tadhkira (collection of biographical anecdotes) on the title page of the lithograph, and this mild inaccuracy has been passed down through Edwards and Storey. No manuscript of the work was known to Storey, nor to my knowledge has any notice of one appeared in any other published catalogue. This is of importance, because in the text of the lithographed edition there is no clear mention either of the name of the author and the date of composition is garbled. However, even without these a close perusal of the anecdotes would have revealed that the author was the grandson or greatgrandson (nabisa-pisare,the phrase leaving room for either interpretation) of a contemporary of the Shaykh, who came from a family which had removed from Lahore to the town of Kara on the Ganges (Anecdote no. 129). Given that the subject of the tadhkira died in 700 A.H.IA.D. 1301, this would suggest a date of compilation not later than the third quarter of the fourteenth century.9 The matter would rest there, but for a manuscript which came to light in an old private collection in Western Europe.iO The text differs very

100

OF PERSIAN

JOURNAL

STUDIES

IsI

\

$.4

c

AN

as

?1

CL< VI

4

-L-

KV$

-

N

N\

;

-

N-'7

~ B

AN

4&:&1 `

;i

It4 QS

isi r~t:?

3- n_

Fig. 1. Ms, Asrar al-majdhiibin,fols.

lb-2a.

`03 1>

4'

OF A PROVINCIAL

ANECDOTES

SUFI

?~'-

DEHLI

SULTANATE,

-c

?t

- ,5 -

GURG

KHWAJA

=

.............................

L N

{N-~

"iw

~-~

9

9:-

N

q\

,•

\ v;24:;,%.

< "g

',

.i-•A

04,

4p.. 4\

SI ~z4r It

vw

OF KARA

xN$

c

?r\

~~,

17:

OF THE

• - A'

Fig. 2. Ms, Asrar al-majdhibin, fols. 2b-3a.

•

.

:

101

102

JOURNAL

OF PERSIAN

STUDIES

little in its readings from the lithograph, though it are sometimes represented as wolves with a single is evidently not this manuscript from which the horn growing from their foreheads.13 In the text of lithograph was copied. From a perusal of the the Asrdr al-majdhuibin,the subject is described by opening it becomes evident that the second folio others and by himself in verses and in conversation was missing from the manuscript from which the as KRK-i gusfandin "sheeplike KRK";KRK-i payband lithograph was transcribed, and the copyist, as in "KRKwith chained feet"; and KRK-ipir "old KRK". such instances is often the case, has not even These epithets make no sense if they are applied to noticed the break. (In effect, he has copied straight "a crack of thunder", but perfect sense when on from the bottom of the original folio 1 verso to qualifying "a wolf' in the phrases "a sheeplike the top of the original folio 3a.) It is upon the wolf', "a chained wolf', and "an old wolf". The last second folio that the present manuscript provides phrase is used in the Persian poetry of this period, us with the name of the author, which we read as as in modern English, for a cunning and Muhammad 'Ali Isma•il (son of?) Mas'id 'Ali, 'urf experienced veteran. One may add that gurg-i pir "Sajjan"Lahawri.It also describes in brief detail the "old wolf' (applied to the Khwdja)occurs in a poem circumstances of the compilation of the work and in the text in mutaqdribmetre which would not scan the date of its completion (khutimabi-tawfiqi'lldh), if one read karak.r4 It may help us to understand the character of this the beginning (ghurra) of Rajab 771 A.H., cor1370.11 A.D. end of to the strange hagiographical work if we next examine January responding We may now turn to the title of the work. In the the information which the Asrar al-majdhubinprolithographed printing, this appears as Asrdr al- vides concerning the author or compiler and his makhdimmin ("Thesecrets of the lords").This has no family. We should probably understand his name obvious relevance to the personality and the activi- as Muhammad Ismai'il son of Mas'fid 'Al. His 'urf ties of the holy man described in the work. In the name, which is shared by earlier members of his manuscript, the title appears in the alternative family, in the manuscript is consistently written in

form of Asrdr al-majdhubin("The secrets of those drawn away"). Those conversant with Arabic script and its handwritten forms will realise that this variation of the name represents a minor alteration in the placing of the dots and the thickening of one initial. The term majdhuibis used in Sufi literature in India and elsewhere (e.g. in a whole section of 'Abd al-Haqq Dehlavi's Akhbaral-akhydr)for a class of "madmen" who are so "drawn away" by divine love that they appear to have abandoned the of Muslim ordinary norms of behaviour believers.12 In this text, the subject is often described, both by others and from his own mouth, as majdhiiband divdna ("mad"). Asrdr al-majdhibin is therefore to be preferred as a more apt title for the work. The last ambiguity of which we can dispose concerns the name of the holy man. In modern times, in the nineteenth and twentieth centuries, he has been remembered as Shah Karak. (It may be noted that he is called Khwdja, not Shah, throughout the text.) Karak with retroflex "flapped" r is an onomatopoeic north Indian word used much like its English homophone, bijlf ki karak, "a crack of thunder". In Arabic script it is represented by the letters kdf rd', kdf (In older palaeography, there is seldom a sign over the rd' to indicate the retroflex r). In most pre-modern handwriting, the Persian letter gdf is not differentiated by a superscript stroke from kdf Kdf rd', kdf, as students of the Shahndma will be aware, may equally well represent karg, a rhinoceros, and gurg, a wolf. Through this amphibology, in mediaeval illustrations to Firdawsi's poem, the wild beasts which Isfandiyair confronts

shikasta script as SJN, which we should probably read as "Sajjan".15 The lithograph reads Sanjar and Sanjari.16 The compiler's nisba is Lahawri, indicating that his family came from Lahore, though, as we shall see below, they had been settled for some generations in Kara. The compiler of the anecdotes states that when he was attending on Sayyid Jalal al-Din Abui Sa'id Bukhari,17 a Darvesh gave him stories and verses about the miracles, sayings and spiritual authority (wildyat) of KhwdljaGurg Abdal. This Darv-sh told him that he had been happily passing his days reading Sufi works when he saw Khwdja Gurg in a dream. Gurg commanded him to give the material about himself which was in his house to Muhammad Ismfil L-hawri. The compiler gladly accepted the material from the Darvesh, but did not feel inclined to proceed with the composition until he also had a dream, in which Khwdja Gurg appeared and said: "Young man, why have you not heeded the words of this madman? And why are you not putting them together?" The compiler began to tremble and woke with a start. After recitation of the Holy Qur'-n and the performance of the dawn prayer, Muhammad Ismia'fl started to compose the work. From the proem and from a number of anecdotes (nos. 59, 51, 104, 117, 129), it is possible to gain some picture of the family of the compiler and their connection with Khwaja Gurg. They bore the nisba Lihawri, but appear to have come from Lahore to Kara on the Ganges at least some decades before the end of the thirteenth century.

ANECDOTES

OF A PROVINCIAL

SUFI

OF THE

that they were sharif There is no indication Muslims of other than South Asian lineage, but it is claimed that the names of 'Ala' al-Din and Mas'Uid had been born in alternation by men of the family for twelve generations (Anecdote no. 117). They had links, not with the capital city of Dehli, but with the Lahore and Multan. Some other anecdotes suggest that this was true of other members of Muslim population of Kara, an indication which may be of relevance when the settlement patterns of the Dehli sultanate are studied. Successive members of the family of the compiler appear to have used the 'urf "Sajjan". They are nowhere claimed to be great Shaykhs with khildfat, but their Sufi allegiance is to the Suhrawardi Shaykhs of Multan. Among the family, Gurg's older patron Mas'Uid 'Ali was a Murid of Bahi' al Din Zakariyya'. At Mas'Uid 'Ali's funeral at Kara, Khwdja Gurg claimed to have seen the three great Suhrawardi Shaykhs of Multan, Bahai' al-Din, Sadr al-Din and Rukn al-Din, invisibly and supernaturally present (Anecdote no. 104). The family appear to have been wealthy citizens, and they a building called the maintained in which Qalandars and other Darveshs received Si.fi-khana, An element of loyalty lodging and maintenance. and pride of family has entered into the compilation of Asrar al-majdhibmn. statements about Khwija Gurg Biographical himself are rare in the work. It is likely that he was a locally-born Muslim who never travelled far from his birthplace of Kara on the Ganges. His parents evidently died when he was still young; one anecdote has him weeping beside the grave of his father (Anecdote no. 175). We may discount the numerous anecdotes, sometimes denied by him and sometimes tacitly acknowledged, of Gurg being seen by travellers from far away at the Ka'ba (nos. 144, 176), or at the Prophet's grave in Medina, or in Bukhara (no. 194). One of the most attractive features of the work is the local feeling which Khwija Gurg and the compiler show for the settlement of Kara and its suron the Ganges. Both Gurg and the roundings to the Muslim sanctification of contribute compiler the local landscape.18 Outside Kara, Gurg had a vision of the Prophet passing by (no. 49). This was a sign of the death of a member of the compiler's house, a disciple of Bahq' al-Din Zakariyyi' of Multan. It was at his funeral that Gurg, seen conversing with invisible personages, claimed that the three great Suhrawardi Shaykhs were present, though invisible to others (Anecdote no. 104). In an anecdote which is historically possible, KhwLja Gurg was urged by Sultan 'Al' al-Din Khalji to come to DehlT (Anecdote no. 57). He declined the invitation, and recited a favourite quatrain of his own composition:

DEHLfSULTANATE,

KHWAJA

GURG

OF KARA

103

qdni' shudam bi-khushkninegu tara mailam nabuvad bi-siiyibirydnu bara dihli u samarqandu bukhdrdu 'irdq har chdr turd bdd u mard bdd kara I am content with dry My appetite is not for DehlT and Samarqand You can take them all

bread and vegetables; roast meat and lamb.19 and Bukhara and 'Iraq, and leave me Kara!

A further instance of Khwdja Gurg's local feeling was when he visited a boys' school (no. 61). He remarked: "The boys of Kara have a good temperament. Whatever they write, they write well! There is no place like Kara (hamchi maqam-i kara jda' nest)!" to the boys that "this Gurg then observed madman" (in diwana) had written a quatrain out of regard for them and they should memorise it. We may quote it as another typical and favourable example of his style of verses: 'ishq-i azalI td bi-abadkhwdhadbfid jilyanda-i 'ishq bei-'adadkhwdhadbild fardd chu qiyamatdshkaragardad ay harki na 'ashiq ast rad khwdhadbiid Eternal love will endure till the end of time; The seekers of love will be without number; On the day when God's judgement is made manifest, Yea, any man who is not a lover will be cast forth. Once, when in a happy mood, we are told, Khwaija Gurg described his entry into the religious life, and what he considered to be his particular vocation (Anecdote no. 102). He was an active youth of some sixteen or eighteen years, and he was accompanying a local armed band encamped not far from Kara at a place called Bhambharoll.20 In the middle of the night he awoke, but he saw no one around, and he began to weep for his (evidently dead) father and mother. He then heard a voice, which told him to go into Bhambharoli, where there was a holy man called Mawlaini Ismi'l1, and to do what he would be told by him. (The name suggests that he was yet another member of the Laihawri family of the compiler of the text.) [After going into the settlement,] Gurg beheld a Pir with spiritual authority (sdhib wildyat) standing before the door. The Pir greeted Gurg as soon as he saw him, and he said: "O brother in the faith, it is good you have come! It is nearly ten years that I have been waiting for you. Take this letter to the hill where the lamp is burning, and deliver it to Khwdija Khizr. He will tell you whatever is to be!" Gurg felt longing and pain and exaltation, and immediately set out. He came to the foot of the hill where the light burnt. Before the hill there was a vast expanse of water. He drank of the water and climbed the hill. There he found an old man performing his prayers on a prayer carpet. He greeted

104

JOURNAL

OF PERSIAN

him and placed the letter in front of him. Khwaija Khiir (for it was he) pronounced: "Verily the time has now come! God has called you His own Abddl.21 Go back to Kara and drink the wine that is reprobated by people, and turn it to rosewater!" Gurg bent his head low, and set off back to Mawlana Isma'fl. [When he entered his presence,] he saw that the latter was seated with two personages. He greeted them and they answered his salutation. He then sat down politely on his knees. After a while he asked his preceptor for wine, who said to him that it was "over there". "In Hairat's cell?" Gurg asked. Mawlana Isma'il replied: "Go yourself and get it!" Gurg, in telling this anecdote, remarked that the wine he found was in the very same vessel that Khwaja Khizr had shown him. After this, Gurg took his leave from his Pir and came to Kara. From that day on, he states, he found the scent of rosewater in wine; and till the day when he was speaking in all that he drank Gurg found the scent of the Friend (God), and this inspired the thought of Him. In the whole of the Asrar al-majdhiibin, this strange account is the only passage of sustained and allegedly autobiographical narrative; and it has to be said that, like almost every other anecdote in the work, it is hard to accept it as a realistic account of objective happenings. We may add here the sole concrete item of information regarding Gurg's life to be found in the work, namely the date of his death as given in the final anecdote (no. 210). On Saturday, 3 Rajab 700 A.H. (15 March 1301), Gurg, when he was brought his morning wine, contrary to his wont, poured it on the ground. He passed away in the middle of the following night. The majority of the anecdotes show the same aberrant sense of mission that is revealed in the narrative of his call to the spiritual life. The formulaic anecdotal pattern displays Gurg seated "in the bazaar", sometimes in the town mosque, occasionally visiting the boys' school or the house of his patrons and devotees of the Lahawri family. Occasionally, he is wandering outside the city or even bathing in the Ganges, which he calls the local Dijla or Euphrates. Most often, he is sitting at the liquor shop. There were evidently two of these alcoholic run by Hindus called Bhola and establishments, Bandi (Pandey?). Bhola, we are told, was childless, of Khwaja Gurg, he but through the intercession later had sons by his three wives, and this led to his conversion to Islam; but he evidently did not abandon his profession of wine-selling. Many of the anecdotes testify to the belief that the wine consumed by Khwaja Gurg changed to guldb (rosewater) and honey. Its sweet fragrance scented whole adjacent areas of the bazaar. Yet in

STUDIES

is other anecdotes inebriated, Gurg, evidently as fallen down unconscious having represented under the and naked (no. 176). His companions influence of shardb sometimes turned violent and disorderly and even laid hands upon the Khwdja himself. The local shardb is therefore represented as a strong physical intoxicant; and if these anecdotes can be taken as historical evidence, there was evidently a problem of lower-class urban alcoholism in northern India in the thirteenth century. As also in a well-attested poem of Kabir which we must assign to around the fifteenth century, in this text there are references to the fiery bhati in which strong liquor is distilled.22 Into the fire of a still from which the liquor was dripping, Gurg is said to have thrust the khirqa (cloak) despatched to him by Shaykh Nizalm al-Din of Dehlf, though it was afterwards taken out and sent back unharmed (Anecdote no. 196, f. 80). As Gurg sits drinking his wine, he is joined by visitors. (More anecdotes start in this way than in any other.) Some of the visitors are hostile. Among them there are men of education (muta'allimzn), a royal official who wishes to chastise Gurg for drinking wine, quarrelsome Qalandars and mere drunkards. What is built into the pattern of these anecdotes is the outcome, that Khwaja Gurg will triumph over his opponents. The arguments and criticisms of the verbose men of learning are confounded. The royal official repents and becomes Saq- or cupbearer to the Khwdaja(Anecdote no. 172). The violent Qalandars and drunkards are paralysed and rendered powerless (Anecdote no. 91). The tendency of the anecdotes is on the whole merciful rather than wrathful. Khwaja Gurg's opponents are usually given the opportunity to repent, and from the paralysis or blindness that has struck them they are often restored to the use of their faculties. Such anecdotes are often accompanied by homilies by Gurg on Islam or tasawwufand spiritual love, or by tales of the Prophet of Islam and his times. The homilies and tales are on the whole of a sobriety which contrasts favourably with the setting and wild superstitious atmosphere and unconvincing detail of many of the anecdotes. Ayas of the Qur'an and hadiths are frequently and correctly quoted. The anecdotes often end with the Khwdja reciting a quatrain of his own composition. Other visitors to Khwdja Gurg, who believed in his paranormal powers, sought his aid. They bore offerings, often money that was immediately spent upon wine, or sometimes wine itself. A Hindu who wanted a son brought a flower; and he is told that he will have two sons (no. 110). Many other men wanted sons, but two other suppliants, one a traveller, came to collect money for the marriage of their daughters (nos. 146, 154). Yet another man asks for money "because he has no sustenance". Gurg gives

ANECDOTES

OF A PROVINCIAL

SUFI

OF THE

him five gold tankas, telling him to spend four and keep one for a day of need (no. 165). A number of anecdotes testify to the belief in Khwaja Gurg's ability to produce gold at need.23 For those who wanted money, the Khwdja used to put his hand into a pot (no. 63) or to strike the ground (nos. 53, 114, 127), and gold or silver coins It is a feature of such goldwere produced. anecdotes that the gold is always producing bestowed or spent on suppliants, never used for the needs of the wonder-worker himself. Conforming to an anecdotal pattern found elsewhere in Indian Sufi hagiography, of Shaykhs displaying a river of gold, are two anecdotes which associate this power of Khwaja Gurg with the most feature of Kara, the conspicuous geographical on whose banks the settlement river great Ganges stood. One of these anecdotes concerns Khwaja Gurg's alleged relations with Sultan 'Ala' al-Din Khaljl, considered below. The other is a long and exuberant anecdote (no. 113), in which Khwaja Gurg is represented as drinking wine with sixteen other people from the mid-morning prayer to the afternoon prayer, without the measure (man) of wine ever diminishing. He then bade them follow him and himself sat down in the water of the Ganges. The sixteen followed him. Then he asked them all what they desired. One said: "I have need of two tankas of gold." [Khwija Gurg] put his hand into the Ganges and drew them out. The second said: "A tanka of silver." He drew it out and gave it to him. ... From the waters of the Ganges, nuts, a cup of sherbet and sugar followed, roasted wheat and kabdb and shardb, a veil and more "current coin" (tanka-i ra'ij al-waqt). One of the sixteen wished to be brought to God and (in a naive and irreverent response to this wish) he was thrown up into the sky. From the river the various demands of the whole band were satisfied.24 In another anecdote (no. 11), Gurg, sitting in the Ganges, brings out of the water for members of the party accompanying him sherbet, halwa and milk, and finally honey, the last at the request of a girl. The Ganges flowed between the towns of Kara and Manikpur. Like the Buddha before him, Khwija Gurg is shown walking upon its water and crossing to the other side (no. 68): It is related that Khwaja Gurg ... was sitting [in his usual place in the bazaar at Kara] when a man came and said, "The Mad Boy has died in Manikpur." Khwija Gurg rose and said, "Come, we shall go to Manikpur and visit [his burial]!" Slave-like, the man fell at the Khwaja's feet; and he left with him, the Khwdja in front and the man

DEHLfSULTANATE,

KHWAJA

GURG

OF KARA

105

following. When they arrived at the crossing there was no boat there. "What do you want with a boat?" the Khwaja said. "Come along with me!" The Khwdja took his hand and they went across the level surface of the river just as if it was dry land. Most of those who were at the crossing trembled and said: "Hurrah for the spiritual authority (wildyat) that the Khwdja has!" They arrived at Manikpur when the Mad Boy was just about to be taken out [to be buried] ... So when the Mad Boy had been taken out and buried, the Khwdja stayed for a while, and then he went back to Kara in the same manner.25 Of Khwaja Gurg's clientele, some visitors resorted to him because they had a wife or relation afflicted by evil spirits, Paris or De-vs (nos. 88, 155, were themselves so 169); or, less commonly, afflicted (no. 92). Khwaja Gurg used to chase these spirits away, in one instance forbidding the spirit ever to come back within the boundaries of Kara. In two cases, the Parn volunteers that he will not remain longer in the town (nos. 88, 169); in yet another, the King of the Paris promises that they will no longer roam "round and about the settlement of Kara" (gird u payramun-i havdli-i kara; no.

205).

the most notable of all the giftPerhaps in mediaeval credited to Sufis bestowing powers India and elsewhere in the eastern Islamic world is that of bestowing kingship on aspirants to power. The tale was widely believed that Khwija Gurg had the rise to the throne of 'Ala' al-Din prophesied Khalji when the latter was Governor (Muqta') of Kara. 'Isaam-, who wrote in the Deccan twenty years before the composition of the Asrar al-majdhuibin, has a story regarding this bestowal.26 Our text has a number of anecdotes about the special relationship between Khwdja Gurg and Sultan 'Ala' al-DTn. When 'Ala' al-Din was first appointed Governor of Kara, he rode past Khwaja Gurg and some Qalandars as they were drinking wine. The future Sultan dismounted and fell at Khwdija Gurg's feet, whereupon Gurg told him that he had given him the of Dehli. After he had attained the kingdom throne, he offered five trays of gold coins and five of silver to Gurg, but the Khwdja took him by the hand and led him to the banks of the Ganges, where he told him to close his eyes. When the Sultan opened his eyes he saw that the whole great river was flowing with gold and silver (Anecdote no. 136). When the Sultan opened his eyes, the Khwa~ja asked: "What do you see?" The Sultan said: "The whole of the river is flowing with gold and silver. It is saying, 'Here I am, Gurg! Here I am!'"

106

JOURNAL

OF PERSIAN

The Khwdajaremarked: "This is the treasure of My Friend (God), so it belongs to me. I do not need to look towards others." In 'Ala' al-Din's mind, great awe took hold, and like a slave he asked pardon at the foot of the Khwdja.To set his mind at rest, the Khwdjaaccepted a single silver tanka from him.27 'Ala' al-Din rose to supreme power by the treacherous murder of his uncle, SultanJal-l al-Din Feroz, at a meeting arranged on a boat on the Ganges. The Asrar al-majdhibin contains a version of this incident, which is a disingenuous attempt to absolve 'Ala' al-Din from all blame while exalting the role of Khwija Gurg. According to this version, Jalll al-Din, being afraid of his nephew (who was also his son-in-law), had tried to persuade his daughter, 'Ala' al-Din's wife, to poison the latter. But husband and wife fled; and with Khwdja Gurg's advice and blessing, 'Ala' al-Din brought back treasure from his successful raid on Devgir (in this source called Kharagarh). After his return, at the meeting between uncle and nephew in the boat on the Ganges, it was (according to this unconvincing account) Jaldl al-Din who had plotted treachery against his nephew. There were eight armed men hidden under his throne on the boat; but when these rushed out, dagger in hand, the appearances ofJal-l al-Din and 'Ala' al-Din were miraculously reversed. It was Jal-l al-Din whom the armed men stabbed to death in error (Anecdote no. 196).28 There are many miraculous anecdotes of familiar patterns in the Asrar al-majdhuibin.Bilocation is even more common than the production of gold. KhwLja Gurg, after dramatic and exhausting encounters in the bazaars of Kara, would suddenly disappear for days on end; and his adherents then believed that he had been away in distant holy places or the world of the unseen. On Friday eves he was supposed to be present at the Ka'ba (no. 144) or the tomb of the Prophet (no. 149). One particularly extravagant anecdote is the alleged testimony of a man at whose house Khwdja Gurg was passing the night. Outside, he suddenly saw a great light, and beheld a camel and its rider descending from the sky, escorted by two angels (Anecdote no. 4, folio 4b). This was to bear Gurg away to Mecca.29 The man hung onto the tail and was afterwards stranded in the holy city. Never mind, the servants assure him, Gurg will be back next of the and he can hang onto the tail for the return Friday,H.aram journey. On distant roads, Gurg would appear before travellers in peril, to save them from dacoits, ghouls or wild beasts (nos. 62, 86, 103, 146). Many anecdotes are part of the extended folklore of Sufi piety and Sufi powers, and they are also found in other hagiographical works and in oral tradition attached to the names of other Shaykhs.

STUDIES

A very widely-distributed story which finds a place in the Asrar al-majdhubin is that of the spiritual challenger who comes with wrathful aspect, riding a lion or a tiger and brandishing a snake as whip, only to be confronted and discomfited by the local resident holy man, who rides forward on the wall on which he has been sitting (Anecdote no. 183). Numerous Indian versions have been recorded, and it has been found far afield, among the Kurds in 'Ir-q, and among the Bektashis in Anatolia.30 There are anecdotes of a talking horse and of a talking snake, each of which obeys the commands of Khwdja Gurg to moderate its behaviour. The anecdote of the talking snake is a long one, recognisably modelled on popular dastan literature, in which Gurg travels to a distant region to recover the mate of the snake from the King of the Paris (Anecdote no. 84). To the historical folklorist, this suggests that the Paris of thirteenth-century Kara were ancient Indian Ndgas but thinly disguised.31 If the Paris are recognisably Ndgas, an identity of name (derived from the common Aryan linguistic heritage) ensured that the Devs ("demons") of Persian folklore were yet more easily identified as local deities (deiv,deva) in worship in the vicinity of Kara. In one anecdote, Gurg is shown seated at the shop of a cloth-merchant (bazzaz) called 'All He reproached him, asking why he kept Hindus in his house. (The implication seems to be that 'Ali himself was a householder who was a convert to Islam, but members of his household had not converted.) It then transpired that 'Ali himself went out every Sunday with flowers and a lamp to worship a Dev, who would trouble him if he did not appear. Khwdja Gurg proceeded to rid him of the Dev, saying that he had sent him off and "burnt" him; and 'Ali was troubled no more by the Dev. In the same way, one may reflect, the Muhtasib had sent troublesome Qalandarsout of town (Anecdote

no. 9).

