IRAN

Journal
VOLUME XIII

of

the

British

Institute

of

Persian

Studies

1975

CONTENTS Page ii Governing Council .. Obituaries . . . . . . . . . . iii . . . Director's Report . v The Merchants of Susa; Godin V and Plateau-lowland Relations in the Late Fourth Millennium B.C., by Harvey Weiss and T. Cuyler . . . . Young, Jr . I Some elaborately decorated Bronze Quiver Plaques made in Luristan, . c. 750-650 B.C., by P.R.S. Moorey . 19 .. Chinese Porcelain and Ardabil, by Margaret Medley . . Islam, 31 The Ardabil Shrine in the Reign of Shih Tahmisp I (concluded), . . 39 by A. H. Morton . .. The Masjid-i 'All, Quhrfid; an architectural and epigraphic survey, by Oliver Watson 59 ? ? " ... The Foundations of Persian Prosody and Metrics, by L. P. Elwell Sutton . 75 The British Museum Mirzinima and the Seventeenth Century Mirz5 in India, by Aziz Ahmad . . . . 99 Dalma Tepe, by Carol Hamlin . . . . IiI Early Fourth Millennium Developments in Southwestern Iran, by H. T. Wright, J. A. Neely, G. A. Johnson and John Speth . .129 Excavations at Haftavan Tepe, 1973; Fourth Preliminary Report, by . . Charles Burney 149 ... Shorter Notices . . . . 165 Survey of Excavations in Iran--1973-74 . . . . 173 .
. .

THE

BRITISH

INSTITUTE
Price: ?5.oo

OF

PERSIAN

STUDIES

c/o The British Academy, Burlington House, Piccadilly, London, WIV ONS

STATEMENT OF AIMS AND ACTIVITIES

I. The Institute has an establishment in Tehran at which British scholars, men 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 at 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 a journal, 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 Honorary Secretary, J. E. F. Gueritz, Esq., M.A., 85 Queen's Road, Richmond, Surrey, TWI o 6HJ. The annual subscription for Membership of the Institute is ?i, while the total sum of ?4 entitles the subscriber to receive the Journal.

IRAN
Volume XIII i975
CONTENTS Governing Council Obituaries Director's Report The Merchants of Susa; Godin V and Plateau-lowland Relations in the Late Fourth Millennium B.C., by Harvey Weiss and T. Cuyler Young, Jr. Some elaborately decorated Bronze Quiver Plaques made in Luristan, c. 750-650 B.C., by P. R. S. Moorey Islam, Chinese Porcelain and Ardabil, by Margaret Medley The Ardabil Shrine in the Reign of Shah Tahmasp I (concluded), by A. H. Morton The Masjid-i 'Ali, Quhrfid; an architectural and epigraphic survey, by Oliver Watson The Foundations of Persian Prosody and Metrics, by L. P. Elwell Sutton The British Museum MirziinAma and the Seventeenth Century Mirzi in India, by Aziz Ahmad Dalma Tepe, by Carol Hamlin Early Fourth Millennium Developments in Southwestern Iran, by H. T. Wright, J. A. Neely, G. A. Johnson and John Speth Excavations at Haftavan Tepe, 1973: Fourth Preliminary Report, by Charles
Burney
149

Page ii iii v I
I9

31 39 59 75 99
III

129

Shorter Notices: An Inscribed Iranian Bronze, by W. G. Lambert A Note on Sasanian Harpies, by Judith Lerner A New Discovery at Bishapur
Survey of Excavations in Iran-1I973-74

I65

I73

THE

BRITISH

INSTITUTE

OF PERSIAN

STUDIES

c/o The British Academy, BurlingtonHouse, Piccadilly, London, WIV ONS

BRITISH INSTITUTE OF PERSIAN STUDIES GOVERNING COUNCIL President *Sir MAX MALLOWAN, C.B.E., M.A., D.Lit., F.B.A., F.S.A. VicePresident BASIL GRAY, Esq., C.B., C.B.E., F.B.A. Members ProfessorSir HAROLD BAILEY, M.A., D.Phil., F.B.A. R. D. BARNETT, Esq., C.B.E., D.Lit., F.B.A., F.S.A. ProfessorJ. A. BOYLE, B.A., M.A., Ph.D. MICHAEL BROWNE, Esq., Q.C., M.A. Professor W. B. FISHER, B.A., D. de l'Univ., F.R.A.I. Dr. ILYA GERSHEVITCH, M.A., Ph.D., D.Litt., F.B.A. Professor A. K. S. LAMBTON, O.B.E., D.Litt., Ph.D., F.B.A. Professor SETON H. F. LLOYD, C.B.E., M.A., F.B.A., F.S.A., A.R.I.B.A. LAURENCE LOCKHART, Esq., Litt.D., Ph.D., F.R.Hist.S. P. R. S. MOOREY, Esq., M.A., D.Phil., F.S.A. RALPH H. PINDER-WILSON, Esq., M.A., F.S.A. B. W. ROBINSON, Esq., M.A., B.Litt. *Sir MORTIMER WHEELER, C.H., C.I.E., M.C., T.D., D.Lit., F.R.S., F.B.A., F.S.A. NEVILLE J. WILLIAMS, Esq., D.Phil., F.S.A. Sir DENIS WRIGHT, G.C.M.G., M.A. Hon. Treasurer STEPHEN C. G. BACH, Esq., C.B.E., M.A., F.S.A. Hon. Secretary E. F. GUERITZ, Esq., M.A. JOHN

Joint Hon. Editors Mrs. GEORGINA HERRMANN, D.Phil., F.S.A. Professor C. E. BOSWORTH, M.A., Ph.D. OFFICERS IN IRAN
Director DAVID STRONACH, Esq., O.B.E., M.A., F.S.A.

Assistant Director ALEXANDER H. MORTON, Esq., B.A. P.O. Box 2617,

c/o The British Academy, Burlington House, Piccadilly, LONDON, WIV ONS

Tehran, IRAN
*DenotesFounderMember

ii

SIR JOHN LE ROUGETEL

OBITUARY
SIR JOHN LE ROUGETEL

Sir John Le Rougetel, KCMG, MC, died on 3 January 1975 at the age of 8o. He served for ten years as Hon. Treasurer of our Institute from the time of its inception in 1963 and was an obvious choice for the post because of the distinguished service which he had rendered in Iran. When ill health made it advisable for him to relinquish office, he continued to attend Council's meetings and to support us with thoughtful advice. His wise counsel and prudent manipulation of the Institute's finances enabled it to undertake the foundation of new premises on land at Gulhak, and it was a special satisfaction to him to know before he died that building had begun. He thus leaves a permanent memorial to himself in Iran, a country in which many friends will mourn his loss. Obituaries published in the Daily Telegraph The Timeson 8 and 9 January 1975 contain a record and of the high offices which he occupied in the Diplomatic Service: they covered a period of 35 years from 1920o onwards. He was Ambassador to Iran from 1946-1950 at a peculiarly delicate phase in its history, and had to overcome the not unnatural feeling of hostility which the Government of the time felt owing to the forceful intervention of the U.K. in Iranian affairs during the Second World War, but his integrity, manifest goodwill and humane outlook gradually enabled him to obtain the confidence of the young Shah who was faced with the difficult task of re-establishingwholly independent Iranian Government. Sir John left the country to take up his next post as Ambassador in Belgium, leaving behind him a fund of high regard and respect which made the tasks of his successors the easier. A man of the highest courage, he was twice decorated for bravery during the First World War and during the Second World War was admired for his exemplary behaviour when he and his wife supported the morale of their companions on being taken prisoner by the Japanese. He was not only a brave man but also one endowed with profound wisdom, and although brought up in the old school was receptive to new ideas and sympathetic to the young. After retirement he led a life dedicated to voluntary public service and maintained his strong interests and links with Iran, a country to which he remained devoted after his term of office as Ambassador. To his many friends his death comes as a grievous blow for he was a warm hearted and loyal companion, always humane and discerning in whatever business he was engaged. No one was more alive to the frailties and imperfections of his fellows, but he was tolerant and understanding and had a far sighted approach to the solution of the problems which confronted him. His modesty concealed his remarkable talent as a linguist. In the course of his career he learnt Russian, Japanese, Hungarian, French and German. Throughout his life he was sustained by his beloved wife who shared the anxieties and the dangers to which he was submitted, especially during the Second World War. He used to say that he could not have borne these trials without her constant support. From the time of his honeymoon in a Buddhist monastery in Japan and for fifty years, marriage was as it should be, a true and enduring partnership, each dedicated to the other. M.M.

OBITUARY
PROFESSOR R. C. ZAEHNER

Robin Zaehner was born in 1913. He went to Tonbridge School and to Christ Church, Oxford, which he entered in 1932. He took Honour Moderations in Classics, then changed his subject to Oriental Studies; he read Persian with Old Iranian as second subject, and took a First Class in the Honour School. Classical studies, however, remained an important part of his scholarly equipment and Classical texts part of his reading. He was awarded a Senior Scholarship at Christ Church. His main interest was always in the study of religion; whilst still an undergraduate he taught himself both Arabic and Sanskrit in order to read texts at first hand. He himself has written of how the vital tool which he required for the study of the Zoroastrian religion was given to him by Professor (now Sir Harold) Bailey of Cambridge with whose encouragement and help he learnt Pahlavi or Middle Persian, and whose friendship he was always to value highly. War service intervened in his career, though he was lucky enough to be able to work in Iran; he was Assistant Press Attach6 and later Press Attach6 in the British Embassy. Whilst in Iran he was received into the Catholic church. He returned to Oxford and became University Lecturer in Persian in 1950. In 1952 he became Spalding Professor of Eastern Religions and Ethics, with a Fellowship of All Souls; he continued, however, to teach undergraduates in the Faculty of Oriental Studies in both Persian and Sanskrit; he also served on the Faculty Board and on its Committees. a In 1952 appeard his Zurvan, Zoroastrian Dilemma,an exhaustive examination of ZurvanismIand in in and Sacred Profane, which he attacked the 1956 TheTeachings theMagi; in 1957he publishedMysticism, of tendency to see mysticism as one and the same in all its manifestations and argued that Sufism had undergone a radical transformation through the influence of Indian ideas on Abti Yazid of Bastam. In Hinduand MuslimMysticism(1960) he continued his thesis and accumulated evidence for the penetration of the doctrines of the Vedanta into Iran.2 His further publications included The Dawn and He lectured widely at home, in the United States and in Iran; he travelled to Czechoslovakia, duly learning Czech before he did so; he was a Fellow of the British Academy. Zaehner was appointed a member of the Council of the British Institute of Persian Studies from its inception in 1963 on the recommendation of the late Sir Maurice Bowra, a personal friend of long standing and an admirer of his work. He took the closest interest in the affairsof the Institute and gave valuable counsel on matters concerning scholarship particularly in the field of religion. Iran, ancient and modern, played a major role in his life. It was there that he was converted to Catholicism in I946; its landscape remained in his mind's eye, as when he writes3of" Mihragdn, the feast of Iran's lovely Autumn ". In addition to being erudite in pre- and post-Islamic Persianliterature, he spoke modern Persian extremely fluently (yet another language in his armoury was French, which he had spoken from childhood). He was generous and solicitous towards younger scholars; he was a natural and skilful teacher, a man of ready sympathy, devoted to his friends.
G.M. Twilight of Zoroastrianism (1961), Hinduism (1962), The ConvergentSpirit (1963), Concordand Discord (1970), Drugs, Mysticism and Make Believe (1972) and latterly Our Savage God.

x He was greatly touched by a fulsome personal tribute from

2

Professor W. B. Henning about this book. I should like to acknowledge the kind help of Mr. Julian

Baldick here and elsewhere in this obituary. p. s The Dawn and Twilight of Zoroastrianism, 14i.

iv

PROFESSOR

R. C.

ZAEHNER

DIRECTOR'S REPORT
NovemberIst 1973 to October 31st 1974

Within the next two years the Institute expects to be able to construct a new building on a site at Gulhak which at present forms part of the British Embassy Summer Compound in northern Tehran. Her Majesty's Government has offered the Institute a generous grant towards the cost of the building and the Institute itself has also undertaken to raise additional funds in this connection. It is hoped that the Institute will be installed in its new premises early in 1976. Lectures The first of the Institute's Visiting Lecturers for the year 1973-74 was Professor Henry Hodges. On the occasion of the opening of the Conservation Laboratory of the Iranian Centre for Archaeological Research early in April ProfessorHodges delivered a lecture on " Conservationin the Museum " under the joint auspices of the Centre and the Institute. Later, as a guest of the Centre, he visited Bishapur, Kangavar, Bisitun and other sites where conservation work is now in progress. The second of the Institute's Visiting Lecturers, Mr. Peter Brown of All Souls College, Oxford, spent three weeks in Iran in April and May. As part of a notable lecture tour which was arranged under the joint auspices of the Institute and the British Council, Mr. Brown spoke before a large audience at the Institute on April I7th on " Sasanian Iran in the Near East ". In addition to visiting Sasanian sites in western and southern Iran, Mr. Brown lectured in Isfahan, Shiraz, Tabriz and Meshed. Other recent lectures delivered at the Institute have included " The Structure of the Safavid State " ProfessorR. M. Savory, " Shaikh Safi and his Family " by Mr. A. H. Morton, " Varzaneh, Mosque by and Minaret " by Mr. A. Hutt, " The City of Rayy as a Scholarly Centre in Mediaeval Times " by ProfessorM. Mohaghegh, " Tepe Nush-i Jan, 1973 " by Mr. Stronach and " Film and Anthropology " by ProfessorNader Afshar. ProfessorAfshar's admirable film " Balut " (" Acorn ") followed the last lecture. Director In addition to lecturing on " Three Seasons of Excavations at Tepe Nush-i Jan " at the Eleventh Annual General Meeting of the Institute in London on November 8th, Mr. Stronach delivered the Rhind Lectures for 1973-74 at the University of Edinburgh on " Aspects of the Archaeology of Iran ". The Lectures covered the period from the end of the third millennium B.C. to the first century A.D. In December Mr. Stronach was made an Ordinary Member of the German Archaeological Institute. Assistant Director Mr. Morton has continued to work on documentary evidence concerning the shrine of Shaikh Safi al-Din at Ardabil and the early history of the Safavi family. While he was in the United Kingdom in the spring Mr. Morton gave a talk entitled " Shaikh Zahid Gilani and some aspects of Sufi organization in I3th century Iran " at the University of Kent.
Fellows Two Fellows were appointed for the year 1973-74. Mr. Peter Andrews, who was made a Fellow for a second period, was able to complete the greater part of his wide ranging study of Nomad Tents. In company with his wife he recorded important elements of tribal equipment, some of it no longer in regular use, in the provinces of Fars and Kerman as well as in parts of north-eastern and north-western Iran.
V

Mr. David Morgan has collected a substantial amount of printed, and some unpublished, material dealing with the impact of Mongol rule in Iran. This material included, among documentary evidence, the Sa'adat-Nama of Ala Tabrizi. Visitors Many distinguished scholars have visited the Institute and made use of its facilities during the year. At a time when most long established fields of academic research in Iran are more active than ever before, we have also provided accommodation for ProfessorN. N. Ambraseys and his colleagues from Imperial College, London, and the University of Cambridge who are investigating seismo-tectonicsin the Middle East and have been able to assist a number of other expeditions concerned with palaeoecological and palaeo-zoological research. Siraf The sixth and final campaign at Siraf ran from October to December 1973. Although the main work of the expedition was concentrated on the conservation of the Friday Mosque and other previously excavated monuments, Dr. Whitehouse was also able to examine a number of rock-cut mausolea of probable Sasanian date which lie inland from the city. At the end of the season the dig house and the whole area of the ancient site-a site which is now registeredas a National Monument-was handed over to the care of the Iranian archaeological authorities. TepeNush-iJan A fourth season of excavations at Tepe Nush-i Jan took place in the summer of 1974. Excavations in the Median levels to the south of the central Fire Temple exposed a series of handsome mud brick arches; work in the Columned Hall revealed the entrance to a deep rock-cut tunnel; and at the western end of the site we now know that the Old Western Building represents a second temple. Finally, in September I974, the National Organisation for the Restoration of Historical Monuments very generously met the cost of a large protective steel roof which has since been erected over the well preserved walls of the Median Fort.

vi

THE MERCHANTS OF SUSA

GODIN V AND PLATEAU-LOWLAND RELATIONS IN THE LATE FOURTH MILLENNIUM B.C.

By Harvey Weiss and T. Cuyler Young, Jr.

In the summer of 1973 fairly extensive remains of Period V (c. 3200-3000 B.c.) were excavated at Godin Tepe in central western Iran (Fig. i). At this time the Kangavar Valley, and, one suspects, neighbouring areas in the Zagros highlands as well, entered a phase of major cultural contact with the lowlands of Greater Mesopotamia. The latter area was on, if not already across,the thresholdof urban, literate civilization. These highland valleys, on the other hand, apparently had not heretoforeparticipated in those developments. One suspects that the dynamics and the form of the interaction between the lowlands and the central highlands during this period of exceptional cultural ferment had much to do with shaping both the contemporary and the subsequent history of the whole area. The data from Period V at Godin provide us with our first opportunity to study this issue in detail.

",,. SEH GABI ' KHAFAJE GODIN

'Tehran

IRAN 0 100 km
o Isfahan

200

300

Baghdad"\
L.,

MISH URUK

MI TAL-iGHAZiR

I SR A Q /S,
.
Yi

*MALYAN
o Shiraz

r
?YAHYA

"&g

Wig-.

Fig.

r Map of western showingthe location theprincipalsites discussed. Iran of 1

2A

JOURNAL

OF PERSIAN

STUDIES

To date only a short note and a popular article have been published on Godin V.1 This paper attempts a more detailed presentation of some of these materials, and a provisional discussion of the chronology and culture of Godin V within the context of late 4th millennium Greater Mesopotamia. STRATIFICATION AND INTERNAL CHRONOLOGY Period V materials were excavated in 1973 over an area of some 550 sq. m. in the Deep Sounding on the Upper Citadel Mound, and in an operation covering about 165 sq. m. at the west base of the Citadel Mound. For convenience the latter excavation is called the Brick Kiln Cut.2 In both locations Period V strata and buildings rested conformably on the uppermost remains of Period VI. This was most easily observed in the Brick Kiln Cut where the walls of private village houses from Period V were only slight modifications of house walls from the upper building level of Period VI. The situation was perhaps less clear in the Deep Sounding. Here the Period V structures are markedly different in character and configuration from what little we know of the Period VI buildings in the area. Even so, there is no clear evidence of any chronological or occupational gap between the two periods. Stratigraphically, therefore, Period V is an uninterrupted continuation of the Period VI occupation. The stratigraphyin the Deep Sounding also shows that the Period V buildings here were set on the summit of the late fourth millennium mound. Both Period V and VI strata sloped sharply down from this summit to the south and west; to the west the strata dropped vertically some 2 m. over a horizontal distance of slightly less than 4 m., and the Period V remains in the Brick Kiln Cut are some 12-13 m. below the Period V structureson the summit of the mound. Due to recent erosion we have no evidence for the ancient slope of the mound to the north. To the east Period V materials are found in small test trenches at the base of the Citadel Mound at a level to suggest that the ancient slope of the site in this direction was as steep as to the west. There is a clear hiatus of occupation at the end of Period V, for the Period IV deposit rests unconformably on the upper stratum of Period V. In the Brick Kiln Cut an erosion surface truncated the upper Period V deposit, and on that surface rested the earliest Period IV walls and hearths. The stratification and the character of the deposits in the Deep Sounding, however, indicate that this hiatus was of short duration. Here several of the Period V walls were still standing when the Period IV occupation began. In one area much ash and trash gathered from Period IV hearths had been thrown into an abandoned but still standing Period V room. The deposit extended down to only a few centimetres above the latest Period V floor. In at least two other areas Period IV hearths had been constructed against the standing stubs of Period V walls. It is hazardous to guess how long it takes a mud-brick or chineh wall to decay once a building is abandoned. Too many unknown factors enter into the equation. Modern parallels from this area, however, suggest that the evidence for the gap between Periods V and IV at Godin is compatible with a chronological hiatus as short as five to ten years, perhaps as long as twenty-five years, and hardly fifty years.3 Finally, it should be noted that there is good evidence for a hasty abandonment of the Period V building complex on the top of the mound, but no evidence for a violent destruction as the cause of that abandonment. The roof at least of room 22 was burned. A woven reed mat on the floor of room 18 was also burned. Numerous whole vessels or completely restorable vessels were found throughout the
complex on floors. A number of the walls were preserved to a considerable height and their lime plaster
ST. C. Young, Jr. and H. Weiss, " Survey of excavations: Godin Tepe ", Iran XII (1974), pp. 207-1 I, and T. C. Young, Jr., "The day we found the tablet", Rotunda 7 (1974), PP. 4-I1. 2For further details on the topography of Godin Tepe and reports on the excavations prior to 1973, see T. C. Young, Jr., The GodinTepeExcavations: First Progress Report,Royal Ontario Museum Art and Archaeology Occasional Papers 17 (Toronto, I969) and T. C. Young, Jr. and L. D. Levine, The Godin Project: Second Progress Report,Royal Ontario Museum Art and Archaeology Occasional Papers No. 27 (Toronto, 1974)3

A small village built on a mound located just east of modern Sahneh was destroyed by earthquake in 1959 and not reoccupied. When Young first saw the remains in 1961 the walls of the houses, whose roof beams had been salvaged, still stood to a considerable height and one could have drawn a complete ground plan of the village. In 1973 it was not possible to tell from a distance that the mound had been occupied recently. Vague outlines of some walls could still be seen walking over the surface of the site. We estimate that in another five years no trace of the 1959 occupation will be visible under any circumstances.

THE

MERCHANTS

OF SUSA

faces were generally in good repair. All these data suggest a rapid departure of the inhabitants rather than gradual abandonment and slow decay. On the other hand, there is no stratigraphicor structural evidence of deliberate or violent destruction. This conclusion is supported by the kinds of small finds found. Pottery, brokenjar sealings, tablets, tablet fragments and other objects of clay are numerous. Yet metal was rare, and precious metals and stones were missing entirely. In short, the valuable items which could and would be carried off in a hasty but planned retreat are missing. Other items of less value, which might have been carried off had the abandonment been more leisurely, were left in situ. Certainly a more complete selection of finds, such as might result from sudden violent destruction, is not present. ARCHITECTURE We shall describe in detail only the building remains from the Deep Sounding since they are most important given the larger purposes of this paper. In this area there was a single Period V building level which had gone through three phases of expansion and alteration. Discussionis limited to the latest of these, Level V: I (Fig. 2). This consisted of a complex of buildings and rooms set around a large central court (area i), with the whole surrounded by an oval wall (Pl. I). Projecting this oval wall along the lines recovered suggests that it originally enclosed an area about 33 x 21 m., of which we have excavated approximately two-thirds. The oval wall was built of sun-dried mud-brick. Its maximum known width (in the area of room Io) approaches I -50 m. We hesitate to call this a fortificationwall, yet it was a good deal thickerand more monumental than a simple enclosure wall needed to be, and certainly afforded a good deal of protection to those living within. Furthermore,with the mound sloping sharply away on all sides, as described above, the effect must have been very much that of a citadel perched on the highest point of the mound. The only known entrance to the compound is from the south into room 4. The south wall of room 4, falling just inside the excavated area, is the inner face of the oval enclosure wall. Room 4, therefore, is a gateroom, flanked on both sides by smaller rooms. Room 5 with its well-constructedcooking hearth could have been a guard room. Room 3 was probably a store room. Indeed, perhaps it was the file room, for the largest cache of tablets and tablet fragments was found in the strange stepped niches at its southern end. Room 2, beyond a blocked doorway, may have been another store room. Passing north from room 4 across a raised mud-brick thresholdwith a drain one entered the central court (area I). The ill-defined walls which delimited area 7 and room 8 and the small curtain wall separating areas 9 and 11 were not there in earlier construction phases. Thus the courtyard originally m. Why the west end of the court was broken up in the last was a large open space some 17 o10 is unclear. The several floors of the court were very hard to follow in excavation, phase of occupation particularly in the centre of area i. At a minimum, however, there were at least six distinct laid and smoothed mud floors. Scattered over the latest of these surfaceswere eleven uneven patches of burning representingirregularly shaped hearths or fire spots. On the north side of the courtyard stood a monumental building (rooms 14 to 2 I) whose configuration changed considerably over the three phases of Period V. Originally in Level V:3 it included only
rooms 15, 17, 18 and 19. Room 18 was clearly the focal point of the structure. Its walls had been laid out with great care and with architectural forethought. Note particularly the balance of the several elements: the two doors equidistant on either side of the central hearth in the north wall, the twin windows in the south wall looking out into the courtyard with their sills at waist height, the opposed niches in the east and west walls with two larger niches flanking two smaller ones. The hearth was built with equal care, its flues having been constructed with the wall (P1. IIa). It is a fireplace, not a which do not extend down to the floor, but rather end cooking hearth. The niches are proper tokhches in a shelf. Rooms 17, 19 and 20 also had fireplaces, none of which was so elaborate as that in room 18. Late in the life of the building the small curtain wall separating rooms 15 and I6 was built. Room I4, which has a crude cooking hearth against its east wall, was also added after Level V:3. A curtain wall at its north end creates a bin or storage slot in which were found several apparently discarded tablet fragments.

GODIN

TEPE

1973

DEEP SOUNDING

PERIOD V:1

remains GodinV in theDeep Sounding. The latestconstruction of phase. Fig. 2. Plan of thearchitectural

THE

MERCHANTS

OF SUSA

In quality if not size this northern building was monumental. In both design and execution it is a remarkable structure. One strongly suspects that it was a public building and perhaps the central focus of the entire structural complex. Rooms Io, 12 and I3 were probably store rooms of some kind, for they have no hearths or other signs of regular human occupation. Their configuration was dictated by the problem of constructing rectilinear rooms within a space defined on one side by a curving wall; hence the triangular shape of rooms io and 13 and the additional inner skin on the north-west wall of room 12. In an earlier configuration rooms I2 and 13 were a single room. In the south-west corner of the compound was room 6, an independent building in its own right. Almost to the centimetre this room maintains the same proportionsof width to length as does room I8. Again like room 18 the two short walls had four niches, two large niches flanking two smaller ones. There were three large niches in the north-east wall. Opposite is a well-made hearth centred between two doors. The hearth, unlike that in room 18, was for cooking, and consisted of an open fire box on one side and a carefully plastered griddle on the other. Beneath the griddle surface was a deep bed of small pebbles (P1 IIb). The southernmost doorway probably led into a small irregularly shaped room (the area was not completely excavated because we could not easily remove a massive mud-brick revetment wall of Period IV). The northernmostdoorwaygave ratherindirect accessto the courtyardvia a narrowcorridorbetween the south-westwall of room 6 and the oval enclosurewall. Cut off as it is from the open, more public areas of the compound, room 6 has all the characteristicsof a private quarter. Unfortunately, we were able to excavate only a part of the large building on the east side of the courtyard (rooms 22 and 23). The floor of room 23 was littered with a thick deposit of oxidized peas, barley and wheat and probably served for food storage. Room 22 may have been the main salon of this building, though unlike room 18 of the northern building it had a doorway giving direct access to the courtyard. The north wall had two large and three smaller niches. A doorway in the rear wall led east, which, by analogy with rooms 6 and 18, was almost certainly balanced by another doorway further to the south. A single row of rooms behind room 22, similar to those on the north side of the northern building, would probably have filled the east end of the oval. Room 22 had been destroyed by fire. Charred roof beams lay on the floor and in the debris of the collapsed walls and ceiling immediately above the floor, and the white lime plastered walls had been burned red and black. The Period V architectureof the Brick Kiln Cut is entirely different. As noted above, it represented only a rebuilding and slight modification of existing Period VI houses. The houses were ordinary in every respect, though sharing many of the detailed characteristicsof the Period V architecture on the summit of the mound, such as white, lime-plastered walls, well-built hearths and carefully-laid brickwork. They remain, even so, parts of simple village houses built by prosperouspeasants. The rooms were rectilinear. Segments of perhaps two separate complexes were recovered, but without a complete house plan we cannot say whether or not the rooms focused on a courtyard. There is no evidence that this part of the Period V town was protected by an enclosurewall. SELECTED POTTERY AND SMALL FINDS
It would be premature and is unnecessary to attempt now any systematic or detailed presentation of the pottery and small finds recovered from Period V at Godin. What should be discussed here are some of the general categories of materials recovered, certain unusual aspects of their distribution in the deposit, and parallels with other sites involving selected items.

Ceramics The pottery of Godin V has been described elsewhere in some detail as regards types and wares.4 For our purposes here it may be divided into two categories: that which represents only a
4 For a preliminary description of Period V ceramics, based on the x965 sounding, see Young, First ProgressReport, pp. 7-8 and fig. 9. Though now out of date and incomplete, this description is still essentially correct.

JOURNAL

OF PERSIAN

STUDIES

continuation of the ceramic traditions of Period VI, and that which is new to Godin and distinctive to Period V. The Period VI ceramic culture, while almost certainly influenced in minor ways by generalized Ubaid features characteristicof lowland Mesopotamia, is in the main a native tradition. It is well established at Godin and elsewhere in the region long before the arrival of the new elements characteristic of Period V, and continues in Period V as a major component of the total ceramic assemblage.' Of the new elements which specifically define Period V pottery and set it apart from the Period VI assemblage, almost all involve ceramic features which can be paralleled in lowland Mesopotamia. The variable distribution of these two categories of pottery in the Period V deposit is striking. Within the oval on the summit of the mound about half of the pottery recoveredis specific to Period V; the other half is typical of Period VI. In the Brick Kiln Cut, however, perhaps as much as eighty per cent of the pottery is of Period VI type, and only twenty per cent is distinctive to Period V. For purposes of comparative chronology we propose to place greatest weight on these ceramics. It seems clear that the stratigraphicand chronological ranges of related epigraphic and glyptic material (see below) are neither as precisely defined, nor as limited in time as are those of related ceramics. For the sake of brevity, we have isolated for discussionfour-luggedpots and bevelled rim bowls from the mass of ceramic materials excavated precisely because of the rapid changes in the presence or absence of these vessels in successive deposits elsewhere. Sherds from horizontally pierced four-lugged pots with " rope "-appliqu6 bands along their shoulders were found in large proportionsamong the in situfloor remains of room 18 in the northernbuilding (P1. III and Fig. 2) and the courtyard. A plain variety occurs in a large-size grit and straw-tempered ware with a red slip (Fig. 3:1) and a medium-size grit and straw-tempered ware with a cream slip (Fig. 3:2a). Comparable vessels are reported from Susa Acropolis level 17 and Warka Eanna IV, and begin to appear in Nippur Inanna level XIX.6 These vessels are not reported in later strata at Susa or Warka. Sherds from a cream slipped variety of this vessel with an incised shoulder band of cross-hatched triangles were found on the courtyard floor (Fig. 3:Ia). Closely similar pots are present in Susa Acropolis level I7A, Warka Eanna IV and Nippur Inanna level XVI.7 These vessels are also not present in succeeding strata at these sites. A red slipped variety, with two bands of black paint, comes from room 19 (Fig. 3:2). Four-lugged jars with painted bands above the " rope "-appliqu6 shoulders are specific to Susa Acropolis level 17, and occur as well in Malyan Banesh TT-F.8 Missing at Godin are four-lugged pots with close geometric painted designs on their shoulders, as occur in Susa Acropolis levels 16-14B and Yahya IVC.9 Neither do we find any of the tall " Jemdet Nasr " related storagejars with geometric design shoulder painting as occur in Malyan Banesh Building Level II, Yahya IVC and " Jemdet Nasr " period levels of Nippur Inanna levels XIV-XII.10 Turning to bevelled rim bowls, the total absence of other mass-produced vessel types within the Godin V assemblage is as significant as the proportionally large number of these vessels present (Fig. 3:3-5). Bevelled rim bowls and " Groben Blument6pfe" occur together in Warka Eanna IV, but by Warka Eanna III bevelled rim bowls have been replaced by " Blument6pfe ".11 Bevelled rim bowls appear in Susa Acropolis level 17, and are still present in level I6, but " Blument6pfe " begin to appear in that level as well.12 In Malyan Banesh TT-F bevelled rim bowls, " Groben Blumentipfe "
and " Blument6pfe" all occur together.18
8

Le Brun, op. cit., Fig. 53:4; W. Sumner, " Excavations at Tall-i Malyan, 1971-1972 ", Iran XII (1974), P. 70o,Fig. 4:h. 9 Le Brun, op. cit., Fig. 64:3-12; C. C. Lamberg-Karlovsky, pottery. 6 A. Le Brun, " Recherches and the Indo-Iranian stratigraphiques A l'Acropole de "Tepe Yahya 1971-Mesopotamia Suse (1969-I971) ", Cahiers DAFI I (1971), Fig. 50:1; D. Borderlands ", Iran X (1972), p. 95, Fig. 2:G. Hansen, " The Relative Chronology of Mesopotamia, Part II ", 10 Sumner, op. cit., p. 171, Fig. 5:a-d; Lamberg-Karlovsky, Chronologiesin Old World Archaeology,R. W. Ehrich, ed. op. cit., p. 95, Fig. 2:G; Hansen, op. cit., p. 207. (Chicago, 1965), p. 203, Fig. iib; von Hailler, Vorldufiger x1 H. J. Nissen, " Grabung in den Quadraten K/L XII in Berichte... in Uruk (hereafter U.V.B.) IV, Taf. I9c:k. Uruk-Warka ", Baghdader Mitteilungen5 (1970), pp. I32-42. SLe Brun, op. cit., Fig. 51:9; von Haller, op. cit., Taf. 2oA:l'; 12sLe Brun, op. cit., Fig. 60:1-4. Hansen, op. cit., p. 205, Fig. 21. 1xW. Sumner, op. cit., p. 170, Fig. 4:b, c, d. ' Ibid., pp. 5-6 and fig. 8, and Young and Levine, Second Progress Report, pp. 12-14, for a detailed description of Godin VI

THE

MERCHANTS

OF

SUSA

cm

10

20

II| I

10

cm I

Fig. 3. Potteryvessels from GodinV whichhavecloseparallelsin lowlandMesopotamia.

JOURNAL

OF PERSIAN

STUDIES

In sum, the absence of " Blument6pfe" and " Groben Blument6pfe" in Godin V, along with the absence of geometric painted four-lugged vessels, suggests that Godin V is earlier than Susa Acropolis level 16, Malyan Banesh Building Level II, and Yahya IVC. The presence of plain and cross-hatched shoulder " rope "-appliqu6 four-lugged vessels at Godin V indicates that the settlement can be no earlier than Susa Acropolis level 17. Tablets A total of forty-threetablets or tablet fragmentswere found (Figs. 4 and 5). Of these, twenty-seven are complete or only minimally chipped. One complete tablet was totally blank and twenty-six were inscribed with numerical notations and one with a pictographic sign as well. Nine of these were rolled, one or more times, with cylinder seals. Twelve fragmentswere large enough to determine two dimensions and/or the tablet's shape. Of these, ten bear numerical notations and two are impressed with cylinder seals. Four tablet fragmentsare too small to provide informationon their original dimensions. None of these bears an inscription, but one preservesa cylinder seal impression. The twenty-seven complete tablets have arithmetic mean dimensions of 5 - 2 x 3 -8 x I 9 cm., with and 0o25. Apart from two tablets with plano-plane respective standard deviations of o'41, o'44 surfaces and four tablets with plano-convex surfaces, all tablets are slightly convex-convex in shape with sides, top and bottom flattened or slightly concave. The concave sides often reveal how the tablet had been shaped by the scribe's fingers. The dimensions and shapes of the fragmentary tablets fall within the ranges of the complete tablets. The notational system appears to use five different numerical signs, all well attested in the protoElamite and proto-Sumerian signaries.14 The thirty-six complete and fragmentary tablets bear inscriptions of wedges and dots made with the same stylus, occurring together and singly; fingernailimpressedcrescents; paired,joined dots; and largerverticalwedges. The one pictograpicsign is similar to a sign present in the Warka IVa and proto-Elamite signaries.15 The largest single cache of tablets came from the recessedniches at the south end of room 3. Several were found in rooms 15 and I8, and the second largest group came from in and around the bin at the north end of room 14. These latter tablets were fragmentary and one suspects that they had been discarded when no longer needed. Complete tablets or fragments were found in several other rooms and in the courtyard. None were found in room 6. Almost all were either on floors or in the primary deposit immediately above the floor. No tablets were found in the Brick Kiln Cut. The presence of a tablet blank from Godin V indicates that at least some tablets were manufactured locally. That none of the Godin tablets are baked might indicate that they were not manufacturedfor transport to another location. Our experimental manufacture of a tablet from local clay, however, indicated that sun-drying alone will produce a product of considerable durability. The manufacture and use of such tablets is a complex cultural phenomenon. The presence of tablets of similar size and shape, with similar sealings and inscriptions, at other contemporary sites indicates the probable source of the tablet tradition of Godin V, if not the source of the Godin scribes themselves. Numerical notation tablets are now known from Khafajah, Habuba Kabira, Sialk, Chogha Mish, Tall-i Ghazir, Warka and Susa. The one numerical notation tablet from Khafajah was located on a trodden floor estimated to be some five building levels below House Level I2 in a test trench (N44:19) just outside the later Temple Oval.16 The tablet is unsealed, roughly formed, possibly with finger-moulded sides like the Godin tablets, and falls within the range of the latter's dimensions. Frankfortand Delougaz date the stratum in which this tablet was found to the Uruk Period.
14

R. de Mecquenem, Epigraphie Proto-Elamite,MMAI XXXI Texteaus Uruk,ADFU (Paris, 1949); A. Falkenstein, Archaische 2 (Berlin, 1936); A. A. Vaiman, " A Comparative Study of the Proto-Elamite and Proto-Sumerian Scripts," VDI 121 (1972), pp. 124-133. Prof. W. W. Hallo of Yale University is working on the Godin tablets. 15 Falkenstein, op. cit., pp. 39-40, sign no. 139; de Mecquenem, op. cit., Pl. XXV, sign no. 1275, PI. LIV, signs 4210-4218;

V. Scheil, Textes de ComptabilitiProto-Elamites,M.D.P. XVII (Paris, 1923), Pl. XLIII: 303r. 16 H. Frankfort, Progressof the Workof the OrientalInstitutein Iraq, 1934/35, OIC 20 (1936), p. 20o, Fig. 19; P. Delougaz, Pottery from the Diyala Region, OIP LXIII (Chicago, 1952), P. 34; P. Delougaz, H. D. Hill and S. Lloyd, PrivateHousesand Graves in the Diyala Region OIP LXXXVIII (Chicago, 1967), p. 2.

Pl. I. View looking east across the architecturalremains of Godin V: I in the Deep Sounding. The northernbuilding with its main salon, room I8, is in

Pl. IIa. Close-up of the hearth against the north wall of room i8.

6 on hearthin room witha fire boxto therightanda plaster laid a of griddle over bedof pebbles theleft. P1. IIb. Close-up the cooking

on and half P1. III. The lowestlayerof thein situ pottery smallfinds scattered thefloorof roomi8. The northern of the roomhad in beencleaned an earlierstageof theexcavations.

Pl. IVa. Impression a cylinder from GodinV, Gd. 73-210. seal of

seal of Pl. IVb. Impression a cylinder from GodinV, Gd. 73-260.

THE

MERCHANTS

OF SUSA

OBVERSE

TOP

..

OBVERSE

(I

BOTTOM

REVERSE

INDICATESMISSINGVIEW

II

cm

Fig. 4. A selection tablets of from GodinV.

10

JOURNAL

OF PERSIAN

STUDIES

0 1

5 I 1 I I
Icm

tabletsandothersmallfindsfrom GodinV. Fig. 5. Selected

THE

MERCHANTS

OF SUSA

11

An unspecified number of numerical notation tablets bearing seal impressions are reported from what appears to be a Late Uruk period settlement at Habuba Kabira.17 This assemblage has also yielded seal impressionssimilar to some found at contemporaryWarka and Susa, and four-luggedjars The eighteen numerical notation tablets and fragments from Sialk IV, include five with protoElamite signs.'8 Although there are glyptic parallels between the assemblagesof Susa Acropolis level 17 and Sialk IV,, and some parallels with the Godin V tablets, Sialk IV, is probably not contemporary with Godin V and dates no earlier than Susa Acropolis 16. Sealed numerical notation tablets are reported from Proto-literate Chogha Mish, but no further information on them is as yet available.'9 The context of a tablet fragment with numerical notations from Tall-i Ghazir is uncertain, although the stratum from which it came is identified as " protoElamite ".20 similar to Gd. 73-403 (Fig. 3:Ia).

Although a large number of numerical notation tablets from Warka have yet to be published in detail, it is abundantly clear that many are rectangular, blunt-edged, slightly convex-convex, and fall within the range of the Godin tablet sizes.21The sealings on these tablets share a number of well-known elements with those of Susa Ca/b and with Godin V as well. Though these Warka-Godin V parallels are clear, yet it is the tablets from Susa which can best be genetically related to those of Godin V. Within the several series of Susa tablets distinguished by Amiet, two present close parallels to the Godin V tablets. Four tablets with numerical notations, one of which is also inscribed with an arrow-shapedsign, conform to the Godin tablets in size and shape. They come from Susa Ca/b.22 The sealings on this group of tablets include two designs present on the Godin tablets, amphikyphelloi archers (see below). Eight tablets of another grouping assigned to and Susa Cc are described as rectangular and " fortement bombees, en forme de coussinets ".23 Three of these bear one pictographic sign in addition to their numerical notations. The dimensions of two of these tablets are given and conform to the sizes of the Godin tablets. The depiction of running caprids is a seal design element common to this group of tablets and to the Godin glyptic assemblage. Some tablets also bear impressionswith " ovals " and " fish " designs which Amiet relates to the impressions present in Susa Acropolis level 17.24 The one numerical notation tablet from Susa Acropolis level 17 is similar in shape and size to those from Godin V.25 Finally, although the Susa Acropolis level 17 assemblage does not include tablets which employ a pictographic sign, the common sealings of the Godin V and the Susa Cc tablets which employ a pictographic sign indicate that the latter are probably contemporary with Susa Acropolis level 17. In sum, a fair number of complex parallels can be drawn between Godin V tablets and tablets from Susa Ca/b, Cc and Acropolis level 17. Significantly, several of these parallels involve design elements which are common to the Godin and Susa assemblages, but which are not found at Warka. Glyptic A total of thirteen seal-impressedtablets, four sealed jar stopper fragments and two cylinder seals were recovered dating to Period V. Except for one cylinder seal, everything came from a good stratigraphic context within the oval enclosure. The seal Gd. 73-21o (Fig. 5:8 and P1.IVa) can be assigned to Period V only on stylistic grounds, for it was found inside a Period IV mud brick. No particular concentration of glyptic materials could be observed in the deposit, though the majority of the jar sealings and sealing fragments,both seal impressedand plain, came from in or around rooms Io, 12 and i 13 and from areas 9 and I1 in the courtyard. The presence of a seal blank indicates that at least some Godin seals were produced locally. No glyptic materials were recovered from the Brick Kiln Cut.
E. Strommenger, " Ausgrabungen in Habuba Kabira und Mumbaqat ", Archivfiir Orientforschung (I973), pp. 168-71. 24 18R. Ghirshman, Fouillesde Sialk I (Paris, 1938), p. 67, P1. XCII. 19 P. Delougaz, Memorial Volume, Vth InternationalCongress of Iranian Art and Archaeology, 1968 I (Tehran, 1972), p. 27. 20 D. Whitcomb, The Proto-ElamitePeriod at Tall-i Ghazir (unpublished M.A. Thesis, University of Georgia, 1971), p. 31, P1. XI:A.
17

21E.g., van Dijk, U.V.B. XVI (I960), p. 57, Taf. 29e, g; 3oe, h; 31a, b. 22 P. Amiet, GlyptiqueSusiennedes Origines a l'Epoque des Perses M.D.A.I. XLIII (1972), Nos. 60o2,604, 629, achme'nides, 657. 23 Ibid., pp. 129, 292; nos. 925-8, 2318-21. 24P. Amiet, "La glyptique de 1'Acropole (1969-i971) ", CahiersDAFI I (1971), p. 225. 2s Le Brun, op. cit., Fig. 44:8.

12

JOURNAL

OF PERSIAN

STUDIES

The following selected tablet sealings, seals and bullae fragments illustrate the chronological and spatial range of possible influences on the Godin V glyptic. Our discussion focuses on only a few specific design elements and their occurrence elsewhere. A systematic and complete analysis of the Period V seals and seal impressionswill be presented in our final report.26 The tablet fragment, Gd. 73-54 (Fig. 4:I), was rolled, perhaps with two different seals, on the obverse and extant three sides. The reverse of this tablet seems to have been scraped clean of any one impression. The obverse, upper left, preservesa still legible impression of an amphikyphellos, of the recurrentfeatures of Warka V-IV and Susa Ca/b impressionson both tablets and jar stoppers.27The recent excavations on the Acropolis at Susa produced one example of this motif on a numerical notation tablet from level 17.28 The Godin sealing is unique in its use of the " drill centred circle " with the One amphikyphellos. other Godin V tablet, not illustrated, has a seal impression with an amphikyphellos One seal was twice rolled over the obverse of the tablet Gd. 73-295 (Fig. 4:2) and once over its two edges and reverse. Seated lions with forepaws set straight, and with raised, curled tails also occur on impressionsfrom the Warka Limestone Temple and in Susa Ca/b.29 The stakes set between the lions resemble those between seated bears on a sealed clay counter from Susa assigned to Susa Ca/b.30 Two different seals may have been rolled over the obverse and reverse of the tablet Gd. 73-320 (Fig. 4:4). A squatting archer with drawn bow occurs on numerical notation tablets of Susa Ca/b;31 a standing archer appears on a numerical notation tablet with one pictographic sign assigned to the same group.32 The complete tablet Gd. 73-64 (Fig. 4:5) was rolled by one seal over all its surfaces. Rearward glancing lions with forward curled tails occur in Warka IV and Susa Ca/b.33 One impressionfrom the latter context presents all the components of Gd. 73-64 with the exception of drill centred circles; another presents a similar lion in association with an amphikyphellos.34 The unbaked jar sealing Gd. 73-326 (Fig. 5:4) bears the impression of running goats similar in position and style to those which first occur at Susa in Cc levels.35 Sin Temple IV and V at Khafajah are the earliest that similarsealsappear in the Diyala.36 A comparableimpressionrecoveredfromWarka was assigned to the " Jemdet Nasr " period. Earlier depictions of running goats occur at these sites, but not in this style. The lines which pass by the horns of the Godin impression's goats do not seem to be related to the streamersattached to the goats on earlier Susa impressions,and remain enigmatic. The tablet Gd. 73-153 (Fig. 5:5) was seal impressedon all extant surfaces. The rearwardglancing lions with intertwined tails occur on Warka V-IV and Susa Ca/b glyptic. The stylized tree, however, only occurs on seal impressionsof proto-Elamite levels, Cc-Da at Susa.37 Two files of bulls and lions, probably led by a human figure, appear on the jar stopper Gd. 73-329 (Fig. 5:6). Similar animal processionsare known from Susa, e.g. on a clay counter bulla dated to Ca/b, and from Warka sealings dated to Eanna IV.38 We have been unable to locate any good parallels for either of the two cylinder seals from Godin V (Fig. 5:7 and 8, P1. IVa and b). The seal Gd. 73-2 o (Fig. 5:8) is particularly interesting in that its drill design is unique, although its individual design elements are well known. Drill centred circles and radial stars are used as fillers in the composition. Indeed, the use of drill centred circles (cercle pointi) as a filler is a significant general characteristicof the Godin V glyptic. It also occurs on four of
thirteen tablet sealings (Gd. 73-54, -64, Fig. 4:1 and 5; and Gd. 73-297 and -319 not illustrated). (Gd. 73-161).

Drill centred circles and radial stars are elements of a " regional" style of seals, usually of glazed
Miss Mina Sadegh is undertaking this study. E. Schott, U.V.B. V (I934), Taf. 26:b; 28:a; Amiet, Glyptique Susienne,nos. 501, 535, 536. 28 Le Brun, op. cit., Fig. 44:i1. 29 E. in Heinrich, Kleinfunde aus den archaischenTempelschichten Uruk, ADFU I (Berlin, I936), Taf. 15:g; Amiet, Glyptique Susienne,nos. 501, 535, 536. so Ibid., no. 574'1 Ibid., nos. 6oi, 602.
27
26

Ibid., no. 604. " Schott, op. cit., Taf. 26:f; Amiet, Glyptique Susienne,no. 521. '4 Ibid., nos. 52!, 53735 Ibid., nos. 92g, 924, 931. Sealsfrom the Diyala Region, OIP Cylinder 36 H. Frankfort, Stratified LXXII (Chicago, 1955), nos. 76, 219. Susienne,nos. 997, o000, 1014. 7 Amiet, Glyptique " Ibid., no. 549; H. Lenzen, U.V.B. XXIV (1968), Taf. Ig:a.
3"

THE

MERCHANTS

OF SUSA

13

steatite, which Le Breton identified at Susa, in the Diyala and in northern Mesopotamia,39and to which distribution we can add the 'Amuq.40 From earlier excavations at Susa the use of drill centred circles is documented in four stylistic contexts, all dated to Susa Cc.41 Recent excavations at Susa produced one seal of glazed steatite, lacking drill centred circles but in this " regional " style from Acropolis level I6.42 One seal which uses a radial drill centred circle star as filler for an animal representation design comes from Acropolis level 17. Amiet assigns this seal to the " regional " style.43 The origin and transmissionof the " regional " style has been discussed by several scholarswithout achieving a consensus.44 More research and discussion is needed. Suffice it to note here that the appearance of the " regional" style in Susa Acropolis level 17 suggests that the style appears in Khuzistan at least as early as it does elsewhere. Until more evidence is in hand, the use of drill centred circles in the glyptic of Godin V should probably be linked with the earliest appearance of this feature in Susa Acropolis level 17. It should be noted that the stylized, almost geometric patternsof plants and animals, which apart from the drill centred circles characterized the " regional " style, are not present at Godin Tepe. This perhaps suggests that thefloruit of this style in Susa Cc is later than Godin V. In the end, if we assume that the Godin V glyptic is the product of a single source of foreign influence, and given the present nature of the data that is perhaps wise, then that source, despite glyptic heraldic lions, seated lions, and animal parallels with Warka, is most probably Susa. Amphikyphelloi, files are all design elements shared by Susa, Warka and Godin V. The squatting archer, stylized plant, and the use of drill centred circles, however, are specific to Susa and Godin. The date of the Susa parallels is, with two exceptions, Susa Ca/b. Running goats such as appear on the Godin jar sealing Gd. 73-326 (Fig. 5:4) and the stylized plant on Gd. 73-153 (Fig. 5:5) are both features of Susa Cc, though the former has antecedents in Susa Cb. It would not be surprising, of course, if certain elements of the Susa Cc glyptic had their origins as early as Susa Acropolis level 17. DISCUSSION AND CONCLUSIONS Period V, as distinct from Period VI at Godin, representsa relatively short occupation. There is no occupational or chronological gap between these two periods. Indeed, the ceramic traditions of Period VI, a local central western Iranian phenomenon, continued into Period V. In contrast, the latter representsa cultural phase with strong connections beyond the Godin region. In a certain sense, therefore, Period V could be described as no more than a phase of major foreign influence at Godin coming late in Period VI. Elements betraying these foreign contacts are found in both the Brick Kiln Cut and the Deep Sounding, but are most frequent and noticeable in the latter context. Furthermore, the two contexts themselves are strikingly dissimilar. In the Deep Sounding we have a large monumental building complex surroundedby an oval enclosurewall standing on the ancient top of the mound. The buildings within this compound show signs in form and content of having had both public and private functions. The northern and eastern buildings were most probably secular, public structures with either a political or economic function,or both.45 Room 6, in contrast,was probably a private apartment. Other rooms served as gate structuresor store rooms. The rooms excavated in the Brick Kiln Cut at the base of the Citadel Mound were quite different. Here are the remains of prosperous,private village houses.
One interpretation which emerges from these data suggests that the foreigners, or the element of the local society most influenced by foreign ways or most closely in contact with foreigners, were living in
L. Le Breton, " The Early Periods at Susa, Mesopotamian Relations ", Iraq XIX (1957), P. o108. 4o R. J. and L. Braidwood, Excavationsin the Plain of Antioch I, OIP LXI (Chicago, I960), Figs. 254:2, 3; 297:5. Susienne,nos. 825, 828, 879, 881. 41 First group, Amiet, Glyptique Second group, ibid., nos. 1036, o1041, o43, 1044, 1055, 1o56, 8o65. Third group, e.g., ibid., nos. xo86, I151. Fourth group, ibid., nos. Io49, 8o55, o63. 42 Le Brun, op. cit., Fig. 59:4. " La 48 P. Amiet, glyptique de l'Acropole ", p. 222; Fig. 44:5.
89

44E.g. U. Moortgat-Coorens, " Bemerkungen zur Glyptik des Diyala Gebietes", O.L.g. LIV (1959), p. 342-54; R. H. Dyson, Jr., " The Relative Chronology of Iran, 6ooo-20oooB.C.", Chronologiesin Old World Archaeology,p. 225; W. Nagel, und Berliner Buntkeramiker, Djamdat Nasr-Kulturen Friihdynastische 8 Beitrage zur Vor-undFriihgeschichte (Berlin, I964), P. 44. 45 It might be tempting to see either the northern or the eastern buildings of the complex as religious structures, but there is no evidence from Godin which suggests this is the case.

14

JOURNAL

OF PERSIAN

STUDIES

somewhat isolated circumstanceson the summit of the mound, while the people occupying the lower town, though touched by these outside influences, generally pursued older, more local cultural patterns, perhaps in more traditional ways. Eventually, the oval enclosureon the top of the mound was abandoned in some haste, but probably not violently destroyed. We may assume the rest of the site was abandoned at about the same time. In a short while a completely new culture, totally unrelated to anything heretoforeknown from Godin, replaced the culture of Period V on the mound.46 On the whole, this much is clear from Godin itself. Our ability to isolate more precisely some of those foreign influences in Godin V, however, makes it possible to carry our interpretationstill further. The closest parallels for the foreign elements in the Godin V assemblage are found in lowland Greater Mesopotamia at Susa, Warka and perhaps Khafajah. Our discussion of pottery, tablets and glyptic, however, has in each instance led us to Susa and Khuzistan as the site and area with the closest and most numerous connections with Godin V. Leaving aside for now the important question of what the relations between the several regions within lowland Greater Mesopotamia may have been, a subject which might profitably be re-examined in the light of recent discoveriesin the highlands, let us assume here that the simplest explanation of our data is that the source for the exogenous influences found in Godin V is Susa. We are then faced with three possible explanations for the presence of these materials at Godin Tepe: production at Susa followed by transport to Godin, local production by the indigenous Godin population, or local production by Susians. On the whole the last explanation seems more probable. We have already mentioned the evidence for local production in the form of seal and tablet blanks, to which might be added the difficulty of transportinglarge quantities of bevelled rim bowls and fourlugged jars from Khuzistan to the Kangavar valley. It is possible that indigenous potters were producing excellent imitations of Susa-relatedceramicsat Godin, but this seems much less likely as an explanation for the Godin V seals and tablets. The recording of economic transactions on sealed numerical notation tablets involves a complex of highly conventionalized actions. The replication in considerable detail of a Susian recording system at Godin by indigenous scribes and seal cutters seems improbable. More likely than not, the Period V " acropolis" tablets, seals and ceramicswere the product of resident Susians engaged in commerical transactionsat Godin Tepe. What sorts of commercial transaction were involved? Again accepting the simplest explanation given the evidence available, we suggest that the Godin V oval enclosure on the summit of the mound was a Susian trading post immediately supported by a local agricultural village or town. Such an entrep6t could have been involved in long distance trade, in strictly local trade, or in both. Let us here consider the former possibility in some detail. A Susian trading post at Godin Tepe would have served ably as a tap controlling and furthering the flow of trade channelled along the Khorasan Road. Although we have not identified within the Godin V assemblage the goods or materials which might have been the object of such trade, certainly the Godin tablets could be records of such trade transactionsand of goods destined not for Godin but for points further west and perhaps for Susa itself. Under circumstancesof deposition and abandonment which might be quite different from those which obtained at Godin, at both Sialk IV and Yahya IVC proto-Elamite tablets occur with such goods within assemblages whose mix of indigenous and Susa-related ceramics could be a slightly later analogue to that of Godin Tepe.47
Susa's sudden and relatively short-lived control of Godin for purposes of trade could be involved in shifting patterns of the lapis lazuli trade which followed the Khorasan Road. Lapis lazuli, as well as ceramic and glyptic styles, passed freely along the Khorasan Road throughout the periods of Gawra XIII-XI. By late Uruk Gawra IX, however, these imports slacken off, almost disappearing by Gawra VIIIC. Even prior to the excavation of Godin V, Georgina Herrmann suggested that the demise of
46 Godin IV represents a penetration southeastwards of the

"Early Transcaucasian Culture" or "Yanik Culture ", centred in the Caucasus and eastern Anatolia and first excavated in Iran at Yanik Tepe in Azerbaijan. See C. A.

Burney and D. M. Lang, The Peopleof theHills (London, 1972), p. 59; First ProgressReport,pp. 9-Io. * J. R. Alden, The Questionof Trade in Proto-ElamiteIran (unpublished M.A. Thesis, University of Pennsylvania, 1973).

THE

MERCHANTS

OF SUSA

15

this trade " was due to Elamite expansion onto the plateau, which blocked the route to the northern (Mesopotamian) cities, and which acquired control of the trade for Susa ",.48 The suggestion that it was through Godin, and perhaps similar sites along the Khorasan Road, that Susa tapped the northern trade route for lapis lazuli and other goods requiressome evidence for a mechanism by which these materials were transferredfrom this area to Susa. Our attention is drawn to the large Uruk period sites which dominate the valleys of western Luristan. Here, according to their surveyor, " the closest parallels for the buff ware assemblages, both in shape and fabric, come from
Godin V

seem to be representedin Luristan by a change in culture" may very well indicate that with the abandonment of sites like Godin V Susa's interest in Luristan ceased.50 The departure of Susian traders from Godin Tepe at the end of Period V could be seen to correspond with the disruption of trade further east along the Khorasan Road caused by the arrival of the Yanik Culture in the area. Moving along the eastern flanks of the Alvand alignment, the people who brought the Yanik Culture south from Azerbaijan and the Caucasus, first crossed the Khorasan Road in large numbersin the Hamadan plain, some Ioo km. east of Godin.51 Shortly after the abandonment of Godin V, the Yanik Culture, represented locally by Godin IV, arrived in the Kangavar valley. Perhaps at about the same time it also appeared at Malayer, some 50 km. furthersouth. This latter occupation is " reported to underlie sherds of Susa D type ", suggesting that the start of the Yanik Culture's control of the Hamadan plain further north could be contemporaneous with late Susa C and thus not very long after the start of the Susian settlement of Godin V.52 This interruption of trade routes further east could be an explanation for the abandonment of Godin V and the retreat of the Susian traders from central western Iran. Yet Susa probably remained interested in the Khorasan Road and northern trade routes. Sialk, safely east of the Yanik Culture's intrusion, was settled by Susians now employing proto-Elamite signs on their tablets and thus coming to the edge of the Dasht-i Kavir shortly after the abandonment of Godin V. Unfortunately, the end of the Sialk IV occupation remains ambiguous. Finally, during Susa Acropolis levels I6-I4B times Susa's persistentinterestin an easterntrade was directed along the southeastern route which passed through Tall-i Ghazir (Stake Io Room), Malyan (BaneshBuilding Level II) and Yahya IVC. In summary, we have presented for discussion some of the exogenous components of the Godin V assemblage in their architectural and stratigraphic setting. We have established the settlement's relative chronology, identified Susa as the prime source of foreign influences and Susians as the local producers of the Godin V assemblage, and have then concluded that Godin V was a Susa-controlled trading post. Following this we have tried to view Godin within a sequential pattern of Susian trade expansion: the tapping of the Khorasan Road, the disruption of that trade flow by the intrusion of the people bearing the Yanik Culture, the re-establishmentof Susa's trade with the east through Sialk, and then the shift to the south-eastern trade route through Tall-i Ghazir, Malyan and Yahya. This presentationis, of course, both sketchy and hypothetical. A complete and systematic examination of the Godin V material will no doubt encourage us to revise these suggestions and may lead to entirely different conclusions. Yet even so preliminary an inspection and interpretationof the Godin V assemblage has allowed us to phrase entirely new sets of questions about lowland-plateau relationships in Late Uruk and Jemdet Nasr times.
On a primary level, through an analysis of the Godin V assemblage and its foreign relations, we are now concerned with a refinement of the relative chronology of the plateau and the lowlands in this period and of the indices by which to evaluate the source of outside influences on the plateau. On another level we are concerned with the nature of the contacts behind those connections. While we
48 G. Herrmann, " Lapis Lazuli: The Early Phases of Its Trade ", Iraq XXX (1968), pp. 36, 37, 53, after P. R. S. Moorey. 49 C. L. Goff, " Luristan Before the Iron Age ", Iran IX (1971),
p. 145. 50 Ibid.
51

.49 Goff's further observation that the "Jemdet Nasr phase in Mesopotamia ...

does not

R. H. Dyson, Jr., " The Archaeological Evidence of the Second Millennium B.c. on the Persian Plateau ", C.A.H. II, 3rd ed., pp. I4-16. 52 Ibid.

16

JOURNAL

OF PERSIAN

STUDIES

have posited a trading function for the Godin V settlement, we are still searching for direct evidence of the objects traded. Thus on yet a third level we are concerned with the system of that trade itself. Lacking tangible evidence of the goods traded, we are not in a position to understand the economics of this trade. Both parties, however, were finding it either profitable or an absorbable expense. Nowhere at Godin do we find evidence for an indigenous resistance to the Susian presence. On the contrary, our data indicate that the children, and perhaps the grandchildren,of the Period VI residents continued to live at Godin Tepe after the Period V traders arrived. This despite the fact that where the houses of an indigenous Period VI population once stood on the summit of the mound we suddenly find an enclosed compound of foreigners. By what means was this at least locally symbolic domination of Godin achieved? From these observations and concerns follows a series of questions. What were the mechanisms by which this Susian outpost was maintained? Are the legumes and grains stored in room 22 local products acquired from indigenous farmers, or local products harvested by landowning Susians? Can we relate the manufacture of bevelled rim bowls and their presence in both the Deep Sounding and the Brick Kiln Cut to the control and distribution of these comestibles? Looking to a still wider horizon our attention is drawn to the need to examine the large Uruk period sites in the valleys of Luristan. Might these, as Goff has suggested, representindigenous, intravalley consolidations of social and economic power, or are they stages on a trade system? Even more intriguing are the valleys west of Godin along the Khorasan Road. The presence of several large Uruk period sites in the Mahi Dasht and Shahabad valleys must be included in any future integration of the data. These large and fertile intermontane plains control the westernmost passage of the Khorasan Road before it reaches the lowlands of the Diyala. In Late Uruk times they stood midway between settlements like Godin and Khafajah, which at least shared the use of numerical notation tablets. Did trade along the Khorasan Road continue west into the Diyala and thence into southern Mesopotamia? And what of connections between Susa and the Diyala, so clearly seen in seal motifs of Susa Cc times? Might Godin V be the result of an expansionof influencefrom the Diyala into the highlands? As the search creates the questions, so the questions direct the search. It is time to do some work in the Mahi Dasht.53

Is

We wish to thank the following people who helped us in preparing this article. Claus Breede, who took the photographs and did the pottery and architectural drawings: Linda Ritchie, who drew the small finds: Louis Levine,

Carol Hamlin, Matthew Stolper, C. C. Lamberg-Karlovsky, Henry Wright,William Sumnerand Irene Winter, who read at and commentedon the manuscript one stageor anotherof its production.

THE

MERCHANTS

OF SUSA

17

CATALOGUE Fig. No. 3: I 3: Ia Field No. Gd. 73-114 Gd. 73-403 Description Red slipped Coarse Ware. Reddish. Burnished. Wheelmade. Cream slipped Common Ware. Cream buff. Smoothed. Wheelmade. Smokeblackened in spots on the inside. Shoulder incised with three bands of hatched triangles. Red slipped Common Ware. Red. Faint bands of red or black paint on carination. Partly smoke-blackened on outside. Smoothed. Wheelmade. Cream slipped Common Ware. Cream. Smoothed. Handmade. Shallow bevelled rim bowl. Plain, straw-tempered Coarse Ware. Tan. Matte. Handmade. Large shallow bevelled rim bowl. Plain, straw-tempered Coarse Ware. Tan. Matte. Smoothed inside. Handmade. Finger sweeps on outside. Deep bevelled rim bowl. Plain straw-tempered Coarse Ware. Red buff. Matte. Large white stone inclusions. Tablet fragment. Well-levigated clay. No visible inclusions. Reverse scraped clean. Tablet. Medium well-levigated clay. Some visible inclusions. Mica flakes. Surfaces fairly rough. Tablet. Well-levigated clay. No visible inclusions. Some mica flakes. Smokeblackened. Tablet. Poorly levigated. Some large inclusions visible. Mica flakes. Tablet. Well-levigated clay. Smoke-blackened and semi-fired from secondary burning. Tablet. Well-levigated clay. No visible inclusions. Mica flakes. Tablet. Medium well-levigated clay. Some inclusions visible. Surfaces fairly rough. Tablet blank. Well-levigated clay. Mica flakes. Human figurine. Probably female. Well-levigated clay. Some visible inclusions. Applique hairdo. Punched eyes and mouth. Pinched nose. Arms broken. Applique navel originally held in place with a stick. Scraped smooth in one area on back. Clay stopper for a wide-mouthed jar. Medium well-levigated clay. Visible inclusions. Mark of jar rim on underside. Tablet fragment. Well-levigated clay. No visible inclusions. Rather crudely moulded. Clay bulla. Poorly levigated clay. Large grit inclusions and mica flakes. Stone cylinder seal. Black stone. Remains of copper or bronze pin in centre hole. Stone cylinder seal. Grey-green stone. Unusually wide centre hole.

3:2 3:2a 3:3 3:4 3:5 4: 4:2 4:3 4:4 4:5 4:6 5: I
5:2

Gd. 73-214 Gd. 73-230 BOl 83 Gd. 73-314 Gd. 73-117 Gd. 73-54 Gd. 73 295 Gd. 73-290 Gd. 73-320 Gd. 73-64 Gd. 73-286 Gd. 73-292 Gd. 73-285 Gd. 73-300

5:3

5:4 5:5 5:6 5:7 5:8

Gd. 73-326 Gd. 73-153 Gd. 73-329 Gd. 73-260 Gd. 73-21o

3A

SOME ELABORATELY DECORATED BRONZE QUIVER PLAQUES

MADE IN LURISTAN, c. 750-650 B.C.

By P. R. S. Moorey
No single " Luristan Bronze " has been the subject of so many diverse interpretations as the tall, narrow rectangular strip of sheet-bronze now in the Metropolitan Museum, New York, decorated in horizontal panels with chased and repousse' scenes of animals, men and demons (P1. I).1 In the past this said originally to come from a grave, has been variously dated between the twelfth and quiver plaque, seventh centuries B.c. In interpreting its complex iconography scholars have usually taken for granted that it illustrated the theology either of Mesopotamians, sometimes specifically Kassites (Godard; Lancaster), pre-ZoroastrianIranians (Ghirshman; Dussaud) or Indo-Aryans (Dumizil). The apparent confusion, particularlyconsideringthe narrow limits of the available evidence, could hardly be greater. The group of decorated quiver plaques to which this one belongs provide a microcosm of the iconography used by the Luristan bronze-smithin Iron Age II-IIIB. In the absence of any written records which bear on the question the following paper seeks to define as narrowly as possible the limits within which interpretation is feasible. Such a quest must take into account both the source and date, not of the Metropolitan Museum quiver alone but of all similar pieces available for study, the historical situation in Luristan at the time of their production, the repertory of motifs employed (its range and ancestry) and only then seek, if possible, to elucidate this iconography in religious terms. Even the most cursory examination of such work in related fields, where some literary evidence may even survive,2 indicates that neither a single simple solution, nor any specific identification of deities or myths, is probable. The aim is not so much to banish speculation entirely, for the very nature of the evidence precludes that; but to reduce its role as far as is presently possible by a rigorous considerationof what actually constitutes the available body of evidence. The Metropolitan Museum'squiver plaque is the largest of a small group, which form but one aspect of Luristan'ssheet bronze industryin Iron Age II-IIIB. The integrity of the larger repertoryof products to which they belong is establishedby a common iconography, style and technique and by close association in the shrine at Dum Surkh in the only controlled excavation where many, often fragmentary, examples were found.3 The whole range embraces decorated sheetmetal vessels, belt and body armour fragments, as well as hammered disk-shaped pinheads and pendants. The last level of use in the Dum Surkh shrine in the Kuh-i Dasht, excavated by the Holmes Luristan Expedition in 1938, yielded a considerable number of these decorated pieces. Van Loon dates this level to the seventh century.4 How late into this century the shrine was in use remains an open question. To judge by the faience objects of Elamite manufacture found with the metalwork in question in this upper level the majority of it was made in the last decade or so of the eighth, or first quarter of the seventh, century B.C.5 The
1 The main references are R. Dussaud, Syria XXVI (1949), pp. 213-7; G. Dum6zil, RevueHittite et AsianiqueXI (1950), 5 pp. 18-37; C. Lancaster, Archaeology (1952), PP- 98-9; P. 8 Ackerman, Archaeology (1955), pp. 26-30; R. Ghirshman, Persiafrom the Origins to Alexanderthe Great (London, 1964), pp. 70-I (hereafter Persia); more generally A. Godard, L'Art de l'Iran (Paris, 1962), pp. 47ff.; for an extremely speculative interpretation outside these categories see P. Ackerman in V. Ferm (Ed.), ForgottenReligions (New York, 1950), P. 15. This paper is based on a lecture given to members of the Iran Centre, Columbia University in New York, on November 12, 1973. I am most grateful for their hospitality and to Professor Edith Porada for her comments; I am also indebted to Professor J. A. Brinkman, Director of the Oriental Institute, Chicago University, and Professor M. van Loon of Amsterdam University, for permission to study the archives of the Holmes Luristan Expedition in the Oriental Institute. 2 For example B. Goff, Symbolsof PrehistoricMesopotamia(New Haven, 1963). For a classic statement of the problems see E. Panofsky, Meaning in the VisualArts (Anchor Books, I955), P. 353 E. Schmidt, Bulletin of the American Institute Iranian Art and for V Archaeology (3) (1938), pp. 205-16. 4 Bib. Or. 24 (1967), p. 24; for modifications see Bib. Or. 29 (1972), p. 69, n. 22. 5 Compare P. Amiet, Syria XLIV (1967), pp. 27-46.

19

20

JOURNAL

OF PERSIAN

STUDIES

quantity of this decorated metalwork in public and private collections has grown steadily since 1938, first through well authenticated clandestine excavations at Dum Surkh itself and then from other illicitly explored sites of unknown location, supposedly in the same general area of western Luristan.6 The historical situation throughout Luristan between about 720 to 650 B.C. slowly becomes clearer. North-west Luristan then formed the ancient non-Iranian state of Ellipi, a buffer region between the spheres of Assyrianand Elamite influence, mentioned in Assyrianrecordsfrom the reign of Ashurnasirpal to that of Esarhaddon.7 Westwardsit ran down to the edge of the Mesopotamian plain, probably embracing many of the important cemeteries recently explored in the Ilam region by ProfessorVanden Berghe,"southwardsto Elamite Luristan (Rumishgan and Kuh-i Dasht). Its eastern limits are obscure. From sometime about the middle of the eighth century it was increasingly in contact with the Medes who had by then establishedthemselvesalong the line of the Great Khorasanroad in the easternZagros. Indeed when last mentioned under EsarhaddonEllipi was on the brink of an alliance with the Medes and their Cimmerio-Scythian confederates.9 It may then be assumed, in the light of subsequent historical developments, that when Assyria first crushed Elam and then retreated from the southwestern Zagros crippled by her own troubles in the third quarter of the seventh century B.C., the Median alliance overwhelmed Ellipi and with it all of Luristan. Now for the first time pottery taken to be typical of the Iranian-speakingintruderswas widely distributed between the Mahi Dasht and the northern bordersof Khuzistan.xo One point, more vital than any other for the present discussion,then becomes clear. The artefacts in question here were made in an area not yet under the control of Iranian-speakingpeoples at the time of their production. Indeed it is likely that it was the arrival of these intruders which saw the end of such local shrines as that at Dum Surkh. Before considering the known local Luristan quiver plaques in detail, the wider Near Eastern tradition of which they form a remote part may be rapidly reviewed. Both Hrouda and Madhloom" have provided concise surveys of the evidence on Assyrian palace reliefs for the forms and decoration of quivers. In so far as this evidence goes it was only in the reign of AshurnasirpalII (c. 883-59 B.c.) that animals and mythological demons were used to decorate the covers of royal chariot and personal quivers. Rosettes and geometric patterns alone appear to have been used in the eighth and seventh centuries B.C. as on actual examples from excavations at Uruk in Iraq12and War Kabud in western Luristan.13No elaborately decorated metal quiver plaques have yet been reported from excavations in Assyria, though plain ones have. A marked contrast is presented by the archaeological evidence from Urartu, where excavations have yielded a whole range of military equipment, metal shields, helmets, belts and quiver plaques, decorated with horizontal friezes. Many of these objects bear royal inscriptions which allow for a reliable chronology.14 The earliest richly decorated quivers, normally 65 to 70 cms. long, belong to the first half of the eighth century B.c. Piotrovskii'5has given details of their various designs. The majority at Karmir Blur have eight bands of decoration filled with friezes of horsemen and war-chariots; but other fragments indicate a wider repertory of motifs, akin to those also used on shields and helmets, in which animals, real and imaginary, as well as ritual scenes appear. Plainer metal quivers are known from a number of sites, notably those from Kayalidere,e1that show clearly how the exposed front only of the quiver was normally covered with metal. These may be taken as the form regularly issued to Urartian troops. It is broadly to this tradition that the armour from the destruction debris of level IV at Hasanlu, c. 8oo00 possibly an Urartian sack, belongs.17 B.c.,
A corpus is clearly desirable, but must wait on Professor van Loon's publication of the Dum Surkh material, and of the two finest collections, those in the Louvre (Coiffard) and the Mus6e Cinquantenaire, Brussels (Graeffe) only partly available in published illustrations. 7 L. D. Levine, Iran XII (i974), pp. xo4-6; T. Cuyler Young, Jr., Iran V (1967), pp. 12-14. 8 See reference under note 13 and Archeologia24 (1968), pp. 53-63; 32 (1970), pp. 65-73; 36 (1970), 10-21; 43 (197i), pp. 14-23; 57 (1973), PP- 49-58; 63 (I973), pp. 24-36; 65 pp. 16-29. 9 For(i973), a possible oblique reference under Ashurbanipal see M. Streck, Assurbanipal (Leipzig, 1916), pp. clxxxii, cccxlv. I 10 T. Cuyler Young, Jr., Iran V (1967), p. 27; C. Goff, Iran VI
6

(1968), p. 128. des Assyrischen Flachbildes n B. Hrouda, Die Kulturgeschichte of (Bonn, 1965), p. 85, pl. 2I; T. Madhloom, The Chronology Art (London, I970), pp. 49-52Neo-Assyrian 11 H. J. Lenzen, Uruk ... WVDOG XXIV (1968), pl. i5a. ... van Luristan ... I, L. 13s Vanden Berghe, Het Archeologisch Kalwali en War Kabud (Brussels, 1964), p. o09, fig. 21.6, pl. 29c. 1 G. Azarpay, Urartian Art and Artifacts (California, 1968); M. van Loon, UrartianArt (Istanbul, 1966). Urartu (London, 1967), p. 47; KarmirBlur (Leningrad, 1970), 16 colour plate. 16 C. Burney, AS XVI (1966), pl. XVIIIc, fig. x8.6-6B. 17For a quiver see Bastan-i-shenasi (Tehran, 1951), p. 72. I

SOME ELABORATELY

DECORATED

BRONZE

QUIVER

PLAQUES

21

A western iconographic source is also evident on the best known group of quiver plaques from Luristan, recently discussed by Calmeyer.18 According to Madame Maleki's original reportl9 a considerable number decorated in derivative Assyro-Babylonianstyles were found in a cemetery illicitly excavated about 1956 near the village of Gilan in north-west Luristan. They were said to have been associated in the graves with arrowheads,shields, iron helmets and swords. Calmeyer defined two groups, one in a reasonably well executed style closest to Babylonian art, the other more ineptly worked that owes more to Assyrian sources. Both types may be broadly dated to the eighth century B.c. The source, at the western end of the Great Khorasan Road through the Zagros, is exactly where such a fusion of influences might be anticipated at this time. Against this backgroundthe distinctive decoration of the quiver plaques reviewed here becomes the more obvious. There are no direct links of style or decoration with any of those already discussed. If anything slightly later than the " Gilan Group ", the "Dum Surkh " group belongs to an entirely different milieu. Apart from a fragment in Tehran20 that may well come from the Dum Surkh shrine, the others have no context; their decorationsuggestsobjects for deposit in shrinesor graves rather than routine military equipment. The seven examples known to me, all illustrated in publications, will be consideredhere one by one, beforean attempt to elucidate the complex sourcesof their decoration. As illustrationsof all but number 6 are included, I have for economy's sake dispensedwith conventional catalogue descriptions. New Collection,21 York, I. The plaque in the Heeramaneck far the most finely executed, has four surviving (fig. I) by registers each occupied by a symmetrical group. The flanking animals, lions and caprids, are in no sense exceptional; only the full-face rendering of the lion-heads in the lowest panel provides a trait distinctive of this sheetmetal work in Luristan.22 The lowest central figure, entirely human, with long robe and prominent forelockhas Neo-Elamite relatives.23 The other figures introduce a sub-human theme. The bullmen of panel 3, two in profile, one full-face, are also in the Elamite tradition, as Miss Porada has pointed out,24and may be associateddirectly with scenes on decorated faience boxes from Susa, some of which reached the Dum Surkh shrine in A fragment in the Louvre, the early seventh century B.C.I2
which may well be from one of these quivers, illustrates a

Fig. I. Plaque in the HeeramaneckCollection, New York; after Mostra d'Arte Iranica (Rome, 1956), pl. XIIL.

18P. Calmeyer, DatierbareBronzen aus Luristan and Kirmanshah 21 Mostra d'Arte Iranica (Rome, 1956), no. 48, pl. XIII: 43 cms. long; E. Porada, Ancient Iran (London, 1965), pl. 15, rt. (Berlin, 1969), pp. 81-7; it should be noted that there are -colour: two registers only. also a number on which the decoration has been added 22 Compare R. Dussaud, Syria XXVI (1949), fig. 6; Barbier recently. 19 Iranica Collection(fully illustrated sale catalogue, Hotel Drouet, Paris, Antiqua IV (1964), PP- 2-3. 20 May, 1970), no. 32, pl. on p. 53. (Hereafter Barbier). Drawing in the Holmes Luristan Expedition Archive, Oriental Institute, Chicago (No. 1859); possibly part of the quiver, 28 P. Amiet, Elam (Paris, 1966), fig. 407. originally in the David Weill Collection, now in the Louvre, 24AncientIran, pp. 70-I. 25 Compare P. Amiet, Elam, pp. 498-501. P1. IIa here.

22

JOURNAL

OF PERSIAN

STUDIES

similargrouping (P1.IVd).26 The figure in the upper panel, human in all save his bull's ears and horns, with curling sidelocks, belongs in the same tradition; but when he is shown, as in panel 2, with lionheads on each shoulder a fresh theme is evident that plays an enigmatic role among bronzes, cast and sheet, produced in Luristan in Iron Age II-IIIB. This triple-headed motif requiresfuller elucidation. The central head, essentially human, may be flanked either by animals, birds or secondary human heads in profile, which often seem actually to spring from the sides of the central head. They appear on cast standard-finials2' and, with greater elaboration of detail, on sheetmetal pinheads and

Fig. 2. Bronze umbo in a private collection in Basle; after R. Ghirshman, Persia, fig. go.
26

AO 20.62i-P. Amiet, Revuedu Louvre(1963), p. I8, fig. s2; A. Godard, L'Art de l'Iran, fig. 38; R. Ghirshman, Bichdpour II (Paris, I956), fig. I3.

P. R. S. Moorey, Catalogueof the AncientPersian Bronzes in the Ashmolean Museum, Oxford, (1971), nos. 182-3 (hereafter Ashmolean).

SOME ELABORATELY

DECORATED

BRONZE

QUIVER

PLAQUES

23

in bronzedisc-shaped Persia, fig. 63. Fig. 3. Part of a fragmentary pinheadin a privatecollection New York; afterR. Ghirshman,

umbos28(figs. 2-4; P1.IIIb). The same device appearson the rectangularsilver fragmentin Cincinnati, sometimes regarded with suspicion though it fits perfectly well, technically and iconographically, into this group of material (fig. 4).29 The ancestry of this motif has been traced into Mesopotamia, but all the parallels there are much earlier and never very close.30 Closest is a creaturecut on a cylinder seal of unknown origin, formerly in the de Clercq collection now in the Louvre. Stylistically it is transitional

Persia, fig. 64. Fig. 4. Central motifon thesilverstrip in Cincinnati; afterR. Ghirshman,
28Persia, figs. 63, 90; Barbier, no. 32, pl. on p. 53 (all human); Godard, L'Art de l'Iran, figs. 43, 45 (animal); eagles or vultures appear on an unpublished piece in the Heeramaneck Collection; also Dum Surkh 1722. 29 Often illustrated, see Persia, fig. 64; for a cautionary note, C. Nylander, Ioniansat Pasargadae(1970), p. 149. p. 0oAshmolean, 159; Godard, op. cit., p. 47,

24

JOURNAL

OF PERSIAN

STUDIES

from " Elaborate Mitannian" to " Middle Assyrian" glyptic and might come from a workshop in Iran rather than Iraq (Fig. 6).31 2. This clue leads naturally to another Luristan quiver, until recently in the BarbierCollection (Fig. 5),32 where again resource must be made to " Mitannian/Middle Assyrian" glyptic for the best source of its primary motifs. Two panels here are occupied by a rampant double-headedleonine monster, with interlaced body and the feet of a rapacious bird, holding inverted goats or gazelles. Such a monster, with single body, appears on the seal of Eriba-Adad I at Assur (c. 139266 B.c.).33 Virtually contemporary are representations of an allied creature, but with single head and body, on Middle Elamite cylinder seals from Tchoga-Zanbil in Iran.34 Exactly the monster of the quiver is shown with the same victims on an electrum goblet in the Louvre (P1.IVc),35 which may be dated by analogy with the fine gold goblets from the Marlik cemetery, where Middle Assyrian influence is very clear, to Iron Age I Though rare, relatives of the bicephalous (c. I300-II00 B.C.). leonine figure appear on the cast bronze pinheads from Dum Surkh and on disk-headedpins in Iron Age IIIB.36 The central panel's figure, holding braces of birds, is not so easily placed. The treatment of head, hair and beard, as also the tasselled skirt, is at home on Middle Assyrian seals."' A fine Middle Elamite cylinder seal illustrating a ritual meal shows an attendant holding a tethered bird.38 The frieze of vultures passant appears on other Luristan sheetmetal whereas the lowest scene on this quiver, of vultures feasting on a caprid or gazelle, appears both in Kassite art and on a Marlik goblet.39 3. To move from this quiver to the well-known Metropolitan Museum example I) revealsfreshmotifs, but not real surprises. (P1. Here, in style and choice of themes, there is a much closer affinity
P. Amiet, Bas-reliefs imaginairesde l'ancien Orient d'apris les cachets et les sceaux-cylindres (Louvre, 1973), no. 424; cover design. 32 Barbier, no. 25, plate on p. 48; W. Culican, The Medes and Persians (London, 1965), fig. 6: 7000 Jahre Kunst (Zfirich, 1962), no. 731, pl. 21b. 33 T. Beran, ZA 52 (i957), fig. 2; see also the twisted stance in Mitannian glyptic, H. Frankfort, Cylinder Seals (London, 1939), figs. 87, 90o; E. Porada, AASOR XXIV (1947), no. 720o. 34 E. Porada, MDP XLII: TchogaZanbil IV, pp. 43-585 Louvre: AO 20281: A. Parrot, Syria XXXV (1958), pl. XV figs. 45, PP. 175ff.; E. Porada, AncientIran, pl. 22b. 36 Dum Surkh I465; Ashmolean, no. 353; cf. such western precursors as G. Loud, The Megiddo Ivories (Chicago, 1939), pl. xIf (Hittite); A. Moortgat, Tell Halaf III (Berlin, 1955), pl. 93A. 7'A. Moortgat, ZA 47 (I942), fig. 57; T. Beran, ZA 52 (1957), figs. 25-7. 38 E. Porada, Expedition13 (3-4) (1971), P. 33, fig. 9; also her remarks in Acts (Nicosia, 1973), p. 267. Archaeological Symposium of the International Reporton Marlik 39 A. Godard, op. cit., fig. 85; E.O. Negahban, A Preliminary Excavation(Tehran, 1964), pl. IV; E. Porada, Memorial Volume the VthInterof nationalCongress IranianArt and Archaeology (Tehran, 1972), pp. 163--82. of Fig. 5. Plaque formerly in the Barbier Collection; after Sale Catalogue, Hotel Drouot, Paris, May, 197o, no. 25, plate on p. 48. de 31 M. de Clercq, Collection Clercq(Paris, 1888), I, no. 357, pl. XXXIII:

Pl. Ila. Decorated quiverplaque in the Louvre. (Photo: Louvre)

Pl. IIb. Decorated quiverplaque in the Musde Cinquantenaire, Brussels. (Photo: Musee Cinquantenaire)

PI. L Decorated Museum,New York. quiver plaquein the Metropolitan Museum.) (Photo: Metropolitan
2B

Pl. IIIa. Disc-shaped sheet bronze pinhead in the Rietberg Museum, Zurich (von der HeydtCollection). (Photo: Rietberg Museum)

P1. IIIb. Disc-shapedsheet bronze pinhead in the UniversityMuseum, Philadelphia. (Photo: University Museum)

Pl. IVa and b. Copper silvergoblet and Iran (Louvre AO. 21625). (Photo: Louvre) from north-west

P1. IVc. Electrum gobletfrom north-west Iran (Louvre AO 20281) (Photo: Louvre)

P1. IVd. Fragment from a decorated quiver plaque (Louvre AO 2o621). (Photo: Louvre)

SOME ELABORATELY

DECORATED

BRONZE

QUIVER

PLAQUES

25

Seal: Louvre AO 22350; Fig. 6. Cylinder

afterP. Amiet,Bas-Reliefs Imaginaires de l'Ancien Orient (Paris, r973), no. 424, covermotif (enlarged).

to the most elaboratelydecoratedof the disk-headedpins. There is also a common horror vacui.The upper and lower registersof winged rampant bulls flanking a central tree are but one variant of a predominant motif among the cast standard-finials. It may also occupy the whole field of a disk pinhead or be constantly repeated as on a remarkable example in the Rietberg Museum, Ziirich (P1. IIIa).4? This pinhead also bears the lion and gazelle motif of the Metropolitan Museum quiver. The narrow band of gazelles passantdraws in another group of sheetmetal objects produced in Luristan in Iron Age IIIB: tall goblets with concave sides decorated in horizontal registerswith animal friezes.41Here also belong two similarly decorated narrow-necked bronze jars,42 rather awkwardly made in two pieces, for the form is not one naturally suited to reproduction in metal, that match the form and decoration of contemporary polychrome glazed pottery jars.43 In the third quiver panel the triple-headed motif reappears. One of the branches held by the figures here terminates in the flattened linear lion's head that was used by Luristan smiths for pinheads and bracelet terminals.44 The master-of-animalsin
40 H. A. Potratz, AFO XV (1945), fig. 3. 41 For example SurveyIV, P1. 68B. 42 O. Muscarella, MetropolitanMuseum Journal 5 (1972), pp. 25-50; and a similar jar in the Heeramaneck Collection; cf. also some Middle Elamite seals: MDP XLII: Zanbil IV, group VIII. ,* For example, Persia, fig. 398. nos. 319-23, 382. "4Ashmolean, Tchoga

26

JOURNAL

OF PERSIAN

STUDIES

profile dominated by lions in panel 4 is unusual only in the marked difference of proportions. On the parallel range of cast standardfinials the anthropomorphic central figure is never so obviously dominated by the animals. This profile figure and those in panel 5 with their rounded hair-styles,fluid featuresand straggly beardsare again familiar participants in scenes on disk-headed pins.45 The central figure of panel 5 has at the centre of his skirt a detached face which has stimulated considerablespeculation. When set in context there seems little doubt that here, as for instance on the Cincinnati silver plaque and a number of pinheads (Fig. 4, P1. IIIb),46 it represents a metal umbo in the form of a face sewn onto the skirt. Such detached face umbos, directly related to those often set at the centre of disk-headed pins, are well known,47and examples were found in the Dum Surkh Shrine. Constantly in Luristan part of an animal may stand for the whole and the same idea applies to these faces which may, as on the Mus6e Rietberg pin (P1. IIIa), be set between rampant animals to represent the whole human figure more usually put there. Quivers 4, in Brussels(P1. IIb), 5, in the Louvre (P1. IIa) and 6, in Tehran,48 equally close to the decoration of the pinheads in are the choice of themes and in the manner of rendering them. Here also appear the floral borders so often used by the makers of disk headed pins either as primary motifs of decoration or as frames for other main scenes.49 Yet again, as Amiet noted in first publishing it, the Louvre quiver draws upon a repertory of motifs shared with Elam. The large Tehran example no. 6 (64 cm. in length) still retains some of the rings in its perimeter holes that secured it to the quiver, and the large chains by which it was fitted to a chariot or human bearer. The style and manner of executing the floral borders, two sets of panels with opposed recumbent winged bulls, two panels with a central tree flanked respectively by winged semi-human genii and rampant winged bulls, and its central panel framed within a large floral ornament, associate this example particularlyclosely with no. 5. It is the last of the published quiver plaques, no. 7 now in Tehran (Fig. 7),50which opens some new avenues for exploration; but only in the upper and lower panels. The central two again show the

Fig. 7 left. Plaque in the Iran Bastan Museum, Tehran; after Sept Mille Ans d'Art en Iran (Paris, 9g6z), no. 227, pl. XXX

"s Godard, op. cit., Tehran 1581-16o6: Zahr! Nabil, Memorial Volume of the pl. 21, figs. 56, 59. Vth International 46Persia, fig. 64. I, Congress Iranian Art & Archaeology p. 29, of 47 Godard, op. cit., figs. 68-74. plate (in Persian). 48Brussels: A. and Y. Godard, Bronzes du Luristan: Collection 11 Cf. Ashmolean, nos. 357-36o. E. Graeffe(Hague, undated), no. 311, pl. 19; G. Goossens, so Sept Mille Ans (Paris, 1961), no. 227, pl. XXX; Persia, fig. Bronze uit Loeristan (Brussels, 1956), pl. 14; E. Phillips in Vivanter (1968), p. 114, fig. 132; E. Phillips 423; Archeologie Vanished Civilizations(London, 1963), p. 227, fig. 18; Vanden in Vanished Civilizations(London 1963), p. 227, fig. I9; G. G. Kunst (Gent, 1966), no. 363; 29.5 cm. Berghe, Oude-Iraanse Belloni, Iranian Art (London, x969), pl. 26; P. Calmeyer, Stil (Munich, 1973), fig. 143; high; 13.0 cm. wide. Louvre AO 25011: P. Amiet, Revuedu Reliefbronzenin babylonischem Louvre(1972), p. 433, fig. 17; 43-9 cm. high; 16.8 cm. wide; 6o.o cm. high; 14.o0 cm. wide.

SOME ELABORATELY

DECORATED

BRONZE

QUIVER

PLAQUES

27

constantly repeated, if pictorially varied, theme of the lion as predator, upon man and beast. In the upper panel a light chariot with six-spoked wheel, drawn by a bovid not an equid, carries a bowman hunting lions, to judge from the creature set in the upper field. This chariot is like those in the upper registerof the Hasanlu gold bowl and on a goblet in the Louvre (P1.IVa and b)51in contrast to the war chariots of Neo-Assyrian type shown on the silver goblet and ivories from Level IV at Hasanlu.52 The cast cheek-pieces for horsebits from Luristan include a few rendered as a very stylized light chariot, but with four-spokedwheel, and a bowman almost standing on the yoke-pole to take aim.53 There the draught animal is a monster more feline than equid. On this quiver, as on the cheekpieces and the Hasanlu bowl, this is a world of myth, not reality. A seated male or female figure, set above a creature more like a leopard than a lion to judge from its body markings, in the lowest panel has no close parallel elsewhere. A disk-headed pin in Brussels shows a female figure, possibly an Elamite deity, holding snakes. She is seated directly on a leopard's back;54 on the Hasanlu gold bowl the " mountaingod" is set on a lion's back.55 Du Mesnil du Buisson's recent detailed interpretation of this quiver, taken in isolation, in terms of Canaanite religious imagery, is based on extremely general iconographic comparisonsunchecked by any archaeological or historical controls.56He offers no reasons why at the time of its production, or at any other time, the inhabitants of Luristan should use a religious iconography, which even in its homeland is still a source of much lively discussion.57 If it may be accepted that these quiversform a homogenousgroup, and that their relationto the whole range of decorated metalwork of the Dum Surkh type establishestheir time and approximate place of manufacture,then some attempt may be made to define the major strainsin the iconographic repertory upon which their makersdrew. The concise investigationofferedhere broadlyindicatesa blend of various strains, some long established,some more recent, which may be crudely classifiedin the following way: I. The Prehistoric legacy: a foundation of imagery native to the region since the prehistoric period that describes the world of the hunter in certain distinctive concepts and images, notably the naturedemon, half-man, half-caprid, mouflon or lion. It well espressesthe intimate, in the case of the lion mutually predatory, relationshipof man and wild fauna in their mountain home. So far as it is possible to judge domestic animals are conspicuously absent. Glyptic evidence is steadily growing for the " master-of-animals" and a close relative, either a man wearing an animal mask and skin or a demon, part man, part beast, in Luristan in the fourth millennium B.C.58 Thereafter it is, in so far as present evidence goes, generally submerged until exceptional circumstances at the very end of the second millennium B.C. allowed for the emergence of a local industry producing bronzes illustrating much of this very ancient repertory of motifs. By then it has been modified by contact with at least one other tradition: " The "Hurrian or Mitannian legacy: in analysing the motifs of the Luristan quiver plaques recourse has constantly been made here first to the imagery of the " Mitannian-Middle Assyrian " glyptic styles, and secondly to the iconography of a series of decorated vessels produced in north-west Iran in Iron Age I under strong western influence. Miss Porada implicitly59 and both Miss Mellink and Amiet60 explicitly have defined the Hurrian character of the stories, and the pictorial means for expressing them, used on the Hasanlu gold bowl and such relatives as the copper and silver goblet in the Louvre (P1. IVa and b). Other vessels from Marlik, on which the iconography has much less
2. narrative, have also been recognized as showing Middle Assyrian traits.6e Any attempt to define further the significance of this relationship in terms of Luristan's history runs the danger of categorizing what is, for the moment, no more than a hypothesis cautiously advanced.
51 Persia, fig. 31; P. Amiet, Syria XLII (1965), pl. XVI-XVII.

52E. Porada, AncientIran, pl. 28; 0. Muscarella, Bulletin of the Museumof Art (November 1966), p. 128, fig. I I. Metropolitan "5Persia, fig. 73. "4Persia, fig. 97. 55Persia, fig. 31. 56NVouvelles Etudes sur les Dieux et les Mythes de Canaan (Leiden, 1973), pp. 21Iff.; pl. XVIII. 5 For example, U. Oldenburg, The ConflictbetweenEl and Baal

in Canaanite Religion (Leiden, 1969). 68 E. Porada, AncientIran, pp. 32-3; R. D. Barnett, Syria XLIII (1966), pp. 259-76; P. Amiet, Revuedu Louvre(1973), pp. 21524. , ExpeditionI (1959), pp. 19-22; AncientIran, pp. 96-102. Iranica Antiqua 6 (1966), pp. 72-87; Syria XLII (1965), 60 pp. 235-51. 61 E. Porada, Ancient Iran, p. 91; E. O. Negahban, JNES XXIV (1965), PP. 309-27.

28

JOURNAL

OF PERSIAN

STUDIES

Did Luristan share this common legacy from the mid-second millennium or was this imagery only introduced there later from Iron I sources in the north? A Hurrian presence in Elam in the earlier second millennium B.C.may be identified from a variety of evidence, though this is too sparse at the moment to reveal their effect on Elamite art or religion;62 the more so northwards into Luristan. The cultural significance of clear affinities between the painted pottery tradition of western Iran and ceramics from Syrian sites, where a Hurrian presence is well established, depends once again on further documentation, as indeed does the possible Hurrian role in the diffusion of metal types postulated by Deshayes.63 Although a Hurrian homeland eastwards of the Tigris is generally accepted, it is more often located in the region of Lake Van or north-westernIran than further south in the Zagros.64 More striking, in the light of Dumezil's earlier argument associating the Metropolitan Museum quiver with Vedic theology, is the extent to which Luristan smiths in Iron Age IIIB may have been drawing on an iconographic legacy of Mitannian origin, albeit modified in the processof transmission. Dumezil's argument in this case rested on no more than an assumption that he had found in this quiver clear illustration of his well-known tripartite division of Indo-European society and religion. There is no direct evidence whatsoever for his very specific identification of Vedic deities in these scenes.65 But, allowing for the purely speculative nature of Dumezil's identifications, did he perhaps direct attention to a very significant cultural substratumin Iron Age Luristan not otherwise evident? It is now widely recognized that the Aryan linguistic traces in the Near East in the third quarter of the second millennium B.c., outstandingly the names of such Vedic gods as Mitra, Varuna, Indra and the Nasatyas in the fourteenth century Hittite-Mitannian treaty of Shuppiluliumashand Kurtiwaza,66 are to be connected with Indo-Aryan, not Iranian. So long as the archives of the Mitannian capital remain unknown the basic evidence is certainly meagre, but this fundamental conclusion has yet to be cogently challenged.67 The small minority of Proto-Indo-Aryans who established themselves, c. 1550 B.c., as a ruling aristocracyamong the Hurrians in the steppe lands of north Mesopotamia came out of Iran, where it is currently assumed they formed the western branch of intruders from central Asia."6 In Iran they were particularlyassociatedwith the central northernpart of the country (Mazanderan) and may have penetrated westwardsby the route from the Urmia basin to the Upper Zab and beyond. Their Near Eastern ascendancy was relatively brief, virtually over by the mid-fourteenth century B.C. By the early 13th century they were Assyrian vassals, but what happened thereafter is unknown.69 They had established a link between northern Mesopotamia and north-westernIran that endured, perhaps not merely for commercial reasons, and has a vivid archaeological reflection in the luxury goods of gold and glass at Marlik, some of them dating from not long after the eclipse of the Mitannian supremacy in the west. Might there then have been a proto-Indo-Aryan element in the population of Luristan by at least the later second millennium B.C. ? It 3. The Elamitecontribution. is particularly difficult with the material to hand to decide exactly how the undoubtedly close relationship between the lowland urban civilization of Elam and their northern neighboursin the mountains of Luristanaffected their respectivecultures. It would be unwise to assume that the conventionally more primitive tribes of Luristanwere necessarilyalways the debtors, as there is enough in Elamite art to suggest that they shared from prehistoric times, if they had not adopted more recently, primary motifs and symbols. The mountain fauna, and the nature-demons, of Elamite art might be supposed to spring most naturally from the life and aspirations of their mountain neighbours. Recent excavations at Tchoga Gavaneh (Shahabad), 70 km. west of Kermanshah, have revealed, by way of inscribed tablets and seals, clear evidence for a strong "Middle
62

R. Labat, Elam, c. i6oo-z20oo B.C. (Revised CAH, 1963), PP- 4-513 J. Deshayes, Les Outils de Bronze de l'Indus au Danube I (Paris, 1960), pp. 426ff. For the pottery see T. Cuyler Young, Jr., Expedition13 (1971) p. 25; C. L. Goff, Iran IX (1971), p. 50. and Subarians(1944); recently 64 Basic study: I. J. Gelb, Hurrians AncientHistory II (1), pp. 417ff. M. Drower, Cambridge 65 For Dumezil's general approach see the comments in R. Frye, The Heritageof Persia (London, 1962), pp. 2o-r.

66

H. Thieme, Journal of the AmericanOrientalSociety60o (1960), pp. 301-17. 67 M. (Wiesbaden, Mayrhofer, Die Indo-Arierim alten Vorderasien 1966); challenged by A. Kammenhuber, Die Arier im vorderen Orient (Heidelberg, 1968); T. Burrow, JRAS (1973), PP. 123-40. 68 I. M. Diakonov, IstoriyaMidii (Moscow, 1956), pp. " E. Weidner, Ugaritica VI (Paris, 1969), pp. 519-31. I24-5-

SOME ELABORATELY

DECORATED

BRONZE

QUIVER

PLAQUES

29

Elamite " presence there.0o By Iron Age IIIB in southern Luristan both historical circumstancesand imported objects at Dum Surkh certainly demonstratean Elamite presence, not so obvious earlierwhen Elam was in eclipse following the Babylonian campaigns of the later twelfth century B.c. With the rich range of illustrative material for the religious imagery of Luristan in Iron II to IIIB recoveredfrom excavations in the Dum Surkh shrine, and other unknown sites in Luristan, the temptation to explain and interpret it in precise and specific terms is strong; so strong indeed that selfdeception is a constant danger. When there are no checks or controls it is all too easy to be persuaded that more has been discovered than is actually discoverable in the present state of knowledge. This paper is predominantly negative in tone. It has attempted to underline the dangers of overconfidence in interpretation by seeking to eliminate certain possibilitieson historical grounds, whilst pointing out the broad range which might still apply in the light of present evidence. Where reliable sourcesfor the interpretation of ancient Near Eastern religious iconography exist they always reveal the extreme fluidity of the underlying religious attitudes and the protean character of deities. Acknowledgments I am most grateful to Dr. P. Amiet, Dr. V. Crawford, Professor R. Dyson Jr., and Professor Dr. L. Vanden Berghe for generously providing photographs, and to them, and to the Director of the Rietberg Museum, Zurich, for permissionto publish objects in their charge; to Mrs. N. Heeramaneck for kindly showing me her collection; and to Mrs. Pat Clarke for the drawings.

70

M. Kordevani, Bastan Chenasiva Honar-1Iran 7-8 (irg7I), pp. 30-5; P. Amiet, RA LXVII ('973), pp. I57ff.

ISLAM, CHINESE PORCELAIN AND ARDABIL By MargaretMedley

In 1972 at the third Percival David Foundation Colloquy on Asian Art and Archaeology, which was I to concerned with The Westward art influence theChinese from the fourteenth theeighteenth of century,1 read a paper that attempted to explain some of the peculiarities of fourteenth century Chinese blue and white porcelain. It was suggested that since the Chinese developed blue and white almost exclusively for the benefit of the Near Eastern market, the designs, far from being Chinese in conception were in fact Islamic, and consequently in terms of the treatment of surface there was a clear dependence upon Islamic traditions. In the year that has passed since then, the opportunitiesof examining this proposition have been many, and the results have tended to support it in a number of ways. For this reason, the following short paper examines the thesis especially in relation to the Chinese material in the Ardabil Collection, a large part of which is now available for study in the Irdn Bastan Museum in
Tehran.2

A fundamental revolution occurred in the history of Chinese porcelain in the fourteenth century under the domination of the Mongols and this revolution had an enormous impact on the Near East, and ultimately on Europe, out of all proportion to any expectation on the part of those who introduced cobalt blue to the Chinese potters as a decorative pigment. The introduction of the new pigment to the Chinese ceramic painter occurred towards the end of the first quarter of the fourteenth century, and from the moment of its appearance an entirely new series of concepts in decoration emerged with a heavy dependence upon alien sources. By the fifteenth century most of the alien influences had been absorbed and transformedto suit Chinese taste and the current of influence reversed so that Chinese concepts, satisfactorilyre-established in a new artistic climate, flowed back westwards into Islam, to Iran, Egypt and Turkey. In the present instance it is not intended to carry the study beyond the end of the fifteenth century except to show how some of the later Islamic wares were affected by this flow back from China. When the Chinese potters were presented with cobalt blue by their potential customers, the merchants of Persia and Syria (for these were the two Islamic groups apparently most interestedin Chinese production), the Chinese were faced with two problems. The first was that the Islamic market demanded vesselsof a totally differentshape and size from any which the potters were accustomedto make for the domestic market, and the second was that the decorationsto which they were accustomed were unsuited, for technical reasons, to the new market at this point in time.3 When to these was added the fact that the potters were required to decorate vessels whose dimensionswere three if not four times the size of anything normally produced, it became clear that the problems could only be overcome by an entirely new approach to decoration and its lay-out on the surface of any given vessel. At the same time, it was essential to achieve an internal harmony in the use of motifs in relation to the extent of the surface. It was this latter aspect which was to present the greatest difficulty and which initially resulted in the castigationof early blue and white by the Chinesegentleman of education with antiquariantastes; he maintained that it was " extremely vulgar ". The other problem, that of the sizes and shapes, was of a purely technical nature and relatively easy to resolve in an industry already well advanced and at
least partially mechanized; but to achieve harmonious balance in the right use of decorative motifs, some of which were unfamiliar, was a major problem in artistic invention and one not uncommonly encountered.
x "Chinese

Ceramics and Islamic Design ", Percival David Foundation on in Colloquies Art andArchaeology Asia, No. 3 (London 1972), pp. i-io. 2 Some pieces have now been transferred to Tabriz University

Museum and to the Azerbaijan Museum in Tabriz. * The painting of porcelain as opposed to stoneware had barely been attempted at this time.

31

32

JOURNAL

OF PERSIAN

STUDIES

In order fully to appreciate the character of the problem with which the Chinese potter was faced, a brief outline of the historical situation which made the Chinese receptive at this point to the alien influences will not be out of place, because political, social and economic factors all had a part to play in the unusual revolution which took place in the fourteenth as well as in the early fifteenth century. Even beforethe Mongol invasion of southern China occurredin the middle of the thirteenth century, the Southern Sung government had initiated a deliberate and carefully calculated interest, amounting sometimesto active interference,in the export trade in both silk and ceramicsfor the sake of the revenue to be drawn from it, and for the first time there was official sponsorshipof much foreign trade. The reason for this was, not unexpectedly, the need for an increased revenue in order to combat the military threat from the north, a need which became increasingly urgent following the Mongol conquest of the north in 1234, when these new invaders replaced the older Jiirchen Tartars of the Chin dynasty. Most of the East Asian trade was carried by Persians and Arabs, and at first the greatest volume of traffic passed through Canton; but after about I2o00 a shift northward began, until from the second half of the thirteenth century up to the middle of the fifteenth century, the more important port was Ch'iianchou, Marco Polo's Zayton. In the thirteenth century the Southern Sung government appointed a number of superintendents of shipping at the major ports, and the one appointed to Ch'iian-chou, a man called P'u Shou-keng, was an official of great importance.4 He was extremely wealthy and was moreover of Persian extraction; of further significance was the fact that he ran his own trading fleet down to the Malacca Straits and back. He is reported to have had a large fleet of big ocean-going merchantmen, which had good passengeraccommodation as well as ample cargo space. It is significant that he was only one of many engaged in a profitabletrade. The Muslim inhabitants of China, especially at Ch'tian-chou,were famous, if not notorious, for their wealth and their houses were lavishly equipped with furnishingsof the utmost luxury. Apparently the wealthy Persiansand Arabs lacked nothing that money could buy. The contacts between the rich foreign merchants and the hardworking Chinese craftsmenwere undoubtedly profitable to both sides, and probably became more so after the beginning of the thirteenth century with the introduction of official sponsorship. After the Mongol invasion of the south, and in the closing months of the campaign in I275, the Sung merchant fleet ruler, fighting a losing battle, made a fatal error; he tried to confiscate P'u Shou-kAng's at Ch'iian-chou, an attempt to which that affluent gentleman took immediate exception and promptly changed sides. He thereby maintained his financial security in as much as his fleet survived, and alsovery largely because he was not Chinese-he retained his post as superintendentof the shipping. The Mongols not unnaturally distrustedthe Chinese as much as the Chinese detested the Mongols, and most of the higher administrative posts were allotted to foreigners, since they could be relied upon to give their allegiance directly to the Mongols rather than to the Chinese.5 Thus the Mongols, being also interested in the revenue from the foreign trade, established conditions favourable to the intake of new ideas and materials from overseas. Moreover, since Ch'iian-chou was well situated with regard to the kilns, and porcelain had long been a major commodity in the trade with the Near East, it was logical to expect it to be subjected to strong external influences. That such a situation had barely arisen in the Sung period was largely due to the distinctive taste and regular patronage of the kilns by the court and members of the aristocracy. There was also fairly strict government control of the kilns to ensure that the court was kept adequately supplied with the most refined and decoratively restrainedporcelains of
the day. It was probably inevitable that, following the Mongol conquest, the patronage of both court and aristocracy should be withdrawn. Certainly, there were to be no more exacting demands from petulant officials for ceramics of subdued colour and restrained decoration, of the kind which had previously been required. The Mongols were not in the least interested in what was produced; all they wanted was the revenue, and so no specific demands were made on the potters, at least in the earlier years of the dynasty. Hence there was now freedom for the first time in two centuries for them to do entirely as they
KuwabaraJitsuzo, " On P'u Shou-ktng, Superintendentof the Trading Ships Office at Ch'iian-chou Memoires the ", of Research of Department theToyoBunko (1928), pp. 1-79, (1935),
pp. 1-104. I Etienne Balazs, Chinese Civilizationand Bureaucracy (New Haven and London, 1964), pp. 79-81.

ISLAM,

CHINESE

PORCELAIN

AND

ARDABIL

33

pleased and to try experimentsof all kinds. It was into this situation of, on the one hand, freedom from restraint, and on the other of positive economic pressure (since taxation was high), that the merchants from the west, from Persia and Syria especially, came with interesting, indeed, exciting demands for colourful wares suitable for their home countries-and came with cobalt blue, which was imported mainly from Kirma-nprovince in Persia. The Chinese response to this direct foreign invasion of the market was to have a world-wide impact, the fruits of which are still with us today. It was not, however, just a matter of introducing a coloured pigment for painting on porcelain; there was much more to it than that. The influences introduced included those relating to forms. And while the province of Kiangsi, where the white porcelain potteries were located, received the cobalt as well as the new forms, the celadon kilns in Chekiang, of which there were at least I50 in operation at this time, received the forms alone.6 But it was the great kilns in the north of Kiangsi at Ching-te Chen that were to be most profoundly affected. These expanded rapidly and drew off skilled labour from many kiln centres farther south until they became the best financed ones, as well as those at which the craftsmenwith the most advanced skills were to be found.7 In the first instance it was the celadon kilns in Chekiang which were affected. This is scarcely surprisingin view of the proximity of the most important port and of the affluent society which was so largely comprised of Muslims. The appearance among the celadons of the massive plate form based on a Near Easternmetal prototype antedates the same form at Ching-te Chen in the adjoining province of Kiangsi, more remote from the ports, by about half a century.8 But in the celadons, with their restraineddecoration, dependence on the sensuous character of the glaze and their subdued colouring, the foreign influence is not so immediately apparent as it is in the brilliant blue decorated white porcelains of Ching-te Chen, where the revolution in design and production was almost immediate and complete. Thus the influence from abroad strikesone with remarkableforce in the colour and richness of the new ware, in contrast to the earlier, more restrained and perhaps over-refined, tradition. The formswere more generally affected in the blue and white, unlike the celadon, but this was to be expected since the blue and white was developed in the first instance for the Islamic market. Thus in order to comprehend the full extent of Islam's impact on Chinese ceramics, it is best to give attention almost entirely to the blue and white porcelains. Prior to the fourteenth century, there had not been much painting on the ceramic wares in south China. Most decoration had been incised or carved before glazing, or else it was moulded. Carved decoration especially appealed to the Chinese educated taste, perhaps partly on account of the freedom of deliniation associated with it, but also certainly on account of the subtle colour variations that occurred when there was thickening of the glaze in the hollows. This kind of decorationwas simple and necessarily cursive. Moulded decoration was generally more complex and exhibited less restraint. Carved designs are likely to have had connexions with contemporaryink painting, while the moulded ones, so much more elaborate and slightly textured, were related to repoussein silver working, and also to textile designs, particularly to embroidered textiles. The introduction of cobalt blue pigment for painting on the admired white porcelain set the potters a special problem, because the types of decoration to which they had hitherto been accustomed were not executed using a brush and the traditional designs were also to some extent unacceptable to the new customers, who had been brought up in a totally different culture and artistic climate.
The Chinese solution to the problem was in fact to use their own decorative motifs, mainly from the repertory of moulding, but to use them in a new way. At this point form became an important factor. Once again the potters had a problem to solve; this time it was that the shapes popular on the domestic market were not those demanded, so they had to accept new forms. Up to this time, the shapes popular among the Chinese were relatively small, but they were now asked to produce very large pieces. Indeed, these last were immense, many of them four times the size of anything they would make for themselves.
6 Win-wu [" Antiquities "] (Peking, 1963), No. I, pp. 27 ff. Chinese). 'Chi-chou yao [" Chi-chou pottery "] (Peking I958). Chinese). The authors state on p. 63 that it is clear that loss of markets here in southern Kiangsi was partly due to
4A

(In (In the the

drift of craftsmen north to the better financed and more prosperous kilns in the great centre of Ching-t6 Ch6n. 8 M. Medley, Metalwork and ChineseCeramics,Percival David Foundation Monograph series No. 2 (London, I972), pp. I213.

34

JOURNAL

OF PERSIAN

STUDIES

They were thus faced with an enormous surface to decorate and relatively little idea of how best it could be done. In these circumstances, what could be more natural than to make use of the same approach as that adopted by their Islamic counterpartsand other Islamic craftsmenworkingin different media? In turning to Islamic lustre wares and metalwork-the latter especially, since it was a major import from the Near East-the potters gradually evolved a style worked out on similar principles, often in concentric rings on a plate (Pls. Ia, b and IIb), or expanding out from a central medallion in a series of arcs and based on a fourfold division of surface (P1. IIa and c).9 The result was a curiously hybrid style, wholly Islamic in feeling although almost all the motifs were Chinese. The richness and unaccustomed elaboration of the effects were not only extremely gaudy but positively vulgar to sophisticated Chinese taste. Such large plates had absolutely no place in a Chinese household and it is significant that no plates of such large size, that is of about 35 cms and more, have been found in China until very recently, and certainly none that can be dated to the Mongol period, which came to an end in 1368. The earliestones to have come to light probably date from about 1380, by which time the style had begun to undergo an internal transformationsuch as to make blue and white more acceptable to the Chinese.10 Parallels of the earlier fourteenth century between China and Islam are not hard to find. Thus the concentric rings, common to both metalwork and lustre in Islamic wares, as figured in the bowl (P1. IIb), is repeated in Chinese terms in the plates decorated in blue and white in P1.Ia and Ib. The parallel is obvious although the motifsthemselvesare distinctivelyChinese in origin. There is, however, one motif, often described as the " lotus panel " seen in one band in each of the plates in P1.Ia and b, which is just as distinctively Islamic."' It will be noticed that the panels are of two types, but both have the same ultimate origin in the Idm-alifcombination so familiar in Islamic art. The type shown in P1. Ia, for instance, can be paralleled by the very similar treatment in one of the illustrations in the Edinburgh al-Birini (P1. IIIb), while that in Pl. Ib can be coupled with not only similar patterns on lustre tiles, but also on the more elaborate type in the bordersof carpets, such as the one illustrated in the Cairo Kalilawa-Dimna(P1.IIIa). The Chinese may very well have seen the first type on metalwork and the second on the carpets such as could be found in the houses of the wealthy Persian merchantsin Ch'iian-chou, who would certainly do everything they could to stimulate both the acquisitivenessof their customersand the ingenuity of the potters. This type of concentric banding incorporating the ldm-alifelement was not the only treatment employed on plates. There are also a number of examples of two dimensional expanding designs worked out in multiples of four, of which the Rayy lustre bowl in P1. IVa is an example. This might by its curvilinear design have inspired the rather elaborate and symmetricalspecimen in Ardabil (P1. IIa). As time passed, the designs became more regular and better controlled, as though the Chinese painters had by the middle of the fourteenth century put together a series of pattern books. Indeed this is the most likely explanation for the constant repetition of a number of themes with only minor variations of detail. Some of the vases and ewers were also alien in form, but because they were sometimes smaller they were generally more elegant in contour. Another contributory reason was perhaps because they were ultimately based on the easily thrown pear-shaped form already familiar to the Chinese. The foreign element recognizablein these is the faceting of the walls instead of the body being left smoothly rounded (P1.IVb).12 When it came to bowls, the Chinese usually remained unaffected by Islamic taste, but one
type is worth comment. This is the type with a relatively small foot-ring, rather wide swelling walls and neatly in-turned rim. The shape, based on Syrian and Persian bronze forms of the thirteenth century, appears first in China in the celadons from Chekiang in the middle or late thirteenth century and is later taken over into the blue and white, where it persistedintermittently into the sixteenth century.
Medley, " Chinese Ceramics and Islamic Design ", p. 3. Medley, " The Yiian-Ming Transformation in the Blue and Red Decorated Porcelains ", Ars OrientalisIX (1973), pp. 89104; see especially pp. 95-6. 11 Lotus panels, or petal panels, two terms for the same motif shown in P1. IVc. The first name was given and defined by
10
'

CenturyBlue and White ... in the John A. Pope in Fourteenth TopkapuSarayi Miizesi, Istanbul (Washington, 1952), pp. 46-7. Petal panels is the name preferred by J. M. Addis, " A Group of Underglaze Red ", Transactions the OrientalCeramic Society of XXXVI (1957-59), PP. 43-72. 1s Medley, Metalworkand Chinese Ceramics, 14, Pls. 46a and b. p.

ISLAM,

CHINESE

PORCELAIN

AND

ARDABIL

35

In the blue and white it was frequently decorated in the rich Sino-Islamic style, the largest and one of the most brilliant examples being that in the Ardabil Collection (P1. IVc and Va). The decoration of this handsome specimen is also a good example of Islamic influence in the use of space with concentrically arrangedbands. The ldm-alifpanelsshould be noted below the rim on the inside, while the true " lotus panel ", or petal panel motif appears in the lower zone on the outside. Unlike most of the other imported forms, this bowl shape seems to have been integrated into the Chinese celadon repertoire, particularly in the smaller sizes, of which a large number have been found in Chinese excavations. It was during the last thirty years of the fourteenth century, however, that a change began to come over blue and white.13 The change coincided with a break in the trade with Persia occasioned partly by the disruptiveactivities of Timur and partly by the replacement in China of the Mongols by the new native dynasty of Ming. The first emperor of this family placed certain restrictionson foreign trade, and he was at the same time very worried about the threatening attitude of Timur. It is significant that the Ardabil Collection only includes two examples of the blue and white produced in the years between c. I1365and 1410o,while many pieces of this period are to be found among the remains at Fustit and there are several also in the Topkapu Saray collection in Istanbul.14 The total output during this period is difficult to estimate, but that the decorative style became sinicized during the last thirty years or so of the fourteenth century is clear; for when, after the death of Timur and the ascent of the emperor Yung-lo to the throne in China, trade was once more encouraged, the porcelainsfrom China displayed quite different characteristicsfrom those formerly seen. Yung-lo encouraged the commercial expansion of China into South-east Asia, India and the Near East by instituting a series of maritime expeditions which were as impressive for their size as for their frequency.'5 They began in I405 and continued intermittently until as late as 1433, the most important of them being under the command of Cheng Ho, the famous eunuch admiral. At the same time, Timur's successors, especially Shth Rukh and his son Ulugh Beg at Samarqand, displayed great interest in re-opening relations with China, and Ulugh Beg in particular acquired a large collection of Chinese porcelain.'e Embassiesbegan to travel regularly during the first third of the fifteenth century between Iran and China, and considerable quantities of high value material undoubtedly passed to and fro either overland by Central Asia, or by the sea routes round the south of India and through the Malacca in Straits. The results of the renewal of trade are to be seen in a seriesof interesting cross-fertilizations the porcelain of China and the later earthenwaresof the Near East. The new sinicized blue and white decorative style was firmly establishedeven before the end of the first decade of the fifteenth century, so that when trade was re-opened on a large scale, most pieces intended for export from China bore designs which were wholly Chinese in taste. The forms also were for the most part Chinese in origin and those of Islamic taste and ancestry are usually easy to identify, even when decorated entirely in Chinese style. Occasionally, however, both form and decoration are closely akin to Islamic traditions, and then such a piece stands out so sharply against the rest that one is probablyjustified in saying that it was made to special order.17An example of this type is a gourd flask of a form familiar in the first half of the century, of which there are several specimens in the Ardabil Collection. The ancestrycan be traced directly to the unglazed water flasksin common use from Syria to Afghanistan. While many of the Chinese blue and white ones are decorated in a predominantly Chinese manner, there are a few decorated differently. On these few, the major decorative element is so far the only one that it is possible to identify directly with a specifically Islamic design. The flask in

Is See n. 8 above. 14 Some from the latter collection are illustrated by Soame Jenyns, " Chinese Porcelains in the Topkapu Saray, Istanbul ", Transactions the OrientalCeramicSocietyXXXVI (1964-66), of PP. 43-72, P1. 45. 16The major expeditions were studied by Paul Pelliot, " Les grands voyages maritimes chinois au d6but du XVe siecle ", T'oung-PaoXXX (I933), PP. 237-452. 16 One of the great problems with regard to the Ardabil collection itself is where the pieces were located between the late four-

teenth and early fifteenth century. It is tempting to suppose that some may have graced the collection assembled by Ulugh Beg, but so far as I know there is no evidence at all of what became of his collection. 17 A particularly good example of a piece made to special order is the massive flask in blue and white, now in the Freer Gallery of Art in Washington, described and placed in context by Pope "An Early Ming Porcelain in Muslim Style ", Aus der Kunst (Berlin 1959), PPWelt derIslamische 357-375.

36

JOURNAL

OF PERSIAN

STUDIES

P1. Vb is one of these with the design on the lower bulb being immediately identifiable with a mosaic tile decoration from Bey?ehirnow in the Karatay Medrese in Konya.18 In other fifteenthcentury examplesof Chineseimitation, or of adoption and modification,the nuances of Islamic form and decoration are much less explicit, as for instance in a group of ewers in which the basic form is copied almost exactly, even to the angled upward slanting spout and the reinforcingcollar at the base of the neck (P1.VIa),19 thus following the Near Eastern metal prototype very closely. The lobing of the body of the metal form, however, is less easily imitated in thrown ceramics and this is thereforeimplied by the vertical division of surfaceinto a seriesof well defined panels. The decoration, unlike the form, is convincingly Chinese with its fluent floral scrolls and sprays. Similar comments can be made of a number of other forms, including a group of rotund jugs whose origins can be traced to both Islamic earthenware, again unglazed, and to metalwork. At this point it is possible to turn to Iran and see something of the impact of China on the potters there, for until the fifteenth century at the earliest, the Iranian potters apparently remained unaffected by the Chinese imports of blue and white. Perhaps the hybrid style was too complex or the conditions and technical knowledge unsuited to the production of similar wares; or perhaps it was simply that the political situation did not favour the exploitation of new ideas. In fact, it is likely to have been a combination of all these factors which inhibited the potters. At all events, the greater assurance of the simplified decorative schemes of the fifteenth century evidently held an attraction, and from now on both form and decoration are taken over with such little modification as may be suggested by the stout jug form or none. It is noteworthy that when the Chinese copied the jug form it was from the metalwork prototype, as the re-inforcingcollar familiar on the metal model is retained and the decoration adjusted to take account of this feature (P1. VIb). On the other hand, the Iranian potters, either ignorant of transmission and transformationof form, or simply ignoring what they regarded as an unnecessary elaboration, omitted the reinforcementand implied the joint by the organization of the decoration. To suppose that the Iranian potters only accepted from China formswhich were specifically Islamic in origin would be a mistake. Bowls dating to the late fifteenth or early sixteenth century are known in which both form and decoration are faithfully adopted (P1. VIIa and b). The best known examples are probably the Barlow and Reitlinger one20 and one in the Victoria and Albert Museum, both of which are exceptionally close to the Chinese originals. Plates, too, are not uncommon, having been by now accepted by both China and Iran into the ceramic repertoire. The Iranian ones are, however, treated in the new sinicized style. This makes the one in the British Museum an exceptionally interesting parallel, based as it is on the fully developed early fifteenth century Chinese style (Pls. VIIc and VIII). The flowing rhythm of the well-spaced design in the central field has been taken over, at least conceptually, as has the series of detached flower spraysin the well, though the Iranian potter has evidently preferredto connect them in some way. The rim is filled in with a somewhat uncertain wave pattern, a common theme on plates in the early part of the century.21 The chief impressionreceived on comparing the two pieces is that the Iranian example displays a preferencefor a degree of symmetrical repetition such as the Chinese designertends to avoid in an explicit form, preferringan oblique reference to this type of treatment. If the Chinese examples of the period, however different their designs and however apparentlyclose to the Islamic copies, are compared, it will be seen that even in the Iznik wares, this tendency remains right down to the sixteenth and seventeenth centuries.
It is precisely in the treatment of surface that China exerted her most powerful impact on Islamic concepts of ceramic decoration. Prior to the fifteenth century, the Chinese working in blue and white had been content to accept the two dimensionality of Islamic design, with its decorations capable of infinite extension into space and so filling it. But once the mathematical division of space which was a concommitant of this approach had been mastered, the Chinese adapted the fundamental need for division of surface to their own ends, and reverted to their own deeply-ingrained habit of using space in
" 18 M. J. Rogers, Seljuk Influence on the Monuments of Cairo ", Kunstdes OrientsVII (1970), p. 54. The Ardabil specimens are Nos. 29.458 and 29.463, the former illustrated in Pope, ChinesePorcelains from the Ardebfl Shrine (Washington, 1956), P1.5519The Ardabil piece is the unnumbered one now in Igfahin; it is illustrated by Pope, Ardebil,P1. 54. 0o Lane, LaterIslamicPottery(London, I957), Pls. I8 and i9. A. 21 Chinese examples in Ardebil,Pls. 37-41.

P1. Ia. Chinese blue and white plate D. 57'5 cm. i4th centuryArdabil Collection.

P1. Ib. Chinese blue and white plate. D. 47 cm. Locsin Collection.

4B

Pl. Ha. Chinese blueand whiteplate. D. 46 cm. i4th century ArdabilCollection.

bowl.Kdshdn.D. ca. 20 cm. Early 13thcentury. Ades. Pl. IIb. Lustre R.

Pl. IIc. Persia. Brass tray. D. 34 cm. 12th or 13th century.

P1. Ilia. Detailfrom MS. Kalila wa-Dimna, Cairo.

al-Birfini. Pl. IIIb. Edinburgh

Pl. IVa. Persia. Lustrebowl. Rayy. D. 49 cm. 13th century.

Pl. IVc. Chinese blueand whitebowl. D. 37 cm. 14th century.ArdabilCollection.

blue Pl. IVb. Chinese and white facetedvase. Japanesecollectio

Pl. Va. Insideof blueand whitebowl reproduced Pl. IVc. on

Pl. Vb. Chinese andwhite blue flask. Ht. 26cm

Pl. VIa. Chinese andwhiteewer. Ht. 34 cm. i5th century. Ardabil blue Collection;Isfahdn.

Pl. VIb. Chinese and white blue jug. Ht.

13.

6 cm. 15th

Pl. VIIa. Blue andwhitebowlD. 15th century.Freer 22"6cm. Galleryof Art.

Pl. VIIb. Blue and whitebowl with blackdetails. From Iznik. D. 17 8 cm. Early i6th century.Victoria AlbertMuseum. and

Pl. VIIc. Chinese blueand whiteplate. D. 37'5 cm. Early z5th

Pl. VIII. Persia. Blue and white plate. Late i5th or early i6th century. British Museum.

ISLAM,

CHINESE

PORCELAIN

AND

ARDABIL

37

such a manner as to imply a third dimension. They achieved this by using decorative elements which could stand independently of each other and yet were spatially related on the surface. Each in some way was an unrepeatablevisual event incapable of geometricextensioninto space. The actual painting of the motifs, with their carefullycalculated modelling (as shown in the Addis plate), effectivelyimplies a third dimension, with the living lotus spray floating out from the surface; a similar effect is recognizablein the flower sprays in the well. In the fifteenth century, it was this concept of the treatment of decorative elements and their organization in space that had an impact on Islamic pottery design which has been irreversiblein its effects. The impact was not confined to Iran, but is seen in Egypt, in the work of Gaibi and his contemporaries,in Syria, and in the Iznik wares of Turkey. Ultimately the same impact, in a slightly different context was made on Europe. The actual elements constituting the Chinese style have now become so deeply imbedded in the traditions of both the Near East and Europe that it is not easy to recognize them, but once seen and their origins understood, they become as obvious as those of the Near East in Chinese traditions.

THE ARDABIL SHRINE IN THE REIGN OF SHAH TAHMASP I* (concluded) By A. H. Morton

ParaC. The Dar al-Hadith (4) and the tombs on either side of it correspondto the building which now occupies the south side of the Area (9), a large iwan flanked by two side chambers. At present it is known as the vault (tdq)of the Mutawalli. Whatever use the Mutawalli may have made of it once, the building now provides storage space for tilework removed during restoration and the like. When Dibaj describedthe shrine, there was no chamber to the west of the iwan. The door in the tiled arch of the fagade, which now leads into the western chamber, simply opened at that time into an extension of the uncovered area occupied by the ruins of the New Chilla-khana (13).1 Although the present western chamber is thus undoubtedly modern, the Sari/ al-Milk shows that there was a structure of some sort, serving as a tomb for members of the Safawi family, in the same position. While the rest of the buildings and the fagade appear to be of some age, it is difficult to be certain that they are those that existed in the time of Isma'il or Tahmdsp; but the description of the Sarihal-Milk, as far as it goes, points to their being similar and very possibly the same. No signs of tombs are visible in either of the side chambers, but it is impossible to see the floors entirely. In the reign of Tahmdsp the Iwan had an inscription including the name of Shah Isma'il. Presumably it was of the usual type, a tilework inscription on the pish-tdqof the iwAn. There is no good early photograph showing the whole of the iwan. Sarre says that there was little tiling left except on the eastern flanking chamber. His photograph of the chamber shows a little of the IwAnand its tilework is mostly missing.2 It seems that Isma'il's inscription did not exist then. The surviving tilework was, however, certainly of the Safawi period. Recently, the whole of the upper part of the facade has been retiled.3 From its name, it appears that the Dar al-IHadithwas originally intended as a place where instruction in the field of religiouslearning concerned with the sayings and actions of the prophet Muhammad, that is the Hadith, could take place. The type of institution is well known.4 While the al-Milkdoes not explicitly say that the Dar al-Hadith was built by Shah Isma'il, it is reasonableto assume that it was. S.arih Such an assumptionwould be extremely appropriate, for we know that at the time when Isma'il was proclaimed Shah in Tabriz the Safawi movement was distinctly short of orthodox Shi'i book learning.5 The creation of the Dir al-Hadith may have been brought about by a desire to alter the situation. A third designation for the buildings on the south side of the area appears in the caption to photograph No. I2 in the Gulistan Palace series. The photograph shows, among other things, the fagade of the easternchamber, which is describedin the caption as " the doorway of the night prayer-hallof the Mosque known as the Hasht Bihisht."6 Presumably the Dar al-IHadithwas used as a mosque in the nineteenth century, but this was not its original function. The houses (8) behind the Dar no longer exist. Nor does the street (7). At present houses abut directly on to the back wall of the iwan. al-.Hadith ParaD. The Area (9) is the present inner court of the shrine. At least part of the site of the court was a garden in Shaikh Safi's lifetime.7 On three sides, the layout is the same as it was in the reign of
* See also Iran XII, pp. 31-64.
1

Dibaj, p. 65. The National University Survey (Weaver, Preliminary Study,p. 133) appears to follow Dibij at this point, but the plan which accompanies the report (p. 176) shows the new western chamber already in existence. * Tafel XV.

None of the sources used mention this piece of restoration. 4 See El2 art. " Dgr al-IHladith" (Fuat Sezgin). 6Aihsanal-Tawdrikh, I, p. 61. * sardar-ishabistdn-imasjid-ima'rnaf-i hasht bihisht. bi SSee commentary on Para F.a.
3

39

40

JOURNAL

OF PERSIAN

STUDIES

Tahm~sp, and the various buildings are discussedelsewhere. On the western side, however, alterations were, it seems, made after Tahmasp's reign. At present, the western facade consistsof nine arched bays of uniform height. The central bay, the arch of which is slightly larger than the rest, forms the gateway leading into the little courtyard which in turn leads into the Yard (26). We lack an early photograph showing the four arches to the south of the gateway. Apparently this part of the faqade was in worse condition than the rest; in 1322/1943 the surviving tileworkwas removed and this section of the faqade demolished and then rebuilt after the former plan. The symmetryof the fagade and the uniformityof the stoneworkon the lower part of the wall make it appear that it was all constructed as a single scheme, in spite of the later doors and windows that have been inserted into the northern part. The gateway has two inscriptions that help to date it. Here again restorationhas confused the picture, and one cannot simply read the inscriptions as they stand today. The first inscription is above the doorway into the little court. Like the inscriptionover the doorway of the Dar it extends round the three sides of the arch above the door. Sarre's photographs shows theal-.Huffdz, section replaced by crudelyhand right arranged octagonal tiles and half of the centre section missing. The left side of the arch is not visible. The surviving section that can be seen contains the words al-Safawi Bahadur Khan, the usual titles of a Safawi Shah, and the very end of the ShAh'sname. An alif is perfectly clear, which al-I.usaini could be part of 'Abbas or Tahmasp. (As will become apparent no later Shah can be considered). In the upper registerof the inscriptionpart of the tail of a second letter can be seen. It is difficult to be certain, but it appearslikely that it is sin rather than bd'. Thus the inscriptionwas probably in the name of 'Abbas. The restorersof the tilework were evidently of that opinion. Dibatj'sdescription dates from the time when only the right-hand end of the central section had been restoredand the name of 'Abbas completed, for he says that the right-hand section was missing.8 Since then a new beginning has been added. The second inscriptionis on the left-hand of the faqadeof the gateway, written horizontally beneath the large Qur'anic inscription that runs round the faqade of the gateway. The first two lines survive. The rest is now lost and has been replaced with plain tiles (P1. Ia). The surviving portion states: " There exerted himself in renewing this noble building the descendant of the saints and the devotees, the scion of the Shaikhsand. . . .'9 It evidently recordsreconstructionor repairs. For the rest, one has to rely on the partial translation of the inscription as it was in 1897, by Hartmann, and on Sarre'sphotograph of the gateway which shows the inscription as almost completely illegible.10 Hartmann's reading of the beginning of the inscriptionagreeswith the survivingpart. The name of the subjectof the opening sentence, the person responsiblefor the renewal of the building, he gives as Shaikh Sharif al-Zahidi, the Mutawalli, and it is possible partially to verify the name on the photograph. The scribe who wrote the
inscription was one Ismi'il, and Hartmann read the date as 1309/1891-2. It cannot be seen on the

photograph but, in any case, it cannot be correct, though it provides one of the clues to the real date of the inscription. The inscriptionitself is a fine piece of tilework mosaic and appears to be of a piece with the Qur'anic inscription round the gate. In Sarre's photograph, the tilework of the gate is in rather decrepit condition. While it had been patched up here and there, these attempts at repair are extremely crude and it certainly had not undergone a major restorationat the hands of skilled mosaicistsonly five or six years before Sarre's visit. Furthermore, we know who Shaikh Sharif Zahidi was. As has been said in the introduction, he was the uncle of Shaikh HIusainZahidi, the author of the Silsilat al-Nasab, and was Mutawalli of the shrine for seventeen years. We know also that he was Mutawalli in 1038.11
It is natural to assume, then, that the date read I309 by Hartmann was in fact written 139, the nought This would being omitted, as it often was, and that the real date of the inscription was I039/1629-30. put it in the reign of Safi, but only just, for Safi came to the throne at the beginning of Io38; one cannot be certain that the inscription in the name of'Abbas and the renewal inscription do not both come from a single campaign of reconstruction. The inscriptions make it fairly certain that the western fagade was

in existence in more or less its present form in 1o39.

8 DibNj, p. 64. * Sa'd ft tajdid hddhih al-'imdrat al-sharifa khalaf al-a(u)liyd wa wal-salikinwa suldlat al-mashdyikh ....
10

"1 See Iran XII (1974), P. 32, n. 13.

Sarre, Abbildung 5, p. 21.

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

41

Its earlier history and the date of its construction remain obscure. The present gateway, built or rebuilt under Shah 'Abbds I, correspondsto the Lofty Gateway at which the doorkeeperssat (11) of the Sarihal-Milk, but is presumablya different building. The Sarihal-Milk, as far as it goes, seems to imply that the faqade did not exist in 977, since it merely mentions the gateway (11) and the two Chillakhanas (12 and 13) as adjoining or touching the area. Della Valle's description of the Area as an uncovered mosque, and above all, his comparisonof it to the Masjid-i Shah of Isfahan, then being built, gives the contrary impression and suggests that in 1619 the western side did have a unified faqade. Olearius remarks that the Area was " ... pav'd all over with little square-stonesof several colours." Perhaps the constructionof the faqadefollowed the disappearanceof the Old Chilla-khina (I2), which itself is not easy to date (see commentary, Para F.a.), and Shaikh Sharif's inscription only records redecoration and restoration and not a major rebuilding. The tomb of Tahmasp's mother (14) is not known at Ardabil today. On the east side of the Area, however, and close to, but not actually touching, the tomb of Safi, is a very low, domed tomb, nowadays said to be that of the mother of Shih Isma'il.12 The tomb itself is now concealed behind the screen wall between the Ddr and the Dar al-IHadith; a small door in the screenwall opens into it. As has been pointed out, the finely carved marble tombstone which is inside it bears a date of the eighth cenal-Huff.z tury of the hijraand cannot be that of Ismd'il's mother.13 The tombstone is also too large for the tomb and must have been put into it at a later date. As far as the Sarihal-Milk goes, the so-called tomb of Isma'il's mother admirably fits the description of the tomb of Tahmasp's. The only other reference to the tomb of Tahmqsp's mother confirms the impression. In the Tdrikh-i account of 'Abbis's 'Abbdsi's visit to the shrine in 1611, 'Abbds is said to have first visited the tomb of Safi, next gone to the kitchen and then visited the tomb of Tahmasp's mother. As he put his hand on the door the lock opened (a miracle repeated several times in different places in the course of the visit). Here the tomb is described as being " below, outside the IHaram."'4 The so-called tomb of Isma'll's mother is on a lower level than the tomb of Safi, which is elevated on a stone base, and it seems probable that it does in fact properly belong to Tahmasp's mother, who, it appears, was a woman of considerablecharacter. As for the hand of 'Ali, at present there is a large piece of stone with the mark of a gigantic hand on it built into the north wall of the interior of Ismd'il'stomb. In the earliest European referenceit is said to be the hand of Isma'il,16 but it is more usually called the hand of 'Ali.8l It is not part of the original wall of Ismi'il's tomb, and it is easy to suppose that it was first somewhere near the tomb of Tahmasp's mother and was later moved. ParaE. For the Jannat-sard (i o), the large octagonal building which occupies the north end of the Area (9), the Sarihal-Milk provides a definite dating. The list of propertiesgiven as Para Y.a.-e. shows that part of the site of the actual building was bought at the end of 1536 and that other propertiesfor the dependencies and garden were bought between that date and 949/1542-3. One can probably date the actual construction to around 1540. It was certainly complete by the date the Sarihal-Milk was written in 977/1570. At that time the pish-tdqof the Iwan of the building displayed an inscription, no doubt of tilework, with the name of Shah Tahmasp, and the large size of the building makes it
12 Dibij, p. 72. 13 Weaver, Preliminary Study, pp. viii, 23.
14

dar pd'in md az biran-i haram (Tdrikh-i 'Abbdsi, f. I96b). The accounts of Tahmdsp's mother are contradictory. According to the Tdrikh-i'Alam-drd,p. 44, she was the daughter of one of the Amirs of the Mausillfi tribe. Giovan Maria Angiolello says she was the granddaughter of Ya'qfib Sultan Aqquyiinlii and was called Taslucanum, i.e. Tajlfi khdnum. Her family was reluctant to have her marry Ism'lil, presumably because IsmA'il was at the time engaged in destroying the Aqquyiinlfi dynasty (A Narrativeof Italian Travelsin Persia (Hakluyt, First Series, No. XLIX, I873), p. io6). The Mausillfi were a cadet branch of the Aqquyfinli, and the two stories can be reconciled by assuming Tajlii khdnum was a direct descendant of Ya'qfib in the female line. A number of stories about Tahmasp's mother are preserved in the anonymous seventeenth century compilation which has
5A

been published under the title 'Alam-drd-yiShah Isma'tl (ed. Asghar Muntazir-i SIhib, Tehran, I349, see index s.v. Tajlfi Begum). In this work she is said to be the daughter of 'Abidin Beg Shamlii. Of surpassing beauty, she was equally remarkable for her physical strength and fighting spirit. She was able to outwrestle her brother, Dfirmish Kh~n Shfmlfi, and made it a condition of her marriage to Ismi'il that she should be allowed to participate in battle. Clad in armour, leopardskin and veil, she played a conspicuous part in the battle of Chdldirdn. While these romantic tales need to be taken with a grain of salt, they may reflect part of her activities which more conventional historians preferred to ignore because she was a woman; this would help to explain the respect in which her grave was held at the time of 'Abbas. on the Caspianshore,p. 39. 15 Holmes, Sketches 16 E.g. Sarre, p. 13.

42

JOURNAL

OF PERSIAN

STUDIES

probable that it was commissioned by the Shah personally. Sarre was badly misled about the date of the Jannat-sard. He did not see the area where the ruins of the New Chilla-khdna (13) stood, and as a result assumed that Olearius's description of the Chilla-khana as a large, domed structure referred to the Jannat-sari. Olearius's incorrect statement that the (New) Chilla-khdna was built by Safi al-Din, together with the similarity that exists between the plan of the Jannat-sard and that of the tomb of led him to date it variously to the early fourteenth or even thirteenth centuries."7 Qljditii at The Sarih al-Milk leaves no doubt that the Jannat-sart was originally covered by a dome. The Sult.niyya, account of the repairs carried out on the building by Shaikh Abdal Zahidi in the reign of 'Abbas confirms the point: The Jannat-sard had not been touched from the time it was built. Water had penetrated the roof and stained the walls black with an efflorescence. Shaikh Abdal capped the top of the dome with brick and mortar. He built a wooden scaffolding inside and whitewashed the interior and made the windows gul-jdm.18The building had not been previously repaired because it was very difficult to erect a scaffolding on account of the great size of the Jannat-sara.19 Dibaj makes the curious statement that the facade of the Jannat-sara, which forms the north side of the Area (9), used to consist of four bays of the same size as the two surviving side bays. In 1307/188990, in the reign of Nisir al-Din Shah, the two central bays were demolished and the present large iwan, designed to correspond with the iwan of the Dar al-IHadith (4), replaced it.20 This is plainly untrue. Sarre's photographs show the iwan as an old building still retaining some of its Safawi tilework.21 On their evidence, the restorations of 1307 went no further than the plastering over of areas where the tiling had fallen off. One of the Gulistin palace photographs shows part of the iwan before the restorations, as does one of de Morgan's illustrations (Pls. III and IV). Finally, the Sarih al-Milk and Della Valle make it quite certain that the iwan did exist in the Safawi period.22 The iwan had to be repaired in 1314 shamsi/1935. Surprisingly, none of the European travellers of the Safawi period has left a recognizable mention of the Jannat-sart, apart from Della Valle's reference to its iwan, in spite of the fact that it is the largest building in the shrine and its dome must have dominated the Area (9). The building which de Bruin describes as covered by a dome, unsupported by pillars, like that of the Rotunda at Rome, but smaller, is wrongly placed to be the Jannat-sari and must be the New Chilla-khana (13). Struys's mention of a building " round like a theatre " may refer to the Jannat-sari, but his description of it is in parts composed of reminiscences of Olearius's account of the Area (9) and the portal of the Dar (2). al-.Huffaz Possibly the dome caused trouble again and fell or was removed not long after Shaikh Abdal's repairs. The earliest firm evidence we have comes, however, from as late as 1843, when Holmes describes the building as roofless.23 At present the Jannat-sara is known as the Masjid or Mosque of the Jannat-sara, and before discussing to what extent the term is correct it is convenient to consider the evidence for the existence of a mosque at the shrine at all. As we have seen Anthony Jenkinson refers to the shrine as a whole as a Meskit. Later travellers did the same and were also liable to call any reasonably grand building which seemed to be designed for some sacred purpose a Meskit, Mosquey or the like. Continuing this tradition, both de Morgan and Sarre refer to the shrine as a whole as a mosque, or, in an attempt to refine the term, as a funeral mosque. In Persian sources, however, from the earliest time to the present day, the shrine is never referred to as a mosque. More remarkably, in the early Persian sources there is no reference to any mosque within the shrine. The evidence of the Sarih al-Milk seems decisive: the one mosque (i9) that is mentioned is only noted as defining the wall of the Shahid-gah (I5), and must have been
9, I6. s8Or gil-jdm? The meaning of the word is not explained and it does not seem to occur in dictionaries. Possibly it refers to the insertion of grilles of plasterwork like those of the Chini-kh~na in the windows. 19 Silsilat al-Nasab, p. 112.
17 pp.

20 pp. 65 n, 86. Della Valle describes two vaulted " tribunes " at the entrance and end of the Area which must be the Iwins of the Jannatsari and Di.r al-IHadith. 22 See note 15 above.
22

21 Tafel V; Abbildung 13.

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

43

outside the shrine since otherwise it would have been included in the list of the shrine buildings. As we have seen, the Dar (4) was used as a mosque in the nineteenth century, but, again, the al-Milk shows that it was not originally one. al-.Hadith As or Sar.h for the Jannat-sari itself, in the Silsilatal-Nasabit is called the 'imdrat building of the JannatThe term 'imdrat a vague one which could be applied to any building of some pretension, but is sarA. sounds odd applied to a mosque as a whole. In the SariJal-Milk the building is referredto several times as Jannat-sari, which in itself is an unusual way of referringto a mosque, and once, at the beginning of the section, as the dome of the maqsara the Jannat-sara.24 The word maqszra frequently applied to is of It may mean an enclosure parts of mosques and in such a context may have two distinct meanings. within a mosqueintended as a place for the ruler to say his prayers,or it may simply referto a " detached part of the mosque which is set aside for communal prayer."25The second meaning might be thought to have some relevance to the Jannat-sara at Ardabil. However, the form of the building makes it impossible that it was designed as a mosque. The plan of the building is octagonal and the facet of the octagon which correspondsmost closely to the qiblahas in its centre not a mihrdb a doorway. The facet next but to it contains the large, open archway leading from the iwan facing the Area (9). Part of any large congregationwould have been looking into the Area while praying. That the buildingwas later used as a was mosque there is no doubt. When Sarre saw it, a minbar in position and apparently the wall with the there. However, the use doorway in it was being used for the qibla.26Until 1971 there was still a minbar of the Jannat-sardas a mosque is a very late development. Holmes says there was no roof at all in 1843,27 and the Jannat-sard probably became a mosque when the flat roof supported on wooden pillars which Sarre saw was built. Parts of this roof collapsed in recent years, and in I972, as part of the currentprogramme of restorations, it was removed. One of the wooden capitals bears a building inscription dated 1299/1881-2 (P1.Ib).28 Very possibly the Jannat-sara only served as a mosque after that date. Its use was discontinued when the roof became dangerous. The word maqsira,applied to the Jannat-sara does not necessarily imply that it was a mosque. In fact Arabic dictionaries give the word as meaning primarily a spacious enclosed apartment, whence it used to refer to structures as different as palaces and dog
kennels.29

Since it was not a mosque, what might it have been? Mr. Weaver has put forwardthe theory that it the may have been a dervish maiddn, hall used for their ceremoniesby some Saifiorders, particularlythe In the absence of much informationabout the practicesof the Safawi order in the sixteenth Bektishis.30 and seventeenth centuries, one cannot be certain. As in the case of the Chini-khana (3), the choice seems to lie between assuming that the building was meant for some such ritual function special to dervishes or that it was a tomb. If it was a tomb, one might even suggest that from its size it was intended for the burial of Shah Tahmasp, who at one time may have thought of imitating his father's example and being buried at Ardabil, although in fact he was buried at Mashhad.31 The Sarh al-Milk'saccount of the Shahid-gah (15) is one of its most complex passages. Nevertheless, by considering other information we have about the Shahid-gih one can make sense of it. Nowadays Shahid-gih is the name given to the large cemetery which occupies the eastern end of the shrine. It appears from Dibaj's account that it properly includes part or all of the small enclosed cemetery area On immediately to the south of the tomb of Safi (I) and the JHaram-khdna. the other side of the eastern wall of this area the continues. It extends past the Chini-khana, and some way beyond the Shahid-gvh
level of the Jannat-sara, the boundary wall turns towards the west and continues until it meets the wall and fence of the Safawi Secondary School for Girls at a point almost level with the east side of the Jannat-sara. The grounds of the school now bound the Jannat-sarai on the north and west. The school
24

In the Silsilat al-Nasab (p. I I2) the Jannat-sard is similarly p. 3. It reads: katabahu al-'abd al-mudhniblibn 'Ali Asghar either called Jannat-sard or referred to as 'imdrat. ft Shfrwadn; tahrtran hashtum-i/ shahr-i Ramaddn al-mubdrak; " 25 R. Hillenbrand, Saljiiq Monuments in Iran: I ", Oriental z299; 'amal-i Karbald'iMuhammad-i Vajjdr.A word to the left of 99 remains illegible. Art, New Series, XVIII, p. 77. 26 p. 9. 29 See the dictionaries of Lane and Dozy. 27See note 15 above. 30 Preliminary Study, pp. I4-16. 28 The inscription is mentioned first by Weaver, Second Study,II, 3' Tdrikh-i'Alam-ard,p. I23.