These exorcisms of vexing spirits, in one case specifically required to dwell beyond the boundary of the town, again show local loyalty to the Muslim settlement of Kara. And it was not only against ghostly adversaries that Gurg's aid was sought. The Shiqqddror governor of Kara was engaged in a campaign against local insurgents at Bhambharoli and Mahoba. Ten days of fighting had brought no gains. He then despatched a messenger to Gurg, saying that he had been told by Sultan 'Al~' al-Din to resort to him in case of difficulties. As a result of Gurg's prayers, he conquered the village which he had been besieging (Anecdote no. 43). Though the terms of Gurg's appointment by Khwija Khizr to sit and drink wine in the bazaar may lack the grandeur of the delegated authority of the great Shaykhs of the major silsilas, it is a claim of the conferral of a territorial wildyat (dominion).32 It is from the

ANECDOTES

OF A PROVINCIAL

SUFI OF THE DEHLI SULTANATE,

KHWAJA GURG OF KARA

107

wilayat of Kara, also called his own wildyat (wildyat-i man) that Khwdja Gurg bids the Paris depart (Anecdote no. 205). A large portion of the anecdotes of the Asrar al-majdhuibin endeavour to present Khwaja Gurg as engaged in maintaining the well-being of the place which has been entrusted to him and of its inhabitants, by his actions, his advice and his spiritual power (barakat). is It cannot be claimed that the Asrar al-majdhuibin of great value for its factual or historical evidence, but it provides some information regarding the

social life of a Muslim settlement in northern India in the thirteenth century, for which other evidence is lacking. The work lacks any profound or subtle spiritual content, while providing many instances of extreme credulity and of antinomian behaviour. Yet it remains of value as a recognisable addition to the early corpus of biographical literature regarding Sufis of the Dehli Sultanate, and as evidence of levels of popular Muslim belief, practice and states of mind in a provincial urban community of the period.33

1 M. Habib, "Chishti mystic records of the Sultanate period", Medieval India QuarterlyI 3, pp. 1-42; repr. in K. A. Nizami (ed), Politics and Society During the Early Medieval Period; CollectedWorksof ProfessorMohammadHabib (New Delhi 1974), pp. 385-433. 2 The Surir al-sudir, written by a grandson of Shaykh Hamid alDin Nagawri, remains unpublished. It is frequently cited by K. A. Nizami, from a manuscript in his personal possession; see idem, On History and Historians of Medieval India (New Delhi 1983), pp. 183-7. Use has been made of another ms. in the Dargdh at Nagawr, Rajasthan, in a modern Urdu hagiography of Shaykh Hamid al-Din; see Pirzada Ahsan al-Haqq Fariqi, Sultdn al-tarikin (Karachi, Da'ira Mu'Tn al-ma'arif, 1963). Amir Khwurd, Siyar al-awliya' (Storey, no. 1250), ed. Chiranji Lal (Dehli 1302/1885); photo-repr. from the last with variant page numbers, Lahore (Mu'assisa-i intisharat-i Islami, 1398/1978). Shaykh Shu'ayb Man-ri, Mandqib al-afiyd' (Calcutta, NUir alafiiq Press, 1313/1895). For a summary notice of the extant literature of the early Firdawsi Shaykhs of Bihar, see S. H. Askari, Islam and Muslims in Medieval Bihar (Patna 1989), pp. 30-6. 4 Fazl Allah Majawi's Fatdwdal-4iifiyya,written in Arabic around 1350 A.D. and extant in ms. at Oxford (Bodleian, Uri 321), still awaits detailed analysis. It contains valuable information on the organisation and regulation of the Suhrawardi Khdnqdhat Multan; see Julian Baldick, MysticalIslam; an Introduction to Sufism (London, 1989), pp. 95-6. The Khulasat al'drifin regarding Baha' al-Din Zakariyya' of Multan (= Storey, p. 1057, no. (33)?) clearly belongs to the class of bogus compilations. An edition has been published; Banui Daktar Shamim Mahmiid Zaydi, Ahwdl u dthdr-iShaykhBahd' al-Din Zakariyya' Multdni u khuldsatal-'drifin ([Rawalpindi], Markaz-i Tahq-qat-i and Majma' alFarsi, 1974), pp. 123-91. For Fatdwd akhbdrfi manaqibal-akhydr(sic, apud M. Shafi), s.ifiyya a work regarding Shaykh Rukn al-Din and inaccessible, if any longer extant, see Nuir Ahmad Khan Faridi, Tadhkira Shdh Rukn-i 'AlamMultdni, (Lodhran [Zila Multan] 1381/1961), pp. 280-1; 'Abd al-Haqq, Akhbdral-akhydr(= Storey, no. 1298 [1]) (Dehli 1309/1892), p. 62. Both of these substantial mediaeval Indo-Persian texts have fairly recently been published from Delhi: Sirdj al-hidaya,ed. Qazi Sajjad Husain (Delhi, Indian Council of Historical Research, 1983 [1984]); Jami' al-'ulifm, ed. idem (Delhi, the same publisher, 1987). 6 For general surveys of this literature, see the following modern works; C. A. Storey, Persian Literature: a Bio-Bibliographical Survey, N. Biography:(b) Saints, Mystics, etc., I, Pt. 2 (London, 1953), pp. 923-1066; Bruce B. Lawrence, Notesfroma Distant Flute; the Extant Literature of Pre-MughalIndian Sufism (Tehran 1978); K. A. Nizami, Supplementto Elliot and Dowson's HistoryofIndia, Vol.III (Delhi 1981), pp. 92-112; idem,On History and Historians of Medieval India (New Delhi 1983), pp. 163-97.

For the literature of the Firdawsis in Bihar, see S. H. malfi.z and Malfuz Literatureas a Sourceof Socio-Political Askari, Maktub History (Patna 1981), pp. 30-64, supplemented at some points by the same author's Islam and Muslims in Medieval Bihar (Patna 1989), passim. For another valuable early fifteenthsee Z. A. Desai, Malfuz Literatureas a century regional and CulturalHistoryof Gujaratand RajasSourceofPolitical, Social malfzi., than (Patna 1991). 7 Nizami, On Historyand Historians of MedievalIndia, pp. 117-78. Cf. idem, Supplementto Elliot and Dowson'sHistory of India, Vol. III, pp. 102-3. 8 Storey, op. cit., p. 1057. 9 For the date of Khwaja Gurg's death, 3 Rajab 700 A.H., see below; Asrdr al-majdhibin, Anecdote no. 210. 10 Formerly in the collection of A. H. Harley, Principal of the Anglo-Arabic Madrasa, Calcutta, circa 1920; now Digby ms. 10. " Though it is not made clear in the text, the date may be related to the death-anniversary on 3 Rajab, see below. The portion of the text which has dropped out of the lithograph, p. 2, line 1, runs from f. 2a of the present ms., line 1, wisdwis-i mukhdlifto folio 2b, line 11, az shanidan-i(in) sukhan-imubdrak. With this omission, it is not surprising that those who have previously examined the lithograph, among them possibly Storey and Nizami, have failed to recognise the date of composition given four lines later, in which the century 700/wa sab'mi'a is garbled to MST'MAYTH. 12 'Abd al-Haqq Dehlavi, Akhbdr al-akhydr,DehlI 1309, dhikr-i ba'iimajddhib, pp. 279-85. The Majdhibs who are described in Uzgani's fantasising Historyof the Uwaysis,translated by Julian Baldick, are not so segregated. Several are described as majdhib from birth. Dr. Baldick renders majdhfibas "enraptured"; see his Imaginary Muslims: the Uwaysi Sufis of Central Asia (London, 1993), pp. 57, 81, 132, 159, 189. " Cf. the enormous single horn on the head of the wolf slain by Bahraim in the "Great Mongol" Shah-ndma dispersed by Demotte; O. Grabar and Sheila Blair, Epic Images and Contemporary History: the Illustrations of the Great Mongol Shahnama (Chicago 1980), Fig. 53 and Colour Plate facing p. 19. See also R. Ettinghausen, Studies in Muslim Iconography:the Unicorn (Washington 1950), pp. 37-43, and Plates 9 (another "Great Mongol" Shah-nanma page), 23, 25 and 26. For the distinction of karg and gurg, cf. the qasida of the thirteenth-century Dehli Sultanate Persian poet Shihab al-Din of Bada'fin, in which by way of artifice every couplet mentions four animals, shir ("lion", or, in the Indian context, often "tiger"), pil (elephant), karg and gurg, see Iqbal Husain, The Early Persian Poets of India [A.H. 421-670] (Patna, 1938), pp. 184-90. One may quote a typical couplet: pil bakhshddar badda'inbdyad-amwirana'1 garchihjay-i gurg u karg u shir bdshadin diydr "Elephant-bestower, I desire a desert place in Badi'ffin, even though these regions are the abodes of the wolf (gurg) and the rhinoceros (karg) and the lion. (I. Husain's translation, op. cit., p. 189).

108

JOURNAL

OF PERSIAN

14 This is not the sole instance from this period in northern

India where ambiguities of Arabic script or of manuscript transmission have led to a corrupt form of a famous name prevailing in popular tradition. Another instance is that of the name of the pious Sultan of Dehli, Shams al-Din Iltutmish ("world-grasper", cf.Jahdngir, which has the same sense), later remembered as Altamish or Altamsh. See S. Digby, "Iltutmish or Iletmish? A reconsideration of the name of the Dehli Sultan", Iran VIII (1970), pp. 57-64. Cf. also the scribal corruption of Sijzi to Sanjari as nisba of Mu'in al-Din Chishti, a form long and firmly established in popular consciousness. 15 Cf. such Indian names as Rdja, Rdwat, etc. borne by Muslims of the early Dehli sultanate. "Sajjan" [usually m., "good man" > "friend" > "lover"] would have been a pen-name (cf. A.P. takhallus). Variously derived from Skt. sant- + jan- or sat- + jan-, the word has been current in this form for two millennia from Pali texts to present-day spoken North Indian vernacular usage; see Turner, CDIAL, 13090, sajjana-2;C. Shackle, A Guru Ndnak Glossary (London 1981), pp. 31-2, 45; R. C. Varma, Saniksipt Hindi sabdasdgar (Kashi, N-gari Pracarini Sabha, 1933), s.v. 16 Cf. note 14 above for Sanjari in place of Sijzi in the nisba of Mu'in al-Din Chishti, also found in that of the poet Amir Hasan Sigzi Dehlavi. The manuscript, written in generally unpointed shikasta, does not vary in the orthography of the name, with an upward curve at the close which is correctly used for nun and not yda'.Mr. Ziyauddin Desai, formerly Arabic and Persian Epigraphist to the Archaeological Survey of India, who read with attention and interest the photocopy of the Ms that I brought to the New Delhi conference, initially suggested "Shaykhan" (an Indianised nickname or pen-name), but later, in view of the consistency of the orthography which lacks a "tooth" after initial sin, agreed that "Sajjan" was the more probable reading. 17 Abii Sa'id is not recorded as a name of Sayyid Jalil al-Din Husayn "Makhdfim-i Jahaniyin" Bukhari, though this great Shaykhwas a Khalifa of Rukn al-Din of Multan, and therefore in the same Sufi lineage as the compiler of the Asrdr almajdhibin. Such an identification therefore remains uncertain but not impossible. need among the growing domiciled 18 On this emotional mediaeval Muslim communities of northern India for a legendary Islamic sanctification of their homeland and the features of its familiar landscape, see Digby 1986 (below, note 21), pp. 58-9. 19 The contrast is a recurrent commonplace in Persian. Steingass notes: "tara dar kih birydnast, Proverb founded on a play of words between tara (tarra), 'herb', and bara (barra), 'lamb', to express that in case of need any food is welcome." (Persian-English dictionary,s.v. tara). 20 I am grateful to Mr. Z. A. Desai (see note 16 above) for this reconstruction of the name, which also appears, corruptly written, in another anecdote (no. 43). Mr. Desai also suggested Mahoba in Hamirpur District, mentioned in proximity to Bhambharoli in a passage discussed below. 21 For ideas and anecdotes of Abdailsin the Dehli sultanate, see S. Digby, "Qalandars and related groups", in Y. Friedmann (ed.), Islam in Asia, Jerusalem 1984, pp. 80-1; S. Digby, "The Sufi Shaykh as a source of authority in medieval India", Purusdrtha IX (Paris 1986), p. 62. One further anecdote indicates suspicion and confrontation (no. 131). A Qalandar brought wine to Gurg, who after sipping it poured it upon the ground. The implication may be that the wine was poisoned or drugged. Gurg then said that he would meet the Qalandar at the prophet's tomb at Medina and the Qalandar fell at Gurg's feet. However, peripatetic Qalandars may be distinguished from settled local Majdhibs. In an anecdote in Asrdr al-majdhubin(no. 9; translated in Digby 1984, p. 98), a group of Qalandars attacked Khwija Gurg while he slept. Other anecdotes show Qalandars and Haydarls (members of

STUDIES

the groups also being called zhinda-posh) consorting with Gurg and showing him respect (nos. 94, 149, 150, 157, 176). In the last of these anecdotes, Khwija Gurg also meets Abddls at the Ka'ba. In another anecdote of Gurg's power of bilocation, Gurg rescues a Qalandar called Shirin Goya ("Sweet singer") from ghouls who are seated beneath a great shady tree on a distant road (Anecdote no. 103). In another anecdote, the same Qalandar is passing time in Gurg's company (no. 153). A Qalandar meeting Gurg addresses him as "Emperor of Abdals ... naked Gurg" (Anecdote no. 147). 22 Cf. Kabir-granthavali,ed. Pairasnath Tiwairi (Prayaig 1961), pada 56, p. 32. Cf. also Kabir Bijak,Sabda12.2, 29.3; ed. Sukdev Sinh (Ilahabad 1972), pp. 115, 122; tr. the Rev. Ahmad Shah (repr. Delhi 1990), pp. 100, 110. 23 Anecdotes of the power of Sufi Shaykhsand of other Indian wonder-workers over gold are considered in a forthcoming study by the present writer. 24 Ms. ff. 49b-50a. 25 Ms. f. 30a. For the Buddha's walking across the Ganges, see Mahdparinibbanasutta,i, 34; Divydvaddna, III, ed. Cowell and Neil, p. 56; J. Ph. Vogel, Indian Serpent Lore, or the Ndgas in Hindu Legendand Art (repr. Varanasi 1972), p. 172. 26 See S. Digby, in Iran XXVII (1990), p. 76; Futiih al-saldt.in, (Madras 1948), p. 229. 27 Ms. f. 57a.

28 The historian Farishta, writing at the end of the sixteenth

century, has an anecdote of Khwija Gurg's "bestowal" of kingship on Sultan 'Ala' al-Din Khalji which does not derive from the Asrdr al-majdhubinor from 'Isaimi. Before the confrontation between uncle and nephew on the boat in the Ganges, 'Ali' al-Din went to pay his respects to "Shaykh KRK the majdhib buried in Kara. The majdhib raised his head [from meditation] and said: har kas kih bikunad bd tu jang sar dar kishti tan dar gang Whosoever makes war upon youHead in the boat, body in the Ganges!" Ta'rikh-iFarishta (Bombay 1831-2), I, 174; ed. Kanpur, Naval Kishor, I. p. 100. Nizaim al-Din Ahmad, Tabaqdt-i Akbari, I, p. 136, merely reproduces the words of Farishta. 29 A similar anecdote of the flying camel standing ready to take Shaykh Nizam al-Din Chishti of Dehli to Mecca is found in an early and reputable source; Mir Khwurd, Siyar al awliyda',p. 143, Lahore repr., p. 153; for Shaykh Sharaf al-Din Ahmad Maneri's presence in Mecca on Fridays, see Mandqibal-asfiyd', p. 142. 30 M. Van Bruinessen, "Hajji Bektash, Sultan Sahak, Shah Mina Sahib and Vavions Avatars of a Running Wall", Turcica,Revue d'Etudes Turques, XXI-XXIII (1991) [Melanges offerts ci Irene Milikoff], pp. 55-69. 31 See J. Ph. Vogel, Indian SerpentLore, passim. 32 See Digby 1986, pp. 62-3. 33 A first draft of this paper was presented at the International Seminar on Sufism organised by the Government of India in New Delhi to commemorate the centenary of Mawlinfi Abu'l Kalaim "Azad"'s birth. I could not rival the lofty and wideranging surveys of Sufi doctrine or of Sufi influence on Indian society boldly offered by other delegates, and instead I drew attention to this surviving hagiographical work which must have seemed less than edifying to my audience in current political circumstances. My paper was submitted late, and I was disconcerted when it was returned from the Secretariat of the Indian Council for Cultural Relations, on the morning of its presentation and distribution, photocopied rather than typed. But I then reflected that this was one of the curious but apt coincidences that so frequently occur in the patterns of Sufi anecdotes. The paper refers to the hazards of manuscript transmission, and it was circulated in reproduced handwritten form.

ANECDOTES

OF A PROVINCIAL

SUFI OF THE DEHLf SULTANATE,

Likewise, the story that I tell in it concerns a missing manuscript leaf, and in the original circulated form there was an error in the enumeration of the pages, whereby p. 11 appeared to be missing. My paper's timing in the course of the Proceedings of the Seminar also had a symbolic value. This paper concerns a majdhisb:and the Indian Sufi biographical tadhkiras sometimes notice this category of devotees in a section towards the end of the work reflecting their lower status, after the notices

KHWAJA GURG OF KARA

109

of grander or more conformist Shaykhs (cf. note 12 above). Hence it was appropriate that my contribution should have been presented at the tail-end of the deliberations of the Seminar. I am grateful to the Convener on behalf of the Government of India, Shri Burney [ = Barani], whose very name evokes a potent shade of the Dehlf Sultanate, and to my Chairperson, Shrimati Kapila Vatsyayan, whose already wide tolerance may have been stretched further by the persona of the man of religion who was my subject.

THE ACCESSION OF ISKANDAR KHAN By Audrey Burton GreenHammerton,York

Iskandar Khan was an Uzbeg prince who acceded to the throne of the khanate of Md war-i' al-nahr in 1561 in rather irregular circumstances. Not only was his accession arranged during the lifetime of his predecessor Pir Muhammad and without the usual formalities, but it seems fairly clear that it contravened the rule of succession by seniority. It should be explained that at that time the ruler of M- ward' al-nahr (the khan or khdqdn)was chosen from among all the members of the Abulkhairid dynasty, i.e. all the descendants of Abu '1 Khair's ten sons, strictly according to seniority, and he remained in power until his death. Once chosen the new ruler was seated on a large square of white felt which was thrice raised into the air by the princes and chiefs of tribes, this being the Mongol equivalent of the anointing ceremonies of Christian monarchs. More far-reaching proof of his accession, however, came when his name was mentioned during the Friday sermon (khutba)and when coins were minted in his name (sikka). If the prince chosen as khan by his peers and the chiefs of the Uzbeg tribes sitting in council (qurultdi)turned out to be incapable of carrying out his duties or leading the armies of the khanate against foreign foes, then this task was undertaken by a younger, abler man, but only if asked to do so by the khan. His brief was carefully defined and he could not expect the princes and generals to follow him if he decided on an independent course of action. Thus in 1536-7 the capable young prince 'Ubaidallih, who headed the army of the khanate on behalf of the khan Kiichkiinji, was unable to persuade his peers that it was worth continuing the siege of Herat and they returned to the khanate.1 Abu '1 Khair himself had come into prominence in the fifteenth century after he united under his command a large number of Turkic, Mongol and Tibetan nomadic tribes camping between the Tobol and Ural rivers. He had not conquered Mi ward' al-nahr, although he did raid some of its cities, for he preferred to live in the steppes. It was his grandson, Muhammad Shaibaini, who conquered the country situated beyond the river Oxus where he then founded the khanate of Ma war-i' alnahr, ruled after him by the eldest representative of the Abulkhairid family.2 Another of Abu '1 Khair's grandsons, Jini Beg,

had twelve sons, and it so happened that after 1556 his sons were senior to the other descendants of Abu '1 Khair. It is among them that we find both Iskandar and PTr Muhammad. The fact the Pir Muhammad was chosen to rule the khanate before Iskandar seems fairly positive proof that he was older, yet in the genealogical table given by DesTanish, maisons, which follows that given by of Iskandar's son Iskandar is .HIfiz 'Abdallah, biographer listed as number five and Pir Muhammad as number eight. How can we explain this? A close study of four Oriental historians of the sixteenth and seventeenth centuries shows that there was some confusion about the order of seniority of Jaini Beg's sons, although there seems to be little doubt that Iskandar was younger than his brothers Kisten Qar- and Rustam. Kisten Qari held the major post of governor of Balkh as early as 1528, having helped to retake this town from the Safawids a few years earlier. He is listed as older than Iskandar, and indeed figures among the three oldest sons of Jani Beg in all the sources checked so far. Both the official historian, Hhfiz Tanish and Muhammad Yar b. 'Arab Qataghdn, who held high office under Iskandar's son 'Abdallah, place him second. The anonymous sixteenth-century author translated by Berezin in his Sheibaniadaplaces him third, and the very reliable seventeenth century historian, MahmUid b. Wall, places him first among Jani Beg's sons.3 As for Rustam, he is always placed before Iskandar in the seniority list. He is given the fourth place among the sons of Jini Beg by both HaIfiz Tanish and Muhammad Yir b. 'Arab Qataghan, and also by the author of the Sheibaniada, but MahmUid b. Wall puts him third. However, they do not enter into the problem surrounding the succession, because they had both died by the time Pir Muhammad came to the throne, and only Pir Muhammad and Sulaimdn appear to have been alive at the time of Iskandar's accession.5 But what about the difference in age between Pir Muhammad and Iskandar? As stated above, the Tanish places Pir Muhammad eighth, as does .H1fiz author of the Sheibaniada,making him thus younger than Iskandar, but Muhammad Yar places him third and Iskandar seventh, while Mahmuid b. Wali

111

112

JOURNAL

OF PERSIAN

places him second and relegates Iskandar to the that Pir is no doubt twelfth place. There Muhammad was chosen to rule the khanate in 1556, and since the order of succession adopted in the khanate was based on seniority, a fact recorded khan supreme chez by Bacque-Grammont-"(le) les Abu'lkhayrides ... etant en principe l'aine de la famille, dans quelque branche que cela soit, elu par mentioned as one of un conseil dynastique"-and the weaknesses of the regime by Dickson, and since Iskandar did not come to the throne until several years later, this must indicate that Pir Muhammad Tanish makes it clear was older.6 In any case, HIaHfiz that Pir Muhammad was elected because "there was no one older among the descendants of Abu '1 Khair, (nor was there any one else) who could of the country or show handle the government such exclusive zeal in (matters of) greatness and fame. Therefore the sultans agreed among themselves and proclaimed him".7 Some years earlier, and perhaps for the same reason, it was Pir Muhammad, and not Iskandar, who was appointed to take over Balkh when Kisten Qara died in 1547. However, the fact that Pir Muhammad was appointed to Balkh is not entirely conclusive, for it could be that he was simply considered more competent and capable than Iskandar, and there is no doubt that Iskandar was held by some, such as Abu 'l-Ghazi, to be a simpleton, more or in praying than in government interested matters of everyday life.8 But to return to the list of Jani Beg's children given by Hafiz Tanish, how could there be any room for error in a list that begins with the uncompromising statement: aulddjadni Big ... ba in tartib, i.e. "the children of Jani Beg in this order"? And how could the historian say that Iskandar, his patron's father, was the fifth son and yet explain, without any semblance of embarrassment, that Pir Muhammad was elected as ruler several years earlier because he was the oldest of the princes? The answer to this mystery seems to lie in the fact that HI-fiz Tanish listsJani Beg's sons in groups as per their mothers. Iskandar is one of the four She was the sons of the second one mentioned. daughter of the famous Mahmiid Khan of Kashghariya, and the eldest of her sons was Rustam, whom HI-fiz Tanish places fourth amongJani Beg's sons. The three princes who preceded Rustam were the sons of an equally well connected lady. She was the niece of 'Abd al-'Ali tarkhan,9 a man who ruled area in the late Bukhara and the surrounding fifteenth century and who was famous for the wealth of his court, the banquets he gave there and his enormous military guard of 3,000 men. Clearly, it is because of their distinguished parentage that Hlfiz Tanish did not think it necessary to alter their place in the family tree and make them follow

STUDIES

the group which included the father of his patron, although he might have been tempted to do so in order to pay a flattering tribute to his patron, 'Abdallah II. Pir Muhammad's mother, on the other hand, was only the daughter of "one of the Mongol amirs" and that probably explains why he and his brother Shah Muhammad were placed after the descenas was Yar dants of better-known ancestors, Muhammad, whose mother was also the daughter of "one of the Mongol amirs",10 and another two sons of Jani Beg's, whose mother was only a "hostage", i.e. a captive presumably the daughter of a conquered chief, and therefore a person of no moment. Was Hafiz Tanish therefore guilty of trying to confuse the reader when he stated categorically that he was giving the list of Jani Beg's sons "in order"? Not at all. He just thought that his readers would be more interested to learn first about those of distinguished parentage. In any case, they would have known that Pir Muhammad must have been chosen first because this was done, as in the case of Kiichkiinji, "in accordance with old laws and traditions", that is to say, following the principle of seniority.loa

Another point that is worth remembering is that Muslim rulers often had several wives or concubines at one and the same time. Therefore the fact that Jani Beg had four sons by a particular wife does not necessarily mean that they were born in quick succession and were close together in age. In fact, the author of the Sheibaniada, who attributes the same four sons as HIIfiz Tanish to Jani Beg's union with the daughter of Mahmild Khan, places the first three, including Iskandar, as numbers four, five and six, but enters the fourth one, Sulaimin, as number nine." Muhammad Yar, on the other hand, places Sulaiman straight after Rustam at number five, and perhaps it is because he wrote his history some six years after the death in 1598 of both Iskandar's son and grandson that he felt it safe to relegate Iskandar to number seven. Mahmiid b. Wall, who began his history even later, in 1630, gives a totally different order for the sons of J-ni Beg. He places Iskandar twelfth, gives Sulaiman the sixth place, and puts Pir Muhammad in a far more important position as second son. HaIfiz Tanish places Sulaiman after Iskandar at number seven, yet it is interesting to note that in 1567, when Pir Muhammad unexpectedly died, took advantage of the fact that both Sulaimin Iskandar's son 'Abdallah and Pir Muhammad's son Din Muhammad were campaigning in Khurasan to seize Balkh. He even issued coins in his own name as khaqan, which seems to imply that he was in fact older than Iskandar, and therefore had a good claim to the throne, although he gave this up when

THE

ACCESSION

OF ISKANDAR

KHAN

113

and his son caused him no difficulty, and even returned from 'Abdallah and Din Muhammad three sons of a former khaqan from Tashkent came neither Khurasan. Since he could persuade over to pay homage in 1564.'15 Only after Pir 'Abdallah to agree with his claim to the throne, nor Muhammad died in 1567 was Iskandar's rule dispersuade Din Muhammad to let him rule Balkh, he of the puted, the first challenge coming from his brother accepted from Iskandar the governorship town of Dabusi near Samarqand, and subsequently Sulaiman, as mentioned above. which were far more The other challenges, of Iskandar be a both to loyal supporter proved and 'Abdallah.12 serious, came from younger men and were really aimed at Iskandar's son 'Abdallfih who was known One point has yet to be cleared up: how could to be the defacto ruler of the khanate and who, it Iskandar replace Pir Muhammad in his lifetime, was feared, might eventually succeed his father as and without going through the usual form of selecde jure ruler. These challenges to Iskandar came tion by the qurultai? This only came about because from the sons of more than one former khaqan and Iskandar's capable and ambitious son, 'Abdallah, took fifteen years to resolve, by which time most of decided that Pir Muhammad had shown himself and their appanages his them had been eliminated of the khanate and enforcing ruling incapable had been annexed by the khaqan. A year later, in decisions. In 1561 he had asked 'Abdallah to let him have the town of Bukhara in exchange for 1583, Iskandar died and 'Abdallfih had himself Balkh,'3 but when 'Abdallah, who had agreed to recognised as ruler, although his brothers and his the exchange, prepared to carry it through, Pir cousin Uzbeg Sultan were older than him.16 The new principle governing the succession was that Muhammad's own son, Din Muhammad, refused to Balkh. was the ruler should be the prince most capable of prothe and attacked 'Abdallah deal accept and decided that Pir tecting the Muslim law. This was spelt out by annoyed understandably 'Abdallah's Muhammad would have to go. He let his uncle powerful supporter, Khwaja Sa'd of know what he was proposing to do, and it seems Jiiibari, leader of the major Siiff brotherhood in the following words "he that the grandees of Bukhara agreed that PTr the Naqshbandiyya, who gets (who should get) overall power is he should be replaced Muhammad by Iskandar. of religion are as Indeed, when 'Abdallfh's brother, 'Ibadallah, was whose justice and encouragement sent to Karmina, near Samarqand, to explain to clear as sunlight". This principle would in due Iskandar what was expected of him, he was ac- course enable 'Abdallah's son and rival, 'Abd alMu'min, to follow his father on the throne in 1598 companied by a number of Bukharan grandees. and to ignore the seniority claims of his father's And when Iskandar came over with them, the brother and cousins. However, the principle of whole population of Bukhara apparently went out in a modified him. Iskandar's to welcome name was then seniority would be re-introduced, in the khutba and the sikka "and he mentioned form, about sixteen months later, by the dynasty which took over Ma war-a' al-nahr after the death of ruled from Kasghar to Khurasan ... from Dasht-i the last two Abulkhairid rulers, and which conto the outskirts of Qandahar".14 Qipchaq tinued to rule until the nineteenth century. His rule was recognised throughout the khanate during Pir Muhammad's lifetime. Pir Muhammad

I Iskandar Beg,

p. 65. The Abulkhairids continued to rule the khanate until 1599. They are often called Shaibanid, but this is strictly speaking incorrect, as none of them was a direct descendant of Muhammad Shaibani. 3 Hafiz Tanish b. Mir Muhammad al-Bukhril Sharafndama-yi shdhi also known as 'Abdalldhnama, BI Or. 3497, f. 21b. Muhammad Yar b. 'Arab Qataghin, Musakhkhir al-bildd, LOIVAN C 465, ff. 108-17b also known as ShaibdnZ-nimain IVAN Uzb. 1505, ff. 74b-83a. Berezin I.N., Bibliotekavostochnykh istorikov. Tom I Sheibaniada (Kazan, 1849), p. LXXVIII. This list of Jhni Beg's sons is the second one found in this work, the first one, on page LI, being rather different. If the same writer was responsible for the whole work, as believed by Berezin, and if the latter part was added after an interval of some years before the end of the sixteenth century, then it seems likely that the second list was a revised version and based on more extensive information. N.B. Kisten Qara is listed second on page LI. For Mahmlid b. Amir Wall, XV see excerpts translated in Materialy po istorii kazakhskikhkhanstv XV-XVIII vekov (izvlecheniyaiz persidskikhi tyurskikhsochinenii), (Alma-Ata, 1969), pp. 356-7.

2

4 J. L. Bacque-Grammont, "Une liste ottomane de princes et d'appanages Abu'l-khayrides", in Cahiers du monde russe et soviitique, vol. XI/3 (1970), pp. 423-53 at p. 438. Kisten Qarai is said here to have died around 1547 or 1548. Salahetdinova in Hafiz-i Tanish ibn Mir MuhammadBukhari "Sharaf-namayii shakhi" chast' 2 (Moskva, 1989), n. 278, says he ruled Balkh from 1526 to 1544, which implies that he may have died in that year. As for Rustam, he was killed in battle near Qarshi in 961/1553-4. HIIfiz Tanish, f. 43b. 5 The other brothers are not mentioned by Hafiz Tanish except in his original list of the sons of Jaini Beg. 6 Bacqu6-Grammont, p. 423. M. B. Dickson, Shah Tahmasband the Uzbeks (the duel for Khorasan with 'Ubayd Khan 930-46/1524-40), unpublished Ph.D. thesis (Princeton, 1958) pp. 25, 28. H HIfiz Tanish, ff. 49b-50a. 8 Abu '1 Ghiizi Bahaidur Khain, Shajara-yiTurk, ed. and tr. P.I. Desmaisons as Histoire des Mogols et des Tatarespar Aboul-Ghdzi Bihadour Khan, (St. P6tersbourg, 1871-2), tr. p. 193. 9 Hafiz Tani-sh, f. 21b; Berezin, pp. LI and LXXIX. Here she is listed first as 'Ali tarkhan's daughter and later as his sister. 10 Hafiz Tanish, f. 21b, gives his name as Sultain Mahmiid

114

JOURNAL

OF PERSIAN

Tanbal. This is slightly different in the version translated by Salahetdinova, the first name being Muhammad instead of Mahmufd. The second name seems identical, but Salahetdinova translates it as Tanbala. In Berezin, p. LI, his name is translated as Muhammad Tibli. 10aSalahetdinova, chast' 1, p. 87. 1 Berezin, p. LXXVIII. Here Iskandar is listed as number five, but in the earlier list (p. LI) Iskandar and Sulaiman were respectively number six and number seven. 12 HIfiz Tanish, ff. 69a-71a. Davidovich E. A., "Serebryanye monety udel'nykh vladetelei kak istochnik po istorii Srednei Azii XVI v.", in Pismennye pamyatniki vostoka. Istorikofilologicheskie issledovanniya,vyp. 6, 1973 (Moskva, 1979), pp. 55-101, at p. 82. " For 'Abdallih's involvement with Bukhara from 1550 and his uncle's preference for this town, see J. A. Burton, Bukharans in trade and diplomacy,1558-1702, unpublished Ph.D. thesis

14

s

16

STUDIES

(Manchester University, 1986), pp. 31-4. Hafiz Tanish, ff. 63b-64b. Ibid., ff. 67a-b. Mahmuid b. Wall, p. 357, says 'Abdallah was the youngest of his brothers. HaIfizTanish, f. 21b, makes 'Abdallah to be the eldest, but this could well be a courtier's way of expressing 'Abdallah's pre-eminence over his brothers. Muhammad Yar, LOIVAN, f. 119a, also lists 'Abdallah first, although he goes on to give a potted history of each of Iskandar's sons, beginning with 'Ibadallah and 'Abd al-Quddiis and only_then dealing with 'Abdallah. Dickson, p. 28, explains that Uzbeg Sultan was a few months older than 'Abdallah. That was why, even when 'Abdallah was "already a dominant Uzbeg and the leader of the Janibegids", he had to follow the old traditions and show respect by going up to Uzbeg Sultain in order to offer congratulations after a particular victory.

THE IJAZAFROM YUSUF AL-BAHRANI (d. 1186/1772) TO SAYYID MUHAMMAD MAHDI BAHR AL-'ULUM (d. 1212/1797-8) By Robert Gleave Universityof Manchester

INTRODUCTION

existed as merely parts of other manuscripts; the naming of the work(s) within the text of the ijaza was deemed sufficient. If this progression from oral pronouncements to written marginal notes and then to separate stylised document is accepted, then the ijaza from Bahr-ni to Bahr al-'Ulum would appear as an example of the last stages in this development. The and analysis of the ijaza following description demonstrates that Bathrni's motives for writing the ijaza are broader than the conveyance of a permission to transmit certain scholarly works. The ijaza functions as a means whereby the "scholarly tradition" in its widest sense is ensured as in the next continuing generation. The ijaza is probably a prototype for Bahrani's biographical work Lu'lu'at al-bahrayn,9 which has been used by many scholars as a primary source for Shi'i history in general and for eighteenth-century Shi'ism in particular.10 This work, though its contents place the work within the genre of biography (tabaqat), has the external form of an ijaza.

In the manuscript section of the Ayatallah Mar'ashi Libr'ry in Qum, Iran, there is a collection of ijazas (manuscript number 5605) by various Shi'i clerics.1 One of these (found on leaves 15b to 22b) is a copy of the ijzza from Yuisuf b. Ahmad alto Sayyid Muhammad Bahraini (d. 1186/1772) Mahdi b. Murtadai Bahr al-'Ulim (d. 1212/1797-8). Nineteenth-century Shi'i biographical works, such as the Qisas al-'ulama' by Muhammad Baqir Tunakabaini (d. 1302/1884)2 and Rawdat aljannat by Muhammad Baqir Khwainsfiri (d. 1313/1895)3 contain much information about both scholars. is described as a member of the Akhbairi Bahrdini school of jurisprudence (fiqh), whose principle doctrine seemed to be the rejection of the interpretative practice of ijtihad. According to these works however he was not an extremist (mutaassib) or pure (sirf) AkhbirL.4 The biographical entries on Bahr al-'Uliim state that he was a student of in Kerbala, before moving to Najaf to Bahra.ni under Muhammad Baiqir al-Bihbahani (d. study 1207/1792). Under Bihbahaini he learned Usiili or which in opposition to the ijtihad-i jurisprudence, THEIJAZA Akhba-ri school endorsed ijtihad as a legitimate exeto getical practice (one practice ualified ijtihaid being called mujtahid). The ijtza can be divided into five sections: In what follows, I argue that the manuscript of (i) Introductory prayers the ijatza between these two scholars represents a (ii) An introduction to the muftz, the mujaiz and the late stage in the development of the ijazza system. purpose of the ijaza An ijzza was a "permission to transmit" given by a (iii) A list of thirty-five isn5ds scholar (the mufTz). The recipient of the ij~iza (the (iv) A list of nineteen works for which the ijfiza is, mujaz) was permitted to transmit (riwaiya) from the in part, given. mufjz a specified work or works (termed the matn of (v) Closing prayers and praise for the mujaiz. the ijiiza).6 Studies of the ijaza system seem to indicate that an ijaza in earliest times was an oral pronouncement by the mujfz specifically using the Sections (i) and (ii): Introductory Prayers and the terms ajaztu la-hu an yarwi 'anni ("I permit him to Function of the Ija-za transmit from me .. .",,).7 As scholarship became more structured and an important written element The introductory prayers, and opening remarks emerged within the system of learning, the ijiiza on the mujiz, mujaz and the purposes of the ijaza are was written in the margins or at the beginning of a written in florid rhyming prose. Phrases are often manuscript of the work in question. In this written grouped in tautologous pairs for the sake of maincontext, ijizas began to take on a formal structure. taining rhythm and rhyme, rather than conveying Goldziher argued that, later, ijzzas became separate meaning. What follows is a paraphrase of sections documents with standard formulae incorporating (i) and (ii), which omits these tautologies. It also both verse and prose styles.8 They no longer omits the extensive prayers which follow every 115

116 mention of God, the Prophet, individual scholars.

JOURNAL

the Imams

OF PERSIAN

and

Section (i) 1. Praise God who has guided us upon the way of the true faith; who has strengthened us upon the straight path; who has secured for us success in obeying the rule of the Blessed Imams, which is the one thing that will secure happiness in this world and the next and is a sure rope that will save us from his just punishment; who has kept us in line with those who transmit the sayings of the Imams (akhbar), by which this life and the next are governed, and in line with those who relate these sayings (athar)preserved by a chain from them [the Imams], from their ancestor [the Prophet], from Gabriel, from God. 2. Prayers and peace be upon the one who carries out his command and vigorously spreads the message of the divine purpose; and upon his family who followed him in what he said and in his great actions, both in public and private; and who, by his grace, taste the bitter drink of misfortune, patiently and thankfully. Conforming to the usual convention, this passage is divided into two sections, prayers to God and prayers to the Prophet and Imams. The tone of of the Imams these prayers and the inclusion locates the text within the Shi'i tradition. Section (ii) 1. And so (amma ba'du), Yuisuf al-Bahrani says that, of all the sciences, the study of hadith ('ilm al-hadith) is my outer and inner garment, and I occupy myself with it night and day, to the point that I know all about it. It has become as clear as the mid-day sun that there is nothing in this world that can equal it; it is the principle which governs the two worlds and the two risings. Any science that opposes it must either be based upon it, in which case it is true, or it is like scattered dust, even if this is denied by those who are accustomed to following famous [opinion/scholars?] (taqlid kulli mashhiLr). 2. Muhammad Mahdi Tabataba'i is someone who has taken from ['ilm al-hadith]a great deal and has been successful in this. He asked for an ijiza for everything which I consider sound from those trustworthy sayings; and for the paths I trace to the great teachers (mashayikh).He was worthy of this, and when I was sure of his trustworthiness and piety, I answered his request quickly and accepted his demand. 3. I begin: There is no doubt that an ijzza is appropriate in these times; even if it is of little use for confirming the akhbir and proving the paths which stem from the illustrious scholars, [it is of little use] because they are so famous that they cannot be denied, particularly the four books which are as clear as the sun in every land." Since our scholarly predecessors have done this [i.e. given ijizas], seeking good fortune and blessing in connecting this noble and holy chain back to the Imams, we shall follow their example, agreeing with what they agree and rejecting what they reject. 4. I give him an ijaza (ajaztu la-hu) to: 4.1 transmit from me all transmissions (riwaya) which I consider sound and all learning (diraya) which I consider sure, consisting of the books of akhbar,other

STUDIES

books from the 'ulama' and also the books of the Sunnis (mukhalifin) [written in] every time and place; for it is a full and complete ijaza; 4.2 and to transmit everything I have written in the form of books, treatise, commentaries, additional comments and answers to questions. My paths (turuqi) to those books are numerous and widespread, reaching everywhere. I will mention only that which is easy [to understand], for, as the saying goes, easy things are not defeated by complexities. Three notable themes may be detected running of 'ilm althrough this section: the importance hadith; the general nature of the ijizza; and the purpose of ij*za as the maintenance of the tradition of past 'ulama'. is exemplified Bahrani's epistemology by the unassailable position he gives to 'ilm al-hadith and a general emphasis on the sanctity of the akhbar (the must be sayings of the Imams). All knowledge based on 'ilm al4hadith, and Bahrani considers himself particularly learned in this most important science. The books of the akhbar are so well known and established that they cannot be denied (ila haddin la-yaqbalu al-inkar), and it is for these books of akhbair that the ijiza was originally requested. Other works, including Bahrani's are seemingly given only secondary importance as the matn of the ijzza (4.2). A second theme is the general nature of the ijzza. It is not uncommon for an ijaiza to be given for a of works to a number of people.12 number However, the general terms used to describe the matn indicate that what is conferred is an honorary position within the scholarly class as well as a permission to transmit certain texts. Bahr al-'Uluim is permitted to transmit the books of akhbar as well as the other books of the 'ulama'. The ijUazais even given for all works by Sunni scholars (mukhalifin). Finally, Bahrani gives an ijdza to Bahr al-'Uluim to transmit everything that he himself has written (ma jar- min qalami). The generality and non-specific nature of these categories, found in this section and repeated later in the text,13 is a marked difference between this ijiza and those described by Vajda.14 The ijzzas described by Vajda are always for specified works. The number of works may be large, but the works are always named. The subsumptive terms used to describe the matn of this ij~iza seem to indicate that this ijzazawas primarily to confer the general approval of Bahr~ni for Bahr al'Ulum. Within the developed system, it is to be expected that Bahr al-'Uluim would require additional ijfzas from other scholars. Finally, the whole purpose of the ijaza is stated explicitly by Bahrani (para. 3). Ijfizas generally, and the isnids that accompany them, are not necessary

THE

IJAZA

FROM

YUSUF

AL-BAHRANI

TO

SAYYID

to establish the reliability of the akhbar of the Imams. These are beyond doubt. Bahrani has produced this ijzza, instead, to maintain the tradition of the 'ulama' of the past. They produced ijazas "seeking good fortune and blessing" (tabarrukan wa tayammunan) in connecting this noble chain back to the Imams (ashab al-'isma). The maintenance of the chain is a task undertaken by each generation for which they seek blessing and reward. This ijaza was given, not to validate the akhbar, but to extend the silsila or mu'an'ana (chain) of the 'ulama-'. Bahr al'Ulum is conceived as the next link in this chain which Bahrani traces through thirty-five isnads in the next section. It might be argued that the ijaza system envisaged by Bahrani here is primarily a celebration of scholarly history and only secondarily a means whereby works are diachronically transmitted. Section (iii): The Isnad Network From 16b.6 to 21b.6 there is a list of thirty-five isnad chains. An isnid, in its most characteristic technical use, is the part of a hadith which aims to secure the reliability of the story or saying being related (matn'5). It is a list of names of individuals stretching back over a number of generations, each name separated by the preposition 'an or some other appropriate formula denoting transmission (i.e. A 'an B signifies that A heard the related story from B). Whether the hadith is to be considered ("sound, reliable") or da'if ("weak, unrelisahi.h able") is, in part at least, dependent upon the integrity of the isna~d.Every person listed must fulfil in certain criteria, such as being trustworthy to and in character, having lived contemporaneous the same geographical area as (or having visited the area of) the people whose names form the links on either side in the isnad. The isnaids listed in Bahrani7's ijfza, unlike those which are found in hadith collections, have no explicit matn, though their probative function is equally clear. By these isnaids, the writer intends to secure his position within the fabric of Shi'a tradition. The isnads stretch back over 900 years. Within these isnafds, Bahrani succeeds in constructing a network of about 160 individuals reaching back to early Shi'a figures. The chains terminate with Shaykh alSadiiq Muhammad b. 'All b. BTibuiya (d. 3811991) and Muhammad b. Ya'qilb al-Kulayni (d. 328/940-1 or 3291941-2).16 Each of the isnaids begins or ends with the name of an 'ilim who has already been mentioned in a previous isnid, thus ensuring the integral nature of the chains. Though new names are introduced midisndd, as it were, an isnad is never opened with an unfamiliar figure. Often an isniid chain ends with what I will call a major figure-that is, one who is

MUHAMMAD

MAHDI

BAHR

AL-

ULUM

117

mentioned repeatedly in many isniids, who is given extensive honorific titles and who regularly begins and ends isnads. There is a collection of formulae that Bahrani uses to guide the reader through this web of Shi'a history: 1. In the text, each chain is separated by an isolated letter hae'.17 This isolated hai' in the current copy clearly stands for the word haylula (meaning break or separation). HIaylila is also used by Bahrani in the Lu'lu'at al-Bahrayn to indicate isnad breaks.'8 As has already been mentioned, the substantial structural and similarities between this ijiza and the Lu'lu'a indicate that this ijiiza was either a prototype by BahranT of the Lu'lu'a. The ha-' is usually followed by wa 'an ... (and from .. .). In only three places is this pattern disrupted: at 12.20a, it is followed by wa bi 'l-isndd almenmutaqaddim 'an ("and by the previously tioned isna-d from ..."); at 11.20b, wa bi 'l-isna-d ilai ("and by the isniid up to") and at 5.20a, wa-mimma dhakarna min al-asainid al-mutaqaddima ... ("and from the previous isnaids we have mentioned ... ). The full isnad break marker and introduction of a new isnad should be considered hae' wa ... (i.e. haylfila wa ...). 2. 'an 'iddat al-mashayikh minhum ... wa minhum ... wa kulluhum 'an ... ("from numerous Shaykhs including ... and ... and all of them relate from ... ). This formula is used when an isniid "fans out" and an individual from one generation is connected with a number of scholars in the who all relate from one previous generation, particular scholar, one generation further back. 3. The phrases al-mutaqaddim (possibly short for the common formula, al-mutaqqadim dhikruhu) or almadhkuir are used to indicate a figure who has mentioned already appeared in a previously chain. They are often used when describing the first name in an isna-d, directing the reader to where the name was introduced, and assuring the reader that this isnad is connected with the established network. 4. ilae akhir mae taqaddama: This phrase is employed when the writer wishes to add to the isniid a list of individuals from a previous isnad (i.e. an isnaed is elongated and reaches further back into history without the need to repeat a list of names since an identical list has already been included). The manner which these formulae are utilised and their importance as tools for the construction of an integrated network of isnads is best demonstrated by an example. Four related isnads run from 16b.10-17a.2, 17a.5-17.11 and 17a.15-17a.17. The 17a.2-17a.5, chains are displayed diagramatically in Figs. 1-4.

118

JOURNAL

OF PERSIAN

STUDIES

'Abdallih b. 'All Bilddi

Shaykh Ahmad al-Madhkiir

Sulaymin b. 'Abdallh

Muhammad Mu'mfin al-Husayni al-Astardibidi

Sulaymin b. 'All

Sayyid Nfir al-Din

'All b. Sulayman

Muhammad Shlib al-Madirik and Husayn Sihib al-Ma'alim

Husayn b. 'Abd al-Samad

'Idda min al-fudala' minhum 'All b Abi Husayn wa-Husayn b 'Abd al-Samad wa-Fakhral-Din Hdshimi

al-Shahid al-Thani (Shahid II) Fig. 1 16b.10 to 17a.2

Kulluhum min

al-Shahid al-Th~ni Fig. 3 17a.5 to 17a.11

Sulaymin b 'Ali al-mutaqaddim

Sulaymin b. 'Abdallih

Ahmad b. Muhammad al-Muqabi al-Bahr~ini

Ja'far b Kamil al-Din wa-Silih 'Abd al-Karim

His father

'All b. SulaymIn Fig. 2 17a.1 - 17a.5

These isna-ds then can be worked together to form a composite picture which is represented in Fig. 5. This example fits into the overall network as

Nfir al-Din b 'Ali (al-mutaqaddim)

ila akhirima taqaddama Fig. 4 17a.15 to 17a.17

one of the many links between Bahrani and Shahid II. 'Abdallahb. 'All Biladi is located by Bahrani as one of his teachers (the other being Husayn b. Muhammadb. Ja'far).