44

JOURNAL

OF PERSIAN

STUDIES

was built in 1328/1949 and its grounds were formed from what was then a further extension of the Shahid-gah, though the western part of it proved to contain the remains of the foundations of various buildings.32 Certain parts of this large, irregularcemetery area can be shown to have served as graveyards for a very long time. In Shaikh Safi's lifetime the Graveyardof the Disciples (Mazdr-iMuriddn) was located just to the south of the house where his tomb later stood. This correspondsto the little courtyard south of the tomb-towers. According to Dibaj, about two metres of soil were removed from around the base of the Chini-khana, in an attempt to combat rising damp in the building. From the present appearance of the area as compared with Sarre's photographs,it appears that most parts of the rest of the Shahid-gdh have also had a fair amount of soil removed.33Yet in spite of this,when, in 197I, a trench was dug round part of the Chini-khdna,once again to combat the rising damp, it was revealed that the earth to a depth of over two metres more was occupied with layer upon layer of burials.34 A sondage carried out by Mr. A. A. Sarfarazof the Iranian Archaeological Service in the same year to the east of the Chini-khdna produced no traces of buildings, but graves and pottery dating from the fourteenth or even thirteenth century.35 Bearing these considerationsin mind, it becomes possible to explain the text of the Sari al-Milk. The statement that the Shahid-gdh is north of the dome of the Jannat-sard is only true of part of it. As has been suggestedin the translation,what the Shahid-gah enclosurebounded on the east was the whole shrine complex. The sides of the enclosureto the north and south representthe boundariesof the enclosure at its northernand southernends. The turquoisedome is not, as one might at firstsuspect, the dome of the Jannat-sard, but the dome of Safi's tomb, to which the same epithet is applied in the introductory paragraphon the whole shrine. The boundary wall started from near Safi's tomb, ran up to Sayyid 'Ali Beg's house (17) which itself formed part of the boundary and then continued to the house of the heirs of Sayyidi Beg (18), which was behind the wall. From there it went to the mosque (19), which too was not part of the shrine. The western part of the Shahid-gdhat the northernextremity ended in the vicinity of the Kitchen (20) and the Bath (21). Exactly where the two housesand the mosque stood is impossible to determine, but they may have been mentioned as marking the points where the wall changed direction and, in the case of the mosque, where it ended. Early in the reign of Shah 'Abbds I, the wall (hisdr)of the Shahid-gah had fallen and disreputable used to be able to get in. Shaikh Abddl Zdhidi built a strong new wall.36 The people (ajldfwa ajdmira) of the shrine published by James Morier shows that the Shahid-gdh had no wall in the early engraving nineteenth century; a pathway crosses its north east corner and sheep are being pastured in the cemetery.37 According to Diba-j,at the end of the Qajir period and later, the Shahid-gdh was being encroached upon by houses and otherwise serving as a rubbish dump. In 1322/1943 and later it was cleared and a new wall built.38 The Farrdsh-khana(16) was evidently located in an ordinary house. Thefarrdshswould have been responsiblefor keeping the shrine clean and performingother domestic services about it. One of the who lit the lamps in the waqf documents in the archives provides for money to be given to thefarrdshs Old Chilla-khana (I2).39 A keeper of the lamps (mash'al-ddr) also mentioned, but it is clear that the is were, at least in part, responsiblefor the lighting of the shrine, for which elaborate arrangements farrdshs existed. The list of propertiesacquired for the Jannat-sara site (Para Y.a.-e.) shows how the shrine expanded
by absorbing neighbouring houses and courtyards. It does not seem possible to recreate the plan of the area before the building of the Jannat-sari, since the description of the boundaries of each of the properties are taken from the deeds of purchase and waqf and are not necessarily consistent with each other. The author of the Sarih al-Milk may not himself have known what the area had been like. The first item
82

Dibaj, pp. 82-3. ** Sarre, Abbildungen 7, I1. 34 See Weaver, Second Study,p. io. 85 The evidence for the early existence of a cemetery in the Shahid-gsh area makes it clear that the story that it was created for those who fell in a particular battle, be it Ismi'il's defeat of

the Shirwinshih or the disaster of Childirtn, cannot be true of the cemetery as a whole. 36Silsilat al-Nasab, p. II 2. 87 A second Journey,facing p. 253" Dibij, p. 83. 3"See note 42 below.

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

45

does present some points of interest. The mention of the Lane of the Shaikhsindicates that there was an entrance to the shrinein this area. Della Valle, indeed, says that there was a little gate which led into the street beyond the kitchen. Perhaps, in the early days, the main entrance to the shrine was here. The 'Imarat-i Shihabiyya must have been some fairly grand building attached to the shrine. It seems likely from the location given for Khwaja Khdn WRhid'shouses that it stood on the easternpart of the site of the actual Jannat-sari and was demolished to make way for it. Its name leads one to suppose that it was built by Shaikh Shihdb al-Din Mahmild, the son of Sadr al-Din, who appears to have been Shaikh of the Safawi order for a period after his father's death.40 Finally, the Jannat-sara garden mentioned in items Y.b. and c. must have stood to the north of the Jannat-sard, which was bounded on the west by the Kitchen (20o) and on the east by the cemetery. When Olearius came out of the Kitchen (20o)he entered " a very fair Garden " in which he saw a number of sepulchres " without any thing over them, but a smooth stone. "41. He must have gone from the kitchen into the northern part of the Shahid-gah and his description, together with the fact that the al-Milk does not mention the garden of the Jannat-sard as existing at the time of composition, Sar4h makes it appear likely that the garden of Jannat-sard became transformedinto an extension of the Shahid-gah soon after its creation. ParaF. Previously,the Old Chilla-khina (12) was only known from some brief mentions in a much in damanged waqf-ndma the Ardabil collection.42 The Sarihal-Milk preservesa more complete text of the same waqf-ndma, which proves to be dated Sha'ban 977/Jan.-Feb. 1570.43 At the stage when the waqf-ndma provided our only information about the Old Chilla-khana, the theory was put forward that the Old Chilla-khdnaof the waqf-ndma was what we now know to be the New Chilla-khana (13) and that the New Chilla-khanaof the waqf-ndma the Jannat-sar (10).44 The Sarikal-Milkshows that it was was not so. The Old Chilla-khanahas vanished, although it is possiblethat tracesof it may still exist in the buildings at present on its site. The indications which we possesspoint to it having disappeared a long time ago. The Silsilatal-Nasab,in the list of ShaikhAbdil's constructionworksat Ardabil, mentions extensive repairsmade to the Noble Chilla-khina,45but does not specify which one. The author knew the shrine well and could hardly have failed to distinguish the two buildings if it had been necessary. Similarly, in the eighteenth century inventory, in an entry on a finial (tilq) of gilded copper, it is merely stated that it was attached to the dome of the blessed Chilla-khana46and was brought down when the dome broke. The Chilla-khdnadescribed by Olearius, although he says it was used by Shaikh Safi, is quite clearly the New Chilla-khdna (i3). No other European of the period mentions the Old Chilla-khdna. Since the Sarihal-Milk states that the Old Chilla-khqnatouched the Area (9) between the main buildings of the shrine, evidently at a point where the Area is now defined by the western facade (which dates from the reign of 'Abbds or earlier), it had probably vanished by the time the facade was constructed. Its disappearancemay thus be put at the end of the sixteenth or beginning of the seventeenth century. The Old Chilla-khanawas presumablybuilt in the lifetime of Safi. At the least, it musthave occupied the site of a building which existed in Safi's lifetime. It could be one of the buildings mentioned in the Safwatal-Safd. Referencesto buildings that could have been the Old Chilla-khanain the Safwatal-Safd fall into two classes: referencesto zdwsyasand referencesto khalwats. To take the zdwiyasfirst, the word means corner and hence, by extension, secluded corner, retreat, hermitage or dervish khdnqdh.It is used on occasion in the Safwatal-Safdto refer to the shrine as a it whole,4' as it is in the Sarih al-Milk and numerous documents. However, in the Safwat al-Safad is
In the al-Milk, documents are quoted in which Shihib al-Din Sar. Mahmfid is given titles which make it clear he was Shaikh. The writer hopes to examine the Sarih al-Milk's information about the genealogy of the Safawi family elsewhere. 41 Olearius says he saw the tombs of Sultan Ijaidar, Shih TahmIsp and several other kings of Persia in the garden. TahmAsp was buried at Mashhad and Ijaidar's grave is now supposed to be in the tomb of Safi. 42 Inventory No. 403. An English translation of part of the docu40

ment by the writer is given by Weaver, Second Study,II, Appendix I. 48 The name of the donor, which is missing on the actual document, was Malik Mirza Beg Mughdni. 4 Weaver, SecondStudy, II, pp. 1-3. sharffa. 45 P. 12, chilla-khana-yi 46 Ganjina-yi Shaikh Safi, p. 62: dar gunbad-i chilla-khdna-yi mubdraka nasb bbad.

"7P. 258.

46

JOURNAL

OF PERSIAN

STUDIES

frequently also used, with or without qualification, to refer to particular buildings within the shrine complex. An anecdote told by Sadr al-Din relates that at the beginning (darawd'il-ihal; i.e. when the present site first became Safi's khdnqdh), Shaikh Safi built himself a zdwiyawhich was rathersmall. In spite of its modest size and the enormousnumber of Safi's disciples, there was never any shortage of room. It was replaced by the large Zdwiya which was built by Amira Pahlawdnof Gilan who had begged the Shaikh to allow him to build it. Although it was a spacious building, even a small crowd of people often found inadequate room in it.48 The story of IHIjjiSam's zdw*ya quoted above says that the Shaikh had built a zdw ya of unbaked brick.49 This is the same building as is referredto in Sadr al-Din's story, for it said to have had the same quality of never lacking space. Yet another designation for what is presumably the same building occurs in a story about how Safi saw to it that the qiblawall was properly aligned.50 Here it is called the Old Zdwiya, as in one other place.51 To sum up, the building in question, which we may conveniently call the Old Zdwiya, was built by Shaikh Safi and was a comparativelysmall structure of mud brick. As for further mentions of Amira Pahlawdn's zdw ya, when the Ilkhan Sa'id Abni (I316-1335) visited Safi he saw a large zdwzyaof baked brick, which caused him to have doubts about the Shaikh's reputationfor sanctity.52 (They were soon dispelled.) In another story, Safi appears resting on the roof of the Large Zawiya.53 The New Zawiya, which is once mentioned, is presumablythe same building.54 Thus we can say that the New Zawiya was a comparativelylarge building of baked brick put up at the Sam's unused zdw ya. Both expense of Amira Pahlawdn. It was, we can suppose, built after Hia-jji yas appear as places in which Safi met his disciples and gave them guidance.55 zdw The word khalwat used in three distinct ways in the Safwatal-Safd. It originally means withdrawal is or privacy and is used by the Silfis to referto the withdrawalof the Silfi fromworldly society to commune with God. Such withdrawalswere normally accompanied by fasting and other forms of self-mortification. In this sense, the word occurs in the Safwat referringto the withdrawalsof Shaikh Safi, which took place regularlyin Ramadan.56 It is alsoal-S.afdreferto the quartersof visitorsto or residents used to at the khdnqdh. an example of this usage, one of the visitorsrelates how Safi ordered that he and two As should be assigneda khalwat front of the tomb of Muhiyy al-Din, and they lodged there.5 in companions Such personal khalwats perhaps to be connected with a practice introduced by Safi and called the are Dervishes' retreat.58 Very large numbers are said to have attended on the occasion of such retreats, which, one imagines, consisted of periods of communal withdrawal. The personal khalwatsreferred to in the Safwatal-Safdwould probably have been the quarters given to particular dervishes on such occasions. The famous story of the ordeal of the poet Qasim al-Anwdr illustrates the use of the word in khalwat a similar sense, though not, it appears, on an occasion of general retreat.59 In the seventeenth century, Qasim al-Anwar'skhalwat,that is, the spot where he underwent his ordeal, was still known in the (New) Chilla-khana (I3). According to the story, after he had served Sadr al-Din for some time, Qasim al-Anwar was told that the only way he could be " cured " was by passing a period of forty days standing all the time, except when the requirementsof prayer made it necessary to do otherwise. He thereforetied a rope to the roof of the khalwat, attaching the end of it to his long hair and so managed to At the end of the khalwat(here referringto the period of trial itself) he was rewardedby a keep standing. dream in which he saw himself standing in the Friday Mosque at Ardabil with a large lighted candle in his hand from which he lit other candles, brought to him by a great throng of people. From this dream
he received the name by which he became famous, Qasim al-Anwar, " Distributor of Lights ".
48 P. 258. nuzal kardim; also pp. 235, 283. 58 Silsilat al-Nasab, p. 38. 49 pp. 299-300. 15 Ibid., p. 50 P. 287. 40o. Whether there is any truth in the story does not matter to a discussion of the meaning of the terms employed. "1 P. 268. 52P. 290. However, it should be mentioned that there are various more or less different accounts of this or similar episodes in the bio8SP. 219. 65 P. 268. graphical material on the poet's life printed by Sa'id Nafisi in his edition of the Kulliyydt(Tehran, 1337). The earliest such 55Pp. 258, 268. al-Shu'ardof Daulatshih (London story, found in the Tadhkirat "5Pp. 195, 303. 11 P. and Leiden, 1901, pp. 346-50), is set in the Masjid-i Jdmi' of 340, ishdratfarmid td khalwat-i md har sih pfsh-i marqad-i munawwar-i Khwdja al-Din mu'ayyankardand wa anjaQazwin and Sadr al-Din is not mentioned. Mu.iyy

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

47

The third sense of khalwat found in the Safwatal-Safais not unconnected with the first two. Khalwat refer to a particular building. For instance, when, three years after Shaikh Zahid's death, may Safi went to visit his master'sshrine, and Shaikh Zahid's son, Jamal al-Din 'Ali, tried to assassinatehim, An there were of wood and khdshdk.60 attempt was made to set Safi's personal the houses and khalwats on fire. Elsewhere it is stated that when Safi withdrew himself during Ramatdin, none of his khalwat human servants or disciples were able to enter the Khalwat some time on account of the number of for appears to jinns already inside it, come to attend their spiritual master.61 The similar term khalwat-sard be used to refer to a particular area of the shrine, for HJ-jjiSam asked to be allowed to build his zdw~ya The in the Shaikh's khalwat-sard.62 public khalwat-sard also mentioned.63 No details are given about is at the construction of khalwats Ardabil. The only description of the relative locations of any of these buildings occurs in the story of Safi's death and burial.64 The house in which the Shaikh died, where his tomb was to stand, stood between and was beside the Shaikh's and the Disciples' Cemetery (mazdr-imuriddn), the public khalwat-sard a and khalwat, covered pool (haud-khMna) a garden. A few sentenceslater, it is said that the house had a stood. We window which looked out on to a garden in which the covered pool and the Shaikh'skhalwat know that at the time of Safi's death the tomb of Muhiyy al-Din stood close to the house on the east. must have been on the north side of the IjHjjiSam's zdwiya,which was replaced by the Dar in which stood next house and the khalwat-sara which it stood must be the same as the public khalwat-sard, al-.Huffaz, to the Shaikh's house. It too was thereforeto the north of the house. South of the tomb there is still a cemetery, part of the Shahid-gdh, though now divided from the main part of it by a wall. This was stood must then have been to thereforethe Mazdr-iMuridan.The garden in which the Shaikh'skhalwat the west of the house. It is possible that the Old Chilla-khana,standing as it did to the west of the tomb, of was the same as the Shaikh'skhalwat the Safwatal-&Safd. Even the site of the covered pool perhaps corto the well, now covered over or filled, the stone coping of which still stands in the centre of the responds Area (9) (See IranXII [19741, P- 45, Fig. I). The change of name can be easily explained. The word chillameaning " a period of forty days " is synonymous with the Arabic arba'in,and both may refer to a forty-day period of retreat. In the story of Qasim al-Anwar, quoted above, arba'inis used in this sense and, furthermore,so is khalwat. Thus is chilla-khdna equivalent to khalwatwhen it refers to the place of retreat. In the Safwatal-&Safd, when Zahid was in his khalwat. The same story occurs in a sevenSafi first arrived at Shaikh Zahid's khdnqdh, teenth century work written in a more popular style, where it derives from the $Safwat al-Safd. There Shaikh Zahid is said to be in his chilla-khdna.65 There is no clear information in the Safwatal-Safdabout the locations of the zdw yas. However, in and one story the same building is referredas both dw?ya khalwat.66 Perhapsthen the Old Chilla-khana, and the Shaikh'skhalwat one of the zdwiyas,presumablythe New Zawiya, are all one and the same building. The two different names would then be accounted for by the building's dual function. When it is thought of as a place of assembly and teaching it is a zdwiya,and when it is a place of retreat it tends to be called khalwat. The Old Chilla-khanawas situated opposite the New Chilla-khana on the other side of the passage (23). None of the early accounts give a clear descriptionof what was in this area. De Bruin appears to locate there an elevated apartment with large glazed windows. One of the photographs published by de Morgan (our P1.IV) shows the wall that joined the gateway (29) that leads from the Yard (26) to the inner part of the shrine and the long arcaded north-easternwall of the Yard. It had no door in-it and was built of rather poor brickwork,varied only by two simple pilasters. The photograph is taken from the top of the south-westernwall of the Yard and it can be seen that there was no building behind the wall until the row of rooms behind the west facade of the Area (9). One of the Gulistan Palace photographs (P1.III), taken from the roof of the Dar al-Huffaz (2), shows the back of the wall. Once again
Pp. 255-6. khdnahawa khalwathd-yianja tamdmanaz chab wa means rubbish, but khdshak bashad. In modern Persian khdshak here it probably refers to brush or reeds used as building materials, as would be normal on the shores of the Caspian. 61 P. 195.
60

62 P.

299.

63 P. 308. 64 P. 308.
65

'Alam-drd-yi Shah Isma'il, p. 9.

66 P. 268.

48

JOURNAL

OF PERSIAN

STUDIES

there is no sign of anything that could be the remainsof the Old Chilla-khdna. Sarre'sphotographof the south-easternend of the Yard shows that by his time a of three arches similar to those of the side facade walls of the Yard had been constructed,67 and Sarre'splan shows that two doors led into rooms behind the new fa?ade, as they do today. The way (24) to the Jannat-sard, the kitchen and the Shahid-gThis probably the same as the road to the kitchen and the bakehousewhich passed to one side of the bath (20). In the seventeenth century the same way to the kitchen was used. Della Valle says that the place where food was distributedwas at end of the same Court (9), there is a little door which brings you to a place where the Royal Alms are distributed to the Poor, Morning and Evening; being just against the King's kitchen." The New Chilla-khina (13) is only so called in the 977 waqf-ndma and the al-Milk. Two chronicles also credit Sadr al-Din with the buildings of a Chilla-khdna at the shrine.68 The Sar.h Safawi original structurecan be dated to the period when Sadr al-Din was Shaikh, sc. to 734-791/2. We do not know how extensive Shdh Tahmisp's repairswere, but they seem to have included the building of a new dome. Further restorationswere carried out by Shaikh Abdal Zahidi, who repairedthe eight windows of the building and replaced the plaster which had been in them by making them gul-jdm.69Sarre did not know that the ruins of the Chilla-khina still exist, but he mentioned the door that led to the site. Dibaj was the first to draw attention to the remains of the building and to the fact that it is still known as the Chilla-khana, also being called the Place of Sacrifice, Qurban-gdh.70 The ruins were first described in detail by Weaver7" and the Sarihal-Milk confirmsa number of his deductions. It is apparent from the ruins that the New Chilla-khana was a large, domed chamber surroundedby two storeys of attached chambers. It is clear that it is the Chilla-khana which Olearius mentioned by name, though incorrectly, as belonging to Safi al-Din, for Olearius says that he entered it by a door at the end of the Yard (26), on the right hand. The door is now blocked up on the Yard side but it can be seen from the interior (Plate IIb). Olearius speaksof it as " a very fair and spacious Vault arched above, paved without with green and blew stones .. ." Mr. Weaver has pointed out that a small piece of the glazed brick facing of the Chilla-khdna is still visible at the south-west corner and that this is the remains of the decoration that Olearius describes. One of de Morgan's plates shows the end of the garden court, and it can be seen that the glazed brick facing survived at the time the photograph was taken in places, up to a height of about six metres (our P1. IV). The door is off the photograph. Sarre's photographs do not show the wall nor does he mention it. Probablythe presentouter skin of simple brickworkwas added on to the survivingremainsof the glazed brick during the restorationsof the early 189os. It is not facade that the was part of the original building, for the use of glazed brick to decorate large unlikely facade expanses of exterior walls was common in the fourteenth century. Olearius mentions priestsin the New Chilla-khdnawho chanted in unison. Describing the IHuffdzof the Dar al-Huffdz, he says that their movements were like those of the people in the Chilla-khana. It is probable that the priests were also IHuffdz. The building which de Bruin compares to the Rotunda at Rome is, from its position, also the New Chilla-khdna,but de Bruin entered and left it by the door on the north side through which Olearius only went out. De Bruin's account shows that the building was still intact in 1703. Some time between then and 1758, the dome broke and the finial was removed.72 Since then it has been going to ruin. Excavation of the remains could be expected to throw light on a very unusual type of mediaeval building, of which, as far as the writer knows, there are no intact surviving examples in Iran. The 1759 inventory contains an entry on a silver grille (mi'jar) which stood in the place where Sadr the end of the Yard (26) on the left hand. Tavernier is more precise: " . . . upon the left hand, at the

al-Din used to teach.'73It is not stated where it was and it is perhaps unlikely that it then stood in the

7 The photographwas not publishedby Sarre,but the negative Jahin-ird, p. 261. is preservedin the archivesof the IslamischesMuseum, East 69Silsilat p. II2. On gul-jim see note I8 above. Berlin, Pl. No. 8879. A print is reproduced by Weaver 70 P. 63. al-.Nasab, 71 Preliminary (Second Study, Appendix3). II, Study,pp. 7-9nach Qeji AhmadQumi, 72 Ganjfna-yi 68ErikaGlassen, Die friihenSafawiden Shaikh p. 62. Freiburg-im-Breisgau, 1970, Persian Text, p. 51; Tdrfkh-i 7' Ibid., p. 6o. .Safi,

tA to

Pl. Ib. Inscribed wooden from theJannat-sa capital

Pl. Ia. The surviving on portionof the renewalinscription the eastern pier of the gateway(II) of theArea (9). OliverWatson.

Pl. Ic. The Tard(26) of theArdabilshrine during afirework display of by University from Reproduced permission theCambridge Library,
1647.

in Pl. Ila. Theformerouter of gatewayof theshrine. Froma photograph thepossession Mr. YahyaZoka.

side Pl. IIb. The western of theNew Chilla-khina(13) ; theinterior, showingthe formerentrance from the Yard(26) andthechamber of thecollectors offerings.OliverWatson. of

from the topof thenorthendof theDar al-HIuffd; : in theforeground, of the western (2) part facade of theArea (9) and the twan of t Pl. III. View N. centre, part of the suffa courtyard of the New Chilla-khdna; behind,the Tard (26) and the Main gate (30). Palaceseries. Froma print in thepossession Mr. rahya Zoka. Oneof the Gulistdn of

the wall the Pl. IV. View complex theshrine: in the centre, gateway(29) to the suff of from thetopof thesouth-western of the Tard(26), showing central from left to right, theJannat-sard(io), the twan of theJannat-sard,the Ddr (13) part of theglazed brick facade of theNew Chilla-khdna ; behind, domeof theChind-khdna rising behindit, the tombof ShdhIsmda'land the tombof Shaikh (I). De Morgan,Mission Scientifique en Perse, I, Plate LI. S.aff

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

49

Chilla-khana, where the dome was already broken; but it may have once marked the place where Sadr al-Din sat in his Chilla-khdna, as mentioned in the Sari al-Milk. ParaG. The various referencesto the chambers (22, 25, 27) attached to the Chilla-khanasare difficult to sort out properly. It is hard to see how, if the chambersof the collectorsof offeringsare attached to the Old Chilla-khdna and are opposite to the New Chilla-khana, the upper chambers can simultaneously be attached to the new Chilla-khina. Secondly, where have the forty and more chambers of the collectorsof offeringsbeen previouslymentioned? Are they the chamberswhich stood between the Old Chilla-khina and one of the gateways? Or are they the forty chambers which surrounded the New Chilla-khana? One can only suggest that the author regardedsome or all of the chambersattached to both Chilla-khanas as a single unit because they were used by the collectors of offerings. Olearius, who provides the only other original early information on the matter lends some support to this view, for he says that the collectors of offeringswere " every day in an apartment on the left hand, as a man
... is evidently on the ground floor, and not one of the upper chambersof the New Chilla-khdnawhich the Sarhal-Milk assigns to the collectors. The chamber in which Olearius saw the collectorsstill survivesin the western side of the New Chillakhdna, beside the now blocked doorway leading into the Yard (26). So does the chamber opposite it (P1. IIb). Remains of a staircase in the north-west corner confirm the Sarihal-Milk's statement that there were two storeysof chamberswhich surroundedthe Chilla-khqna. Whether there were really forty of them, or whether the figure of forty includes chambers attached to the Old Chilla-khana,one cannot tell. The custom of presenting offerings provided an important part of the shrine's income. Olearius gives an admirable account of the property and revenues of the shrine. After describingits real estate and the income obtained from certain tax grants he continues: goes into the Metzid Tzillachane ." Olearius is referring to the New Chilla-khana and the apartment

Commissaries every day in an apartmenton the left hand as a man goes into the MetzidTzillachane, are and are set on both sidesof a Chestor Box, cover'dwith crimsonVelvet, into which they put the Money that is broughtthem, as they do alsothat whicharisesby the sale of thoseHorses,Camelsand Asses,which are bestowedon the Sepulchre: for the Oxen and Sheepare kill'd,and distributed amongthe poor. They thosewho bringthem a small Present,which is a handfulof Anniseed; and they are given to undergive stand thereby,that their Souls shall enjoyserenityand blissein the otherWorld. The Persian of Olearius's Nesurtziahn must be nudhairchkydn,more colloquial designation for the a nadhr the of al-Milk. The farman appointing Shaikh Abdal Mutawalli orders the Mushddbitdn-i Sari rifs, i.e. the financial overseers,of the shrine to present the daily account of the offerings (rizndmacha-yi
nudhirdt) to the Mutawalli for him to seal each day.75 Presumably the term Mushrif here means the collectors of offerings as well. Para H. The mention of the two gates (11 and 29) makes it clear that the Suffa (28) was situated in the area of the little courtyard between the two Chilla-kh.nas (12 and 13). Beyond the Sarih al-Milk's account we have no helpful early information about this part of the shrine. De Bruin, at about this point, mentions a library, a chapel and an elevated apartment with large glazed windows, but is certainly confused or mistaken.
u'

sacred place, and they are called ANessurtzchan,74 the word Nesur, which signifies a Vow, ... These from

of and who Religionsendthither, ... not accountingwhat the Tartars Indians, makeprofession the Persian nor the Presents, of which are broughtfrom all parts,in consequence the Vows, which they arewont to make, in greatJourneys,in their Sickness,nay indeed in any businessof Importance,which they very Religiouslyperform. Besidesall these, there are so many Gifts,Donationsand Legaciesmade to it, that therepassesnot a day, but a man shall see going thitherHorses,Asses,Camels,Sheep, Money, and other things. All these things are receive'dby two Persons,who are oblig'd by an Oath to be faithfulto that

In the first edition " ... Nessurtziahn (vom Arab: Nesur so ein Geliibte bedeutet) ..."

"

Silsilat al-NVasab, 1o9. p.

50

JOURNAL

OF PERSIAN

STUDIES

The courtyard today is much the same as it was when Sarre saw it. The of the two longer facades sides are each composed of four arched recessesof different sizes. The third bay from the entrance on each side contains a doorway, that on the north leading to the open area behind the Jannat-sard (I o), which is now part of the school grounds, while that on the south opens directly on to the ruins of the New Chilla-khana. The floors of the other bays are raised above ground level and carved tombstones stand in a number of them. At least one of the tombstones is of the nineteenth century. The earliest useful descriptionof the courtyardis that of Holmes, published in I845.76 He mentions the deep arched recesseson each side of the court as having tombs in them and briefly notes the tiled decoration. When Sarre saw the court it was still largely decorated with mosaic tilework.7 The greater part of the tilework has since disappeared and a programme to restore and replace it only began in 1972. The existence of mosaic tilework of good quality appears to be sufficientevidence that the construction of the court in its present form is to be placed in the Safawi period and so raises the possibility that it is, in fact, the Suffa of the Sarihal-Milk. The identification cannot be certain, but the word sufa, among other meanings, is used to refer to vaults and arched recesses." Possibly, in the Sari1i al-Milk, it refers to the ranges of arches on either side of the court: the tombs would then have been under the shelter of the arches and the Way (23) would be the actual courtyard. Such a hypothesis would solve the problem of where to fit in the Suffa between the two Chilla-khanaswhile still leaving room for the main route to the inner part of the shrine. The gateway at the western end of the Suffa Courtyard evidently correspondsto the gateway (29) leading to the Yard (26) in the reign of Tahmasp, but one doubts if much of the present structure is Safawi. In de Morgan's illustration (P1.IV), while the actual doorway may be old, the brickworkof the remainder, though not brand new, does not give the impressionof being of any great age. The point was first made by Sarre and by the time of his visit the doorway itself had been altered. Its top had been given a more sharply pointed shape and the arch of the gate above it was filled with decorative pseudo-vaulting of brick and plaster which survives today.79 Para L With the Yard (26) we come to a point where great changes have taken place since the Safawi period. To explain them requiresa brief account of part of the wider history of the shrine. During the time of the Safawi Shahs, the shrine, while it may have had its ups and downs, was on the whole extremely well supported. Its income was large; pilgrims and visitors were numerous; in short, it operated on a grand scale. Its favourable position, however, depended to a very great extent on royal support. When the Safawi dynasty fell in the early eighteenth century, the shrine lost its special place and went into a rapid decline from which it has never recovered. The inventory of 1759, in spite of its limited purpose, gives a clear enough picture of the inability of the shrine'sofficials to protect its possessions in the disturbed conditions of the mid-eighteenth century. As has been said, the inventory notes the shrine'slosses in the preceding thirty years. In the casesof the more important items, quite detailed accounts are given of how they were taken.80 For the same period, a few trivial acquisitionsare recorded, but the list of lossesis far greater. Almost every invader, pretenderor passingcommander seems to have stolen or " borrowed " things which he felt he needed or which took his fancy. The shrine was not, during this period, plundered on a large scale, but it was subject to continual minor robberies and extortionswhich the officialswere unable to prevent. It is surprisingthat so much survives today. The shrine'slack of effective protection must have led to similar losses in areas which the inventory does not cover. Presumablyit was at this time that the shrine lost most, if not all, of its lands. For the
present purpose, the important point is that the scope of the shrine's operations was drastically curtailed. With a few exceptions, the buildings in the outer part of the shrine in the Safawi period existed to provide for its administration and its charitable distributions or to supply facilities for people living within the precinct, particularly those who had taken asylum there and could not safely go outside. In its state
Sketches, pp. 36-9. 77 Sarre, p. 8. 78 See 'Abd al-Razz.q Samarqandi, Matla' al-Sa'dain, ed. Mubammad Shafi', Lahore, 1946, PP. 304-5, for the use of suffain this sense in a description of the shrine at GAzurgshand, for an English translation of the passage, Lisa Golombek, The
76

TimuridShrineat Gazur Gah, Toronto, 1969, p. 82. 7* Sarre, p. 7; for Sarre's photograph of the gate, which he did not publish himself, See Weaver, Second Study,Part II, Appendix 3, p. 2. (Islamisches Museum, Berlin, Pl. 8879). 80 Ganjina-viShaikh Saff, pp. 80-6.

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

51

of decline the shrine could not afford its former widespread charity. Its popularity as a refuge and the extent of its administrative requirements were much more limited. The buildings which had been necessaryduring its time of prosperitywere superfluousto its diminished needs. Every one of them had disappeared by the late nineteenth century and probably much earlier. The buildings themselves are likely to have been less solidly constructed than those in the heart of the shrine. Only one Safawl structurecan be precisely located in the outer shrine, the former outer gate, which is much later than the Sarihal-Milk and which has now been demolished (Pl. IIa). The presententrance gateway is relatively late and cannot be proved to stand on the site of the Main Gate (30) of the Sarih al-Milk. Otherwise, there exist rather vague indications in the European accounts, Olearius's possibly inaccurate picture of the Yard, and the information of the Sarih al-Milk and the Silsilat al-Nasab. These data are not in conflict, but they leave many things uncertain. Excavation would appear to be the only way of establishing the layout of the outer shrine more precisely. The Yard (26) certainly correspondsin part to the present first courtyard of the shrine, for which Sarre provides the apt name of Garden Courtyard. Its appearance nowadays owes much to Mirza Nasrulldh Sadr al-Mamalik Ardabili who, as Dibij informsus, built the brick walls divided into arched recesseswhich now enclose the two long sides of the court.81 Dibaj says that the walls were built in the early Qajar period, but it can be argued that they were probably built in the reign of Muhammad Shah (I834-48). Sadr al-Mamalik's family came from Ardabil. He was a Silfi and a Pir. His Sifi title was Nusrat 'All. His Sfifismdid not prevent him from more mundane ambitions and he was one of a number of people of similar attitudes who benefited from MulhammadShah's devotions to the Sufi path. He is chiefly famous nowadays as a perpetually unsuccessfulcandidate for the position of Sadr-i A'zam or chief minister, but during most of Muhammad Shdh's reign he was in charge of the supervision of waqfs and the allowances made by the government to the religiousclasses.82This was the high point of his career, and since it coincided with a time when Siifism enjoyed royal favour, it is likely that he was then in a good position to honour the Sufi saint of the town from which his family came by repairing his shrine. The south-westernwall, which was brokenand collapsing, was completely removed P1. IV shows that in the eighteen-eighties there was a house built on to the long south-western wall, with windows and a doorway opening through the arches of the wall. Probably it was the home of one of the attendants of the shrine. Sarre did not mention the house, and the door and windows have now vanished, perhapsin the rebuilding of the nineteen-forties,but there is still a small extension of the shrine compound in much the same area, entered by a differentdoor through the wall and known from its function as the Hiyat-i Mustarah or Lavatory Courtyard.84 The house on the right hand side of the Garden Courtyardprovides a link with Morier's description of this part of the shrine, which precedes the enclosure of the court by Sadr al-Mamalik.85 Morier does not notice the tiled outer gateway and may not have passed through it. His description begins with a gate correspondingto the present entrance, which, in Morier's words, " leads to the street, composed of a brick wall on the left side, and of the habitations of the priests attached to the foundation on the right." Morier'suse of the word " street " certainly means that there was then no garden at this point, and probably implies that there were buildings on part of what later became the Garden Courtyard. The Yard of the Safawi period differed again in that it was lined on both sides by the buyatdt or shrine offices (which are discussedseparately below). The shape of the actual courtyard did however,
to some extent, resemble the Garden Courtyard of the late nineteenth century: Della Valle describes it as being longer than it was wide; Olearius as being at least as long as the Outer Court (corresponding to the Space (58) of the Sari4 al-Milk). Olearius's picture of the Yard (P1. Ic) shows the length of the court as several times its breadth. As well as the Main Gate (30), the picture shows, rather indistinctly, the buytatt along the south-western side of the Yard, which according to our reconstruction are the
81 83

and then rebuilt in 1312-3/I943-4-83

Dibaj, p. 62. The title is incorrectly written Sadr al-Milik. On Sadr al-Mam~lik see Muhammad Ma'sfim Shirizi, Tard'iq al-Haqd'iq, ed. Mul.hammadJa'far Malijlb, Tehran, I339-45, III, PP. 240-2; Mahdi BtmdAd, Shariz-iIfal-i Rijdl-i

Irdn, Tehran, I347, IV, pp. 336-788 Dibdj, p. 62. 84 See the plans of Diba-j and the National University Survey. 86First Journey,p. 253.

52

JOURNAL

OF PERSIAN

STUDIES

Hawij-khana (46), the Daftar-khdna (42) and, perhaps, the Sharbat-khana (37). They appear as a row of attached two- and three-storey buildings and are probably not very accurately represented. Sarre believed that the Garden Courtyard was the site of both of the first two courtyards which Europeanvisitorsdescribedas existing in the Safawi period, and also that the gate of the Garden Courtyard was the one by which they entered the shrine. The reasonfor his belief appearsto be that when he visited Ardabil, the Outer Court was known as the Maidan or Square and was surrounded by shops, which led him to identify it with the Maiddn of the Safawi period.86 Thus he had to fit in two courtyards after the Outer Courtyard. For several reasons, the identification of the two Maidans must be wrong. Without exception the European travellers of the Safawi period place the Maidan outside the shrine, while the existence of the former outer gate, which was built by the Mutawalli and cannot be anything but an entrance to the shrine, shows that Sarre's Maidan was inside the shrine. The SafawI Maiden led to the bazaar and was surroundedby shops. Olearius'sestimate of its size, three hundred by one hundred and fifty paces, makes it far larger than the Outer Courtyard. Olearius'sview of Ardabil shows the Maidan outside the shrine and the Silsilat al-Nasab'saccount of the enclosing of the Space (58) to form the Outer Courtyard speaks of the Maidan as being beyond the Outer Courtyard.87 There seems to be no real objection to assuming that, by and large, the Outer Courtyardwhich Sarre saw correspondsto the Outer Courtyardformed by Shaikh Abddl Zahidi and the Garden Courtyard to the Yard. Olearius states that the Yard, or in his terminology, the Great Court, was added to the shrine by Shaikh Junaid. Since Junaid spent very little of his adult life at Ardabil and was probably never in undisputed control of the shrine, one doubts it, but perhaps the shrine was extended in this direction Della Valle saw the Yard when it was being paved at ShTh'Abbas'sorders. Olearius and Tavernier also mention the stone paving. The Spring (31) is mentioned elsewherein the Sarhal-Milk among the did propertiesof the shrine for which waqf-ndmas not exist.89 Della Valle talks of a canal at the sides of the Yard, used for washing. Tavernier calls it a rivulet and puts it in the middle of the Yard. Olearius describes the water supply as a " fair Fountain ", the water of which was brought from a league outside the city and entered the Yard through a " brass-Cock". Struys transmutes the brass into silver, but may be right in placing it next to the New Chilla-khana (i3). In such a position it would be well placed to supply the Sharbat-khana (37), the Bath (20o)and the Kitchen (21). As for the Shop (32), there does not appear to be any furthermention of it. It seems most unlikely that it is the butcher'sshop of Para. Yg. Olearius, however, does speak of shops in the Yard. Para.J. There is a wealth of information about the distributionof food at the shrine, but here the institution will be treated very briefly. From the beginning, the shrine appears to have had extensive alcatering arrangements. The cook and the baker are incidentally mentioned in the 3Safwat Safd.90 Shaikh Safi himself was a strong supporter of the practice of providing hospitality for the dervishes. He is quoted as saying, in disapprovalof the over-lavishdecorationsof HaIjji Sam's zdwiya,that what was needed was a well-supplied table, not a painted house.91 Rashid al-Din's gift of food to Shaikh Safi demonstratesthe lavishnessthat was practisedon occasions. If the gift was seriouslyintended for a single feast, there would have to have been several thousand guests.92 Between Shaikh Safi and Shah Ismd'il we know nothing directly about the shrine kitchen, but no
doubt the distributions continued. Shah Ism-'il I showed his approval of the custom by his foundation of the trust to provide for it known as the Ash-i Halal. For the seventeenth century, the European travellers have left good descriptions. Olearius gives the number of those fed free as over a thousand daily; others, ashamed to accept charity, bought their food from the kitchen. The Silsilat al-Nasab preserves a detailed list of the quantities of foodstuffs consumed in the charitable distributions (shildn-i khairdti).93
86 Sarre, pp. 6-7. 87 Silsilat al-Nasab, p. i 12. 88 Irans

during the time when Junaid claimed to be Shaikh of the Safawl order, ca. I448-6o0.88

9x

Hinz,

Aufstieg,p. 146.

89 F.I5oa.

al- $qaf, p. 3o0, better put in the words attributed to the Shaikh: mdrdbdyadkik khwdnrangfnbdslhad khdna. na S.afwat " Mukdtabdt-i Rashfdf,pp. 27o-I. 9 P. 112.

90 Pp. 207, 280.

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

53

According to the Tdrikh-i'Abbdsi,Shah 'Abbas built a kitchen at the shrine in Io4/1605-6.94 Perhaps,as in the case of the Chini-khana, 'Abbas only refurbishedan existing building, for the locations of the kitchen of the al-Milk and that existing in the seventeenth century are certainly similar. Sari The Sarihal-Milk describes the Bath (20o) and the Kitchen (21) as bounding the western side of the Kitchen and the Shahid-gdh, went out of the Yard between the Old Chilla-khana (12) and the Bath, that is, somewhere at the eastern corner of the Yard. (Paras. Fa, W) Khwaja Khan Wahid's houses, which stood on the site of the Jannat-sard,were bounded on one side by the Yard ('arsa)of the Kitchen. (Para. Ya). Thus the kitchen must have stood somewherebehind the Bath, separated from the Jannatsard by its own courtyard. Seventeenth-century indications of the location of the Kitchen correspond with ealier ones very closely. Della Valle says that food was distributed at a place in front of the Kitchen at the end of the left hand side of the second court, that is the Yard (26). Tavernier's account is even closer to the Sarihal-Milk; ... and upon the left hand, at the end of the sameCourt(the Yard),thereis a little Doorwhichbringsyou to a placewherethe Royal Almsare distributed the Poor,Morningand Evening; beingjust againstthe to King's Kitchen. The foundationsof the Kitchen and its dependenciesmust be among those which came to light when the western extension of the Shahid-gah (15) was cleared to make way for the school.95 Details concerning the kitchen building itself are sparse. Tavernier speaks of " ... about thirty Ovens contriv'd in the Wall, with as many great Cauldrons to dress Pilaw and other Food,..." Olearius also mentions the cauldrons, adding the detail that they were supplied with water by a system of piping. De Bruin too mentions the ovens, and the cauldrons set on top of them. The ovens must either be the dig-khdnas the Sarihal-Milk or their replacements. Otherwise, de Bruin mentions two of wells in the Kitchen and says that the wall of the Kitchen was quite high. One other feature which all the Europeans except Della Valle mention is the pair of silver-plated doors belonging to the Kitchen. References in the Tdrikh-i to 'Abbdsi Shah 'Abbas I's gift of the doors have been quoted above."9 It seems, particularly from Tavernier's account, that the doors were not those of the actual kitchen building but that they hung in the gate leading from the Yard (26) to the Kitchen. Interestingly,the doors are not listed in the 1759 inventory, either as existing or as having been lost in the preceding thirty years. Probably the large-scale distribution of food had to be discontinued soon after the fall of the Safawis. The doors may have been casualties of the Ottoman occupation of Ardabil in the years I725-30o.97 There is still a kitchen at the shrine, housed in the most northerly of the chambers behind the western fagade of the Area (9), next to the Jannat-sard ( o). It is not known when it was made into a kitchen. It is a far smaller affair than the seventeenth-centurykitchen and is now unused.98 Para. L. The Sharbat-khana(37) stood on the south western side of the Yard (26) beside the Daftarkhana (42), with the spring (31) on the other side, between it and the New Chilla-khana (13). When Shah 'Abbas was at the shrine in Io21/1612, he visited the Sharbat-khanaof Safi (sharbat-khdna-yi Qutb
al-'Arifin)"9 The wording may imply that the building which existed at that time was actually built in Safi's lifetime, but buildings at the shrine erected after his death could well have been spoken of as belonging to him also. The Sari al-Milk provides the only information about the appearance of the Sharbat-khana but the production of sherbet and confectioneries at the shrine are recorded from the beginning. Safi was accustomed to break his fast during Ramatadn with sherbet.100 Among the supplies
94F.14o. 95 Dib-j, p. 83. 96 Part I, Commentary on Paras A and B. 97 On the occupation, see Laurence Lockhart, Nddir Shah, London, 1938, pp. 46-51; idem, The Fall of the Safavi Dynasty, Cambridge, 1958, p. 266; R. L. Tapper, The Shdhsavanof (Ph. D. thesis, University of London, 1971), pp. Azarbdajan 432-6. 98For a description, see Weaver, Preliminary Study,pp. 14-599 Tdrikh-i'Abbds[,f. 197a. 100 Safwat al- afd, p. 303.

Shahid-gah (15;

Para E). The way or road (24) from the Yard (26) to the Jannat-sara (io), the

54

JOURNAL

OF PERSIAN

STUDIES

given to Safi by the vizier Rashid al-Din Fadllullahto make a feast were large quantities of rose-water and sugar, to be delivered by the Royal Sharbat-khana at Tabriz.xox A more precious commodity, of fifty mithqdls ambergris,intended to be used in sweets, was to be supplied by the treasury. Among the Olearius provides a glimpse of the operations of the Sharbat-khana, mentioning that the Europeans, food provided for the ambassadors of Holstein from the shrine kitchens on one occasion included " Conserves,in nine great Vessels of Porcelain."'02 No sherbetsor sweetmeatsoccur in the descriptions of the food distributed ordinarily. Possibly, they were only provided for important visitors and on special occasions, such as the feast for which Rashid al-Din's gift was intended. In the seventeenth 103 century, the rose-waterused in the Sharbat-khanawas supplied from the garden of the Madrasa (65). When ShTh 'Abbas touched the lock on the door of Safi's Sharbat-khana, it immediately opened. The same miracle occurred at the Privy Sharbat-khina (sharbat-khdna-yi khassa)which Shah 'Abbis had given to the shrine as waqf, willing the benefit for the pious act to his mother. What special purpose the Privy Sharbat-khanaserved is not explained. Para M. The Daftar-khAna(42), between the Sharbat-khana and the IHawij-khana(46) was, one can assume, where the scribes worked and the records and archives were kept. It would be the place where the property registerswere compiled. During the 'Ashiird'ceremonieswhich Olearius witnessed in the Yard, the banner supposedly made by Muhammad's daughter, Fatima, and brought by Sadr al-Din from Medina, was planted " near the Chancery ", that is, the Daftar-khana. The standard shown in the left foregroundof Olearius'spicture of the Yard is, no doubt, the one in question. (P1.Ic). Para N. The Hawij-khana (46), the last of the offices on the south western side of the Yard, was primarily a store for food, as its name implies.'04 The Silsilatal-Nasabreports that after Shaikh Abdal Zahidi had put the finances of the shrine on a sound basis there was always the wherewithal for a year's distribution of food in the Hawij-khana.x05Even if this is an exaggeration, the scale of the catering at the shrine must have demanded a considerable amount of storage space. Tavernier notes the presence of " Granariesfor Rice and Corn " on the other side of the Yard from the Baths (2o). The details of the properties added to the stores in the reign of Tahmasp are given in Para. Yf. Para 0. The Naqqtra-khana (49). The playing of kettle-drums and assorted wind instruments at or is certain times of day, known as naqqdra naubat, usually a symbol of royal or princely power in Iran, as in other parts of the Islamic world. Its use at shrines, in recognition of their spiritual power, is not At unknown; the shrine of the Imam Ridt at Mashhad still has a functioning naqqdra-khdna.?06 Ardabil the privilege was early acquired; Sadr al-Din relates that he was once being given instructions He by his father on how to spend the future income of the khdnqdh. showed surpriseat the scale of expenwould be played to sumditure envisaged and was informed that the day would come when the naubat ba naubat mon the dervishesto meals. (hangdm-i zanandta darwishdn sufradyand) And, indeed, after sufra ten or twelve years, Shaikh IHasanChilbani sent the shrine a drum and flag and begged that the naubat be played at Safi's gateway. The custom was adopted and persistedfor a long while but, it is implied, then died out.'07 The marking of meal times with the playing of the naqqdra appearsto be exceptional. it was played at set times of day without any connection with meals, and so it is at At the Qajatr court, Mashhad today. The Ardabil tradition confirms the importance that was placed on the provision of food at the shrine. The Sarihal-Milk shows that the naubat must at some time have been revived. When 'Abbas visited
the shrine in I021/1612 it was still being played, but only at soup time, presumably meaning only once a day. This was due to the laxity of the musicians. To mark the miracles that occurred during his visit,
Rashidi, p. 272. 1ol Mukdtabdt-i
104

102During the 'Ashfri' celebrations and firework show which the ambassadors attended in the shrine, " several Vessels of Porcelain, with Suger'd and Perfum'd waters " were placed on a cloth before them. Olearius seems to imply that the Ambassadors were being entertained by the Governor and not the shrine on this occasion, but the Porcelain vesselswhich appear in Olearius's illustration of the scene (P1. Ic) are likely to be from the shrine collection.

106
107

plural of bdjja, HIaw'jis an abbreviated form of Arabic hawd'Uj, are meaning need or necessity. Iawd'ij and bawy" used in a restricted sense to mean materials necessary for cooking, aux stores of food. See Dozy, Supplement dictionnaires arabes,s.v.; Farhang-i Mu'fn; Farhang-i Anandrdj; Tddddshthd-yi Qazwinf, IV, Tehran, 1337. P. 106 III.
Rdhnamd ..., pp. 31 1-2.

Silsilatal-Nasab, 113p. 1Os

744/1343),see EI' art. " 6(ibanids " (R. Savory).

S.afwat al-Safd, p. 242.

On Shaikh Ijasan Chiibani (d.

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMXSP

55

the Shah ordered that it be played according to previous custom, and " This was a cause of joy and happiness to all the people of Ardabil, male and female. " The instruments mentioned here are kettle-drums," hautboys ", trumpets and horns.108Olearius speaks of the playing of two " Timbrels ", said to have been brought from Medina by Sadr al-Din with the banner of Fatima. They were played when food was distributed, at that period, three times a day. Neither Tavernier nor de Bruin mention the naqqdra. The descriptionof the location of the Bath (Para W) establishesthat the Naqqara-khanastood next to the Bath (20) on the north-easternside of the Yard. Since the furnace of the bath was behind it, it must have been closer to the Yard than to the Kitchen, in spite of the fact that the playing of the naqqdra were announced meals. To enable the sound of the music to travel further, naqqdra-khdnas located in in or on other structures; they do not appear to have existed as separate buildings.109 elevated positions The one at Ardabil conformsto the pattern, standing on top of the cistern. Possiblyit was to some extent like the one at Mashhad, which has a roof supported by pillars and open sides. If so, the sound of the band would easily have reached the kitchen. The last sentence of the paragraph perhaps makes best sense if taken as referringto the way to the woodstore (53)Para P. The Woodstore (53) cannot be precisely located, but stood conveniently near the Bath (20o) and Kitchen (21), where fuel would be most needed. Paras Q, R andS. It is convenient to treat the Main Gate (30) and the chambersto its right and left (54, 55) together, since the gate at this point in the shrine, from the time we know anything about it, was set in a substantial building and had chamberson either side and, at times, above it. The Sarihal-Milk is uninformative about the appearance of the gate, but the fact that the interior of the gateway and the passageway, presumably the passage in which the actual gate was set, are considered worth mentioning seems to show that in the sixteenth century too the gateway was a tunnel-like one passing through a building. Della Valle describes the gateway as having a number of apartments and open balconies built over it in which people who had taken refuge at the shrine and some of the officials lived. Tavernier's account is very similar. He says that the door which led to the Yard (26) " brings you presently under a Portico,where there are fair Balconies rais'd after the fashion of the Country. Those Balconies are full of several People; either Pilgrims,or personswhose Crimesconstrain them thither for Sanctuary." Both Tavernier and Olearius mention that at this gate the visitor had to surrenderany arms he was carrying. Olearius's engraving (P1. Ic) shows the gate as having two rows of windows above the large archway of the entrance. On either side of the building which encloses the gate and occupying the rest of the end of the Area are ranges of buildings, lower than the gate and uniform in height. A single row of windows can be seen in the upper part of these buildings; the lower part is not visible. The gateway of the end of the nineteenth century had no upper story. It consisted of a vaulted passageway, the outer and inner fagadesof which were formed by simple brickpishtdqs(Pls. IIa, IV). The gate itself was set into a smaller arch in the centre of the passageway. The building through which the entrance ran also had one storey. Sarre's plan shows three chambers to the left of the gate and one to the right. The present gate building once again has an upper storey, both above the actual gate and above the side chambers. In 1305 A.S.H./1926 Sayyid 'Abd al-Rahim Khalkali constructed a building in order to provide the shrine with some income. His description is rather vague.'1 He says only that he built five shops and an upper building ('imirat-ifauqdni) which were leased for four hundred and four tamdns. The gate building today has a ground floor containing, besides the gate, five shops which face the open space in front of the shrine. Its upper floor contains a set of rooms which serve as the officesof the Waqf Organisation. There is no other building at the shrine which answersto Khalkhali's description,which has to belong to the present gateway.

108Tdarkh-i f. 'Abbdsf, I96b. 109 of Naqqdra-khana,course,oftenreferssimplyto the band with-

out any implication of the existence of building to house it. 110" Buq'a-yi Shaikh Safi al-Din ", p. 400.

56

JOURNAL

OF PERSIAN

STUDIES

How much, if any, of the previous building was incorporatedinto the new structure,it is difficult to say. At least we know it was on the same site. With the earliergate or gates, even that is uncertain. The problem of the definition of the limits of the Safawi Yard (26) has already been discussed. The evidence of Olearius's picture would seem to show that the Main Gate was on a different site, for it places it in the centre of the north-easternend of the Yard. The present gate is not in the centre, but is only separated by a single chamber from the north corner of the Garden Courtyard. But, of course, the picture could be inaccurate. There does not appear to be any other mention of tombs in this part of the shrine. Para T. In the reign of Tahmdsp there was an area in front of the shrine which was evidently considered as being in some way part of the shrine, while remaining outside the real enclosure. The Silsilatal-Nasabgives an account of how it was brought fully within the shrine; this gives some information about its previous layout and deserves to be quoted in full:"' The area in frontof the door (simplydarfin the text, but obviouslycorresponding the Space of the to of the blessedshrine: he (ShaikhAbddl ZAhidi)made the river frontageinto twenty-one Sarihal-Milk) shopswith his own money. The Maidan side, which had not been built with arches,he made into shops with an archedfrontage. Sincethe foursidesof the areahad been built up, he erectedfour gateways,hung doorsin them and made it secure. Also he providedthe lane which leads to the door of the Sultanof the Saints,Sultan Haidar,with an arch and a gateway. The enclosure of the Space (58) to form the Outer Courtyardof the shrine took place after Shaikh Abddl's appointment as Mutawalli in Rabi' I Ioo9/Nov.-Dec. I6oo. It was complete by 1619, when Della Valle described it as a large courtyard with a great number of shops around it, selling food and clothing to pilgrims and the inhabitants of the shrine. Olearius mentions that the court was paved. The suffa(64) mentioned here is probably to be imagined as a raised platform of masonryor brick with the pool in its centre.112There is no other referenceto it and it seems that it had been removed by the time the courtyard was formed. One of the Safawi gates of the courtyard survived into the twentieth century. It was an imposing arched portal, decorated with tiling and with elaborate stalactite work in the vault above the door (P1.IIa). The inscription, dated 1057/1647-8, records that it was built in the reign of ShTh 'Abbas II by the Mutawalli, Nazar 'Ali Khin.l13 In the late nineteenth century, the portal was still almost complete. A series of photographs illustrate its decay,114and in 132I/I942 the Archaeological Service removed what remained. Only the Silsilatal-Nasabmentions that the court had more than one gate. At that time it appears that the Court had six gates, if we include the already existing Main Gate (30) leading to the Yard (26) and if the gate leading to Sultan Haidar's house stood on the Outer Courtyard. On Sarre'splan, three lanes lead into the Outer Courtyard which could also be entered by Nazar 'Ali Khan's gateway. In most cases it is difficult to be quite certain which gate the European travellersentered by. It does seem possible that there was one more important ceremonial gate through which the visitor normally entered and that the other entrances were less grand. Tavernier places the gate by which he entered on the south side of the Maiddn. Nazar 'Ali Khan's gate answers roughly to this description, facing N.N.E. Della Valle's entrance gate was located in an alley near the Maidan and is presumablythe predecessor of Nazar 'Ali Khin's gate.'15 Olearius and Tavernier also made their entrancesnear the Maidin. Thus
possibly they all entered at the same point. De Bruin describes the gate as large, vaulted in the upper part and decorated with coloured tiles. All these details correspond with those of Nazar 'Ali Khan's gate. Most of the other features mentioned in the Sarnh al-Milk and the Silsilat al-Nasab cannot be precisely located. The Maidin side of the courtyard, which Shaikh Abdal made into shops, must be the north
x"l P. I112.
112

Lisa Golombek, op. cit., p. 85; the inscription on the platform in which the tombs of various members of Sultan Ilusain, the Timurid ruler's family, stand at Gazurgrh refersto the platform as suffa. 11aFor the inscription, see Sarre, p. 21; Dibaj, pp. 61-2.

11x

I; Najafi, p. 313. PI. IIa; Sarre, xl5 There is some information in marginal additions to the

Tafel

Sarithal-Milk about alterations made to the Maidan itself which may explain the discrepancies between Tavernier and Della Valle at this point.

THE

ARDABIL

SHRINE

IN THE

REIGN

OF SHAH

TAHMASP

57

al-Milk places side and presumablystood in much the same place as the lane and shops which the Sari the Main Gate (30). Olearius puts a garden, presumably the Garden of the Madrasa (63) opposite somewhere " backwards" in the courtyard. As for the River (61), in the seventeenth century the Balikhlii River, which now passes through Ardabil in a single channel, forked into two channels which rejoined each other after passing through the town. Many minor canals were led off the river through the streets and gardens. Della Valle goes so far as to compare Ardabil and Venice, implausible as this sounds to one who visits Ardabil today. Probably the Sarihal-Milk's river is one of the lesser canals. The house of Sultan IHaidar referredto elsewhere. Shah Isma'il stayed in the houses of his father is when he came to Ardabil in the course of the rebellion that brought him to the throne.n6 So did Tahmasp on a visit to the shrine."'7 The Sarikal-Milk mentions the Garden of Sultan Haidar, which may be the same thing."s De Bruin, after visiting Safi's shrine, went to visit the royal gardens which were near the Maidan. In them stood a stone building, falling into ruins. According to be Bruin, Shaikh Safi once lived there. The gardens are mentioned by Tavernier, but not the building. There is no other source that states that Shaikh Safi lived anywhere but in what later became the central part of the shrine enclosure. Possibly de Bruin is mistaken and the ruined building in the garden was really Sultan Haidar's house. Some time after the demolition of Nazar 'Ali Khan's gate, road improvementsbrought about a new change. Part of the Outer Courtyard was used for one of the modern avenues of the town. A kind of roundabout occupies most of the remaining area. The entrance to the shrine alone survives, although it is possible that the relatively recent buildings to the northeast and southwest may preservethe memory of the older layout. Para U. In the manuscriptof the Sarihal-Milk, the entriesfor the Madrasa and the enclosure beside it are each given half a page, most of which is blank. One receives the impressionthat the original intention was to describe them in greater detail. As it is, we have little information. In 1021/1612 the Madrasa (65) was in a ruinous state. Shah 'Abbas ordered it to be rebuilt.119 Here it is called the Madrasa 'Abbdsi-and of the Shah whose place is in Paradise (Shdh-i In Jannat-makdni). the usage of the Tdrikh-i in many other works-this means Tahmdsp, which makes it all the more likely that the author of the Sarihal-Milk intended to devote more attention to it. One can imagine that the Madrasatook the place of Shah Isma'il's Dar al-Hadith as the centre for religious instruction at Ardabil. It no longer survives. Para V. The Sarih al-Milk is normally reasonably consistent in its terminology and perhaps the Enclosure (66) next to the Madrasa (65) should be distinguishedfrom the Garden of the Madrasa (63) mentioned in Para T. On the other hand, they may be the same. Shaikh Abdal Zahidi found the garden ruinous and its enclosurewall completely disappeared. He built a wall round it and turned it into a rose garden. On the side towards the doorway of the shrine he built an openworkwall (mihjar)so that the garden was visible. It became a favourite spot for the people of the city to visit, and it also supplied
rose-water for the Sharbat-khana (37).120

Olearius noticed the garden in the outer courtyard and described it as " very fair ". He noted that it was open to all. Struys also describeswhat must be the same garden in glowing terms. It was open to everyone, provided they did not pick the flowers or fruit. The penalties for damaging the garden
were severe: " . ... a drunken Persian had hewn down a great bough with his Scymiter, but being taken

was immediately beheaded with the same Weapon." Struys is, however, confused about the location of the garden which he places inside the Yard (26).
Para W. As has been previously stated, the Bath (20) is not described in the main section on the shrine,.but with the other baths owned by the shrine in Ardabil. It stood on the north-eastern side of the Yard (26) next to the Naqqara-khana (49). The Furnace of the Bath (51) was actually behind the Naqqara-khana. The Bath was separated from the Old Chilla-khana (I2) by the Way (24) to the Kitchen (21I), the Jannat-sara (I o) and the Shahid-gah (15), which thus went out of the east corner of the Yard (26). The foundations of the Bath must have been among those which came to light when the
" 6s1E. Denison Ross, The Early Years of Shah Isma'il ", JRAS (1896), pp. 325, 339. x17 Tadhkira,p. 38. 6
118

F. 32b. 119Tarikh-i 'Abbdsf,f. 197b. 120o Silsilat al-Nasab, pp. I12-3.

58

JOURNAL

OF PERSIAN

STUDIES

area was cleared to make way for the school.121 The Bath was originally made over to the shrine by Shaikh Safi himself; possibly he also built it. It must be the same as the Old Bath mentioned in Para E, since there was no other bath within the shrine. Presumablyit was called " old " to distinguishit from other baths in the town. Tavernier mentions the Bath as standing on the other side of the Yard from the Granaries,that is the (46) and store (47). The existence of the bath underlines the fact that the shrine provided all the necessaryfacilities for .Hawij-khdna people who wished to live wholly inside it, although the apparent absence of a mosque seems to mean that their religious needs were not wholly catered for. No doubt the baths and the other services available were used by the shrine officials, but they were really essential for the other important class of residents,those who had taken bastor asylum at the shrine and were only safe so long as they did not go outside the compound.

121

Dibaj, p. 83.

THE MASJID-I 'ALI, QUHRUD:

AN ARCHITECTURAL AND EPIGRAPHIC SURVEY*

By Oliver Watson
Quhrfadis a small village on the summer caravan route between Kashan and Isfahan. It lies at a distance of 45 km. from Kashan, towards the head of the pass through the Quhriid mountains from which it takes its name.' During the winter months it is frequently cut off by snow, and the longer route round the mountains is now followed by the modern highway. Many European travellerspassed through Quhrfid on their way to Isfahan, and the pleasant climate and picturesque houses straggling up the steep hillsides usually elicit a few words of praise.2 From external appearances the village seems to have little to commend it in the way of" cultural remains ", the Safavid caravanseraibeing the only substantial building. However, two small mosques, interesting in themselves, contain remarkable features that deserve to be better known.3 They both exhibit a style of building that seems to be typical of the region. The larger contains a considerable number of Kashan lustre tiles which provide much interesting information, and a carved wooden door of superb quality.4 The smaller mosque contains a ceramic plaque signed and dated by one of Kashan's most famous ceramicists. The Masjid-i'Ali The Masjid-i 'Ali is the larger of the two mosques to be discussed. It is now used as a Friday mosque by the inhabitants of the village. Its ground plan is square (Fig. i). It is roofed with nine flattish domes (Fig. 2) supportedon a system of cross arches of roughly keel-arch shape. These arches are supported in the centre of the building by four pillars, and by piers projectingfrom the walls. These piers also support at a height of approximately
* I should like to take this opportunity to express my indebtedness to the following persons and institutions: to the British Institute of Persian Studies, which awarded me a Fellowship for this and other work in Iran, and to the Central Research Fund of London University which provided funds for travel to Persia and to Quhrfid; to Dr. Baqerzadeh and the officers of the then Archaeological Service of the Ministry of Art and Culture in Iran who made this study possible; to C. Pancheri, the architect who accompanied me to Quhriid and made the measured drawings; to R. Hillenbrand and R. Pinder-Wilson who read the finished manuscript and made many valuable suggestions; to Professor A. K. S. Lambton, Dr. Gandjei and M. Bayini for their help with the inscriptions; and especially to my supervisor Dr. G. Feh6rvari, who relinquished prior claim to the mosque and whose constant guidance and enthusiasm have been of great help.
1

M. Siroux, " Anciennes Voies et Monuments Routiers de la Hunar u Mardum (May 1972). RWgiond'Ispahan ", Mimoires de l'InstitutFranCais d'Arche'ologie 3 The monuments, apart from the caravanserai, have received Orientaledu CaireLXXXII (I97I). Carte II shows a detailed no detailed attention in European writings. Wilber, The Architecture Islamic Iran, The Il-Khanid Period (Princeton of map of the Kashan to Isfahan caravan routes. On pp. 2o-I he gives a description of the route via Quhriid (Kouhroud), and 1955), P. 129, refers to the Masjid-i 'Ali on the authority of A. Godard, but did not visit the site. He mentions neither the on p. 142 he gives a description and plan of the Safavid door nor the tiles, and gives a garbled version of part of the caravanserai at Quhrfid. in his Athdr-i historical inscription. In Persian, H. 2Chardin says of the village, which he calls " Carou ", " On ne Kdshdn wa Nartqi(Tehran I348), Tdrikhi-yi Shahristdnhd-yi peut trouver un plus charmant et agr6able endroit dans le tems Na.tanz chaud," Voyagesen Perse (Amsterdam 17I1), I, p. 216. He pp. 347-51, gives a brief description of the village and the spent the night in the caravanserai. Other later travellers Masjid-i 'All. He publishes the historical inscription on the included J. Dieulafoy, La Perse, La Chaldieet la Susiane (Paris door, but fails to read the signature of the artist. numbers of lustre 1887), p. 212; Curzon, Persia and the Persian Question II 4 To my knowledge, the only sites where large tiles are preserved in situ are the shrines at Meshhed (D. M. (London 1892), p. i8; and Browne, A YearAmongthe Persians Donaldson, " Significant Mihrabs in the Haram at Meshed ", (London 1893), p. 186. The Comte de Sacy who passed Ars Islamica II (1935), pp. 120-27) and at Qumm (Nardqi, through in 1839-40 is alone unenthusiastic, and describes op. cit., pp. 364-5). Both these places are difficult of access for Quhr-id as ".. . un mis6rable bourg ", La Perse (Paris x928), European and Persian scholar alike. Woodwork is sadly p. 229. A lithograph by Flandin (Voyagesen Perse, Paris 1851) and a photograph of the village are published by Nariqi in neglected in the study of Persian minor arts.

59

60

JOURNAL OF PERSIAN STUDIES

new

mosque

new porch

qibla

,2

,3

5M.

Plan, Masjid-i 'Alf. Fig. i. Ground

,1

,2

,3

5 M.

Fig. 2. Elevation,Masjid-i 'Al.

THE

MASJID-I

'ALI,

QUHRIJD:

ARCHITECTURAL

AND

EPIGRAPHIC

SURVEY

61

1*75 m. a gallery which runs round all four walls (P1. VIIb). The supporting piers are pierced at the level of the gallery to afford uninterrupted access, which is now only broken by the new door in the south-westcorner of the building, leading to the new mosque. The central pillars are octagonal, tapering slightly towards the top before spreading to a square from which the arches spring. Unshaped wooden beams are set between the pillars and the piers at the springing of the arches.6 Apparently the only original entrance to the building was the door in the middle of the west wall.6 Access to the gallery was by a separate outside doorway in the east end of the south wall where the slope of the land raises the ground level to that of the gallery. This entrance has now been blocked and a door into the main part of the mosque and a stairway up to the gallery have recently been constructedin the north-east corner of the building. Windows are pierced between the piers in the north wall on the ground floor and in both north and south walls at gallery level. These appear to have been enlarged in modern times. The inside of the mosque is heavily plastered and whitewashed; decoration is found only round the and mihrab door. The mihrab comprisesa deep rectangularniche flanked by two smaller niches (P1.Ia). The various panels of this mihrab covered with tiles which are discussedbelow. The side niches have are decoration, and in the head of the central niche is found plaster moulding. An inscription muqarnas (dated 1317/1900) runs over the top'of the mihrab.The door is recessedin the west wall; on the exterior at either side are set panels of tiles above stone benches. Above the panels further plaster inscriptions are again dated 1317/1900. Above the entrance arch are more tiles in the spandrels. This entrance now opens into the porch of the new mosque, which runs the length of the west wall. Local memory helps to elucidate some of the mosque's recent history.7 Apparently the new mosque attached to the south-west corner of the old was erected on the foundations of a zamistdn(said to be 200 years old) which collapsed some 14 years ago. The new doors in the south-west and north-east corners of the old mosque were installed then. There had been originally a courtyard in front of the west entrance and the door and tiles were open to the elements. If this is true, one can only be astonished at the excellent state of preservationof both door and tiles. It is difficult to determine the material used in the interior constructionof the mosque owing to the heavy plaster overlay. On the outside the south wall is of fired brick (20 x 4 cm.). The east wall is of mud brick and the north wall has fired brick for the projections (correspondingto the engaged piers on the inside), but the parts in between are apparently of mud brick. The whole is set on a foundation of stone rubble, the height of which varies and is difficult to assess,as the ground rises steeply towards the south.

The Masjid-iKalah It is appropriateto discuss this smaller mosque here since it is of exactly the same construction and elements as the Masjid-i 'Ali. It has an irregular ground plan, necessitated perhaps by the steep slope of the land downwards towards the east, or by buildings and roads that surroundedit (Fig. 3).8 The
5

Many other buildings in the Jibil area exhibit the same basic characteristics, sc. covered buildings, with vaults supported by pillars and piers, and some are of considerable age. Siroux (" L'Evolution des Antiques Mosqu6es Rurales d'Ispahan ", Ars AsiatiquesXXVI (1973), P. 23 No. 2 and p. 81, and figs 17 and 19) dates the mosque at Shapurabad to the 9th century A.D., and that of Eskarand to the Ioth century. Covered, pillared mosques are being built to this day in the Kashan area, though now with steel pillars and concrete (the Masjid-i Ftima and the Masjid-i Warqada in NIshabhd to the north of Kashan were built within the last fifty years and show these features). The Masjid-i QAldiin Baidgul, also to the north of Kashan and probably of Safavid date, is a rectangular vaulted hall with cross arches supported by pillars identical in type to those of the mosque in Quhriad. The square ground plan is more difficult to match. The Masjid-i Kucha-Mir in Natanz approaches it (Godard, Athdr-6Iran I/I (1936), p. 82), but the building is smaller and the vaulting system different. Mr. R. Hillenbrand informs me that the Seljuq hypostyle mosque at

Sujas shows some of the features of the Masjid-i 'Ali, including the four central pillars and a raised gallery, which to this day is used as the women's part of the mosque; for this information I am most grateful. 6 The convention of designating the qibla wall of the mosque as the south is used here, and the other directions are correspondingly designated. 7 I am most grateful to 'Abd al-Karim Karimi Quhrfidi, a senior and cultivated member of the Quhrfid community, who provided me with this and much other information about the history of the mosque. 8 The Masjid-i Kalah is found at a little distance from the Masjid-i 'Ali. Because of the value of the lower lying land in the valley bottoms for cultivation, the village is built on the side of the hill, with the houses built virtually one on top of another. The Masjid-i 'Ali has houses built up to it only on the east side, the Masjid-i Kalah has houses built up to it on three sides. Its free side, the north wall, faces on to a narrow alley that drops away steeply to, the East.

62

JOURNAL

OF PERSIAN

STUDIES

qibla

,o

,2

,3

,4

,5 M.

Plan, Masjid-i Kalah. Fig. 3. Ground

system of vaulting, the piers, pillars and gallery are identical to those of the Masjid-i 'Ali, except that the gallery does not run all the way round the building. It does not continue over the recess immediately to the west of the mihrab(P1. VIIIc). The main entrance is in the north wall where a door is dated 1024/1615. In the north-east corner stairs lead down to the street, and attached to the east end of the north wall is a barn where a nakhlis stored. The interior is undecorated apart from crude muqarnas decoration and some carved plaster in the mihrab.However, one item is of great interest, sc. a ceramic panel set in the back of the recess to the west of the mihrab. TheFoundation Panel,Masjid-iKalah This is a moulded blue and black under-glaze painted tile measuring 46 x 76 cm. The inscription and the scrollworkare in relief and reservedin white on a dark blue ground; black is used to outline the white areas. The design (P1.VIIIb) is standard for ceramic tombstones and mihrabs the I3th-I4th of centuries A.D. A few centimetres of it have been set into the plaster floor, rendering the first and last words of the outer inscription illegible. The outer inscription reads:

9~1;11 i

le

0 **

iw

;*A__w

THE MASJID-I 'ALI, QUHRUJD: ARCHITECTURAL AND EPIGRAPHIC SURVEY "....

63

this blessed mosque in the year eighty-three from 'Abd al-Nabi, may God bless him and give him peace, the poor slave Muhammad b. Fadl All~h b. 'All Quhrudi ordered the renovation of the structure of this blessedmosque in Jumida I of the year ... " " This formula is most unusual. The opening word may have been bunfya it was built ", though I have not been able to find any parallel example. The reading 'an" from " is dubious. The date 83 may either refer to a local tradition of the founding of the mosque, or more likely the hundreds have been left out. Across the top of the inner inscription is written the shahdda," There is no god but God, Muhammad is the Prophet of God ". 12. Written on the capitals of the inner pair of columns is found the Under this is inscribed Sura I of the potter: signature

" The work of Yfisuf [b.] 'Ali b. Muhammad [.....]

the potter of Qashdn."

Yalhy in Veramin,10 dated 705/1305, and the large lustre mihrabfrom the Imdmzdda Ja'far in Qumm dated 734/1334.11 This signature helps us to read what is left of the date on the blue-and-white tile: (sab'a mi'a). Four possibilities result: 716, 717, 726 or 727 (1316-27

The inscription is chipped in several places, but there is no doubt that this is the last recorded member of the famous Abli Tahir potting family, whose surviving work covers more than a century.9 from the Imamzdda Yiisuf, the last member, is known for two outstanding pieces-the lustre mihrab we have units beginning with sin which must be either six (sitta)or seven (sab'a); tens beginning with or 'ain representing either ten ('ashara) twenty ('ishrin); the sin of the hundreds must be read as seven
A.D.). The tile is not of the highest

quality, and it would be pointless to try to decide between them. The special interest of this tile is that it is, to my knowledge, the only piece signed by a member of this lustre-pottingfamily that is not lustre ware.12 Blue and white tiles of this kind are not common. Perhaps the same potter made the four other
known dated pieces of this type: the frieze tiles dated 705/1305, the mihrabs dated 719/1319 and 722/1322,13 and a piece more recently published, the bottom half of a mihrabdated 71 1/131 1.14

Tiles in theMasjid-i 'Ali

The tiles are set in panels in the mihrab,on either side of the door and in spandrels above the entrance arch. The lustre tiles are six-pointed and measure 18-19 cms. across the largest diameter; the lustre is brownish and still lustrous, the patterns are emphasized with touches of blue and turquoise. They alternate with hexagonal tiles glazed in opaque turquoise. Round the panels on either side of the door is

a border of dark-blue eight-pointed stars and turquoise crosses. Set in the mihrab some 78 whole are lustre tiles and over 90 half pieces, and in each panel by the door are 40 whole and Io half tiles. In the spandrels are set some 25 whole and more than 6 half tiles; others were hidden by the pieces of cloth draped over the beam just in front of them and these proved impossible to record.15
Keramik(Hetjens Museum, Dusseldorf I973), p. 138, Genealogical Tables of the Principle Faience Workers of 14 Islamische Kashan ", Survey,p. 1666. One can still make out the remains entry I85. of a few letters where the tile is chipped between " Mulham- 15 By " half tile " I mean a large fragment, usually exactly a half, mad " and "the Potter "; they would certainly suit a reading broken from a complete tile. There are no tiles made as half of Ibn Abt Tdhir. stars. Numbers given for the total of stars, the numbers in the 10 Ettinghausen, " List of Dated Persian Faience ", Survey, various panels, and the numbers of each of the various inscriptions will be found not to tally. Various tiles proved impossible p. 1683, No. 87 (hereafter referred to as " List ..."). 11Y. A. Godard, " Pihces Dat6es de to read, either from excessive damage, removing most of the CUramique de Kishn & D6cor Lustr6 ", Athdr-e'Iran (1937), PP- 309-17, figs. 139-43. II inscription, or by their inaccessibility for photographing. In the course of study I have tried to match up the half tiles, 12sBu Tdhir Husain, who signed a mina'i bowl, may be an earlier but out of a total of more than a hundred pieces, I was able to generation of this family; no other pieces of his work are match only three pairs. Others may have escaped my notice, known, see Wiet, " L'epigraphie Arabe ... ", Memoires de 1'Institut but a better idea of the original total is perhaps gained by d'ligypteXXVI (I935), No. Io and pl. III. 1i " List ... ", Survey,p. I690, No. 157, p. 1691, Nos. 163 and adding the numbers of wholes and halves, resulting in a total of about 300. Of these, the inscriptions of about 50 have not 164. This last was more recently published by Ettinghausen, " Comments on Later Islamic Ceramics ", ArtibusAsiaeXXXV been identified for the reasons given above. (1973), p. 169, fig. 7.
9 "

64

JOURNAL

OF PERSIAN

STUDIES

Their fragmentarystate is explained by the fact that, shortly before I900, thieves stole both the tiles and the door. A local " posse " was formed, and the tiles and door recovered. The inscriptionsdated I317/I900 are of the resetting.le Many of the tiles must have been severely damaged when removed. We know nothing of their formerpositions,and it is very possiblethat the mihrab may have been modified when the tiles were reset-in the head of the central niche is a plaster moulding of a corded ring with a protruding central boss that would well suit a 19th century date (PI.I a)."7 The only exception to the six-pointed lustre stars is one, set into the spandrels, of the usual eightpointed variety. This singleton is inexplicable in terms of the decoration as it now stands. Presumably it either formed part of a panel of such stars which were not recovered with the others, or it entered the building for the first time with the others in 1900oo.18 The tiles divide into two main groups; the larger containing Quranic quotations consists of some 125 whole tiles, 75 half tiles and numerous small fragments. They are dated as a group 700/1300. The second smallergroup comprises34 whole tiles, 20 half ones, and fragments,and is inscribedwith Persian poetry. As a group they are dated 707/1307. The decoration of the two groups is closely related. TheFirst Group These are dated 46 times in numeralsand 19 times in words. In words the date reads sannasab'ami'a " the year seven hundred ", occasionally prefixed by kutiba" it was written ". It occurs once as kutiba rabi'al-awwalsannasa'bami'a " written (in the month of) Rabi' I of the year seven hundred ", giving the more precise date of November-December 1300. The tiles dated in numerals are mostly marked Yo but Yo* and y ** occur often enough to show that the signs of the circle and the dot both equal zero (Pls.Ib and c, II a and b). The group is certainly of one date; and the numerals cannot indicate
dates as widely separated as 700, 750 and 755 (1300-1354).19

The Quranic quotations are: Sura

I.

Verses
(Opening) all

No. of times occurring

24

2.
2. 3. 5. 18.

(The Cow)
(The (The (The (The Cow) House of 'Imrin) Table) Cave)

255-6 (Throne verse)

285 I90-I 55 1-2

5
6 I 2 3

18. (The Cave) 36. (Ya-S.n)

40. 41. 48. 57. 76.

22 I-6
15-16
30-1

I 11
3 8 27 I 24
I

86. 97. io8.

I I0. 112.

(The Believers) (The Distinguished) (Victory) (Iron) (Man)

1-3 1-4 1-3

(The Night-Star) (Power) (Abundance)

(Help)

i-6 all all

all

57
to
2

(Sincere Religion)
(Men)

all
all

6
6

14. I
16

17

My thanks to 'Abd al-Karim Karimi for this information 18 We do not know how many tiles were originally in the mosque, or what proportion of them was recovered. In view of the (note 7). Such a deep rectangular recess with side niches would be scarcity of this type in European public and private collections, unusual for a 14th century mihrab; the mihrabof the Masjid-i one may perhaps presume that most were recovered (see below, Kalah is by no means as deep. The shape and decoration of note 37). The eight-pointed star is of a well-known series dated both would however suit a Safavid or later date. The recesses variously in the second half of the 13th century. between the piers facilitate the construction of a new mihrab 19See appendix below. at a date later than that of the building.

THE

MASJID-I

'ALi,

Q UHR.fD:

ARCHITECTURAL

AND

EPIGRAPHIC

SURVEY

65

Study of the use of Quranic inscriptions in buildings is still at an early stage, and it is therefore difficult to determine the significanceof this choice of verses. However, it is interesting to compare this selection with that of the lustre tiles in the mausoleum of Pir-i Husain at Bakik,20 where the commonest scarce amongst the tiles of the Masjid-i 'Ali, Quhr-id, (apart from the opening sura). Those that occur frequently at Quhriid (suras 36, 48, 76, and 97) are relatively uncommon in the Bdkfi mausoleum (3, 7, 2 and 4 times respectively). The suras of the Pir-i Husain are almost entirely from the last 15 suras of the Quran. The tone of these short powerfulverses, with their emphasis on man's dependence on God, the horrorsof Hell, and the need for repentance, seems eminently suitable for a tomb. The versesin the Masjid-i 'Ali emphasize rather the benevolence of God and Man's duty towards Him in life, and are therefore more suitable for a mosque. Orthographical mistakes are not rare and words are occasionally omitted or repeated, but the mistakes seem to be caused by hasty writing rather than illiteracy. Inscriptions frequently end in the middle of a phrase or even of a word. Where the inscription is too short, the space is filled with a section of the pious phrase: " God the Almighty speaks the Truth, His noble Prophet speaks the Truth." This is twice followed by: " We (are witnesses of) what our Lord said and what our Master (Muhammad) said ".21 The style of writing is uniform on all the Quranic tiles, and seems to be by a single hand. It is a free cursive with many extra ligatures, at times degenerating into complete illegibility. The designs (Pls. I-III) of the tiles show considerablevariety, but the number of motifs from which the patterns are composed is limited: stiff intertwining palmettes, crude flowers, borders which divide the inner area into a hexagon, and intertwining bands. There are no animals or figures among this set. Many of the patterns are very similar to those in the Pir-i Ifusain Mausoleum, which are dated 682-4/ I283-5.22 Pearl borders, rosette patterns, and half-palmettes match closely those in the Quhrfid mosque. Various of these elements are found in lustre tiles of a much later date. In a group dated 738/1338, the pearl border is found with degenerate formsof the central rosette.23 Such conservatismis characteristicof the Kashan lustre potters, from whose ateliers these tiles evidently issued.
suras are I12 (27 times), I 14 (23 times), i 10 (10 times) and sura I (19 times). All these are relatively

TheSecond Group
The tiles of this group, unlike those of the first, are inscribed with Persian verse and bear the date mi'a " At the beginning sannasaba'wa-sab'a rabi'al-dkhir 707/1307, which in its full form reads: Fi ghurra of the month Rabi' II of the year seven-hundredand seven " (Pl. Vc). This occurs in full five times; a curtailed form only giving the month or ' ... the year seven ', occurs some fifteen times (Pl. IVb and c, Va and d). None of this group is dated in numerals. The tiles are of exactly the same shape and size as the Quranic group, and their decoration is closely related though not identical. The parrots and phoenix well known on other lustre pieces appear on some examples of this group (Pls. IVc, Vd).24 The hand in which the poetry is written differs distinctly from that of the Quranic group. It is smaller and more angular, neater though no more legible. The poetry falls into two classes, both previously known on lustre ware; the erotic quatrain, and selections from the Shdhndma Firdausi.25 Of the nine inscriptionsoccurring more than once, two are of
and two are erotic quatrains. Two single tiles contain further quotations quotations from the Shdhndma
tiles Pls. IVb and Vb; compare also Pls. Ha and Vc. 2oV. A. Kratchkovskaya, Les Faiencesdu Mausolle de Pir-Houssein 25 See especially, M. Bahrami, Gurgan Faiences (Cairo (Tbilisi 1946, in Russian with a French r6sum6), pp. 8o-i. 1949), The analysis given there is of a total of Io2 tiles. sur de pp. I 14-20, and by the same author, Recherches les Carreaux 21 This formula is Lustre' Revitement commonly found on lustred tiles and mihrabs. (Paris 1937), pp. 65-9, 93-103. For selections See e.g. A. Lane, Guide to the Collectionof Tiles (London, see from the Shdhndma, Gyuzalyan, " Frieze Tiles of the 13th III HMSO, I96O), pl. I, where it occurs twice. century with Poetic Fragments ", EpigrafikaVostoka (I949), 22 Kratchkovskaya, op. cit., Pls. X, XIV, XVIII, XIX. pp. 72-82, and by the same author, " An Extract of the " Evidence for the Identification of 23Ettinghausen, Shahndma Pottery of the I3th and I4th centuries ", Epigraon KdIshan Pottery ", Ars IslamicaIII (1936), fig. 21. fika Vostoka(1951), pp. 40-55 (in Russian: review and r6sum6 24 The one Quranic tile that seems to have served as a model for in English by O. Grabar, Ars OrientalisII (1957), p. 55I). many of the poetic group is shown on Pl. IIc. Cf. the poetical

66

JOURNAL

OF PERSIAN

STUDIES

from the Shdhndma. remaining inscriptionsremain for the moment undecipheredofr The indecipherable, and for the most part occur on single tiles. From the phrases that are legible, they would appear to be further examples of erotic quatrains. From the Shdhndma:26
(I) On a single tile (P1. IVa).27 Chapter I, lines 1-2, 4-5:

In the name of the Lord of soul and wisdom Than Whom thought can conceive nothing greater Lord of name and Lord of place Lord the Provider and Guide He is above name and sign and imagination Beyond the painter's art, He is the Essence You will not see the Creator with your eyes (so do not hurt them by trying).
(2) On

eight tiles (P1. IVb).28 Chapter I2c, lines 23-4, 26: From the priest we remember in this way That at dawn Rustam prepared His heart was troubled, he prepared for hunting He fastened his belt, his quiver was full of arrows He turned towards the borders of Turan Like a lion fierce and vengeful. (3) On three tiles, (P1.IVc).29 Chapter 15, lines 2507-8, 2511: Now one must drink delicious wine For the scent of musk comes from the river The sky full of noise, the land full of commotion Happy is he who has a happy heart in drink Every garden is beneath the leaf of the rose Every mountain full of tulips and hyacinths.
(4) On a single tile (P1. Vb).30 Chapter I2d, line 17:

Everything that you sow you will reap, Everything you say you will hear. Quatrains (I) On seven tiles (P1. IVd).31 Satiated men of the world eager for your kindness Heroes of the world frightened of separation from you What have the gazelles in the desert (to compare) with your eyes The lions of the world are caught in your hair.
26

References are to the edition of the Shuhndma Siydqi, in His figs. 3, 7-10o, show three tiles and a vessel whose inscriptions by 6 vols. (Tehran 1956). Variant readings found on the tiles are reproduce exactly those lines that appear on these eight tiles sometimes found in Shdkh-Ndme.Kritichesky Tekst ... ed. E. G. from Quhrid. Bertel (Moscow i960), and in the Epigrafika Vostokaarticles 29 GyuzalyaAi, op. cit., p. 79, and fig. 5, shows a frieze tile in the mentioned above. British Museum that has the first couplet of this extract. These 27 C.f. Gyuzalyan, verses are also found on a tile from Ddmghan in the Louvre, see op. cit., p. 75 and fig. I (a frieze tile in the .. sur Hermitage that shows an extract from these lines). Bahrami, Recherches les Carreaux ., p. Io7. 2s Gyuzalyan, op. cit., p. 76, fig. 72 (frieze tile in the Museum for 30 This line is unrecorded elsewhere. Eastern Culture, Moscow). His article in Epigrafika Vostoka 31 See Bahrami, Gurgan Faiences,p. 12o, for other vessels inscribed with this verse. It occurs very frequently on tiles of the late (1951) is entirely concerned with tiles that have extracts of the from a small portion of the text, starting with the I3th- early 14th century. Shdhndma extract given here (section 2) and continuing for some 30 lines.

THE

MASJID-I

'ALI,

QUHROfD:

ARCHITECTURAL

AND

EPIGRAPHIC

SURVEY

67

(2) On four tiles (P1. Va).32

Oh heart do not seek help from any mortal Or seek shade from a bare branch Glory is from contentment, distressfrom desire Be content with your happiness, seek not distress. The choice of Shdhndma quotations shows that the verseswere not chosen to form part of a continuous narrative. The same verses are repeated on several different tiles, and the different sections are from unrelated storieswidely separated in the text. The fact that the very same verses are noted to occur on other tiles, and in one case on a vessel, of Kashan manufacturewould suggest that significanceof choice lay in something other than narrative quality. The first section of the Shdhndma religious in character and it is not surprisingto find such a quotais tion in a mosque. The remaining poetry is distinctly secular, like the many other similar examples that have been recorded on Kashan lustre tiles. This has led to the general assumption that such tiles were intended for secular use;33 there is however a certain amount of evidence that this was not the case. Tiles with " secular " poems have been recorded in at least two other religious buildings: the mausoleum of Pir-i Husain at Bkfi34 and the Imizmdda Ja'far in It has been claimed that the Dtmghmn.35 tiles in the ImdmzddaJa'far were transferredfrom a ruined palace to the tomb for safe keeping.36 In the case of Quhriad, there are strong arguments against a similar explanation. Firstly, lustre tiles of this particular shape (six-pointed stars formed by two equilateral triangles superimposed) are rare. To my knowledge there are only a handful of tiles of this shape known, mostly smaller (14 cm) than the Quhrfid group and distinctly differentin design.37 There is only one comparablein size and shape to the Quhraid tiles sc. a tile in the Victoria and Albert Museum (No. 561-I900). Its inscriptionand decoration can be directly paralleled among those in the Quhrfid group.38 It is written in the same hand as the Quhriid tiles and is dated 700/1300 in words. It was acquired by the museum in 1900 from the Mayer collection.39 It is reasonableto regard this as one of the tiles stolen from Quhriid either in I900 with the other tiles or possibly earlier. Secondly, the poetic group at Quhriid is identical to the Quranic group in size and quality, and very similar in decoration (cf. note 24). This fact, when considered with the rarity of the shape, and despite a difference of seven years in date of manufacture, would suggest that they were intended to form part of a single decorative scheme. The hypothesisthat a wealthy individual in Quhriid ordered a set of lustre tiles for his private abode which happened to be very similar to those in the mosque, and that at a later date his tiles were installed alongside those in the mosque, and that a similar process of transferringlustre tiles from a private residence to a religiousbuilding took place in at least two other cases many hundredsof miles away, presents such a number of coincidencesthat the theory becomes suspect. The implications of the contraryviewthat tiles with poetical inscriptionswere intended for religious buildings-are beyond the scope of this
paper.40

This verse is unrecorded elsewhere. the list of plate illustrations. 38 Cf. P1. Ic. The Victoria and Albert Museum tile contains sura Bahrami, " Le probleme des Ateliers d'toiles de Faience Lustr6e ", Revuedes Arts AsiatiquesX (1936), p. 187; A. Lane, 18, verses 1-2, which occurs three times among the tiles in Early Islamic Pottery (London 1947), p. 39; Wilber, The Quhrfid. Architecture IslamicIran (Princeton 1955), p. Ii entry I3. I am unable to discover when it entered that collection. of 3S I, The objection that, in the case of the Imamzdda Ja'far, tiles 84 Kratchkovskaya, op. cit., p. 82-4. 40 with pictures of animals and persons cannot have been intended 83 Bahrami, op. cit., pp. I86-89. 36 for a religious building, is not sustainable. On the tomb of Wilber, op. cit., entry 13 and bibliography. Keramik(Vienna Habib b. Mfisd in Kashan, where a number of lustre tiles are 37 One is published in W. Hein, FriiheIslamische the I963), Tafel XVIII, another in E. Atil, Ceramicsfrom World preserved, are several pictures of animals and two with persons, to which there is no objection at the present time (A. U. Pope, oflslam (Freer Gallery, Washington 1973), fig. 34. A tile in the " New Findings in Persian Ceramics of the Islamic Period ", Mus6e des Arts Decoratifs in Paris (No. 2252), whose decoration is similar to some stars in the mosque at Quhrfid, is smaller Bull. Amer.Inst. IranianArt and Archaeology (1937), pp. 155V/2 cm) and has no inscription border. See also Denkmdler 6, fig. 7). Tiles painted with animals and quotations from the (I6"5 Persischer Baukunst(Berlin 90oi-io), Abb. 75; and Riviere, La Quran are common, though as yet no tile with persons and Musulmane(Paris 1913), tile illustrated at the end of Ciramique Quranic quotations has been brought to light.
"2 33

68

JOURNAL

OF PERSIAN

STUDIES

Other Ceramic in Inscriptions theMihrab In the main mihrab niche are set two ceramic slabs (Pl. VIIIa). The first is a turquoised-glazedtile, of a well known type.41It measures44 x 36 cm. and contains two inscriptions,one being sura 112 of the Quran. In the spandrels of the arch and continuing in the field between the columns is the following:

-'Lbm1m3w*

.0Lm*_LI

f*I ~J

~l*J

\isJA,*-n

" The weak slave 'Abd al-Wdhid b. Muhammad ordered this mosque to be built in the year Quhrfidi seven hundred and eight ". The year of the hundreds is somewhat puzzling. The glaze has run thick at the bottom obscuring the relief moulding. The date must read " seven hundred ' as the name is the same as that of the donor of the door, also dated seven hundred. Two extra letters seem to have been included at the beginning of this word. The differencebetween this date and that on the door is perhaps explicable by the time taken to erect and decorate the monument. Above this is another tile with inscriptionsin lustre on an opaque white background. It was originally much larger: the text is in Persian verse, and on the left hand edge (bottom edge of P1. VIIIa) can be seen the beginnings of yet more verses, so at most only half the complete tile remains. The tile has a large fragment missing,replaced by bits from another part of the tile. The text is in a naskhhand tending towards nasta'liq,and generally rather illegible; few phrasescan be read. At the end is the date: wa-sab'ami'a. bi 'l-khairwa 'l-gafar/li-(sanna) sitta wa thalathin wa-kutiba awdkhir f .Safar\khutima " finished well and with success, written in the last days of Safar in the year 736/ September-October "1335 TheDoor The door consists of a series of wooden panels with inscriptions. It frames the entrance, and has a total height of 2 *75 m. and a width of 2 Io m. Two side panels stop at a height of 70 cm. from the ground where stone benches project from the wall (P1.VIa). The panels arejoined by border elements through which at a later stage large-headed metal nails have been driven, perhaps in the resetting after the robbery (Pls. VIb and VIIa). The inscriptions are each carved from single planks of wood, each cm. (P1.VIb). The background to the inscriptions leaf of the door is a plank measuring 170 X 48 x " miniature-bevelled" style. The border elements are and the leaves of the door are carved in a fine 7"5 carved in a style that became the norm in later wood-carving; intertwining palmettes and scrolls are raised from the background, but their surfaces are not modelled, leaving a " carving in two planes ". The leaves are carved to imitate the lattice work frequentlyfound in wood-workof this and later periods. in The minbar the mosque at Nd'in42dated 711II,eleven years later than the Quhrfid door, provides an interesting comparison. It is constructed from lattice work but many of the panels contain carving in the " miniature-bevelled" style of Quhrfid.43 A door with mock lattice work, and carving in the background similar to that at Quhriid is found at the entrance of the Shah Kamaliyya Madrasa in Yazd. This door is uninscribedbut the building is of the 14th century A.D.44 The main historicalinscriptionof the Quhrid door is written in a monumental naskh. The other inscriptions are either in the standard Kufic of the period, or in rather more advanced hands. The variety of scripts is remarkable,and their quality extremely high. If this is-as it appears to be-the work of one calligrapher, one can only be astonished at such a virtuoso display.
41

42

For comparative material, see Iraj Afshar, Yddgdrhd-yi Tazd (Tehran 1348), pp. 50-1, and plates on p. 472-3; and " List Nos. 142, 152, I58, 159 and 161. ... ", Survey, M. B. Smith, " The Wood Minbar in the Masdjid-i Djimi', NAin ", Ars IslamicaV (1938), pp. 21-35.

** Smith, op. cit., cf. pls. 3, 4, 5 and I2. "*Wilber, op.cit., entry I 3, gives the date as 792/1390. The building is earlier, dating from the first quarter of the 14th century A.D. For this information I am grateful to Professor Renata Holod.

THE

MASJID-I

'ALI,

QUHRUD:

ARCHITECTURAL

AND

EPIGRAPHIC

SURVEY

69

(I) The historical inscription reads:

d?.e 3Jmwm~~~LmsYI

eJCI Jt

CLlslJ

muIJ~

~\

.,

J.A..*J

jiIj

Id

I JI

C-1

csr?J eA~i
%AJ .j W

LJJ

4.p
") l

c~l~

[,

W-

p4U
1
it jmljli

~LJp..miJ13?

-I I-

CIL

--It,-~c?

" Ordered the building of this mosque and ablution place the Leader, the Imam, the Wise, the King of the Virtuous and the Khatibs, the Glorious One of the Qa7dis, Mufti of the Age, the Nu'mdn of the the Era, the Model of the Shaikhs and the Verifiers, the Crown of the Community and Religion, 'Abd-al Wahid son of the Happy, the Leader, the Imam, the King of the 'Ulama' and the Virtuous, the Glory of Iraq, the Seal of the Age, the Star of the Community and Religion, Muhammad son of the late, the Happy, the King of the 'Ulama' and the Virtuous, Ornament of the Pilgrims and the Two Sanctuaries, the Star of the Community and Religion 'Ali son of the Happy, the Leader, the King of the Great and the Virtuous, the GloriousOne of the Qadis, the Mufti of the Sects, Ornament of the Pilgrims and the two Sanctuaries,the Guide of the Religion Ahmad son of Muhammad son of 'Abdallah, son of 'Ali the Quhrfidi, may God forgive their sins. The poor slave Muhammad son of 'Ali son of Mujib the carver from Isfahan wrote this ".45 (2) The topmost panel is in Persianverse, and recordsthe vision seen prior to the erection of the Mosque. By the Grace of God, the Pure, the Judge Noble, Creator, Living, Powerful. One night in sleep I saw Murtatad Prince, glory of the Family of the Prophet (Xl-i 'Abd). He said 'Abd Wahid, be a master I command every sort of building to you. The foundation of a mosque, mihrab and minbar Also an ablution place and all that pertained thereto. When the Commander of the Faithful had given his order At once the means were prepared for me. Suddenly some stones appeared Some young men were there to help.
41 can find no reference to the patron of the building. The I calligrapher (who may or may not be the carpenter) is not recorded by Mayer, Islamic Woodcarvers their Works(Geneva and 1958); two carpenters sign themselves as naqqdr " carver" (pp. 38 and 42). In the second inscription is the signature of who is equally unknown. Possibly he carved the work to the other's design, as the latter only records himself as .HIjji, the calligrapher.

Ustid

70

JOURNAL

OF PERSIAN

STUDIES

And for Hd-mim Td-ha and Td-sin. and

By his order I placed the foundation in marble So that it rose up level. When I had to fulfil this sign I constructed this building to the limit of my means. The hope is that it will be a monument I shall not remain, may it remain for a period. Perchance a pious man and a man of pure faith Will recite a prayer for this place. Lord, You are the Knower of secrets That I have many faults and transgressions. I have not a good foundation for worship Except faith and submission and contentment. By Your grace make this slave happy By Your justice make me free. Make my hope of approach to Your court permissible Make Mustafa an intercessorfor us. For the Sake of the Walis and the people of rd-sin

These last letters refer to formulae found at the beginning of various suras of the Quran: Ya-sin = = Sura, 20, and Ta-sin = Sura 27. In this connection, Sura 36, Ijt-mim = Suras40-6 inclusive, there is obviously referenceto some esotericinterpretationof these prefixeswhose meaning is now lost.46 T.-ha Each line of the panel contains four couplets. Set between the lines in the bottom right-hand corner is " Work of the Master Hajji ". the phrase 'amal Ustdd On either side of the.Hdjji,are panels containing religiouspoems in Persianand Quranic and religious door quotations in Arabic: (3) Right-hand side a. I turn to God from Satan the Accursed b. Quran, sura 68, verses 51-2 c. Quran, sura 38, verse 42, This is a washing place, cool and a drink. d. The root and branch of prayer is ablution The prayer is " There is no god but He ". Lawful prayer cannot go towards God Unless you have performed ablution. When by the grace of God it is finished By the leave of Mustafa a prayer of peace. The man who made this pleasing building Oh Lord, by Your grace accept [it] from him. His name derived from the stem of unity His abode founded on eternal grace. He asks from noble pilgrims One Fdtihafrom a pure heart. When our desire had come to completion The hijri year had come to dhdl(=700ooI300).47
46

4' Dhll equals 700 in the Abjad system of counting (E.I.x, art.

E.Lx, art. " IKur')n

",

section 15.

"Abdjad "). This is a very early use of this system in dating buildings.

THE

MASJID-I

'AL1,

QUHRUD:

ARCHITECTURAL

AND

EPIGRAPHIC

SURVEY

71

side (4) Left-hand

a. Indeed God is the All-Hearing, the All-Knowing.
b. Quran, sura 19, verse 98.

c. And with God is the best Protection. d. ? (reading undecipherable) If you wish the Ka'ba to become prosperous Make the hearts of the pious happy. With love be loving to every person Be like the sun of the world to everybody. Do not fault a person, for that is not pleasing For God knows how He created him. Be generous, for any person who is generous I am not allowed to call unbeliever. While the master was erecting this door The owner was asking for one Fatiha. Oh God, forgive that slave Who sincerely recites the Fdtiha. In the month of Rabi II of the year 7oo/December 1300. (5) Above the door in Kufic script: a. Quran, sura 9, verse I8 b. There is no god but God the Explicit Truth, Muhammad is the Prophet of God, the sincere, the faithful. (6) On the leaves of the door are carved two traditions: a. The Prophet, upon whom be peace, said, Whosoever builds a mosque for God, be it only like a nest of a sand grouse, God will build him a house in Heaven. b. The like of a believer in a mosque is the whale in the sea, the like of a hypocrite in a mosque is a bird in a cage.48 These inscriptionsprovide us with much information. The patron is a local religious leader of some note, as were his father, grand-fatherand great-grand-father. Their titles reveal some connection with would suggest wa Sufism; " Model of the Shaikhs and the Verifiers ", qadwat al-mashdyikh 'l-muhaqqiqin, this. Their Shi'ism is revealed by the vision of 'Ali, the first Imam (Murtadd and " Commander of the Faithful " can only refer to him in this context; reference to the " family of the Prophet " meaning Muhammad, 'Ali, Fdtima, IHasanand IHusainconfirms this). The patron's concern with ablution is reflected in the inscriptions. The ablution place, siqdya,is referred to in the historical inscription (Inscription No. I) and in the record of the vision (Inscription No. 2), and is referredto in the Quranic quotation and poem on the right-hand side of the door (InscriptionNo. 3/c and d). To construct the mihrab he has been ordered, he acquires tiles from Kashan (a strong Shi'ite as the door he orders from Isfahan, which was strongly Sunnite.49 centre); The poems on the tiles and the door show something of the spirit of popular Shi'ism at this time, perhaps explicable by the influence of Sufism. The contents and spirit of inscription No. 2 are closely matched by those of a lustred faience plaque recording the foundation of a Shi'ite religious building in
48

The first tradition occurs in the major collections of the IHadith,though without reference to the sandgrouse, see A. J. et Wensinck, Concordance Indices de la Tradition Musulmane I (Leiden 1933 ff.), p. 221. A contemporary, though slightly abbreviated example of the version on the Quhrfid door, is found on Uljaitu's mihrabin the Friday mosque at Isfahan,

M. van Berchem, " Une Inscription du Sultan Mongol Uldjaitu," Me'langesDerenbourg(Paris 1909), p. 370. I have been unable to trace the other tradition. 49A. Bausani, "Religion in the Saljuq Period ", Cambridge " History of Iran V (1968), p. 285, and Religion under the Mongols ", op. cit., p. 543.

72

JOURNAL

OF PERSIAN

STUDIES

Kashan itself.50 Visions of Shi'ite saints seem to have prompted the foundation of numbers of holy buildings of Shi'ite faith in Iran. The evidence that the mosque is Shi'ite dispenses with one problem that might have otherwise arisen, as surviving evidence shows that when lustre tiles are found in a religious context, they are found almost exclusively in Shi'ite buildings. Though Kashan was one of the Shi'ite centres of this period, Mustawfi notes in 740/1339 that the surroundingarea is Sunnite.51 The lack of lustre tiles IHamdalldh in buildings in the immediate environs of Kashan is noticeable. The Masjid-i 'All is certainly Shi'ite, but the Masjid-i Kalah possessesnothing to indicate its creed. Might one hypothesize that this latter was the Sunni mosque of a village in which both sects lived ? That lustre tiles are found in a mosque is also noteworthy; of the dozen or so buildings in which lustre tiles have been recorded, only one is secular (the Kiosk on Takht-i Sulaimin),52 and only one other mosque is known (the Masjid-i Maydin in Kashan).53 The remaining buildings are all mausolea of one sort or another.54 Quhriid probably had close connections with Kashan as the nearest large Shi'ite centre, and may even have had special connections with the potters. Though Quhriid is not mentioned in Abfi 'l-Qasim's text on pottery production, he names Qamsar as the source for cobalt which the potters used to produce blues.55 Quhrtid is but a few kilometresfrom Qamsar over the mountains. Wulff refersto the cobalt mines as being "... between the villages of Gujar and Kohrrid (Quhraid)".56 The locals of Quhriid rememberlead mines in their hills which they say only closed down a few years ago. Lead was an indispensablesubstancein Kashan pottery production, as Abai'l-Qasim's text shows.57 Any connection with either of these substances in mining or transport would have brought the villagers in close contact with the potters. Dating of theMasjid-i'Ali There is nothing absolutely to indicate that the structure of the building is contemporary with the tiles and the door, these being easily removed and replaced in a later building. However, all the inscriptions are of a limited period (70o-736/1300-1335), except for those installed when the tiles were reset. The Masjid-i Kalah, built of the same elements and of essentially the same type as the Masjid-i 'Ali also has its major inscription dated in this period (716-27/1316-27). The keel-archesare characteristic of the Mongol period. This would all support the argument that the buildings are contemporarywith their inscriptions,though it is by no means conclusive. Until contrary evidence is produced it remains, however, the logical conclusion. The chronology is then as follows:
700/I300 707/1307 708/13O8 736/1335 c. 1200/1785, 131.7/1900
c.I380/I960

Founding of the mosque, door and tiles installed. Installing of second set of tiles.

Second foundation inscription installed.

Panel of poetry installed. Zamistdn built on in south-west corner.

Tiles and door stolen, recovered and reset, possible modifying of the mihrab.
(1339 Shamsi) Zamistan collapses, new mosque and porch built on its foundations.

Door and stair-case built in north-east corner of old mosque.

" Un

50 C. Adle,

Isfahan; Shah Kamil, Yazd; 'Abd al-Samad, Natanz; Disque de Fondation en C6ramique ", Journal Asiatique (1972), PP. 277-97. Imrmzida Ja'far, Dimgh~n; 'Ali b. Ja'far, Qumm; Ijabib b. Mfasi, Kashan; Yahyd, Veramin; and Ja'far, Qumm. 61Nuzhat al-Qulifb,ed. Siyaqi (Tehran 1336), p. 74" 52 R. Naumann, Ein K6sk im Sommerpalast des Abaqa Chan 15 Abi'l-Qaisim is the brother of Yiasuf b. 'Ali (note 9) and wrote zur his treatise in 700oo13oo, see " Abfi'l-Qasim's Treatise on ... ", Forschungen KunstAsiens (Istanbul x969), pp. 35-65. 53 Dieulafoy, La Perse ... (Paris 1887), p. 204. On p. 206 she Ceramics ", trans. J. W. Allan, Iran XI (1973), p. I 12, para 8. publishes a picture of it still in situ. See also " List ... " 5 Wulff, The Traditional Crafts of Persia (Cambridge, Mass., Survey,No. 38 and pl. 704. 1966), p. I63. Shrines at Meshhed and 57 Iran XI, pp. 112-3, paras. 13 and 19. 24 A preliminary list is as follows: Qumm, Mausolea at Najaf; Pir-i IHusain, BS.kfi; Pir-i Bakrin,

THE

MASJID-I

'ALI,

QUHRfJD:

ARCHITECTURAL

AND

EPIGRAPHIC

SURVEY

73

Conclusion The lustre tiles in the Masjid-i 'All and their inscriptionshave provided much informationon the use of this ware. It suggests that tiles with poetical quotations were intended for religious buildings; it confirms the Shi'ite connection of such wares; it demonstratesthe conservatismof the potters in their decoration and choice of inscription; and it shows that Quranic quotation is not necessarilyrandom or insignificant. It removes some uncertaintiesof numeral dating in this period. The mosque itself introduces a new type of small local building with several interesting features that persist in modern times. The door by its early date, its signature, its calligraphic and literary interest, its excellent craftsmanship and state of preservation, ranks among the most important pieces of Iranian woodwork yet to have come to light. Numeral Appendix: Dating on Lustre-Ware A circle in modern Arabic numerals always means five, and zero is representedby a dot. The use of the circle in the dating of lustre vessels,has however, always provided problems. Kiihnel referredto two pieces dated with this sign-a fragment in the Berlin Museum dated 7 Y, and a tile in the Boston Museum similarly dated (Fig. 4, 2 and 3),58 and read both dates as 657. Ettinghausenin his article on refers Kashan wares repeats the 657 date for the Boston tile, but in his list of dated faience in the Survey
to the possibility of the circle equalling zero as " ...

(2) so that o might have stood for modern."59 On a blue and white tile dated ,Y g he reads the

another form of 5 seems to have been customary

.
2.
3.

6
4oV
k

8.
ov
10o.0

VoO

I9s. 4

6.
7.

12.

'(A)
..

Fig. 4. Tracingsof Dates on LustreWarein numerals showing fives and zeros; not to scale.

date as 705 rather than 755 (or 754 as it had been read by Wiet60) because " .

the date given in numbers is so small that it should be read as o ".61 Bahrami in 1936 had already decided that the tiles dated 7 b V must be read as 607, as one is signed by a potter who signed and dated works in the first decade of the 7th century A.H. A working life of nearly 50 years he considers too long. 6 The style of these pieces dated in numerals correspondsexactly with those dated in words to the first decade of the 7th century A.H. We know of no pieces signed by him, or in the same style, after about 612. We cannot expect that after 45 years the artist produced a few more tiles in the style of almost half a century before, and dated them with numerals. Stylistically, the tiles belong to the first
'8 Kiihnel,

the second figure of

" Dated Persian Lustre Pottery ", Eastern Art III (1931), p. 227. 59Ettinghausen, " Evidence for the Identification of Kashan Pottery ", p. 46, and " List ... ", Survey, 1677, note 4. p. 7A

60

Wiet. Albumdu Musle Arabe du Caire (Cairo 1930), pl. 64. " List . .. ", Survey,p. 69o0,note 2. 62 Bahrami, " Le Problme des Ateliers ... ", pp. I84-5.
61

74

JOURNAL

OF PERSIAN STUDIES

decade of the century, and the circle must be read as zero. This reading raises no problems when considering other pieces dated in numeralsof this period. There are known to this author 28 dates between 598/1202 and 738/1337 that are represented on lustre ware by numerals, of which ten contain five or
zero:63

598 (2) 607

(I)

Lobed plate"4 Tile65

(3) 607 (4) 654

(5) 665 (6) 665

tile66 Fragmentary Plate67

^Startile, DImghdn68 Mihrab tile69

(7) 685 (8) 700 (9) 701

(10) 705

Star tile70 Star tiles, Quhrid71 Star tile72

Mihrab, Veramin73

(11) 707 (12) 7Io

Mihrabtile74 Star tile75

The crucial dates are Nos. 2, 3, 8, 1o and I I, where the small circle might be argued to represent five. However, in each case the circle must be read as zero for stylistic reasons. The argumentsfor so reading the circleson Nos. 2, 3 and 8 have been given above, and similar arguments apply to the other two cases. Both pieces suit a date in the early 8th century A.H. and to try to fit them into a mid-8th century date would create enormous problems in the stylistic development of this ware, not least from the fact that dated pieces of lustre ware of this type after 740 are unknown.76 To explain two singletons dated in numerals after this date would stretch any hypothesis, and in view of the evidence provided by the Quhrid tiles, and that of the 607 tiles, one cannot doubt the reading of the circle as zero. If this is accepted, then it will be seen that no case occurs where a five is written as any kind of circle; it always appears in the double-loop form.77 The zero is either a circle or a dot. The modern heart and circle shaped fives evidently developed from the double loop, at what date remains to be determined. The evidence shows that in this period five was written only as a double loop, and zero as a circle or dot. This holds true for lustre, mina'i and under-glaze painted wares. Other objects dated in numerals are much rarer and the problem seems not to have arisen yet.7'

und Steinbacher Persische (Istanbul 1935), pl. IV. Fayencetechnik s9 Mina'i and under-glaze painted wares are also known dated in numerals. These are not taken into consideration, but they 74 " List ... . ", Survey, No. 1oo. Ettinghausen, " Important raise no difficulty for the argument here given. Pieces of Persian Pottery ", fig. I8. The last digit would seem to be 8 or 9 rather than 7. "*Bahrami, " A Master Potter of Kashan ", Trans. Oriental Ceramic 7s " List ... ", Survey,No. io6. Ettinghausen, " Dated Persian Society(1944-5), pl. i6a. Ceramics in Some American Museums ", B.A.LP.A.A. IV/4 65""List... ", Survey,pl. 722e, No. 44. 66 Bahrami, op. cit., pl. LXVIb. " List... ", Survey, No. 45. (1936), fig. 9g. 7 No photograph of this has ever been published. Ettinghausen 76 That is to say, that star tiles and mihrabsof the Kashan type dated in words do not continue after this date (" List ... " (" Important Pieces of Persian Pottery ", Ars IslamicaII (1935), No. 133). The I5th century A.D. pieces are of a different Survey, p. 58) describes it as similar to the Victoria and Albert Museum kind. The mihrab published as dated 751/1350 (Melikian, " The polo-player plate dated 604 (see Survey, 703b). It is now in pl. the collection of G. Reitlinger Esq. The second digit is a small Sufi Strain in Kashan Art ", OrientalArt (1966), pp. 251-58) is circle and should, as in the other cases be read as zero. Mr much later, possibly I9th century A.D. a view with which Mr. Melikian now agrees. Reitlinger kindly allowed me to see the piece, but it was too late to obtain photographs in order to trace the date for this article. 7 The five of the lobed plate dated 598 is not written with a circle 68 " List... ", Survey,No. 58 pl. 721. but with a double loop in which the upper loop has been drawn 69 LustredWall Op. cit., No. 57. H. Wallis, The Thirteenth very small. The loop five in the star dated 685 is similarly Century Tiles (London 1894), Fig. 39. distorted and the bottom loop is shown as a line. 70 " List... No. 8i. Lady Evans, LustrePottery(London ", Survey, 7s Mr. Lowick of the British Museum informs me that coins of ShAh Rukh are known dated in numerals in the first decade of 1920o),pl. 71 See above, II/5. Ic, Ha and b, IIIa. Pls. the 9th century hijri, where the zero is also represented by a small circle. 7s Victoria and Albert Museum No. 734b/I888. Unpublished. T' "List ... ", Survey, No. 97. H. Ritter et alii, Orientalische

P1. Ia. Mihrab of the Masjid-i 'AlI, Quhruid.

Quotations, Masjid-i 'Ali. P1. Ib and c. Tiles with Quranic

Pl. II. Tiles with Quranic Quotations, Masjid-i 'Al.

Pl. III. Tiles with Quranic Quotations, Masjid-i 'Al.

Pl. IV. Tiles with Poetic Quotations, Masjid-i 'Ali.

Pl. V. Tiles with Poetic Quotations, Masjid-i 'Ali.

Pl. VIa. Door of the Masjid-i 'Ali.

P1. VIb. Detail of the d

Pl. VIIa. Detail of thedoor.

Pl. VIIb. Interior, Masjid-i 'Ali.

Pl. VIIIa. Ceramic Tiles in the Mihrab Recessof the Masjid-i 'Ali.

Panel, Masjid-i Kalah, Quhriid. P1. VIIIb. Foundation

Masjid-i Kalah. P1. VIIIc. Interior,

THE FOUNDATIONS OF PERSIAN PROSODY AND METRICS By L. P. Elwell-Sutton

The poetry of the early Islamic period in Iran is marked by some obvious points of departure from the past. The language is coloured by Arabic (though perhaps less than contemporary prose); the sentiments are influenced by religious and philosophical ideas developed under Islam; even the new Arabo-Persianscript has its effect on the poet's style. All these influences were absent before the Arab invasions of the Ist/7th century. Is it legitimate then to regard this poetry as a complete break with tradition, an entirely new development in Persian literature? In seeking an answer to this question we are severely handicapped by the absence of source material. It is only in fact from the 3rd/9th century that we begin to find any substantialvolume of Persianverse in the written records, and even this consistsof small isolated fragments quoted in the body of historical, literary and lexicographical works of later date. Nevertheless, it is possible to judge that by this time Persian versificationhad already reached a fairly well-developed stage, from the points of view of both prosody and subject-matter. There are several possible explanations for this sudden leap into the light of day. Some scholars, both Western and Iranian, maintain that poetry was not composed in Persian prior to the Islamic era, and that the earliest efforts were those modelled on Arabic verse, which so far as can be judged had achieved a high degree of technical achievement at least as early as the century preceding Muhammad's mission and the Arab conquest of Iran. This theory is based on the supposed identity of Arabic and Persian " classical " metres, from which premise it is argued that the Persians, having no poetic tradition of their own, merely adopted that of the Arabs. This could of course account for the technical maturity of their earliest works. This view finds some support in a statement by the 3rd/9th century writer 'Amr b. Bahr al-Jdhiz to the effect that the versification (kaldm)of the Iranians was not to be regarded as poetry (shi'r).1 However, a modern scholar, Muhammad Qazvini, is certainly going too far when he argues that a verse in a " classical " metre cannot have been composed earlier than the 3rd/9th century, because the rules of the Arabic metrical system were only " laid down " at the end of the 2nd/ 8th century.2 We shall revert to this point shortly, and will do no more at this stage than suggest the innate improbability of the belief that the Iranians, famous now for more than a thousand years for the wealth and profundityof their poetic expression,were prior to that devoid of any poetic instinct and started as mere imitators of the literary genius of another culture. We know in fact that poetry was composed in Iran long before Islamic times. True, we can hardly expect to be able to trace any direct relationshipin form or content between Islamic Persianpoetry and the Gathic hymns of the Avesta, composed not later than the 7th century B.c. and probably much earlier; the gap is too wide. But the existence of these hymns is clear enough evidence that poetic genius was not a sudden mutation in the Iranian character dating from an influx of foreign invaders, but a trait as old as the history of the people themselves. In Parthian and Sasanian times the evidence becomes still stronger. Benvenistehas shown that the Pahlavi books Drakht-iAsarik,a religious polemic, and Ayydtkdr-i a Zarjrdn, secular epic, were almost certainly originally written in verse, sometimeswith rhyme or assonance,3while the Manichaean documents discovered in Turfin in Central Asia in the early years of this century contain many hymns in metrical form.4 Isolated samples of verse occur in other Middle Iranian works.
al-JAbii, Kitdbal-Bayanwa'l-Tabyin (Cairo, I932), Vol. I, p. 295. Qazvini, " Qadimtarin shi'r-i farsiy-i ba'd az Mub.ammad ", Bfst Maqdla-yi Qazvfin, 2nd ed. (Tihrdn, 1953), Islim Vol. I, p. 35. 3 E. Benveniste, " Le texte du Draxt Asfirik et la versification
1

pehlevie ", JA (1930), Vol. II, pp. 193-225; ibid., " Le M6morial de Zarer, poeme pehlevi mazd6en ", JA (1932), Vol. II, pp. 245-93-

in (London, Hymn-CyclesParthian Mary Boyce, TheManichaean 1954), PP. 45 if.

75

76

JOURNAL

OF PERSIAN

STUDIES

Apart from such concrete evidence, the historicalworksof the early Islamic period constantly repeat the tradition that poetry was sung and recited at the Sasanian court, as well as on a more popular level. statement referredto above, that this was not true poetry, may be discounted as the reaction of a protagonist of Arabism against the claims of the Shu'fibiyya. The names of famous minstrels of Al-J.hiz's Sasanian times are recorded-Bdrbud, Bdmshdd,Nagisi-and we may be sure that what they sang was poetic in form. We even know something of the different types of poetry composed. The sarfidwas a or was hymn addressedto a king, or used by the priestsin their rituals; the chakdmak chdmak a love-lyric or romantic story, while the tardnak a feasting or drinking song. All these terms are still in use, and was the same classificationcould equally well be applied to the poetry of Islamic times. The same historical sources also provide us with frequent referencesto poetry and poets of the early years of Islam, and even sporadic examples of such verses. It seems clear then that we must look for some explanation other than a lack of poetic genius for the virtual absence of recordedpoetry prior to the 3rd/9th century. The most obvious one is that the written recordswere destroyed. There is certainly reason to think that many Pahlavi manuscriptsdisappeared during and after the Arab conquest. Nor was this the end of it. The library of the Samanid court is related to have been destroyedin an accidental fire about 388/998, while Mahmfid of Ghazna is said to have demolished the Bfiyid library of Ray in 420/1029. Nor can we overlook the wholesale destruction of life and property during the Mongol invasions of the 7th/I 3th centuries, in spite of what tradition says about the efforts of rulers like Hfilgni to preserve libraries. An alternative, equally likely, explanation is that poetry was rarelycommitted to writing in Sasanian times. The poets and singers we hear most of were court officials whose task it was to entertain the monarch, and their workwas not intended for an audience outside the court. Popularverse, on the other hand, would not-as indeed it scarcely is even today-be considered of sufficient importance to be preserved. Moreover, the art of writing was a monopoly of the Zoroastrian priesthood, and there is evidence of their hostility towards poetry,s an attitude that was shared by the religious authorities in early Islamic times. Although poetry would thereforehave been preservedin the memories of the professionalreciters,the attitude of the priesthood, and-in Islamic times-the dominating status of Arabic as the language of literature as well as of religion and science, would together have militated against efforts to make a more permanent record. Bausani, noting that this " second-class" status of poetry continued into much later times, links it with the constant referenceby Siifi poets to rusvd'i(ignominy), rindi (rascality), kufr (idolatry), even in a religious context.8 However, even if we abandon the theory that the Iranians composed no poetry before the 3rd/9th century, it is still possible to argue convincingly that there is no continuity between the classical poetry that dates from this time and the pre-Islamic and early Islamic verse that must have preceded it. Our reactions to this argument must depend on the criteria of comparisonthat we adopt, and it is probably right to lay the main emphasison formal prosody rather than on style or subject-matter,which must be subject to extraneousinfluences. Even so, the materialsfor the earlier of these two periods are so scanty that it is difficult to observe any consistent pattern or thread of development within itself, let alone in comparisonwith the later period. We may thereforedo better to consider the latter first, and to take a short leap forward in time to the point where the earliest examples emerge of poetry written in accordance with the " rules " of classical quantitative verse. It is all the more profitable to do this because it

seems that, once this tradition of versification had become established, it ruled unchallenged for centuries, and indeed has scarcely been abandoned even by the "free verse" practitioners of the present day. So the output of the poets of the period up to the end of the Ghaznavids will provide us with ample, consistent material for the formation of a judgement as to the nature of Persian prosody. Having said this, we must add that this material has not hitherto been particularly accessible. The

is practice of gathering the writings of an individual poet into a divadn a relatively late one, and it is not
6

Idem, "The Parthian gosan and Iranian minstrel tradition ", b. JRAS (April 1957), P- 31; Shams al-Din Mul.ammad Qais al-Rizi, al-Mu'jam Ma'dyfrAsh'dral-'Ajam,ed. Muhamff

mad Qazvini and Mudarris (TihrAn, 1935), PP. 149-50. Persiana Ra.avi 6 A. Pagliaro, and A. Bausani, Storia della Letteratura (Milan, g96o),p. 243.

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

77

until the poets of the Ghaznavid period itself, like 'Unsuri, Farrukhiand Manuichihri,that we have the advantage of these handy compendia. For the earlier poets we must comb a wide range of sourcesnotably works on rhetoric, poetics and literature like the Tarjumdn al-Baldghaof Radiiyani (5th/I Ith the Qdbfisndma cent.), Maqdlaof Nizdmi 'Arilli (6th/I 2th cent.), the HIadd'iq (5th/I Ith cent.), the Chahdr al-Sihrof Rashid al-Din Vatvat (6th/I2th cent.), and al-Mujamfi Ma'dyirAsh'dral-'Ajamof Shams-i Qais (7th/I3th cent.); anthologies like the Lubdbal-Albdbof Muhammad 'Aufi (7th/I3th cent.), the of al-Shu'ard Amir DaulatshTh(9th/ 5th cent.), and the Majma'al-Fusaha Rii~ Quli Hidayat of Tagkirat i 9th cent.); and lexicographical works like the Lughat-i Fursof Asadi Tisi (5th/i Ith cent.), and (i3th/ its much later successorsof the i I th/1 7th and 12th/18th centuries. To these should be added a few historical works, notably the Tdrikh-i Baihaqiof Abu'l-Fail Muhammad Baihaqi (5th/I Ith cent.) and the Tdrikh-i Sistdn(5th/I Ith cent.). The first thing that is to be noted about all these sourcesis that even the earliest post-date the beginnings of " classical " Persian poetry by some two centuries; moreover, most of them provide only isolated fragments or even single verses, used as illustrations of lexicographical, grammatical or prosodic points. Fortunatelythe taskof gatheringtogether these disparatefragmentshas now largely been done for us, notably by Sa'id Nafisi and Gilbert Lazard,7and so it is not too difficult to gain an overall view of the field. Several poets compete for the distinction of having composed the earliest extant verses in the classical form. We must discount the claims of Abu'l-'Abbis Marvazi, who is stated by 'Aufi to have addresseda laudatory ode to the Caliph al-Ma'mfin in 193/809, both because the style of the few quoted versesis obviously of later date, and because the poet is also stated to have died in 300/912-13. We must also regretfully exclude such figures as Ibn al-Bu'ais (d. 235/849-50) and Abu'l-Ash'as al-Qummi (fl. 200/815), since, though their names are given by the historianTabari and others, their poems are not quoted. Our most informative source on the early development of Islamic Persian poetry is the Tdrikh-i Sistcn, an anonymous work the first part of which was completed in 445/1053. According to this, the honour of having composed the first poem in dard (i.e. " court ") Persian-the form of the language that came to prevail for literary purposes-must be given to Muhammad b. Vasif, who is credited with having broken into Persian verse when his patron, the Saffarid ruler Ya'qiib b. Lais, objected to being praised in Arabic.8 This was in or shortly after 251/865, and the poem refersto the defeat and death of one of Ya'qib's enemies, the Khtrijite 'Ammdr. It begins: 'ai 'amiri ki 'amir~n-ijahdn khdssava 'am banda vQichdkir u maulat va sagband u ghulim O prince, to whom the princes of the world both great and small Are but slaves and servants, clients, dog-keepersand house-boys! His successis said to have inspired another of Ya'qab's enemies, Bassdmthe Kurd. Suing for peace, he addressed the victorious Ya'qiib in these terms: har ki nabtid fly bi-dil muttaham bar 'asar-i da'vat-i tfi kard na'am All those who have no suspicion in their hearts Have replied to your call-Yes! The third poet mentioned by the Tarikh-i Sistdnis Muhammad b. Mukhallad, whose panegyric to Ya'qtib was composed on the same occasion:
'Sa'id Nafisi, Ahvdlu Ash'dr-iAba 'Abdalldh Ja'far b. Muhammad RaidakiSamarqandf(Tihrdn, Vol. I, 1930, Vol. II, 1931, Vol. III, I940; reprinted 1962); I. S. Braginski (ed.), RudakiStikhi (Moscow, 1964); Lazard, Gilbert, Les premierspoltes 84 persans (Tihrin/Paris, I964). p. 209.

8 Tdrikh-i Sistdn, ed. Malik al-Shu'ard Bahdr (Tihrin,

1935),

78

JOURNAL

OF PERSIAN

STUDIES

juz tu naztd IHavvava 'Adam nakisht shir-nihddibi-dil u bar manisht Eve never bore nor Adam begat one like you Lion-like in heart and in greatness! All these pieces must have been written shortlyafter 251/865, the year of 'Ammar'sdeath. Curiously Sistdndoes not mention another poet, HIanzalaof Badghis, who must at least have enough, the Tdrikh-i been contemporarywith the foregoing, for the following versesfrom his pen are said to have inspired a local chieftain whose military career began not later than 259/873mihtari gar bi-kam-i shir dar ast shau khatar-kunzi kim-i shir bi-jiy ya buzurgiy u 'izz u ni'mat u jih yd chu mardant marg-i riiydrfiy Even if chieftainship be in the mouth of a lion, Be reckless, seek it from the lion's mouth! Either greatness, glory, wealth and pomp, Or meeting death face to face like a man! All these poets are quoted in the context of historical anecdotes, and may well have been of only local reputation-in which case we may assume that there were many others in all parts of Iran whose verse did not happen to lend itself to such quotation. Leaving aside for the moment questionsof style and subject-matter,the first point to be noted about this group of poems (whatever may be said of their predecessors)is that they are all finished examples of common classical metres, as used throughout the history of Persianliterature. Muhammad b. Vasif's is in ramalmakhban au and those by Bassim the Kurd and Muhammad b. Mukhallad in mahdhzf sari' ma.twi while IHanzalahas chosen khafif makhban au maksiifau mauqff,maq.sar, malhdhif maqsar. The terminology is derived from the system of Arabic prosody devised by al-Khalil b. Ahmad al-Farhildi (or al-Farthidi) of Basra about the year 175/791-2, and first described (so far as extant works are conThe fact that these are cerned) in al-'Iqd al-farid of Ibn 'Abd Rabbih of Cordoba (246/860-328/940).9 the earliest surviving Persian verses to follow the rules of classical verse, that the terminology of these rules is derived from this Arabic system of prosody ('arfid, and whence the Persian term 'arf~i?), that the traditional sources describe them as " the first verses in dard Persian ", has hitherto led scholars, both eastern and western, to the conviction that this was an entirely new departure, and that these poets and their contemporaries, abandoning the poetic traditions of the past, spontaneously and simultaneously decided to copy an alien system of prosody that had grown up around an entirely different language. This question is so fundamental to any discussionof the origins and development of Persian versificationthat it is worth while looking at it in rather more detail. What do our sources mean by " the first Persian verse " ? We have already seen that verse of one kind or another had been known to Persian literature for centuries, and indeed the passage from the Tdrikh-i Sistdnreferredto above, while saying: " Muhammad b. Vasif then began to recite poetry in Persian (shi'r-i pdrsi). He was the first to recite Persian poetry in Iran, for no one had recited it before continues: " . . . because in the time of the Persian kings speech in their presence was him...", accompanied by the lute in the khusravdni (royal) manner." The whole episode of the performanceof the poets before Ya'qfib b. Lais need not be taken too seriously as history; it has rather the ring of a story composed to account for a phenomenon, in this case a new landmarkin the development of Persian poetry. But exactly what was this landmark? Possibly it was linguistic. The author of the Tdrikh-i Sistdnrefers to the language as " Pirsi ", and elsewhere we find it called " parsi-i dar!", court Persian.
9

Ibn 'Abd Rabbih, al-'Iqd al-farid (Cairo, x928), Vol. IV, pp. 34-87.

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

79

into mutaharrik (" moving ", vowelled) and sdkin (" silent ", unvowelled), and then grouped into (" elements ") and arkdn (" feet "). It is considerably simpler, however, to classify the syllables us.il according to quantitative length. Each language has three short vowels and three long, together with

Discussionof this falls outside the scope of this essay, and we may content ourselveswith the thought that an explanation was wanted for the domination in Persian literature of the " court " language of the Siminids. Or the landmark may have been the adoption of a new system of prosody. If we admit that there was already in the 3rd/9th century a long-standing tradition of oral poetry in Iran, both popular and courtly, it is stretching credulity a little to suppose that the prosodic techniques of this poetry were suddenly jettisoned in favour of an alien system. Yet we could accept this if there was found to be a striking resemblance between the Arabic and the Persian metres. The only evidence for such a resemblance is the traditional descriptionof the latter in Arabic terms, for when we examine them objectively we find that the two are far apart. Of al-Khalil's system it is sufficientto say here that he laid down sixteen basic metres (to which three more were added by later prosodists),together with a list of modifications (zih?fdtor 'ilal) by the application of which virtually any other metre of a quantitative nature could be incorporatedinto the scheme (others indeed were invented by his successorsto explain some of the Persian metres). In spite of the ingenuity of this system, it had one surprisingfailure; the syllable was disregarded, and indeed left unrecognised. The prosodistsworked rather with the letters of the written script, which they classified

two diphthongs also classifiable as long vowels. The possible syllable structuresare therefore: short U long-

Consonant + short vowel Consonant + long Consonant + shortvowellongconsonant vowel +

Consonant + short vowel + consonant + consonant

Consonant long vowel + consonant +

overlong(in Persianonly)

sometimes of ' U, separated by one or two (occasionally three) variable syllables (see Table on next page). The Persian system is entirely different. Stress does not seem to enter at all (or if it does, provides a contrasting pattern to the quantitative one); nor is (virtually) any variation permissible in the quantitative value of the syllables. What we find are a number of regular patterns of short and long syllables, in which two consecutiveshortsyllablesmay be replacedby one long, and a long and a shortsyllableby one overlong. In addition the final syllable of a misrd'(hemistich) may be either long or overlong (never short), while an initial pair of short syllables (where this occurs) may be replaced by long + short, never by one long. Persian verse is therefore more precisely framed and balanced than Arabic, and yet succeeds in achieving a high degree of flexibility owing to the possibilityof varying the number of syllables in a hemistich without altering the total quantitative length (in contrast to Arabic, where the converseis the case). A table of the Arabic metres follows:
10

up around a series of invariable stressed " nuclei ", consisting normally of the sequence U -', and

A long syllable is equal in value to two short syllables, an overlong to three (but in practice only in the sequence long + short). This last type of syllable does not exist in Arabic, and so is not provided for in the rules of prosody, even when applied to Persian-of which language it is a marked feature. It is worth noting in passingthat experimentwith samples of modern Persianverse-speakinghas shown these values to be approximately correct in practice. In both languages metres are formed by arrangingsyllablesof differentlengths (ortheir equivalents) into regular and orderly patterns. But it is at this point that the two systems part company. Among recent theories on Arabic metrics, the most convincing is that of Gotthold Weil,xowho has shown (basing himself largely on a new interpretation of al-Khalil's system) that the Arabic metres are built

Gotthold Weil, Grundriss Systemder altarabischen und Metren (Wiesbaden, 1958).

80
i. mutaqdrib: ux

JOURNAL

OF PERSIAN STUDIES

mutaddrik:

x x
X

u-

ux
X UX UUU-

x x
X X

u-

Ux uX UX UUU-

x x
X X

ux UX UX UUU-

u-

x x
X X

ux
X UX UX

2. hazaj:
rajaz: ramal: 3. wifir: k~mil:

U-

u-

U-

XXX XXX

XXX XXX

XXX XXX

XXX XXX

U-

4. tawil:
madid: basit:

U-

x
x u-

- x
x x

x U-

U- x
uUx X

x
x UX UX U-

x x U- x xu

5. mudari': muqtadab: mujtathth: sari': munsarih: khafif:

U-

x x

X- -U x - u xu

X x x x

Ux x x uuu-

X x x x x

X x x X x x

UUUu-

x X x x x x x - U -u -u x x x x uU-

Like Arabic, Persian prosody reveals five distinctive patterns, but this is purely coincidental, since the five patterns in each language are quite different. The Persian patterns are as follows:
I2345678...
I. U--U--U--U--*

2. U---U---U---UU--3. UU--UU--UU--UU--* 4. U-U-UU--U--UU--U 5. --UU-U-U--UU-u-u

***.. . *. .*. ....

By starting the verse at various points in the pattern, and by varying the total number of " basic " syllables, an almost infinite range of distinct metrescan be evolved. In fact, something like two hundred metres are known to Persian poets and prosodists,and samples of most of them are to be found in the poetry of the early period with which we are concerned. A statistical count shows that all but an insignificant minority fall into one or other of the five patternsset out above, Pattern 4 being the commonest, followed by 2, 3, 5, and I. The twenty-four commonest metres are given in the following table together with an illustration from an early poet (the first figure in the code number indicates the pattern, the second the point in the pattern at which the metre begins, and the remaining two the number of" basic " syllables):
I. II

: U --U
U-

--U

-- U--

UUU- UU -

mutaqaribmuthamman mahdhiuf
-U -U

pay-i mahd-i 'atfal-i jdhat sazad * ki 'iqd-i suraiyd shavad bS.z-pich

U -

It is fitting that the necklace of the Pleiades should serve as a plaything on the cradle of the children of your greatness. (Shahid Balkhi)

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

81

I. I12: U -U-

- - U-- U -U-U -U - UU - -U - -

sdlim muthamman mutaqarib

U --

magar ghaib u 'aib-ast k-izad naddidat* digar har-chi bdyist diniy u dari Except for the unknowable and the unworthy, which God has not bestowed upon you, you know and you have all that is proper. (Firiz Mashriqi)
2. I 1: U---U--U -hazaj musaddasmahdhaif

nigdrind bi-naqd-i jant nadham * girani dar bahR 'arzant nadham -- U-U -U ---U - Beloved, I will not sell you even at the price of my life; you are of great worth, and I will not sell you cheaply. (Mahmiid Varraq)
2. I116: U - - U-- - U--- - - U- - -U -sdlim hazaj muthamman

bi-tir az chashm-i nibina sapidi nuqta bar darad * ki nah didah biyazirad na nabini khabar dirad
U U--U --U-

With his arrow he can remove the white spot from a blind man's eye, without the eye being harmed or the blind man being aware. (Shahid Balkhi)
2.316: --U---U---- U-U--U-

sdlim rajaz muthamman UU-U -U-

khaishanabidh-I ghariji ba dtistan-i yak-dilah *

-U-U-

giti bi-'aram andariin majlis bi-bing iRvilvilah

-

How good is the morning draught of wine with friends of like mind! The world is at peace, while the gathering is full of noise and merriment. (Abai Salik Gurgani) 2.41 I: -u---u---uU -U--UU---U-U--U

ramalmusaddas mahdhiaf

khtin-i khud-ra gar birizi bar zamin * bih ki 'ab-i rily rizi dar kinar If you shed your own blood on the ground, it is better than throwing aside your honour. (Abti Salik Gurgani)
2.415: - U - - - U - - - U - - - U ramal muthamman mahdhi7f U- -U -U U-U- -U -U

har ki bdshad tishna vii chashmah nayabad hich jay *

U U -

bi guman rail biyayad gar biyabad 'abkand

U -

A man who is thirsty and cannot find a spring anywhere will surely be satisfied if he finds a stagnant puddle. (Shahid Balkhi)

82

JOURNAL

OF PERSIAN

STUDIES

ramalmusaddas makhbfin 3. II : U U - - U U - - U U mahdhkaf * shabam az y~dash chrin shTvgara riizam az dardash chiin nim-shabast
- U -U U -U UU --

My days, from yearning for her, are as midnight; my nights, from the memory of her, are as the desolate land of Sh~vgard. (Abu'l-'Abbis Rabinjani)
3. 15: -U U U--U -U UU U-U uU U--u U Uramal muthammanmakhbiin mahdhif

nauha-gar karda zabdn chang-i hazin 'az gham-i gul * mily bugshdda va bar ruiyzanan ndkhiind

-U U

--u U-

The mournful lute raises its voice in lamentation at the sorrow of the rose, its hair dishevelled, beating at its face with its finger-nails. (Firtiz Mashriqi) u-- U U -3.314: -- U U-'ai 'az rukh-i til tafta zib!'iy u 'aurang *
U U - -UU --U U-U -- U U U U -U

akhrab makfaf hazaj muthamman mahdhaf

'afrfikhta'az tal'at-i til masnad u 'aurang

- -U U

O thou from whose face shines beauty and splendour, by whose countenance throne and palace are illuminated! (Shahid Balkhi)
3.307(2) : --

U U ----

I--

U ---

td kai kunad Rikharam

-

ti kai zanad ri shangam *s ---UUal-.arbai farsilda shavam 'Akhir gar 'Ahanu gar sangam
--U U-U U -

al-darbain sdlim

hazaj akhrabal-sadrain

How long will he humiliate me ? How long will he lash me with this trunk (sic) ? In the end I shall be worn out, though I be of iron or stone. (Abri Shaktir)
3.41 I: - U U - - u U -- U UU UU - U UUsari' musaddasmatwi maksaf -U "-U

kiishish-i bandah sabab az bakhshish ast * kar-i qaid buidu turd 'aib nist Man's efforts are worthy of pity; (Muhammad b. Vaslif)
4. 15: UUU-UU-uu --U - U - uuU-U U-U-

it (this disaster) was the work of Fate, and was not your fault.

muthamman mujtathth makhbian

mara bi-jdn-i tu saugand u sa'b saugandi *

madhuf

ki hargiz az tu nagardam na bishnavam pandi

U U-

U-

U-

I swear by your life an oath, a solemn oath, that I will never turn from you nor listen to advice. (Shahid Balkhi)

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

83

4.116: u-u-u U U-

u--u-u-u
U U --

u-U -UU U -U UU-

makhbin muthamman mujtathth

hamah niyisha-yi kha-jahbi-niku'iy u bi-sulh ast * hamah niyisha-yi nidan bi-jang u fitna va ghaugha-st
U UU U -U U U

All my lord's intent is on goodnessand peace; all the intent of the ignorant is on war and rebellion and strife. (Abti 'Abdallah R-idaki) uu-u 4-413: -u u-uhizum khiham hami du 'amna zi jaudat *
-U UU U-

muthamman mankhir ma~twi munsarih

chrindu jarib i du khumm-isiki chainkhiin

U U-U U--

I seek from your generositytwo bundles of firewood,two hundred-weightof grain, and two jars of bloodred wine. (Abu'l-'Abbas Rabinjani)
4.407(2): - U U -U-

j-

U- - U-

matwi maksaf munsarihmuthamman

qand juda kun 'az fly -U U- - U -U

dir shav az zahr-i dand * - U -U -U U har chi bi-'akhir bih ast jan-i tura 'an pasand
UUU UU- U-

Put sugar far from him, keep away from the poison of wormwood; whatever in the end is best will be most pleasing to your soul. (Abil 'Abdallah Rfidaki) makhbzin mahdhaf khafif musaddas bi-muzhah dil zi man bi-duzdidi * 'ai bi-lab qaiiy ii bi-muzhgan duzd
4.511: U U U U U-U U -

U --

- - uuU

With your eye-lash you stole away my heart; your lips are the judge, and your eyelashes the thief. (AbUiSalik Gurgani) -- u - u 4.714: -- u u - uU akhrab muthamman muddri' makfif mahdhuf

yaram sipand agarchi bar atish hami figand * - UU U-U-U'az bahr-i chashm ta narasad mar vira gazand
U -U UUUU-

My beloved may throwwild rue on to the fire,lest the evil eye should bring her harm. (Ijanzala BNdghisi) - - u-u-akhrab muddri' al-sadrain --u-u-raftam bi-mah-i rauzah bizar-i marsmandah * ---u--U ta gfisfand 'aram farbih kunam bi-randah
4.707(2): ---u -uU -

I went during the month of fasting to the bazaar of Marsmanda to get a sheep and fatten it on clover. (Abu'l-'Abbas Rabinjani)

84
o: 5. II - - UU UU -

JOURNAL

OF PERSIAN STUDIES

akhrab al-sadrain hazaj musaddas al-hashawain maqbfid mahdhaf al-darbain

-u

dar janb-i 'ulyiv-i himmatat charkh *

U UU U-

mananda-yi vushm pish-i chargh ast --u U U -U -U

In comparison with your vaulting ambition the heavens are like a quail before a hawk. (Abfi Salik Gurgini)
5. I: - - U U - U - U- - hazaj musaddasakhrabal-sadrain

sdlim al-hashawain maqbfid al-darbain murghist khadang 'ai 'ajab didi * murghi ki hamah shikir-i 'i jinid
-U U -U UU U UU -

The arrowis like a bird, but pray have you ever seen a bird whose whole prey is souls? (Firiz Mashriqi)
akhrab 5.113: - - u U - U - U - - u u hazaj muthamman (- u) makfif majbfib(rubd'i) maqb.ad (3.313) 'ai qimat-i toi bi-siirat-i kivanjak * hasti tu bi-chashm-i har kasi bulkanjak
- -UU-U-U-----UU
-

U----

Your body is as bent as a cucumber; to every eye you look like a clown.
5.208(2):
-

U - U - U --U -

UU- U- U U -

makhban rajaz muthamman maztwi

-U

bi tu padid mikunam hIl-i tabih-i khish-rd *

U -U U U

td tu nasihati kuni
- U U- UU-

chashm-i siyah-i khish-rl

U UU -U

I reveal to you my distressedstate, so that you may admonish your dark eyes. (Nizimi Ganjavi) 5.316: u u - u - u - sar-i zulf-i ti! na mushk ast
U U U U -

u u - u - u - u bi-mushk-nabmdnad *
U U U U --U U -U U -U U -UU

ramalmuthamman mashkal slim al-'arid wa'l-darb

rukh-i raushan-i tu 'ai duist bi-'ift~b mdnad

--

The curls in your hair are not musk, and yet they are like the finest musk. Your shining face, O beloved, is bright as the sun. (FarrukhiSistdni) The complex Arabic terminology required to describemost of these metres shows how far they differ from the Arabic " originals ". Only four of the latter, indeed (hazaj, rajaz, ramal,and mutaqdrib), are in their sdlimor " sound " form in Persian-and there is general agreement that the mutaqdrib duplicated are hazajand ramal each applied to metres in three differentpatterns, and rajazto two; while even when the name is confined to a single pattern, it is not the same as the Arabic pattern. Finally, one may well ask how it was that, if the Persian poets were in fact copying the Arabic metres, they failed entirely to copy those most commonly used in Arabic (tawil, madid,basit,wdfir,kdmil).
at least is of Persian origin, and that if there was any copying it was in the other direction. The names

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

85

We are bound to conclude that these five patterns bear some relation to the natural speech-rhythms of Persian, and that their origin must be presumed to lie, even if it cannot be traced today, deep in the sources of the language itself. Their most characteristicfeature (and the one that distinguishes them most clearly from the Arabic patterns as described by Weil) is their careful balance and symmetry. This is particularlynoticeable in Patterns 3, 4 and 5, where every short syllable is balanced by a corresponding long syllable. This proportionis maintained even when two syllables are according to the rules replaced by one, since the latter will equal the quantitative value of the two for which it has been substituted. This does not, however, lead to monotony. Apart from licence to treat many specified syllablesin the language as either long or short,the operation of the substitutionrulesjust referredto can make a marked difference in the shape and rhythm of two lines in the same metre. In the comparatively simple mutaqdribmetre (I.III)

example, we may have a light, almost tripping verse of eleven syllables: pariy ai palang anjuman kard u shir
U-UU U -U

adopted by Firdausi (322/934-415/1025)

for the Shdhndma,for

He gathered together fairies, leopards, and lions. or a heavy, sonorous verse of only eight syllables: bi-padraidkardansh raftand pish
U -U - --u - "U -U

They went forward to bid him farewell. The more complex patterns allow of even greater variety; instances may be found in the examples syllables). In all these, the total quantitative length remains constant. It remains to be consideredwhether we can throw any light on the origins of these metres, now that we have satisfied ourselves that they are not copied from Arabic. Our task is complicated by the fact that, until recently, specialists in Iranian studies have tended to confine their researchesto one side or other of the Islamic watershed. So far as the examination of prosody is concerned, this has resulted in two quite distinct approaches. Islamic scholars, alerted by the work of al-Khalil, have recognized that New Persianverse is quantitative, but largely ignore questionsof stressand syllable; their colleagues on the other side of the fence are convinced that Avestan and Pahlavi verse is " syllabic ", and that neither quantity nor stress enter into the matter. Where it can be shown that a text is divisible into lines of approximatelyequal syllabic length, this is accepted as sufficientevidence that it is verse, and no further enquiry is deemed necessary. Thus two entirely different types of verse are observed, and it is easy to assume that there is no connection between them. Lazard, for example, considersthat the fragmentsof early popular verse quoted in workslike the Tdrikh-i Sistdn,which " are not composed according to the of quantitative metrics, on the Arabic model ", are to be regarded as relics of the traditional principles " accentual " system of pre-Islamic poetry." If, however, we recognize the quantitative poetry as a native growth, we are almost bound to assume an intimate relation with whatever went before. Let us begin by questioning the view that verse can be purely syllabic. The essence of rhythm is contrast-a regular " swing of the pendulum ". A constant syllabic length may or may not be a prerequisite of metrically balanced verses, but it is not enough by itself; some pattern must be imposed on this, whether of quantity (time lapse), stress,pitch, or even vowel colour. One problem is that, when we are consideringtexts available to us only in written form, we cannot do more than guess at the extent to which any or all of these are present. The rules for determining quantity in Persianverse have become stabilized over the course of time, but there are many ambiguities even at the presentday, and the early poetry we have been considering contains many examples of apparent breaches of the rules. We can
Lazard, cit.,Vol. I, p. 1o. xxGilbert op.
9

quoted above under 3.115 (i5 and

12

syllables), 4.407(2) (12 and 14 syllables), and 5.11I (9 and II

86

JOURNAL

OF PERSIAN

STUDIES

assume that the " rules " were even less carefully observedin earlier times, or rather, that the rhythmic pattern was not necessarily represented in the written script. In any case we may be mistaken in assuming that, because the later poetry was clearly quantitative, a purely quantitative pattern must be sought in its predecessors. The important thing is the pattern, and there is no reasonwhy a quantitative pattern should not derive from an accentual one, or vice-versa. The Pahlavi and even Avestan verses may well have had some such pattern, no longer discerniblein the form in which they have reached us. Apart from anything else, some explanation is needed for the fact that in these verses the syllabic length does not in fact remain constant; and we have already noted that this is also true of the classical quantitative verse. All this is not to say that Arabic poetry had no influence on Persian. In matters of form and structure, and even perhaps of rhyme, there may well have been some copying; but it is not reasonable to suppose that this copying could have extended to anything so fundamental as the speech rhythms. In point of fact, if we look back over the rather meagre remnants of pre-classical and pre-Islamic Persian poetry, we find that the scene is not quite so featurelessas might at first glimpse appear. The Russian orientalist N. Marr takes the view (for which he claims the support of Krachkovskyand Salemann) that the mutaqdrib metre is derived directly from the eleven-syllablemetres used for much of the Middle Persian poetry that has survived.12It is certainly true that this metre is unknown in pre-Islamic Arabic poetry, and is extremely rare in the poetry of the Umayyad period. It occurs more frequently in 'Abbdsidtimes, a fact that suggeststhat the copying was from Persianto Arabic. It would seem unlikely that Firdausi,who composed his Shdhndma a monument to the greatnessof Iranian culture,would have as chosen a foreign metre for the purpose, and the same applies to his predecessorin the task, Daqiqi. and the Benveniste has pointed out13 that parts of the two Pahlavi poems, the Ayydtkdr-i Zarradn Drakht-iAsarikseem to be in a metre very close to the mutaqdrib, he even quotes a verse from the and former that must have been a prototype for a verse by Daqiqi: hach En khdk khezet 6 kay gah nishinat (Ayydtkr-iZarirdn) tu zin khdk bar khiz u bar shau bi-gih (Shdhndma) Rise from this land and sit on the royal throne! Ibn Khurdadbih, author of the Kitdb al-masdlikwa'l-mamdlik 230/844), quotes a popular verse (c. (which he attributes to one Abu'l-Yanbaghi 'Abbds b. Turkhan), both the reading and the meaning of which are doubtful, but which neverthelessseems to have something of the mutaqdrib rhythm: samarqand kand mand 'az chich ta bihi bi-zinat ki 'afgand hamisha ta khuhi

The verses are said to be a lament for the ruin of Samarqand, and are interpreted by Zabihallih Safi as follows: " O prosperousSamarqand,who has brought you to this state? You are better than Chich, you are always good."'4 Another early verse of less certain authenticity is quoted by Daulatshdh in his Tagkiratal-Shu'ard (892/1487), on the authority of a late fifth/eleventh century poet, Abil Tihir Khittini, as having been inscribed on the ruins of Qasr-i Shirin. The metre is quite clearly mutaqdrib, though the original may well have been corruptedin transmission. The meaning is obscure:
12

N. Marr, " Vazn-i Shi'r-i ShihnSima ", Hazdra-yi Firdausf (Tihrin, 1944), p. 195 (Pers.). 13 Benveniste, op. cit., pp. 225, 293.