THE

IJAZA

FROM

YUSUF

AL-BAHRAN

I TO SAYYID

MAHD

MUHAMMAD

r BAHR

AL-cULUM

'Abdallih b. 'Ali Bilidi

Sulaymin b. 'Abdalldh

Ahmad b. Muhammad al-Bahraini

Sulaymin b. 'All Muhammad Mu'min al-Astaribadi His father

'All b Sulaymin

Ja'far b Kamil al-Din

Silih b 'Abd al-Karim

Nfir al-Din Muhammad al-'Amili Sihib al-Madirik

S hib al-Ma'ilim

'idda

HIusaynb 'Abd al-Samad

'All b. Abi Hasan

SHAHID II

Fig. 5

Fakhr al-Din

119

Sahib al-Ma'dlim

x

x

Yiusufal-Bahrani Husayn b MuhammadJa'far

M. Baqir Majlis

'Abd Allah Bilidi

x

x

x

xMull

al-Hurr

x Musin xMuhsin

x

Yuisufal-Qatiff Fakhr al-in

Nasir al-Din al-Tasi

Fig. 6 (M.replaces Muhammadx

x

x

M. Taqi Majlis

Karaki

x

x Shahid II x

x Shahid I

X

x

7 pupils of 'Allama

x X

Kulayni

x

x

x

Shaykh al-Bah'

M.Amin al-Astarabadi

Rashid al-Din Abi 'Ali Hasan

Mufid

Shaykh al-Ta'ifa

X

al-'Allama al-Hilli

Hasan b. Shahid II

Sahib al-Madarik

'Abd al-Samad

'Ali Mazidi x

Ibn Ma'iyya x

x x

x x 'Abd al-Jabbar M b 'Ali Karajiki Murtada Ibn Babiiya

THE

IJAZA

FROM

YUJSUF

AL-BAHRANI

TO

SAYYID

The thirty-five isnad chains fit together in a similar manner to produce the network depicted in Fig. 6. The major nodal points are unsurprising. They are established important figures in Shi'i scholarship. The figures mentioned towards the end of the network are the major figures of early Shi'i b. jurisprudence: Shaykh al-Ta'ifa Muhammad Murtadd Hasan al-Tusi (460/1067); al-Shaykh (436/1044); Shaykh Mufifd (413/1022); Ibn BThbuiya, Shaykh al-Sadiiq (381/991); and Muhammad b. Ya'qib Kulayni (328/940-1). The network as a whole serves to integrate Bahr-ni into the mainstream of Shi'i history. By tracing these paths (turuq) from himself, back to the earliest scholars, he affirms his membership of the "saved sect" (alfirqa al-nnajiya)and identifies himself with their achievements. Furthermore, he is qualified to give this ijaza to Bahr al-'Uluim because of his ability to trace these isnads. These "paths" make Bahr-ini a legitimate mujiz and the ijaza a valid ijiiaza. The nature of the connections between these 160 scholars is never explicitly stated. Most chains However, it is posbegin wa-'an ... ("and from..."). of the sible that Babhrani saw the mentioning nature of the connection (qirai'a, sama', ijaiza etc.) as unnecessary because he has described the first link (that is, from Bahrani himself to his teachers) in detail: Shaykh Husayn b. Muhammad al-Mah-izi told me (akhbarani), by qirda'a, sama' and ijaza; and Shaykh 'Abdallah b. 'All Bilddi relates to me by sama' and zijza from ('an).... Qira'a, sama' and zijaza are categories used in of texts. both 'ilm al-hadith and the transmission The relative reliability of these modes of transmission differs between author; however, they are all generally considered reliable modes, as long as a number of conditions are reached.'9 Qirai'a 'alarefers to the process whereby a text is copied by an then reads the text The individual individual. aloud, in the presence of one who has the authority to judge whether it is an accurate copy. This authority is based upon being the author of the text, or one who is part of a sound chain (isnad) which leads to the author. Sama' refers to the process whereby the author, or one with the requisite authority, dictates the text whilst one or more people commit it to writing. The copies, generally are not then verified. Ijaza as a mode of transmission has been mentioned earlier. Bahrani's use of these terms as a description of the first link in the network clearly implies that all subsequent links are deemed to be one of these types. The isnad network is a literary construct. The such as author is writing with set objectives, tracing the network back to the earliest scholars,

MUHAMMAD

MAHD

BAHR

AL-

ULUJM

121

integrating himself and Bahr al-'Uliim into the mainstream tradition, etc. To what extent this ijaza, or documents like it, are useful in reconstructing the history of scholarship is not calculable.20 As a however, three features are literary construct, apparent: the inclusivity of the network; the avoidance of khabar wahid; and the occasional incorporation of biographical detail. This network of 160 scholars is impressive by its The well-known and breadth. universallyrecognised (amongst Shi'a) scholars are included as major nodal points. In addition, a large number of lesser-known and more contentious scholars are included. In places Bahrlini explicitly states that he disagrees with a particular person; yet the person is still included in the network. The aim here appears to be to demonstrate the breadth of Shi'a tradition and to accept a certain amount of difference of opinion (ikhtilaf) within it. The motive is perhaps eirenic, intending to affirm Shi'a community unity through relating a shared history. An interesting aspect of this inclusivity is Balhr-ni's incorporation of scholars from Bahrain. Bahrani, himself born and trained in the Bahrain area, attempts to raise the region's academic by including profile For scholars in each major period. Bahraini whom is to (reference example, Raishid al-Bahrlani three generations found 21a.7-8) is positioned from the Shaykh al-Ta'ifa. Rashid al-Bahr-ni is not included in any major Shi'i biographical work. The second feature is the avoidance of al-khabar al-wa-hid within the network as a whole. A khabar wahid hadith is one which has reduced reliability because, at one point in all the isnaids which support it, there is only one link from whom all the other chains relate.21 The result is that a hadith's upon this one reliability is entirely dependent is Even if this just or trustworthy, person person. the hadith is considered potentially dubious. As can be seen from the diagram of the network, there are major nodal points; however, there is never a total reliance on one person, and there are always alternative routes. By successfully avoiding khabar wah.id type linkages in the isnads, Balhrani avoids unreliability in his network. Finally, within the text as a whole, Bahr-ni occaan isnad chain to include sionally interrupts biographical detail about a person. This happens only four times in a network of 160 names, and the principle behind including this material (if there is one) is not clear. A possible explanation would be that these figures are particularly controversial and For example, in need of further justification. Rishid al-Bahrani's name is followed by the following detail: This shaykh was a jurist, a writer of adab, a linguist; he transmitted via qira'a from the Shaykhs of Iraq, staying there for some time. His tomb, famous in the peninsula

122

JOURNAL

OF PERSIAN

STUDIES

nephews. It maintains the same structure outlined above as sections one to five. Since the Lu'lu'a is principally a biographical work, the isnad section is expanded by the inclusion of biographical details (place and date of birth and death, character traits and lists of works) of 130 scholars who make up the isnads. As noted above, the ijzza to Bahr al-'Ulfim incorporates some biographical material but only of The greater amount on four individuals. biographical detail in the Lu'lu'a is the principal difference between the two works. the Apart from these structural similarities, tioned are MuhammadJumhiirl al- Ahsa'l (located formulae haylula, al-mutaqaddim and ila- akihir ma at 19a.16-18) and the Mu'tazili scholar 'Abd al- taqaddama are found in both works. Furthermore, Jabbdr b. 'Ali al-Raizi (located 21a.9-10). The the biographical notes given in the ijdza are former is a Bahraini scholar and the later is reproduced in the Lu'lu'a, either with similar described as a member of the ashafbal-ra'y ("the phrases or occasionally with exactly the same adherents of opinion"). Both might be described working. The most striking example of this is b. Hasan the entries concerning Muhammad unusual, perhaps weak, links in an isnad. Section (iii), the isnaidnetwork, ends with a state- Jumhiiri Al-Ahs'Ti. Except for one word they are ment that these isnads are well known (mashhuir) identical: of the Prophet, is situated with the tomb of Ahmad alBahrani (who has already been mentioned). (21a.7-8) Being a Bahraini scholar, who is also not included in previous biographical works, Rashid alBahrani's inclusion may be in need of explanation. A similar example is that of Muhammad Amin alAstarabadi, who is criticised in latter biographical works as having founded the maligned Akhbari school of fiqh (biographical details are located at 19a.4-8). Perhaps, inclusion of such controversial or lesser-known figures could not pass without The other two figures mensome explanation.

and can be easily found in works such as the ijazas From Muhammad b. Hasan b. 'Ali b. Ibrahim b. 'Ali of 'Allama, Shaykh Hasan (ibn Shahid good man, a theologian, he had a al-Thn-ni?) Jumhir- al-AhsaS'i-a and MuhammadBaiqirMajlisi.22 strong friendship with the previously mentioned Sayyid

Section(iv) TheList of Works From 21b.9 to 22a.17, Bahraini lists nineteen worksfor which this ijazawas given. The ajaztula-hu formula is repeated. The first five books are followed by a precise account of the contents. For example: The book al-Shihabal-thaqibon the meaning of the term and what is mentioned about it in the sciences: alnamsib23 it is a good [book] which includes noble proofs [derived from] the holy akhbar (21b.9 to 21b.11).

The remaining books are merely listed with no appended comments. A complete list of the books mentioned is:

al-Shihadbal-thaqib; Salasil al-hadid; al-Durr al-najafiyya; 'Aqd aljawa-hir; al-Hada'iq al-naiira; Sharh al-madarik; Sharh alfaqih; Risalat al-sala-t;Risala qati'at al-q-il wa'lqil; Risala kashf al-qina'; Risa-lat al-kunuiz al-mawdi'a; Khutub yawm alJuma'; Anis al-musiifir wajulits al-hadir; al-Risala al-Muhammadiyya; Risa-lat mzzan al- tarjfih; Man-sik al-hajj; Risa-lat al-layalf; al-Naffahat; Risiila tahrim aljuma' bayn alfatimayn. The ijaza ends with section (v), praises of the mujiaz and dedicating the work to God. THE RELATIONSHIP

BETWEEN THE IJAZA

AND THE LU'LU'AT AL-BAHRAYN

work the Lu'lu'at alBahrlini's biographical bahrayn bears certain significant similarities to the ijiza to Bahr al-'Ulfim. The Lu'lu'a is a tabaq&twork written in the form of an ijaza to Bahrlini's two

Muhsin. He wrote the book [commentary] Ziid almusafirin and he used to argue with Mawla al-Harawi in his house in Tuis, demonstrating to him what was obligatory and binding. His dispute with him was well known.24 That the two works are connected, with a similar structure, utilising common material, is clear. The relative dating of the two works indicates that the ijaza predates the Lu'lu'a. The Lu'lu'a is dated as having been finished in 1182 A.H. The ijzaza is not dated; however it can be shown to have been written before the Lu'lu'a by a comparison of the following two sentences in the lists of books in each work: The Ijaiza to Bahr al-'Ulum: [Concerning al-Hladii'iq al-nidira] the kitab al-Zakaithas now been written, and most of the kitdbal-Hajj (22a.3-4).

The Lu'lu'a: and amongst them (meaning the works of BahrdnDi)is the Kitab al-HadZi'iq, up to (the end of) the kitab al-Hajj, and I am now occupied with the kitab al-Matajir.2s In the printed edition of the Hada'iqj6, the kitab al-Matajir follows directly on from the kitab al-Haij. Presuming that Bahrani wrote the Hacdai'iq in the order he states, which is also the order they appear in the edition, one can assume that ijaza was written before 1182. Evidence from biographical works point towards Bahr al-'Ulim being Bahrini's student from, at the earliest, 1175 A.H. The izaza, it can be reasonably assumed, was written between 1175 and 1182 (A.D. 1761-8) and is therefore a forerunner or prototype of the Lu'lu'a.

THE IJAZA

FROM

YUSUF

AL-BAHRANI

TO SAYYID

CONCLUSION This ijaza came into existence as a document separate from its matn (the works to which it refers). It has been copied, together with a number of other ijazas into the volume where it is now found. It is definitely an example of the later stages of the ijaza system when ijazas were produced as documents separate from the works they referred to. However, the motives which Bahraini cites for giving the ijzza are not primarily concerned with the transmission and preservation of texts. Instead, it is to continue the chain of transmission (silsila, mu'an'ana) which leads back to the earliest 'a"lims. Fihrist-i Nusakh-h-.-yi khatt-yi Madrasayi 'umum-yi Najafi Mar'ashl (Qum, 1976-92), vol. XV, no. 5605, p. 9. 2 Tehran,

n.d.

Qum, 1390-1/1970-1. 4 Rawda, vol. VIII, p. 204; Qisas, pp. 168-75, and see also Muhammad 'All Kashmiri, Nuji?m al-Samai' (Lucknow, 1303/1885), p. 203. 5 Rawda, vol. VII, pp. 203-16; Qisas, pp. 168-74. 6 matn is also the technical term used to describe the story (a saying of, or an incident in the lives of the Prophets or Imams) contained within an hadith. The two uses of the term should not be confused. 7 See articles, I. Goldziher, "Idjiza" in El, vol. II, pp. 446-7 and G. Vajda, "Idjaiza"in EI. vol. III, pp. 1020-2, and the Variorum collection of his articles, La transmissiondu savoir en Islam VIIe-XVIIIEsiecles (London, 1983); see also N. Calder, Studies in Early MuslimJurisprudence(Oxford, 1993), pp. 171-4. 8 MuslimStudies,S. M. Stern and C. R. Baker (London, 1967-71) vol. II, pp. 175-80. 9 Lu'lu'at al-bahrayn(Tehran, 1386/1966). 10 See, for example, H. Algar, "Shi'ism and Iran in the Eighteenth Century", in T. Naff and R. Owen (eds.), Studies in Eighteenth CenturyIslamicHistory (London, 1977), andJ. R. Cole, The Roots North Indian Shi'ism in and Iran (London, 1988). 11 This isofa reference to the fourIraq "canonical" books of akhbar: Man la yahdiruhu alfaqih by al-Shaykh alSadiiq; al-K/ifi by Ya'quib al-Kulayni; and Tahdhib al-Ahkadmand al-Istibsitr by Muhammad b. Hasan al-Tiisi, Shaykh al-Ta'if-i. 12 An example of an ijiiza to two people for a number of works can be found in F. Kern, "Tabari's Ibtilaf alFuqaha", ZDMG, LV (1901), p. 74. 13 Namely, at the end of section (iii), 21b.8. 14 See for example, "De la transmission orale du savoir dans l'Islam traditionnel", in L'Arabisant, IV (1975), and other works in the collection, La Transmissiondu Savoir. 1S See note 7 above. 16 For example, 21a.16-17 and 21a.18-19. 17 The possibility that this is an innovation by the copyist is indicated by a mistake made at 5.20b where he writes al-ha and then an isolated ha'. In the manuscript he is using (either

MUHAMMAD

MAHDI

BAHR

AL-cULUM

123

In giving the ijaza he is maintaining this chain and "seeking good fortune and blessing". Ijazas formally require a matn; however, the matn of this ijhza is described in such general terms that one cannot help but conclude that the matn here is really the whole scholarly tradition. If these features are common to other ijdzas in the later period, then one might postulate a further stage in the development in ijazas. Having become divorced from the texts, they become signs of approval from one generation to the next. A matn is formally required, and thus a large body of texts, a token of the tradition as a whole, are cited.28 the original, or more likely a copy) it is probable that isniids are separated by the formula al-ha, rather than the isolated initial. Since this is the only copy of the ijzza known to be existent, there is no means of verifying these presumptions. 18 The haylula terminology is used throughout the Lu'lu'a, see for example pp. 189 and 185 and passim. The initial hi' is used also in the ijiizas found in Muhammad Baqir Majlisi's Bihair al-anwair(Tehran, 1391), kita-bal-ijizi2t, in vol. CVIII, p. 67. al-riwd-ya(Beirut, 1969), 19 Qadi 'Iya;d,al-Ilma' ilai ma'arifat p. 88. us.ll 20 J. Robson, "The Transmission of Nasa'i's Sunan", JSS, I (1956), passim; particularly the General Remarks, pp. 57-9; Vajda, "La Mashyakha d'Ibn al-Hattab al-Razi. Contribution A l'histoire du Sunnisme en Egypte Fatimide", Bulletin d'Etudes Orientales, XXIII (1970); for more pessimistic assessments of the strict historical utility of these documents, see L. Massignon, "Chaines de temoignages fondamentales dans la tradition musulmane hallagienne", in Opera Minora de Louis Massignon, vol. II (Beirut, 1963), pp. 61-2, and N. Calder, review of La transmissiondu savoir en Islam, in BSOAS (1985), p. 202. 21 For an analysis of al-khabaral-wahidin early Muslim jurisprudence, see J. Schacht, Origins of MuhammadanJurisprudence (Oxford, 1950), pp. 41-2. For al-khabar al-wi~hidin a ShVi' context, see Calder, "Doubt and Prerogative: the Emergence of a Imdimi Sh1' Theory of Ijtihaid",SI, LXX (1989), passim. 22 Majlisi's ijaizas can be found in vol. CVIII of Biha.ral-anwar. 23 Cf. E. Kohlberg, "Non-Imaimi Muslims in Imaimi Fiqh", in Studies in Arabic and Islam, VI (1985), pp. 99-105. Jerusalem * This passage is found in the ijiiza to Bahr al-'UlHm 19a.16-18, and in the Lu'lu'a, p. 166. The word in brackets [] is found only in the Lu'lu'a. 25 Lu'lu'a, p. 446. 1363. 26 27 Qum, 28

See references under note 6 above. I would like to thank Mr. Reza Arjomand of the Mar'ashi Library for his help in locating and copying the manuscript, and Dr. Norman Calder of the Department of Middle East Studies, University of Manchester, for his extensive remarks on a number of drafts of this article.

JAMES BAILLIE FRASER: TRAVELLER, WRITER AND ARTIST 1783-1856 By Denis Wright Haddenham,Buckinghamshire

James Baillie Fraser was a man of many talents and great energy, a bold and adventurous traveller in unexplored of India, Persia and corners a writer of travel books and Mesopotamia, prolific artist. romances, an accomplished quasi-historic He also undertook a number of special assignments for the Foreign Office under Lord Palmerston. Lord Curzon, in his day the acknowledged expert on Persia, considered him "as incomparably the best authority upon the northern provinces"' of every and praised his "faithful portraiture aspect of modern Persian life".2 He was born in Edinburgh in 1783 and spent his early youth in the family home at Moniack (now known as Reelig House) deep in the Scottish Highlands near Inverness, where this land-owning branch of the Fraser clan had been lairds since the fifteenth century. His father, Edward Satchwell Fraser, had taken part as an officer in the Grenadier Guards in the American War of Independence, while his paternal grandfather, another James, had spent seventeen years in India with the East India Company, learnt Persian, collected oriental manuscripts,3 and written the History of Nadir Shah, but had never visited Persia. James Baillie, like his four younger brothers,4 was tutored at home before attending school in Edinburgh. In 1799, aged only sixteen, he went to Berbice in what is now Guyana to manage the family's sugar and cotton plantations. The venture was not a success and the plantations were sold in 1817. Apart from a visit home in 1809, James remained in Berbice until 1811 when he returned to Scotland, while his brother Edward, who had joined him in Berbice in 1803, went to India. Back in Scotland, James spent his time shooting, touring the Western Highlands and Skye with a cousin, and pondering about his future. Seeing little promise at home, he decided to try his luck in India and family's fortunes help restore his debt-encumbered with remittances home. of Charles Grant Thanks to the patronage (1746-1823) a family friend, Member of Parliament for Inverness and long-serving director of the East India Company, all four of James's brothers were already employed by the Honourable Company in India. Now, thanks to Grant's help, James was found a passage on H.M.S. Daedalus, a frigate

convoying Indiamen from Liverpool, at the end of January 1813. After nearly five months at sea, culminating in shipwreck off Ceylon, James landed at Madras in July. Luckily there had been no loss of life and, apart from a few trifles, James lost none of his luggage. He spent some weeks travelling in South India before moving to Calcutta, where he entered into a short-lived and unsatisfactory business partnership with a Mr. Beecher. They lost money speculating in indigo but when the partnership was dissolved, James received some 35,000 rupees as his share of the profits5 and felt free to 1815 for Delhi in leave Calcutta in January his brother with third Alexander, who company was employed there. From Delhi, James travelled north to Nahan in the Punjab hills, reaching there on 13 March. A few days later he was joined by his second brother, William, whom he had not seen for sixteen years and was now Political Officer with the East India Company's forces, then engaged in the Nepalese War. James spent April in Nahan, fascinated by all he saw, sketching and witnessing the siege of nearby Jytock. Early in May he and William embarked on a three months' tour of the lower Himalayan hill Commisstates, William having been appointed sioner in Garhwal. They were escorted by some six hundred Irregulars, recruited by William, in what was a display of force among the restless hill people. The two brothers spent most of May, June and July 1815 travelling through unknown regions of the Himalayas. At one point, James went off without William to find the sources of the Jumma and Ganges rivers. Throughout this period he kept a diary, made detailed notes of all he saw and learnt, mapped uncharted country and sketched. James and William were back in Delhi by midAugust, James remaining there until the following June when he returned to Calcutta. During these eight months in Delhi he continued to draw and, with William, commissioned Indian artists-as he later did in Calcutta-to paint Indians of all classes in their infinite variety of dress. Back in Calcutta, where he spent the better part of four years (1816-20), he made drawings of the expanding city and its environs; he also sought to improve his own artist, technique with the help of a professional William Lord Havell, who had accompanied

125

126

JOURNAL

OF PERSIAN

Amherst's mission to China in 1815-17, and by

STUDIES

known painter, who was then living in Calcutta.He also spent time in working up his Himalayan sketches into finished water colours which he sent

father that he had decided tell his understanding to quit India by the end of the year. Shortly afterwards, when it was too late to change his mind, he had qualms about doing so on learning that the family debts, at ?21,000, were heavier than he

home with his journal. With his father's help, both

thought.10

taking

lessons

from

George

Chinnery,

the well-

were published in London in 1820. Although

in the preface to his published Journal

Fraser modestly claimed that, owing to his own

"fatal deficiency of information in every branch of physics", he was unable to contribute "to science in those branches which are now so interesting as

geology, mineralogy,botany, etc.",6theJournalwith its thirteen appendices provided just this and is

... an amazing record of a region which remained little known until the 20th century. It provides meticulous information on a wide variety of subjects-history, agriculture, flora and fauna, geology, ethnology, commerce, manufactures, mineralogy; information which was collected under conditions of extreme physical discomfort and at times even danger.7 Along with the Journal of over five hundred

pages, twenty aquatints of Fraser's water colours

were published in a folio edition entitled Views in the Himala Mountains. Subsequently, between 1824

and 1826, twenty-four aquatints of Fraser's Calcutta drawings were published in London in eight parts. The two sets have justly been acclaimed as "amongsthe finest engravings ever made of Indian scenes". Fraser also sent to the Geological Society in London specimens collected during his Himalayan travels, together with notes about them, duly published in the Transactions of the Geological Society (1819), being the first of several papers with specimens presented by Fraser to the Society, by whom he was elected a Fellow in 1823. The coloured drawings commissioned by James and William from Indian artists only came to light in 1979, when more than ninety of them were discovered in Reelig House, together with James' diaries and a mass of his letters and other papers. A selection of these Indian drawings and Fraser's own aquatints were published in 1989 by Mildred Archer and Toby Falk in their absorbing India Revealed. The Art and Adventures ofJames and William Fraser 1801-35. During his second stint in Calcutta, James again tried his hand at commerce but without much success orjoy. He joined his friend Aeneas Mackintosh and other fellow-Scots as junior partner in a trading house, being mainly occupied on the shipping side. But he was not optimistic about his prospects. His heart was more in painting, writing and the colourful Indian scene than in business; he was also homesick for "the life, the friends, and the country that I love".9 Filial duty had kept him in India so long but now, in October 1819, he wrote to

James finally left Calcutta in May 1820 but did not reach home until over two-and-a-half years later. First, he spent four months travelling north to see his brother William, now stationed in Delhi. He spent two months there before travelling south through Rajasthan to Bombay, a journey of some three thousand miles through little-known country. He was accompanied by his negro servant from Berbice, fifty Indian servants and sowdrs and a train of horses and camels. He admired the mosques and palaces, sketched and kept a diary as well as recording latitudes and longitudes, and collecting geological specimens." However, hopes that he would find a publisher for his journal and drawings again were disappointed. James reached Bombay at the end of March 1821. There he joined forces with Dr. Andrew Jukes of the East India Company, who was bound for Persia on a delicate diplomatic mission. Together they set sail on 14 May and touched at Kishm island (whose Company's sepoy garrison was the subject of Jukes' mission) and Bandar Abbas before landing at Bushire on 4 August. It had been an unpleasantly slow journey, bedevilled by calms, contrary winds and intense heat. By this time James, who had originally thought of travelling home via the Red Sea and Egypt, intended after reaching Tehran to make for Persian Kurdistan "a country little known, not at all described either by pencil or pen" and then through Azerbaijan to Turkey, where he hoped to see "as much as I may be able of the country lying between Antioch and Smyrna and Trebizond on and Constantinople, the Black Sea, a large sweep that is very imperfectly known", with "many Grecian and Roman monuments to give employment to the pencil".12 He was probably thinking of producing an illustrated book. However, by the time he reached Shiraz he learnt that the way into Turkey was blocked by fighting between Turkey and Persia. His thoughts then turned eastwards to Nishapur, Mashhad, Bokhara and Samarkand.13 Remote places no less than ancient monuments attracted him. Cholera was raging in Bushire and delayed Jukes' and Fraser's departure for Shiraz, which they did not reach until the end of September, a few days before the death there from cholera morbus of Claudius Rich, the East India Company's young and brilliant Resident in Baghdad. Fraser read the Anglican burial service over his grave. Two months later in Isfahan, Jukes also succumbed and was buried in the Armenian cemetery there. Fraser,

JAMES

BAILLIE

FRASER:

TRAVELLER,

"acting the Elchee [ambassador] here for the last ten days",14 took it upon himself to take his late friend's place and negotiate with the Persian Prime to be in Isfahan-an Minister who happened initiative that earned him a mild reproof from the authorities in Bombay.15 From Isfahan, Fraser travelled to Tehran with Dr. John McNeill (later Sir John), the Mission doctor, who, on hearing of Jukes' illness from a by Fraser, had ridden messenger despatched virtually non-stop from Tehran in less than four days in a vain attempt to save his life. This was the beginning of a life-long friendship between Fraser and McNeill, his junior by twelve years. after a month in At the end of December, Tehran, Fraser left for Mashhad dressed as a Persian merchant, not so much as a disguise as to avoid undue attention. His entourage consisted of his "faithful negro John" a Christian in Fraser's employ since his Berbice days twenty years earlier, four Persian servants including cook and grooms, and an educated young Persian as interpreter.16 A which train of horses carried their impedimenta, a variety of instruments included brought by Fraser from India to help ascertain latitudes, longitudes and altitudes. They spent over a month on the well-travelled road to Mashhad through Semnan, Damghan, Shahrud and Nishapur. As in India, so on this and all his later journeys Fraser assiduously collected information about places and people and sketched. to miss nothing, whether He was determined "some curious old monuments" in Bastam or the turquoise mines near Nishapur. He spent six weeks in Mashhad, where he was regarded with deep suspicion by the mullahs yet somehow, with the help of a Shrine official whom Fraser thought had seen through his disguise, was able to enter the tomb chamber of the Shrine. Equally remarkable, he was able to sit and sketch from an upper row of the Shrine's arches.17 But because of political turmoil he had to abandon his plans for Bokhara and Samarkand; instead, he decided to turn westwards and make for Tabriz and home by way of the Caspian provinces of Mazandaran and Gilan, then scarcely known to Europeans. Fraser left Mashhad in March 1822 and reached Tabriz four months later. The first stage of his by an armed escort, was journey, accompanied across the remote and lawless territory of the Turkomans, no respecters of human life, through to Astarabad and Bujnurd Radkan, Quchan (Gurgan). From Astarabad he went to Amol, Sari and along the Caspian shore to Rasht, then the centre of the silk trade. There he was suspected of being a Russian spy, arrested, escaped, robbed and recaptured before eventually being allowed to continue his journey to Enzeli (where he noted the