14

Zabilhall.h,

Safi, Tdrfkh-i Adabiyyat dar Iran (Tihrin, I956, 2nd ed.), p. 145.

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

87

hizhabrd bi-gaihdn antishah bi-zi jahdn-ra bi-diddr-i til shah bi-zi May the lion live prosperouson the earth! May the world gain a king at the sight of you! Finally, Christensensuggests that a fragment of a Zurvanite hymn detected in the Pahlavi book the Bundahishn Nyberg is in somethinglike the mutaqdrib metre, though Nyberg himself regardsit as being by nearer to the New Persian dabaitimetre (Pattern 2).15 Pattern 5 is even more clearly of Persianorigin.16 The most noteworthymetre in this class is that used for the rubd'ior quatrain (5.I I3), which is normally found (unlike other metres, where mixing of the patterns is the exception) in combination with the correspondingmetre of Pattern 3 (3.3I3):
--UU-U-U--UU-

--UU--UU--UU-

This metre (or pair of metres) is almost exclusive to the rubd'iform. Although the earliest extant example is by Shahid Balkhi (c. 225/840-3o10/922), even the prosodistsaccept that it was developed by the poets of Iran-though perhaps we should take with a pinch of salt the legend recorded by DaulatshTh that it was originally inspired by the excited cry of the small son of Ya'qfib b. Lais as he was rolling nuts into a ditch: ghaltin ghalt~n hamiravad ta lab-i gav
.- U-U-U U -

It goes rolling, rolling, to the edge of the ditch! Certainly no such metre is found in Arabic, even when the rubd'iform is consciously copied. On the other hand, according to Rempis, there is a discerniblerelationshipwith the metre used for some parts of the Avestan Yasna. If this particular metre then is Persian, it seems likely that the pattern as a whole is from the same source; other metres in the group are to be found even earlier than the rubd'iitself (Firiiz Mashriqi, Abti Salik Gurgani). Krachkovsky, quoted by Marr, finds similar metres in the Arabic poetry of the 'Abbdsid period (he quotes an example from Abu'l-Faraj Dimashqi, a 4th/ioth century poet), but is convinced that these were copied from Persian rather than the reverse, since they are not found in Umayyad times at all." There is an obvious resemblancebetween Pattern 5 and Pattern 4, the latter being the samesequence of syllablesin the reverseorder. Pattern 4 is in fact the favourite, accounting for nearly half of all poetry in Persianliterature, and one metre alone (4. I 15) being chosen for some 150% of the total. It is less easy to detect examples of it in pre-classicalpoetry. Some versesin the Drakht-iAsarik might be thought to fit: evom apartar hach t0 drakht asfirik I alone am superior to you, O Assyrian tree! The Tdrikh-i quotes a hymn recited by the Zoroastrianpriestsat the fire-templeof Karkily, which Sistdn Safa regards as syllabic,18but which could equally be fitted into a prototype of this pattern:
15

Arthur Christensen, Les Gestes des Rois dans les Traditionsde I'Iran antique (Paris, 1936), pp. 48, 53; H. S. Nyberg, " Ein Hymnus auf Zervan im Bundahishn ", ZDMG LXXXII (1928), p. 225. 16 L. P. Elwell-Sutton, " The Rubd'i in Early Persian Literature ", Cambridge History of Iran, Vol. IV (Cambridge, 1975),

pp. 633-57. 17 C. Rempis, " Die Metrik als sprachwissenschaftliches Hilfsmittel in Alt-iranischen ", ZDMG CV (1955), pp. *64*-*65*; Marr, op. cit., p. 197. 18 Tdrikh-iSistdn, pp. 35-7; Safd, op. cit., p. 143.

88

JOURNAL

OF PERSIAN

STUDIES

frukhta bddhi rufsh khunidha garshasb huish,etc. May the light be kindled, may the soul of Garshasbbe praised! Both patterns seem to be related to, possibly even derived from, Pattern 3; the alternation characteristic of the rubd'imetre is also found, though not so normally, in other metres of both patterns. Several pre-classicalverses show the features of Pattern 3. The Tdrikh-i Sistan,as well as three Arabic sources, attributes to one Yazid b. Mufarrigh, in about the year 6o/68o, the followingjingle: 'usar~t-i zabibast sumaiya roisbidhast There's water and date-wine; there's dregs from dried raisins; there's fat and a sheep's fat tail-and Sumaiya is a tart! dbast u nabidhast u dunba farbih u plyast In IO8/726 a defeated general is said to have been thus satirized by the people of Balkh: ba-ri tabdh madhiya bi-dil fardz madhiya He's come back from Khutldn; he's come with a sour face; he's come back on the run; he's come down sick at heart. az khutlan amadhiya avar baz madhiya Several different versions are found in the sources of a verse attributed to the Sasanian king Bahrim Gar. The attribution is unlikely, but the baititself must be quite old, probably in the following form: manam an shir-i shalanbah u manam babr-i yalah I am that lion of Shalanbah; I am the roaming tiger. The verse attributed to Abu IHafsSughdi (4th/Ioth cent.), but probably of much earlier date, also fits Pattern 3: ahuy-i krhi dar dasht chigUnahdavadhi ydr naddrad bi yir chig-inah davadhi How does the mountain deer run in the valley ? He has no companion, how can he run without a companion? Braginskiifinds some resemblancesbetween this and a modern popular poem.19 Lastly, we may note the curiosity that, according to the Greek scholiast Cheiroboscus, the metrical foot lonicusa maiore (- - U U) was known as the persikos;and certainly Aeschylusused this pattern for the opening section of his play ThePersians.20 Pattern 2 comes nearest to an Arabic prototype; but even this has its Iranian forbears. Still commonly used today for the popular dabaiti,a quatrain form known in many Persian dialects, is metre and other literary forms. Adib TUsi differs from 2.11 ; and it is equally frequent in ghazals, masnavis Asirik can be scanned in this way; while Benvenistein consideringthat most of the versesin the Drakht-i is a hymn to Zurvan found in the Bundahishn also of this type.21 Nyberg suggeststhat
21 Adib-i Tfisi, " 19 Braginskii, Iz Istorii Tadzhikskoi Narodnoi Poezii Danishkada-yi Tarina-hR-yi Mahalli", NashrDya-yi (Moscow, Adabyadt-iTabriz, V/2 (1954), PP. I38-68; Nyberg, op. cit., 1956), p. 260. 20oSee Ulrich Wilamowitz-Moellendorf, Griechische Verskunst p. 225. (Berlin, 1921), p. 336n.

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

89

We must now consider the other distinctive feature of Persian verse-rhyme. As with the metres, it has long been assumed that this too was derived from Arabic. It is certainly the case that rhyme is the exception rather than the rule in Middle Persian verse, though examples of it are by no means cited by Nyberg is one instance, while many of infrequent; the Zurvdnite hymn from the Bundahishn the verses in the Ayydtkdr-i Zarirdnshow, if not strict rhyme, assonance (bushn-burz, mikh-arvish, zarir-kunit, etc.). Nearly all the examples quoted above of pre-classical verse also use rhyme-the Karkily hymn, the Yazid b. Mufarrighjingle, the lament for Samarqand. However all these pieces are too short to enable us to judge whether the most characteristicfeature of Islamic verse, the monorhyme, was present. This is certainly found in pre-Islamic Arabic poetry, and also in the earliest examples of" classical " Persian poetry. Christensensuggests that the monorhyme at least must have been copied from Arabic, since only that language has the facility of forming words of similar pattern in sufficient quantity to sustain the rhyme throughout poems of the length of the qasida(see below).22 But in fact, investigation reveals that the overall ratio in rhyming words is four or five Persian to one Arabic, while even in long qasidas, depending on the rhyme chosen, the ratio is rarely lower than two to one. For instance the philosophical qasidaby Abu'l-Haisam Gurgani (early 4th/Ioth cent.), rhyming in -dr,has 6o Persianrhymes to 30 Arabic. It is also worth noting that certain archaic pronunciationsfor instance, khvard, khvash, etc., for later khurd, khush--which were already dying out in late Sasanian times, were retained by convention in poetry for many centuries; a possible implication is that such rhymes as zard-khvard had become well-established at least in Sasanian times. Although the grammarians and prosodists evolved an elaborate terminology to describe all the elements in the rhyme, in practice all that needs to be said is that the rhyming element in the rhyming word may consist of anything from one to four syllables (five is theoreticallypossible, but is not actually found). Several examples of monosyllabic (e.g. 4.407(2) and 4.714 above) and bisyllabic (e.g. 2.316, the longer rhymes:
3.307(2),

4.115, 5.113) rhymes have already been given. The Shdhndmaprovides plenty of examples of

na gasht-i zam~na bi-farsdyadash na 'in ranj u timar bagziyadash (3 syllables) Neither the passing of time wears it out, nor does pain and sorrow injure it. bidin gfina 'az charm-i pilyandagan bi-piishid bildy-i griyandagan (4 syllables) In this manner, with the skin of running creatures, he clothed the bodies of speaking creatures. One feature of Persian rhyme is certainly of native origin, the radif. In this technique the rhyming word proper, which is set back in the verse, is followed by an additional word or words repeated without change in each rhyming line. This repeated element may be of any length; indeed prosodistshave composed verses in which only the first word changes from line to line. Here is a 7th/I3th century example: 'ai dist ki dil zi banda bar ddshta-i nikfi-st ki dil zi banda bar dishta-i O beloved, you who have stolen my heart from me, It is good that you have stolen my heart from me !23 and 5.316, and in the following verse by Mus'abi (5th/I Ith cent.):
2a

More normal examples may be seen in the verses quoted under metrical patterns 2.11 I, 2.116, 5.20o8(2),

Christensen, op. cit., p. 55.

28

b. Shamsal-Din Mulhammad Qais al-Rizi, op. cit., p. 195.

90

JOURNAL

OF PERSIAN

STUDIES

shakkar-shikanast ya sukhan-gfiy-iman ast 'anbar-zaqan ast ya saman-btiy-i man ast Is it a sugar-tongued one or my eloquent one? Is it an amber-chinned one or my jasmine-scented one ? This last verse also contains an example of another type of rhyme, in which the word preceding the actual rhyming element-in this case g/b7ay-i-also rhymes. Another common, rather looser form of assonance is exemplified in the following verse from the Shdhnima: judi kard g~v ii khar Uig-ispand bi-varz avarid ~nchi bud sfidmand He distinguished cow, ass and sheep; he put to good use whatever was useful. It may be noted that in the satiricaljingle from Balkh referredto earliera form of radifisused (dmadhfya), tabdh,bdz,fardz-nizdr in another version). preceded by assonant rather than rhyming words (khutldn, Some of the later poets employed a variant of the radif,known as the hdjib,in which the repeated word or phrase precedes the rhyming word: guftam gul ast yd saman ast in rukh uizaqan gufti yaki shigufta gul ast Uiyaki saman guftam dar an du zulf shikan bish yAgirih gufta yaki hamah girih ast fi yaki shikan I said, Is that face and chin a rose or jasmine ? She said, One is a blooming rose; the other, jasmine. I said, In those two locks of hair are there more curls or ringlets? She said, One is all ringlets; the other, curls. (Farrukhi Sistani) The surviving works of the earlier poets of the Islamic period are so fragmentary that it would be difficult on the strength of them to form any picture of the types of poem composed by them. It is only when we move into the 5th/I Ith century that we find collections of the work of individual poets sufficiently complete to enable us to classifytheir poems accordingto type. By this time, as with the metres, the poetic forms had settled down into the mould that they retained until modern times. They may be divided into three categories-those in which each couplet has its own rhyme, those in which there is no internal rhyme in the couplet but the same rhyme is preservedat the end of each couplet throughout the whole poem, and stanzaic poems in which the rhyme varies to a greater or lesser extent with each stanza. The unit of any poem is the bait or couplet, divided into two hemistichs (misrd')each of the same metrical pattern and length. In the magnavi rhyming couplet form each mrird' or rhymeswith its partner, and the rhyme changes for each bait (indeed it is considered bad style to repeat the same rhyme too is frequently). The masnavi thus capable of expansion to any length, and is thereforepre-eminentlysuitin able for epics, romances, and didactic, philosophical and mystical poems. The best known masnavi Persian literature-the Shdhndma Book of Kings of Abu'l-Qasim Firdausi, begun by his predecessor or Daqiqi (c. 330/941-370/981)-runs to over 50,000 couplets. This was by no means the first of its kind, and indeed was certainly based on Pahlavi sources-though whether these were in verse-formwe have no means of knowing. Three verseshave survived-somewhat corrupt-of a Shdhndma composedtowards the end of the 3rd/9th century by Mas'fidi Marvazi, and isolated verses by Rabinjani (fl. 331/943) and

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

91

(d.329/940-I) appear to have come from poems of the same class. The earliest surviving fragRfMdaki ments of masnavi poems, however-those by Shahid Balkhiand Faralavi-seem to be ratherfrom romantic narratives. According to the prosodists,it is possible to tell the nature of a masnavd poem from the metre in which
it is composed: metre I.i I I is used for razm (heroic epics) and bazm (festive poems), 2.111 and 5.o10 for 'ishq (love poems), 2.4II, 3.111, and 3-411 for pand (homilies) and tasavvuf (Sfifism), and 4.511 for

form, though examples in bazm. These seven metres are the only ones normally used for the masnavi other metres are found by both Rtidaki and Daqiqi. Nor does it seem that these conventions were established in Sdmdnid times. Abfi Shakiar's philosophical Afarin-ndma (336/947-8) is in metre I.I I I, while IHakimMaisari's medical treatise, the Ddnish-nama-the earliest full-length masnavi have-is we in 2.11 I. Rfidaki chose 2.411 for his version of Kalila vaDimna. Moreover, at least two romantic poems u from Ghaznavid times were composed in metre I.1II- 'Unsuri's Vdmiq 'Agrd and 'Ayy-iqi's Varqa va Gulshdh.These latter, like two other poems by 'Unsuri of which virtually nothing has survived u and Khingbut Sorkhbut) Visu Rdmin Fakhr al-Din Gurgani (composed u and by (Shddbahr 'Ainal-haydt
433/1040-446/1054),

origin, and are typical of the " star-crossedlovers " theme that reached its height in the romancesof the metres are of the short 6th/I2th century poet Nizdmi. It will be seen that all the conventional masnavi eleven-syllable type (except for 5.1IO0,which has ten). This strongly suggests a connection with the Middle Persian poetry of the Pahlavi books, where the eleven-syllable line is also the norm, though the metrical pattern is less clear. The monorhymeforms, on the other hand, tend to use the longer lines, though there is no restriction on choice such as is observedin the case of the masnavi.The " standard" form of the monorhyme poem retains the same rhyme at the end of each couplet, and also internally in the first couplet or matla'. However, most poets seem to have composed poems in which this initial internal rhyme does not occur, and such pieces are known as qita'dtor occasional verses. The " standard " forms include the qasida and the ghazal, and the rubd'imay also be considered as falling within this cateogry. The qasida a comparatively long poem, ranging according to the authoritiesfrom thirteen couplets is to as many as two hundred. The three qasida-writers the early period whose works have come down of to us in a bulk sufficient to allow us to draw general conclusions-'Unsuri, Man-ichihri and Farrukhi, all poets of the Ghaznavid regime-seem to have favoured poems of 40 to 50 couplets, though Mantichihri has a fair number of shorter ones (if they are complete), while Farrukhi has one, celebrating to 175 couplets. It seems to have been in Ghaznavid times that the qasidadeveloped its characteristic form. Only two earlierpoems of any length (other than masnavis) survive, the philosophical and didactic of Abu'l-Haisam Gurgdni (early 4th/Ioth century) running to 89 couplets, and an incomplete qaida lyric and possibly panegyric poem by Daqiqi amounting to 55 couplets. All the elements of later Persian poetry are present in embryonic form in the fragments from Sdmdnid times-panegyric, lyric, didactic, elegiac, satirical; but they appear to be treated individually rather than within the framework of a formal scheme. This is perhaps characteristicof the period, which was permeated with a spirit of free thought and scientific enquiry. With the arrival of the Turks during the 4th/Ioth century, and more especially with the establishmentof the Ghaznavid dynasty after 365/975, religiousorthodoxy and
authoritarianism began to prevail, and with this a great degree of formality and artificiality in poetry. gained a reputation as a great patron of poets and writers, but his motives Mahmid (388/998-421/Io30) were not purely altruistic; he employed them as propagandists for the glories of his empire. The best succeeded in rising above this limitation. 'Unsuri was the panegyrist par excellence,but his poems are replete with historical, scientific and religious allusions. Farrukhi's verses are simpler and more lyrical; Mantichihri excels in descriptions of nature. But all qasdasby this time were primarily panegyric in purpose. is The qas~ida generally believed to have been of Arabic inspiration, and certainly of all the Persian poetic forms this comes nearest to its Arabic parallels. It normally falls into two parts-the nasib, consisting usually of the tashbibor taghazzul (erotic prelude), and the maqsad,the central element of which the is the madihaor panegyric. In the Arabic qasida theme of the tashbibis nearly always a lament over Mahmfid of Ghazna's successful campaign against Sfimnat in India (415/1024-417/1026), that runs

are literary versions of popular folk-tales, some of Persian and some of Arabic

92

JOURNAL

OF PERSIAN

STUDIES

the deserted encampment of the beloved, a natural Arabian setting that was sometimes copied rather open with a lyrical self-consciouslyby the Persianpoets. For the most part however the Persian qasidas Persianfestivals (both purely Persian description of spring or autumn, or a welcome to one of the great themes); or the poet may praise his mistress,or the joys of wine-drinking. At the transitionalpoint in the poem, the takhallus guriz-gdh, poet turns his attention more or less abruptly to the object of his or the eulogy, and thereafterthe merit of the poem lies in the ingenuity with which the changes are rung on the accepted laudatory images and epithets. By comparison with the qasida,the ghazal is a much less formal, more individual means of poetic expression. It is a short poem in lyric form, with anything from five to seventeen couplets, and like the qasidais unrestrictedin choice of metre. Later on, in the hands of masters like H~fiz, it became the vehicle for the noblest expressionof Stfif mystical experience; but in the early period Siafismwas not yet a feature of Persian poetry, though much of the raw material was already present. For instance, in the fragmentsattributed to Daqiqi we find referencesto wine and the cupbearer,the Magian contrasted with the orthodox believer, the beloved described as an idol (but,sanam).24 The origin of the ghazal has been the subject of much speculation. Traditionally it has been reand in favour of its Arabic origin garded as stemming from the erotic section of the qasida(the tashbib), is the name, which appearsto be Arabic, though it is interestingto note that the firstrecordeduse of it in who if he existed the sense of " love-song " occurs in a poem attributed to the Umayyad poet at all was probably of Persian origin.25 Yet this form is so typically Persian that one hesitates to accept Wadd.ah, its foreign derivation without question. The Tajik scholar A. G. Mirzoev takes the view that the form in general was based on popular folk-songs, but that its development as a technical form, which was not complete until the time of Sa'di (7th/I3th century), may possibly have been influenced by Arabic models.26 If so, we may assume that the ghazalsof the Sdmdnidand Ghaznavid period owed nothing to alien sources. However, it was not at this period as popular as it later became. Farrukhi, generally looked on as the master of the ghazal at this period, has no more than twenty-eight in his divdn,while Manfichihri and 'Unsuri number barely a dozen each.27 The following ghazal by Farrukhiis characteristicof the genre, with its naturalisticimagery and its pessimisticoutlook. Gardener, why do you sit beneath the tall cypress? Is there no space but where the cypress stands? Other trees will give you shade enough; Why not sit beneath the jasmine bough ? Your garden is full of shady trees; Go and choose whichever one you wish. Do not linger near that half-grown cypress! Forget the bright hues of that new-grown cypress! Free me from the toils of that tall cypress! Have mercy on this poor tormented lover! The rubd'i distinguished from all other Persian verse-formsby the fact that it is strictly confined is as to both metre and length. The classicalform consistsof only two couplets, the first of which (as with all monorhyme forms) has an internal rhyme as well as rhyming with the second. The name rubd'i (quatrain) however suggests that it must originally have been seen as a four-line poem, and indeed metres examples are common in which all four hemistichsrhyme. Although single couplets in the rubd'! are known as early as Shahid Balkhi and Fardldvi, the earliest complete quatrains are those attributed
2' 25

Pagliaro, op. cit., p. 244. 1957). R. Blachre, " The Ghazal in Arabic Poetry ", ElP, Vol. II, 27 Dfvdn-i HakfmFarrukhf Sfstdnf,ed. 'All 'Abd al-Rasili (Tihrin, der p. Io28; Carl Brockelmann, Geschichte Arabischen 1933); Dfvdn-i Ustad ManachihrfDdmghdn!,ed. Muhammad Litteratur, Supplement-band I (Leiden, 1937), PP 3o, 82. Dabir-SiyRqi (Tihrqn, I948; Divdn-i Abu'l-Qdsim Hasan b. 26 Abdulgani Mirzoev, Radakf va Inkishdf-i Ghazal (Stalinabad, ed. Yal.by Q arib (Tihrin, 1944). Ahmad 'Un.urf,

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

93

to Rfidaki-though the authenticity of many of these is doubtful, for only one occurs in a source earlier than the Ioth/I6th century, and even that is in a 7th/I3th century work: vajib nabuvad bi-kas bar aftil u karam vajib bashad har ayinah shukr-i ni'am taqsir nakard khaja dar na-va-jib man dar vajib chigfina taqsir kunam Bounty and generosity are not binding on any man; What is always binding is gratitude for favours. My lord has not failed in that which is not binding; How then shall I fail in that which is binding ?28 The rubd'idid not really come into its own until the end of the Ghaznavid period with the appearance of the mystical poet AbUiSa'id b. Abi'l-Khair (357/967-440/Io48).29 Although in later centuries many spurious quatrains were attributed to him, the following is probably among the authentic ones: dar dida bi-jay-i khab 'ab-ast mara zir~ ki bi-didanat shitib-ast mard gilyand bi-khib ta bi-khab-ash bini 'ai bi-khabaran chi jiy-i khib-ast mara In my eyes tears ('db) have taken the place of sleep (khdb), Because I am in desperate haste (shitdb)to see you. They tell me, ' Sleep, so that in sleep you may see her (Him).' O misguided ones! What sleep is there for me? Both these examples illustrate the epigrammatic quality of the rubd'i,with its extreme conciseness and the " sting in the tail ". Its brevity, neatness and metrical flexibility made it particularlyappealing as a vehicle of expressionto those who were not professionalpoets. The quatrain form was not always restrictedto this type. The Balkh verses quoted earlier, and those Yazid b. Mufarrigh and Abu'l-Yanbaghi are examples of a looser mould, while both the poems by attributed to Badghisi are quatrains, but in metres 4.511 and 4.714. One metre, however, which has remained a favourite in popular poetry till the present day, is 2.1 I I; quatrainsin this metre .Hanzala are known as dabaitis, and one of the most famous practitionerswas Bdba Tahir 'Uryan (early 5th/I Ith a dervish poet of Hamadan, whose verses are still the stock-in-tradeof the dervishfraternityin century), Iran: gil-i 'idam sirishti dar chihil rfz ki ta takmil gasht fi shud dil-afrfiz riyaiat bayad i-ra ta chihilsal ki ta gardad qarinash bakht u piriiz (0 God), Thou formedst man of clay in forty days, Till he was perfect and rejoiced the heart. Yet he must train himself for forty years, Till he achieve good fortune and success. The stanzaic forms seem to have been an innovation of the Ghaznavid poets. Mantichihri uses the in musamma.t, which each stanza consists of six hemistichs, the first five rhyming independently in each while the last has a common rhyme throughout the poem, which may run to as many as 35 stanza,
28 Shams

al-Din Mulhammad b. Qais al-Razi, op. cit., p. I75; Nafisi, op. cit., Vol. III, pp. o037-47.

29 Sukhanan-i Manpzim-i Aba Sa'id Abu'l-Khair, ed. Sa'id Nafisi

(Tihrin, 1955).

94

JOURNAL

OF PERSIAN

STUDIES

stanzas (band). In one poem the independent rhyme is used for all six hemistichs, so that there is no rhyme-linkbetween the stanzas. Variants used by later poets include stanzas of from three to ten hemistichs (though the number must remain constant throughout the poem), and rhyme-schemes such as aaabb, cccbb, dddbb..., or aaabb, cccdd, eeeff... Farrukhi's divdncontains three samples of the form; in this the stanza consists of from five to ten couplets, using either the normal monotarji'-band rhyme rhyme-schemeor a common rhyme for each hemistich, with the exception of the final couplet of each stanza, which has its own rhyme and is repeated as a refrainat the end of each stanza throughout in the poem. A variant on this, of somewhat later date, is the tarkib-band, which the final couplets differ or are merged with the rhyme-scheme from stanza to stanza, and either have their own common rhyme of each stanza. The poetry of Iran, like its art and architecture,has long been characterizedby a love of formal ornamentation and decoration, and the art of poetics has been a favourite study of prosodists. The earliest book in Persian on this subject is the Tarjumdn al-Baldghaof Muhammad b. 'Umar al-Raduyamni. Though nothing is known of its author, and the sole extant manuscript (itself of very early date) gives no indication of the date of its composition, it seems clear that it must have been written shortly after 481/1 o88.30 Certainly its poetic examples are all, so far as their authors are identified, drawn from the poets of the Sdmdnid and Ghaznavid period, and it therefore provides us with a useful survey of the poetic techniques and figures of speech in use at the time. The book contains sections on seventy-three such figures, some of them subdivided still further. Here we can refer only to some of the more noteworthy. These may be conveniently grouped into categories according to type (these are not the categories used in the book itself). One type is based on orthography and spelling; so we may have verses that are muqa.tta', composed entirely ofjoined letters. composed entirely of unjoined letters, or muwassal, or A poem may be composed without using a particular letter (mujarrad); two parallel couplets may be using the same letters, but with different dots or vowels, so changing the meaning (musahhaf). composed a kind of visual punning. The pun is indeed found in a More often this last device is found in humorous number of different forms, and has nomud.dra'a,associations. Tajnisincludes many different types, all based on resemblance between words of different meanings: 'ab-ast u za'faran hasad-i t i ki h5sidat bar chashm chashma darad u bar chihra za'farin Envy towards you is both water and saffron, for your envier and in has a spring of water (chashma) his eye (chashm), saffron (the colour of woe) on his face. ('Unsuri) are Anagrams and palindromes (maqliabdt) often found: ki 'az dad didi buzurgiy u riaz yaki pAdshabUd dar nim-rfiz andarash sdkhta khdsta bi-jangandarash lashkar ardsta bi-ganj There was a king in the East, whosejustice brought him greatness and good fortune; his treasury was given over to gifts, and his army was ready for battle. ('Unsuri) Sometimes rhyming puns are found: shahi vaqf karda bar dmal mil chun fi nai bi-mardi kasi z-il-i zil
al-Baldgha, so al-Ridfiyini, Mulammad b.'Umar, Kitab Tarjumdn ed. Ahmed Ate? (Istanbul, x949); Ateg, "Tarcumin alBaldga, das friiheste neupersische Werk uiber rhetorische Figuren ", Oriens,Vol. I/I (1948), pp. 55-7.

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

95

A king whose wealth is devoted to the satisfactionof hopes; there is none like him in generosity even from the family of ZIl. (Yazddni) Ingenuity in this field extends to acrostics,word squares, " circular " verses, and the like; but these are the products of prosodistsrather than of poets, who neverthelesswere not far behind in their delight in clever play with words. 'Unsuri, who was much regarded for his skill in this respect, wrote verses like the following: agar chi bashad tanha hamah jahin bd 'fi-st va gar chi bd 'fl btshad hamah jahdn tanhd-st Even when he is alone, the whole world is with him; And even when the whole world is with him, he is alone. Tarsi'is the art of composing verseswith parallel arrangementsof words; the verse by Mus'abi quoted earlier is one example, and the following is by a poet of the Ghaznavid period, 'Abd al-Jabbir Zainabi: riiz-i bazmat namdird fikhtah 'anbiz-i baz rfiz-i razmat kdmktirt shir shdgird-i shabdn On the day of your feasting, O illustriousone, the ring-dove is the partner of the hawk; on the day of your warring, O mighty one, the lion is the servant of the shepherd. In the figure known as siydqatal-a'ddd poet contrives as long a list of words as possible: the dfist~n fi dushmanin-ri riiz-i bazm ii rfiz-i razm sh~nzdah chiz ast vaqt-i kdm kardan vaqt-i kdr n~m u nang Rifakhr u 'ar ii 'izz u zull i nfish u zahr shddiy Uigham sa'd u nahs fi taj u band Ritakht u ddr For friends and foes on the day of feasting and on the day of battle there are sixteen things at the moment of pleasure and the moment of strife-fame and disgrace, glory and dishonour, nobility and baseness, a joyful toast and a poison draught, happiness and grief, felicity and misfortune, crown and fetters, throne and gibbet. al-Baldgha) (Tarjumdn These verses also provide examples of mutadddd, combination of opposites, and of tafsir (separation the of the subject and its complement, e.g. friends-feasting-pleasure-fame, etc.) and taqsim (distribution of attributes). In the figure i'ndtthe poet displays his skill by repeating the same word in different contexts: mukh5lif~n-i tu mlirdn budand mir shudand bar avar az sar-i mflr~n-i mir-gashta damdr makun darang 'az in bish u raizgarmabar ki 'azhdaha shavad ar rfizgir yabad mdr Your enemies were ants and have become snakes; bring down destruction on the heads of these ants turned snake. Delay no more and waste no time, for if it is given time the snake will become a dragon. (Mas'idi) Under the headings of isti'dra,kindya are and tashbik included various types of metaphor and simile:

96

JOURNAL

OF PERSIAN

STUDIES

khudayagind farkhanda mihragn aimad zi bdgh gasht bi-tahvil-i 'aftib ahvil sarRy-parda-isuhbat kashid sib u turunj bi-tabl-i rihlat bar zad gul fi banafsha duvil 0 Lord, the joyful autumn has come! With the passage of the sun the state of the garden too has changed. The apple and the orange have drawn across the curtain of privacy, while the rose and the violet are beating the drum of departure. (Munjik) In later periods the metaphor was to become a normal and even hackneyed element in Persian poetry, and in this early stage too we find many expressionsused in a way that makes it clear that their metaphorical meaning was fully understood by the listener. Thus the eye is already called the narcissus or khurshid);the cheeksare disguised as tulips (nargis) the magic spell (jddfi),and the face the sun (dftdb, the hair as chains (zanjir)or chain-mail (zirih), the teeth as pearls (gauhar, the khushdb), lips as (ldla), rubies (la'l, ydqiit) or jujubes ('unndb), and the skin as ivory ('dj) or silver (sim). A beautiful woman is a painting (nigdr),an idol (but,sanam),a Turk (turk),a moon (mdh,qamar),or a houri (har). She is as tall as a cypress (sarv)or a teak-tree (sdj), above which her face shines like the moon; her hair is as and black as the night (shaba-rang) scented with musk (mushk) amber ('anbar). Meanwhile her woeand lover is saffron-coloured(za'fardn) from grief; the tears in his eyes are the blood of his begone (khdr, zdr) he is tightly bound by the chains (band, heart (khlin-i dil), but halqa)of her hair. The hero-figureon the other hand is elephant-bodied (piltan)and lion-hearted (shirdil) in battle he lays the dustwith the blood ; of his enemies, and scattersheads as freely as he does gifts in peace-time. In the next category of poetic figures come those that depend on play of syntax and subject-matter. the Especial ingenuity is called for in the manipulation of the takhallus, bridging section in the qasida between the introductory erotic passage and the main, panegyric part of the poem: khujasta bishad rtiy-i kasi ki dida buvad khujasta-riiy-ibut-i khish bamdad pagih agar nabfidi bar man khujasta didan-i til khuddy shid nakardi mara bi-didan-i shah Happy is the face of one who has seen the happy face of his beloved in the early morning; if I had not had the happiness of seeing you, God would never have blessed me with the sight of the king! (Farrukhi) The term tafsirhas already been mentioned; here is another example in which the objects are listed in the first couplet, and the correspondingverbs in the second: hamah nIm-i kinshin bi-parkhish-i mard dil-i jang-jily u basij-i nabard hami tfikhtand u hami takhtand hami sfikhtand u hami sakhtand A name for vengeance, an assault on the enemy, The heart of the warrior, the armour of battle, They have gained, they have launched, They have inflamed, they have donned. ('Unsuri) Although originality in the broadersensewas frowned on even at this early stage in the development of the poetic arts, originality in the handling of conventional material was matter for praise. The author of the Tarjuman al-Baldgha quotes this verse by 'Unsuri as an example of ibdd' (invention):

THE

FOUNDATIONS

OF PERSIAN

PROSODY

AND

METRICS

97

hami bi-kushti ta 'ddami namand shuji' hami bi-dadi ta 'ddami namand faqir You killed until not a brave man was left; you gave until not a poor man was left. Rather similar is ta'ajjub(surprise): bi-gah-i khanda nami'i hami du saf gauhar tur~-st gauhar u chashm-i man ast gauhar-batr When you smile you display two rows of pearls; you have pearls (teeth), and my eyes too are filled with pearls (tears). (Qamari) Eulogy of the patron or the beloved provided opportunities for ingenious twists of meaning, in which ta'kidal-madh seeming disparagementor reproach is deftly turned to praise (istidrdk, bi'l-dhamm): asar-i mir nakhdham ki bi-mdnad bi-jahan mir khiham ki buvad mdnda bi-jiy-i 'asard I do not want any trace of my prince to remain on this earth; I want my prince to remain instead of his memorial. (Rildaki) bi-zulf kazhzh valikin bi-qadd u blla rast bi-tan durust valikin bi-chashmagan bimdr Twisted as to hair but straight as to stature; Healthy as to body but languid as to eyes. (Raidaki) Among other more elaborate figures of speech developed during the Ghaznavid period were ilghdz and muhdjdt propounding of riddles), su'dlujawdb (question and answer), mulamma'dt (macaronic (the poems consisting of alternate Arabic and Persian verses), and tadmin(the incorporation of verses by other poets (barsabil-imihmdn, duzdida," as guests, not plagiarized ", as the author of the Tarjumdn na al-Baldgha puts it). Many poets liked to use paraphrasesof Qur'anic verses, sayings of the Prophet and other traditions, folk-tales and proverbs. Such " plagiarism " was not frowned on but rather admired, the point lying rather in the skill with which the borrowed material was blended into the body of the poem. Indeed, much of the beauty and effectivenessof Persian poetry lies in what is not expressed; the traditional imagery carries with it a vast substratum of associated ideas, familiar at least to the educated reader, and it is the interplay of these ideas, rather than the elegant but conventional surface, that has given the best of this poetry its assured place in the imaginative literature of the human race.
SUPPLEMENTARY BIBLIOGRAPHY OF WORKS NOT LISTED IN THE NOTES Agha Ahmad Ali and H. Blochmann, A Treatise on the Rubd'l entitledRisdlah i Tardnah(Calcutta, 1867). Shi'r-i Fdrst (MashMuhammad Taqi Bahdr, Tdrikh-i Ta.tavvur-i had, 1955). H. Blochmann, TheProsody thePersiansaccording Saifi,Jami and to of otherwriters(Calcutta, I872; reprinted Amsterdam, 1970). E. G. Browne, Literary History of Persia, Vol. II, pp. 19-89. (London, i906, reprinted Cambridge, 1928). Mas'uud Farz-Ad, The Metre of the Rob$a'ii (Tihrdn, 1942). PersianProsody(Coventry, 1966). Idem, Concise Idem,PersianPoetic Metres (Leiden, 1967). G. W. Freytag, Darstellungder arabischenVerskunst (Bonn, 1830; reprinted Osnabrtick, 1968). on Francis Gladwin, Dissertations the Rhetoric,Prosodyand Rhymeof the Persians (London,i 8o ). Mahdi Hamidi, 'AriL2-i Hamidi (Tihrfn, I963). P. N. Khdnlari, Vazn-i Shi'r-i Fdrsi (Tihrdn, I958). A. F. Motamed, De la mitrique(Tihrin, 1962). PoetikundRhetorik Perser(Vienna, der Friedrich Rickert, Grammatik, 1827-28; Gotha, 1874; reprinted Osnabriick, 1966). Jan Rypka, " La metrique du Mutaqarib ", Hazdra-yi Firdausi (Tihran, 1944), PP. 75-87. Joseph Garcin de Tassy, Rhitorique et Prosodie des Langues de I'OrientMusulman (Paris, 1873; reprinted Amsterdam, 1970). Gotthold Weil and G. Meredith-Owens, " 'Arfil ", in El2, Vol. I, pp. 667-77.

THE BRITISH MUSEUM MIRZANAMA AND THE SEVENTEENTH CENTURY MIRZA IN INDIA1 By Aziz Ahmad
I The British Museum Persian MS. Add. I6, 817, which is a collection of nine treatises and collections of letters, contains in its folios 89b to 95b, a manual of manners called Mirz~ndma an by c. anonymous author. The MS. was written in Shikasta-dmin I 152/1739;2 but the treatise itself was almost certainly written a few decades earlier. This is borne out by the date of completion given as one Io070/66Oin the only other known copy of this Mirzdndma, in the library of the Asiatic Society of Bengal.3 but These are two of the three manuscriptsnoted by D. N. Marshall4under the title Mirzdndma; the third of these is quite a different work, the title of which bears the name of its author as one Mirza Kamran, not to be confused with BTbur'sson. It was found in the private collection of Muhammad Mahmiid of Chakdah in the Nadia district of Bengal by M. Hidayat IHusaynand published with an English translationby him in TheJournalof theAsiaticSociety Bengalin 1913.5 The date of the compoof sition of Mirzi Kdmrin's work has been suggestedby Hidayat Husayn as c. I o17/1608, which is the date on which Mirzi Rtfi' Shirazi, the author of the Tadhkirat al-mulak, began work on that book. This date seems erroneous,judging by the internal evidence of a comparison of the contents and style of B.M./ A.S.B. anonymous6Mirzdndma and Mirzd Kimrdn's Mirzdndma.Compared to the former, the latter shows a great deal of vulgarization in theme as a manual of manners, as well as in its style, which is comparatively more demotic and inferior. The date suggested for the composition of Mirzi Kamrin's work by Hidayat IHusaynrests solely on his assumption that the Mirza Rafi' Shirazi mentioned by which is al-mulak;7 Mirzi Kdmrdn as his contemporary is the same person as the author of Tadhkirat doubtful. Mirzi Rdfi' was a common name; and two persons of that name from Shiraz could have made India their home at different times. Both the Mirzdndmas, of the B.M./A.S.B. manuscriptsand that of Kimran, define themselvesas that of the criteria and conduct for the gentleman who called himself a mirzain the seventeenth manuals8 century; this person will be studied in some detail in the latter part of this article. In both cases, their have assumed the authors alleged that they are writing their treatise because a number of pseudo-mirzds title mirzdwithout adhering to the standard of requisites, qualities and accomplishmentsrequired of a true mirzd; but whereas the B.M./A.S.B. Mirzandma lays down the necessary guidelines for a mirzd in a general sequential order, Kamran's Mirzdndma reduces them to the mathematical exactitude of ten usal (principles) and sixty- fourfura' (minutiae), borrowing these terms from Islamic jurisprudence.
xI
am grateful to the Trustees of the British Museum for their permission to publish the facsimiles of the Mirzdndma. My warm thanks are due to my colleague, Professor G. M. Wickens, for his useful comments in tracing the etymology and history of the term mfrzaand on my synoptical translation of the work. 2 Charles Rieu, Catalogueof the Persian Manuscriptsin the British Museum(Oxford, 1866), II, p. 826 a-b. aW. Ivanow, A ConciseDescriptiveCatalogueof the Persian Manuscripts in the Collection the Asiatic Societyof Bengal (Calcutta, of 1924), p. 926. '4D. N. Marshall, Mughals in India, a BibliographicalSurvey I, Manuscripts (London/Bombay, 1967), p. 241. SMawlavi M. Hidayat Husain, " The Mirz1 Ndmih (The Book of Perfect Gentleman) of MirzI K~mrin with an English Translation ", Journalof theAsiatic Societyof Bengal, New Series, IX (1913), 1-13. 6 While Marshall notes the BM/ASB Mirzandma as the work of an anonymous writer (op. cit., p. 241), on p. 326 he erroneously lists it as the work of Muhammad Khalil. The confusion seems to be due to the fact that in the BM Pers. MSS. Add. 16, 8x9, Muhammad Khalil's Inshd'follows the anonymous Mfrzdndma. The ornate style of Mulhammad Khalil's letters is very different from the matter-of-fact prose of the Mirzdndma,and the two works could hardly be by the same author. ' JASB, pp. 1-2. 8 BM, Add. 16,819, f. 89b; JASB, pp. 8-9.

99

100

JOURNAL

OF PERSIAN

STUDIES

II The following is a synoptical translation of the B.M. Mirzdndma: [F.89b] This Mirzd Ndmais planned as a manual which would serve as a test to distinguish between the " true " and the " false " mirzd,as well as to distinguishmirzds from non-mirzds;so that those people of" irregulargait " and " canceroustemperament" who claim to be mirzdsshould forego their baseless claim. Mirzd-hood is tied to certain conditions, without the fulfilling of which one cannot be called a mirzd. The first of these conditions is a pure and well-known pedigree. The second is that in the eyes of the people a mirzdshould have a position of dignity. Therefore, in the third place, he should have a should be able to commanding at least one thousand dhdtand one thousand horses. The mirzamansab9 meet the expenses expected of his position. If he is not a mansabddr, should be an affluent merchant. he These are the three basic requirementsof mirzd-hood. If one is born to this much economic fortune, one should not be greedy for more. The fortunate one who has the necessary means should set out in the path of becoming a mirzdwith wisdom; and he should carefullyobserve the rules set forth below in this manual. In short, one should start being a mirzdwith a mansab a thousand, or the capital of Io,ooo of without which the title is meaningless. On the other hand, a non-mirza, whether he holds a minor tamdn, rank or is a mansabddr five thousand horse, is mean and worldly, if he does not possessthe character with and temperament of a mirzd. The purity of soul and uniquenessof attributes distinguisha mirzd,whose actions are wholesome, who is pure in outward appearance and virtuous in habits. All his qualities are as elegant as his appearance; on his stature his dress looks becoming; his disposition draws others to him; in generosity he is ahead of others so that his expenses sometimes exceed his income. If [F.90oa] a mirzdhas enough for his status, he should be grateful for it more than for being a commander of seven thousand (haft-hazdri), because service and subordination is degrading for a mirzd.10 He should not be beguiled by the attraction of the greater mansabs the multitude of horsemen and and foot-soldiersin the service of great nobles."1 The lesser the headache of high status, the better. There are among the mirzdssome descendants of the nobles12who hold lower mansabs and are not affluent, who are by the richnessof their own temperamentsattracted to mirzd-hood; their plight deservescompassion. When one lacks personal affluence, if one has inherited some money or property from one's father, one should not spend it or dispose of it, as such money or property is unblessedand soon comes to an end. Some others hang on to the status of mirzd-hoodby virtue of being the foster-brothersof a king, a noble or a khdn;they are like calves intoxicated with milk, though their milk is diluted with water; their position is one of insecurity and dependence on the fortunes of those to whom they are related. Others who claim to be mirzdsare administrativeofficerswho spend on themselves what does not belong to them, and who will have to face the consequencesone day. Other aspirantsto the title of mirzdare people who have accumulated some wealth and are in quest of digging more of it out of the earth; their mirzd-hoodcan have no durability; because for the love of money, the father disinherits the son, or the son kills the father and wastes the wealth in a few days. The real mirzd-hoodis a very different thing altogether. It is not merely pinning flowers to one's headgear or wearing a greenish or semi-greenishturban and strolling through a garden. [F.9ob] It is rather to inhale and imbibe the fragranceof the flower. A mirzdshould spend some of his time at day and night in the study of ethics. He should not be
ignorant of the questions offiqh (jurisprudence) or oblivious of the study of Qur'gnic exegesis; for an irreligious mirzd is even more insignificant in the eyes of the accomplished ones than an impecunious mirzd. He should also study works of history. He should memorize the rules of prosody and rhyme, so
9 An office in the Mughal bureaucracy or army. For the early though responsible to the wakfl or wazfr in administrative and revenue matters. organization of the institution see Abfi'l-Fadl 'Allfmi, A'kn-i AkbarfI, tr. by H. Blochmann (Calcutta, 1939), PP. 241-3, 1xThe mfrzd of the manual is thus a " middle class " gentleman 247-9, followed by Blochmann's useful notes, pp. 249-59. distinguished from the higher nobility, a few members of which 10 There is an obvious contradiction here. The commander of also bore the title mirzd. I,ooo, the status recommended for a mirzd, was much more 1' Mansabs and landholdings (jdgfrs) were not hereditary under subordinate in the Mughal official hierarchy than that of a the Mughals. commander of 7,000, who was subordinate only to the emperor,

SEVENTEENTH

CENTURY

MIRZA

IN INDIA

101

that he may not be incapable of understandingpoetry, which is a great accomplishment. He should try to read collections of epistolography and to express himself with coherence; he should know correct orthography. He should strive in learning the science of etiquette. Keeping in view the manuals for the training of falcons and horses, he should be an expert in judging horses and the good and bad points of falcons. He should use the shikasta(broken script) for worldly purposes and the naskh script for copying the Qur'an and other religious writings; and he should spend some hours in this occupation. He should not be inclined towards writing the nasta'liqscript, which is pointless and a script of low category. When talking, he should not speak at length; he should be brief and concise. If someone else speaks to him at length, and he has to listen to him, then he should only outwardly be attentive, but inwardly he should occupy his mind with something better, so as not to let time be wasted. He should learn to recognize the qualities of swords, and acquire the science of archery. He should not be all that inclined to use a matchlock musket, so that the unpleasant smell of its fuse may not reach his nose. Of all the sports he should prefer playing chawgdn (polo). He should urge his horse forward. A soldierly mirzd is better than an unsoldierly one. He should be not inclined to enjoy (merely?) watching marksmanship. On the day of battle he should not choose the ignominy of running away; at the time of action he should remain firm like a soldier, even though he be killed. One who claims to be a mirzdwelcomes such a death; for an honourable death is better than a dishonourable life. He should not dance attendance at the houses of great men, since this is undignified. He should know the requirementsof musical recital, so that he does not sing automatically in every gathering or assembly. Perfectionin the art of music is very difficult to achieve; and if he sings badly, his listeners will be disgusted. This noble science (of music) was formerly held in great esteem; now, with the passage of time, it has declined. If he decides to listen to musical songs or instrumental music to dispel the oppressionof his spirits,the singers [F.9 Ia] and musiciansshould sing and play in plaintive tunes. For his musical assemblies, held at day or night, he should choose the qdnan, chang,dd'iraand and bin. as tanbar musical instruments. Of the Indian musical instruments, he should prefer the rubdb as He should regardpakhdvaj the musical mode appropriatefor festivitiesand weddings; and he should which are musical instrumentsmore not favour the use of the dholki(a kind of drum) and the khanjari, suited to the assembliesof widows. Of the Indian musical modes, he should listen to and understand, he of the dhurpad Tinsan and Nd'ik Bakhshwar. If he likes the khaydl, should appreciate the compositions of Amir Khusraw and other moving compositionsin his style. He should abstain from listening to the of of khaydl Shaykh ShErMuhammad Hindi, and the tappaof Shaykh IHusaynFaqir and the chutkula worksof art, can be sung well by very few singers; also, Jawnptiri, which though perfect Shih because of unjustifiedinterpolations,they are not what they used to be. The same can be said about the .Husayn and Khayribidi Khaydl other such compositions. He shouldratheravoid the company of thosewho enjoy and chutkula dholakand khanjari;for such people are shallow and and listening to the Khayrdbddi Khaydl and ignorant and lack dignity. He should avoid the shows of bhdnds jesters, but he should well enjoy the who realistically satirize all sorts of people. If one wants to learn about one's performancesof the bhaya own faults, one should once in a while sit in one's own ambush. Knowledge and understandingof music is a great art; but the mirzdmust confine himself only to the knowledge of the harmony and musical tones, words and their meanings, which cannot be regarded as disgraceful. He should under no circumstances indulge himself in singing, but leave this rather to the professional musicians. Singing can lead to dancing, and that necessarily to other disgraceful and ignominious actions. He should totally abstain from giving a chance to his male friends and companions to listen to the singing of his private concubines; otherwise, it will amount to pandering, and may lead to a great deal of mischief. If he has the bad habit of consuming alcohol, and cannot get rid of it by any other alternative medicine, he should indulge in it wisely. As far as possible he should drink at home, and in the company of his private sweethearts. If he happens to be in the company of other friends and acquaintances, and has to be considerate to them, he should not join in the circulation of the bottle, but rather keep his own bottle separate. The wine he drinks should be perfumed. He should consider the wine which upsets the stomach as undesirable as tobacco with a bad smell. He should not drink daily [F.9Ib], as it is the habit of the rabble of the market-place. He should drink when the sky is cloudy in spring [monsoon] time
10

102

JOURNAL

OF PERSIAN

STUDIES

and when it is drizzling; he should not expose the daughter of the vine to the sun. He should regard a bearded cupbearer better than an unbearded, handsome cupbearer; he should rather place the bottle and the cup in front of each one of his guests separately, so that everyone may be his own cupbearer. He should let everyone drink according to his own choice and should not drink according to the urging of another, for this can lead to evil; and someone may not like to have more of that food and drink. He should not praise his own wine, as this has an odour of self-praise. He should adorn his drinking feast with a gold-embroidered table-cloth, pleasing glass bottles and cups, jewel-studded jugs and golden as goblets, as much as he can afford. He should not eat kabdb a savoury with wine as the gluttons do; it besmears the hand. If he does not like quince and pomegranate as relish, he should eat salted pistachios and similar things (with his wine). Plain wine looks transparent in a glass, bottle or jug. He should always provide perfumesin his parties; and try to keep his party fragrantwith them. All sorts of vases full of flowers in every season should be on view. Without them, he should consider the luxury of living as forbidden. He should keep his feast colourful; so that whoever departs from it may feel that he has been to the feast of a mirzd; that is to say, he should depart bearing the fragrant smell of scent and flower After smoking the tobacco from a pipe, and after the perfumeshave been served in the party and he has listened to some music, he should rise, making the excuse that it is time to go to bed; and then he should say farewell to his [departing] guests. When going to sleep, he should unfasten the knot of his robe and straighten the lower garment; otherwise he will not sleep comfortably. If he needs to borrow money, he should borrow it from a Hindu mahdjan (moneylender) whom he should prefer to a Muslim Mughal merchant, even though the latter lends money without interest. He should totally avoid purchasing anything from the shop of a Mughal, as it means a loss from beginning to end; for in the beginning, it means paying four times the cost of the thing purchasedand suffering great loss, and in the end it means listening to four-foldharangues of these [Mughal] merchantsin the market-place. On the other hand, a Hindu is content even if he reduces the interest, considers'the little he gets as plenty and is thankfulfor it. The Hindu money-lendersaskfor their due with salutations and an attitude of submission. The Muslim merchants make their claims [for the return of the debt] like equals and lord over the debtor. saying saldm'alayka The etiquette of dining. The table cloth for dining should be spread in the garden [in front [F.92a] of the house] at the edge of the pond. If one cannot afford a cloth of gold for a dinner-cloth, one should use the chhint with decorated design, provided it has no oily stain. If one cannot afford that every day, he should use a good white linen dinner-cloth, provided it is not spread repeatedly [without being washed]. The attendant at dinner should be a young man of wholesome appearance, who offers his prayers regularly. One must not employ a handsome person or a good-looking boy as a dinner attendant, as this leads to well-known errorsand is imprudent. For it may induce the guests to eat the bread but to break the salt-bottle [i.e. abuse hospitality]; and that would be damaging to the dignity of a mirzd. If all the plates on the table-cloth are of the same set, so much the better. If he cannot afford it, he should try to have the drinking glasses, plates, pots and demi-pots similar, of the same colour and of the same kind. The colour scheme of the crockery on the dinner-clothis an important matter. If he cannot afford the white china manufactured in the capital, he should regard the china of
Mashhad as better and more elegant than the china of any other manufacture. If he cannot afford even that, then he should never affix the title mirzd to his name. He should not use broken earthenware, as it would be hopelessly defaming for a mirzd. But he can, according to his need, use pretty earthen-ware. He should have the sour things for his meals, according to the season, such as clean achdrs (pickles) with vinegar; and should have glasses of the juice of pomegranate, mango, lemon and orange, especially the Kashmirijuice. He should ensure that a coloured or golden spoon is not used for food or for stirring a drink; he should be careful of the amber-coloured oiliness of a coloured spoon. He should be content with well-chiselled white spoons, and if he can afford china or glass spoons, so much the better. Of the varieties of food, he should prefer boiled rice with spices. He should not be an eater ofparkoko [?] and dizpydza (" double-onioned " curry), as it is the way of the gluttons. Instead of diipydza, he should

SEVENTEENTH

CENTURY MIRZA IN INDIA

103

eat qalya.13He should regard boiled rice and tender kabdbs the diet of the mirzds. Of the varieties of as he should preferthe ones made withyakhni(meat juice) or halwdn barbari or with the meat of a [?] pilaf, and he should eat them freshwith oil [presumablyghee]. He should regard fat, healthy sucking lamb; the qdbali[of rice] with green peas better than bdqlapilaf. Likewisein the season of mangoes, he should prefer the qalyaand pilaf containing mangoes. He should eat less of the food which greases the hand, such as pilafs,14 but eat rather boiled rice and kabdb; though it makes a choking morsel, it can be dealt with by taking that morsel and eating a spoonful of qalyawith it. I had the occasion to be present at the meals of an amir; [F.92b] and I noticed again and again, that though all sorts of food were there, he was inclined towards eating pilaf. One day I sedately inquired why he was not inclined to eat other things and liked the pilaf so much. In a tone most befitting the great, he answeredthat he preferredthe pilaf becauseit did not grease the hand and he did not have to try hard to cleanse his hand with a towel, which was disgusting. In fact, it is a blemish to cleanse the hands with a towel after having washed them.15 In eating, this is a primary condition. Although among the achdrs (pickles) those with mango and oil are quite familiar in India and are generally praised, he should not give them any credit. [If he eats them at all], he should mix them with the Indian khhichri (rice with lentils); for theyakhnipuldw (pilaf with meat juice) needs gheewith it, and not oil. This is the opinion of an expert who knows; it is a science, not just a saying or opinion. He should regard the colourfuljuice of pomegranate better than other juices. In season he should regard beetroot pilaf with relish as food suitable for a mirzd. Among the varieties of broth (dsh),he should prefer the one with delicate relish. He should also like the ashjaw(broth of barley) compounded with lemon juice, sugar, herbs and rose-water, but he should regard the meat in that broth as tasteless, and be content with the meat juice in broths like that. In winter, he should eat the top leaves of turnips, and always have them on the table-cloth; but he should eat turnip itself as little as possible openly; because the dignity of a mirzddoes not withstand the eating of a cooked turnip. He should regard the proverb " cooked turnip is better than unalloyed silver ", as inappropriate to the dignity of a mirzd. Instead of turnip shula[pilaf], he may eat beetrootsin shulawith relish, or in pickles or with coagulated milk, or in qalyd; all this would be all right as beetroot is a food fit for a mirzd,being agreeable, colourful and sweet. In winter, he should serve on his dinner-cloth winter foods like goat's jaw, harisa(rich, thick, wheat soup with meat, cinnamon, and herbs), barley broth and shab-bakht shab-pukht], according to [or his means; but he should himself eat only the sarpdcha and foot-joint) sprinkledwith vinegar, (boiledjaw lemon juice and mint, touching no other food, regarding this restraint as a requirement of the dignity of a mirzd,for the goat harisaand barley broth are not the foods for mirzds. He should not trust the cleanlinessof his cooks; but should take every possible care in the investigation that they are clean. If he cannot afford [to have it cooked in his own kitchen] he may use the cheese-bread sold in the bazaars; but he should regard other foods sold in the Indian bazaars as unwholesome. Of all the varieties of halwa (sweetmeats) and Indian preserves, he should like the mango preserve with fresh, sweetened [F.93a] herb juice, provided it is not more than a week old. He should not even look at other Indian pickles or the pickle of green mango made with filtered vinegar. He should eat a few spoonfuls of delicate, fragrant, perfumedpdlada(a beverage of water, flour or honey or other components). If it is possible, and if he can afford it, he should not eat the padlda without ice, or at least cooled by saltpetre. If this is not possible, he should not eat pdladaor even mention its name; for without ice and without being chilled by saltpetre it does not chill the teeth. He
should like fragrant Indian boiled rice. Bread, cheese and melons are becoming eatables for a mirzd. He should offer milk syrup to his friends. Milk mixed with sour milk is the drink for a mirzd. He should leave the milkfirni (rice-pudding) to the gluttons; and regard the firni without milk as suitable for a mirzd. He should consider drinking water without bid-mushk (a fragrant ingredient) as a thing forbidden. He should regard the eating of dalama (new cheese) wrapped in the skin of a chicken or with a sweetener as unwholesome. He should find pleasure in cheese made of fermented milk. Of the varieties he should prefer that of nuts with fragrance and tablets of amber, and pieces of lemon and of .halwd,
13

14But see the next

Originallyfriedfood, but in India a highly spicedcurry. paragraphwhere he contradictshimself.

15

Probably because the hands were dried in the smoke of incense in a censer after being washed at the end of the meal.

104

JOURNAL

OF PERSIAN

STUDIES

sandal-perfume. He should not be unmindful of the rishta(or sweet) of the style of khita (Turkestan) perfumed with amber. He should form the habit of eating with the tips of his fingers. He should stop eating before he is full; even though he remains hungry; for it is not easy to be a mirzd. If he does remain hungry, he can eat [a little more] in the interior of his house, out of the sight of the critical. But this has a great risk of bad reputation, involving a loss of the prestige of the mirza. It is better to remain hungry. He should not eat outside the dining room;16 for eating outside it is like eating at a shop in the bazaar. Of the fruits, after melon and grapes, he should regard first-classmangoes, which can be obtained only rarely, as the best of fruits. But he should not call it [i.e. the mango] superior to melon, as that would amount to apostasy. For the [cooked] foods he should regard unripe mango as a better ingredient than apples and other sour fruits. He should not be unmindful of relish made of pineapple mixed with sugar and rose-water; for like the fruit of Kuch-Bihar and the Sumatra of AkbarTb~d,it is a delicacy made especially for the mirzds. The sugarcane of the province of Bihar and of Akbarabid is befitting food for a mirzd,on condition that he does not pile the chewed refusein front of himself. Of the greens he should like lettuce, green coriander and mint. He should regard leek and radish as enemies of God. From the belch which follows the eating of radishesand is worse and more unpleasant to the mind than the sound of a gunshot and the smell of gunpowder, he should preserve himself. He should not eat with someone who is ignorant of the manners of dining and who is gluttonous. If by chance he has such a guest, he should soon finish his dinner and rise, making some such excuse which does not hurt the feeling of the guest. But after that he should shun the company of such a person, though he can occasionally send him a tray of food. Keeping company with a gluttonous person is not free from hazards, one of which is belching [F.93b]. Similarly, he should avoid someone who does not pick his teeth after washing his hands [after a meal] or does not clean his mouth and hands with water after picking his teeth, since the company of such a person is distastefuland disgusting. After the meal, he should wash his hands with fragrantsandal powder compounded with other perfumes. After chewing the pdn (betel-leaf) he should rinse his mouth, as it [i.e. the stain of pdn on the lips and the teeth] is unwholesometo look at. He should avoid a companion who talks at length after eating pdn,as the spray from his mouth may soil the mirzd'sdress. He should considerit obligatory that perfume is brought in in after the meal. In winter he should burn fire in an iron censer and put aromatic lakhlakha it. In winter, if he can afford it, he should keep the house fragrantwith the perfumesoffitna,flower and argaja. He should constantly put aloe-wood in the fire. In summer he should consider it obligatory to have a khaskhdna (thick curtains of straw on the door and the windows, which are sprinkledwith water to keep the room cool) prepared, to have fans, to have the floor-spreadingof gangdjalsitalpdti(a variety of' North Indian matting) and to arrange vases with fragrant flowers of that season. A white linen floorspread [on the top of the matting] is best for this season; and the best perfumefor this time of the year is high-classJahdngiri argaja. It is the best perfumefor this season, except for rose, which is also the perfume for this season. He should regard the enjoyment of the monsoon season as a compound of winter and summer. The floor-spreadof this season is sometimesa good mat of split-reedsand sometimesa good woollen broadcloth. In this season also he should consider it necessary to have a censer and lakhlakha. In the monsoon season he should occasionally sit elegantly on a wooden takht(wooden, platform-like sofa), but not without a carpet covered with quilting or a white linen-spread. This would be very inappropriatein winter-time. He should use carpets of good variety in winter, those made in Kirman if he can afford them; otherwise, those of Kashmiripatta. A house without curtains and screens is like an open shop in a bazaar. In every season he should consider it unwholesome to eat without a magaskhana net hanging all around for protection against (a the flies); otherwise, it would be like eating in the bazaar. He should regard the beauty of flowers as better than the green beauty of grass, though he should appreciate that also. [F.94a] He should appreciate more the green grass under the flower-shrubs. Of the jewels, he should like rubies and pearls; of the fruits,pomegranate. A house which does not have
16

Presumably during the winter and the rainy season, for earlier he had recommended that the dinner-cloth be spread in the

garden of the house near the pond, presumably during the summer.

SEVENTEENTH

CENTURY

MIRZA

IN INDIA

105

a pond and a fountain surroundedby flowerpotsfilled with flowers of every season, is a house without enjoyment. He should set a garden wherever possible in the compound of the house, for the mirzdis to equal in numbers (ham-'adad) a gulzdr (a rose-garden); he is bound to be attracted by a garden. In every corner of his garden there should be colourful chirping and singing birds like nightingales" and parrots. He should hear unpleasant voices of other birds from a distance, because a mirzd'stemperament cannot bear listening to such noises. The beauty of these flowers and birds is not merely for external view; the beauty of every bird leads one to the contemplation of its Maker, and its singing leads the heart to the anguish [of divine love]. Of the fighting animals, he may like to view deer- and camel-fights, provided that he does not befriend deer- or camel-keepers. But like a good mirzd,he should avoid viewing ram-fightsor bull-fights, since the people of the bazaar like these. For his journey he should have camels ready with all their equipment. He should avoid the company of such [self-proclaimed], self-opinionated, bastard mirzdswho tie their turbans with great delicateness,who talk with the movement of head or with the gesturesof body or of eyebrows, who are over-emphatic in speech, who praise halwasand greens from among the foods, and do not like the Kashmirijuice from among the drinks, who turn away with affected delicacy from whatever is invigorating, who do not clean their teeth without looking in the mirror, who clad themselves in the single layer of a thin and transparent upper garment and wear trousers of satin and kamkhdb (many-coloured, embroidered cloth), and who have the habit of eating pan frequently and are blackening their teeth with missi. Such mirzds no good. Mirzd-hoodis to be mirzd khdnor mirzd-beg;1' not to be a mirzdda-begum mirzdda-khanum effeminate]. or [i.e. Theetiquette wearing of apparel.In winter, he should wear a datahi(sc. a garment with two folds and made of various kinds of Indian material). He should use pearls for buttons, [F.94b] for pearl is natural while other jewels have to be cut. In winter, he should wear a shawl, either plain or imprinted with gold and silver leaves. He should wear gold-embroideredturbans (chira)which though of high quality are inexpensive, and come from the fatah workshopswhich produce high quality striped cloth for low prices. In summer, when he sits on a wooden seat with a white covering, he should wear the silverthreaded cap round the head and the ears (gash-pich), a silver-threadedupper garment (bdla-band). and He should never wear brocade or cloth of gold. He should never be inclined to wear a golden or embroidered turban, or a gash-pich embroidered brocade, or trousers (shalwdr) cloth of gold or satin, of of as this would be below the dignity of a mirzd. He should use the cloth of gold only for pillow-cases and curtains. If he has the means he can indeed give away cloth of gold and satin as a khil'at (robe of honour). He should not wear embroidered and gold-threaded turban as it may give the mirzaa headache, and it may also fall down. Theetiquette ridingandhunting.Of all the means of riding, he should regard the pilki (palanquin) of as the best. It is most harmlessof all the kinds of rides available to a mirzd. In riding an elephant or a horse there is the risk of falling off. But, in the rainy season, an elephant ride is the best; as both in the pdlkiand on the horse there is the risk of getting soaked in the rain, or of the mud and dirty water from the rooftopssoiling the head and headgear and trousers,which would be against the dignity of a mirzd. But for visiting gardens and viewing flowers and flower-gardens,he should ride a flower-colouredor black and white (ablaq)horse. For shooting cranes or herons he should ride a black and white horse with long mane. Such horses should be available and ready in his stable every day....
[F.95a] Of the hunting birds, he should prefer the sparrow-hawk (bdsha). He should enjoy the sight of hunting with a bahri-hawk. He should adorn the wrapping linen and the headgear of the hunting animals, and should fully apply his aptitude for decoration and art in this matter so that the hunting animals of the mirzd remain distinguished from those of a non-mirzd. These are his associates, and the associates should be dressed like the man. He should know that hunting with a falcon is allowed for this reason, so that one may not remain empty-handed until one obtains a bahri. One should not gallop one's horse impatiently after the bahri, but should ride slowly and let one's companions gallop; for there is danger of falling off the horse in galloping on such occasions. One may well be enamoured of hunting
17

This is an euphemism; there are no nightingales in India.

106

JOURNAL

OF PERSIAN

STUDIES

and absorbed in it; but life is more precious than the spectacle of hunting, which is not worth falling from one's horse, or the falling off of one's headgear (chira),or breaking one's neck. One should only spend one pahr(a quarterof the day, or the first quarterof the day) in hunting; for continuing after that is doggish, and the hunting animals lose their lustre [after the firstpahr]. He should leave prolonging of the chase to the dogs and should not join them. If he is hunting with a matchlock, he should not fire more than two shots. If he is hunting a heron or a lark, he should decorate his headgear with their feathers [presumablyas a decoy], and give some of these to his companions. When he returnsfrom the chase, he may rest by a pool or stream of water under the shade of a tree. A white linen sheet has to be ready on this occasion to be spread [overhead] under the tree, so that the dressof the mirzdmay not get soiled by the droppingsof the birds. After drinking coffee, he should distributeportions of the kill of the chase during the previousnight and during that morning among all his companions,high and low. If he had hunted any deer, he should have kebdbs made of the venison, as they are delicious. [F.95b] He should have the hunted birds like bustard, partridge, heron or duck paraded in ones or twos by the foot-soldiers for general view. He should enjoy unfamiliar and half-coloured (nimrang) flowers. If he wants a flower to yield fragrance, he should himself pluck it from the bough. He should not accept it from the hands of the gardener, for there is no hand cleaner than the hand of a mirzd. He should not wear flowersin his turban, as it is effeminate to do so. It is a blemish for the mirzd,who is a [masculine] lover. But he can, as a blessing, put a gul-i mutlaq[rose] which is made of the holy sweat drops of the Pride of All Creation on his head. Occasionally, in privacy, he may put a bunch of ndfarmdn flowers,which is like a plume of feathers, in the corner of his turban; it would look becoming. Etiquette bathing. In the bath he should not use the same brush for rubbing the body and for the of rubbing of the feet. There should be different brushesfor the purpose. He should not allow a bearded bath-attendant to rub his body with the brush; for the sweat which falls from his hair and beard is as unpleasant as water with brimstone. He should not allow a mean person (pdji) to be his companion; he should not look at such a person if he stands in front of him; and he should regard his presence as disturbing to the mind. He can have one or two servants, whom he must employ, also serving in his parties, on condition that they do not remain standing in front of him. He should not utter anything vulgar or obscene, as this is degrading for a mirztd. Similarly,he should avoid ill-mannered and short-temperedpersons. He should not speak to a person of low or mean origin without necessity; and he should try to communicate with him [only] by signs or gesture. If such a person does not understandhis gestures,he should not retainhim in his service. When ill, he should not mind paying the physician handsomely; for life is more precious than money. He should avoid a sweetheart who has another lover running after her [or him]. He should stay a mile away from a mad elephant or camel, or a bullock attacking with his horns or an intoxicated person who has a sword in his hand. To be involved with them is foolishness. If he is with the imperial army, he should cross bridges, corners and narrow passes either before everyone else or after everyone else. He should not think of running away or being absent on such an occasion; for a living hero (pahlawdn-i zinda) is commendable. He should not become warmly familiar with someone destitute, be he a khdnor sultan; for such people professfriendship only to borrow money or for the attainment of some other end. Their friendship is tainted with selfishnessand has no durability. But if a person is
genuinely in want and in distress, he should help him in gratitude of the means which he has; and he should not be one of those mirzds who "spend little and sit high ". He should not accept anything from anyone, in return of which he does not intend to give something else of greater value. He should consider it obligatory to wear a dagger orjamdhar (an Indian dagger). He should regard as obligatory the wearing of rings of ruby, emerald, turquoise and cornelian on his fingers, as they have different properties. III Mirzt Kimran, the author of the other Mir~dndma,came in the fourteenth year of the reign of one of the Mughal emperors [Awrangzib?], by way of Kashmir to Lahore where he found a number of " undeserving " persons calling themselves mirzd, and decided to write a manual (dastir al-'amal) of

SEVENTEENTH

CENTURY

MIRZA

IN INDIA

107

mirzd-hood. His own claim to be a mirza was based on his " personal efforts and the acquiring of pleasant manners ", though he confessesthat his words (or verses) are inclined towards wit and jesting. and Indeed, some of the points in his Mirzdndma some of the requirementsfor mirzd-hoodare quite trite to the B.M. Mirz~ndma. compared has and translated (it was published over sixty years ago, Although his Mirzandma been published"8 and in a journal in India which is not easily obtainable in the West),19it will not be out of place now to give a brief summary of his treatise here for any easy comparisonwith the B.M./A.S.B. Mirzdndma. Mirzd-hood depends upon ten (principles) and a certain number [actually 64] of furi' are that: a mirzdshould believe in God; he should acquirelearning, especially us.al (minutiae). The ten that of grammar; heus.l should have read in his youth the Gulistan Bzistdn Sa'di; he should have the and of capacity of judging men; he should not engage in scholarly disputes with a student who has just completed his education; he should not engage himself in conversationwith every unworthy person but should associate only with those who are his equals; he should learn Arabic, Persian, Hindi and Turkish; he should be an expert in composition and a fast worker; he should have some knowledge of accounts; and he should have some knowledge of philosophy, but should not indulge in fruitless discussions regardingfree will and determinismor finite and infinite. The [sixty-four] fura' of mirzd-hoodare that a mirzdshould stay a mile away from a mad elephant; he should prefer chinaware to gold or silverwarein crockery; if there is one evil person among ten in a place, he should give up the company of those ten; he should shorten his name (for instance Mirzd Thsufinstead of MirzAJaldl al-Din Yfsuf) ; he should not haggle when buying; he should regard ruby as the best of the jewels; he should regardpdlkias the best of all conveyances; he should regard melon as the best of all fruits; if he does not want to break his limbs he should not travel with the imperial army to Kashmir;20 he should preferboiled rice with spices to other types of food; he should leave pure love on a high shelf, but at the same time he should not chase a capriciousbeauty; he should not smoke too much tobacco; he should be on his guard from diseasesin India; when introduced to a house, he should firstvisit it on the occasion of introduction, then for a second time to pay his compliments, then a third time without compliments (i.e. informally) if he is welcome; he should regard Lahore as the best of all towns in India; he should regard the Agra fort as unique in the world; he should regard Isfahdn as the best city in Iran; in a crowd he should be mindful of the swords kept under their arms by ahdis (foot-soldiers); he should pay up to twenty rupees a day for payment to his palanquin-bearersand porter; he should not take a room in a sard (inn); he should not take part in a battle, and if he has to participate, he should not stand within the firing range, should never pursue a defeated enemy, and if in a defeated army, should make haste to run away;21 his conversationshould be eloquent and expressive; he should like narcissus,violet and orange; he should know who is bad, but wisely not talk badly of him; he should know how to play nard,but must not become a gambler; he should not bet heavily in chess; he should not admire obscureverseswhich take time to understand; he should not keep company with a fast rider; he should not reveal his income to someone who could be coveting his job; at dinner he should not sit next to a mufti(jurist); he may sing occasionally if he has a good voice, but avoid proand fessionalism; he should regard Girahrtidi,Mirza Rafi' Shirdzi22, Jaldld-yi Ardastini as the ablest of his contemporaries; he should not be impressed by everyone who puts on the robe of Q:.4i-zdda learning; in a discussion,if he succeedsin making a point, he should do so fully, and if the other person gains the point he should be courteous; where people talk of their income or means, he should get the
1s JASB IX (1913), 8-13.
20

22

19 Ibid., pp. 2-8.

21

For a description of the imperial army's march to Kashmir, see F. Bernier (English tr.), Travelsin the Mughal Empire,A.D. r656-r668 (London, 1891), pp. 351-428; also Niccolao Manucci, Storia do Mogor, English tr. by W. Irvine (Indian reprint, Calcutta, 1965-67), II, 60-69. This cowardliness advised by MirzA KdmrAn is exactly the opposite of the standard of courage enjoined for a mirzd, who must never run away from a battlefield, according to the BM Mfrzdndma; see above.

As mentioned earlier in the text, Hiddyat HIusaynhas identified this Rafi' Shirazi with Rafi' al-Din Ibrahim Shirizi, the author of Tadhkiratal-mulak (begun in o0I7 1608-og and completed in o2o/ 16I -I12), a history of the 'Adil Shihs of Deccan and of contemporary Indian and Persian dynasties (Cf. Storey, I/ii, 742-3). But the latter never went beyond Ahmadnagar to the north of India. There is no evidence that he visited the Mughal Empire, especially Lahore, where alone he could have met MirzA Kgmrqn. One may safely conclude that Mirzd K~mriin's Rafi' Shirazi is a different person.

108

JOURNAL

OF PERSIAN

STUDIES

topic changed, otherwiseleave the company; he should visit Egypt; he should preferSyria for making a home as bread, cheese and melons are nice there, even though these are the food of the Rdfidis;23 in India, he should not wear the chdqshar;he should not be fond of his own words; in an assembly he should not get involved in discussion, especially about his religion, which he may conceal to avoid the chance of bodily injury;24 unless he has a mansab 5oo, he should not invite a guest; when in financial of straits, he should not be guest of someone mean; he should regard watching gambling as the best of shows; he should not talk scandal; he should not indulge in jokes with an improper person; he should not repeat the visit [too soon]; with an income of a hundred rupees he can spend ten on belt and chira, but even fifty on a fur [jacket]: he should not talk when people are playing nard; he should not take his sweetheartto someone else's house, especiallyif he is richer; he should wash the dirt on his face with salt and vinegar; he should trust only the dervish; he must recite the firstverse when reciting a poem, but should not interrupt to complete the second half-line if someone has recited the first; in India he should pay more attention to the strength of a house against rains than seek architecturalbeauty; in a bath he should not allow himself to be served by a bearded bathing attendant; he should not sit in a prominent place in an assembly,in order to avoid being snubbed; in India he should not expect wisdom from those who wear large turbans; he should not have any expectations by being familiar with [rich] he persons; he should not eat fdlada andfirn in India, as they make one lazy; if he has a low mansab should not have costly trappings; riding with his superior,he should be a pace or so behind him, except when crossing a bridge [which is risky] and which he should be first to cross; even if impecunious, he should not use engraved brassvessels, as the dirt is unwashable in the engravings; he should not talk at Sunbal KhAn'sinn, which is full of argumentative men; he should not be drawn in by someone who talks at length; he should avoid argumentative people; he should avoid the company of Muhammad Beg and Mirza Fadli Beg, who have written [a version of] Kalila wa-Dimna; when invited, he should not be the first to go to a party, as thus he will have to greet many more people. S.lih

IV
Who was a mirzi, and what was the category of mirzd-hoodin the latter seventeenth century when these two manuals, and possibly others now lost or hitherto undiscovered,were written? Before one answers this question, one has to investigate the etymology of the title mirzd,its various meanings and various applications. It is generally considered to be an abbreviation of amirzdda, though this derivation has been doubted by some."2 Mir, as an abbreviation of amir,was quite common from the sixteenth century onwards in Iran as well as in India as a title of amirsand sayyids,and for officialdesignationssuch as mirmunshi, than in Mughal etc. mirshikdr, In SafavidPersiamorecommonly26 used the prefix mirzdrather than mir. India,'7 many sayyids In eighteenth century India, and almost certainly earlier, in the seventeenth century, the term mirzdhad a primary and a secondary meaning. In the primary sense it has been defined by Arzfi as " used formerly in the titles of kings and princes, and now by noblemen and their sons; also used in Irin for the sayyids."28 has been defined in the secondarysense by another eighteenth century Indian It as " a manly person held in esteem by people ".29 A similar definition occurs in a lexicographer, Shad, as dictionary compiled by Nafisi in Iran, who also defines mirzdmizsaj one who appreciatessubtleties, is of an independent temperament and haughty, and mirzdmanish one of exalted nature, graceful, as
23This polemical referenceto the Shi'is may either mean the larly abbreviated. Druzes of Syria, or Shi'is in general in which case Mirz! '* MirzN Mulammad TAhir Nasribldi, Tadhkira (Tehran 1317 Kdmrin (who came to India through Kashmir) might be of shams!), pp. 65, 71, 95, 99 et seq. and passim. 27 Shaykh Farid Bhakkari, Dhakhiratal-Khwanin(Karachi 1961), TirAni origin. But see below, n. 24. 24This advice for taqiyya curious,coming from some one who is I, 170-72, 230; Shah Nawiz Khan, Ma'dthir al-umard' polemically calls the Shi'is Rdfidfs. But some Sifis having (Calcutta 1888-9o), I, 222-4, 479-81, 739-48; II, 736-8; heterodoxor pantheisticviews also concealedtheir faith. III, 218, 214-21, 590-94, 633-521 Sirajal-Din 'Ali "Arzfi" Akbarlbldi, Chardgh-i hiddyat,p. 297. Khan hiddydt, 28 Chardgh-i ed. Dabir (Tehran i96o), 293-98. Professor 9*Muhammad PAdsh-h " Shid ", Farhang-i Anand Rdj, ed. Muh.ammad Wickens,who also doubtsthis etymology,pointsout that other Multammad Dabir (Tehran 1335 shamsi), VI, 3939. titles like shdhzdda, and have pfrzdda shaykhzdda not been simi-

SEVENTEENTH CENTURY MIRZA IN INDIA

109

leaders of the various Tatar tribes in Crimea were known as mirzds,and the Ottoman sultan installed or deposed a Girty Khan in Crimea generally in consultation with them. Four of these mirzds (dirt Karafibegi), the heads of ?irin, Argin, Barin, and Kipchak tribes held a more honourable position by tradition than the rest.37 his and assumed the title Bddshdh,38 sons Kamran and Although Babur discarded the title mirZd others retained the title mfrzd. Even in the sixteenth century, Akbar'shalf-brotherwas known as Mirzd Muhammad IHakim,though by this time the sons of the Mughal emperor came to bear the title shdh, or titles shldhzdaor bddshdhzdda, which had already come to be used along with the title amirzdda mirzd Timiir's reign.39 during Even in the higher echelons of Mughal nobility, the title mirzdhad considerablydeclined as early as the sixteenth century and the official title khdn,bestowed by the emperor, was taking its place. Shaykh Farid Bhakkari,whose tadhkira compiled in 1060/1650, has written notices on 186 nobles of Akbar's was of the of which only 13 bore the title mirze.40The Ma'dthiral-umard', other famous tadhkira the court, which was compiled, revised and completed during the late eighteenth century, gives Mughal nobility, notices on 722 nobles from the sixteenth to the eighteenth century, of whom only 55 bore the title mfrzd.41Most of these mirzesbore the title in the primary sense, but some in the secondary one. A brief survey of the mirzds the upper echelons of Mughal nobility might be interesting. To begin in there were the Timirid princes, distant relations of the Mughal reigning family, such as Muhamwith, mad Sultan Mirza,42a grandson of Sultan IHusayn Bayqara, and the Timarid princes of Balkh, Mirz~ and Mirza Shih Rukh and the latter's son, Najabat Khan Mirza Shuja'.43 Then there were Sulayman the Safavid refugeeprinces MuzaffarHusayn Mirza and Rustum Mirza who had surrenderedQandahar to Akbar, and their descendants in India, Mirza Shah Nawaz Khan, Mirzi JHasan, Mirzi Nawdhar, Mirz~ Sultin, and Mirza Mukarram Khan.44 To these one may add Mirza Safavi Khan 'Ali Baqi, who came to India during the reign of Awrangzib, who claimed to be of Safavid descent, and who received a title to that effect from the Mughal emperor.45 Several Iranis with real or pretended claims of Safavid ancestry came to India during the eighteenth century, after the fall of that dynasty.46 Among the Irani higher nobles of Mughal India was Mirza Ja'far Asaf Khan who rose to be wakil, the highest office in the state.4' Another equally illustrious mirzd was Nfir Jahan's father I'timid
Akbar Nafisi (Ndzim al-atibba'), Farhang-iNaffsi (Tehran 1343), V, 3257. 3' For a detailed history of the Sarbadars, seeJohn Masson Smith, The History of the SarbaddrDynasty, 1336-r381 A.D. and its Sources('s-Gravenhage I970). ed. 'Abbisi 3s Sharaf al-Din CAliYazdi, .Zafarnama, Mul.ammad (Tehran 1336 shamsi), I, 223, 232-7. 33 Thus, for instance, Amirzada Jahangir (ibid. I, i8o and passim); Amirz~da MirdnshAh (ibid. I, 336, 445 and passim); and AmirzAda 'Umar Shaykh (ibid. I, 473). 34 Ibid. II, 30 and passim. 35 Kamal al-din 'Abd al-Razzdq, al-sa'dayn wa majma' Ma.tala' B.M. Or. MSS. 1291, passim. al-bahrayn, 36 Nasrqbqdi, pp. 69-125. 3 TAhir Max Kortepeter, OttomanImperialismduring the Reformation:
30o'All

successor-states. By the fifteenth century, mirzd was quite a widespread title in the Islamic world. After, and possibly even before, the extension of the Ottoman suzeranity over the Giriy Khans of Crimea,

delicate and having delicacy.30 The two mirzdndmas concerned with the mirzdin this secondary are sense. In the primary sense, the title mirz~is said to have been used first by the Sarbadtrs of Harit,31 or though the last Sarbadar used the title malik.32The sons of Timiir were called amirzddas33 mirzds, which is the strongest evidence for the theory that the latter term is an abbreviation of the former. Shah Rukh was referredto both as amirzdda34and as mirzd,the latter title being applied to him more generally after his accession.35 Among the Timiirid princes, the title mirzdcould be a prefix or a suffix; and officers of the in the case of the Safavids it was always a suffix. In Safavid Iran, wazirs,mustawfis affixed the title mirzd to their names.36 So did also some of the nobility in the Timafrid daftarkhdna

Europeand theCaucasus(New York 1972), pp. 8, 13'Abd al-Bdqi NihAwandi, Ma'dthir-i Rafhmi (Calcutta 1924), I, p. 4953* Yazdi, ZafarnamaI, p. 573 and passim. al-khawdnin (Karachi 1961), I. 40 Shaykh Farid Bhakkari,Dhakhirat 41 Shih Nawdz KhAn, Ma'dthir al-umard,' (Calcutta, I888-90). 42Ibid. III, I92-9. 43 Dhakhirat [Dh], I, 20-23; Ma'athir al-umard[MU], al-khwadnn III, 264-76; 329-34; 821-8. 44Dh I, 99-IOI; MU III, 296-30I, 434-42; 477-8, 555-7; 581-2; 583-5, 692-4; II, 670-76. 45 MU III, 653-446 Ibid. III, 681-2. 47 Dh I, 170-72; MU, I, 107-15.
38

110

JOURNAL

OF PERSIAN

STUDIES

al-Dawla Mirzt Ghiyath BEg Tehrani, his eminent son Asaf Khan and his descendants who rose to high offices.48 Of the Tfirdni mirzdsin the high Mughal nobility, only the Khan-i A'zam Mirzd 'Aziz Koka, Akbar's foster-brother,may have received the title mirzdfrom the emperor or assumed it as a token of his close relationship with the emperor. Bayram Khan Khin-i Khanan, a descendant of the illustrious Turkmen line ofJahdn Shih, did not use the title mirzd,but his son 'Abd al-Rahim Khan-i did, as did also his sons and grandsons.49Another distinguisheddynasty of mirzdswas that of Khdnmn Jani Beg Tarkhan, who claimed descent from the distinguishedTiirni clan of Arghiin and was ruler of Sind, but later submitted to Akbar and became one of the high nobles of his administration.50 Other illustrious nobles of his house such as 'Isd Tarkhdn and Ghazi Beg Tarkhan continued to use the title and mirzd.51 Notices on three sayyidmirzdsappear in the first volume of the Dhakhirat al-khwainin, on eleven of them in the Ma'dthiral-umard', including Mahabat Khan Sipahsilar, who held the person of the EmperorJahangir in detention and who though a Kabuli, claimed Sayyid descent. Of the Sayyid while his brother 'Abd al-Hadi called mirzds,Nawazish Khan 'Abd al-Bdqi called himself a mirzd,52 himself mir.53 Most interesting is the case of Hindu mirzes. Of these, Bahadur Singh was given the title of mirzd rdjaby Akbar;54 and the famous Rajp*ftgeneral Jay Singh was given the same title by Shah Jahan.55 Thus it seems that in rare cases, mirzdcontinued to be a title bestowed by the Mughal emperors until the middle of the seventeenth century. in But the usual title for the nobles and mansabddrs, and low, was khdn, both Mongol and Pathan high we tradition. This title became exclusive; and the small percentage of mirzdsin both the tadhkiras are examining, shows that the later title was merely an ancestral affix or suffixsurviving in only a few cases. In most of these few cases, the nobleman also bore the more prestigioustitle of khdn. some were called mirzds. Among the lesser nobility and among the more insignificant mansabddrs, In the case of some of them such as MirzAQuli Khan and Mirza 'Ali Khan of the reign of Akbar, we know hardly anything beyond their names and some minimal details.56 Eight such " insignificant " It mirzdsare subjects of short notices in the Ma'dthiral-umard'.57 is possible that these " insignificant " mirzdsrepresent the " vulgarization " of the title mirzd. What began as a royal title, became in due course, not only in India but also in Iran, an affix chosen by a gentleman or a lesser nobleman. Hence Mirzd became a category of gentlemen of taste and culture in the secondary sense of the term. We meet the new mirzd,though in one case of high descent, already in the early seventeenth century: Mirza AbG-Sa'id,a grandson of I'timad al-Dawla and a nephew of N&rJahdn. He was famous for his " good looks and delicate personality "; fastidious and " ceremoniousin matters of dress and food and floor-spread"; showing off such tastefulnessand pompousnessin everyday life that his equals could not rival him; and " of such a proud and snobbishnature that he held the high heaven and the angels as of no account."58 Mirzds of this type, and of the secondary sense, multiplied during the seventeenth century. It was for their edification that the two mirzdndmas, perhaps other such manuals, came to be written. and

8 MU I, I27-33, 152-60, 180-82, 230-32; III, 431-3. 49Dh I, 11, 31; MU II, 14-17, 645--8; III, 586-9. Dh 0o I, 176; MU III, 302-13. R1MU III, 345-8, 485-8. For a detailed study of Gh~zi B-g and his intellectual patronage see iHusAm al-Din Rgshidi, awr uskiBazm-i adab (Karachi, 1970). Mirza Ghdz Big Tarkhdn 52MU III, 828.

Ibid. I, 167-7I. "4Ibid. III, 360. " Ibid. III, 570-71. D I, 212, 232. Dh 6" MU I, 504-05; III, 459-60, 557-8, 351-4, 200oo-o4, 354, 258-61, 66o-6i, 482-5. 58 MU III, 513-14.

53

DALMA TEPE By Carol Hamlin

Excavations at Dalma Tepe were conducted for the Hasanlu Project for brief periods during 1958 and 1959 (by CharlesBurney) and in 1961 (by T. Cuyler Young, Jr.).' The aim of the Dalma excavation was to retrieve an artifactual sample of early (" chalcolithic ") materials of a type recognized but not easily accessible at Hasanlu Tepe, the major focus of the project's efforts. The Dalma excavations were thereforeoriginally designed to obtain vertical samples from small areas; in 1961 shallow soundings over an enlarged area were made to obtain a somewhat larger sample of the architecturalremains. This brief account of the results of the excavations is based on the notes of Burney and Young, supplementary information provided by Young and Dyson, and an examination by the author of Dalma sherds and registered objects currently in the University Museum in Philadelphia.2 Dalma Tepe is located at the southwest end of Lake Urmia, in western Azerbaijan, approximately 1350 metres above sea level. The site is about 5 kilometressouthwestof Hasanlu and 4 kilometresnorth of the Gadar River. It is a small and nearly circular mound measuring roughly 50 metres at its base and rising approximately4 metres above the modern plain level. The nearest source of surfacewater is a spring in the modern village of Dalma, a few minutes' walk from the mound. The brevity of the three excavation periods (totalling less than one month) and the small staff freed from the work at Hasanlu Tepe combined to produce small soundings and, therefore, relatively small samples. Although the data from Dalma are limited, they are nonethelessof interest: the site represents one of the earliest sedentary occupations in the Ushnu-Solduz valley system, and Dalma-like ceramics have subsequently been reported from an area extending from Lake Urmia south to the KermanshahHamadan region, with additional limited and scattered occurrences as far south as the Khorramabad valley in Luristan.3 Since Dalma Tepe is one of the few excavated sites with stratified material of this type, it provides some basis for assigning a relative date to surfaceoccurrencesof Dalma pottery. There are two reasons for this. First, Dalma material (designated Hasanlu IX in the Ushnu-Solduz valley sequence) is securely placed stratigraphically between the earlier Hajji Firuz (Hasanlu X) and later Pisdeli (Hasanlu VIII) materials. Second, there is, apparently, a change within the Dalma ceramic assemblage; an earlier preponderance of painted ceramics appears to give way to an increasing, relative frequency of distinctive impressed wares. This apparent temporal variability is of potential value in dating surface occurrencesof similar ceramics elsewhere. The recent excavations of the Royal Ontario Museum at Seh Gabi, in the Kangavar-Assadabadvalley, produced a stratified sample of Dalma wares closely resembling those from Azerbaijan.4 The detailed analysis and publication of the Dalma-like materials from Seh Gabi, which in this stage of their analysis appear to corroborate our understanding of the sequence at Dalma itself, promise to shed further light on the range of Dalma ceramic variation.
the time of the Dalma excavations,the projectwas cooperaoccurrences in the Khorramabad area; the survey collection made by Frank Hole of Rice University is being prepared for tively sponsoredby the UniversityMuseum of the University of Pennsylvania,the MetropolitanMuseum of Art, and the publication, and will provide interesting comparative material. Iranian Departmentof Antiquities. 4 For preliminary accounts of the Seh Gabi excavations, see T. 2 Approximatelyhalf of the registered of objects (by definition, Cuyler Young, Jr. and Louis D. Levine, Excavations the Godin exclusive of sherdageand bone) are now in North America, Report,Royal Ontario Museum, Art and Progress Project: Second being housedlargelyin the UniversityMuseum. The illustraArchaeology Occasional Paper 26 (1974); Louis D. Levine tions shown here were prepared by the author, T. Cuyler and Carol Hamlin, " The Godin Project: Seh Gabi ", Iran XII Young, Jr., Maude de Schauensee,and other staff members (1974), PP. 21x-13; Carol Hamlin, "Seh Gabi, 1973", of the UniversityMuseum. 27 Archaeology (1974), PP- 274-77. The Dalma material at Seh sherdsfromKunji Cave. Gabi comes almost exclusively from the lower portion of 3 The authorhas seen a few impressed These were excavatedby John D. Speth, of the Anthropology Mound B; additional small samples were found in a test trench at Godin Tepe. Departmentat Hunter Collegeof the City Universityof New
1 The Hasanlu Projectis directedby Robert H. Dyson,Jr.; at York, where they are currently stored. There may be additional

111

112

JOURNAL

OF PERSIAN

STUDIES

An outline of the recordedstratigraphy,architecture,burials, and artifactsof Dalma Tepe is followed by brief references to other known Dalma occurrences and to the chronological status of the Dalma assemblage insofar as it is currently defined. Stratigraphy As is indicated in Table I, a total of four trenches was opened at Dalma Tepe.5 In combination, these covered an area measuring approximately Ioo sq. m., less than one-tenth the mound's surface (see Fig. i). Young's 1961 operations III and IV were laid out north-east of Burney's Trenches A (1958)

Fig. I. Dalma Tepe, 9g61. Contourinterval I metre. Scale I : 2oo.

G

(After roung.)

In 1958, Trench A (5 m. x 2 m.); in 1959, Trench B (7 m. x 3 m.); in Ig96, Operation III (7 m. m.) and Operation X4"5 IV (7 m. x4' 5 m.). The contour map (Fig. I) indicates the

size FollowingYoung, approximate and locationof the trenches. TrenchesA and B are also designated OperationsI and II respectively.

DALMA

TEPE

113

Table I
at Summary Stratigraphic of Sequences Dalma Tepe.

1958

1959

196I

AI

BI

A2
- -

B2
B3 B3A

2
3 4 5 5A 6 (water table at
--3" 2 metres)

- (floor)
A4

A3

---

(floor)

A5 - - - (floor) A6 (water table at

--2 8 metres)

and B (1959). Every year it was necessary to cease excavation before reaching virgin soil because the water table was at a depth of 3 to metres from the mound's surface. In all areas the uppermost 3"5 metre of deposit consisted of loose topsoil disturbed by numerous post-Dalma graves. These were cut into wash which contained some Dalma sherdage (Young's stratum 3). Below this wash was a series of deposits containing architecturalremains. The two architecturalstrata recorded by Young (his strata 4 and 5) overlay a deposit containing some Hajji Firuz sherdage (stratum6) but without clear architectural remains. Hence, for the purposes of this review, the significant deposits (i.e., those relatively undisturbedand yielding Dalma ceramicswithout substantialadmixture of either earlieror later sherdage) are Young's strata 4 and 5 (see Fig. 2). Architecture The only coherent architectural units were those cleared in 1961; while Burney recorded portions of walling both in plan and in section, these remains do not relate clearly to those exposed in I961. Fig. 2 reflectsthe relatively complex architecturalsituation at Dalma; there is evidence for at least four building phases in strata 4 and 5 (see the southeast baulk of Operation IV). The plan in Fig. 3 illustrates structuresexcavated in Young's strata 5 and 5A; these consistof portions of two free-standingstructures separated by a large open courtyard." In one of these, only one area (perhaps roofed) was partially cleared; the more extensively-excavated northern unit contained at least three enclosed areas. A fourth central area (Operation III [I]) may have been an interior open courtyard. A large thin-walled bin-like structureextended to the south of this building, and a curtain wall with a rounded buttress-like end ran off to the south. A hearth was exposed in the open area between operations III [3] and IV [31]. The walls at Dalma were constructed of chineh(layers of packed mud), separated in some cases by thin mortar-like clay layers, and were preserved to heights ranging from 15-6o cm. The remains of earlier wall stubs tended to be shorter than those of later deposits. On some walls and floors, relatively thin plaster coatings were found; sometimes these were white, but more often consisted of uncoloured mud plaster (gach). There was no evidence of either coloured plaster or wall painting.
6

On the plan (Fig. 3), Roman numerals indicate Young's 1961 Operations; Arabic numerals in squares are excavation areas

within these operations.

32

I CI'A

IA

AA

AASH

SA
,** n

i n~

w -or

z
uIas

so

III

NORTHWIEST

III

NCE

H AS

*)

3A
38 DATUMPU

3 4 A 4 ASH 4
ASM 4

3 0I LENSES
4

BURIAL

L.NS'IS$

AS S

ASH HAND
S IV IA NO0R1T
T

ASH
AB SOUTHEAST

Fig. 2. Dalma Tepe, r961. (After roung.)

DALMA

TEPE

115

rI

/-

B /

,11S

DABMA

TEPE

19B61

1]00

CM.

STRATUM

5. Fig. 3. Dalma Tepe, 9g6z. Plan of Stratum (After Toung.)

Burials Sixteen burials of Iron Age (Hasanlu V) and later date were found above the uppermost Dalma architecture. These burials invariably contained adults, interred individually in diverse positions,with few grave goods. In addition to the I6 post-Dalma burials, cut into Young's strata 3 and 4 by people who apparently preferredto use the mound only as a cemetery, 14 Dalma-period burials were cleared during the 1961 soundings. These primary burials consisted of tightly contracted single infants placed in ceramic vessels. The depth and location of the burial pots varied; while none was actually found clearly associated with room-floors, several of the 14 were recorded in areas interpreted as outside courtyards. There is no clear evidence of grave goods in the form of artifacts, animal bone, or remains of shroudsor clothing. The infants were buried neither with adults nor with juveniles; remains of adult occupants of the Dalma settlement were never found. Lacking the excavated faunal collection, which was lost in shipment and therefore never analysed, it is impossible to be sure that human bone was not also mixed in with animal remains. Artifacts As noted above, the brevity and exploratorynature of the Dalma excavations produced a relatively small artifactual sample. Horizontal distributions of most artifacts were not recorded, because they occurred, for the most part, in secondary deposits rather than on primary occupation surfaces. At the time of the Dalma soundings, limitations on time and labour precluded sieving, and flotation had not

116

JOURNAL

OF PERSIAN

STUDIES

yet been developed as a retrieval technique. Objects of small size, such as debitage, plant remains, and small bones, were thus generally not obtained. The largest artifact category retrieved consists of ceramic vessels and sherds; small quantities of bone, ground stone, chipped stone, and miscellaneous clay objects were also retrieved. No metal objects can be attributed with certainty to the Dalma period occupation. Stone Chipped A few pieces of chipped stone are mentioned in Burney's notes; there is no record of attributes or provenience. Of the 45 to 50 pieces of chipped stone (exclusive of obsidian) recorded as coming from Young's strata 4 and 5, approximately half are currently in North America; the following remarks thereforepertain only to this portion of the assemblage. All the pieces examined were relatively large blades and flakes (see P1l.I). The sample contains neither cores nor " debitage ", nor any identifiable stone-knappingobjects other than one or two possible hammerstonefragments. Of the 16 pieces examined, seven are blades; four of these have sheen on at least one edge, and may have been used as sickles. The remainderof the non-obsidian sample consists of flakes,of which only one has a slightly shiny edge. Few of the i6 pieces were struckoff the same parent rock (in most cases flint or chert); almost all have some form of edge wear, occasionally consistingof tiny irregularnicks but more often of steeper, deeper, and more regularflake scarson one or both edges. All occurrencesof sheen are on edges with such scars. One rather thick flake is heavily battered along both edges, and resembles an outil 'cailli.7 There is no clear evidence for snapped blades; no microliths, points, scrapers, burins, or other clearly recognized " types " occur in the sample examined. Approximately 15 pieces of obsidian were found in 1961, supplementingthe tiny collection retrieved in the earlier excavations. In contrast to the flint artifacts, those of obsidian tend to be relatively small (the longest piece, which falls in the size range of the majority of the flint blades and flakes, is 7 -6 cm. long). There are no cores in the obsidian sample, most of which is black or greyish-black in colour; one piece has a slightly greenish cast. Three of the obsidian pieces now in North America are small blades, all from stratum 5; the remaining 3 are flakes, one of them possibly a point. Stone Ground The 1959 excavations produced a fragmentary pestle, a hammerstone, and part of a stone vessel. No stone hoes or grinding stones (querns, mortars, pestles) were found during 1961. The apparent absence of such objects suggeststhat the grinding of foods, pigments, and other items was not carried out within the area exposed. Recorded ground stone objects (see P1l. include a " loom weight ", a " whet I) stone ", a flattened pebble interpreted as an " awl sharpener", a " chisel ", and a green pendant. In contrast with the Dalma assemblagefrom Seh Gabi, none of the spindle whorls is of stone. Bone Relatively few bone artifacts (9) were retrievedin the 1961 sounding; most of these came from stratum 4. Most of the bone objects are awls; each of these is partially polished, and one of them bears, in addition, short oblique scratches on its convex longitudinal surface. There are recorded examples of awls both with and without articularsurfacesof the foot bones employed. In addition to the awls, there
is a notched rib fragment, and a long bone with rounded tip which may have been a scraper (see P1. I). No bone beads, pendants, or shell objects were recorded from any of the Dalma soundings. Clay Spindle whorls of baked clay comprise the largest group of clay objects; all but one of the 9 found in 1961 came from stratum 5. Most of the spindle whorls are plain conical objects (see, for example, P1. I), but one has fingernail incisions on its base. A whorl now in Tehran has a rounded biconical shape,
SThe piece is 61-163; see Michel Br6zillon, La Denomination des Objets de Pierre taill6, IVe Suppl6ment A Gallia-Prthistoire (1971), p. 288.

DALMA TEPE

117

with finger impressionsin a horizontal row around its circumference. In addition to spindle whorls, the clay artifactsinclude a " clay weight ", a "sling pellet ", a perforatedsherd disc, and a small piece of clay, possibly of the sort occasionally referredto as " gaming pieces " for lack of a more secure interpretation. No animal or human figurines were reported in any year. Ceramics The bulk of the Dalma ceramics is in sherds. In addition to the sherdage not discardedin the field, the excavated sample includes 23 vessels recorded in Burney'snotes and 31 complete or near-complete vessels obtained in 1961. Fewer than half of the complete vessels are now in North America.8 Both Burney and Young observed a change from relatively high proportionsof painted pottery in the lower levels at Dalma to higher proportionsof surface-manipulatedwares in the upper levels. A second noteworthy feature of the ceramic assemblage recorded in both sets of field notes is that ceramics of " Dalma " type were stratifiedbelow Pisdeli and above Hajji Firuz ceramic types.9 There are no clear variations in the samples obtained in different soundings. The following brief description and accompanying illlustrations of ceramics indicates the range of wares and shapes in which they were fashioned. All of the Dalma pottery is handmade, most frequentlyof heavily chaff-temperedclay with small grit inclusions. While generally fired to a pink- to orange-buff colour, cores or cores' centres are often grey. The pottery tends to fracture unevenly, and is relatively friable. A few sherds are marked by a greenish tinge, and in some cases the paint became slightly vitrified in firing. There are four major categories of surface treatment in the ceramics of strata 4 and 5. These are (i) painted, (2) surface-manipulated, (3) red-slipped,x0and (4) plain. A large proportion of the available ceramic sample consistsof diagnostic body rather than rim sherds. On the basis of those rims now in the University Museum, there appears to be a positive correlation between specific shapes and specificsurfacetreatments. A few shapes were produced in several size/volume classes; for the majority of shapes so few sufficiently complete examples exist in the collection that variations in volume or the frequenciesof vesselsin their original cultural contexts cannot be determined. The main characteristics of each category of surface treatment are outlined below; some examples of shapes associated with these categories are illustrated. Dalmapainted: Painted decoration occurs on a relatively large number of vessel shapes. The most frequently painted shape, which occurs in several sizes, is a flat-based roughly globular pot with plain pinched rim, slightly inverted (see Fig. 4). Other shapes are footed " chalices " (Fig. i oB), flaring everted bowls (Fig. 6C), straight-sided bowls with everted flat-bevelled rims (Fig. 5F), and pots with high angular shoulders and short, relatively narrow necks. In this last category a few examples have a small shoulder perforation, presumably for suspension or the attachment of a lid (Fig. 6D). A few miniature bowls are illustrated here (Figs. 7E, IoA, C). Painted decoration is often applied over a thick matt cream or white slip. It is invariably monochrome and is most usually restricted to vessels' outer surfaces; occasionally, in flaring open-mouthed vessels, the interior is also painted with a geometric motif and, rarely, the motif used on a vessel's outer wall continues over its base. In most cases, however, the interiors of painted vessels are either left untreated or covered with a thick plain red " slip " or paint.
Decorative motifs, in matt red paint ranging to a purple or brown shade, usually cover the entire vessel wall. The motifs consist of a limited number of geometric shapes, repeated alone or in various combinations. The geometric motifs themselves are painted solid, cross-hatched, striped, filled with lines paralleling the outer edges of the shape, or merely drawn in linear outline. In one case, shown in Fig. 4D, the vessel exterior is decorated with two completely different sets of motifs. While the variation
8 The total size of the sherd sample examined by the present author is approximately 550. 9 For information concerning Pisdeli Tepe, see Robert H. Dyson, Jr. and T. Cuyler Young, Jr., " The Solduz Valley, Iran: Pisdeli Tepe ", Antiquity 34 (196o), pp. 19-28. The Hajji Firuz ceramics will be extensively described in the disser11

10

tation of Mary M. Voigt, to be submitted to the University of Pennsylvania during 1975This " slip " is sometimes quite thick and uniform, and might equally be classed as " paint ". This distinction might be clarified by thin-sectioning.

118

JOURNAL

OF PERSIAN

STUDIES

in painted decoration at firstappearsalmost infinite, the most common motif combinations may actually be clustered in four groups: (a) inverse triangles suspended from the rim, with the rest of the vessel wall filled with horizontal bands of lozenges, either free-standingor connected at their corners, (Figs. 4C, E, 6F, H). Rim triangles and body lozenges are wholly or partially filled with stripes, solid paint, or cross-hatching. (b) inverse triangles suspended from the rim, with horizontal rows of pendant inverse triangles below (Fig. 4D right; Fig. 6B). (c) oblique stripes or thick irregularslashes, extending from the rim to the vessel base (Figs. 4B, 5). These are more often arrangedin clusterswith unpainted spaces between them than in bands uniformly distributed over the vessel wall. A variation on this theme is seen in one motif consisting of large sloppily painted and internally striped zigzags which extend from the rim down the wall toward the base (Fig. 4A). (d) inverse triangles suspended from the rim, with thick zigzags below them on the body (these are either solid, filled with parallel lines, or cross-hatched;see Figs. 4D left, 4F, 7E). Several sherds painted with Dalma motifs are so highly fired as to have a " clinky " sound when lightly struck; these have a dark red body, are relatively thin-walled, and are of slightly denser clay than most of the Dalma ceramics. Also noteworthy is the unusual and possibly imported vessel illusDalma surface-manipulated: There is a wide variety of surface treatmentswhich break up the smooth vessel surfaces; these appear restricted to a small range of shapes. Because the various surface treatments are not limited to impressing, the more inclusive if awkward term " manipulation " is used here to cover the wide range of treatments. Many examples of surface-manipulatedsherds also have red slip or paint, usually applied over the textured surface. There are no recordedexamples either of a combination of different techniques of surface-manipulationon a single vessel, or of a combination of surfacemanipulation with geometric painted decoration. Many examples of the outer walls of surfacemanipulated vessels are blackened in patches, suggesting that they were used in cooking. There are, parenthetically, no vessels comparable to Hassuna " husking trays ". Techniques of surface-manipulationare illustrated in Pls. I and II, and include (a) shallow finger-tip impression (see also Fig. 8D), (b) jabbed oblique fingernail impression, (c) jabbed shallow holes, which appear to have been made with a twig, (d) comb-impression (see also Fig. 8E), (e) reed-impression, (f) small straight slits, sometimes arranged in oblique rows, and (g) slug-like applique lumps (termed " barbotine " in some earlier publications; see also Fig. 8A). It is possible that those specimens formerly classified as examples of" stick-impression" may have been " textile-impressed" and some, at least, of those termed " comb-impressed" may have been decorated by pressing the soft clay with a cord-wrappedstick." While individual " motifs " on a vessel wall are often arranged in irregularrows, they always cover relatively large zones of the vessel wall, though they are occasionally restricted to the shoulder. The interior walls of surface-manipulatedvessels are either untreated, or covered with a matt red slip. In no cases are both the interior and exterior surfacesmanipulated, or the interior surface alone surface-manipulated. Macroscopically, the fabric of surface-manipulatedvessels is in the same range as that of painted
vessels, though a few pieces appear to be made of somewhat denser clay containing less chaff, and a few pieces are somewhat more friable and soft than the majority of painted vessels. Shapes whose surfaces are treated in any of the ways enumerated above include large pots with slightly inverted pinched rims (Fig. 8A, B, C), vertical-sided pots (Fig. 8D, E), small globular pots which often have vertical lugs or handles, and a relatively shallow vertical-sided heavy tray-like bowl, whose outer base bears shallow fingertip impressions. trated as Fig. IoI.

xxThis in the opinion of Frederick Matson, according to Dyson (" Early Cultures of Solduz, Azerbaijan ", in A Survey of Persian Art, VolumeXIV, edited by Arthur Upham Pope and

of Phyllis Ackerman(Proceedings the IV InternationalConPartA (1967),PP. 2951gressof IranianArt and Archaeology, 29o70)

DALMA

TEPE

119

Dalma red-slipped: The fabric of vessels covered with a solid red pigment is characteristicallyfired to a salmon colour; the clay includes both chaff and small white grits. Fracture is sometimes quite even, and the pottery in general is somewhat less friable than that of surface-manipulatedvessels. Exterior surfacesand, often, interior walls, are covered completely (if sometimes by varying thicknesses of the coloured solution) with a dark red slip or paint. This tends to be maroon, but is sometimesso dark as to be termed " plum " colour. Frequent mottling on the surfacesof red-slippedvesselsis presumably due to variations in oxidation during their original firing and/or to their use in cooking. The coloured surface coating is usually quite thick, often crackled and partially flaked away. In a few cases, a thin solution of the pigment appears to have been smeared on the vessel with a rag. The examination of broken edges suggests that the red pigment was occasionally applied over an undercoat of white or cream slip. A few rare cases of burnishing have been observed in the Dalma red-slipped ceramics. A large variety of shapes is treated with this solid red slip or paint: pots with short everted necks (Fig. 9D), large globular pots with inverted rims (Fig. 9E, F), small globular pots with plain and slightly inverted rims (Fig. 9C), small pots with interior ledges just below the everted neck, flat-shouldered hole-mouth pots, very shallow trays or plates (Fig. 7G), bowls with flaring walls and pinched rims (Figs. 7F, 8F, ioF, G), and the problematical " decanting vessels " (Fig. 9M). Flat objects, probably lids, and the so-called " horn " lugs (Fig. IoD) also usually have a solid red slip covering all surfaces. In general, there appears to be a larger number of necked and large-size vessels decorated with red slip than with other forms of surface treatment; possibly this was a preferredtreatment for storage vessels. In addition, several of the miniature bowls excavated at Dalma are decorated with red slip (Fig. 7F). In one of the rare combinations of different forms of surface treatment, a flaring-sidedbowl with an external red slip has a large cross painted in red over its inner surface (extending from the rim across the inner base and up the opposite side). Dalmaplain wares: The surfaces of several vessels are neither painted, covered with red slip, nor display evidence of any form of surface treatment besides smoothing. While it is possible that some of matt, some of them smoothed, and some quite rough-surfaced,but all of pale buff colours. Surface coatings would have to be identified in microscopic analysis. Shapes subsumed in this somewhat catch-all category include trays (Fig. 9J, K, L), everted bowls (Fig. 7H, I), large pots (Fig. IoH), elongate globular short-neckedvessels in several sizes (Fig. 9A, B), and both deep and shallow inverted bowls (see Fig. 91). Chronology
Table 2
Radiocarbon Determinations to Relevant theDating of Dalma Tepe. (uncorrected)

the vessles included in this category were either " self-slipped " or even " cream-slipped ", most are

Period

Site

Sample No.

5730 Half-life (B.c.)

Pisdeli Pisdeli Pisdeli Pisdeli

Pisdeli (?) Dalma Hajji Firuz Hajji Firuz

Pisdeli Pisdeli Pisdeli Kush Ali Yanik Tepe

Dalma Hajji Firuz Hajji Firuz Yanik Tepe

P-i57 P-504 P-5o5 P-866

P-1254

3666 ? 165 77 3734?- 8 3857-? 74 3658?

3293? 58 84

Pisdeli (?)

P-I246 P-50o3 P-I843 P-502

3475? 75
4215

5128?1 07 79 5152?

Hajji Firuz Hajji Firuz (?) Hajji Firuz (?)

Hajji Firuz Yanik Tepe Yanik Tepe

P-455 P-I343
P-1244

5536? 88
5184? 5297? 82

71

120

JOURNAL

OF PERSIAN

STUDIES

Only one carbon sample obtained at Dalma Tepe itself has been analysed. Several samples from the Dalma component on Mound B at Seh Gabi should be helpful in establishing the absolute date of this assemblage. The best data currently available for the chronology of the Dalma assemblage consists of a series of carbon samples for both the Hajji Firuz period (from Hajji Firuz and the Yanik Tepe " neolithic ") and the Pisdeli period (from Pisdeli and Yanik's " chalcolithic "). Hence the relative stratigraphyin the Ushnu-Solduz valley system permits the use of earlier and later materialsto bracket, in absolute time, the material from Dalma Tepe. Relevant available carbon dates are shown in Table 2. Using these uncorrected carbon determinations, Hasanlu VIII (Pisdeli) may be placed roughly in
the period c. 3900-3200 B.C. and Hasanlu X (Hajji Firuz) in the period c. 5600-5100oo B.c. This allows a placement of the most poorly-dated period, Hasanlu IX (Dalma), in the period c. 5000-4000 B.C.

Material Comparative A review of the available literature suggests that the Dalma ceramic assemblage is confined to the northern and central Zagros mountains. Dalma-like pottery of various kinds has been excavated at Hajji Firuz, Pisdeli, Hasanlu, Dinkha Tepe, Yanik Tepe, and Tepe Seavan in the Lake Urmia basin; Tepe Siabid, Seh Gabi, Tepe Giyan, and Baba Gassem in the Kermanshah region; and Kunji Cave in the Khorramabad area.12 Dalma-like pottery is also known from Takht-i Suleiman, Nuzi and Tepe Gawra in northeastern Iraq, and possibly from Soviet Azerbaijan.13A more distant and possibly fortuitous series of ceramic parallels occurs in the eastern European Starievo complex.14 In Mesopotamian terms, Pisdeli painted ware is generally treated as correspondingchronologically and to some extent typologically to late Ubaid, and Hajji Firuz ceramics to Hassuna.15 If only by default, Dalma thus falls into the Halaf time slot. However, no Halaf-type sherds have been reportedin published Iranian sites having Dalma material. There are, interestingly,a few sherdsfrom the Iraqi site Tell Arpachiyah which bear painted Halaf motifs on one side, and Dalma-like surface manipulations on the other.e6 The analysis of Dalma-like ceramics, seals, and other artifacts from Seh Gabi near Kangavar should allow us to pinpoint more specific typological parallels which presumably exist between the assemblages of Dalma Tepe, northern and southern Mesopotamia, and the Iranian plateau, and will be reported in a future publication.

x1

On the material excavated by the Hasanlu Project, see preliminary accounts in Dyson op. cit.; T. Cuyler Young, Jr., " Taking the History of the Hasanlu Area Back Another Five Thousand Years ", IllustratedLondon News, No. 241 (1962), pp. 707-709; T. Cuyler Young, Jr., " Dalma Painted Ware ", Expedition5 (1963), PP. 38-39. For Yanik Tepe, see Charles Burney in Iraq XXIV (1962); for Baba Gassem and other sites in the Kangavar-Assadabad area surveyed by Young, see his article in Journal of Near EasternStudies25 (1966). Information concerning Dalma materials at Tepe Seavan was conveyed in a personal communication by Drs. Rose and Ralph Solecki; concerning Tepe Siabid, by Frederick Matson; concerning Tepe Giyan, by Louis D. Levine and T. Cuyler Young, Jr., concerning Kunji Cave, see note 3 above. 13 For information concerning Takht-i Suleiman, see Dyson op. cit., p. 2953; on the Soviet material, see Charles Burney and David Lang, The Peoples of the Hills (1972), p. 38. The Nuzi material, like Dalma surface-manipulated wares, is illustrated in Richard F. S. Starr, Nuzi II (1939), pl. 45. The Tepe Gawra parallel is a painted bowl, illustrated in Arthur J. Tobler, Excavationsat Tepe Gawra II (1950), pl. CXXVIII,

no. 187. A number of additional surface occurrences of Dalma ceramic types are recorded in the project notes of The Hasanlu Project and were kindly provided by R. H. Dyson, Jr. 14 See for example "Die Star'evo-Kultur im SlawonischSyrmischen Raum und das Problem des Ubergangs vom Alteren zum Mittleren Neolithikum in Serbischen und Kroatischen Donaugebiet ", in Neolit i Eneolit U Slavoniji (edited by Stojan Dimitrijevj', 1969); also, Vladimir Milojai6, Chronologie (1949) derjiingerenSteinzeit Mittel und Siidosteuropas The impression that there are parallels with Dalma finger-tip impressed wares, in particular, has been confirmed in discussions with Ruth Tringham and Steven Kaufman. 15 For example, Robert H. Dyson, Jr., " Problems in the Relative in Chronology of Iran, 600o-2000 B.C.", in Chronologies Old WorldArchaeology (Robert W. Ehrich, editor, 1965), pp. 215256, and particularly p. 249. Certain typological parallels between Hassuna and Hajji Firuz ceramics have been pointed out to the author by Mary M. Voigt (see note 9). 16 Observed by the author in the study collection of the Institute of Archaeology, University of London; permission to look at this collection was graciously given by David Oates.

DALMA TEPE

121

Fig. 4. Dalma Tepe,paintedwares (approximately i:4).

122

JOURNAL

OF PERSIAN

STUDIES

Fig. 5. Dalma Tepe,paintedwares.

DALMATEPE

123

Fig. 6. Dalma Tepe,paintedwares.

124

JOURNAL

OF PERSIAN

STUDIES

Fig. 7. Painted,red-slipped G), andplain (H-J) wares. (F,

P1. Ia and b. Dalma Tepe,smallfinds.

Pl. Ic. Dalma Tepe,sherdswith surface manipulation.

with manipulation. P1. IIa-d. Dalma Tepe,sherdsdecorated surface

DALMA

TEPE

125

and (F) Fig. 8. Surface-manipulated red-slipped wares.

Fig. 9. Plain and red-slipped (C-F, M) wares.

Fig. so. Painted(A-C), red-slipped F, G), andplain wares; paintedbowl (I), possiblyan import(rimdiameter 27.5 cm.). (D,

EARLY FOURTH MILLENNIUM DEVELOPMENTS IN SOUTHWESTERN IRAN By H. T. Wright,J. A. Neely, G. A. Johnson and John Speth
That urban society and states arose in Mesopotamia during the Uruk period is widely accepted. The end of the 'Ubaid period and the beginning of the Uruk period have been the object of much work. In fact, though individual workers have attributed particular manifestations to such a transitional period, there is no adequate taxonomic or chronologicalframeworkfor this period, much less an understanding of its fundamental changes. In this paper such a framework is outlined in the hope that it will facilitate future efforts toward such an understanding. Our geographical focus is what is now south-westernIran, including both the valleys of the central Zagrosand the broad upper Khuzistan plains (Fig. I).1 Within this area four important stratigraphies, all of which are at present published in preliminaryform at best, span the period of interest. Those at Tall-i Ghazir2, Susas and Farukhabad4 are in the plains. That at Godin Tepe5 is in the highlands.
2
2

Ay

,
Site.

_ r--r - ,-.-L.L r ,
r.

A3P,

,>,

A
Gulf
A

AB

m5
c,

Gulf

->..A^

-~-N-

?A

A-

?~G

I f

Fig. r. South-west Iran in the early Fourth Millennium B.C.: I, Kunji Cave; 2, Baba Jan Tepe; 3, Kuh-i Dasht; 5, KS-269; 6, Tall-i Ghazir.
x Speth's work at Kunji Cave was supported by National
2

4, Sargarab;

Science Foundation Grant GS-2402. Neely's at Sargarab was supported by GS-2194; and Johnson's at KS-269 was supported by GS-3147, both from the same source. Wright was able to visit the first two projects with support from Grant FRR-835 of the University of Michigan Horace Rackham School of Graduate Studies. None of these projects would have succeeded without the full cooperation of the Iran Archaeological Service, and the encouragement of the staff of the British Institute of Persian Studies and the American Institute of Iranian Studies.

der und J. R. Caldwell, " Tall-i Ghazir ", Reallexikon Assyriologie Vorderasiatischen III Archeologie (Berlin, 1968), pp. 349-55. 3 Robert H. Dyson, Excavationson the Acropolisat Susa and the Problems Susa A, B, andC, (Ph.D. thesis, Harvard University, of 1966); Alain Le Brun, " Recherches Stratigraphiques . L'Acropole de Suse (1969-71) ", Cahiers de la Diligation Archiologique Franfaiseen Iran I (1972). 4 Henry T. Wright, " Farukhabad ", Iran VII (1969), pp. 172-3. 6 T. Cuyler Young, Jr., "Excavations at Godin Tepe: First Paper No. 17 of the Royal Ontario Progress Report ", Occasional Museum (Toronto, 1969).

129

130

JOURNAL

OF PERSIAN

STUDIES

We owe a fundamental debt to the excavators of these stratified sites and to Dr. Frank Hole and Dr. Clare Goff whose surveys in the Pish-i Kuh region of the central Zagros, made available as we were writing this paper, clarified our thinking on many problems. However, the ceramics which are illustrated here derive primarily from soundings and surface collections from a number of other small sites in south-western Iran, none of them previously published. Before turning to these data, let us briefly summarize what is known of settlement organization and ceramic craftsin the Late 'Ubaid period and the Middle-to-Late Uruk periods in south-western Iran. By Late 'Ubaid times human settlement had spread to every part of greater Mesopotamia. Every lowland river bank and every highland valley contained small villages or hamlets, and most larger tracts of easily cultivable land had one or more small centres covering up to ten hectares. The occupants of both the centres and the villages were herding sheep, goats, and cattle, and cultivating wheat, barley, and other crops. Ceramics and stone tools were certainly produced in both villages and centres, although the production of goods with special social functions may have been restrictedto the centres. Differences in residential architecture in both types of settlement and differentation in burials suggest the presence of social ranking. The relative size and elaboration of the centres probably indicates they were the seats of the richer and politically more influential kin groups. Craftsmenof Late 'Ubaid times probably used some type of potter's wheel to produce many of their vessels. This is suggested by the symmetry of the bowl forms and the structureof the clay body, as most vessels were subsequentlyscraped and smoothed until all trace of surfacewheel scoring was obliterated. Many vessel forms were individually slipped, or painted in bold geometric and naturalistic motifs, or both. The end product of this individualized approach to vessel production was a series of simple basins, bowls, and high-neckedjars with a wide variety of designs. By Middle to Late Uruk times, the inhabitants of Greater Mesopotamialived in a different manner. Communities were less evenly distributed and more differentiated. In several areas of the lowlands, there were settled enclaves centred around large towns of 20-50 hectares. In south-western Iran the main enclave was centred around Susa and Chogha Mish on the Susiana plain. There were several types of smaller settlements including smaller towns, centres involved in administration, villages involved in specific crafts like stone-working,and villages probably involved solely in food production. Portions of the towns were devoted to large buildings in and near which there is clear evidence of record-keepingand other administrative activities, but not of direct involvement in agricultural work or even in food preparation. Most types of ceramicswere made solely in the towns and were distributed through smaller towns to the villages. Craftsmenof the Middle and Late Uruk periods used both the potter's wheel and rough press moulds. Vessels were seldom subsequently finished or embellished. The end product was a variety of standardizedbowl and jar shapes with little decoration in comparison with earlier 'Ubaid assemblages. Without question Middle and Late Uruk societies were highly differentiated and were coordinated by some kind of bureaucracy.6 In contrast, the smaller lowland plains and the nearby highland valleys of lower Luristan contained little Uruk settlement. Only isolated small villages, perhaps important as herding centres or as transferpoints in the nascent trade network,' occurred. There are several series of Carbon-14 datings on later 'Ubaid assemblagesin south-western Iran, for instance on the earlier Late 'Ubaid Bayat phase of the Deh Luran plain,8 and on the Susa A phase
of the Susiana plain. It is now possible to determine absolute ages from such datings.9 Taken together, these suggest that the 'Ubaid period ended c. 4000 B.c. rather than somewhat later as had been thought. There is no such series of dates on any phase of the Uruk period; however one can extrapolate from the
" Local Exchange and Early State (Ann Arbor, I972). Development in South-western Iran ", Anthropological Paper 8 Frank Hole, Kent Flannery, James Neely, and Hans Helbaek, No. 51, Universityof Michigan Museum of Anthropology " Prehistory and Human Ecology of the Deh Luran Plain ", (Ann Memoir No. I of the Universityof Michigan Museum of AnthroArbor, 1973). 7 Henry T. Wright, "A Consideration of Inter-regional Expology(Ann Arbor, 1969). 9 Ingrid U. Olsen, RadiocarbonVariationand AbsoluteChronology change in Greater Mesopotamia 4000-3000 B.C." in Edwin Wilmsen, " Social Exchange and Interactions ", Anthropological (Nobel Symposium No. 12, New York, I970). Paper No. 46, Universityof Michigan Museum of Anthropology
6 Gregory A. Johnson,

EARLY

FOURTH

MILLENNIUM

DEVELOPMENTS

IN SOUTHWESTERN

IRAN

131

better dated Early Dynastic period. Such extrapolation and the few available Uruk determinations suggest that the Middle Uruk period began about 3400 B.c. Thus, what has been called the Early Uruk covered the surprisinglylong span of 600 years. This can be subdivided into two portions. The later is the Early Uruk period proper, in which such forms as the bevel rim bowl and the straight spoutedjar appear throughout the lowlands in quantity. We consider the ceramics from Warka Eanna XIII to be the type assemblage.10 The earlier subdivision, which may be called the Terminal 'Ubaid period, is one of considerable diversity and it is impossible to speak of a type assemblage. It is on this Terminal 'Ubaid period that we wish to focus for the rest of this paper, though we shall refer to Early Uruk developments when relevant. The Transition Luristan from 'Ubaidto Urukin Highland We focus here on the valleys of Khorramabad, Kuh-i Dasht, Rumishgan, and Tarhan. The small rockshelterof Kunji Cave faces west acrossthe KhorramabadValley. The valley floor is about I200 m. above sea level, while the cave is about Ioo m. higher at the top of a steep talus and near the base of a limestone cliff. The cave measures about 4 m. in height, 18 m. in width and 28 m. in depth. Because of a large rockfallduring the Pleistocene period, only the front half of the cave was occupied during the Late and Post-Pleistocene periods. Excavations have been 'onducted at the cave by Frank Hole of Rice University in 1963 and by John Speth, of City University, New York, in 1969. Fortunately, the occupational layers in which we are interested were relatively undisturbed by the animal burrows which elsewhere in the cave destroyed important features. During the transitional period after the end of the 'Ubaid, a series of at least four small circular arrangements of stone along the edge of the rock fall were built. From these constructions the floor sloped downward towards the south side of the entrance. A thin layer of yellow clay which contained almost no artifacts was deposited in front of these stone arrangements. A low terrace or wall of rough stones marked the front of this relatively clean area. Below this wall was a thicker deposit of yellow clay which contained a number of broken ceramic vessels, described below. The wall or terrace later collapsed down the slope on top of the clay layer, sealing the ceramics from later disturbance. There were few artifactsother than ceramicsin this sealed layer. Similar ceramicswere also found in disturbed layers in other portions of the cave. The activities carried out in the circular stone arrangementsare not at present known, although they may have served as animal pens, or shelters for small groups of people, or both. The clay body of the ceramics from the sealed layer was tempered with crushed straw. Minor inclusions of crushed limestone and sand could have been accidental. Most vessels are probably wheelturned. The surfacecoloursare very light brown to gray. The larger examples have a coarserpaste and scraped or smoothed surfaces. The smaller examples have a fine paste and are often slipped and burnished. It is not yet possible to say which bases match which upper bodies. In the following descriptions the number in parentheses preceding each statement refers to a unified numbering system used for forms from all the sites. The following bowl rims occur in the sealed layer. (i) SimpleRoundLip Bowls(Fig. 6a, b). The upperbody of theseopen formsis conical. The rimis bent
slightly outward. The form is often burnished. (5) Simple Beaded Lip Bowls (Fig. 6n, o). These shallow open vessels have a heavy thickening on the outside of the rim. Few examples are burnished. The second example is atypical. (7) Small Incurved Bowls (Fig. 6m). This shallow form has a markedly closed rim with rounded lip. The sides are relatively thick. Few are burnished. (9) Small Incurved Beaded Lip Bowls (Fig. 6i, j). These have a globular body with a slight closing of the rim. The lip is thickened on the exterior. The form is usually burnished.
10

A. von Haller, "Vierter Vorlaufiger Bericht iiber die von der Notgemeinschaft der Deutschen Wissenschaft in Uruk Unternommen Ausgrabungen ", Abhandlungen der

AkademieDer Wissenschaft:Philosophisch-Historische Preussischen Klasse 6 (1932), PP. 38-42.

132

JOURNAL

OF PERSIAN

STUDIES

The following jar rims are present in the sealed layer: (i) Fine Jars (Fig. 6c-h). The lower body of this form is probably globular; the upper body is concave with a slightly flared rim. It is often burnished. Several sherds of this vessel form, unfortunately from disturbed contexts, have black or dark brown painted designs. Some (Fig. 6g, h) have two horizontal bands, one near the rim and one near the point of maximum body diameter. Various vertical elements are placed between the two bands. One rim (Fig. 6f) has a chevron design reminiscent of Siyalk II11 and Godin VI.12 (4) Flared Neck, Round Lip Jars (Fig. 6k, 1). These also have a globular body, but the sides are thicker. On this body a high, narrow, constricted neck is added, as can be clearly seen in the second example. Few of these are burnished. Several body sherds suggest a smaller low necked jar form, but rim sherds of such jars have not been found in the sealed layer. A variety of base forms occurs. (I) Flat Bases (Fig. 6p, q). Some are scraped and some are wheel finished. (2) Ring Bases (Fig. 6r, s). These are clearly wheel finished. (3) Pedestal Ring Bases (Fig. 6t). This base flares downward. It has a distinctive flattened vertical lip. Preliminary counts of the total sample of 446 rims from the disturbed parts of the excavations by Speth, taken as a whole, revealed that 56 per cent were of the six categories defined above on the basis of the sample from the sealed layer. Counts of the rim sherds from both the sealed yellow clay layer and from the total sample are presented in the following table. CERAMICS FROM KUNJI CAVE Sealed Layer (I) (5) (7) (9) (I) (4) Round Lip Bowls Simple Beaded Lip Bowls Small Incurved Bowls Small Incurved Beaded Lip Bowls Fine Jars Flared Neck Jars Total Counts
21 2 2
I3

Total Sample 57
20

13 9 60o

34 44 83 30 251

In summary, the ware from Kunji Cave can be divided into two types, a thin burnished type, used for small to medium sized bowls and jars, and a thicker unburnished type, used for medium to large sized bowls and jars. A ceramic assemblage similar to that from Kunji Cave has been reported by Goff'3 from the Central Mound of Baba Jan, 70 km. to the north-west of Khorramabad. Four bowls from Baba Jan also occur at Kunji Cave, (i) Simple Bowls with Round Lip,14 (5) Large Simple Bowls with a Beaded Lip,15 (7) Incurved Bowls with a Round Lip,le and (9) Small Incurved Bowls with a Beaded Lip.17 In addition, there are (8) Large Incurved Bowls with Rounded Lip and Impressed Strip or cordon round the vessel below the rim,18 (Io) Incurved Bowls with a Heavy Beaded Lip,19 and (1i ) Incurved Bowls with a Beaded Lip, sometimes slightly flattened, and an impressed strip.20 The latter example could be categorized as a neckless jar. Among the jars from Baba Jan are (I) Fine Jars, some examples of which are painted in a fashion similar to the painted examples from Kunji,21 and (4) Flared Neck Round Lip Jars.22 In addition, there are (2) Neckless Jars with Beaded Lips.23 A straight spout, presumably from a jar,24 and ring or disc bases25complete the illustrated assemblage.
I1 Roman Ghirshman, Fouilles de Sialk I (Paris, 1938), pl. 69:

S 1695. Cuyler Young, Jr., op. cit., fig. 7: 1. " 1x Clare Goff, Luristan Before the Iron Age ", Iran IX (x971), PP- 313-51. 14 Ibid., fig. 7: 1. 15 Ibid., fig. 7:25-7. 16 Ibid., fig. 7: 2-6. 17 Ibid., fig. 7: I1.
12 T.

Ibid., fig. 7: 17- 8. 19 Ibid., fig. 7: 12. 20 Ibid., fig. 7: 9-20. 21 Ibid., fig. 7: 14-16. 23. 22 Ibid., fig. 7:7-8. 28 Ibid., fig. 7:24. 24 Ibid., fig. 7:30. 25 Ibid., fig. 7:7-8.
18

EARLY

FOURTH

MILLENNIUM

DEVELOPMENTS

IN SOUTHWESTERN

IRAN

133

In summary, all Kunji vessel shapes are representedat Baba Jan. However, at Baba Jan a number of other vessel shapes are reported including some larger or heavier forms with supporting impressed strips. Most of these are made with a straw-temperedclay body similar to that from Kunji, but there are a few examples of other wares including heavy grit and mica tempered wares. This greater diversity is probably a result of the more diverse activities carried out on a village or town site as opposed to a small special function site like Kunji Cave. Goff26has defined the broad outlines of Terminal 'Ubaid and Early Uruk settlement patternsin the highland valleys of Luristan. Each larger valley contains one or more larger centres covering 3-10 hectares and most valleys contain one or more smaller village sites covering about I hectare. These sizes suggestthat valleyswould have had only a few thousand settled inhabitants. However, the existence of a site like Kunji may indicate the existenceof a fully transhumantpopulation even at this early date. Goff has suggested that there are geographical variants in ceramic style within Luristan at this time.27 If so, this suggests that the inhabitants of the different parts of Luristan interacted relatively infrequently, and that there was little overall political integration. Period VI at Godin Tepe provides some evidence of the development of this ceramic complex. Young28notes a decline of the finer painted ceramics during the course of this period. The ceramics from Kunji Cave at Baba Jan are most similar to those from Godin, operation B, Layers-2o and 21, late in Period VI. The parallels have been noted by Goff. However, most of the forms have antecedents in the earlier portion of Period VI, though these are often painted. Additional data from Godin should elucidate this development. In any event there does seem to be a highland antecedent for the ceramic complex found at Kunji and Baba Jan. The Transition from 'Ubaidto Urukon theDeh LuranPlain This small lowland plain in the foothills of the Kabir Kuh, the last massive ridge of the Zagros, had a long and complex settlement history during the 'Ubaid period. Its closest affinities during the later stages of this history were with the larger Susiana plain of central Khuzistan.29 Deh Luran is about 90 km. south of Kuh-i Dasht, and I Io km. north-west of the Susiana plain. An unprecedented settlement density on the plain was reached in the Bayat phase equivalent to Late Susiana c on the Susiana plain. In the succeeding Farukh phase, equivalent to Susiana d, there was a slight drop in settlement density (Fig. 2). Ceramics perhaps equivalent to the succeeding Susa A have been found only on a portion of Tepe Musiyan and on a few sites to the east of this large centre. The formerly densely occupied west portion of the plain was abandoned.
N

I
'

Deh Luran
Survey

-J" ,,i!:i:::..

km.
* Terminal Suso A 'Ubaid Related Settlement Settlement

o
45

Site Number

Edge of Present Flood Plain

sites on theDeh LuranPlain. Fig. 2: Susa A andSargarab

26

27

Ibid., fig. 4, page I45. Ibid., fig. 4, page I43.

28 29

T. Cuyler Young, Jr., op. cit., p. 6, Table III. Clare Goff., op. cit., p. I45-

12

134

JOURNAL

OF PERSIAN

STUDIES

For the succeeding Terminal 'Ubaid period, intensive survey by Neely has revealed only two sites. One formersettlement, Sargarab,is located high on the piedmont slopes of the plain close to the mountain wall. This is one of the few known settlements on the piedmont proper before to Sasanian times. The other settlement, Chakali, is about 5 km. downstream from the first at the juncture of the plain and the piedmont. Sargarab, numbered DL 169 in Neely's 1969 survey, stretches for 190 m. along a narrow ridge defined by two dry gulleys immediately east of the present Ab-i Garm or Sargarab stream (Fig. 3). The ridge averages 35 m. in width and 15 m. in height above the present level of the stream. The site covers about 7 hectares. A short distance above the site, the Ab-i Garm issues from a series of o. sulphuroushot springs. Though the vapours are noxious and even native-born Deh Luranis dislike the taste of the water, the springs are among the few year-round water supplies on the plain and are still frequented by Luri transhumants. Two km. to the north is the mouth of a gorge which is a commonly used route up into the higher pastures of Dinar Kuh, Ab Danan, and Kabir Kuh. Approximately 3'5 km. to the north-west is the beginning of a Sasanian or Early Islamic roadway still used today. m. It covers about Chakali, numbered DL-19, is c. 16o x 250 m. with a maximum height of c. 3"5 4 - o hectares. It is about 500 m. east of the present bed of the Ab-i Garm at a point where the water has lost much of its noxious gas and where it can be easily used to irrigate the plain. There has been settlement at this location in almost every period since the aceramic period.30 The stone foundationsof Sargarabhave been revealed by erosion (Fig. 3). The sides of the piedmont spur on which it sits were terraced with rough stone walls. The accessiblenorth-east upper end of the spur was protected by a rough stone wall with a narrow stepped entrance flanked by standing rough slabs. Within this space were a successionof four or more open spaces bounded by stone-footed rooms. Additional rooms are situated on the terrace walls below these courts. In the erosion gulleys flanking the architecturallyterracedsite are traces of dams and canals. Those to the east of the site are probably contemporarywith the Terminal 'Ubaid occupation at Sargarab. One of the springsmay have flowed through the gulley when the site was occupied, and rain fed agriculture is also possible. The ceramic assemblage from this site is similar to those previously discussed from the highlands. Most of the vessels have a clay body tempered with straw and some crushed calcareous rock: a few special types have only straw, and a few large vessels have mostly rock. The interior surfacesand rim areas usually exhibit wheel-scoring, but the lower exteriorsof some forms have been scraped. Slipping is common but burnishing and painting are both rare. The firing is sometimes poorly controlled and partially oxidized dark or red cores are common while surfaces are usually very light brown or green. The bowl forms attested at Sargarab are as follows: (i) SimpleRound Lip Bowls (Fig. 7a). One is slipped. (2) SimpleFine Bowls(Fig. 7b). Some of thesehave lower body scraping,some are slipped,and some are lightly burnished.Note the distinctiverecurvedtaperedrims. (3) Simple Flat Lip Bowls (Fig 7c). This may be an extremevariant of the beaded rim bowlsnoted below. (4) SimpleFlat Lip Bowlwith Impressed Strip(Fig. 7e). This uniquevesselwas lightly temperedwith strawand sand,its surfacewas carefullysmoothed oxidizedin firing. In these and it was thoroughly
technical features it approaches Terminal Susa A vessels discussed subsequently. (5) Simple Beaded Lip Bowls (Fig. 7f). Many exhibit lower body scraping and some are slipped. Most beading is on the exterior of the rim, but some is oblique to the rim and might better be termed Bevelled Lip (Fig. 7g). (6) Simple Beaded Lip Bowls with Impressed Strips (Fig. 7h). (7) Incurved Round Lip Bowls. (8) Incurved Round Lip Bowls with Impressed Strips (Figs. 7i). (9) Small Incurved Beaded Lip Bowls (Fig. 7d). This small vessel is slipped. (io) Incurved Beaded Lip Bowl (Fig. 7j).
so Hole, Flannery, Neely, and Helbaek, op. cit.

SARGARAB
//m

DL 169 , --

io~~
ml+++l

_\O ..
qb/

tp.~
.
?'
+

000
?9

0"

?Oo.

.. . o S~o?460

"?' "? ? +?'"" a a + 10 00 00o

0 o ................ OGO o oj OG 0 0 .................... qqOCP QQ dDd 00 Yo P0 a O .' a 0 +. ro,0ai%'00r 0 o.~O 10.019

. .....0
v00'~2

cup

00e,, 0040 0a 0 ca a a a
CA 0coo* 4* ..........

9000 a 0 oootb o

ID 0D~b 100 p 000R c 00O q 0 bc q

0 01-1 0

Fig. 3: The site of Sargarab(DL I69).

136 (1i )

JOURNAL

OF PERSIAN

STUDIES

Incurved Beaded Lip Bowls with Impressed Strips (Fig. 8a). These are very large and exhibit body scraping. The impressions on the strips tend to be circular rather than oval, possibly an earlier characteristic.1' The following jar forms are attested at Sargarab. (i) Fine Jars (Fig. 8b). These are smoothed but not burnished. One example has simple painted bands on the rim (Fig. 8c). (2) Neckless Jar with Beaded Lip (Fig. 8e). Some approach the neckless ledge rim forms common in the subsequent Uruk manifestations in Central Khuzistan.32 The illustrated example has painted black bands on a red slip. (3) Straight Neck Jar (Fig. 8d). All have rounded lips except for one which is slightly flattened. One jar has an attached oval nose lug. (4) Flared Neck Jar with Rounded Lip (Fig. 8f, g). (5) Flared Neck Jar with Expanded Rim and Convex Lip (Fig. 8i, j). This common form is often slipped. It occurs in small, medium, and large varieties. Some large examples are tempered predominantly with calcareous material. Two medium sized examples exhibit a burnished red slip with painted black bands (Fig. 8h) and two, a cream slip with similar bands. (6) Flared Neck Jar with Expanded Rim and Concave Lip. Other jar parts include additional oval nose lugs and a fragment of a straight spout. Specialized forms of interest include a shallow tray with slightly incurved sides and inwardly bevelled lip (Fig. 81) and fragments of perforated strainers. The ceramic assemblage as a whole has few resemblances to the earlier Susiana assemblages from the Deh Luran plain. Almost every feature is paralleled in the central Luristan ceramics previously discussed. The distribution of ceramic types in samples from three portions of the site of Sargarab are given in the following table. CERAMICS SimpleBowls Round Lip Bowls (i)
(2)

FROM SARGARAB South-west

2 I6

Fine Bowls

Flat Lip Bowls Same with Impressed Strip (4) Beaded (5) Lip Bowls (5a) Bevelled Lip Bowls (6) Beaded Lip Bowl with Impressed Strip Incurved Bowls (3)
(7)

5 I
24

Centre 2
-

North-east -

Total
2

3
2
-

I9
9 I 67 13 3
I

13 2
I

io

I
-

33 -

(8) Same with Impressed Strip (Io) Incurved Beaded Lip Bowl (11) Same with Impressed Strip (i)
(2)

Incurved Round Lip Bowl

2 3 7
I

I I
I

2 8 4 2 3
22
-

3 5 16 16 3 4 57 I 6 9
I

Jars
(3) (5)
(6)

Fine Jars Neckless Jars with Beaded Lip Straight Neck Jars Flared Neck, Expanded Rim, with Convex Lip
Same with Concave Lip

I
25 I

IO

Nose Lugs Trays

Strainers

2 5
II

--

4 4

31

Frank Hole, ed., Preliminary Reportsof the Rice University Project in Iran I968-9 (Houston, 1969).