WRITER

AND

ARTIST

1783-1856

127

price of naphtha from Baku used in lamps)18 and through the densely wooded Talesh mountains to Ardebil, where he saw Shah 'Abbas's great collection of Chinese porcelain lying covered in dust on the floor of the Safavid shrine. Although it was mid-July by the time Fraser reached Tabriz he was in no hurry to get home, despite the prevalence of cholera in and around the town. Early in August he went with Major William Monteith of the British Military Mission then in Tabriz to help survey the country around Lake Urmia. He took note of the Christian Armenian and Nestorian villages in the region and was one of the first travellers to describe the hereditary Nestorian patriarch, the Mar Shimun, "a sort of prelatical chief, whose dignity is hereditary in the family although the chief himself, being set apart for the church, cannot marry".19 Fraser returned to Tabriz from Urmia and then at the end of August, unable to enter Turkey, made for Tiflis and home, travelling by way of Odessa, Vienna and Munich to Paris, which he reached in January 1823. Though having only a few hours in Munich he visited the picture gallery there, as also the Louvre in Paris, such was his keen interest in painting. There was a happy reunion, after ten years, with his parents in St. Omer, then home. Later in the year Fraser married his cousin Jane a Tytler, younger daughter of Lord Woodhouselee, distinguised Scottish lawyer and university professor. The Frasers had no children.20 For the next ten years James divided his time between Scotland and London, writing seven lengthy books, all but one and doing what he with a Persian background, could to help restore the family fortunes. Of these books the two most interesting were in 1825 and 1826, viz. Narrative of a published Journey into Khorasan in the Years 1821 and 1822 and Travels and Adventures in the Persian Provinces of the Southern Banks of the Caspian Sea. Both books are of it, as Fraser packed with information-some gathered from earlier writers but acknowledges, much of it collected by himself-on the state of the country, its government and leading personalities, revenues, army, religion, buildings, etc. He is very critical of the Persian character, blaming it on their hated Qajar rulers. He gives a rare first-hand description of the shrines in the holy cities of Qum and Mashhad, both of which he had entered in disto do so. guise, probably the first European the remote he did not visit and wonderful Though natural fortress of Qalat-i Naderi in Khorasan, he describes it in some detail in one of his many appendices which include, inter alia, the latitudes of places where he stayed or and longitudes camped, together with temperatures and weather reports, routes, prinicpal exports and imports to and from Europe and neighbouring countries,

128

JOURNAL

OF PERSIAN

local manufactures and natural resources, together with information about Bokhara and the other Trans-Oxus khanates. He lists the most popular colours for imported woollens, lamenting that "the French and German manufacturers have been much more successful than the English in hitting the Persian taste; no doubt because they took care to have the best information upon the subject".21 He also advocates-being one of the first persons to do so-the use of the Black Sea routes to Persia via Trebizond and Redout Kaleh as cheaper than the customary overland routes through Bushire and Constantinople. With a fine eye for detail, he recommends that ships using these Black Sea ports should not draw more than three and a half feet of water. Fraser's next two books, published in 1828 and 1830, were The Kuzzilbash. A Tale of Khorasan and its sequel The Persian Adventurer, each book consisting of three long, rambling volumes. It is unlikely that they were, as Fraser states, translated and adapted from the for the English reader by himself memoirs of a man who had been in the service of Nadir Shah. Fraser was not a Persian scholar and his knowledge of the language was limited. It seems more probable that he based his story on a variety of sources, including his own grandfather's History of Nadir Shah. All six volumes are heavy reading, a mixture of history and romance centred on Nadir Shah with descriptions of his rise to power, his wars against the Afghans and Turks, his coronation and his invasion of India, with grim details of the Delhi massacre and looting of the royal treasure. According to this account, the Peacock Throne was stolen before reaching the safety of Qalat-i Naderi. There followed in quick succession three more lengthy books, each very different from the other. in 1832, is a The Highland Smugglers, published colourful story about the old Highland life of smuggling and illicit stills. The following year, Tales of the Caravanserai. The Khan's Tale appeared as a volume in a London publisher's popular Library of Romance: this was another rambling story set in Fraser's favourite Turkoman country during Nadir Shah's reign. Next, in 1834, came An Historical and Descriptive Account of Persia as Volume XV in the Edinburgh Cabinet Library. This last book was, in its day, the most complete and informative account of Persia in the English language, ranging from history and religion to tribes, geology, flora and fauna, together with a chapter on Afghanistan and illustrated by thirteen engravings, all but one made from Fraser's own drawings. Fraser again wrote scathingly about the Persian character and their Qajar rulers, whose early downfall he predicted. He was equally wrong in believing that Christianity would eventually displace Islam in Persia. Fraser never overcame his

STUDIES

early dislike of the Persian character. In writing home soon after his first arrival in the country he had described the Persians as "cringing, mean and abject, when in the presence of one they fear; insoand over-bearing when they lent, disrespectful, dare to shew their teeth" and "a people without truth or principle, cowed and degraded".22 These books established Fraser as an expert on Persia. Not surprisingly, therefore, when alarm bells were set ringing in London early in 1833 by Russian moves in Turkey, Fraser, on the recommendation of Charles Grant (later Lord Glenelg), son of the Charles Grant already mentioned, was invited by the Foreign Office to return to Persia and to report on the Russian position there.23 His instructions, signed by Lord Palmerston the Foreign Secretary and endorsed by King William IV, were wide-ranging and were evidence of the growing concern at the highest level in London about Russian ambitions.24 Fraser was called upon to report on "the actual condition of Russian intercourse and influence, political as well as commercial, with Persia" and with other countries lying between Russia and the Indus. He was to ascertain the extent of Russian influence at both the Shah's Court in Tehran and that of the Crown Prince, 'AbbSis Mirzai, in Tabriz and to discover the attitude towards Russia of the leading personalities at both Courts; also, to assess the prospects of the Crown his father, Fath 'All Shah Prince succeeding of either Russia or the "without the intervention United Kingdom". Were there any other claimants to the throne? Could the Crown Prince count on the loyalty of his European-trained troops? In addition, Fraser was instructed to visit the Caspian provinces and to ascertain the extent of trade carried on there by Russian subjects, and also to judge whether the Shah was right in fearing that the Russian wish to establish consuls there was a prelude to annexation. Further, he was to return home via Hamadan, Kermanshah and the Ottoman Pashalik of Baghdad in order to assess both the degree of the Persian Government's authority over the Kurdish tribes and the possibility of Baghdad of the unstable Ottoman becoming independent Empire. Finally, the purpose of his visit was to be kept secret. He was "sedulously to avoid the appearance of being invested with any diplomatic character or mission whatsoever. You will always present yourself in the character which you really possess-, namely, that of a gentleman, in the confidence of the Government, to whom important despatches for the Envoy [in Tehran] and Resident [in Baghdad] have been entrusted and who is desirous of recommending himself to the further notice of his Government by obtaining correct information respecting the Courts and countries which he has

JAMES

BAILLIE

FRASER:

TRAVELLER,

visited". Almost too subtle a cover story to fool anyone! Before leaving London, Fraser was paid ?700 to cover outfit, loss of time and expenses already incurred; in addition, he was to be paid ?1,200 a 1 December 1833, while on his year, beginning mission together with full reimbursement of travelling expenses of every kind against submission of by vouchers when regular accounts supported This was a arrangement practicable. generous which would have pleased Fraser's father, who had advised his son to guard against "being seduced by some vain honour, some silly title ... to the detriment of your pecuniary compensations".25 After over six months' irritating delay, which he blamed on the Foreign Office rather than the East India Company's Board of Control, Fraser left London at the end of December and was away all but a year and a half. During this time he covered well over ten thousand miles, all of it "Tartar or post", i.e. horseback, once he reached the end of the carriage road at Semlin (Zemun) on the AustroSerbian frontier, and mostly without the company of a fellow-countryman except on the outward and Tabriz. Both between journey Constantinople going and returning he faced vile winter weather and rode along rough and often dangerously precipitous paths, deep in mud or snow, spending nights in verminous hovels, yet somehow managing to write up his diary, make copious notes and keep his family, the Foreign Office and Charles Grant, a Member both of Parliament and of the East India Company's Board of Control, informed of his progress and ideas. In addition, he sent the influential Grant a number of long drafts which formed the basis of some of the memoranda which he submitted to the Foreign Office on his return. It was a remarkable feat for a man of over fifty. Fraser had met Edward In Constantinople, Bonham and the pair decided to travel together as far as Tabriz, where Bonham was in business and was to become the first British consul. They left the Turkish capital on 16 January 1834 with nine horses, grooms and an Armenian servant, taking the ancient route through Amasya and across the bleak Anatolian plateau to Erzerum; then via Karakilisa (Agri) to Bayazit and past Mount Ararat to Tabriz-five weeks of "discomfort and fatigue and anxiety". When at long last he reached Tehran on 8 March, Fraser had this to say about his journey: The first seven hundred miles, from Constantinople to Amassia, were performed within six days, in bitter weather, and in spite of mud, and rain and snow; but for the last seven weeks ... we never saw the colour of the earth. During the whole of this period we have been wading night and day through interminable wastes of deep snow, exposed to all the violence of storm and drift,

WRITER

AND

ARTIST

1783-1856

129

and wind, with the thermometer frequently 150 to 200 under zero. Our clothes and face, and beards, were clotted into stiff masses of ice. Our boots hard as iron, frozen to the stirrup, and our limbs tortured with pain, or chilled into insensibility by intense cold.26 Tehran served as Fraser's base for the next year, the starting and finishing point of two long journeys which he undertook before returning home. First, however, he spent two months in the capital, He was studying local customs and ceremonies. received in audience by Fath 'Ali Shah and met Abu 'I-Hasan many old friends including Mirza. first Persia's Khan, Foreign Minister, who had taken London by storm twenty years earlier. He was especially pleased to find John McNeill was now right-hand man to the British Minister, Sir John Campbell. He shared McNeill's distrust of the latter and wrote a number of letters to Charles Grant about Campbell's ineptitude and low standing with the Shah and his ministers. These letters to Campbell's recall the following contributed year.27 Fraser left Tehran for Mashhad on 8 May 1834, following his old 1822 route through Shahrud with a train of horses and mules, servants and muleteers and a well-connected young mirzd to help improve his Persian and make contact with local officials. En route he met Prince Muhammad Mirza, the eldest son of the late Crown Prince whose death a few weeks earlier caused Fraser to worry about the succession when the ailing Shah died. He considered Muhammad Mirza the worthiest of Fath 'Al1 Shah's numerous descendants and hoped that the Shah would nominate him as his successor (which he in fact did). He found Mashhad, like much else in the country, very run-down and full of beggars. He spent two weeks there collecting information about conditions in Afghanistan before venturing westward across the turbulent Turkoman country him on his previous which had so fascinated journey. But once again the troubled state of the countryside prevented him visiting Qalat-i Naderi; for the same reason, he was unable to reach Astarabad and was obliged to hurry back to Tehran for fear of possible chaos should the Shah die. He was also having trouble with his servants. He reached Tehran at the end of July and was disappointed to find the British Mission shut. Most he enjoyed robust good healthunusually-for Fraser had to take to his bed with a high fever. However, he soon recovered and on 17 August left Tehran on the second and far more trying leg of his mission. This time he was away six months. He first went north to the Caspian by a route which took him past Mount Demavend and down the valleys of the Lar (where he caught trout) and Haraz rivers to Amol, where he was back on familiar ground. From Rasht he turned north-west

130

JOURNAL

OF PERSIAN

through the Elburz mountains to Turkomanchay, reaching Tabriz on 14 September. There he found members of the British Mission including his friend McNeill. After a week's much-needed rest, he left for Urmia and then went south to Sauj Bulaq (Mahabad), whence he crossed the heart of the Kurdish mountains to Sulaimaniya and Baghdad, passing through areas where no European had travelled. The proud, previously independent Kurds reminded him of the old Highland clans. At the beginning of November Fraser reached Baghdad, which became his base for the next three months while he explored the countryside and collected information about the political situation, the Arab tribes and their shaykhs. He found the Pashalik of Baghdad in a state of desolation and of misery due, he believed, to a "combination causes, of which pestilence and famine with the oppression of their rulers are the principal",28 and the depredations of the Arab and Kurdish tribes. In Baghdad he saw signs of Russian "secret workings"29 and made the unrealistic recommendation that the British Government should rent the Pashalik from the Ottoman Government in the laudable hope that Mesopotamia and Syria might thus regain their lost prosperity and "become not only valuable producers but great consumers of our manufactures".30 In company with two fellow countrymen then living in Baghdad, he visited the site of ancient Babylon and other ruins and unexcavated mounds, being, in the words of Professor Seton Lloyd, "the first [European] to venture across the terra incognita between the two rivers and in the process discovered and recorded the names of mounds which we now know to be the remains of some of the most famous Sumerian cities.... He was also probably the first visitor to the ruins of Ur". 31 Abandoning an earlier plan to go south to Basra and Khuzistan, Fraser left Baghdad for Tehran at the end of January 1835, taking the direct road and Hamadan. through Khanaqin, Kirmanshah he when in Tehran the had arrived Although, previous year, Fraser had sworn that "no consideration upon earth, short of absolute and imperative duty"'32 would induce him to undertake another

such winter'sjourney, he was now doing so again. In letters home he complained of heavy rain, deep mud and snow, and bitter cold before reaching

Tehran in late February. He could rightly congratulate himself on having travelled some four thousand miles through lawless country without loss of man, beast or property. Winter was not yet over when he left Tehran for home on 8 March 1835, returning via Tabriz and Turkey as he had come. There was more bitter cold, mud and snow causing "mental weariness and bodily distress..,. sitting ten or twelve mortal hours

STUDIES

in cold and wet, flogging on a parcel of miserable yaboos, galled and lame, and lean and lazy".33 Despite this he covered the eleven hundred miles where (sic) between Erzerum and Constantinople, he arrived on 11 April, in eleven days, of which "two were lost from delay for horses so that I rode for nine days at the rate of 122 miles a day".34 After a few days in the Turkish capital as the guest of Lord Ponsonby, the British Ambassador, Fraser left for Semlin on 22 April; then by carriage through Vienna, Cologne and Brussels to board the Dover-bound boat on 23 May. He had been away a year and five months and had covered over ten thousand miles. His travel claim for ?1,253 was accepted by the Foreign Office without demur. Fraser's account of these travels was published in 1838 and 1840 in two two-volume books in the form of letters home: A Winter'sJourney (Tartar)from Constantinople to Tehran and Travels in Koordistan, much detailed Mesopotamia, etc. They contained about places, scenery and, above all, information peoples (Turkomans, Gilanis, Kurds, Arabs) almost unknown to the West. The books also contained brief but valuable accounts of the problems caused by the deaths of the Crown Prince and the Shah and the struggle for the succession. There followed in 1842 a second volume by Fraser in the Edinburgh Cabinet Library, Mesopotamia and Assyria from the Earliest Ages to the Present Time. This was an encyclopaedic work drawn from all available sources. His travels also provided background for two more excessively long romances (in three and four volumes respectively), Allee Neemro, the Buchtiaree Adventurer. A Tale of Louristan (1842) and The Dark Falcon (1844). Fraser combined this prolific authorship with work at the Foreign Office. Within a fortnight of his arrival home he had submitted to Lord Palmerston the first of six long memoranda, mostly about the Russian peril and the way to deal with it. One memorandum followed another in quick succession, as follows: 1. On the political state of Persia in March 1835 and the means of preserving it from falling into the hands of Russia. 8.6.1835. 2. On the designs of Russia on Persia, Turkey and India and the means of counteracting them. 23.6.1835. 3. On the creation of an Oriental Department in the Foreign Office. 20.7.1835. 4. On the powers to be invested in, and the instructions to be given to the British Minister at the Court of Tehran. September 1835. 5. On Lieut. Rawlinson's letter to SirJohn Campbell of 8 June 1835. 3.10.1835 [in which Fraser urged British help in settling Turco-Persian frontier problems]. 6. On that part of Mr. Wherry's letter of 8 July which relates to the means of disposing the Arab and other tribes on the banks of the Euphrates to assist rather than impede the progress of the Euphrates Steam Expedition

JAMES

BAILLIE

FRASER:

TRAVELLER,

under Colonel Chesney. 4.10.1835 [in which Fraser provided detailed information about various Kurdish and Arab tribes]. The theme running through the three principal memoranda was the dire threat to British India from Russia. To prevent an eventual Russian occupation of Turkey, Persia and Central Asia, Fraser with scant concern for the political recommended, or financial implications, that the British Government should subsidise and become the main influence in "every Asiatic state from the Indus to the Bosphorus" and at the same time conclude defensive alliances with the then independent khanates of Bokhara and Khiva. In proposing "vigorous action" while Russia was still weak, he even seemed to be thinking of war. Fraser saw Afghanistan and Persia as the main barriers against Russian advance. The purpose of the British Mission in Tehran was "resistance to the power and progress of Russia" and "the annihilation of Russian influence" there. The succession of Muhammad Shah under the strong influence of his pro-Russian first minister or qa'zm-maqam had, he argued, tipped the scales in Russia's favour. The British, having helped the Shah to the throne, were now blamed for the miserable state of the country. Strong action was therefore necessary if Britain rather than Russia was to become the protecting power in the weak, demoralised country. To this that the British Governend, Fraser recommended ment should interfere "as delicately as may be but and directly" in the government of decidedly Persia; that the pro-Russian qa 'zm-maqamshould be dismissed; and that the British Minister, aided by the appointment to the Court of a carefully-chosen should become the power English physician, behind the throne. In order to hold his own "in the eyes of the natives" against the Russian envoy, the British envoy should be "a nobleman or gentleman of rank and character", backed by a good staff and impressive Lancer escort; also, like the Russian Ambassador, he should have at his disposal funds for intelligence gathering and the suborning of Court personalities. Because communications were so slow, Fraser also recommended that the British envoy should be given "more unlimited powers" to act on his own and be empowered to let both the Shah and his Russian opposite number known that Britain was "resolved at all hazards to maintain the and integrity of Persia against independence Russian encroachment". Additionally, he proposed that British officers should train, administer, and regularly pay Persian forces in the key frontier provinces of Azerbaijan and Khorasan. With his fourth memorandum he included a draft AngloPersian Defensive Treaty under which the Shah would take up arms against any Power seeking to attack India through Persia or Turkestan in return

WRITER

AND

ARTIST

1783-1856

131

for Britain undertaking to uphold the independence and integrity of Persia, and in the event of attack by a European Power, to compel that Power to withdraw. This last proposal, which begged the question whether Russia was a European Power, never saw the light of day but several of Fraser's other ideas were adopted during the nineteen century as moves in the "Great Game". Fraser's views on Persia had thus come a long way since his first visit to the country when he had wanted "an end to our senseless, useless and degrading relations with this country, from which we receive no advantage and continually are incurring thankless expense".36 He was ahead of his time in proposing the creation of an Oriental Department in the Foreign Office "to meet the new and rapidly increasing pressure of business arising out of Asiatic affairs" and the compilation of a large-scale map and a of the countries updated regularly gazetteer between the Bosphorus and Indus. Fraser would have liked this job for himself, and in a letter to Palmerston, enclosing a copy of his memorandum, discreetly proposed himself for it while disclaiming any "selfish wish of insinuating myself into the service of the Government".37 Although nothing came of the idea at the time, Fraser continued working at the Foreign Office, writing unsigned articles and pamphlets for the press designed to awaken the public to the Russian menace. In this he was helped by John McNeill, temporarily back from Persia, and David Urquhart, "an extravagant Turcophil", both of them as paranoid about Russia as himself.38 Also, in June 1836, he was given the job of dancing attendance on three Persian princes, sons of one of the unsuccessful claimants to the Peacock Throne, who had come to London at their father's behest in order to seek help and protection. The British Government, seeing them as a potentially useful card up their sleeve against the pro-Russian Muhammad Shah, treated them as honoured guests. Fraser, as their mehmandar, paid them assiduous attention during their three months' stay in London. He arranged private apartments for them and their three servants and Christian Syrian interpreter in the fashionable Mivart's Hotel in Mayfair, while he lodged nearby in Davies Street. He obtained Foreign Office approval of a generous budget to cover their daily expenses, including the hire of a carriage and two footmen; and he attended daily on these three young men, of whom he became fond, though both they and their servants could be troublesome. He took them sightseeing: to the Zoo and Kew Gardens, St. Paul's and Mme. Tussauds, the Bedlam madhouse and the

132

JOURNAL

OF PERSIAN

new tunnel under the Thames. He travelled with them on their first railway journey and accompanied them on their social rounds, to the opera and theatre, and to balls and receptions, among them one given by the Duchess of Kent, where the seventeen-year old Princess Victoria was dazzled by their good looks. Fraser's advocacy was an important factor in the British Government's decision to support them with an annual pension of ?2,000 when they left London in September to join their families in Baghdad. He had argued that "in the present aspect of Eastern affairs these men are worth any pecuniary sacrifice to this country" and in Russian hands would provide a pretext "for introducing her troops into Persia". 39 When the time came for the princes to leave London, they flatly refused to return by sea as they had come. Reluctantly, the Foreign Office agreed that they might travel overland. To ensure that they did not fall into unfriendly hands en route Fraser agreed to travel with them as far as Constantinople. The Foreign Office allowed both his wife and her sister (in place of a lady's maid), together with a manservant, to accompany him and the Persian party. Two of the servants having been sent back by sea, the party of nine were able to crowd into two horsedrawn carriages when they landed at Calais. They spent over three months on the journey travelling via Brussels, Vienna and Bucharest, where they were delayed by illness. In mid-November, leaving his wife and sister-in-law in the care of H.M. Consul and his sister, Fraser left Bucharest with his charges for Galatz on the Danube. There they were again held up by illness before boarding an Austrian on 6 steamer. Constantinople They reached December 1836. Fraser's instructions were to return home once he had arranged the safe onward journey to Baghdad. Owing either to Ottoman bureaucracy or, as Fraser believed, Russian intrigue it was over two months before he did so. The princes gave him a diamond ring as a parting gift. on 19 Fraser eventually left Constantinople February 1837. Instead of taking the more direct route for Bucharest through Adrianople (Edirne) he went, at the request of Lord Ponsonby, H.M. Ambassador in Constantinople, along the Black Sea coast through Bourgas and Varna to Kustendji on what was another intelligence(Constantza) gathering journey.40 Once again he was travelling "Tartar or post". After crossing the Danube at Czernavoda, he was held up on the Wallachian frontier by three weeks' enforced quarantine, contentedly spent writing, before rejoining his wife and Miss Fraser Tytler in Bucharest and hurrying home.