3s

G. A. Johnson, op. cit., pp. 54, 181-2, pl. III L.

EARLY

FOURTH

MILLENNIUM

DEVELOPMENTS

IN SOUTHWESTERN

IRAN

137

The only evidence that Sargarab was occupied after the Terminal 'Ubaid period is the presence of four fine sand-tempered Ledge Rim Jar rims in the north-east area of the site (Fig. 8k), which could be of Early Uruk manufacture. If so, they represent only a brief utilization of the site, as other diagnostic Uruk forms are not present. One other artifact from Sargarab requires comment. From a terrace near the north-east corner of the site came a stamp seal made of a granular green stone (Fig. 4). The back of this rectangular seal is stepped up to a central ridge. The front portrays two dogs, recognizable from their curved tails, apparently pursuing two caprids, probably wild goats, recognizable by their curved horns. This fairly complex design has parallels in Gawra X133and Susa.34

Fig. 4. Sealfrom Sargarab.

In the succeeding Early Uruk period proper on the Deh Luran plain ceramicscontinue in the tradition establishedat Sargarab, but new featuresare added. Straw and limestone tempered clay bodies are predominant, but sand tempered vessels occur. Some of the former bowl and jar forms continue, but some become more common such as the incurved bowl form and the jar with flared neck, expanded rim, and concave or grooved lip. Among new forms are the flat or bevelled lip bowls, often carinated, small globularjars with strap handles, ledge rim jars, and bevel rim bowls. Stylistic elements such as impressedstrips and red on black painted motifs continue. However, among the former there is a shift from predominantly round impressions to predominantly oval oblique impressions. This ceramic complex has been recovered in the excavations at Tepe Farukhabad from levels below those containing typical Middle and Late Uruk assemblages, and its characteristics will be discussed at length in the forthcoming final report on the excavations at Tepe Farukhabad. In this early Uruk period, settlement pattern changed markedly. Sargarab was abandoned. A small town of about 5 hectares grew up along the Mehmeh River on the west edge of the plain. Five small villages were scattered across the west half of the plain. Tepe Musiyan itself and the east half of the plain were apparently unoccupied, the reverseof the situation at the end of the Late 'Ubaid period. In summary, communities producing Susiana ceramics declined or moved eastward. In the Terminal 'Ubaid period two communities using ceramics of highland affinity are established in a limited area of the plain: one is a well defended settlement near routesinto the mountains; the other is a village suited for small scale irrigation. These communities grow, divide, and relocate during the Early Uruk period, and settlement spreads throughout the west half of the Deh Luran plain. The Transition Plain from 'Ubaidto Urukon theSusiana The Susiana plain (Fig. 5) reached its greatest 'Ubaid settlement density in the Susiana d period, and the Susa A population was somewhat less. Susa itself with its central platform covered with large buildings and with an area no less than Io hectares was one of the largest centres in Greater Mesopotamia. The precipitous decline evidenced in the Terminal 'Ubaid period is at present completely unexplained.
"8 A. J. Tobler, Tepe Gawra II (Philadelphia, 1950), Nos. 158, 34 Pierre Amiet,

La Glyptique Mesopotamienne(Paris, 1961),

I63.

pl. 5, Nr. o9g.

138

JOURNAL

OF PERSIAN

STUDIES

;i.

J/; t r ...

fi': Khuzistan, Iran ..usiana Survey 1970-71

i
/,

,,
\

67t ' . I , .

...
!.

/
412 ,

'

:':!. , .

ti,. ;~i,

iiiiii.
1"2

'OSUSA108 11 S

6f

;;:;::: ? , . ...

ix.

i
2691

KEY Highlandsr s SiteNumber: 1: SmellCenter Village PresentFlood Edgeof

94 -.-.. ...."

Pi,

Susa A sites on theSusianaPlain. Fig. 5. Terminal

We present as an example of what we call " Terminal Susa A " the ceramics collected by Johnson and Wright in 1972 from KS-269, a small unnamed site 13'5 km. south-east of Susa, first reported to us by Frank Hole in 1969. The clay body of these ceramics with the few exceptions noted below is untempered or tempered only with a fine sand. The vessels were wheel-made, judging from the internal structure of the clay body, but most surface traces of wheel-turning have been eliminated. The firing is well controlled and the vessels are completely oxidized. There are two distinct types of ware, as in Susa A itself: a buff ware often greenish in colour and a red ware often with a light buff slip. The following bowl forms are attested at KS-269. (i) Simple Round Lip Bowls (Fig. 9a). The only example has incisions on its rim. (2) Simple Fine Bowls (Fig. 9b). These are somewhat thicker than those noted previously. The exteriors seem to have been shaved down while the clay was leather hard, leaving a fluted surface. One example is slipped. (3) Simple Flat Lip Bowls (Fig. 9c-e). These medium to large vessels, termed basins by Hole, Flannery, and Neely, are often oval in plan. Lower body scraping is common. The lip is often grooved. This form had been made with little change for centuries.35 (4) Simple Flat Lip Bowls with Impressed Strips (Fig. 9f-i). This stylistic variant of (3) is the hallmark of Terminal Susa A. It differs from the above only in that it is often slipped and that it has an applied strip or cordon with roughly circular impressionsbelow the rim. Note the resemblance to the example from Sargarab (Fig. 8e). A number of jar forms occur at KS-269. (i) Fine Jar (Fig. 9j). This is one of the few straw and limestone tempered vessels from the site. The few examples of this vessel shape found at other Terminal Susa A sites on the Susiana plain were similarly tempered. The surface was smoothed but not slipped or burnished as were highland examples. (3) Straight Neck Jar. The example has a rounded lip. The shoulder has bands and panels painted in a (4) Flared Neck Jar with Rounded Lip (Fig. gIk). Some of these well-made vessels are of a red-bodied ware, well known from Susa A sites.s3 (7) Heavily Flared Neck with Rounded Lip (Fig. 9n). (8) Heavily Flared Neck with Concave Lip (Fig. 90). These two new rim forms probably occur on large jars. They are at present known only from the Terminal Susa A phase on the Susiana plain.37
35 Hole, Flannery, Neely, and Helbaek, op. cit., p. 127. " " 3* G. Dollfus, Les Fouilles A Djaffarabad de 1969 1971 ", Cahiersde la Dilegation Archeologique Franfaise en Iran I Paris,
37

dark paint (Johnson, I973:

Pl. 8d).

1971). M. J. Steve and H. Gasche, " L'Acropolede Suse: Nouvelles Fouilles ", MDAI XLVI (Paris, 1971), P1. 34: 1o.

EARLY FOURTH MILLENNIUM DEVELOPMENTS

IN SOUTHWESTERN

IRAN

139

Other ceramics made with the Susiana clay body are a rough conical spout, also shaved like the Fine Bowls, an oval nose lug, and some animal figurine fragments. The following table presents rim counts. CERAMICS FROM KS-26938
(i)

(2) SimpleFine Bowls (3) Simple Flat Lip Bowls Strips (4) SimpleFlat Lip Bowlswith Impressed
(I)

Simple Round Lip Bowls

3
II

Fine Jars

13
I
I

(3) (4) (7) (8)

StraightNeckJar with Round Lip FlaredNeckJar with Round Lip Heavily FlaredNeck with Round Lip Heavily FlaredNeck with ConcaveLip

8 3 3

There are several items on other clay bodies. There are three straw tempered bevel rim bowl fragments, two ledge rim jars (Fig. 9m, p), and a small neckless flat lip jar (Fig. 9q), also straw tempered, and a sand tempered clay sickle fragment. All of these are probably later than the Terminal Susa A period. This ceramic assemblage contrasts strongly with those previously discussed. The recently reported excavations on the Acropole of Susa39indicate that it is a development out of Susa A. The diagnostic bowl shapes are rare in contemporaryhighland assemblages,though the method of decoration is somewhat similar. That such a development out of lowland tradition is not merely a local development is demonstrated by Caldwell's report of a similar assemblage of unpainted Susa A forms from Tall-i Ghazir on the Ram Hormuz plain 90 km. south-east of the Susiana plain.40 Johnson's 1970-71 survey of the Susiana plain revealed only twenty-two Terminal Susa A sites covering a total of 33 hectares, indicating about 6ooo people. In contrast, Susa A occurred on about forty-one sites covering about 59 hectares. These Terminal Susa A sites are divided into at least three discrete clusters (Fig. 5), two of which are dominated by small towns of about 5 hectares in area. The period was thus one of decentralization as well as one of declining settled population. The succeeding Early Uruk period is not well represented either by excavated samples or surface samples from single occupation sites. Hopefully this deficit will soon be eliminated by the work of the French Mission on the Acropole of Susa. In the interim we can note that the statistical studies undertaken by Johnson indicate the following to have been characteristicof the ceramic assemblage of the period. Sand is increasingly used as a tempering material and straw is used primarily for specialized shapes. The bevel rim bowl, primarily the tapered rim variety known as the " proto-bevel-rim bowl "
was commonly used. Neckless Ledge Rim Jars, Flared Neck Expanded Rim Jars with concave or

grooved lip, Flared Neck Jars with High Expanded Band Rim, and Jars with long Straight Spouts are all common. Population increasesgreatly during this period. There are fifty sites covering 97 hectares. Perhaps 20,000 people were settled on the plain. Susa emerged as a large centre once again, and there is a variety of evidence indicating that it is the centre of a small state.41 This state was probably far more powerful than any of the contemporarypolities of the highland valleys or the marginal lowland valleys.

Conclusion
lowland communities used pottery either untempered or with fine sand tempers. However, there are
a8These counts differ somewhat from those in Johnson 1972, Table 35, because we have subdivided" round rim bowls" into categories (2) and (3), and " flared rim jars " into categories(4) and (7)Ibid., Alain Le Brun,op. cit. J. R. Caldwell,op. cit. 41 Gregory Johnson,op. cit., p. 101. A.
39 40

Between c. 4000-3700 B.C. highland communities used pottery with a crushed rock or straw temper and

This paper has focussedon what we have called the Terminal 'Ubaid Period in south-westernIran.

140

JOURNAL

OF PERSIAN

STUDIES

certain broad similarities in ceramic complexes. First, the potter's wheel is in general use. Second, functionally similar series of large basins, bowls, and jars are used everywhere, perhaps as a result of similar domestic tasks and organizations. Third, similar stylistic embellishments, such as impressed strip applique and red slip, are widely used. Though each valley and plain seems to have a local pottery tradition, sites of the period can be easily recognized everywhere. The available data on site size and spacing indicate that the collapse of the lowland groups coupled with the growth of highland centres left lowland and highland enclaves closely matched in terms of population. A few communities with highland technologies appear on a marginal lowland plain. If these represent actual immigration, then the possibility of inter-regional conflict must be considered. Is there evidence for such phenomena elsewhere in Greater Mesopotamia? In southern Iraq there is little evidence of the Terminal 'Ubaid period of any kind. Though it was here that Woolley first the recognized a Terminal 'Ubaid assemblage in what he termed the Ur-'Ubaid III Graves,42 assemblage has not been widely recognized in either excavation or survey, perhaps because the period was one of low population densities. The best evidence appears to be that from a survey of the Nippur area by Robert McC. Adams. Here a channel of the Euphrateswas first occupied at the end of the 'Ubaid and was completely abandoned in the Late or Middle Uruk period.43 The decline of this settlement enclave might have resulted from a river channel shift. In northern Iraq little survey data of any sort is available. However, the excavations at Tepe Gawra reveal evidence of unsettled conditions during the transition from 'Ubaid to Uruk.44 Late 'Ubaid Gawra XII suffered from raiding in spite of its defensible position and protecting blockhouses. Succeeding Gawra XIA was dominated by a massive central redoubt. It is interesting to note that the ceramics of Gawra XII are of Northern 'Ubaid fabric, form, and design, with clear lowland affinities, while the ceramics of Gawra XIA are burnished brown and grey wares with few local antecedents. Though the technical and stylistic proto-typesof these wares have yet to be demonstrated,the situation is reminiscent of that on the Deh Luran plain. In the highlands there are two widely separated manifestationswhich may be Terminal 'Ubaid in age. To the south-west on the Marv Dasht of Central Fars, the Lapui Red Wares,45first recognized from Layer V of Tall-i Bakun A,48share many forms with the Terminal 'Ubaid Ceramics previously discussed, though the ware itself is distinct. However, the relative chronological placement of the Lapui communities is unknown, making both external comparisonsand the assessmentof local demographic trends difficult. To the far north-west, on the Upper Euphrates near Elazig in south-central and other yet unreportedsites show a Anatolia, the Late Chalcolithic ceramics from Fatmah-Kalecik47 to the Terminal 'Ubaid materials of Luristan. The clay preparation, firing, and surprisingsimilarity vessel forms are like those from Kunji Cave and Baba Jan. The relative chronological placement of these ceramics and the demographic trends during their period of use are unavailable at present. The distance between this area and Luristan,or even the Mosul Plain is sobering. However little is known of south-westernAnatolia, and future work may document a broad zone with similar ceramic techniques from the Central Zagros to the Anti-Taurus during the Terminal 'Ubaid Period. These broader problems aside, this brief survey suggeststhat the root of south-westIranian state and urban development lies in a period of contractinglowland settlement, leading to equally small highland
and lowland societies. There are widespread suggestions of conflict. To go beyond these simple propositions requires additional background work on the absolute chronology, on a better technical understanding of ceramics, and on surveys involving closer examination of the terrain to estimate population. Several small excavation programmes would be useful. All the Uruk phases in Luristan require better definition. Elsewhere the Terminal 'Ubaid and Early Uruk proper require better definition. Such
4 Sir Leonard Woolley, Ur: The Early Periods, U.E. IV (Philadelphia, 1956). 43 Robert McCormick Adams, personal communication. 44 A. J. Tobler, op. cit. "6William Sumner, Cultural Development the Kur River Basin in (Ph.D. thesis, Philadelphia, 1972).
46 Alexander Langsdorf and D. E. McCown Tall-i Bakun A, OIP LIV (Chicago, 1942).

4 Robert Whallon and Henry T. Wright, " I968 FatmahKalecik Excavations: Preliminary Report" in "1 968 Cahsmalan" Middle East Technical University Keban Project PublicationsI, Publication No. I (Ankara, 1970).

EARLY

FOURTH

MILLENNIUM

DEVELOPMENTS

IN SOUTHWESTERN

IRAN

141

local excavations would produce faunal remains useful in the elucidation of the impact of new transport animals such as the ass and the onager as well as of new herding techniques. Some means must be devised to recover the material remains of unsettled nomadic groups, who may have first begun to play an independent political role in this period. Only when some of this work has been undertaken can we answer the intriguing questions of social and political process.

142

JOURNAL

OF PERSIAN

STUDIES

CATALOGUE
from KunjiCave Fig. 6: Ceramics (a) Round lip bowl (L-I/6) :48 straw temper, paste 5 YR 6/6, burnished. (b) Round lip bowl (L-I/3): straw and limestone temper, paste 7.5 YR 6/4, burnished. (c) Fine jar (N-I/2): straw temper, paste 5 YR 7/6, burnished. (d) Fine jar (L-I/ -): straw and limestone temper, paste Io YR 7/3, burnished. (e) Fine jar (F-6/I): no visible inclusions, paste 5 Y 7/2, slight burnish. (f) Fine jar (H-5/I): no visible inclusions, paste Io YR 8/4, slight burnish, paint Io YR 3/2(g) Fine jar sherd (N-1/2): straw and limestone temper, paste 7.5 YR 6/4, burnished, paint 'o YR 4/x. (h) Fine jar sherd (L-5/5): straw and limestone temper, paste 7.5 YR 7/4 paint 2.5 YR 4/2. (i) Small incurved beaded lip bowl (L-I 6): straw temper, paste 7.5 YR 7/5, burnished. (j) Small incurved beaded lip bowl (L-I/4): straw temper, paste 5 YR 7/6, burnished. (k) Flared neck jar with round lip (N-I/6): straw temper, paste 7.5 YR 7/4. (1) Flared neck jar with round lip (L-I/6): straw and limestone temper, paste 7.5 YR 7/4(m) Small incurved bowl (L-I/4): straw temper, paste io YR 8/4(n) Simple beaded lip bowl (L-I/4): straw temper, io YR 5/4(o) A typical beaded lip bowl (L-I/6): straw temper, paste Io YR 6/4, burnished. (p) (q) (r) (s) (t) Flat base (N-I/2): Flat base (L-i/4): Ring base (L-I/4): Ring base (N-I/2): Pedestal ring base straw temper, paste 5 YR 5/3straw and limestone temper, paste 2.5 YR 6/4, burnished. straw and limestone temper, paste 1o YR 8/3, burnished. straw and limestone temper, paste 2.5 Y 8/3(L-I/3): straw and limestone temper, paste 5 YR 7/6 burnished.

from Sargarab Fig. 7: Ceramics Round lip bowl (191):49 straw and limestone temper, reduced core, paste 2.5 YR 6/2, surface 5 Y 7/3.5, (a) slip 2.5 Y 8/2. (b) Fine bowl (155): straw and limestone temper, reduced core, paste 2.5 YR 3/1, surface 5 Y 8/I. (c) Flat lip bowl (213): straw and limestone temper, paste 5 YR 6/5, surface 2.5 YR 7/4, slip Io YR 8/3(d) Incurved beaded lip bowl (99): straw and limestone temper, paste Io R 4.5/4. surface 2.5 YR 6/4, slip R xo 7/3. (e) Flat lip bowl (141): rare straw and sand inclusions, paste 2.5 Y 7/5, surface Io YR 7/4(f) Beaded lip bowl (92): straw and limestone temper, reduced core, paste 2.5 YR 6/2, surface 5 YR 6/5, burnished white slip 2.5 YR 8/4. (g) Bevelled lip bowl (214): straw and limestone temper, reduced core, paste .5 YR 6/5, surface 5 YR 6/4, slip Io YR 7/3(h) Beaded lip bowl (398): straw and limestone temper, reduced core paste Io R 5/2, surface 2.5 YR 6/2, slip io R 7/2. (i) Incurved round lip bowl (454): straw and limestone temper, reduced core, paste 2.5 YR 6/4, surface Io R 6/3, white slip 2.5 Y 8/3(j) Incurved beaded lip bowl (I): straw and limestone temper, reduced core, paste 1o R 5/5, surface Io R 5/6 to 6/8.
4a The numbers in parentheses are marked on the sherds. The

letter number combination preceding the slash designates a grid square: the number following the slash designates a locus

in that square. 16The numberin parenthesis Neely's serialcataloguenumber. is

EARLY

FOURTH

MILLENNIUM

DEVELOPMENTS

IN SOUTHWESTERN

IRAN

143

Fig. 8: Ceramics from Sargarab (a) Incurved beaded lip bowl (124): straw and limestone temper, paste 5 Y 6/2, surface 5 Y 7-5/2. (b) Fine jar (473): straw and limestone temper, paste 5 Y 7/2, surface 5 Y 8/2 to 6/3(c) Fine jar (80): straw and limestone temper, paste 5 Y 6/2, surface" Y 8/2. 5 neck jar with round lip (381): straw and limestone temper, paste 5 YR 5/3, surface 5 YR 5/4(d) Straight (e) Neckless jar (394): limestone and straw temper, paste Io YR 7/3, surface Io YR 6/3, paint 2.5 YR 4/3, red slip. (f) Flared neck jar with rounded lip (356): straw and limestone temper, reduced core, paste 5 YR 4/I, surface 7.5 YR 6/3. (g) Flared neck jar with rounded lip (19): straw and limestone temper, paste 10 YR 7/3, surface 2.5 YR 7/6, slip Io YR 8/3. (h) Flared neck jar with expanded rim and convex lip (81): limestone and straw temper, paste Io YR 4/3, surface 5 YR 6/4, burnished red slip Io R 5/4, black paint. (i) Flared neck jar with same (389): straw and limestone temper, paste 5 YR 5/5, surface 2/5 YR 5/5, white slip 5 Y 8/I. (j) Flared neck jar with same (I6): straw and limestone temper, paste 2.5 YR 6/6, white slip 5 Y 8/4. (k) Ledge rim jar (366): sand temper, paste io YR 7/3, surface 2.5 Y 7/2. to 5/6, surface 2.5 YR 6/6, slip 5 YR 7/2 to 8/3. (1) Tray (353): straw temper, reduced core, paste 2.5 YR
3/I

Fig. 9: Ceramics from KS 269 (a) Round lip bowl: no visible inclusions, paste 7.5 YR 6.5/5, surface io YR 8/3(b) Fine bowl: rare limestone and sand inclusions, paste 7.5 YR 7/5, surface o YR 8/3, light slip. (c) Flat lip bowl: rare sand and limestone inclusions, paste Io YR 8/5, surface 2.5 Y 8/2. (d) Flat lip bowl: no visible inclusions, paste 2.5 Y 8/3, surface 2.5 Y 8/2, D. c. 40 cm. (e) Flat lip bowl: no visible inclusions, paste io YR 8/3, surface Io YR 8/4, D. c. 58 cm. (f) Flat lip bowl: no visible inclusions, paste 2.5 Y 7/3, surface io YR 7/3, vessel oval in plan, average D. c. 55 cm. (g) Flat lip bowl: no visible inclusions, paste io YR 7.5/4, surface 2.5 Y 8/3, light slip. D. c. 6o cm. (h) Flat lip bowl: no visible inclusions, paste 7.5 YR 7/5, surface Io YR 8/3, D. c. 38 cm. (i) Flat lip bowl: no visible inclusions, paste 7.5 YR 8/5, surface Io YR 8/4, D. c. 52 cm. (j) Fine jar: rare straw and limestone inclusions, paste Io YR 8/4, surface Io YR 8/3.5. (k) Flared neck jar with round lip: no visible inclusions, paste 2.5 YR 6/6, surface 2.5 Y 8/1, red slip. (1) Flared neck jar with round lip: no visible inclusions, paste 7.5 YR 7/4, surface 2.5 YR 8/2, possible slip. (m) Ledge rim jar: no visible inclusions, paste 5 YR 8/4, surface Io YR 8/4, light slip. (n) Heavily flared neck with round lip: no visible inclusions, paste 2.5 Y 7/3, surface 2.5 Y 7-5/3(o) Heavily flared neck with concave lip: no visible inclusions, paste 5 YR 7/6, surface Io YR 8/4, sand grains impressed into interior surface. (p) Neckless jar: straw temper, paste 7.5 YR 8/5, surface io YR 8/3. (q) Ledge rim jar: no visible inclusions: paste io YR 7/4, surface Io YR 8/4.

144

JOURNAL

OF

PERSIAN

STUDIES

Fig. 6. Ceramics from KunjiCave.

EARLY

FOURTH

MILLENNIUM

DEVELOPMENTS

IN SOUTHWESTERN

IRAN

145

Fig. 7. Ceramics from Sargarab.

146

JOURNAL

OF PERSIAN

STUDIES

from Sargarab. Fig. 8. Ceramics

EARLY

FOURTH

MILLENNIUM

DEVELOPMENTS

IN SOUTHWESTERN

IRAN

147

Fig. 9. Ceramics from KS 269.

EXCAVATIONS AT HAFTAVAN TEPE 1973:

Fourth PreliminaryReport By CharlesBurney

1973.1 The members of the expedition staff were Charles Burney (Director); Rachel Annand, Stuart Brown, Martin Charlesworth,Altan gilingiroglu, David Kennedy, Frank Kinder, John Manley, Mrs. Barbara Manley, Swatantra Pidara, John Sharples, Frances West and Christopher Wolsey (archaeological assistants,with Mr. Sharples and Miss Annand also responsiblefor object photography); Mr. Hasan Talai, of the University of Tehran and a native of Shahpur, who for three weeks from mid-August kindly helped the expedition by supervising the excavations in Area AA, before joining Professor Negahban's expedition at Sagzabad; Mr. Michael Upton (architect); Mr. Velson Horie (conservator); Miss Louise Hare and Miss Laura Mackenzie (drawing). Of the above, ten were from Manchester, with the universitiesof Edinburgh, Keele, London, Oxford, Philadelphia and Toronto also represented. The Director would like to thank his staff for their effortsin a season when suitable labour was hard to find and thinly extended; and Mr. Henry Hodges for his good offices in securing a successionof conservators from the Institute of Archaeology. The average number of workmen employed was about one hundred. Akhough new trenches were opened up (TT6, TT7 and TT9 and a small trial trench (TT8) far out on the north-west fringe of the mound), the main effort was in areas already opened up (CI, C2, C4, TI on the citadel and YI to the east) and in extensions to existing areas (Y2, X2, X3). Eight cultural periods have now been distinguished, with the evidence from the 1973 season confirming that of 1971, in that Haftavan VIB has every appearance of having been the most populous and wealthy period in the long history of the settlement. Remains attributable to this period have now been found in four separate parts of the site (Areas YI-Y2, XI-X3, ZX and TT6D), surely indicating that the town had expanded considerably both to east and to west since Period VII. Though the excavations at Dinkha Tepe were of limited extent, the burnt buildings of the same period at that site, contemporary with Hasanlu VI, reveal a phase of urban communities on a larger scale than ever before in north-west Iran.2 The recently discussed pottery from Dinkha Tepe has links with northern Mesopotamia in the form of Khabur Ware not yet identified at Haftavan Tepe, whose affinities lay more with TransCaucasia to the north.3 There is no necessity to repeat the general classification of the cultural periods at Haftavan Tepe given in the last report.4 Accordingly there follows a summary of the season's results, period by period.

The fourth season of excavations at Haftavan Tepe lasted from July 19 to September 19,

1 The expeditionwassponsored supported the University and by of Manchester, with generous grants from the Calouste GulbenkianFoundation,the Russell Trust, the BritishAcademy, the BritishInstituteof PersianStudies, the Ashmolean Museumand the BritishMuseum. The expeditionis indebted to H. E. the Ministerof Culture,Mr. MehrdadPahlbod,and to the Director-General the Archaeological of Service,Mr. A. Pourmand,for renewalof the permitto continueexcavations. Dr. FiruzBagherzadeh owed a debt of gratitudefor his help is and encouragementand for the organizationof the Second Annual Symposium of Archaeological Research in Iran.
18A

Mr. MahmoutKhordvaniassistedus with his long professional of experiencein the capacity of Representative the Archaeological Service. The village schoolat Haftavanwas once again made available to the expeditionthroughthe courtesyof the Directorof Educationin Shahpur. 2 See the writer's forthcomingcontribution to the Grahame on ClarkFestschrift, settlementpatternsin north-westIran. 3 Carol Hamlin, " The Early Second Millennium Ceramic of Assemblage DinkhaTepe ", IranXII (1974), pp. 125-53. 4 IranXI (I973), p. 155-

149

150

JOURNAL

OF PERSIAN

STUDIES

HaftavanVIII (Fig. I and P1. Ia) The 1973 season saw the excavation of the later phases of the greater part of a massive round house comparable with similar structuresfound at Yanik Tepe, near Tabriz, I 0okm. as the crow flies east of Haftavan Tepe,5 and also at sites over much of Trans-Caucasia."The discovery of a round house was not wholly unexpected, given the probability that excavations had in effect begun in a level preceding VII. Advantage was taken of the trench cut in 1934, along and into the south side of the citadel; and, as suggested by P1. Ia, this allowed of a restricted excavation in TT9, overlooked by the heavily eroded sides of the 1934 trench. The disadvantagesof such an excavation are, of course, the difficultyof removal of spoil and the virtual impossibility of any extension. The only direction in which an extension might be possible, to the south, would certainly reveal a steep descent down the side of the mound, at a time m. when the settlement was probably more or less equivalent in area to the citadel mound. Only 60o to the south-west of TT9 lies Area X2 with its Period VIB building level, at an absolute level not very different from that of the round house. If Yanik Tepe is any guide, this round house endured through more than one phase; and, as at that site, there is evidence of burning. The wall face of the round house was easy to distinguish during excavation, with continuous heavily burnt debris among much fallen brick. Burning was also found on parts of the interior wall faces, including the partition wall. Along a west to south-west arc a reinforcement was found built along the exterior. Two structural phases of the circular wall were distinguished. On the south segment, where a doorway is tentatively indicated on the plan (Fig I), perhaps more plausibly than on the north segment, the main wall continues, but it is narrower and surviving only to a height one metre below the segment to the left of the doorway (?). The transverse,or partition, wall and the west half of the round house have survived well, with good mud plastered wall faces. The dating of this round house in TT9 to the very end of the sequence of twelve levels containing such houses at Yanik Tepe, and thus to the third quarter of the third millennium B.C.or even later, is confirmed not merely by the scale of the house but also by the pottery, with the absence of incised decoration and the occurrence of the distinctive graphite burnished ware.7 HaftavanVII (Fig. 2 and Pls. Ic-d, IIa) Since this was a peaceful period in the history of Haftavan Tepe and perhaps of a wide region of north-west Iran, this level (CI/9 and C4/9) can be subdivided into several phases not necessarily representedin each self-contained room, making any indisputable reconstructionof the comparative stratigraphy within the total duration of this building period impossible. The narrow street or alley running from north to south along the west side of the building in Area C4 (henceforthtermed Building A) is significant in showing the means of access to that building; but in effect it separates Building A from Buildings B and C. The absence of any access to the latter from this street makes any precise correlation between the buildings in these two areas impracticable. The absolute levels of the floors in Rooms B, BI, B2 and A5 suggest that they are all more or less contemporarywith the surface of the streetindicated in the plan (Fig. 2), although the passagewayis more easily comprehensibleas providing access, by steps, to the later phase of Room A5, as excavated in 1971.8 Building C seems to have stood at a slightly lower level, in conformity with the general downward slope from north to south. Every indication suggests that the highest point of the citadel in Period VII was very much as it was in the
present century. The question of access lies at the heart of the problems of interpretation posed by these buildings which, in contrast with Haftavan VIB, have left such meagre traces of their contents. Though the use of ladders may seem too easy an explanation, unsupported by impressions in floors or on wall faces, much depends on the function of the excavated rooms. Their walls are certainly massive enough to have supported an upper storey, so that it is more than likely that the ground floor would have been devoted
5 See especially C. A. Burney, " Excavations at Yanik Tepe, North-West Iran ", Iraq XXIII (1961), pp. 138-53. 6 C. A. Burney and D. M. Lang, The Peoplesof theHills (London, 1971), pp. 43-85. 7 Ibid., pp. 66-7 and Iraq XXIV (I962), p. XLIV, nos. 17-18. 1 Iran XI, p. 156 (fig. 2).

EXCAVATIONS

AT HAFTAVAN

TEPE

1973:

FOURTH

PRELIMINARY

REPORT

151

All- 566

? doorway

3 LEVEL
1 -8 3
oven,

5 LEVEL
A 11-456

limit of
,otion

eq
c " *

11.615
burning on wal facesh etpke

crkC

5 LEVEL

? d6orwa

A10.455

4 LEVEL

TEPE HAFTAVAN
AREA TT9 0 o,5 1 2.

1973
LEVELS L53,4,5 3

house(HaftavanVIII). Fig. z. TT9: round

07-

1.615

S G AY PA SA W .

oS I

5 ROOM

oRoom

C koo l . A -569 9

"19.1

WAFTAVAN AREA CI,C4

LEVEL 9
z

TEPE
mudbrick ier-

1973 trou,

0
o

Fig. 2. Ci andC4: Level9 (Haftavan VII).

EXCAVATIONS

AT HAFTAVAN

TEPE

1973:

FOURTH

PRELIMINARY

REPORT

153

wholly or largely to store-rooms. This theory is supported by the presence of a pithos (J4) set into the thin partition wall between Rooms B and B2 (P1. Id). Neither this partition nor the wall closing off the small compartment (" Room " BI) from Room B are bonded into the main load-bearing walls to north and south. It is indeed hard to see what other purposeRoom B can have served than for storage. The floor shown in the plan (Fig. 2) is that contemporarywith the two transversewalls. It is probable that the first floor level associated with the two majorwalls has not yet been reached, a symptom of the undoubted fact that the earliest phase of Haftavan VII has hardly yet been exposed. The great depth of the floor of Room 2 in Area C4, I m. below that of Room 5, was already in 1971 an indication that the excavation of all phases of Period VII was going to be an onerous undertaking: at that time it was found that Room 2 had a doorway in its south wall, into a room lying beyond the present limit of excavations.9 There was, however, no floor in Room 2 at a level close to that of Room 5, and it is more than possible that Room 2 belongs to a building on a lower level. It is this possibility of terracingdown towards the south which makes any southward extension of the excavations from Areas CI and C4 an unattractive proposition. There was evidently only a very slight terracing north-eastwards,down to Area R. Building C is representedwithin the limits of the excavations only by the successivephases of Room C, Room CI to the west apparently being outside this building: the great depth of the floor reached there is yet further evidence of the work which would be involved in any southward extension of these excavations. The rectangular pier in Room C has been uncovered down through three phases, of bricks, mud and bricks respectively, each with its own floor (P1.IIa). The earliestphase is associated with brick paving, or alternatively the top of an earlier wall, to the north-east (Fig. 2). The third and latest phase is capped by irregular flat stones probably serving as the base for a wooden column, supporting either the roof or a staircase: similar stones may have been removed from the stump of each preceding phase when the floor was raised to a higher level. The pier obviously served a similar function throughout these phases. In the third phase there was a trough nearly three metres long against the east wall of Room C, suggesting that by this time the ground floor of Room C may have been a byre or stable; but the proximity of the brick pier would have left little room for animals other than sheep or goats. With the use of the ground floor of Building B for storage this is yet further evidence that the living quarters were in an upper storey. Moreover, the pottery comprised sherds largely of storage vessels. Area TT6, at the west end of the mound, was opened, like TT9 after it, within the confines of a recent cut, this one being very deep and steep-sided, with narrow access from the north-west for the carts used to remove the soil. It was therefore a stroke of good fortune that here was found a wall of mud brick extending north into TT6B and standing about m. high at its lower end. At its upper 3"30 end it is preservedless than I m. high; and here another wall runs north-east. There is every reason to indicate that the large wall was part of the perimeter of the Haftavan VII town, whose preservation may be explicable by its being a salient or tower projecting out and down the slope of the mound, here very steep (P1. Ib). The foot of the wall at the north corner of TT6 is clearly, from the permanent dampness of the fill, not far above the present water table. HaftavanVIB (Figs. 3-6, and Pls. IIb-IIIa, IIId-IVc, IVe-i)
The excavations at the west end of the site, in TT6D, revealed evidence refuting the writer's previous theory of continuity from Haftavan VII to VIB, for here was found a layer of silt overlying the " town wall " of Haftavan VII, and above this was a building level whose pottery made it clearly of Haftavan

VIB. The building may have been of some size, from the presence of a store room with pithoi; but
there was no trace of a town wall corresponding to that of the preceding period. To the lacuna thus implied belong the " painted orange ware " and herring-boneincised ware found in 1971 in Area C4, immediately above the Haftavan VII building level. There seems perhaps little advantage in suggesting an absolute chronology, in view of adjustments to radio-carbon determinations, giving a correction
* Ibid. and Pl. Id.

bot om

of

ad-ch

Aff3l?
Pwey 0'7 eve A. .9

L (Asc 4F2) FLDOK Y214A 11om C-owh LEVEL Qer (-)9< uhole LLEVEL*AOa LEVEL
t

SY21Y bricK Vhelf

r FLOM LEVEL 44
LO R

She
5

J?
r,

owni ,' -..

f-sction ofFloor kriecl phose LUrdrin-situ LEVEwall FLoo (KLEVL2) Y214A t) 1 ScP(0 ~~~~~uern eo okr i er A9758 t . LLE 00 LEVELEV5cofee a Y2LA in-stulue d At, I suem Z 07511 dw pit LIELLEVEL WON cxc /Iwrfr IpiciatI t300was t- ---- SBooFL_ e level4) ,halg 4~4LNikling
n
?

Y2)4 F TjPE ALOORV

os p, th*
o

50 LEVEL LO

it

093wi

---Macke

,p~~pHA

of Y/5k LvELTzIARno Y/Y 12 A ja ?mc n-A thcis t A471 hearth 50 LEVEL RLOOK IVA" of fine fact or"DFe EE
1973~aO
TAVA T

-ll~t

~f

h zw

-- -- --- --

Y21Y

Fig. 3. Tz and r2: plan of rT/5, r214 etc. (Haftavan VIB).

EXCAVATIONS

AT HAFTAVAN

TEPE

1973:

FOURTH

PRELIMINARY

REPORT

155

factor of 250 years to be applied to this period.10 If this correction factor is applied consistently, in accordance with the available data, it would put Haftavan VII back to c. 2350 B.C. The principal investigation into Haftavan VIB was in Area Y (renamed YI) and its extension south westward in 1973 (Y2). With the additional evidence from the south side of the site (X-X3) and the north-east side (ZX) it has become apparent that there were two main building periods within the time-span of Haftavan VIB. The earlier one (YI/5 and Y2/4), with better preservedmud brick walls, some quite massive, was the chief objective of the 1973 excavations in YI and Y2: the state of the excavations in this period of the town's history is indicated in the plan (Fig. 3), whose complexities emphasize the problems of distinguishing successive phases of terracing and of the buildings on those terraces. In ZX remainsof a wall preservedabout 2 *50 m. high, with another one partly contemporary, confirm the extension of the VIB town that modest distance from YI-Y2: there is no mistaking the similarity of the buildings. There are problems concerning the extent and preservation of the Haftavan VIB town, whose summit may have been up against the south side of the citadel of Haftavan VII. The uppermost two levels in Y2 comprised mainly silty material washed down from the top of the site, and suggesting abandonment of that part of the settlement at the end of Haftavan VIB or else the complete destruction of overlying strata of Haftavan VIA. One aspect of the extent of the built-up area of the VIB mound will be a principal objective of next season's excavations: this is the relationship between the deep level reached in ZX (ZX/9), contemporary with YI/5 and Y2/4, and the earlier building remains, of Haftavan VII, excavated in 1969 in Area R, only about 60 m. distant and yet Io m. higher in absolute
level."

The massive walls uncovered in 1971 at the western extremity of YI (YI/5B and 5D) gave every promise of belonging to a major building, almost as well preserved as the buildings of Haftavan VII and much richer in contents. This was the primejustification for the decision to extend the excavations westward into Y2 at the opening of the 1973 season. In the event these expectations seem to have been unfulfilled, although continuation down to the earliestphase ofYI/5 and Y2/4 may yet prove a rewarding, if laborious, undertaking in terms of architecture. Three clear phases were distinguished in Y2/4, most of poorly constructed brick and mudplaster walls, closely related to YI/5B and YI/5D. The excavations in YI will therefore be summarized first. The excavations in YI and Y2 have revealed a series of terraces, whose functions changed in successive phases, and which were subject to frequent alterations and repairs, as terraces are particularly liable to be. The sequence of phases distinguishedin YI and Y2 may thereforenot entirely apply to other parts of the town. Within this earlier building period of Haftavan VIB four phases have been detected, though the earliest has hardly been exposed at all, and awaits future excavation. The first phase is indicated by the evidence that the " Phase I " walls originally continued uninterrupted right across the width of YI and Y2, where the two areas meet (Fig. 3). The great wall with a marked batter running parallel with this across the north-east end of YI and associated with the Level 7 floor to the east could belong in the first instance to this first phase but seems, on present evidence, more probably part of the second phase. Through this east wall there seems to have been a narrow gateway, subsequently blocked in haste with a loose filling of flat stones and carelessly laid brick-work above. The second phase includes the large, well preserved " Phase i " walls and their contemporarysurface
(YI/5D), together with a more modest wall about i *25 m. east of the north " Phase I " wall and continuing south beneath the later " Phase 2 " wall, thus forming the other side of a narrow passage. Anyone using this passage would have to mount one step, where a post-hole on the left (east) suggests a door, and advance about 4 m. farther south, before turning right (west) up three steps and through a door: the impression of a door-jamb on the left is matched by a quern on the right, which could have served as a pivot stone. This door gave entry to a quadrilateral courtyard (Y2/4B), contemporary with and continuing the surface of YI/5D. It is unfortunate that pits of YI/4 have destroyed some of the

10

of For a discussion the applicationof MASCAdates to Dinkha Tepe and Hasanlu, CarolHamlin, IranXII, pp. I29-31.

11

IranX (1972), p. 129 and P1.I (Haftavanreport).

:I

,D

Soven

.
LEV-

2 hse wall

ViI

stike holes

-T2: Fig. 4.

Level 3 (later Haftavan VIB).

AREAYL. WI

AAFTAVAN TEPE
o. 5
I

1973
LEVEL 3

Pl. Ia. TT9: theround house,lookingnorth.

Pl.Ib. TT6: the " townwall ", lookingnorth. N

Pl. Ic. C1g9, looking east towards C4/9.

Pl. Id. C1/9, looking no

and P1. IIb. TI/5 (inforeground) Y2/4: generalview

Pl. Ila. C1/9, RoomC, with brick pier and trough; RoomCI inforeground.

with P1. IIc. r2/4 (west): courtyard stonestepsto higherterrace.

Pl. IId. Y2/4: generalview lookingeast-

Pl. IIIa. r2/3: view lookingeast, withfurnaceinforeground.

wall and drain,looking south-south-east. P1. IIIb. T : Urartiancasemate

Pl. IlIc. C2/4: the bricksocle,lookingnorth-east.

ware vehicle. a P1. IIId. X2/2: polychrome sherdrepresentingtwo-horsed

15'5X 13o0 cm.

Pl. IVa. Unbaked numeration tablet. clay bowl D. 28 4 X P1. IVb. Spouted
21.5

cm.

Pl. IVc. Stonemould a for

Pl. IVd. Sasanian(?) button stampseal. 8x I. o.9 cm.

Pl. IVe. Hole-mouth bowl. 9g6x

8 cm. io.

P1. IVf

Pl. IVg. Polvchrome jar.

12.

ox o205 cm.

Pl. IVh. Polychrome jar.

iz.

6x 25-5 cm.

P1. IVi. Polychro

EXCAVATIONS

AT HAFTAVAN

TEPE

1973:

FOURTH

PRELIMINARY

REPORT

157

evidence in the vicinity. The large " Phase I " wall north of this entrance is distinguished by a structural technique apparently peculiar to Haftavan VIB, the use of beams of modest cross-sectionboth set horizontally into the foot of the wall just below the level of the floor and higher up and transversely through the wall (Pl. IIb). In two places in Y2, as shown on Fig. 3, beam impressionsindicate a rather greater thicknessof wall than the poorly preserved brickworkmight by itself suggest. The third phase is representedby a narrow brick wall, probably a rebuilding of the massive wall beneath, though hardly directly associated with it, and a good clay floor running up to this wall (YI/6) and evidently laid on top of the remains of this much larger wall. With it is the base of a pithos, reused as a bowl. This third phase has a surfaceover the stump of the east-west stretch of" Phase I " wall and post-dates the deeper floor level of YI/5D; but it continues beneath the " Phase 2 " wall. The fourth and final phase, with which the two phases of Y2/4A are at least approximately contemporary,is representedby the " Phase 2 " wall, whose relationship to the earlier" Phase I " wall acrosswhich it cuts was never in doubt; but in its turn it has two phases, the battered exteriorbrickworkforming an addition or reinforcementto the earlier wall behind it. Thus it could be said that there were as many as five phases of this whole early VIB building period. The buildings in Y2/4B may have been of a scale and solidity approaching those of YI/5D; but, if so, nothing of any size remains. A stone-built terrace revetment at the south corner of Y2 seems to have been built in the preceding phase. Nearly 4 m. in front of this was an addition to this terrace, with stone steps up from the courtyard area in the north-west of Y2/4. This addition was apparently made by cutting into fill accumulated against the stone-built revetment: the courtyard entered from Yi (Y2/4B) was contemporary with it. Whatever the precise relationship of the stone revetment to the " Phase I " walls of YI/5D, it marks the highest terrace, as YI/7 marks the lowest, and it suggests that this building level comes fairly close to the present surface of the mound in the unexcavated area south of Y2. It also shows a change in orientationof about 800 compared with the lower terraces. The narrow entry of the first (Y2/4B) phase was widened by demolition of a stretch of the east-west wall on the north side, opening up direct access from the courtyard northward into an open-fronted structure which in plan resemblesnothing so much as a shop. The layout and flimsy constructionof these walls of bricks set irregularlyin mud helped the formulation of the theory, independently arrived at by two members of the expedition, that this may have served as a bazaar quarter, naturally requiring the security provided by a narrow, indirect access from outside the enclosure. In the south corner of YI has been uncovered a small portion of a structure perhaps the major building in the whole vicinity of YI and Y2. Courtyardsand shanty shops occupied nearly all of Y2. In Area ZX a very limited sondage was all that was practicable in exposure of this earlier VIB building period. Here a wall at least 2 50 m. high was found (ZX/9), the floor not being reached. This may representtwo successivephases (ZX/Io and 9), since an adjacent wall was related to a higher surface. Enough has been found, however, to indicate a contemporary level probably at least as well preserved as in YI-Y2. Above this was a level (ZX/8) of a different architectural character, with a stone wall surviving two courses high and forming three sides of a rectangle, probably a courtyard. Here, and in a level overlying this, ZX/5B, was found a quantity of painted pottery (Pl. IVg) and the heavy burning likewise occurring in YI-Y2 and XI-X3. Some of this seems to indicate industrial activity, the rest being destruction debris. The town representedby YI/5 and Y2/4 eventually ended in general burning, clearly to be seen in
an ash layer over most of the area. In the burning was found one stone wall (Y2/3E) with associated burnt beams and roof rafters, this wall being parallel with the later east-west wall of Y2/3: an intermediate building level between Y2/4 and Y2/3 is thus indicated. The following level, Y2/3 (Fig. 4, P1. IIIa), resembles ZX/8 and XI-X3 in the character and preservation of its buildings. At first glance the plan of Y2/3 (Fig. 4) may suggest one simple layout; but in fact there is indisputable evidence of at least five phases. This does not, however, imply a long time-span for this building period, the later of Haftavan VIB, since only the second of the five sub-phases represents a building of any pretensions. The first phase comprises stones belonging to an earlier structure, indicated as " earlier phase " on Fig. 4, underlying the east and west walls and the south-west corner of the rectangular room. The second, and most substantial, phase comprises the west-east wall in the west sector of Y2/3, with its

LEVEL 2.
10i486

LEVEL
bench?

LEVEL 2 5 LEVEL

on mudbrick
50 9'9 stone foundations

o'

H 1973 TEPC I-AFTAVAN 2& LEVEL,3 3

AREA X2,X3
0 oS 1 z 5

Fig. 5. X2 and X3: Levels2 and 3 (laterHaftavan VIB).

EXCAVATIONS

AT HAFTAVAN

TEPE

1973:

FOURTH

PRELIMINARY

REPORT

159

return northwards, and also the earlier phase of the corresponding wall to the east. With this phase belong the oven or furnace near the west corner of Y2/3 and the other oven or furnace about 7 m. to the east. The western oven comprised three rounded compartments side by side with a stoke-hole in the central compartment, this being found filled with ash. The third phase is marked by a rebuilding of the east stretch of wall running into the east corner of Y2/3. The fourth phase comprises the flimsy wall built to join the two earlierwalls into one long east-west wall and also the parallelwall to the north, completing the rectangular room. The absence of burning inside the room is one indication that this fourth phase representsa crudely built structureon top of the burning of the previous buildings. Latest of all is the high-walled oven overlying part of the north-west corner of this room. The general impression of Y2/3 is not inconsistentwith an open-air industrial area, with a later building occupying part of the ruins. The burning found in ZX/8, with its courtyard, could have a similar significance. The excavations of the 1971 season in Area X to the south of the citadel were extended north (X2) and then west (X3). The architectural results can be summarized as the discovery of a building on a terrace in the north sector of X2-X3, overlooking a lower room or courtyard to the south (Level 3), its surface over I m. lower than the floor of the room to the north. Open areas on either side, to east and west, slope down the side of the mound. The brickwork was well preserved, through being heavily burnt, at the south-west corner of the terrace building and along the west wall of the lower room or courtyard: the bricks are almost square and thin, sample sizes being 31 X 31 x 9'5 cms. and 39 x 36 x 8-5 cms. Two coats of mud plaster were found in places. A ledge, probably a bench, runs along the foot of the north wall of the lower room. Among charred roofing timbers were hollow beams, evidently rotten before being burnt. Industrial activity, as in Y2/3 and ZX/8, is hinted at by the discovery of a stone mould for casting arrowheads (P1. IVc). Further extension of the excavations in X-X2-X3 would face similar problems to those arising from the work in TT9, both areas having been sited to take advantage of Butamin's trenches of I934Associated with the few stone footings preservedof a later phase (Y2/3A) overlying Y2/3 was a clay " tablet ", found in the east corner of Y2 (P1.IVa). It certainly is of a shape and size to be reckoned as a tablet; and its material, unbaked and friable clay, may partially explain the absence to date of other examples. The dots on its surface resemble simple numbering, either for accounting purposes or as a school exercise. No account, however brief, of Haftavan VIB would be complete without reference to the commonest category of finds, the pottery. The forms of vesselsinclude a variety of bowls, beakersand plates (Fig. 6). There are also jars of varying size and form, from coarse ware storagejars to crudely made miniature globular jars, as well as globular and straight-necked jars with painted decoration.12 Inevitably the smaller pots, particularly the beakers, are better represented among intact or restored vessels than the jars or larger bowls: in Y2 were found several groups, termed batches, of pottery, especially beakers, of which Y2/Batch 3 was the largest. This group, found immediately north-west of the entry from YI/5D into Y2/4B, is representedby Fig. 6, nos. 2, 6, 10, I I, 15, 18 and 20. Beakers are the most distinctive form of Haftavan VIB, surviving spasmodically in the following period, VIA.13 They may possibly derive from a cylindrical form found in the Early Trans-Caucasian II period at Yanik Tepe, plain or with incised decoration; and the form may also have been introduced from northwest Iran into the repertoire of Khabur ware represented at Dinkha Tepe.14 Most of the beakers of Haftavan VIB are undecorated, though dark brown or black painted patterns occur on some examples, one of the distinctive designs being a ladder pattern (P1. IVf). The basic shape of the beakersis cylindrical, with everted rim and base. Three types were distinguished, with some variations: rounded profile (Type i), carinated profile above the base (Type 2) and fiat based examples (Type 3), this last being the most numerous class, including ten among the list of registered finds for the season, not illustrated in Fig. 6. One fine beaker, so far unique, had a handle attached (Fig. 6, no. 16). Nine types of bowl were discovered in all, four of these including sub-types. The commonest were Types I, 2, 4a, 4b and 7. The hole-mouth bowl (fig. 6, no. 20 and P1.IVe) is peculiar to the earlierbuilding period
is IranXI, Pl. IVa (Haftavanreport). s8IranX, Pl. IId (Haftavanreport).
14 cf. Iran XII, p. 133 (fig. i, nos. 12-13).

160

JOURNAL

OF PERSIAN

STUDIES

7Ii:31
77116
7(

170

14

12

16

17l7
19 2,

18

21

CMS
Fig. 6. Pottery forms of Haftavan VIB: bowls,beakers, plates.

EXCAVATIONS

AT HAFTAVAN

TEPE

1973:

FOURTH

PRELIMINARY

REPORT

161

of Haftavan VIB, not being found in the later period, unlike the other painted pottery forms: this and the other hole-mouth form (Fig. 6, no. 8) are rather rare, all examples being of fine ware and nearly all decorated. The inverted rim bowl (Fig. 6, nos. 7 and I2 and P1. IVb), distinctive of Haftavan VIB frequently has a black painted band round the rim, with simple triangles. Plates were rare, four types being identified (Fig. 6, nos. 19, 21, 22, 23), mainly by profiles of rim and base: their typology is not yet wholly certain. Jars are less easily classifiable,most examples being coarse ware storagejars. One type is a miniature globularjar, of very coarse ware similar to the bowls of Types 7 and 9 and crudely made. Another type, with many variations in their proportions from squat to taller, is the globular jar, frequently of decorated fine ware (e.g. P1. IVg-h). Another type is the straight-neckedjar, also in fine ware and with painted decoration. The painted decoration is the most striking feature of the Haftavan VIB pottery. The simplest style is bichrome, with straight and zig-zag lines in dark brown to black paint, the ware being either finely or coarsely tempered; if the latter, there is often a slightly burnished red slip. Another popular design in bichrome ware is of pendant triangles from the rim. (cf. also P1. IVe). The more typical painted pottery is polychrome, classifiable into two main groups, bowls and jars. The bowls are of fine-temperedred ware, often slightly burnished: the decoration runs in a band round the upper body outside, between rim and carination or widest diameter; it comprisespanels in a white or cream slip, usually bordered in black, and divided into rectangles, diamonds or other geometric motifs by lines of black, red, brown or shades of orange; some of these compartmentsare then filled by painted hatching. This somewhat stereotyped decoration on bowls is familiar at Geoy Tepe D.15 The decoration on jars is on the whole much bolder, the ware being coarse, the colour being brown, buff or pink and the section quite thick; the surface is slipped in a light shade of pink, cream or yellow, slightly burnished. The polychrome decoration comprisesthick lines in brown to black. Lines of diamonds, elongated triangles and panels of hatching were clearly traditional, as was the use of the favourite colours of red, orange, brown and black. It is in this class of pottery that figurative representationsoccur, of which the most interesting so far found is the fragment depicting two horses drawing some kind of wheeled vehicle, with the end of the shaft between the two animals visible (P1. IIId). This sherd is here illustrated upside down, in order to show the better preserved of the two horses to best advantage. The scene is viewed from above, but the triangle of cross-hatched lines shows that it was the other way up on a painted jar. The remainder of the design is difficult to interpret. The birds suggest the herring-bone elements to be trees or bushes; but the gaping toothed jaw (?) could indicate a hail of arrows. The historical implications of this design are wide, since this is an early representation of a horse-drawn vehicle. The writer is inclined to see this as suggesting the presence of Kassites in this region in the Haftavan VIB period: the mean radio-carbon date of 1772 B.C. for this period is perfectly consistent with such an attribution, since the Kassites were invading and settling in the plains of Babylonia during the seventeenth century B.c., beforegaining control of the whole land."6 Highly stylized animals and men also appear on sherdsof similarpolychromeware, the greatestdevelopment of this style coming in the later of the two main periods of VIB distinguished to date. More typical are birds (P1. IVh), perhaps ultimately derived from the incised decoration of the wide ceramic repertoire of the Early Trans-CaucasianII period; but this derivation would be reinforced by discovery of incised pottery in Haftavan VIII or perhaps earlier levels.1 Possibly the appearance of birds in the Khabur ware of Dinkha Tepe is another indication of northern influence on the Ushnfi valley.s18
The Haftavan VIB pottery is either hand- or wheel-made, most of the hand-made vessels being of the undecorated coarse ware types. Plain wares include some red, slightly burnished and finely tempered. Grit is the temper normally used. The above brief summary has touched only on salient features and forms.'9
T. Burton-Brown, Excavations in Azerbaijan 1948 (London, 1951). 16 Thanks are due to Dr. Mahdavi for this radio-carbon date. The Kassites are also strong candidates as producers of the most exotic of the " Luristan bronzes ", as discussed in P. R. S. Moorey, " Towards a Chronology for the Luristan Bronzes ", Iran IX (i97i), pp. I i3-30, especially p. 117.
15 17 cf. Iran XI, P1. IIId (Haftavan report). 1s Carol Hamlin, Iran XII, pp. 144-5, figs. XIId, XIIf, XIIIs. 19 Acknowledgments are due in the first place to Mr. Charlesworth for preliminary study of the Haftavan VIB pottery, with further work subsequently carried out by Messrs. J. Parkhouse, N. Sekunda and G. Summers, of the University of Manchester.

162

JOURNAL

OF PERSIAN

STUDIES

HaftavanV In Area PQ on the west side of the mound the Iron I period was investigated early in the I973 season. Nothing sufficiently important was found underlying the major level (PQ/I), excavated in 1971,20 to justify continuing excavations. Here the Haftavan VIB levels are too deep to be exposed at all extensively. IV Haftavan An extension northwards(TT6D) of the new area on the west side of the mound led to the discovery of an interesting building level (TT6D/IC), of the Iron II period, hitherto all too little known at Haftavan Tepe apart from the burials found in the 1968 and 1969 seasons.21 Here were found fairly large stone wall footings, pottery of the familiar spouted and other Iron II forms and also crucibles, suggesting copper or bronze smelting. III Haftavan (P1. IIIb) In Area TI a limited sector of the Urartian citadel was uncovered, revealing a stretch of the perimeter wall of the whole complex. Any defensive wall would have been further out, and thus destroyed by recent damage to the site. With its surviving brickworkand footings of small stones set in mud this wall is the best preserved part of the Urartian citadel so far uncovered, standing over I -50 m. high. Elsewhere the Urartian level is rather poorly preserved, with hardly any brickwork surviving. The main line of the perimeter, from north-north-westto south-south-east,is in the form of a casemate wall, the floor level to the west, inside the citadel, being higher than outside it, to the east. Part of a stonecovered drain was discovered running between the two walls of the casemate (P1. IIIb). II Haftavan (Fig. 7, P1. IIIc) The fill of the brick tower was further emptied, down to about 9 m. from the top of the brickwork, through the shaft below, cut into the underlying strata of the citadel.22 But proper equipment will be required, before the work can be safely continued down to the water table, about 12 m. below. The mean radio-carbondate of 492 B.C. from a sample of the 197 1 season, kindly supplied by Dr. Mahdavi, satisfactorilysupports the previously proposed attribution to the reign of Darius I (522-486 B.C.).23 Excavations were resumed in Area C2, on the summit of the citadel, from the level reached in the 1968 season. The objective was to reach at least the uppermost building phase of Haftavan VII, thus extending the excavations in CI and C4; but the complexities of this area, disturbed by pits, made this impossible. The principal discovery was of an extensive brick structure in C2/4, well preserved and with stone foundationsalong its south side. No original face, however, could be defined along the north side of Area C2: here, and also to the east, there appears to have been a large cut down from above. A fairly consistent curving line resembled the edge of the pit more than any face. In the south-west part of Area C2 a thin ashy layer was found overlying brick paving related to a later addition to the brick structure. Beneath the paving and some underlying fallen brick was another thin ashy deposit over the original surface related to the brick structure. Although the stone foundations of the wide,
shallow projection or buttress on the south side were laid after those of the main part of the structure, this was contemporary, since the brickwork of the buttress was bonded into the main mass. The precise function of the brick structure remains enigmatic; but it seems to have served as a socle or foundation platform for a structure above it, now completely destroyed. The dating of this level is still far from certain.24

21

The best examples were found in 1969 (Iran X, figs. 8 (p. 136) and 9 (p. 137)). 2s Iran XI, p. i68 (fig. 9).

so Iran XI, pp. 162-4.

23 Ibid., p. 169. 24Acknowledgment is due to the patience and skill of Mr. Stuart Brown in the supervision of the excavations in Area C2, which were among the most difficult undertaken at Haftavan Tepe.

EXCAVATIONS

AT HAFTAVAN

TEPE

1973:

FOURTH

PRELIMINARY

REPORT

163

Removal of the brick socle exposed a burnt surface without building remains. Beneath were uncovered mud brick walls of two adjacent rectangular buildings (C2/5), almost certainly the first building period after the abandonment of Haftavan VII. I Haftavan (P1. IVd) In Area AA was found a deposit of two jars, a spouted bowl and a water jug, well fired and slightly grit-tempered. This group of vessels was too close to the surfaceof the mound for the context to provide certain dating evidence. They are either Sasanian or possibly later. The stamp seal (P1. IVd) is similarly from unreliable context.
mud4brick A

limitof excavation

26

S2422518

El/2.

4LEVEL

LF1E

TLP 1-4AFTAVAN

197

Fig. 7. C2: Level4.

164

JOURNAL

OF PERSIAN

STUDIES

APPENDIX

VIB 6) ofPeriod (fig. Pottery

Bowl, Type 9B. Y2/4, Batch 8. Medium coarse lightly grit-temperedbrown ware. Undecorated. Bowl, Type 9A. Y2/4, Batch 3. Medium coarse ware very lightly grit-tempered. Orange-red inside, red and black outside. Ring base. 3. Bowl, Type 7. YI/5. Very coarse ware, thumb made. Undecorated. 4. Bowl, Type 5A. Y2/4, Batch I. Coarse ware. Undecorated. 5. Bowl, Type I. X2/3B. Fine ware. Decoration: occasional painted black band below rim. 6. Bowl, Type 5B. Y2/4, Batch 3. Medium coarse ware. 7. Bowl, Type 4A. X2/2B. Fine ware. Decoration: rim painted in black triangles. 8. Bowl, Type 6B. Y2/4, Batch I. Coarse ware. Undecorated. 9. Bowl, Type 3. YI/5. Fine ware. 10. Bowl, Type 8. Y2/4, Batch 3. Medium ware. Undecorated. I I. Beaker, Type 3A. Y2/4, Batch 3. Fine to medium coarse pink ware, drab outside. Flat rectangular base profile. Traces of red-brownslip on some exx. 12. Bowl, Type 4B. X2/2B. Medium coarse ware. Inverted rim with spout. Decoration: dark brown painted rim. See P1. IVb. 13. Beaker,Type I. Y2/4D. Medium coarse buff-black ware tempered with sand and grit. Undecorated. 14. Beaker, Type 3B. Y2/4, Batch 6. Fine to medium coarse ware. Thin red slip over buff body. Undecorated. Type 2A. Y2/4. Medium coarse very gritty pink ware. Carinated profile. Undecorated; I5. Beaker, but other exx. occasionally decorated with painted dark brown vertical lines, wavy lines, pendant triangles. Also of this type: lightly grit-temperedorange ware, thin red slip, black-paintedladder decoration. 16. Beaker, Type 2B. Y2/4, Batch I. Fine black ware slightly grit-tempered. Carinated profile with handle. Undecorated. Unique to date. Carinated profile. Decora17. Bowl, Type 2. YI/5. Fine light brown ware slightly grit-tempered. tion: painted black swags. Thin red slip, burnishedin and out. In other exx. bands etc., between rim and carination are very common. i8. Jar, Type 5A. Y2/4, Batch 3. Very coarse grit-tempered red and buff ware. i9. Plate, Type 2. YI/6. Medium coarse ware. Undecorated. 20o. Bowl, Type 6A. Y2/4, Batch 3. Fine ware, buff and light red, burnished outside. Decoration (dark brown on cream): line below rim and pendant triangles, cross-hatched. Other exx. in plain ware. See P1. IVe. I. 2.
21.

22. 23.

Plate, Type 3. Y2/4, Batch 3. Medium coarse green-red ware, grit-tempered. Undecorated. Plate, Type I. Y2/3. Medium coarse red ware. Ring base. Undecorated. Plate, Type 4. Y2/4, Batch 7. Medium coarse ware. Bevelled rim. Undecorated.

Provenance objectsin P1. IV. of IVa, Y2/3a: IVb, X2/2B (see fig. 6, no. 12): IVc, X3/I: IVd, AA/I: IVe, Y2/4, Batch 3 (see fig. 6, no. 20): IVf, Y2/4, Batch 3: IVg, 2X/8: IVh, X2/3A: IVi, X3/I.

SHORTER NOTICES AN INSCRIBED IRANIAN BRONZE VESSEL By W. G. Lambert

This vessel (P1. I) turned up recently on the London fine art market and is published by kind permission of its owner. It is of simple, hemispherical shape, 8-5 cm. in diameter, 4 6 cm. deep, having been beaten out of a sheet of bronze. While other similar vessels have been found in excavations,' the very simplicity of the present piece cautions against trying to assign a specific date or area within Iran for its manufacture. Items so lacking in style could have been made in many centres over many centuries. The distinctive and significant feature of this vessel is a three-line cuneiform inscription engraved downwards on the side (P1. I). While the signs are well cut, and save for some damaged ones in the first line, easy to read, the inscription is in reverse. It reads:
ba -x-x-x-LUM
mdr diamal-ga-mil

warad dnin-lubur

Ba... . /hum son of Samas-gdmil slave of Nin'ubur

quarter of the Third Millennium B.c. and those covering a few centuries around 1000 B.c. This inscription fills a little of the gap. That it is a seal inscription in reverse is very rare, but not unique. One other ancient Iranian bronze vessel, published by P. R. S. Moorey and the present writer in this journal, volume X (1972), pp. 161-3, offers a reversed seal inscription which dates from early in the First Millennium. In both instances everything points to an illiterate Iranian craftsman. A smith in Old Babylonian Mesopotamia, even if illiterate, could hardly not have been aware that seals were for rolling on soft clay, and that the design on the stone was therefore reversed. Old Babylonian seal inscriptions are often very clearly cut so that a conscientious but illiterate craftsman could copy one without difficulty. That the inscription, though reversed, is not upside down or on its side suggests that an actual seal rather than a seal inscription on another medium was used, since Old Babylonian seals bear human figures which indicate the way up. The present writer would, then, interpret this vessel as further evidence for his view that inscriptions on ancient Iranian bronzes were copied in Iran from stray Mesopotamian objects. The date of the original from which the inscription was copied proves nothing of course about the date of the vessel. Seals circulated as curios or charms long after their manufacture.2

This is a typical Old Babylonian seal inscription such as occurson seals to be dated between c. 2000 and I600 B.C.: the pattern is distinctively Old Babylonian, the father's name is common in that period, and the last line in fact frequently occurs. The problems which arise concern first the genuineness of the vessel and secondly the significance of the object. It has been submitted to A. D. McQuillan, Professorof Physical Metallurgy in the University of Birmingham, and he reports that from a minute examination of the surface nothing was found to cast doubt either on the antiquity of the vessel or on the originality of the inscription. Assuming, then, its genuineness, though the report of course declares lack of objection rather than positive proof, what is its relevance? First, it seems to bear the only Sumero-Babylonianinscription from the Old Babylonian period on an Iranian bronze. The datable inscriptions otherwise fall into two groups, those from the third

xAt Tepe War Kabud.

See also P. R. S. Moorey, Catalogue the of AncientPersian Bronzes in the Ashmolean Museum (1971), p. 262, no. 495. Previously known inscribed bowls of this type bear Old Akkadian inscriptions, see P. Calmeyer, Datierbare

Bronzen (1969), pp. 27-8. 2 Note, for example, that a Hellenistic grave at Nimrud con-

tained an Old Akkadian seal among other more recent specimens (Iraq 24 (1962), p. 34, ND. 6098).

165
14A

Pl. L An inscribed Iranian Bronze vessel, 8.5 cm. in diameter, 4.6 cm. deep.

166

JOURNAL

OF PERSIAN

STUDIES

A NOTE ON SASANIAN HARPIES By Judith Lerner

Like earlier peoples in the Near East, the Sasanians favoured composite creatures,such as humanheaded bulls, griffins, and grylli,3 for the decoration of objects, particularly seals. One species of fantastic creaturewhich has hitherto gone unrecognized, and has even been denied4as belonging to the Sasaniarirepertory of artistic motifs, is the Sasanian version of the human-headed bird or harpy. On two seals in the British Museum the upper half of a nude male figure, holding a ribboned ring, is joined to the rear portions of a bird (P1. II I; catalogue nos. I and 2, p. 170). Seals in-the State Hermitage Museum, Leningrad, the New York Metropolitan Museum of Art, the Mohsen Foroughi Collection, Tehran, and from the site of Qasr-i Abu Nasr near Shiraz depict a bearded or beardless male head attached to the body of a bird (P1. II 2-5; catalogue nos. 3-7). Other seals combine the lower parts of a bird with the head of a ram (P1. II 6; catalogue nos. 8 and 9), an ibex (catalogue no. human and animal forms manifests something more than a fondness for bizarre animal and human mixtures. Indeed, it suggests that these creatures possessedfor the Sasanians some symbolic value. In the seventh- to eighth-century Sogdian palace at Varakhsha,V. A. Shishkin found stucco fragments of several bird-like creatures with female heads and breasts that had belonged to a large composition which probably depicted a mythological hunting scene.5 As reconstructed by their excavator, these creatures wear some of their hair in braids or long curls that fall on their chests and down their backs, while the rest is gathered into a low knot on the top of their heads; their breasts are covered by a short, draped garment. Their lower bodies are feathered with large, leaf-like forms, and fragments of their raptorial claws suggest that they originally stood on the tops of trees. Two similar beings, with female human heads but not torsos, are represented in clay sculptures at the nearly contemporary Sogdian site of Pjandzhikent, where they formed a symmetrical group within the niche of a small sanctuary, presumably decorating its arch or ceiling.6 These figures are more crudely done than those at Varakhsha; their hair is short at the sides, but like the Varakhsha creatures,is gathered above their foreheads into a top knot of Indian fashion. Both Shishkin and the excavator of the Pjandzhikent sculptures, A. M. Belenitskij, relate these creatures to the garudaand kinnara Indian art. The first, orginally a sun-bird, was the mount of the of god Vishnu, and, by the fourth century A.D.,had become the imperial emblem of the Gupta dynasty.7 In late Gupta times, Garuda was pictured as a hybrid being with a plump bird body and human face
3 Curiously, sphinxes do not appear in Sasanian art. The

Io), a stag (P1. II 7; catalogue nos. 11-13) or an antelope (catalogue no. 14). The choice of only these

creature described by A. D. H. Bivar in his Catalogueof the Western Asiatic Seals in the British Museum: StampSeals II: The SassanianDynasty (London, I969) as a "sphynx " (P1. 14: EH I-8) is actually a reclining winged human-headed bull, its body clearly bovine, not leonine. The same confusion occurs in E. Baer's monograph, Sphinxesand Harpies in Medieval Islamic Art. An Iconographical Study (OrientalNotes and Studies 9, Jerusalem, 1965), even though she notes that the bearded humanheaded creature has a bull's body (p. 22 and n. 85; also P1. XXI: 38). It is indeed odd that the Sasanian artists ignored or consciously rejected the sphinx despite its popularity (or because of it?) in Achaemenid, Parthian and GrecoRoman art. Its reappearance five centuries later in Islamic times speaks for the persistence of images in the Near East. 4 In trying to establish the iconographic prototype for the harpy, Baer cites only examples from Central Asia and India as immediate antecedants and suggests " an additional source of inspiration, rooted in classical antiquity " (op. cit., pp. 47 and 82) because " there is little evidence so far of representations of human-headed birds in the metropolitan art of Iran from

Achaemenid to Sasanian times which ultimately might have served as a prototype for the Islamic artists " (p. 25). A. M. Belenitskij claims that there are no human-headed birds in the art of pre-Islamic Iran (" Resul'taty raskopok na gorodishche raboty v drevnogo Pendzhikenta v i960 g ", Arkheologicheskie Tadzhikistane 8 (1960), p. IIo). R. G6bl does not include human-headed birds as a distinct type in Der Sasdnidische Siegelkanon (Handbiicherder Mittelasiatischen Numismatik IV, I973). On P1. 27, " Fabeltiere ", human-, ram- and stagheaded birds are classified as " Hahn mit .. ." (No. 74) and " Sonstige Fabeltiere " (No. 75)5 V. A. Shishkin, Varakhsha(Moscow, 1963), pp. 175-6, figs. 89-91 (reconstruction in fig. 90). 6 Belenitskij, op. cit., pp. xo8-Io and fig. 18; and idem, " Iz istorii kul'turnykh svjazej srednej azii i indii v rannem srednevekov'e ", Kratkie soobshcheni ja Instituta istorii material'noj kul'turyA N SSR 98 (1964), PP. 33-41 and fig. 9. (second ed., 7 J. H. Banerjea, The Development HinduIconography of Calcutta, 1956), pp. 116-17, 154, 429-532, passim; Baer, op. cit., pp. 44-5.