STUDIES

Less than a year later, Fraser's account of the princes and their stay in London appeared in two volumes, dedicated to Lord Palmerston. Though an invaluable record this book, like his others, was pedestrian in style. One critic, the British Consul in Constantinople, thought "it would stand a better chance of going through a second edition if he had put a little more bawdy and a few choice Persian expressions into it".41 Back in London, Fraser continued his job at the Foreign Office, preparing pamphlets and press articles about the Eastern Question. He kept in touch with the Near and Middle East through with Ponsonby in Turkey and correspondence McNeill and members of his staff in Tabriz and Tehran. These friends provided him with material for the long memoranda which he produced for the Foreign Office, whether they welcomed them or not. One was headed "The expediency of losing no time in occupying a commanding position in some 270 another, Afghanistan"; pages of closelywritten foolscap, described Russia's alleged designs on the Ottoman Empire and the necessary counteraction to be taken by the British Government.42 For financial if no other reasons, Fraser hankered after permanent employment by the Foreign Office but in 1838, when he enquired about the prospects, was curtly told that, while Lord Palmerston was entirely satisfied with his was not possible.43 conduct, further employment Rather surprisingly he was not thanked for past he had told his services, perhaps because-as are unfavourable to my views".44 mother-"many was shortly followed However, this disappointment by the East India Company's payment to him of for the assassination in ?5,000 as compensation Delhi three years earlier of his brother William. This made James financially independent, so that he could now "snap my fingers at the Foreign Office, whether they want me or not".45 He had, in fact, parted with the Foreign Office on the best of terms: "they have behaved honourably to me ... I am now a free man immersed in the important matters of turnips and potatoes, sale of wood, the encroachments of tenants, and the trespasses of neighbouring farmers".46 a year later Fraser accepted an Nevertheless, invitation from the Foreign Office to return to London and act as mehmdnddr to an important visitor from Persia. This was Husayn Khan, Nazim sent to England by the al-Dowla, the djaudadn-bashi, Shah ostensibly to offer congratulations on Queen Victoria's accession but primarily to seek the recall of John McNeill, now Minister in Tehran, and to settle a number of differences between the two governments following a Persian attack on Herat. Khan arrived in London early in June H.usayn 1839. Fraser, who was paid three guineas a day

JAMES

BAILLIE

FRASER:

TRAVELLER,

to cover all his expenses except travel, had the unenviable task of telling him that he would not, as was customary, be the guest of the British Government but would have to pay his own expenses. Also, because diplomatic relations were broken, the Foreign Secretary Lord Palmerston would receive him at his private residence in Stanhope Street and not at the Foreign Office. There they had two long, unsatisfactory meetings attended by Fraser to make the record; this proved no easy task, the conversation being, as Fraser complained, "so desultory and changeful, and the Khan so pertinacious in his repetitions and attempts at evasion and petty quibbles that it was very difficult to recollect the various terms of dialogue".47 Although Fraser spoke some Persian he did not interpret at these meetings, this being done by a Persian, Mirzai Ibrahim, a teacher of Arabic and Persian at East India College, Haileybury, who had also helped Fraser with the Persian princes. The departure of Husayn Khan in July 1839 marked the end of Fraser's work for the Foreign Office. Apart from one or two health-seeking trips to the Continent he did not travel abroad again. Instead, he busied himself with writing, local of his much-loved affairs, and the rehabilitation Scottish estate having, on the death of his father in 1835,. become the fifteenth laird of Reelig. There, in the words of the Inverness Courier, he became "a zealous of his Highland estate",48 improver farming, planting trees, reviving an abandoned lime works and rebuilding the south front of the Books were home. eighteenth-century family written and published; the last, in 1851, was a twovolume Military Memoir of Lieut. Col. James Skinner, CB, being an account of the life and career of of the Frazer's old Delhi friend, the founder famous Skinner's Horse, taken in part from Skinner's own journal written in Persian. Almost to the end Fraser kept up a lively correspondence with his great friend John McNeill, who had settled in Edinburgh. They corresponded about books, their health, local affairs and, above the Russian all, about the Eastern Question,

Abbreviations F-Fraser of Reelig papers (B = Bundle; V = Volume) IO-India Office Library and Records PRO-Public Record Office SRO-Scottish Record Office

1 G. N.

Curzon, Persia and the Persian Question(London, 1892), 2 vols. I, p. 356. 2 Ibid., I, p. 24. Some 200 Mss, mostly Persian but also Arabic and Sanskrit. They were sold by his widow to the Radcliffe Trustees and entrusted to the Bodleian Library in 1872. 4 James outlived them all. William (1784-1835) was assassinated in Delhi; Edward (1786-1813) died in St. Helena;

WRITER

AND

ARTIST

1783-1856

133

menace and the possibility of war. The outbreak of the Crimean War came as no surprise. In what was with the Foreign probably his last communication of his Office Fraser, drawing on the experience 1837 travels, suggested that a canal should be cut between Czernavoda on the Danube and Constantza on the Black Sea in order to free shipping from Russian interference on the lower Danube. Fraser died on 23 January 1856 at Moniack and was buried in the nearby churchyard at Kirkhill. In the garden of his home there still stands a simple granite obelisk, erected by his mother, recording his death and that of his four brothers, all of whom had died in foreign parts. James had been a devoted son, corresponding and regularly with his parents from London abroad, and helping them financially. Though his romances have long since ceased to be read, his travel books remain an important source of information. Russia's Politically, he over-played territorial ambitions, while some of his ideas were neither sound nor practical. He is best remembered as an intrepid but careful travelller who, through his curiosity and diligence, contributed much to contemporary knowledge of those lands between the Indus and Bosphorus which interested him so much.

ACKNOWLEDGEMENTS I am grateful to the following for permission to quote from papers whose copyright they hold: Her Grace the Duchess of Buccleuch (Sir John McNeill papers), Mr. Malcolm Fraser (Fraser of Reelig papers), the Controller of H.M. Stationery Office (papers in the Public Record Office and the India Office Library and Records) and Messrs Cassell plc. (M. Archer and T. Falk's India Revealed). My thanks, too, to the staff at the Public Record Office, the Scottish Record Office, the India Office Library and Records, the Geological Society, and, above all, to Mr. Malcolm Fraser for their help and kindness during my researches.

Alexander (1789-1816) died near Delhi; George (1800-42) died at Aurungabad. 5 F/B46. Alexander Fraser to his mother. Hoogly, 30.1.1815. 6 Journal of a Tour throughPart of the Snowy Range of the Himdla Mountains (London, 1820), p. vi. 7 Mildred Archer and Toby Falk, India Revealed. The Art and AdventuresofJamesand WilliamFraser 1801-35 (London, 1989), p. 35. 8 Ibid., p. 9

9 F/V23. JBF to his father. Calcutta, 4.10.1819. to his brother William. Calcutta, May 1820. 10 11 F/V58. JBF to his mother. FV24. JBF Lucknow, 2.8.1820, and his father. Bombay, 30.3.1821. '2 F/B348. JBF to his brother William. Bushire, 21.8.1821.

134

JOURNAL

OF PERSIAN

13 Ibid. Shiraz, 6.10.1821.

Ibid. Isfahan, 20.11.1821. 15 FIB93. Elphinstone to JBF. Bombay, 22.1.1822. 16 Fraser, Travels and Adventures in the Persian Provinces of the SouthernBanks of the Caspian Sea (London, 1826), p. 2. 17 JBF's half-finished water colour of the Shrine courtyard is among some seventy sketches and water-colours done on this journey, in his outsize sketch book at Reelig House. is Curiously, he makes no mention of the oil seepages noted by earlier travellers near Daliki on the Bushire-Shiraz road. Among his papers (FIB93) is a note about a reported naphtha sping near Shustar. 19 Travels and Adventures,etc., pp. 325-6. 20 F/B421.JBF to his wife, Tehran, 20.4.1834, on hearing that his friend McNeill had lost a fourth child: "I have often thought that the pleasure of children is more than balanced by the pain and anxiety that is incident to having them ... I declare to you that I am half tempted not only to be satisfied, but to bless God we have none". 21 Travels and Adventures,etc., p. 367. 22 FIV24.JBF to his father. Shiraz, 7.10.1821, and F/B440,JBF to his brother William. Mashad, 16.2.1822. 23 FlB97. JBF to his father. London, 12.5.1833. 24 PRO/FO60/33. Palmerston to JBF. FO, 4.12.1833. 25 FIB97. E. S. Fraser to JBF. Moniack, 17.5.1833. 26 Fraser, A Winter'sJourney(Tartar)from Constantinopleto Tehran (London, 1838), 2 vols., I pp. 253-4, 417-18. 27 F/B92. JBF to Grant. Tehran, 29.4.1834, and 31.7.1834, and Tabriz, 20.9.1834. This bundle also contains draft memoranda sent by JBF to Charles Grant. 14

STUDIES

28 PRO/FO60134. JBF 29 Ibid.

to Palmerston. Baghdad, 12.11.1834.

30 PRO/FO60/35. Memorandum by Wm. Cadell, Board of Control, 18.12.1834. 31 Foundationsin the Dust (Pelican edn., Harmondsworth, 1955), p. 101. 32 A Winter'sJourney, etc., I, p. 419. 3 Travels in Koordistan,Mesopotamia,etc. (London, 1838), 2 vols., II, p. 309. 34 FIB82.JBF to his father. Constantinople, 12.4.1835. s5 PRO/FO60/38 for all but no. 3. 36 FIB440. JBF to his brother William. Tabriz, 3.8.1822. 37 FIB272. JBF to Palmerston. London, 17.9.1835. 38 F/B100 and H. Rawlinson, England and Russia in the East (London, 1875), pp. 53-4. 9 PRO/FO60/54. JBF to Palmerston. London, 11.7.1836, and JBF memorandum, 27.6.1836. 40 SRO/GD371/277/1-3. JBF to McNeill. Malvern, 26.7.1850. 41 SRO/GD371/121/22. Cartwright to McNeill. Constantinople, 12.3.1838. 42 F/B271 contains several of his drafts. 43 FIB100. Backhouse to JBF. FO, 23.6.1838. 44 F/B16. JBF to his mother. London, 15.7.1836. 45 F/B294. JBF to Mrs. McNeill. Moniack, 7.10.1838. 46 Ibid., 26.7.1838. 47 PRO/FO60/68. JBF to Palmerston. London, 15.7.1839. 48 Obituary quoted in The Gentleman'sMagazine (March 1856), p. 307. 49 F/B539. Clarendon to JBF. FO, 18.9.1853.

BRITISH INDIAN VIEWS (NINETEENTH AND EARLY TWENTIETH CENTURIES) OF THE LATER FOLLOWERS OF THE RAUXANIYYA SECT IN AFGHANISTAN AND NORTHERN INDIA By Sergei Andreyev St. Hugh'sCollege,Oxford

The Rau'ini movement, which spanned the period from the second half of the sixteenth century to the first four decades of the seventeenth century, flourished in Afghanistan and Northern India. The movement had a dramatic impact on the cultural and social life of the mediaeval Pashtun tribes. The Rauxani movement was headed for many years by Bayazid Ansari, also called Pir-i Raux'in ("Light Mentor"/"Mentor of Light") by his followers and Pir-i Tirik ("Dark Mentor"/"Mentor of Darkness") by his adversaries, and this is what gave the movement its name. This movement can be considered as the strongest upsurge of the original Pashtun thought, which was extremely free and radical.1 The Raux'niyya are still almost unstudied. Since 1810, when the first article on the Raux'ani movement was published by John Leyden, a few scholars have written about the movement, mainly repeating the ideas and data contained in the first ever article on the problem. Leyden's article, as well as the works of his successors, was based mainly on mediaeval Indian sources written in Persian, with some examination of Akhiind Darwiza's books, while the writings of the followers of the Rau'aini sect itself remained unstudied. The first scholars who began to use Rau*dini sources were Kh. A. Nizami (Nizami, 1960; Nizami, 1989), S.A.A. Rizvi (Rizvi, 1961; Rizvi, 1965-8), Muhammed Shafi (Shafi, 1961) and Tariq Ahmed (Tariq Ahmed, 1982). However, they have concentrated on the

which contain some indirect information of relevance

are the Tadhkirat al-Abrdr wa' 1-Ashrar by

Akhuind Darwiza2 and the Tdrikh-i Murassa' by Afzal Khan Khaffak.3 However, it is possible to find some relevant information in British Indian sources of the late nineteenth-early twentieth centuries. Later British data on the followers of the Raux'ani doctrine are very controversial,4 since British Indian officials were in the main not familiar with the history of the Rau•aini movement and did not focus on this specific problem. However, given the considerable amount of information provided on this subject in administrative reports, it seems possible to assume that this problem was regarded as rather important. In the absence of evidence based on modern field-work, the problem of later adherents of the Rau'ini sect has to be left at present unsolved. Some British colonial sources (Ibbetson, 1883; Ridgway,

1910; Gazetteer of the Peshawar District,

1898) tried to associate the Shiah groups of the Orakzai, Bangash and Turi tribes of Pashtuns5 with the Raux'ini doctrine. The Pashtuns are known to be mainly Sunnites. At the end of the nineteenth century, however, there were 10,591 Shiites counted in the Koh-t district (the main area of the Orakai and Bangash's residence [S.A.]).6 They were confined to a portion of Samilzai and Hangii (administrative units in the Kohait district [S.A.]) and to the adjoining region of Raushdn The Orakzai clans of the Shiah faith of the Tirdih. (also Hadlnama-yi Bayazid study known as Hdlndma-yi Pir-i Dastgir or Hadlnama-yi belonged to the Muhammad Khel tribe. They were BayazidAnsdari)by 'Ali Muhammad Mukhlis, a biog- the following: the Sipaihs* or Sipayas* (perhaps a raphy of Bayazid Ansarli and his descendants in derivation from sipdh or sipdhi [S.A.]), Maini Khel, Persian, having left aside the Pashto sources. None Bar Muhammad Khel, 'Abd al-'Aziz Khel and a of these historians has given an account of the later portion of 'Ali Khel, Tazi, Bar And* and Lar And*; followers of the Raultini creed. Scholars working three hamsdya or vassal sections of the Tira-h Sturi* in modern Afghanistan and Pakistan concentrate (alias Afzal Khel) are also Shiites (Rose, 1914, vol. on the mediaeval history of the movement, its III, p. 177). All these tribes resided in Tirah and doctrine and literary heritage, and do not even bordered on the Shiah portion of Samilzai. Rose, noted that, according to the tribal pedigree, all mention its later followers. There is no direct evidence for the survival of people of the Shiah Muhammad Khel were descenfollowers of the Rau'iini doctrine after the general dants of Biyazid Ans~iri (loc. cit.). These Orakzai defeat of the Rau•aini movement in the first half of clans are said to have been converted by the Tirih the seventeenth century. The only original sources Sayyids about the beginning of the nineteenth 135

136

JOURNAL

OF PERSIAN

century (Kohat Gazetteer, 1884, p. 69; White King, 1984, p. 151). The Shiah Bangashs of the Samilzai clan were probably converted a little earlier, but could not freely admit to being Shiites during rule from Kabul (Kohat Gazetteer, 1884, p. 69). All the Bangash of the Upper Kur(r)am(a)* valley and a portion of Samilzai and Hangui were Shiites, while those of the Lower Kur(r)am(a)* and the remainder of the Kohat district were Sunnites (Ridgway, 1910, p. 76). All the Turi Pashtuns (6,000 in number in 1910, resident in the Kur(r)am(a)* valley) were reported to be the disciples of the Tirah Shiah Sayyids (Ridgway, 1910, p. 159). All the above-mentioned Orakzai Shiah tribes were under the religious influence of a Sayyid7 family residing at Kilal*, and generally known as the Tirah Sayyids. The Sayyids of Pir Khel and Mansuir Khe- (probably the Bangash clans [S.A.)8 were said to be descended from Bayazid Ansari (Pir Tarikhi in the Gazetteer) (Kohat Gazetteer, 1884, p. 69). Raverty stated that the descendants of the famous Raua-ani poet Mirza Khan Ansfirl, a grandson of the founder of the Rau'aini movement, still (sc. in the second half of the nineteenth century [S.A.]) dwelt in the Tirah district, amongst the clan of Miyan Kheel (probably the Bangash clan [S.A.]), "and have the repute of being quiet and well behaved. There is generally one of the family who follows the life of an ascetic; and is allowed, by the simple people, to have the power of working miracles" (Raverty, 1978, p. 55). However, White King argued that the Tirahi Sayyids, the leading Shiah Sayyid clan in the area in from a Shiah priest question,_ were descended Fakhr-i 'Alam, who came from Gardez about the middle of the seventeenth century and settled in Kur(r)am(a)*. His grave is in Kirman (Kur(r)am(a)*), and is held in much veneration by the local Shiites. Another separate branch of Shiah Sayyids, descendants of Shah 'Isa Bukhari, settled in Marai* from Peshawar. There was another Family of Shiah and Sayyids, known as Tal(l)a Nm(a)(n)(d)zi*, in the Bar Takki Nm(a)(n)(d)zi*9 Muhammad Muhammad Khel country (White King, 1984, p. 150). As one can see, neither Bayazid Ansari nor his known descendants were mentioned by White King among the founders of these Sayyid families. The only information available on the conversion of the Pashtuns to Shiism in the Middle Ages is the following. During the reign of the Mughal emperor Jahangir and ca. 1620 A.D., Sayyid Mir 'Aquil settled in the Marai* District with the Bar Muhammad Khel and converted many people to Shiism. These people were driven by the oppressive rule of Malik Toi, the Khan of the Orakzais, 0 to seek asylum with Mir 'Aqiil. Consequently, Malik ToF attacked Mir 'Aqiil, who was slain, and whose family, on their ejection from Tirah, settled in the Samilzai valley of

STUDIES

Kohat. After a time, this family won over the Mishtis*, Mulla Khels and Shaykhans, who, with the Bar Muhammad Khel, seceded from the rule of the Orakzai Khan (White King, 1984, pp. 150-1). Unfortunately, White King did not mention the source of his information; perhaps he relied on the oral tribal tradition, as many other authors often did. According to the Gazetteer oJ the Kohat District (1884) there were no acknowledged members of the Rau'ani sect remaining (Kohat Gazetteer, 1884, p. 70). It is noteworthy that White King, Deputy Commissioner of K6hat and in political charge of the Orakzais (period of service 1897-1900), also never mentioned the Raujxani doctrine in connection with the then contemporary Orakzais. Some authors did not however agree with the statement of the Gazetteer. Ibbetson and Ridgway the contemporary considered Sayyid, Orakzai, of the Bangash and Turi Shiites as followers Rauxani doctrine (Ibbetson also wrote about a few adherents of the Rauxani doctrine scattered along the Trans-Indus Salt Range from the Kohat District to the Dera Isma'il District) (Ibbetson, 1883, p. 146; Ridgway, 1910, p. 31). The authors of the Gazetteer of the Peshawar District generally supported this view: "At the present time (1897-8 A.D. [S.A.]) its [the Rau~iani sect (S.A.)] tenets are professed only by the immediate descendants of the founder in Tirah and Kohhatand by some of the Bangash and Orakzai. The ancestors of those members of the Orakzai tribe, who are popularly known as Shiites were probably of this sect." (Peshawar Gazetteer, 1898, p. 60). According to Ibbetson, these people were rather called Shiites because they were rejected by the Sunnites than because they followed the Shiah doctrine (Ibbetson, 1883, p. 147). Sir Olaf Caroe disagreed with the previous statements and, referring to his experience with the Orakzai and Bangash Shiites, stated that he had never heard this identification publicly made. According to him, there was no affinity between the Shiah beliefs professed in this region and either the Rau3'iani religion or the Khariji and Isma'-li sects, with which the followers of the Rauiiani movement were sometimes associated. Caroe admitted that the final discrediting of the Rauxiani movement generated a political reason for covering up these matters, still operative to some extent in his own day; moreover, even if the suppressed followers of the Rau~aini doctrine had taken cover as Shiites in the Mughal times, their descendants, forgetting their doctrinal origin, would have tended more and more to have merged with the general Shiah confession (Caroe, 1964, p. 202-3). to Leyden, the adherents of the According who still were confined to sect, existed, Rau3aini the wildest and most inaccessible districts, concealing their books and their tenets with equal care.

BRITISH

INDIAN

VIEWS OF THE RAUX2ANIYYA SECT IN AFGHANISTAN

Amir Muhammad, a native informant from Peshawar, told Leyden that some followers of Bayazid Ansari were still to be found, both in Peshawar and Kabul, but that they were reckoned to be even more numerous among the Yuisufzai tribes (Leyden, 1810, p. 79). However, Bellew, the author of the detailed monograph on the Yuisufzai tribe written mainly and from native from his personal observation information, argued that the Rauiiani creed disappeared among the Yuisufzais after Bayazid Ansaari's death (Bellew, 1864, p. 73). According to Leyden's informant, the then contemporary followers of the Rauxani doctrine were supposed to hold secret meetings in Peshawar, by night, in an ancient edifice, with a dome, where Bayazid Ansari formerly resided, and at which the pious Muslims, as they pass by the ruins, generally cast stones in token of their abhorrence, accompanied by fervent imprecations and curses on the founder of the Rauxani movement (Leyden, 1810, p. 79). However, the question arises, who were these followers of the Rauxiini sect descontemporary cribed by the British authors?. There is some very limited information in the original sources about different trends of the Rauxiini doctrine professed by the YUisufzais, who were described by Leyden as followers of the Rau'iani the then contemporary

i

Note on transliteration: The standard Iran system is used for the transliteration of Persian and Arabic words. For the additional Pashto characters the following system is used:

0e

t-

6

x-

)

Raui~inis.

s

.

Words with doubtful and/or various transliteration are marked by * 2 This voluminous book ascribed to Akhiind Darwiza was written in Persian in 1021/1612-13. In this treatise, the author gives an account of orthodox and heretical teachings of Darwiza's time. He condemns the teaching of Baiyazid Ansaariand devotes a special section to the description of the Rawifini tenets and activity of the followers of the movement. Unfortunately, his account of the Rawxani movement is extremely unsystematic and even contradictory. In the concluding part of the Tadhkirat it is stated by Darwiza's son Mulla 'Abd al-Karim, better known as Karimdad, that the book was dictated to him by Darwiza and he wrote it down, arranged the text in proper order and entitled it (Darwiza, Tadhkirat,fol. 202b). 3 The Tdrikh-iMurassa' is a voluminous history of the Pashtuns written in Pashto by Afzal Khan Khaffak (lived approximately 1072-3/1661-3 to 1161/1748), a grandson of the famous Pashtun poet and warrior Khushhal Khan Khatfak. The narration is brought up to the year 1136/1723. The Rawiiini movement is mentioned in the fifth and sixth daftars of the book. 4 British colonial sources are highly accurate in their descriptions of geographical conditions and tribal statistics; however, in the course of historical narration they rely mainly on oral traditions, which sometimes contradict the original written sources. Yet, given the lack of proven data, the student of the Pashtun tribal affairs is forced to use these

INDIA137

sect. According to the Tadhkirat al-Abrar wa 'l-Ashradr "One Miyan 'Isa, a native of Peshawar and follower of Bayazid, alias Pir-i Tarik, the founder of the Raushani heresy, went to Swat (the main area of the Yiisufzais' residence [S.A.]), and preached its docsuccess amongst the trines with considerable Yiisufzais. Those who became converts to his doctrine were called 'Isa'i." (Akhiind Darwiza, fol. 80a [Plowden's translation in (Plowden, 1893)]). of this Plowden wrote in 1893 that descendants in the of Peshawar still resided 'Isa'i (Plowden, city 1893, p. VI). According to the later evidence of the Tdrikh-i Murassa', there was a distinct difference between the genuine Rauxainis and the followers of 'Isa: "After that the pride and impiety of the YUisufzais increased; heterodox practices and sects sprung up among them. Some, together with their head-men, became 'Isa'i (Christians in English translation [S.A.]); others followed the tenets of Pir-i Rauxan, whom people call Pir-i Tarik." (Afial Khain, 1893, p. 221; English translation, p. 183). This is the only information available about the so-called 'Isa'i sect. It seems therefore at present impossible to define any more closely the original affiliation of the followers of 'Isai and their connection with the later

-

-d-

AND NORTHERN

6

8

9

10

sources, although certainly with a reasonable degree of reservation. Such modern authors as Sir Olaf Caroe and Akbar Ahmed have confirmed the existence of Shiah clans among various Pashtun tribes (Ahmed, 1984, pp. III, VI, VII, X; Caroe, 1964, pp. 26, 202-3, 308). It seems that these Shiites were in a difficult position living among their Sunnite neighbours. Ibbetson wrote: "the Sunnite Pashtuns abominate the Shiites and all their works, insomuch that in old days a man hardly dare admit that he belonged to that sect" (Ibbetson, 1883, p. 146). Perhaps this tradition of dissimulation (taqiyiya)is one of the reasons for the discrepancy in defining the real religious affiliation of the tribes in question. In British colonial sources, Muslim clergy were often called Sayyids regardless of their origin. It is noteworthy that, according to Ibbetson, "many of the tribes (perhaps groups/clans, there are no Pashtun tribes claiming Sayyhid origin [S.A.]) which claim Saiyad origin are Shiahs" (Ibbetson, 1883, p. 201). If we read the word nmazi given by White King without transliteration as nmdndzi(praying [man]) (z and dz are often confused in Pashto, especially in Persian loan-words, nmanndz, sometimes nmuindz,from Persian namaz) we get an interesting name of this Sayyid family-"The praying people from Tdl(l)a" (Ta(a)l(l)a is a geographical name especially popular in the country of the Haz~iraPashtuns (Adamec, 1972-85, vol. VI, p. 778) and "The praying people from the Takki" (Takki (Zam) is a river not far from Kainiguram in Wazirist-n (Caroe, 1964, p. 405)). Malik To_, once a committed follower of the Rawxaini doctrine, left the movement about 1620 and joined the Mughal forces. It seems that his conflict with the Shiah preacher in the Orakzai had a political character or was caused by the traditional Sunnite-Shiite hostility. At all events, there is no mention of the Rawxaniyya in White King's narration.

138

JOURNAL

OF PERSIAN

Bibliography Adamec, Ludwig, W., 1972-85. Historical and Political Gazetteerof Afghanistan, Graz, 1972-85. Afial Khan, 1893. Tdrikh-i-Murassa',in "The Kdlid-i Afghdni", Being Selectionsof Pushto Prose and Poetryfor the Use of Students; Compiledand Publishedby the Revd. T. P. Hughes C.M.S., Afghan Mission, Lahore, 1893. English version: Tdrikh-i MuraSSa' or "The Be-Jewelled Chronicle"Composedby Afzal Khan Son of Ashraf Khan Son of Khushhal-KhdnKhattak, in "Translationof "Kdlid-iAfghani" the Text Book for the Pakhto Examinations, with Notes, Historical, Geographical, Grammatical, and Explanatory by T. C. Plowden, Lahore, 1893. Ahmed, Akbar, S., 1984. Introduction to King's Orakzais, in L. White King, The Orakzai Countryand Clans, Lahore, Vanguard Books Ltd., 1984, being a reprint of L. White King, Monograph on the Orakzai Countryand Clans. A GovernmentReport, Lahore, 1900. Akhiind Darwiza. Tadhkirat al-Abrar wa'l-Ashrar,manuscript of the British Library, shelf-mark Or 222. Bellew, H. W., 1864. A General Report on the Yusufzais in Six Chapterswith a Map, Lahore, Government Press, 1864. Caroe, 0., 1964. The Pathans, 550B.C.-A.D, 1957, London, 1964. Ibbetson, D. C. J., 1883. Outlines of Panjab Ethnography,Being Extracts from the Panjab Census Report of 1881, Treating of Religion, Languageand Caste by Denzil CharlesJelf Ibbetson,of Her Majesty'sBengal Civil Service, Calcutta, 1883. Kohat Gazetteer, 1884. Gazetteer of the Kohat District, 1883-4, Calcutta, 1884.

STUDIES

Leyden, J., 1810. On the Roshaniah Sect and its Founder Bayazid Ansari, in Asiatick Researches,vol. XI, 1810. Nizami, Khaliq Ahmad, 1960. Hal Nama-i-BayazidAnsari, in Fikru-Nazar,January, 1960. -, 1989. Akbar and Religion, Delhi, Idarah-i-Adabiyat-i-Delhi, 1989. Peshawar Gazetteer, 1898. Gazetteer of the Peshawar District, 1897-98,

Lahore,

1898.

Plowden, T. C., 1893. Translation of "Kdlid-i-Afghani"the Text Book for the Pakhto Examinations, with Notes, Historical, Geographical, Grammatical,and Explanatoryby T. C. Plowden, Lahore, 1893. Raverty, H. G., 1978. Selectionsfrom Pushto Poetry, Lahore, 1978; reprinted from idem, Selectionsfrom the Poetry of the Afgans, London 1862. Ridgway, R. T. I., 1910. Pathans, Handbookfor the Indian Army, Calcutta, 1910. Rizvi, S. A. A., 1961. "Available Works of Bayazid Roshanai", in Procs. of the Indian History Congress,1961. 1965-8. "Rawshaniyya movement", in Abr-Nahrain, VI, -, 1965-6, 63-91; VII, 1967-8, 62-98. Rose, H. A., 1914. A Glossaryof the Tribesand Castes of the Punjab and North-WestFrontier Province, vol. III, Lahore, 1914. Shafi, Muhammed, 1961. Art. BdyazidAnsari, in EI2. Tariq Ahmed, 1982. Religio-PoliticalFerment in the N.W. Frontier During the Mughal Period. The Raushaniya Movement, Delhi, Idarah-i-Adabiyat-i-Delhi, 1982. White King, L., 1984. The OrakzaiCountryand Clans, Lahore, Vanguard Books Ltd., 1984, being a reprint of L. White King, Monograph on the Orakzai Country and Clans, A Government Report, Lahore,

1900.

PRODUCTION OF SUGAR IN IRAN IN THE NINETEENTH CENTURY By A. Seyf StaffordshireUniversity

It is generally true that the economy of Iran in the nineteenth century was predominantly agrarian based with an insignificant non-farm sector. There is no doubt that the underdeveloped state of the non-farm sector had its roots in the internal class structure of the society, but, it is also true that, especially in the late eighteenth and during the nineteenth century, external factors contributed to its underdevelopment. Two interrelated problems that became apparent in the course of the nineteenth century may be summarised in the following terms: For a variety of reasons, production techniques in the economy remained primitive. This was partly due to the predominance of despotism which acted as a disincentive, curtailing investment. The primitive techniques undoubtedly adversely affected the surplus generation. It is the utilisation of this relatively smaller surplus which was distorted by the predominant despotism for and further the purpose of reproduction regeneration. Given an open-door commercial policy which had largely been imposed on Iran by her main trading partners, namely Britain and Russia, this sector had to face fierce foreign competition. This was despite the fact that the economy, and particularly its non-farm sector, was least able to compete. It is not surprising, therefore, to see a gradual decline of activities in that sector during the course of the nineteenth century. There is little doubt that relatively speaking, the Iranian economy was more dependent on imports at the end than at the beginning of the nineteenth century. On the face of it, there may be nothing wrong with this type of development. There may be some who might even argue that the economy would have benefited from greater integration into the world capitalist system. The crucial question, however, is: how did the economy transform itself in order to accommodate these changes? The available evidence shows that before the discovery of oil, the agricultural sector had to feed the population and ensure the financing of an increasing volume of imports. If there had been sufficient investment in agriculture, or, if there had been the necessary reforms to make this sector

more efficient and productive, that could have led to a different situation. That was not to be the case. Agriculture remained as backward as ever, with little or no investment and no reforms. It is, therefore, possible to argue that this pattern of development, far from being a natural course of progress, originating from the forces from within, was to a large extent a by-product of Iran's greater participation in international trade, i.e. caused by forces from without. By the end of the nineteenth century, the decline of handicrafts produced two sets of results: (i) An increasing number of Iranians were effectively forced to emigrate, mainly to the southern provinces of Russia, but to elsewhere also, in search of employment and more tolerable conditions of life.' The surplus that this segment of Iran's labour force could have produced, had they been employed in the economy, was, therefore, siphoned out of the country. (ii) Agriculture, while having to operate within an overall constraint of reduced surplus for investment, had to cater for a larger proportion of the labour force. It is likely that less surplus was generated within that sector itself, thus, perpetuating the primitive techniques of production. Concerning the non-farm sector of the economy, the only exception was carpet making, which appears to have expanded.2 In this short article, I would like to examine sugar production in Iran in the nineteenth century. Of course, in examining sugar production in Iran, distinction must be made between the different types of sugar, i.e. boiled, and refined or centrifugal sugar. In this case, the available evidence seems to suggest that sugar production in Iran during the period under review was a combination of both, and both were subjected to competition from imported sugar, predominantly refined. Iran imported a very small amount of boiled sugar from India and this appears to have been the only source supplying foreign boiled sugar into the Iranian market. Although we have no data to substantiate this, the impression is that during the first half of the nineteenth century the greater share of the locally produced sugar was boiled sugar, whilst in the second half, despite the overall

139

140

JOURNAL

OF PERSIAN

the greater prodecline of sugar production, was Whatever the refined. portion type of sugar, the technique of production was most primitive. The following some of the passage describes methods used to produce boiled sugar in Mazandran. In 1843, Consul Abbott gave the following description: The rude machinery for expressing the juice from the cane would be difficult to explain ... the sugar cane being cut into short pieces is placed upright in a wooden receptacle followed in the form of an inverted cone. A pole sharpened at the end is made to turn in this hollow pressing the cane against the side of the receptacle. The juice escapes into a trough from whence it is conveyed to a cauldron where as it boils a man stirs it constantly with a pole. It is afterward removed to cool but undergoes a second boilings-and lastly being strained through linen or calico it poured into trays. On cooling it acquires the consistency

of a cake ...3

added that treacle was made from the superior canes, and from 15 mans (6.5 lbs each) of juice about 5 mans of sugar were obtained, and "one machine," Abbott continued, "expresses about 30 mans" of juice per day.4 It follows that the was about 65 lbs. daily boiled sugar production Fifteen years later, in 1858, Consul Mackenzie He

wrote

that the cane

...

is chopped up into small pieces which are placed upright in a primitive mill which is on the principle of a pestle and mortar. The pestle however does not hit the canes, but is rubbed against them ... and the moving force is an ox. The mortar is usually a portion of the hollowed trunk of a tree in the form of an inverted cone; one end of which is sunk in the earth ... the pestle is a sharpened piece of wood fixed in a horizontal beam resting on a pivot which turns round about the mortars and to which the ox is attached ... the boiling apparatus is just as barbarous, there is a cavity for fire dug in the ground, above which is a large cauldron in which the juice undergoes two boilings-it is then strained.5 In his view, "it must leave a lar e portion of the richest juice in the cane trash". He spoke of a state-owned "sugar refinery" in Barforush, but gave no further information as to its condition or mode of operation. Nevertheless, he added that "the consequent obligation under which the peasants have been placed to take all their sugar there for sale" led to a negative response from the peasants, who reduced "the cultivation of the cane" and, as a result, "the price of sugar has increased".7 The only information on the refinery was that the superintendent told him: If furnished with sugar cane on its primitive state, he could manufacture 5mans of refined sugar from 10 [mans] of raw sugar, but as he was under the necessity of purchasing the raw sugar from the villagers who make it of very inferior quality, he could not manage to extract more than 5 mans from 20 ....

STUDIES

On the basis of this evidence, it then follows that 5 mans (32.5 lbs) of refined sugar, to produce 20 mans (130 lbs) of boiled sugar were needed, which would have taken any one of these production units two days to make. Small wonder that could not sugar produced by such methods compete with those produced in the large factories of Moscow or Marseilles! in former times, Iran seemed to Nevertheless, produce and even refine the greater part of sugar consumed locally by itself. For reasons already however, the imported sugar increasmentioned, ingly replaced home produce. It may even be argued that in some parts of Iran, the decline of irrigation facilities affected the cultivation of sugar cane and influenced the production of sugar in the country. In this respect, it has been reported that "sugar cane was much cultivated in Khuzistan" and "had its water from the Karun River by the means of canals cut from the right bank some distance above Shushtar, and from the Diz River by canals cut from the left bank, near the town of Dizful", but with the decline of Jundi-Shapur (the principal town of the province) in the thirteenth and fourteenth century, "The canals were neglected, and the cultivation of sugar cane necessarily ceased".9 Even as late as 1889, it was asserted that, "hundreds of millstones or wheels, formerly used for squeezing the juice out of the cane, are laying about in all direction".1 Referring to sugar cane cultivation in Mazandaran, the report went on, "in some towns of Persia, principally Yazd and Isfahan, Jaru[?] raw sugar was, up to a few years ago, refined and made into loaf sugar ... [But] the imported loaf sugar very cheap, sugar refinery in Persia becoming ceased to be profitable".11 To demonstrate this point, reference can be made to a report prepared by Consul Abbott in 1849 in which, he asserted that there were "ninetysix small factories to refine sugar in Yazd and thirty-five similar factories in Isfahan".12 For the town of Yazd, the total quantity of raw sugar refined in those factories was about 53,000 cwt per annum and he went on: "This article (sc. refined sugar) is sent to almost every part of the country and is used by many people in preference to that of Whatever the Europe from religious scruples".13 motives of the consumers, the fact remains that ninety-six small factories of Yazd produced a little less than 150 cwt refined sugar on average daily in around 1850. In 1893, Consul Preece put down the number of refineries in Yazd at fifteen14 and a year later, at twelve for Isfahan.s15 Furthermore, as we have no data on the total production of sugar in Iran (boiled or refined), nor is there any evidence suggesting any specific increase in the consumption of sugar, we are led to believe that the enormous increase in the amount

PRODUCTION

OF SUGAR

IN IRAN

of imported sugar may serve as a proxy indicator of the decline in local production of sugar, thus giving support to the argument. in Despite primitive techniques of production Mazandaran, that province supplied Gilan and Tehran with locally produced sugar during the 1850s. I'timad-us-Saltaneh estimated the quantity of refined sugar, "of very good quality" exported to the latter place at 12,000 mans (about 150,000 lbs) per annum.16 In the latter part of the century and especially towards the end, however, the Mazandaran province from itself was totally supplied Russia.17

J. P. Ferrier, writing on the trade of Khorasan in the mid-1840s, argued that "sugar, which is brought [into Khorasan] from the refineries of Yazd, forms a considerable article of trade".18 But in the last quarter of the century, Khorasan was entirely supplied by Russia, too. It is probable that Mashhad (the provincial capital of Khorasan) became a distribution centre for Russian sugar, as the import of sugar from Russia seemed to increase five-fold during 1889-1906:

1

TABLE

The Value of Refined Sugar Importedinto Khorasanfrom Russia, 1889-190619 Annual average

Value in ?

Index 1889-91 = 100

1889-1 1892-4 1895-7 1898-1900 1901-3 1904-6

39,800 24,000 46,700 87,000 161,000 193,000

100 60 117 218 404 485

2 shows the estimated value of annual of refined sugar into Tabriz. Until the late imports 1870s, it was predominantly European (namely, French) sugar, but for the last quarter of the century the market was supplied by Russia. Furthermore, while during 1834-8, the annual average quantity of sugar imported from India via Table

TABLE

2

IN THE

NINETEENTH

141

CENTURY

and Arabian Gulf was about the Persian 24,752 cwt21 (the type of sugar was not specified, but it was probably boiled), for the period between 1868 and 1906 that average was more than double, i.e. 58,038 cwt, to Iran alone. The greater part of this sugar was "foreign produce" (British and colonial) and almost all of it was refined. For some of these years, the breakdown of different types of imported sugar is given below: TABLE

3

The Breakdownof Sugar Importedfrom India, 1881-9522 (quantity in cwt) Annual average 1881-3 1884-6 1887-9 1890-2 1893-5

Indian Boiled Refined 751 697 899 1,835 1,863

57 142 154 174 237

Foreign Refined 50,451 107,078 79,956 77,967 104,798

As can be seen, the import of boiled sugar was negligible. It must be added that on the basis of Indian official statistics, the import of boiled sugar from India into Iran was also periodic. The whole of sugar imported during 1876-80 or 1901-6 was said to have been refined. However, the following table compiled from this source shows the quantities and values of sugar imported from India into Iran during this period, i.e. 1868-1906. TABLE

4

The Quantities and Values of Sugar Importedinto Iran from India, 1868-1906 Annual average

Quantity in cwt

Value in ?

1868-72 1873-7 1878-82 1883-7 1888-92 1893-7 1898-1902 1903-6

34,843 41,228 42,294 94,616 78,309 94,283 45,285 33,524

45,319 53,695 71,718 131,646 97,788 126,219 52,127 36,318

The Estimated Value of Refined Sugar Importedinto Tabriz20 Annual average

Value in ?

1850s 1860s 1870s 1880s 1890s 1900-7

100,200 123,632 74,696 44,692 78,641 315,355

It must be borne in mind that, as we have already asserted, towards the end of the nineteenth century Russia almost monopolised the Iranian market, and in many parts, for example Azerbaijan, the Caspian provinces and Khorasan, Russian refined command of the market. sugar held complete from Consul Yate noted that Khorasan, Reporting "at present (1896) there is no competition between Russian sugar and sugar of Indian and French

142

JOURNAL

OF PERSIAN

in Khorasan", and that "The only manufacture article procurable in the market is that imported from Russia".24 It is interesting to note that, during the first half of the nineteenth century, no sugar was apparently imported from Russia into Iran, and it was even suggested that, during the early 1840s, some sugar had been exported to the former country from the latter.25 Whether it was native produce or that originally imported from Europe, or whether it was refined or boiled, is impossible to say and is of because the fact remains secondary importance, that no Russian sugar was brought to Iran. During these years, and indeed during the early decades of the second half of the nineteenth century, the sugar imported into Azerbaijan came from Europe (mainly from France)26 whereas by 1879 "The bulk of it [sugar] is now supplied from Russia ... the government grant a bounty of ... one pence per pound weight of sugar [more than ?9 per ton] .. ."27 In addition to this shift, as we have already seen, the quantity and value of imported sugar into Azerbaijan also increased. Compared with 1859, when the market in Azerbaijan was fully supplied by French and Dutch the value of produce,28 imported sugar increased by 190 per cent from about ?106,000 in 1859 to more than ?305,000 in 1907, and the whole lot came from Russia.29 The sugar trade in the Caspian provinces That is to say, it exhibited similar development. showed the decline of local produce and the increasing importation of Russian sugar. We have some relevant information from the province of Gilan, which further demonstrates the point. It was said that in 1858, ?17,364's worth of sugar was imported into Gilan, out of which twenty-six per cent, or ?4,500 was native boiled and refined produce [from Mazandaran, Yazd and Isfahan] and only seven per cent came from Russia,30 the rest being supplied by Europe. By 1874, the quantity to of imported and value sugar increased 31,696 cwt and ?89,600 respectively and yet it was wholly French produce brought to Gilan via Russia.31 For the last decade of the nineteenth table speaks for itself. century, the following TABLE 5

The Quantities and Values of Russian Refined Sugar Importedinto Iran via the Caspian Sea, 1892-1905.32 Year

Quantity in cwt

Value in ?

1892 1894 1895 1902 1905

383,339 467,882 516,728 729,590 894,780

504,670 727,815 915,700 643,154 783,959

STUDIES

The imported sugar solely came from Russia and were distributed to the interior, i.e. to the other Caspian provinces and to Tehran. As can be seen, in a space of thirteen years, the quantity of imported sugar from Russia into Gilan more than doubled. The increasing imports of sugar from Russia were, perhaps, an indication of in that rapidly expanding sugar production In the last of the nineteenth country. forty years century, sugar production in Russia rose 8.5 fold, from 41 million poods (36 lbs each) during the early 1860s to about 350 million poods in the late 1890s.33 Looking at the value of imported sugar into other trading routes, all this seems to confirm our conclusion that foreign produce increasingly replaced the native sugar. The table below demonstrates this point very clearly: TABLE

6

ComparativeTable Showing the Value [in Sterling] of ImportedRefined Sugar in 1863 and 1905.

Via Tabriz Via Kermanshah Via Bushire

1863

1905

Index 1863 = 100

150,000 7,000 6,000

320,151 114,706 126,636

213 1,639 2,111

Finally, let us recall that it was not only the native producers of sugar who, in view of their primitive and insufficient techniques capital, could not compete in the market and hence were driven out. The same thing happened to a Belgian Company which, upon obtaining the monopoly for the sugar making in Iran in 1891, established a reasonably near Tehran in 1895. factory well-equipped M. to A. costs of raw Jamallzadeh, "high According materials and unexpected forced it to expenses suspend operation in 1899".3 Lorini, an Italian economist at the turn of the century, referred to the above factors and added "The only buyers of beets were the Belgians; and as they wanted to set and impose prices on sharecroppers, conditions the latter preferred to revert to their grains, to barley and run the risk of the free market. ... Moreover, in fact it turned out that the natives do not know how to grow beet, sometimes watering them too much, sometimes too little, and so on ... 36 There can be no question that most of these factors contributed to the failure, but we are inclined to believe that the main factor was the fierce and not always fair competition offered by the Russians, who often went to a great length to capture a greater share of the market. On this aspect of the question, Consul Churchill remarked: "The price of sugar in the Persian market is regulated by

PRODUCTION

OF SUGAR

IN IRAN

prices in Russia, depending on the success or failure of the beet crops etc. When foreign competition is feared in Persian markets a bonus is offered, and other customs facilities afforded.... The sugar bounty has not, as far as I have been able to ascertain, been renewed, but there is a possibility of it being offered at any time".37 CONCLUSION In the preceeding pages, an attempt has been made to examine the production of sugar in Iran in the nineteenth century. It has been shown that its production appears to have suffered from importations from abroad. This was partly due to the primitive techniques used in its production, 1

2

4

5

7 8

9

See, A. Seyf, Some Aspects of Economic Development in Iran, 1800-1906, Ph.D. Thesis (University of Reading, 1982), pp. 161-4, unpublished. For further information see, idem, "Carpet Manufactures of Iran in the Nineteenth Century", Middle Eastern Studies, XXVI/2 (April 1990), pp. 202-13, and see also "The Carpet Trade and the Economy of Iran, 1870-1906, Iran, XXX (1992), pp. 99-105. K. E. Abbott, "Narrative of a Journey from Tabriz along the Shores of the Caspian Sea to Tehran (1843-44)" in UK Foreign OfficeArchives,251-40, p. 141 (hereafter, FO). Ibid., pp. 146-7. C. F. Mackenzie, Narrative of a Journey from Rasht ... through Mazandran to Asterabad (1858-59)", in, FO 60-245, pp. 90-1. Ibid., p. 90. Ibid., p. 91. Ibid., p. 120. Anonymous, "Cultivation of Sugar in Persia", Journal of the Societyof Arts (July 1889), p. 685.

1o Loc. cit.

"

12

Ibid., pp. 685-6.

K. E. Abbott, "Trade, Manufactories and Production of Various Cities of Persia, (1849-50)", in FO 60-165 (not paginated).

'1 Ibid. 14

J. R. Preece, "Journey made ... to Yazd, Kerman

...",

in UK

ParliamentaryAccountsand Papers, 1894, LXXXVII, p. 20 (hereafter, PAP). 15 Preece, Consular Report (hereafter CR), "Isfahan", in PAP, 1896, LXXXVIII. 16 Tdrikh-iMuntazam-iNdsiri, 3 vols. (Tehran, 1883), III, p. 224. 17 See H. L. Churchill, CR "Gilan", in PAP, 1904, C. 18s CaravanJourneys and Wanderingsin Persia, Afghanistan ... in 1845 (London, 1856), p. 124. 19 Compiled from, Maclean, CR "Khorasan" in PAP, 1890, LXXVI, 1892, LXXXVIII; Elias, CR "Khorasan", in PAP, 1893/94, XCX, 1895, XCII; Thomson, CR "Khorasan", in PAP, 1893/94, XCV; Yate, CR "Khorasan", in PAP, 1894, LXXXVII, 1898, XCVII; Temple, CR "Khorasan" in PAP, 1899, CI, 1900, XCV; Whyte, CR "Khorasan", in PAP, 1900, XCV, 1902, CIX; Minchin, CR "Khorasan", in PAP, 1903, XCI; Sykes, CR "Khorasan", in PAP, 1906, CXXVII, 1907, CXIV, 1908, CXIV, 1909, XCVII; Kennion, CR "Khorasan", in PAP, 1906, CXXVII.

IN THE

NINETEENTH

CENTURY

143

but, equally, unfair practices of foreign competitors contributed to the failure. The latter factor was mainly responsible for the failure of the attempt made by the Belgians to set up a sugar-producing factory near Tehran towards the end of the century. In the early years of the century, the state showed some interest in sugar production, but all activities seem to have come to a halt because the domestic producer could not compete with the imported sugar. It is to be noted that the government of Iran at no stage tried to use tariffs as a means of protecting the infant sugar industry, or industry in general. That importance may be attributed to the constraints imposed by the treaty restrictions which, in turn, were the outcome of the disastrous politics of the successive governments of Iran during the whole of the nineteenth century. Compiled from estimates given by: Dickson, CR "Azarbijan", in FO 60-263; K. E. Abbott, CR "Azarbijan", in PAP, 1863/64, LXI, 1864, LIV, 1865/66, LXX, 1867/68, LXVIII; Jones, CR "Azarbijan", in PAP, 1871, LXVI, 1872, LVIII, 1873, LXV, 1875, LXXV; W. G. Abbott, CR "Azarbijan", in PAP, 1878, LXXV, 1880, LXXIII, 1888, CII, 1890, LXXVI; Stewart, CR "Azarbijan", in PAP, 1890/91, LXXXVII; Wood, CR "Azarbijan", in PAP, 1894, LXXXVII, 1895, XCIX, 1897, XCII, 1899, CI, 1901, LXXXIV; Wratislaw, CR "Azarbijan", in PAP, 1905, XCI, 1906, CXXVII, 1909, XCVII. 21 Board of Trade, Statistical Abstract on Revenue, Population, Commerce,in PAP, 1841, XXIV, p. 144. 22 Compiled from Official Statistics published by the India Office, and the Board of Trade, in PAP, 1874, LXX, 1875, LV, 1876, LVII, 1881, LXXXVIII, 1886, XLIX, 1890/91, LVIII, 1896, LXII, 1897, LXV. 23 Ibid., see also PAP, 1898, LXIV, 1899, LXVIA, 1900, LVIII, 1901, 1902, LXXIV, 1907, LIX. 24 Yate, CR "Khorasan", in PAP, 1898, XCVII. 25 Board of Trade, Tablesof Revenues ..., in PAP, 1851, LV, pp. 49-50. 26 See H. M. Jones, CR "Azarbijan", in PAP, 1871, LXV. 27 W. G. Abbott, CR "Azarbijan", in PAP, 1880, LXXIII. 28 K. E. Abbott, CR "Azarbijan", in PAP, 1864, LXI. 29 See Dickson, op. cit., in FO 60-263. Wratislaw, CR "Azarbijan", in PAP, 1909, XCVIII. 30 Mackenzie, CR "The Province of Gilan", in FO 60-235. 31 Churchill, CR "Gilan", in PAP, 1876, LXXVI. 32 Compiled from, Churchill, CR "Gilan", in PAP, 1893/94, XCV, 1895, XCIV, 1897, XCII, 1904c H. L. Rabino, CR "Gilan", in PAP, 1907, XCI. 3 V. Lenin, The Development of Capitalism in Russia, Moscow 1977, pp. 296-7. 34 For 1863, see Thomson, CR "Persia", in PAP, LXI. For 1905, see, Chick, CR "Bushire", in PAP, 1909, XCVII; Wratislaw, CR "Azarbijan", in PAP, 1906, CXXVII; Gough, CR "Kermanshah", in PAP, 1906, CXXVII. 35 Ganj-eShayagdn(Berlin 1915), p. 94. 36 E. Lorini, La Persia economicacontemporana... (Rome, 1900), partly repr. in C. Issawi, The Economic History of Iran, 1800-1914 (Chicago, 1971), p. 308. 37 Churchill to Lord Rosebery, despatch 1 March 1894, quoted by Picot, CR "On the Proposed Establishment of a Sugar 20

Industry

in Persia ...", in PAP, 1895, CII.

SHORTER NOTICES HAJJI 'ABBAS

Allan ByAshmolean J. W.Museum One of the mysteries of the later Persian steel industry has been the identity of Hijji 'Abbais. Mayer in his roll of metalworkers published in 1959 lists nine objects made by this steelworker, and in passing he describes him as Safavid.1 Mayer also includes in his roll a son of Hajji 'Abbais, a man who signs himself pisar-i Hdjjfi 'Abbas but offers no further details of his, or his father's, identity.2 Little additional light was thrown on the problem of 'Abbais's identity until 1982, when the steel in the Nuhad Es-Said collection was fully kashkil H.ijji published. Here Haijji 'Abbais gives the name of his father, the late Aqa Rahim the armourer (yaraq-sdz), and dates the piece to 1015 A.H. (A.D. 1606-7).3 Since then the only further word on this craftsman seems to have been my own article in EIr, also published in 1982.4 Here I suggested that, on the available published evidence, there seemed to be three different craftsmen by the name of Hjjli 'Abbas in later Islamic Iran. The first lived in the seventeenth century, witness the Nuhad Es-Said collection kashkiil. A second HIjjji 'Abbdis seemed to have made animal and bird figures, undecorated apart from a roundel bearing his signature,5 and I suggested that they might date from the eighteenth century. A third seemed to have made kashkils and other heavily decorated objects in the latter part of the nineteenth century. In September 1993 I visited Iran and was able to make a brief visit to Isfahan.7 The previous year I had made the acquaintance of Mr. Husayn Alaghmand, the most respected silversmith and qalamzan in the city, a very cultured man, knowledgeable about the history of the city and about the people who live in it. We once again visited him to discuss and the craft metalworking indussteelworking tries. During the course of conversation I asked if he had ever heard of a craftsman of the name of HiIaji 'Abbas, and it was soon apparent that he knew Hdjji 'AbbSs's grandson. Half-an-hour later we were standing in the courtyard of the house where HIIjji 'Abbas himself had lived (P1. IXa). His grandson, Hiijj Muhammad Bayzii 'alamatsaz (P1. IXb), continues to work in steel, as did his father, Hjjj Karim 'atzqasaz. He of his grandfather as 'Abbas talked MirzS_ 'alamatsaz, and attests, too, to a great-grandfather, 'Abd al-Wahhab, who was also a steel craftsman. Hajji 'Abbis, according to IHIij Muhammad Bayai~,

had first had his workshop (kdr-khana) in the bazaar of the dazvdt-kdrdn (the inkwell makers), had then moved into the Caravanserai Fakhr, and in old age had worked in his house. He had died some years before at the age of ninety-five. HIjij Muhammad Bayi-i also told us that his grave was near the Takiyya Mirzai Abu 'l-Ma'a-i in the Takht-i Pilald, the historic cemetery of Isfahan. A visit to the Takht-i Piilfid sadly failed to produce any evidence of the gravestone, despite a long search in the area of Isfahan, it of the Takiyya. The municipality area the cemetery had bulldozed that of appeared, the previous year to make room for another generation of graves. Returning to the hotel that evening, we cast an eye over the tiny bookstall in the lobby and a a tiled building cover showing colour-printed caught my attention. It was one of the small buildings of the Takht-i Puilfid. The slim volume proved to be some poems by a local Isfahani in praise of various people buried in the cemetery, amongst them none other than 'Abbais.8 Not only this, the author had included.HIjji biographical information about HIjji 'Abbais, including the vital date which had been on his gravestone, the date of his death: 1380/1960-1. His entry reads: "Ustaidh IHijji Mirzai 'Abbas (Baydi) son of 'Abd al-Wahhfib, famous maker of 'alams. Died: lunar year 1380. Place of burial: in the area of the Takiyya Mirz- Abu 'l-Ma 'all. Said by Aqa-yi Mahmuid Mu'azzam son of Ha-ijj 'Ali Kal-dar to be one of the famous 'alam-bearers of Isfahan".9 If one assumes that he was 95 when he died, he must have been born in 1865 or 1866. A few days before, I had been working in the Iran Bastan Museum, which has a fine collection of Persian steelwork. In particular, on show in the gallery, are three signed pieces by HiIjji 'Abbais: a watered steel kashkil, a ghalyan body and a peacock. The ghalyan body is of particular interest, since it is to dated, bearing the figures 1292, equivalent A.D. 1875. If our computation of his date of birth is correct he would have been ten years old when this object was produced! The kashkill is worked with very fine chiselling, showing a scene of dervishes so similar to that of the dated ghalydn just mentioned-even down to the appearance of a kashkil in the picture-that it can only be a contemporary piece by the same craftsman. Two such high-quality pieces bearing the name of a ten-year old make the

145

146

JOURNAL

OF PERSIAN

dating somewhat problematic. For the moment, the only likely solution seems to be a mistake in the date on the ghalydn, which should perhaps read 1296. This would make our craftsman 15 years old, and one could conceive of a 15-year-old having his name put on objects being produced in the family workshop, particularly if by then his father 'Abd alWahhab had died and he had inherited the business. It is of some interest at this point to record other pieces of information about Hji 'jAbbas. According to his grandson, his kdr-khanaemployed two other men, Tahir and Hdjj Muhammad Hasan. What their .H@ji particular skills were is not clear, but Muhammad Bayazi also gave us the very Hajj important information that Hajji 'Abbis was particularly famed for the shapes he worked in steel. Hijj Muhammad proudly showed us animals made by Hijji 'Abbas (P1. IXb), and it is clear from him and from others that HIjji 'Abbas in fact confined himself to that craft, doing no chiselling or gold working. This information was confirmed by a visit to the tea-house of Mahmuid 'Alf Kal-dar, in the Khiyaiban-i Haifiz. Here are kept a large quantity of flagella and steel standards for use in the Muharram processions. According to the owner, one of these standards was made by Tahir, but the animals, on it were the work of.Haijj Hajji 'Abbas, while a second standard was made by two members of the Fiilafdgarfamily (another Isfahani steelworking family), now both dead, except for the figures of deer on it, which again were the work of Hajji 'Abbas. There seems little reason to doubt this information: the tea-house owner even possessed a photocopy of Hajji 'Abbas's identity card from the time when the master had been a regular customer of the tea-house (P1. IXc). The specialisation of crafts within the steelworking industry was confirmed by a number of other factors. Firstly, Hajj Muhammad Bayazi himself confirmed that none of his ancestors did their own goldwork. Secondly, we were able to visit another Isfahani craftsman whose basement room was a goldworking workshop to which steel makers brought their products for adornment. Thirdly, we met Hajj Muhammad Sadlq Shiranl 'aldmatsdz, an eminent craftsman trained by Fuiladgar. (Fuiladgar died four or five years ago, and IH-ijjMuhammad Sadiq is now sixty-two.) He told us that the goldwork for the steel objects manufactured in his workshop is all done by his son Hujjat Allah. Fourthly, the previous year we had been introduced by Mr. Alaghmand to the Bahman family, three brothers making 'alams in the Isfahan bazaar. For fourteen years they trained under IIajj

STUDIES

Muhammad Sadiq, but now have their own workAkbar Bahman 'alamatsdz is the shop. Ustadh eldest of threeHajj brothers, and is the only one with the title ustdd. The title ustdd is so hard to acquire that it is very rare to find a craftsman who is an ustdd in more than one craft, but Hajj Akbar Bahman is in fact an ustdd in three specific trades, sc. gold incrustation, piercing and calligraphy. While he specialises in these three fields, his two brothers are responsible for the construction and shaping of the 'alams, and of the animals which adorn it, and for other less important activities like etching. It seems certain to me now that all the undecorated animals and birds I had thought of as late eighteenth century are in fact late nineteenth century or early twentieth century examples of the master's work. The difference in decoration on the kashkils and candlesticks is a reflection of their quite different functions in the traditions of which they are a part. As the steel standards for use in Muharram processions have become progressively larger and more impressive, so their effect has depended more and more on qualities designed to catch the eye at a distance-striking animal and bird shapes, large areas of openwork, and plenty of gold. Objects for use or display in the home have tended to be far more ornate, and decorated with extensive areas of fine chiselling, as well as fine gold work, requiring close observation and appreciation. And what, then, of the Nuhad es-Said kashkil? There has always been a problem with this piece, for although the inscription points to the seventeenth century, the chiselled decoration on the base looks far more like nineteenth century work.10 The solution must surely be that it reflects the taste for the antique prevalent in the nineteenth and early twentieth century. That taste is exemplified Karim, by the title given to Hajji 'Abbas's son, .Hajj 'athqasdz"maker of antiques", and is discussed by the first American ambassador to Iran, S. W. G. Benjamin. Having mentioned "the very beautiful imitations of the ancient work which they now produce at Ispahan for the foreign market", Benjamin continues: "For those who cannot find or cannot afford antique examples of the fine Persian metal work of former ages, it may be granted that these comparatively inexpensive imitations offer a tolerable substitute". Such a fashion, of course, would then account for Hajji'Abbas's kashkil in the Nuhad Es-Said collection, even if we might consider its "date", and the maker's claim to be the son of the armourer of Shah 'Abbas, a little over the top!

HAJJ r 'ABBAS

I L.

A. Mayer, Islamic Metalworkersand Their Works (Geneva, 1959), pp, 19-20. 2 Ibid., p. 47. 3 J. W. Allan, Islamic Metalwork: the Nuhad Es-Said Collection (London, 1982), pp. 114-17. 4 Art. "Hijji 'Abbas", in Vol. 1, fasc. 1, pp. 76-7. 5 Mayer, op. cit., p. 20, nos. VIII and IX, both in the British Museum. 6 Mayer, op. cit., pp. 19-20, nos. II-V. 7 I am most grateful to the British Institute of Persian Studies for a grant enabling me to visit Iran and undertake research into the history of traditional steelworking, in both 1992 and 1993. I am also most grateful to Parviz Tanavoli for all the

147

help he gave me during my visits, to Mr. Rustamian and Mr. Alaghmand for their kindness, time and assistance in Isfahan, and to Mr. Tehrani Moghadam, Director of the Muzeh Melli (Iran Bastan Museum), and Mr. Ghouchani in the Islamic Department of the latter, for permission to study the steel objects in the Museum. 8 M. Shafiq, Takht-ifildd-i I?fahan (Isfahan, 1372/1992). 9 Ibid., pp. 46-7. 10 This is a point which I failed to appreciate when I published the piece, but a number of other scholars (the first was Boris Marshak) have convinced me that it is correct. " Persia and the Persians (London, 1887), pp. 304-5.

DIZ By Mary Burkett Along the banks of the river Zayandeh Rud, running from west to east from Isfahan to the source of the river at Gav Khaneh, are villages and the remains of villages. Some 60 km. to the east (as the crow flies) is Faizabad, then, moving further east, Kaftun, Partun, Urun, Pitfirz, and then (some 35 km. from Faizabad as the crow flies) Varzaneh. There are six unnamed sites between Faizabad and Varzaneh. Still further east, tracks lead into more mountainous territory. On one of these tracks is the site of Diz. Could this have been an old route from Isfahan to Nain? The size of the pigeon towers in the area would seem to point towards this having once been very fertile and populous territory. The area is known to have been well populated in Seljuq times. It was largely destroyed by the Mongols and later suffered plague and salination of the soil. Maxime Siroux noted that, east of Varzaneh, in the direction of Yezd, a rather arid track passed by Yegu Mish, led around the foot of the Kuh-e Shiraz and climed a hill pass.1 It then rejoined a group of and houses at Gel-c Shur and Anarastan, remounted to Khalilabad, where there is ajunction with Ushn. This desert crossing, with only a single well en route, leads to Aghda. It was too dry to have been used as a route for large military or commercial convoys. From Varzaneh, another route led to Yezd. It was even worse than the first and Siroux suggested that it would only have been used in emergencies, for example at the end of the eighth century A.D., during the fifteenth century by the Muzaffarids and in the eighteenth century by the Afghans. It rounded the north-east side of the Gav-Khaneh, eventually arriving at the Zoroastrian village of Taft further east, 25km. from Yazd.

In 1974 we were attempting to relocate Achard, where the previous year I had found masses of dark green glazed pottery, later identified as Sasanian. Regrettably, we could not find it. However, driving along the Zayandeh Rud, some way east of Varzaneh, I noticed a minaret, which proved to be that of the mosque of Diz (P1. Xa). There was no road to it, but at the point where we waded the stream there was evidence of an old road coming down to the stream from the opposite bank. The minaret was attached to a mosque surrounded by the remains of ruined mud brick buildings. There was a probable cistern, of baked bricks, and baked bricks appeared in some other buildings, one a very large one. Some 750 yards from the ruined town, at map reference 232.405, stands a fine, almost square caravanserai or walled fort measuring approximately 500 x 600 m. on the outside. The surrounding walls were intact, as were many of the interior walls, some of the domed chambers still being used to house animals. At the far end, several domed roofs suggested the remains of a sunken hammam. At the entrance gate (Pl. Xb) there was still a fine domed roof. The interior did not entirely resemble that of a caravanserai with rooms opening onto a central courtyard, but this may have been due to the uneven piles of rubble which had fallen on the ground. The ground was littered with sherds. Most of these were blue and black, and appeared to be fourteenth century.

149

Anciennes voies et monumentsroutiers de la region d'lspahan, in Memoires publies par les membres de 1'Institut Frangais

d'Archiologie Orientale du Caire, LXXXII,Cairo 1971, pp. 31-4.

Pl. Ia. Aerialviewof theNorthHillfrom thenorth-east,showing OperationsB2/C1 (centreright),A2/3 (uppercentre),C2 (centre), C3b/5 (upperleftcentre),and DI (centreleft).

Pl. Ic. Urartiancosmeticjar,Room3, OperationB2.3.

Pl. Ib.Aerialviewof theSouthHill from thesouth-east,showing M3/4 (upperleft). OperationM1/7 (centre)and the 1992 Step-Trench Pl. Id. Fragmentof an Urartianbronzequiver, Bl, North-WestGateway,excavated1992.

Pl. le. Urartianperiodseal impressiononfragmentof largestoragejar, Room3, OperationB2.3.

from thesouth,showingRoom5 (centre),Room4 Pl. Ha. TheB2 Complex (rightcentre),and Room3 (far right),as wellas shallow "Late to the north. Urartian"B2.5 architecture

P1.Ilc. Doorwaysin Room3, B2.3. Rightdoorwayis throughWall6, connectingRooms3 and 4; leftdoorwayis throughWall5, connecting Room3 to an unexcavatedroomto thesouth.

Pl. Hb.Room2, B2 Complexwithremainsof breadoven.

P1.lId. Room2, OperationDifrom its south-eastern corner.Flagstone flooring, 3 columnbases,feeding troughs,hearth,oval stone able, and stoneplatformwith waterbasin areshown.

Pl. He. Ironkeyfound near blockeddoorwayonfloor ofRoom5, B2.4.

Pl. IIIa. Mixed "EarlyIron"and EarlyBronzepotteryfrom Operation C3b/5. Uppermost eightfragmentsarefromLocus8 and bottomtwo from Locus 9.

Pl. IIIb.E2 Stonecistburial,noteiron braceletaroundupperarm.

Pl. IIIc.EarlyBronzevesselsfromEarlyBronzetomb,OperationCl, excavated1992.

Pl. IIId. OperationM7from thewest,showingFeatureI (wall in background)and Feature3 (largecut basin with channeldraining intopit).

Pl. Ille. Boneartefactsfrom roomenclosedbyFeaturesI and 4, Operation M7. Twopoints to therightwerefound togetherwitha cacheof astragali and metatarsalbonesin a small rectangularstoneinstallation.

Pl. IIIf Satkhe,OperationBI--Deposit of EarlyBronzevesselsand antlerswithin interconnected domesticcomplex.

P1.IVa. Merv:Erk and Gyaur-Kala.Viewof the north-eastcornerwithErk-Kalaat therear right. The largelow mound in the centreis MGK:5; theexcavated"ovalbuilding"is visible near thebottomright and theeast wall of MGKis in theforeground.

Pl. IVc.Mero:Gyaur-Kala,Trench5. Generalview to thesouth, showingsomeof the areas. surface-scraped

P1.IVb.Mero:Erk-Kala,Trench1. RoomA, articula

Pl. IVd.Merv: Gyaur-Kala,Trench5. Detail of m area. surface-scraped

Pl. Va. MGK SurfaceSurveyType17 (seeFig. 1A).

Pl. Vb.MGKSurfaceSurveyType2/11/29 (seeFig. iB).

Pl. Vc.MGK SurfaceSurveyType3/22/4 7 (seeFig. 1C).

P1. Vd.MEK, Trench1. Mould, madefrom a re-usedjar handlesherd.

P1. VI. Selectedexcavatedcoinsfrom MEK, Trench1 (1992). Nos. 1-3, Parthian "bronzedrachms"of theMargiana mint, latefirst-secondcenturiesA.D., No. 5, ArdashirI, Type111/2;No. 6, ShapurI, Type2; Nos. 12-13, ShapurIII; No. 14, BahramIV, No. 16, anonymousfifth centurySasanian.

Pl. VII.SelectedMiddlePersian ostracafrom MEKITrench1: a. SmallFind No. 231, interiorand exterior: b. S.F No. 238; c. S.F No. 237.

Pl. VIlla. Merv:Gyaur-Kala.Surfaceremainsof Islamic cruciblesteelproduction.

of steeldropletsin the Pl. VIIIb.Photomicrograph glassygreenslag adheringto the cruciblewall. The two indentationsof theprill arefrom microhardness testing (Hv = 320).

Pl. VIIIc.Detail of cruciblewallfragmentshowing differencesaboveand belowthe 'fin" of glassygreenslag.

P1. VIId. Interpretationof the crucible fragmentpositionsfor steelproductionat Merv:Gyaur-Kala.Thefragmentsare arrangedfrom the baseto the lid.

Pl. IXa. Courtyardof the house of 'Abbas, now ownedby his grandson,H1djjMuhammad IH.djfi Baydi (Photo:J. W. Allan).

P1. IXb. IHdjjMuhammadBaydi with objectsmade 'Abbas by his grandfather, I.djjl (Photo:J. W. Allan).

Pl. IXc. Photocopyof the Identity Card

P1. Xa. The mosqueat Diz (Photo:M. Burkett).

Pl. Xb. Diz, entranceto the caravanserai(Photo:M. Burkett).

ABBREVIATIONS AARP AASOR AfO AJ AJA AJSL AK AMI ANET AO Arch Anz ArO AS BA Besch BASOR Belleten BGA Bib Or BMMA BSA BSOAS CAH CDAFI CHIr CIA CII El' El2 EIr EW IA IIJ IJMES ILN Isl JA JAOS JCS JFA JHS JNES JRAI JRAS JRCAS JSS KF LAAA MAOG MDAFA MDAI MDOG MDP MJ NC OIC OIP PZ

Art and Archaeology Research Papers Annual of American Schools of Oriental Research Archiv ftir Orientforschung Antiquaries' Journal American Journal of Archaeology American Journal of Semitic Languages and Literatures Antike Kunst Archaeologische Mitteilungen aus Iran Pritchard, Ancient Near Eastern Texts Ars Orientalis Archaologischer Anzeiger Archiv Orientalni Anatolian Studies Bulletin van de Vereeniging ... de Antieke Beschaving, The Hague Bulletin of American Schools of Oriental Research Tiirk Tarih Kurumu: Belleten Bibliotheca Geographorum Arabicorum Bibliotheca Orientalis Bulletin of the Metropolitan Museum of Art Annual of the British School at Athens Bulletin of the School of Oriental and African Studies Cambridge Ancient History Cahiers de la Delegation Archeologique Frangaise en Iran Cambridge History of Iran Corpus Inscriptionum Arabicarum Corpus Inscriptionum Iranicarum Encyclopaedia of Islam, Ist edition Encyclopaedia of Islam, 2nd edition Encyclopaedia Iranica East and West, New Series Iranica Antiqua Indo-Iranian Journal International Journal of Middle East Studies Illustrated London News Der Islam Journal Asiatique Journal of the American Oriental Society Journal of Cuneiform Studies Journal of Field Archaeology Journal of Hellenic Studies Journal of Near Eastern Studies Journal of the Royal Anthropological Institute Journal of the Royal Asiatic Society Journal of the Royal Central Asian Society Journal of Semitic Studies Kleinasiatische Forschungen Annals of Archaeology and Anthropology, Liverpool Mitteilungen der Altorientalischen Gesellschaft M6moires de la Delegation Archeologique Frangaise en Afghanistan M6moires de la Dl66gation Arch6ologique en Iran Mitteilungen der Deutschen Orientgesellschaft M6moires de la Mission Arch6ologique de Perse Museum Journal, Philadelphia Numismatic Chronicle Oriental Institute, Chicago, Communications Oriental Institute, Chicago. Publications PraehistorischeZeitschrift

RA REI SAA SAOC Sov Arkh SS St Ir

Revue d'Assyriologie Revue des Etudes Islamiques Soviet Anthropology and Archaeology Oriental Institute, Studies in Ancient Oriental Civilisation Sovietskaya Arkheologiya Schmidt, H., Heinrich Schliemanns Sammlung trojanischer Altertiimer Studia Iranica A Survey of Persian Art from Prehistoric Times to the Present, ed. A. U. Pope, Oxford, 1938 Die Welt des Orients Wissenschaftliche Veroffentlichungen der Deutschen Orientgesellschaft Zeitschrift flir Assyriologie Zeitschrift der Deutschen Morgenliindischen Gesellschaft

Survey WdO WVDOG ZA ZDMG

151

NOTES ON TRANSLITERATION I.

OLD

AND

MIDDLE

FOR CONTRIBUTORS

TO IRAN

PERSIAN

It is recognized that no rigid lines can be laid down here, but it is suggested that the Old Persian syllabary should be transliterated according to the table in Kent, Old Persian. Grammar, Texts, Lexicon, p. 12; that for Manichaean Middle Persian and Parthian, the transliteration system given in AndreasHenning, Mitteliranische Manichaica, vol. III, p. 66, should be used; whilst for Pahlavi, the table of alphabets given in Nyberg, A Manual of Pahlavi, new edition, p. 129, may be used as a reference for transcription. II.

ISLAMIC

AND

MODERN

PERSIAN

Tlihe system used fior the CambridgeHislornvo/fIslam should be used here as far as possi)ble.

Consonants (a) Arabic b.

b St Sth

C

,

s sh

k J 1 m

s

hh

w kh d gh y ,; -a (in construct state: j dh gh t. f -at) Sr (b) Persian additional and variant forms. The variant forms should generally be used for Iranlian names and for Arabic words used in Persian. p g zS v . s 3 zh ch zi (c) 'I'he Persian "silent h" should be transliterated a, e.g. nama. ,'owels Arabic or Persian Short: a

Long: I or j

Pu

a U

Doubled -. iyy (final form 1) au Dipthongs O ? .ai

NO TESJ 1. The iIzafashould be represented by -i, or after long vowels, by -yi, e.g. umard-yijankf. 2. The Arabic definite article should be written as al- or 1-, even before the so-called "sun letters", e.g. 'Abd al-Malik, Abu 'l-Nasr. 3. The macrons of Abia and Dhfi (Zui) should be omitted before the definite article, e.g. Abu 'l-Abbas (but Abfi 'Ubaida). It is obvious that for the rendering of linguistic and dialectical material, and possibly also for contemporary literary and spoken Persian, this rigorous system of transliteration is inappropriate; contributors should use their discretion here. III.

GENERAL

POINTS

1. Names of persons should be rigorously transliterated. 2. Conventional English equivalents (without macrons or diacritics) should be used for the names of countries, provinces or large towns, e.g. Khurasan, Shiraz. Otherwise, all place-names should be rigorously transliterated. Archaeologists are asked to be especially careful in representing the names of little-known places at or near sites. 3. Modern Turkish names and words should be written in the current romanized Turkish orthography. 4. Where classical Greek and Latin renderings of Old and Middle Persian names exist, these familiar forms should be used for preference. 152

NOTES FOR THE CONTRIBUTORS Iran is a refereed journal with a board of editorial advisers. The editors are Professor C.E. Bosworth and Dr. Vesta Curtis and the editorial advisers are Dr. P. R. S. Moorey, Professor J. M. Rogers and Professor David Stronach. In addition, articles are sent to other scholars as appropriate. Articles for Iran should be typewritten on one side of A4 paper or the nearest North American equivalent size with double spacing and generous margins. Carbon copies or photocopies of typescript are not acceptable. Notes should be numbered consecutively and placed at the end of the article. Photographs for reproduction should be, as far as possible, in the form of bright and sharp glossy black and white prints, and should be original photographs. In case of reproductions, permission by the author or publisher must be obtained before hand. Whenever possible, contributors should submit original line drawings rather than photographic or other reproductions. Authors of articles will receive 25 offprints free and may order additional ones, at reasonable prices, in multiples of 25. The Editors should be informed of any extra offprint orders when the first proofs of articles are returned. Authors of contributions in the Shorter Notices will receive 15 copies of the complete ShorterNotices section.

THE BRITISH INSTITUTE OF PERSIANSTUDIES BIPS PUBLICATIONS Rare copies-

(1-9) inclusive

?40 each

Volumes

I-IX

Volume

XII

(12)

?40 each

Volume Volume Volume

XV XXII XXIII

(15) (22) (23)

?40 each ?40 each ?40 each

Complete set

Volumes

I-XXXI

(23)

?40 each

Xerox copies-

Volumes I-V inclusive Volumes VII-IX inclusive

?15 each ?29 each

For a limited period we wish to offer at bargain price per copy, the following IRANjournals and other publications subject to availability. Iran

X (1972), XI (1973), XVI-XXI inclusive (1978-1983)

?9.50

Siraf Report Fasc. III: The Congregational Mosque by David Whitehouse Fasc. XV: The Coins and Monumental Inscriptions by Nicholas M. Lowick

?5.00 ?12.00

Nush-IJan Report Fasc. III: The Small Finds byJohn Curtis

?10.00

These Three Reports may be purchased as a set for the additional discount price of ?20 Copies may be obtained from the Publications Secretary, Miss Mary Totman, 63 Old Street, London EC1V 9HX. Answerphone/fax. number 071 490 4404 (UK). Single copies of publications are sent post free to UK addresses. Postage and packing is extra for multiple copies and complete sets to addresses in the UK and all orders for addresses overseas. Postage and packing for one volume overseas ?4. Payment should accompany orders, please. Those ordering from overseas may pay by draft drawn in

London, international money order, Eurocheque or payment in dollars at an exchange rate of $2 = ?1 on orders over ?15.

250 soRites 1

?

I

z7 Scp sP J.hrmica

I

._-. .Gorgan

.:. ... QaZV* .4NDARAN

-*•-

*TEHRAN

4.

'

I"

"es

(

1

I

?TURKMENISTAN

j RBAI~~RBAJAN sh.. .

I I 4, ooKilZmetres

r

Damghan, DMeshed

,

kRa adaiLKKHURASAN

p

."

A I..R

N

7

i

aasan \

I

,i

AAN -Per$usaous

Shir

.

.-

S"-..

SA

i DI

•"6

""

I.K

ARABLA

ERAiA%

4f

.••s

A

, .

.

Printed in England by Stephen Austin and Sons Ltd, Hertford

A Ni

I