SHORTER

NOTICES

167

and their female companions, kinnaris, a second hybrid form in Indian are (P1. II 8). Male kinnaras are the celestial musiciansof the gods and appear in the early Buddhist monuments iconography. They of Central India-and later, in Central Asian Buddhist caves-with full bird bodies or simply the lower parts or tail, human heads, sometimes with human torsos, and wings attached to human arms which may hold garlands or other offerings.8 The ivory hoard from Begram, Afghanistan, which is believed to be an Indian import of the first centuries A.D., contains representationsof human-headed, bird-bodied kinnaras,9 a glazed zoomorphic pitcher from the site is in the form of a bird with the and torso and head of a woman.10 Paired kinnaras with completely avian bodies, and heads as their upper human element, occur in the fifth and sixth centuries, in the late Gupta period." only Belenitskijviews the appearance of the creatures at Pjandzhikent and Varakhsha as influenced by Gupta India,12but he considersthem in their Sogdian context to be symbols of the human soul.13 The private chapel at Pjandzhikentto which the paired human-headed birds belonged was dedicated to an ancestor cult that was broadly diffused in Central Asia.14 This funerary association recalls the halfbird, half-human female harpies and sirens of Greek mythology who serve as vehicles for the deceased; the most notable depiction of these beings who are shown bearing away the soul of the deceased is found on the " Tomb of the Harpies " from Xanthos, Lycia, dated to the late sixth or early fifth century B.C.15 In Roman times, sirens or harpies appear in early Imperial coinage16and on the contemporary Arretine ware." They had already been known in the East in the second century B.c. among the Parthians at Nisa (Pl. II 9), and later in Parthian Iran.s8 In all these representations,the human head, and body when it is included, are female, although androgynous sirens had occurred earlier in the Hellenistic period.1" In Sasanian glyptic art, the transformationof the harpy-siren into a male being may be due to the Sasanian partiality for masculinefabulous creatures. Except for nursinganimals, those species generally representedon the seals seem to be male.20 The nude winged figure that bears the ribboned ring, as in P1. II I, is always masculine;21 with the exception of the late Sasanian Nikes of TRq-i Bustan, the wreath-bearingNike or Victory that was used throughout the Seleucid and Parthian periods has been abandoned.22 On the other hand, the prototype for the Sasanian harpy may have come from the East in the form of the Indian Garuda or kinnara. In the Cave of the 35-metre Buddha at
Banerjea, op. cit., pp. 351-2. For examples from the Indian sites of Sdfichi, Mathurd and Sarnath, see G. Combaz, L'Inde et l'Orient classique (Paris, 1937), pp. 139-40 and Pls. 66-7; for some later Central Asian examples, see ibid., P1l.69 (" Cave of the Kinnari " near Qumtura), and A. von Le Coq, I Chotscho (Berlin, 1913), P1. I5 C. archeologiques Bdgram, 9 J. Hackin et al., Nouvelles recherches d MDAFA XI, (1954), p. Ioo and figs. 8 and 495. 10Idem, Recherches archeologiques Bigram, MDAFA IX, (i939), d figs. 241-2. M. M. Hallade, Arts de l'Asie ancienne. Thdmeset motifs, I: xx L'Inde (Publicationsdu Musie Guimet,recherches documents et d'art et d'archeologie Paris, 8957), P1. IX: 82. 5, " 1a Iz istorii kul'turnykh ", pp. 36-7. Many motifs-notably some of western origin-found in India appear to have influenced artists in Central Asia long after these motifs had ceased to be employed in Indian Buddhist art. For example, Dionysiac images continue in Central Asia after their disappearance from North-west India at the end of the fifth century (M. A. Carter, " Dionysiac Aspects of Kushdn Art ", Ars VII (1968), p. 146). For traces of Gupta influence in Orientalis Sogdian and Choresmian art, see M. Bussagli, Painting of CentralAsia (L. Small, trans., Geneva, 1963), p. 15. 11 This parallels the ancient Egyptian concept of the soul, or ba, that escapes from the body of the deceased in the shape of a bird. According to F. Cumont (After Life in RomanPaganism, London and New Haven, 1922, p. I57), this belief was held in other parts of the Eastern Mediterranean. x4 Belenitskij, " Resul'taty ", p. IIo. He sees the bird as personifying the soul of the deceased and finds confirmation of this
8

idea in the symmetrically paired birds that decorate the front panels of Pjandzhikent ossuaries. Iranfrom Its Originsto the Time 15 R. Ghirshman, The Art of Ancient the of Alexander Great(New York, 1964), p. 338, fig. 425 (now in the British Museum). The connection was suggested also by Shishkin (op. cit., p. I75), who further identified the Varakhsha creatures with the Iranian homa. Belenitskij refutes this last interpretation, citing lack of evidence in written sources and artistic monuments (" Iz istorii kul'turnykh ", p. 40), but he was unaware of the Sasanian human-headed birds (ibid., p. 42 and supra,n. 4). auf S16F. Imhoof-Blumer and 0. Keller, Tier- und Pflanzenbilder Miinzen und Gemmendes klassischenAltertums(Leipzig, 1889), Pl. XII: 38 (denarius of Augustus). 17 G. Weicker, Der Seelenvogel der alten Litteratur und Kunst: ein in Untersuchungen (Leipzig, I9o02), fig. mythologisch-archaeologische
I02.

18 Baer, op. cit., P1. XXIII:

43 (presumably found in Iran and similar to P1. II 9). 19 Weicker, op. cit., p. 182, fig. 92. 20 Even nursing lions are accorded manes (Bivar, op. cit., P1. Io: DF I, 4-6). 21 See ibid., P1. 6: BK 1-5; Gdbl, op cit., Pl. 7: No. 13 (" Nike; Victoria; ' Putto '; " only this last identification is pertinent). 22 Though widespread on Seleucid coins, the Nike was not too common on those minted in Iran (G. Le Rider, Suse sous les et et Sgleucides les Parthes. Les Trouvaillesmonetaires l'histoirede la ville, MDP XXXVIII, (1965), p. 360; an exception is Pl. III: A). For the Parthian coins, see ibid., Pls. XI-XXXVIII, on [continued nextpage

168

JOURNAL

OF PERSIAN

STUDIES

the painting of the moon-god's quadriga depicts in attendance two kinnara-like beings Bamiymn, with bearded and nimbate heads.23 They appear to carry torches which may connote ideas of
glorification.24

Another source may be the ancient Mesopotamian figure of the bird-man Zu that occurs in seal compositions and is associated with the liberation or rebirth of the vegetation god.25 Although this half-bird, half-man figure is most prevalent in the third millennium B.c., references to the myth of Zu seem to have survived on seals of the Late Assyrian period.26 It is at this time that human-headed birds appear in seal designs that resemble Greek sirens, but with male features."7 On some gems of the creatures with their bearded Achaemenid period-apparently carved under Greek influence-these heads wear the Persian dentated crown (P1. II Io).28 In pre-Achaemenid Iran, a bearded bird-man occurs on a bronze vessel from Luristan,29 and others are found on oriental bronze cauldrons, where they correspond to the female siren handles of Greek examples.30 A bronze statuette discovered at Begram portrays a human-headed cock. O. Kurz recognized the figure as Mercury in the same form as he appears on a Roman gem, with petasus and caduceus(P1. II I I).31 In Greco-Roman religion, Mercury is affiliated with the cock in his role as a psychopomp, who enables the soul to triumph over death by assuring it safe passage to the after life; the cock, as herald of the sun, and thus of resurrection, serves as the god's emblem.32 Mercury's funerary role is emphasized by his association with a siren or harpy on coins and gems.33 The bird bodies of the creatures that are depicted on the Sasanian seals are those of birds that are duck (catalogue commofily found in Sasanian representations: cock (catalogue nos. 6, 8 and 10-12), no. 9), peacock (catalogue no. I4), eagle or some generalized species (catalogue nos. I-5, 7 and 13). Except for those representations of Mercury as a cock, the bodies of the non-Sasanian examples are mainly of an aquiline form. The visual resemblance of the Sasanian creatures to the human-headed birds of western and eastern traditions appears to transcend mere form to partake in similar meanings and functions. In the
A. Godard et al., Les Antiquitisbouddhiques Bdmiydn,MDAFA de II (Paris, 1928), p. 21, fig. 6 and Pl. XXII; A. C. Soper, "The Dome of Heaven in Asia ", The Art Bulletin XXIX (1947), pp. 232-3 and n. 39. The painting is of disputed date, but is certainly contemporary with the Sasanian dynasty. 4 Cf. H. Seyrig, " Antiquit6s syriennes, 47. Antiquit6s de BethMar6," Syria XXVIII (1951), pp. II6 and 118; G. M. A. Richter, The EngravedGemsof the Greeks,Etruscansand Romans, Part Two: EngravedGems of the Romans. A Supplement the to Historyof RomanArt (London, 1968), No. 164 (the Triumph of Dionysos). 2- H. Seals. A Documentary Frankfort, Cylinder Essay on the Art and Religionof the AncientNear East (London, p939),pp. 135-6 and P1. XXIII: a-i. a8 Ibid., p. 198. 27 E. Porada, Corpus AncientNear EasternSeals in North American of I: Collections, The Collection the PierpontMorgan Library(New of York, 1949), P1. 92: 633-4 (linear style of the ninth-eighth centuries B.C.). asJ. Boardman, " Pyramidal Stamp Seals in the Persian Empire ", Iran VIII (1970), p. 28, notes that " the siren is not a normal Achaemenid creature " and that all the examples that he cites, including our P1. II Io, "have plump egg-shaped bodies like sirens on Late Archaic Greek gems from East Greek
2a

foundon Cyprus, ". development, workshops A contemporary into a beardedmale the transforms Greekfemale siren-harpy creature(Weicker,op. cit., p. 12). 29For example,Survey IV, PI. 63. s0 Ghirshman, cit., p. 338; cf. Baer,op. cit., pp. 25-6, no. o02. op. Oriental male-headed examples are discussed by O. W. Muscarella," The OrientalOrigin of Siren CauldronAttachXXXI (1962), pp. 317-29. ments", Hesperia, 31 In Hackin et al., Nouvelles recherches, 147-8 and figs. 455 pp. (statuette)and 458 (seal in Berlin,our P1. II Ix). Kurz cites cock without any anothergem that depictsthe human-headed attributes(fig. 456), and a Roman statuetteof an ithyphallic man with cock's crest and attributesof Mercury (fig. 457). A human-headedcock also appears among the Palmyrene de des tesserae(H. Ingholt, Receuil tessires Palmyre, Paris, 1955, P1.XXIX: 603). dans 22 A. Maury, " Des Divinit6set des g6nies psychopompes (1844), l'Antiquit6 et en Moyen-Age", Revuearcheologique Lectures and Life, Three p. 503, and E. S. Strong,Apotheosis After in Phasesof Art andReligion theRoman on certain (New Empire York, 1915), p. 215. 33Imhoof-Blumer Keller, op. cit., Pls. XXI: 36 (coin from and Asia Minor), and XXVI: 33 (scaraboidin Berlin).

continued from previous page] I: passim, and J. de Morgan, Monnaiesorientales, Numismatique de la Perse antique (Paris, 1933), Pls. X-XXII, passim. The wreath-bearing Nike also appears on the reliefs of Mithradates II (c. 123-87 B.c.) and Gotarzes II (A.D. 38-51) at Behistun aus (E. Herzfeld, Am Tor vonAsien, Felsdenkmale Irans Heldenzeit (Berlin, 1920), p. 37, fig. Ii and P1. XXIII: bottom, respectively). Only one Nike-like figure among the Sasanian seal

designs is known to me. On an impression from Qasr-i Abu Nasr a winged long-skirted figure holds leafy fronds before a male profile bust; the late cursive inscription on the sealing would relate the figure to the wreath-bearers at TAq-i Bustin from Qasr-i Abu Nasr: Seals, (R. N. Frye, ed., SasanianRemains Sealings, and Coins, Harvard Iranian Series I, Cambridge, Mass., 1973. D. 95).

Pl. II. Seal impressions a gilded silvervesselsupport(9) illustrating'A Note on SasanianHarpies'(pp. I66-'I72). and

SHORTER

NOTICES

169

birds or bird-like beings, both real and mythic, are described as opponents of evil, " created Bundahiln, in opposition to noxious creatures".34 In particular, the cock, according to Zoroastrian belief, was created to oppose demons; as herald of the dawn, his crowing was believed to " dazzle away the fiends ".35 The combination of a male human head with a cock's body in P1.II 5 may representsuch a benevolent being-a Sasanian counterpart to the western image of Mercury, the leader of the dead,
as a cock (P1. II 11).

The joining of the harpy-siren-kinnara with the male Victory (P1. II I) enhances the emblematic value of the Sasanian harpy. In the triumphal scenes on the royal rock reliefs,36and the hunting scenes on some silver plates,37this nude winged figure bestows the ribboned ring or pearl necklace on the Sasanian monarch. In other reliefs, and on the coins, the ribboned ring is a symbol of investiture, the means of conferringkingly glory and the right to rule.38 Translated into the more popular medium of glyptics, the ring may be seen as a universal symbol of glory and of good wishes, a fitting gift to be conveyed by a bird-likecreature,39 especially if it is one that functions as a psychopomp. The combination of the moufflon, ibex, stag and antelope heads with a bird's body-in particular, the mouffionheaded cock of P1. II 6-further stressesthe value of the Sasanian harpy: as avatars of the Iranian god of victory, Verethragna, these horned creatures could ensure the attainment of glory in this world and beyond.40 These ideas of victory and glory which are associatedwith Verethragna may representthe popularization of a royal theme. One of the incarnations of Verethragna, the ram or moufflon, is described in the Karnamak-i as Ardashir Pdpakan, a personification of the xvarror kingly glory.41 The diffusion of or official symbolism seems to have occurred with the proliferationof seals in the last originally royal centuries of the Sasanian period.42 Just as the ownership and use of seals was expanded among the extended to all people-not as populace, so were the imagery of the seals and such concepts as the xvarr the auspicious royal glory but as a general wish for good fortune.43 Later, in Islamic times, the presence of certain creatures, including the sphinx and harpy, evoked prospectsof happiness,successand longevity for the owners of the objects that they adorned.44To depict the human-headed bird that would performsuch a function, the Islamic artist did not have to rely upon comparable beings from Central Asia, India or the West, but was able to find a direct model in the Sasanian harpy.

Bundahiln XIX, 19-21 (E. W. West, The SacredBooks of the 40 Specifically, the ram (moufflon) and buck (stag) are mentioned East V, Oxford, I88o, p. as the eighth and ninth incarnations of Verethragna in the 7I). Bahram Yalt VIII, 23 and IX, 25 (Darmesteter, SacredBooks '6 Vendidad,Fargard XVIII, 15-6, 23-4 (J. Darmesteter, The SacredBooks of the East IV, Oxford, 1895, p. 197 and n. 199); XXIII, pp. 237-38). 41 Ibid., p. 237, n. 3. The identification of Verethragna may have BundahifnXIX, 33 (West, op. cit., p. 73)36 IAE, Pls. CXV-VI (Shipiir I's reliefs at Bish~pui: here the been extended to other horned animals, such as the antelope, whose head is joined to a peacock's body in No. I4; the ring is open, as if it were a untied diadem). 7' Sasanian Silver. Late Antique and Early Medieval Arts of Luxury fluttering ribbons about this creature's neck are no doubt further evidence of its special quality. from Iran (The University of Michigan Museum of Art, Ann Arbor, x967), No. 4 (loan, Metropolitan Museum of Art), and 42 A heightened demand for seals may have been prompted by Kav~d I's or Khusr6 I's administrative and fiscal reforms. IV, P1. 229 B (Museum fir Islamische Kunst). A similar Survey Cf. Frye, op. cit., p. 47. winged figure bearing a necklace appears on some plates with banquet scenes, but these are post-Sasanian, such as ibid., 43 Evidence of the extension of a private xvarris found in ninthcentury Zoroastrian writings, although it is not present in P1. 230B (State Hermitage Museum). earlier literature. See H. W. Bailey, ZoroastrianProblemsin 38 From earliest Sasanian times: IAE, Pls. CVIII and CX Books (" Ratanbai Katrak Lectures ", Oxthe Ninth-Century (Ardashir I's reliefs at Firaizdbdd, Naqsh-i Rajab and Naqsh-i Rustam). ford, I943), p. 33. 39 I hope to discuss the significance of the ribboned ring and 44 See R. Ettinghausen, " The Wade Cup in the Cleveland related symbols, and their association with the beings that Museum of Art, Its Origin and Decorations," Ars Orientalis, hold them in a future article. II (1957), PP. 360-I; cf. Ja. Orbeli, SurveyI, p. 741,

34

170

JOURNAL

OF PERSIAN

STUDIES

CATALOGUE OF SASANIAN SEALS WITH THE REPRESENTATION OF A HARPY* I. British Museum, London, Department of Western Asiatic Antiquities, II19376. Winged male figure, with the lower parts of a bird, to the right, bearing a ribboned ring. Chalcedony ellipsoid. Bivar, op. cit., P1. 6: BK 6 and p. 60; our P1. II I. London, WA I19519. Same descriptionas No. I. Almandine bezel. Ibid., P1.6: BK 7 and p. 60. State Hermitage Museum, Leningrad, GL 1242. Figure to the right, with human head, and bird body with outspread wings. Fragment of a chalcedony ring. A. Ja. Borisov and V. G. Lukonin, Sasanidskie Gemmy (Leningrad, 1963), No. 537, P. 155 (not illustrated); P1. II 2. Leningrad, GL 1410. Figure to the right, with a beardedhuman head and bird body. Chalcedony ellipsoid. P1. II 3. Metropolitan Museum of Art, New York, Ancient Near Eastern Department, 62.66.24. Beardless male-headed bird to the right, within a wreath; crescent in the field, to the left. Banded agate ellipsoid. P1. II 4. Mohsen Foroughi Collection, Tehran. Bearded male head with earrings attached to a cock's body, walking to the right. In the field, above its tail, a six-pointed star. Chalcedony plaque
with rounded corners. P1. II 5.

2.

3.

4. 5. 6.

7. Iran Bastan Museum, Tehran, acc. no. unknown, from Qasr-i Abu Nasr. Bearded male, humanheaded bird, to the right. Carnelian; shape of stone unknown. Frye, op. cit., No. 20 and p. 398. Museum fiir Islamische Kunst, West Berlin, VA 1486. Moufflon-headed cock, to the right. Carnelian bezel. P. Horn and G. Steindorff, Sasanidische (" Siegelsteine K6nigliche Museen zu Berlin. Mitteilungen aus den orientalischen Sammlungen," IV, Berlin, 1891), Pl. V: 1486; P1. II 6. 9. London, WA 120298. Moufflon-headed duck, to the right, surrounded by a Pahlavi inscription. Agate ellipsoid. Bivar, op. cit., P1. 26: MG 3 and p. I I0. o. Ashmolean Museum, Oxford, Department of Antiquities, Queen's College loan S6I. IbexI headed cock to the right. Almandine bezel with convex face. I I. Formerly Ackermann Collection, New York (present location unknown). Stag-headed cock to the right. Survey P1. 255: PP. IV, I2. Moore Collection. Stag-headed cock, to the right. In the field, above its tail, a crescent; to its left, a six-pointed star. Haematite ellipsoid with back carved in a lotus design. G. A. Eisen, and AncientOriental of of Cylinder OtherSeals with a Description the Collection Mrs. Wm. H. Moore
(OIP XLVII, Chicago, 1940), P1. XI: I II.

13. New York, 38.40.99, from Nishlipir. Frontal figure of a stag-headed bird, with outspread wings, and head to the right; the lower portion may representpendent ribbons instead of the bird body. Carnelian bezel. P1. II 7. 14. Prof. K. J. Muiller,Bonn. Antelope-headed peacock, to the right; the double ribbons around its neck flutter out behind. Leafy branch in the field, to the right. Material and form of seal unknown. Gobl, op. cit., P1. 27:75a. LIST OF ILLUSTRATIONS PI. II P1. II I. 2. British Museum, WA I I9376 (courtesy, the Trustees of the British Museum) = No. I. State Hermitage Museum, GL I242 = No. 3.

* This does not claim to be a complete catalogue; I hope that other seals that will enlarge this corpus will be found and published from other collections.

SHORTER

NOTICES

171

P1. II

3.

State Hermitage Museum, GL 1410 = No. 4-

P1. II P1. II P1. II P1. II Pl. II P1. II

4. 5. 6. 7. 8. 9.

Metropolitan Museum of Art, 62.66.24 (courtesy, The Metropolitan Museum of Art, Anonymous Gift, 1962) = No. 5. Mohsen Foroughi Collection, Tehran (Photograph, courtesy of R. N. Frye) = No. 6. Museum fiir Islamische Kunst, VA 1486 = No. 8. Metropolitan Museum of Art, 38.40.99 (courtesy, The Metropolitan Museum of Art) = No. 13. Seal impression of KumSragupta I, c. A.D. 530 (Banerjea, op. cit., P1. XXVIII: I). Gilded silver vessel support in the form of a siren, from Nisa, now in the Historical Museum of the Turkmenistan SSR, Ashkabad (G. A. Pugachenkova, IskusstvoTurkmenistana (Moscow, 1967),

fig. 48). II io. Achaemenidstampseal, Metropolitan Museumof Art, 41.I60.268,with a beardedand crowned P1. siren sniffingat a flower (courtesy,The MetropolitanMuseum of Art, Bequestof W. Gedney Beatty, 1941). II i 1. Romangem with Mercuryas a human-headed Die I cock (A. Furtwangler, antike Gemmen,(Berlin, P1. I900), P1.XLVI: 29). taken by the author,who wishesto thankthe following Figs. 2, 3 and 6 are made from seal impressions scholarsfor generouslyallowing her to study the Sasanianseals in their respectiveinstitutions: K. Brisch, MuseumffirIslamische Museumof Art; V. G. Lukonin,StateHermitage Kunst; P. O. Harper,Metropolitan Museum.

A NEW DISCOVERY AT BISHAPUR

In January and February 1975 Mr. Ali Akbar Sarfaraz,Director of excavations at Bishapur,made some important discoveries in the Tang-i Chogan. On removing the stone conduit, which is to be replaced by water pipes, he revealed hitherto unsuspected portions of some of the existing Sasanian reliefs. The most intriguing discovery is that of a prone figure beneath the king's horse in the investiture scene of Bahram I (Pls. III and IV). This prone figure appears to be a secondary addition to the relief. This information has been generously supplied by Mr. Sarfaraz. The removal of the conduit further revealed an additional register at the base of the concave bas relief of Shahpur I showing his victory over the Romans. This register shows Iranian cavalry and Roman captives.

I the revealed Mr. A. A. Sarfaraz.(Photograph, the by figurebeneath King'shorse, recently of Pl. III. Bishapur.TheInvestiture Bahram showingtheprone

the addition. (Photographs, Royal the Pl. IV. Bishapur.Details of thepronefigure beneath horseof BahramL It appearsto be a secondary Road Expedition.)

SHORTER

NOTICES

171

P1. II

3.

State Hermitage Museum, GL 1410 = No. 4-

P1. II P1. II P1. II P1. II Pl. II P1. II

4. 5. 6. 7. 8. 9.

Metropolitan Museum of Art, 62.66.24 (courtesy, The Metropolitan Museum of Art, Anonymous Gift, 1962) = No. 5. Mohsen Foroughi Collection, Tehran (Photograph, courtesy of R. N. Frye) = No. 6. Museum fiir Islamische Kunst, VA 1486 = No. 8. Metropolitan Museum of Art, 38.40.99 (courtesy, The Metropolitan Museum of Art) = No. 13. Seal impression of KumSragupta I, c. A.D. 530 (Banerjea, op. cit., P1. XXVIII: I). Gilded silver vessel support in the form of a siren, from Nisa, now in the Historical Museum of the Turkmenistan SSR, Ashkabad (G. A. Pugachenkova, IskusstvoTurkmenistana (Moscow, 1967),

fig. 48). II io. Achaemenidstampseal, Metropolitan Museumof Art, 41.I60.268,with a beardedand crowned P1. siren sniffingat a flower (courtesy,The MetropolitanMuseum of Art, Bequestof W. Gedney Beatty, 1941). II i 1. Romangem with Mercuryas a human-headed Die I cock (A. Furtwangler, antike Gemmen,(Berlin, P1. I900), P1.XLVI: 29). taken by the author,who wishesto thankthe following Figs. 2, 3 and 6 are made from seal impressions scholarsfor generouslyallowing her to study the Sasanianseals in their respectiveinstitutions: K. Brisch, MuseumffirIslamische Museumof Art; V. G. Lukonin,StateHermitage Kunst; P. O. Harper,Metropolitan Museum.

A NEW DISCOVERY AT BISHAPUR

In January and February 1975 Mr. Ali Akbar Sarfaraz,Director of excavations at Bishapur,made some important discoveries in the Tang-i Chogan. On removing the stone conduit, which is to be replaced by water pipes, he revealed hitherto unsuspected portions of some of the existing Sasanian reliefs. The most intriguing discovery is that of a prone figure beneath the king's horse in the investiture scene of Bahram I (Pls. III and IV). This prone figure appears to be a secondary addition to the relief. This information has been generously supplied by Mr. Sarfaraz. The removal of the conduit further revealed an additional register at the base of the concave bas relief of Shahpur I showing his victory over the Romans. This register shows Iranian cavalry and Roman captives.

i\.
<

) q, Bastom

.
i.

U. S. S. R
CASP/AN SEA
.-

L '"'" "-, *MESHED ? "

'

0"

.Jurjan Tepe * * Sordlor Tepe Tureng rdlarTee HosonlOu oTABRIZ iTakhLb Sulaiman . TEHRAN SSagzabad * HAMADAN (3" *
<(

Kangavar 0, *
-Ir

*Nush-i Jan

Ganj Dareh

ISFDarabadAHAN
Djarab*d

SusdC g1U0ghaMish
*Haft <o S,

Shahr*,,Sokhta * *oShahdad ,e. ? zAHEDAN _._ Ghubayrd . , .

Tlep TallMlyKERMAN S Tall-i Malyan *SHIRAZ Bishapur ..Persepolis

I

,Z
SAUDI D--I ARABIA

<%
SITE LOCATION MAP

0 1 o

500

iortr

opposite) of (see Survey Excavations

SURVEY OF EXCAVATIONS IN IRAN -

1973-4

This year we provide a list of the principal excavations carried out between September 1973 and December I974. The list includes the approximate date of each season, the name of the site, the name of the Director and the name of the sponsoring Institution.' Those excavations that are described in greater detail below are marked with an asterisk. There are in addition three reports of surveys carried out in 1973-74: in the Hulailan valley, in the Kangavar valley and in the Khana Mirza plain. Autumn 1973 Autumn 1973 Sagzabad; E. O. Negahban; University of Tehran. *Siraf; D. Whitehouse; British Institute of Persian Studies. Autumn 1973 Shahr-i Sokhta; U. Scerrato; IsMEO. Autumn I973 Jurjan; Y. Kiani; Iranian Centre for Archaeological Research (hereafter ICAR). Spring 1974 *Chogha Mish; P. Delougaz and H. Kantor; Oriental Institute of the University of Chicago and U.C.L.A. Susa; J. Perrot; Ministbre des Affaires Etrangeres Frangais; CNRS; National Spring I974 Organization for Restoration of Historical Monuments (hereafter NORM) and ICAR. Spring 1974 Spring 1974 Spring 1974 Spring 1974 Summer I974 *Djaffarabad; G. Dollfus; CNRS. Bishapur; A. A. Sarfaraz; ICAR and NORM. Shahdad; A. Hakimi; ICAR.

Haft Tepe; E. O. Negahban; ICAR. Jurjan; Y. Kiani; ICAR. Summer 1974 *Ganj Dareh; P. Smith; Universite de Montreal. Summer 1974 *Hamadan, Parthian cemetery; M. Azarnoush; ICAR. Summer I974 *Hasanlu; R. H. Dyson; University of Pennsylvania. Summer 1974 Summer 1974 Summer 1974 Kordlar Tepe; A. Lippert; Innsbruck University. *Ghubayra; G. Fehervari and A. D. H. Bivar; School of Oriental and African Studies.

Institut. *Bastam; W. Kleiss; Deutsches Archdiologisches Summer 1974 *Takht-i Sulaiman; R. Naumann; Deutsches ArchiiologischesInstitut. Summer I974 *Nush-i Jan; D. Stronach; British Institute of Persian Studies. Autumn 1974 Pusht-i Kuh; L. Vanden Berghe; Ghent University. Autumn I974 Kangavar; Kambakhsh Fard; ICAR. Autumn I974 *Malyan; R. H. Dyson and W. Sumner; University of Pennsylvania. Autumn I974 Sagzabad; E. O. Negahban; University of Tehran.
1

We are very grateful to the Iranian Centre for Archaeological Research for their cooperation in supplying details of the past year's excavations.

173
15

174 Bastam

JOURNAL

OF PERSIAN

STUDIES

Die Grabungen in Bastam wurden in der Zeit vom Io. Juli bis zum 20. September 1974 in der fiinften Kampagne weitergeftihrt(siehe Iran VI (1968), 166; VII (1969), 188; VIII (I970), 176-78; IX (1971), 165-68; XI XII (1974), 205-07.) (i973), 185-88;

Es wurde wieder an zehn Stellen die Arbeit der Vorjahre weitergefuihrt. I.) Das Gebiet siidlich des Siidtores wurde in mehreren Grabungsschnittenuntersucht und die Hallenbebauung dieses Bereiches konnte weiter im GrundriBgeklirt werden. Das Gebiude oberhalb am Hang, siidlich des Stidtores,wurde in seinem architektonischenBestand vollstandig geklirt. Es besteht aus einem mittleren Hof mit 2 iwanartig iiberdeckten Teilen an den beiden Schmalseiten und einer Reihe von Raiumen, die auf allen vier Seiten den zentralen Hof umgeben (Abb. I). Auf der Ostseite sind es 9 kleine Kammern, in denen-wie in den Vorjahrenauch I974 wieder einige reichere Funde gemacht wurden, darunter Keramik und Bruchstiickeeiner weiteren urartaiischen Tontafel, sowie Metallfunde, vornehmlich Bronze- und Eisengerdite.

DORF

"

.? ,9FWG

\\

'$ \l

~d-

SUD

r0

R
6

..
. . . -

suoroR ,
1 i k

- .

..

~
. I'?:.iB"B..,

\W _02__

10

2 ?

30

5O-"1

1/

Fig.

I.

2.) Im Bereich des Siidtores wurden sowohl innen wie auBen weitere Flaichender Steinplattenpflasterung des urartaiischenAufweges aus der Ebene hinauf auf den Festungsberg freigelegt und n6rdlich des Tores konnte ein Teil des unmittelbar an den Pfeilersaal neben dem Tor angrenzenden Gebiudetraktes ausgegraben werden. Kleinere Schiirfungen brachten dariiber hinaus weitere Klrung der Topographie des Bastamer Burgberges,speziell im Bereich des Stidtores. 3.) Im Pithosraum neben der Tempelterrasseauf halber Hohe des BastamerBurgbergeskonnten Pfeiler der inneren Stiitzreihe freigelegt werden. 9 4.) Die Grabungim Bereichdes Tores in die OberburgergabdurchgehendeBruchsteinfundamente,
aber, an Hand von sehr gut gearbeiteten Lavagestein-Quadern und Bronzefunden, die in Zusam-

SURVEY

OF EXCAVATIONS

175

menhang mit dem TorverschluBzu stehen scheinen, auch einen Hinweis auf die Lage des Tores vom Pithosarsenalder mittleren Burg in die Oberburg. 5.) Im Bereich der Innenburg wurden 3 Raume ganz oder teilweise freigelegt, die bei sehr starken urartaischen Mauern ein System kleiner Raiume an der Burgmauer anzeigen. Die Raume sind auf Grund der Funde als Wirtschaftsraiume bezeichnen. zu 6.) Im oberen Teil der Oberburg konnte dieses System kleiner Kammern an Hand spairlicher Steinmauerreste und an Hand der Felsabarbeitungen und Felsabtreppungen sicher nachgewiesen werden und der Gesamtplan der Bastamer Burg vervollstaindigt werden. 7.) Die Grabungen am Hallenbau neben dem Nordtor erbrachten mehrere kleinere Rdiume,die mit der Halle in baulichem Zusammenhang stehen. Der Komplex des Hallenbaus vergr6Bert sich durch jede Grabungskampagne. Gleichzeitig wird auf Grund der Keramikfunde immer deutlicher, daB der Komplex die ailtesteurartaiische Anlage in Bastam ist und ins 8. Jahrh. v. Chr. zu datieren ist
(Abb. 2).

SIEDLUNG

10

1 I I

20

I I

30o

YO 5so

NORD-

IH

I
f..-:I"
Fig. 2.

Haus in der Siedlung konnte das normale urartaiische 8.) Durch Erweiterungder Grabungsflaiche nordlich der Burg weiter freigelegt werden und mehrere R~aumevollstandig untersucht werden, wobei sich wiederum 3 Bauphasen gegeneinander abhoben. 9.) Im Jahre I974 wurde mit der vollsta*ndigen Freilegung der Innenraiumedes Nordgebaiudes begonnen. io.) Die Risalitmauerim Tal gegentiberdem Burgbergkonntein ihrer Ausdehnung (157 X 125 m.) durch zusammenhaingendeGrabungsflichen und punktf6rmige Grabungsweise geklirt werden und auch das auf der Westseite, auf den BurgbergausgerichteteTor konnte freigelegtwerden. Die Funktion der Anlage bleibt aber weiterhin ungeklirt. Die Anlage ist gesichert urartaiisch,in ihrem Bereich wurden mehrere Graber des 2.-3. Jahrh. n. Chr. (sptitparthischeZeit Armeniens) gefunden.

(i
0o5

1 0wi

.
01 KHOY
OMARAND

M
o

AHAR

8 SHAHPURO
/

U /

?F

oTAORIZ

.0 ,

*'

SARA.B REZAIYEH

TORKEI j

MARACHEH

.2 ? GOJER KALE

-I ,?
S
/

..
o . ?USHNUIYEH i nm I R 0H

U O OMIANDUAB
1INS(HRIFTEN * SHEYTANASADMAHABAD

BURGEN UND SIEDLUNGEN INSCHRIFTEN

TEPES UND FUNDPLA;TZE

2g

IRAQ,.

Fig. 3.

Die Datierung aller urartaischenAnlagen-mit Ausnahme des Hallenbaus-ins 7. Jahrh. v. Chr. konnte durch die Keramikfundeweiter gefestigt werden. Im Rahmen der Bastam-Grabung wurden von Bastam aus vier urartaiischeBurgen vermessen und ihre H6henlage nivelliert. Es handelt sich um Cheragaiyeh Amir zwischen Djulfa und Marand, um Danalu nardlich Maku, um Kuh-e Sambil norddstlich Rezaiyeh am Ufer des Urmiasees und um am Kale Sarandj am Araxes in der Ndihedes Dreilainderecks Ararat. Cheragaiyeh Amir und Danalu sind Neufunde, zu denen 5 weitere Plitze als Neufunde dazutreten, einer (Stinat) in der weiteren Umgebung von Bastam am Wege nach Van (Tuspa) und einer (Kishmish Tepe I) westlich von Maku sowie zwei (Boji und Kale Sangar) siidwestlich von Ahar, nord6stlich von Tabriz (Abb. 3). Ein ffinfter Platz mit urartaischer Keramik (Achudarreh) liegt am antiken Passweg zwischen Shahpur und Rezaiyeh.
WOLFRAM KLEISS

Mish Chogha The eighth season of excavations at Chogha Mish by the Joint Iranian Expedition of the Oriental Institute of the University of Chicago and the University of California at Los Angeles lasted from its higher part a new strip along the north edge of the previously-excavated Protoliterate house area was dug. Close to the surface were floors, a hearth, and a kiln of an Achaemenid Persian settlement
Work was concentrated on the eastern part of the terrace. In December 21, 1973 to April 9, 1974-

SURVEY

OF EXCAVATIONS

177

from which a good sample of the ceramics of the period was obtained. The Protoliterateremains were extensive and complex, with pits from destroyed higher levels dug into lower rooms and kilns. An elaborate baked-brickinstallation associated with drain pipes seems to have served a specialized craft. Some distance to the south of the Protoliterate sector, on the lower slopes of the terrace, three areas with for the most part Susiana remainswere dug. The southernmostwas a new trench, XXXVII, 20 x 2 m., intended to check the extent of Early Susiana architecture and of Archaic Susiana settlements in that direction. Near the west end was a deep pit dug in the Achaemenid period and filled with mixed debris. Elsewhere Middle and Early Susiana layers were traced. Even more important was a level with pottery of the final stage of the Archaic Susiana period. The sterile soil below it indicated that there was no habitation at this spot in Archaic Susiana I and 2. In order to connect the stratified Early and Archaic Susiana structures found in the extensive Tr. XXI sector and the nearby Tr. XXXII, the space between them was opened up. At one spot near the eroded surface five skeletons lay in debris attributable to the Archaic Susiana 3 stage. In the vicinity were found two substantial stone maceheads and two bituminous stone bowls nested under the sherds of straw-temperedvessels. Well-built walls continue to the south the substantial Early Susiana architecturefound previouslyin Tr. XXI. In general, the architectureand artifacts found this season in the Tr. XXI-XXXII area provide important evidence for the transition between the Archaic and Early Susiana periods. The Gully Cut located very low on the east side of the terrace has been providing stratifieddebris of the three phases of the Archaic Susiana period since 1969-70. In that same season the small Sounding G to the north-east showed traces of architecture. Since more extensive excavations promised good architectural evidence on still outstanding problems of the Archaic Susiana period, the entire space between the old Gully Cut and Sounding G was opened. Only two Protoliterate pits in the southwestern corner of the new area had disturbed underlying Middle Susiana floors and wall foundations. These levels were well-provided with kilns and hearths. Below appeared the walls of rooms of the preceding Early Susiana period. The walls of the lower of two levels are particularly substantial. The greater part of a room some 5 m. long lies within this year's excavations, but its south-western corner disappears under still unexcavated soil. Nearby is a rather thick buttressedwall, which forms a corner. Both its ends are still unexcavated. The architecture and the extensive finds of pot sherdsin what is an outlying part of the site, indicate that the density and importance of the Early Susiana settlement at Chogha Mish were considerably greater than originally anticipated. The eighth season provided excellent evidence for some of the more elaborate manifestations of Early Susiana culture. A 2 cm. high miniature figure of a seated woman is surprisinglywell-delineated for its size. Clay pellets represent details such as the large breasts and three beads of a necklace. A slight groove at the neck may indicate where a head, now missing, was once attached. Fragments of the cylindrical bodies of larger female figures had details of the sex and dress in paint. Larger in scale than any of the torsosis a head with appliques, including " coffee-bean " eyes, and painted bands. In addition to the abundant sherds and to the figurine fragments, the earth in the vicinity of the buttressed wall yielded a finely-stylized stone pendant in the shape of a pig. It is the earliest example of advanced representationalart so far found at Chogha Mish and puts the culture of the Early Susiana
period in a new perspective. P. P. DELOUGAZ and HELENE J. KANTOR Djaffarabad La 66 campagne de fouilles sur le tdp6 Djaffarabad s'est ddroulde du 22 Janvier au 20 Mars I974, subventionnde par le Centre National de la Recherche Scientifique dans le cadre de la RCP 362.2
SM.

Kordovani repr6sentait le D6partement des Antiquit6s d'Iran. Ont particip6 aux travaux: Mlle S. Baykan (Univ. d'Istanbul) et P. de Miroschedji (A. R. CNRS) archdologues, MM. A. Hesse (C.R. CNRS) et A. Lecaille (ITA CNRS) g6ophysiciens, J. Trichet et P. Poupet (Univ. d'Orl6ans-La

Source) gtologues, Mme J. Pires Ferreira (G. Washington University) palkozoologue, a 6tudi6 la faune des 6 campagnes. Les dessins et photographies d'objets ont 6t6 ex6cut6s respectivement par J. Decroix et A. Rempfer.

178

JOURNAL

OF PERSIAN STUDIES

de la premiere agglomeration (phase Djaffarabad: d6but du Ve millknaire) afin d'en reconnaitre le plan et d'en rep6rerles limites. La superficie de la fouille a 6te portee a 650 m2. En outre, des sondagesont 6t6 ouverts g l'Est du chantier (T. 961 en FG 12-13) et au Sud-Ouest (T. 962 en A 6-7); enfin, au Sud-Est une tranch~e (T. I062 en F-AT 12) recoupant non seulement les couches archeologiques mais aussi geologiques, a livr6 des renseignementssur la topographie du lieu et la nature des sols au moment de la fondation de l'6tablissement. L'analyse stratigraphique de ces niveaux profonds (6-5 b1-2) a montr6 que l'6tablissement s'est forme peu " peu, mais dans un temps qui parait court, par l'adjonction successived'unit6s d'habitation
allongee plus souvent deux, dispos6esen 6querreet plusieurspieces rectangulaireset petites " reserves" de plan carr (I m.-i,io m. de c6t6) (P1. Ib). Plusieurs foyers et un four (P1. IIa) ont t6 d6gages; il est apparu que les premiers sont le plus souvent de plan quadrangulaire et constitues par une semelle de galets couverte d'un lit de torchis tres souvent refait (P1. IIb). Trois s6pulturesd'enfants ont ete mises au jour dont deux dans l'angle de pieces. Un des squelettes 6tait couvert du thorax aux genoux d'ocre rouge devant provenir de la decomposition d'un textile color6 qui, t l'origine, devait envelopper le corps; a proximit6 de son crane une hachette en pierre dure polie a 6te retrouvie. Lors de la phase " Djaffarabad ", l'agglom6ration, dont le plan est tres compact, couvrait un minimum de 1500a 1800 m2. Au Nord et au Sud-Ouest, ses limites semblent acquises. Au Nord s'6tend un vaste espace d6couvert oi se sont accumul6es des couches de debris. Au Sud-Est, la tranch6e 1062 a permis de constater en A 12 la limite des d6p6ts archeologiques. Non loin de celle-ci, J. Trichet a reconnu d'anciens chenaux d'6rosion existant au moment de la premiere occupation du site. La proximit6 du Chaour et d'un vallon profond a l'Ouest ne permettent pas d'envisager de ce c6te une extension de l'6tablissement; vers l'Est, en revanche, I'agglomerationparait s'6tendre.3 Aprbs la phase de Djaffarabad, dans le cours du VWmill6naire, l'occupation du site a 6t6 insignifiante: seul un atelier de potier qui pourrait dater du milieu du Ve mill6naire a 6te mis au jour. Au debut du IVe millknaire,Djaffarabadrenait sur un modble trbsdiff6rent (phase Suse): au plan compact des niveaux anciens se substitue une agglom6rationdont les habitations ceinturent un espace decouvert ouiest creus6 un trbs large puits. Djaffarabad n'est pas abandonn6 ensuite comme nous l'avions pens6, lors de la " phase Suse ". Les vestiges de construction mis au jour en F-G 13 montrent simplement que l'6tablissement du milieu du IVWmill6naire s'est d6plac6 vers 1'Est. La fouille du tep6 Djaffarabad n'a pas apport6 les e16mentsqui auraient permis de mieux comprendre l'6volution culturelle de la phase " Djaffarabad " la phase " Suse ". Une r6ponse devra ^tre recherch'e sur d'autres tep6s du voisinage comme Djowi et Bendebal par exemple qui semblent avoir 6t6 habites durant cette periode interm6diaire. (P1. Ia): chacune comporte au minimum une piece (11,50 m. x3,I5 m.; 6 m. X2 m.), le

L'objectif des travaux conduits cette annee 6tait de degager aussi largement que possible les vestiges

G. DOLLFUS

GanjDareh Tepe
The fourth season of excavations at Ganj Dareh Tepe (Kermanshah District) lasted from 26 May to 18 August I974. The research was funded by the Canada Council, and sponsored by the Universitt de Montreal in collaboration with the Archaeological Centre of Iran. The group consisted of four graduate students from Canadian, American and Danish universities, Dr. F. Ikawa-Smith in charge of the laboratory, and the writer. Mr. Aghil Abedi represented the Archaeological Centre of Iran and his participation in the excavations and in other aspects of the research is greatly appreciated.
En F-G 13 a 6t6 d6gag6 un mur Nord-Sud que nous pensons pouvoir rattacher au niveau 4. Les structures du niveau 4 recouvrant habituellement d'autres plus anciennes, nous pouvons envisager de ce c6t6 une extension de l'agglom6ration. Signalons, en outre, que dans la partie orientale de cette tranch6e ont kt6 mis au jour des vestiges d'une occupation du milieu du IV6 mill6naire.

SURVEY

OF EXCAVATIONS

179

In previousseasons (IranV, VI, VIII, X) we had concentratedon the four principal 8th millennium levels (A, B, C and D) containing brick and mud architecture. In 1974 the work had three main objectives: (I) to expose on the west side of the mound where overlying deposits were thinnest a larger sample of the earliest occupation surfacein Level E; (2) to examine more closely in the central area of the mound the brick and other structuresof Level D and the nature of the interval or transition from E to D; and (3) to excavate a number of test pits outside the mound so as to determine the original extent of occupation under what are now cultivated fields. In the excavations on the west side about 36 sq. m. of Level E occupation surface were cleared. The most interesting features are the " firepits " first reported in the 1971 season. These are oval or circular depressionscut into the virgin soil, sometimesfilled with limestone cobbles and ash, ranging in diameter from about o08o to I *70 m. and up to nearly o*50 m. deep (P1. IIIa). Their functions are not clear but they representthe earliest occupation at this site and apparently precede any attempts at mud or brick architecture. Parts of this occupation were also excavated in the central area of the site as well as in one of the test pits (no. 5) immediately south of the mound. About thirty of these firepits have now been found, and their distribution suggests that originally there had been an extensive occupation comprising perhaps several hundred such depressions. Most of the chipped stone artifacts from Level E are morphologicallyindistinguishablefrom those in the later levels, suggesting that there was no important break in this form of behaviour; in addition, this year for the first time we recovered a single sherd with distinctive punctate impressions,and several clay "Venus " figurines,from this level. Ceramics thus go back even earlier than thought at Ganj Dareh. Interestingly, however, we find no mortars, pestles or polished stone bowl fragments in Level E, and a preliminary study by B. Hesse of Columbia University of the abundant faunal remains collected in I974 indicates that the animals were larger and/or older than in later levels where domesticated goats have been identified. This may reflect a lesser degree of selectivity or of control over the animals in the basal level, and a shift in some of the subsistence patterns in subsequent levels. In the excavations in the central area of the mound, where most of the work had been concentrated in previous seasons, several interesting items were clarified. First, the Level D brick structuresextend furtherdown than we had believed, to within o050 m. of virgin soil at least, thus shortening the vertical distance between E and D. The structuresof Level D were apparently built on a thick accumulation of burnt earth, ash and bones, etc. which had accumulated as the upper part of Level E, above the firepits. Second, many of the Level E walls hitherto thought to be of chineh(packed mud) are now seen, after exposure to the elements for several years, to be made of very long, thick, cigar-shaped or plano-convex bricks. Brick walls are thus as early as, or possibly even earlier than, mud walls at this site. Some of these walls have circular or oval openings, often carefully sealed with clay plugs (P1. IIIb). Third, near the centre of the mound an interesting stone-lined kiln or oven (?) was cleared (P1.IVa). It was filled with grayish powder which is probably burnt limestone, but it also contained a large number of small clay fragments which are probably potsherds. Fourth, re-examination of the bases of some of the very large clay vessels found in situ in Level D shows that contrary to previous impressions, they were originally free-standing rather than integral parts of the walls and floors to which they had been later sealed; that is, they could have been manufacturedand baked outside the structureswhere they are found. Only four test pits (nos. 4, 5, 6 and 7) could be excavated outside the mound this season since the fields were planted with sugar beet rather than wheat and irrigated all summer. The soundingssuggest, nevertheless, that the present mound correspondsapproximately to the extent of original occupation of the site. Inside the mound another test pit carried to several metres below virgin soil revealed a thick gravel concentration containing many chert nodules and fragmentsand a few worked pieces of vaguely Middle Palaeolithic type; this may represent colluvial deposits or gravels of an Upper Pleistocene stream terrace formerly present in the valley. Fourteen human burials were found this season to bring the total now recovered to forty-one. The ages of individuals range from new-borns or foetuses (sometimes buried in specially plastered niches under the floors) to adults. Most of the latter are very tightly flexed. Perhapsthe most interesting is an adolescent burial in Level D with an elaborate necklace of seventy-one shell and stone beads

180

JOURNAL

OF PERSIAN

STUDIES

(P1. IVb); among them are five gastropodstentatively identified as Olivawhich seem to be of marine origin, possibly from the Persian Gulf or, less likely, from the Mediterranean. This is virtually the only indication so far at Ganj Dareh of imported materials, since no obsidian, asphalt or metals have been found here. It is perhaps significant that most of the grave goods at Ganj Dareh seem associated with child or adolescent burials. Near one of the adult burials there were impressionsof matting or coarse textile which may have covered or underlain the body. Some experimentswere carried out during the field season on the production of gloss during cutting of grains with flint blades, and with the making and sun-bakingof plano-convex brickslike those found in Level D. We also attempted to duplicate on these experimental bricks, at several stages of sunbaking, the animal hoofprintsfound on the prehistoricbricks, using local goats and sheep of both sexes and of various sizes. Ethnographicinformationwas also collected from the nearby villagers on modern and brick walls), seasonality of constructechniques of house construction and repairs (for both chineh tion for each type, and the factors involved in damage and destruction. This may help us to understand better the very complex suite of repairsand rebuilding found throughout the prehistoricsequence at Ganj Dareh.
PHILIPE. L. SMITH

Ghubayra The third season of excavations at Ghubayra took place from July 7 to August 25, 1974. We continued to work on the Citadel mound and also cleared buildings at Sites F and E. On the western slope of the Central Citadel Mound we cleared Room I8 and found Islamic pottery and bronze objects. A passage was found behind the chamber which led on the left to the tunnel system, Pit A, and on the right to another room under the Central Mound called Room 2o. Further south, on the Citadel Platform, the removal of fallen walls revealed a shaft which gave access to a cellar, Room 19 (P1.Va). The interior of Room 19 had its plaster coating preserved in situ, so helping to explain the many plaster fragments found in other chambers. South of the Citadel Room 17 of the Palace was excavated. On floor level at its southern end we cleared two pits. An anonymous bronze coin, probably Muzaffarid, was found in the western pit. A similar coin was found below floor level in the centre of the room. It may have been out of position due to late disturbances. Beneath the northern end of the room was a large cave, with traces of the usual elliptical accessshaft. Whether or not the cave had existed before the constructionof the building, it seems to have been re-occupied after its abandonment, since there were traces of pastoral occupation with an anonymous bronze coin dated 795 A.H., the exact year of the defeat (e.g. a large stock of keshk), of Sh~ihManstir, the last Muzaffarid Sultan, at the hands of Timmir.This was the first precisely dated coin to be recovered. Further south, beyond the area of the Palace, work was resumed in " Tunnel F ", where a hoard of 14 Sasanian silver coins had been discoveredin 1972. The tunnel and chamber were cleared, and an exit staircaseon the other side traced up to the surface. In this tunnel system a small pear-shapedgold pendant or cloison (P1. Vd) was found, with a quantity of slip-painted pottery and blue-glazed ware. Within the core of the Palace complex, north-west of Tower no. 5, a well-preserved chamber, Room 21, was cleared. Here a number of hexagonal turquoise, and astragaloid cobalt wall-tiles were
uncovered with some fragments of lustre-ware, and quantities of underglaze-painted pottery. These, like the finds in many other parts of the site, suggested a Seljiq structure with some post-Seljiaq occupation. A remarkable find, however, in a cavity at the northern end of the chamber, was an unengraved but typically Sasanian conoid sealstone shaped from onyx. At the northern end of Room 21 was exposed a doorway, leading to another small chamber, Room 22, in which could be seen some large fragments of glazed jars. The lower walls of Room 21 had been carved from the natural tafk-sandstone, but higher up the foundations had been levelled off with a fill of rubble, which thus formed the arch over the entry to Room 22, and provided a foundation for the flimsy floor-level further north, which was evidently that of an upper storey (P1. Vc). Below Room 22 there descended a pit nearly 5 m. deep, in which were discovered the most important finds of the season. These included a complete facetted glass bottle and part of a wooden casket, lacquered in red

SURVEY

OF EXCAVATIONS

181

and black, with incised scroll decoration (Pl. Vb). The latter object is probably Ilkhdnid Persian rather than of Chinese manufacture, and is likely to be the oldest surviving example of lacquer-work so far known from Iran. There were also fragments of two interesting lustre vessels and several monochrome-glazed sherds. Site F is to the east of the Citadel and consistsof a large building, formerlycalled " The Gatehouse " (Pl. VIa and b). The ground-plan and architectural details suggest that it may have been some kind of audience-chamber or perhaps a meeting-hall (majlis). On the western side of the main room, raised to a higher level than the adjoining or facing the entrance,was revealeda small alcove, exedra divdn, room and with a plastered floor. This exedra was entered through an arch, under which was a semicircular step up to a platform. A large number of glazed tiles were found outside the walls; they had probably once decorated the outer walls. Some were in monochrome turquoise, with fragments of arabesque decoration and naskhiinscriptions; others in lustre-painted glaze. Along its southern wall, certain secondary partitions of baked brick and mortar were located, in association with Safavid blueand-white, and some underglaze-painted pottery. Further to the west stood a baker's oven and a coppersmith'sfurnace. Excavation at this site established three periods of occupation: (i) The walls associated with the elaborate roofing system which may indicate a post-Seljiiq or Ilkhanid building-period. (2) A secondary floor-level in the hall, with baked bricks, together with the baker's oven and furnace, which probably dates from the late Muzaffaridperiod,just beforeTimiir's invasion in A.D. 1393. (3) The brick partitions with later pottery which may represent a late Safavid or even early Qajar settlement of limited extent. Site E lies several hundred metres east of the Citadel. It consistsof the ruins of an hexagonal shrine or tomb-chamber called by the excavators " The Imamzadeh " (Pl. VIc), though naturally the exact function of the building is unknown. The hexagonal base survives with its six angled piers. Indeed the lower part of the zone of transition is still extant at some points, as well as the arch between piers no. I and 2, although the dome itself had collapsed. The interior was cleared down to floor-level. Close to the corner of pier no. i a fine lamp of Seljfiq white-glazed pottery was uncovered in mint condition. At the exterior beyond pier no. 3, a mud-brick wall was encountered running from north to south. One side of the trench was extended to determine its purpose. A baked brick structurewas found which appeared to be the principal burial. When the loose soil was cleared from below, the edge of a wooden board was exposed. It was recognised as part of a wooden coffin, a feature which local reports claim to be typical of burials from the Seljiq period. The presence of such a tomb confirms the initial surmise that the building was indeed a funeral-monument,rather than merely a cenotaph. There are some marked similaritiesto the tomb-tower of Shdh Firiz at Sirjdn,which itself is of arguable date but both buildings may have been constructed before the Mongol invasion. A. D. H. BIVAR and G. FEHERVARI Hamadan The excavations of the Iranian Centre for Archaeological Research at the Parthian cemetery at
Hamadan lasted for two months from June 15 to August I5, I974. The cemetery lies on the southeastern edge of the modern city, in the vicinity of the renowned Lion of Hamadan, and it was already the scene of rescue excavations that were conducted by Mr. A. A. Sarfaraz in 1970. This year's work exposed fifteen baked clay "slipper coffins " in which the dead lay on their backs in an extended position. The coffins were orientated on a north-east-south-west axis with the heads pointing either to the north-east or the south-west. As a rule the coffins were sealed by stone slabs although some of them were sealed by a combination of stone slabs and mud bricks. In one case only an individual was found to have been buried, lying on his back, in a makeshift coffin composed of four large jars. The present work also revealed a number of skeletons which lay in simple earth graves in either flexed or extended positions. The heads of the flexed skeletons pointed towards the west or the north-

182

JOURNAL

OF PERSIAN

STUDIES

west while the heads of the extended skeletons pointed towards the north-east. In one instance a clay coffin was found to lie above an example of the latter type of grave-and this may be an indication that extended burials in clay coffins succeeded extended burials in earth graves. The date of the cemetery appears to fall within the ist century B.c. and the ist century A.D. Excavated objects include two sherds with red painted designs on a buff slip and a pair of gold earrings that probably date from the first century B.C. The earringspossesshandsome lion-head finials complemented on the back by a representationof a human face. A Parthian coin of GudarzesII (c. 38-5I A.D.) was also found not far from a flexed skeleton, a circumstancewhich suggests a date in the Ist century A.D. for this third type of burial. The cemetery was reused by the Islamic inhabitants of Hamadan in the I3th and 14th centuries A.D.
MASSOUD AZARNOUSH

Hasanlu The 12th season of excavations by the Hasanlu Project was carried out during July and August 1974 at Tepe Hasanlu (markingthe 9th season of work at the central site). The Hasanlu Project of the University Museum, the University of Pennsylvania,was carriedout with the cooperationof the Iranian Centre for ArchaeologicalResearch (representedon the site by Mr. Mansur Sadjadi) and with financial support from The Metropolitan Museum of Art, the Kevorkian Fund, the Ford Foundation, and the Museum of Fine Arts, Boston. The season had the four following objectives: (I) collection by flotation methods of palaeobotanical remains for all periods from X (Hajji Firuz) through IIIA (Achaemenian) to provide a sequence of
ecological and agricultural remains from c. 6000 B.C. to c. 300 B.C. (see Tosi report below); (2) limited further exploration into strata of periods VII (c. 2400 B.C. )and V (c. 1350-1100 B.C.); (3) extension

of clearances in Burned Buildings V, IV-V, and IVE in the palace area and exploration of the wall and road system on the west slope of period IV (c. I 100-800 B.C.); (4) further re-examination of the stratigraphic position and date of the main fortification wall and ceramic assemblages relating to PeriodVII (Painted Orange Ware). A small test area in U22 was carried down through trash strata with occasional small brick walls to a depth of 3 -8o m. below the bottom of the period VI deposit. The slope of the various strata suggest that the area was not far from the edge of the mound at the time. The upper strata contained sherds of typical Painted Orange Ware from globular shaped vessels with short necks and out-turned rims. Designs were of cross-hatchedlozenges in a band around the shoulder of the vessel bordered below with a wavy line between parallel straight lines. This standard variety may be tentatively assignedto ceramic phase VIIA. Below this lay several strata in which there was a change in shape to a more hole-mouth jar type with similar painted cross-hatchedlozenges but now flanked by a row of birds beneath. This phase may tentatively be assigned to VIIB. Below this the Painted Orange Ware disappearedin the test area and several sherdsof a related but differentware appeared in the form of small bowls with painted designs consisting of parallel vertical lines interspaced with circleswith a dot in the centre. This ware is not yet of Pisdeli type and must belong to the earlier centuries of the third millennium B.C. if not to the late fourth (Phase VIIC). No evidence was forthcoming for the presence of black wares of Kuro-Araxes (Geoy K or Yanik EB) type. The test failed in its purpose to document the transition from VII to VIII on the Citadel mound.
Period VI (Dinkha Painted Ware). Remains of this period were accidentally encountered immediately west of the central section of Fortification Wall II (the large stone wall surrounding the citadel) at its base. Evidently a wall system of major proportions existed suggesting the probability that a period VI fortification wall exists around the earlier citadel mound. Remains of this period already explored in U22 on the east side of Fortification Wall II (and encountered in a small test trench in V22) lie at a considerably higher level indicating a sharp drop in the contour of the mound at this point in period VI along the west slope. This contour was an important factor in the locating of the period IV structures in the area which in turn influenced the line of Fortification Wall II. Period V (Iron I Grey Ware). The area around the house of period V found in 1972 was enlarged to the east and north with the result that the plan was completed with the addition of the end walls of periods IIIB (c. 750-600 B.C.) and IIIA (c. 6oo-?c. 300 B.C.).

SURVEY

OF EXCAVATIONS

183

the ante-room lying immediately east of the columned hall, and the recovery of the line of the eastern and northern side walls of a long narrow storeroom which flanked the columned hall on the north. These additions to the plan confirmexpectations that the structurewas in its essentialdetails comparable to B.B. III of period IV. Three stoneswhich may be the remains of a wall setting off a stairway room at the south end of the anteroom are all that is left of the robbed structureto suggest that it may also have had a stairway to its roof or upper floor. Immediately overlying the north-east corner of the period V building in a transitional V-IV structure below the burned IVB level was found just such a standard spiral stairway with central column and brick steps. Period (Iron II Grey Ware). The excavated area of B.B. V was extended eastwardwith the result IV that the entire plan of the building has now been recovered except for the south-east corner of the south storeroom. As expected the door from the anteroom to the columned hall appeared in the north wall of the latter room offset to the east from the main door in the north wall of the anteroom. The east wall of the columned hall was unique in that it consisted of a central section in which the foundation was only about half as high as normal in the rest of the building. On this foundation stood four pairs of wooden posts. Between the first and second, and the second and third pair from the south stood a poorly built curtain wall of brick considerably less thick than the normal wall. The thinness of the wall and its poor constructionindicate that it is secondary to the original constructionwhich must have had an open fagade. The opening appears to have centred on the brick paving between the two eastern central columns, behind which stood the raised central hearth. The closing of this opening (at the end of IVC perhaps) left only a narrow door opening through the wall to the east. Along the eastern fagade were additional room areas formed by walls of structuresbuilt just to the east but not original to the building. These areascontainedjars and hearthsand appear to have served as storage and kitchen areas. Small objects from B.B. V included bits of ivory, beads, etc., fallen into the south-east corner of the columned hall and adjacent room. The items included half a stone macehead with a single line of cuneiform reading " Palace of Ashur-Uballit " (the reading kindly supplied by Dr. Erle Leichty of the University Museum). The object, like others found in period IV, would seem to be an heirloom surviving from period V as it no doubt refers to the first Assyrian king of that name who ruled in the fourteenth century B.c. The position of this material, stratified in the second floor collapse, in this particular corner is of special interest in that the comparable but much richer material found in B.B. II came from the same relative location-indicating that in buildings with this general plan the function of the left rear rooms on the second floor must have been in some way similar. The corridorbuilding (B.B. IV-V), constructed at a later date to fill the space between B.B. V and B.B. IVE, was excavated further to the east. As in 1972 the collapsed second floor level yielded a mass of smashedpithoi and burned textile fragmentsas well as a great quantity of bronze horse gear including about ninety bronze bells of various sizes of the rattle type, plaque pendants, tube pendants, bits and assorted other items. Most important was a decorated bronze breastplate bearing the figure of the helmeted storm god holding two bulls by the forelegs and flanked over the shoulders by two birds of prey. The heads of three figures are in the form of protomes. The dark corridorwhich apparently lead to rooms further east was illuminated by lamps set in niches along the walls. A stairway leading to the second floor level was preserved almost to that level and revealed a small brick platform at the top which had been supported by a corbelled arrangement of the brick.
In B.B. IVE excavations were carried out east of the previously excavated portion of the columned hall in order to determine whether or not this hall conformed to the standard plan of the other period IV buildings. Approximately half of the building has been uncovered leaving the eastern end still unexcavated. To the I972 plan were added three stone column bases aligned with the previously uncovered rows. Along the north wall of the hall a plastered bench was found to continue. Centrally located within the hall a large plastered brick hearth was found. These finds confirmed the plan as similar to that of the other columned halls of this period at Hasanlu. Excavation of the floor of B.B. IVE revealed that this building had also been burned at the end of period IVC, then rebuilt, and then burned again at the end of IVB (c. 800 B.c.). Most of the building's brickwork had been reconstructed after this earlier fire. The entire north wall and bench were new in both foundation and brickwork, and even the column bases had been replaced. Excavation of the IVB floor of the building produced

184

JOURNAL

OF PERSIAN

STUDIES

equestrian trappings predominantly of bronze with several pieces of iron. Distinct from artifactsfound on the floor of the building were those found in the second storey collapse. These consisted mostly of decorative items such as beads of stone, bronze, carnelian, shell, frit, buttons of polished and decorated wood and bosses of bronze. Excavations parallel to, and east of, Fortification Wall II on the west slope of the citadel mound revealed the northern and southern ends of a long structure (The " North Building ") excavated partially in earlier seasons. At the southern end was found a passageway room designed for pedestrian passage from west to east through a single room with a door at either end. On its south side this structurehad been built against an older building which consistedof a columned hall with benches and paved area. This structure,B.B. VI, was cut away on a diagonal line running north-west to south-east by the foundation trench of Fortification Wall II and its adjoining walls. The eastern fa?ade of B.B. VI was enclosed by an extension of the pilastered east wall of the North Building and Passageway structure to make the whole appear as a single unit when viewed from outside. Both structures contained bronze and iron artifacts and several skeletons while other skeletons lay on the surface outside the eastern walls. B.B. VI had been burned previously as indicated by an older floor. This discovery confirms the evidence found in other buildings and in the wall system of the western slope that an earlier partial destruction had occurred which led to rebuilding and modification of various structures on the citadel. This event may be assigned to the end of period IV C, with the major destruction at the end of IVB (c. 800 B.c.) and a final burning of the remains of squattersat the end of IVA (early eighth century). The Passagewaystructureand B.B. VI capped the rise formed by the older period V and VI debris at this point. Immediately south against the south wall of B.B. VI and set into the lower slope of this rise was the north-eastern quarter of the Upper Gate of period IV. Its remains consisted of a single guardroom with pilastered fagade and one preservedjamb on the north side of the east door. The jamb was faced by a long narrow stone slab in which had stood a pair of round wooden columns (indicated by burned out holes with charcoal). The base of the columns was buried by about a half metre of accumulated clay and pebbles raising the surface in the gate to its position at the time of the IVB fire. The slope of this surface changed sharply as it descended into the gate from the east. At this point a square-cut beam was set into the surface to prevent erosion. A similar system was seen in structuresnext to B.B. III in I962. From the Upper Gate a pebble surface (part of which was paved with small cobblestones) led east almost in a straight line toward the gate to the Lower Court of the Palace area. A matching stone slab for a second pair of wooden columns was found in position opposite the existing pair although all the rest of the structurehad been cut away by FortificationWall II. It is, however, apparent that the guard-room on the north was repeated by one on the south making a gate plan closely resemblingthe one already known leading into the Lower Court. It is through this entrance that horses must have entered and left the central area of the Citadel. The Passageway structure higher up was apparently designed specifically for foot traffic. Taken all together, the north-south row of buildings (North Buildings, Passageway, B.B. VI, Upper Gate) formed a " wall " which separated the Palace area from the approach system on the western slope. Excavations on the west slope have now revealed remains of all three phases of period IV. In IVC the slope was enclosed on the west initially by a narrow wall (90 cm. wide) decorated by evenly
positioned pilasters and two small towers. At a slightly later time the upper half of this wall (which stretches for about 80 m. downhill from Tower 6 of Fortification Wall II) was strengthened by enclosing

it in a second wall along the west. At the end of phase IVC with the first destruction this wall system
went partially out of use. In phase IVB the upper half of the wall remained standing and the enclosing wall to the west was rebuilt. Both of these walls ended at the Western Enclosure, a rectangular enclosed area which was roofed on the northern side by a shed roof. From the Enclosure downhill ran two parallel roads. The earlier pilastered wall on the west was now replaced with a plain wall running immediately above. The eastern face of this wall supported a brick bench or footing set on a stone pavement edged by a stone drain. This arrangement was repeated on the east side of the West Road: drain, pavement, bench, wall. The entire system was then repeated again for the East Road beginning with brick bench, pavement, drain, road, drain, pavement, bench and wall. The double road system

SURVEY

OF EXCAVATIONS

185

ran from the foot of the slope over 40 m. up to the rectangular enclosure. The east end of the Western Enclosureopened eastward under the later FortificationWall II just opposite the Passagewaystructure which lay further east. Somewhere partway up the slope the East Road may have opened eastwardto allow access to the Upper Gate. Unfortunately time did not allow clearance of the two squarescovering this area. Nevertheless, the basic plan of the ninth century entrance is clear. Following the destruction of IVB the survivors abandoned the elaborate constructions, robbed much of the stone from the walls and built a simplified system retaining the basic plan with a single wall running down the upper slope to a wall which ran across the back of what had been the Western Enclosure. A further wall ran down the western edge of the slope enclosing the former road area. The interior of this area was covered with clean soil possibly as a work area. It was finally burned over at the end of IVA. PeriodIII. After a period sufficient for a hard erosion surface to form the Urartian occupation of IIIB took place with the construction of the garrison plan of Fortification Wall II and associated rooms. Furtherinvestigation of the wall during the current seasonin the areas now reached by excavation at the north and south edges of the Citadel Mound demonstrates conclusively through stratigraphic evidence that the entire wall belongs to the IIIB period and not just a section of it along the west side as previously hypothesized. An extension of the area of B.B. III to the north proved that remains of period IV ran under the wall at that point. On the south side the foundation trench discovered in 1972 behind B.B. II may be seen to tie the Fortification Wall to period IIIB as well. The round plan of the wall with towers (reached by a series of sondages in 1964) projecting at intervals and with strengthening piers between with only one small gate reflects the architectural practices of Urartian sites of 8th-7th century date recently surveyed by Dr. Kleiss in northern Azerbaijan. This was pointed out to me by Dr. Kroll in 1972 while discussingthe problems involved in firmly dating the wall. The associated rooms along the inside of the wall evidence a main building stage contemporary with it followed by secondary changes. Occupation following the IIIB period appears to have been largely continuous with walls going out of use and others being built in different areas. At some point in time a large slab of brickworkfrom the inside face of the Fortification Wall fell over-both on the west side of the mound and on the south (where it formed part of the base for the large building of period II). During this period, or in period II, long after FortificationWall II had fallen into disuse, FortificationWall I was constructed around the Citadel on a line outside that formed by the older wall. The ceramics appear to indicate a slow evolution of forms and painted types through these levels all of which may be classed as IIIA ceramically as they are broadly consistent over the excavated areas and date essentially to the Achaemenian period or shortly thereafter. Two glazed Hellenistic sherds suggest that the terminal date for the sequence was just at the beginning of this period. ROBERT H. DYsON, JR. and VINCENT C. PIGOTT
Hasanlu Project 1974: PalaeobotanicalSurvey

A general surveyof microscopicremainswas conducted on about Ioo samplesof soil by means of dry sieving and water separation. The samples were taken from strata assigned to the ten periods of the
archaeological sequence as follows: from Tepe Hasanlu, periods III-VII (Ist-3rd millennia B.c.); from Pisdeli Tepe, period VIII (4th millennium B.C.); and Hajji Firuz Tepe, periods VIII-X (4th6th millennia B.C.). The collection thus provides data covering a span of over 5000 years. A simple device was built close to the expedition house to provide running water, otherwise not available at the house. The methods used were with minor variations those successfully employed at Shahr-i Sokhta in 1972. Sieve meshes measuring I and 2 mm. were used to guarantee a large spectrum of finds. The method excluded all mechanical devices as they usually decrease the chances of recovering delicate items such as insects, etc. Apart from four large hoards of wheat and barley grains and lentils, about 2500 intact seeds were recovered. Others are still to be found in the unchecked samples of soil which will be processed under safer conditions at the botanical laboratory in Rome.

186

JOURNAL

OF PERSIAN

STUDIES

The finds also included fish bones, mice and bird bones, and a good number of egg shell fragments and microscopic fresh water snails. This data will provide major information on the environment as such as well as on the degree of food-gathering in the local subsistence pattern from 6000 to 300 B.C. The most important crop throughout was barley, the two-rowed type being progressivelyreplaced would appear on preliminary examination) during the 3rd and 2nd millennia B.c. by the six-rowed (it type. Both emmer and bread wheat were present, the latter being more abundant from the 3rd millennium onwards. Rye appears at the end of the 2nd millennium (Hasanlu V) and appears to have been a staple crop throughout the Iron Age. This fact is not surprising considering the apparent importance of horsesin that period. In earlier periods (VI-IX) goat-faced grass was a possible alternative crop occurringpartly together with barley in the field. A single charredgrain of rice was found in a pit datable to period III. Pulses were intensively cultivated and appear to have represented about 15 per cent of the total harvest in period IV. They included lentils, chickpea, horsebean, and bitter vetch. The pattern of cereal and pulse cultivation in the Iron Age at Hasanlu was closely similar to that presented by the palaeobotanical remainsfrom Nimrud as described by Dr. Helbaek. Marked distinctions are, however, to be found among the oil-bearing plants and fruit. Whilst at Nimrud flax was a prevailing crop, at Hasanlu it appears to have been very rare; on the other hand, black poppy was well represented at Hasanlu in period IVA. Fruit found in Hasanlu IV included local products such as grapes, apples or pears and products brought in from the surroundingmountain slopes such as pistachios and almonds. Farming appears to have been a highly diversified and specialized activity in periods V and IV with little difference from the pattern of present day agriculture in the Solduz valley.
MAURIZIO TosI

Malyan4 The I974 work was carried out in three locations: the Middle Elamite building, the ABC stratigraphic trench, and the small Banesh mound in the north-east corner of the site. The MiddleElamiteBuilding.An additional 6oo sq. m. were excavated in this large building. It is laid out around a rectangular courtyard surroundedby a corridor giving access to a number of small rooms. Finds include cuneiform tablets, glazed tiles and knobs, stone pegs, and a massive deposit of flint flakes and debitage, as well as animal bone and grain. The ABC Stratigraphic Excavation was continued in Building Levels II through V, all of Operation. the Banesh period. A large cache of sealings, along with several Proto-Elamite tablets came from an early phase of Level II. Building Level III proved to be a fine structureof some sixteen rooms with red or white painted walls and more than twenty fragments of polychrome fresco. This building also contained sherds of beautifully executed relief pottery showing cows, bulls, and goats as well as ProtoElamite tablets and sealings. Building Level IV included a well made small building oriented as in Level III. Abutting this building were a series of rooms, apparently added later, in which the walls are curved and irregular. Level IV also produced Proto-Elamite tablets and sealings as well as a trash deposit outside the building and a pit inside with the best collections of Baneshpottery and animal bone found in the ABC operation to date. Level V was investigated in a small sounding. The SmallBaneshMound. Some 800oo m. were opened here and Building Levels I and II were sq.
extensively investigated. Level I was a badly preserved structure of uncertain function. Level II includes one very large room and several complexes of small rooms, some with hearths, storage jars, and grinding installations suggesting domestic functions. In addition to a large quantity of pottery and animal bone this level yielded several Proto-Elamite tablets. Level III, investigated in several small soundings, is of special interest because it appears to be stylistically distinct from other Banesh levels investigated to date.
* Sponsored by The University Museum, University of Pennsylvania with additional support from The Metropolitan Museum of Art, The University of Michigan, The Ohio State University, The Boston Museum of Fine Arts, and The University of Oregon.

SURVEY

OF EXCAVATIONS

187

We were fortunate this year in having a Paleoethnobotanist and a Paleoethnozoologist, both from the University of Michigan, who initiated studies of the ancient environment and particularly the ancient subsistence technology with its many general cultural implications. We were also pleased to have an Epigrapher on the staff for the first time. Finally, I must mention with sadness and regret that Michael Nimtz, our friend and colleague during all three seasonsof work at Malyan, died in a tragic accident at Mehrabad airport at the end of the 1974 season. His steadfastnessand his smile will be sorely missed.
WILLIAM M. SUMNER

Siraf
The seventh and final season of the British Institute of Persian Studies' excavations at Siraf took place in October and November 1973. During the season we worked at Sites H, O and Q and carried out simple conservation of the mosque and Sasanian fort at Site B. At Site H, we investigated the buildings which underlie the mausoleum described in IranX (1972), p. 82. The structures,which were difficult to interpret, may have had a religiousfunction. At Site O, we excavated two rock-cut chamber tombs and excavated other structures in the monumental cemetery described in Iran XII (1974), pp. 23-30. Each tomb contained several rectangulargraves. All but one of these had a single human skeleton. The exceptional grave, whichwas undisturbed, contained no human remains but several bones, including the mandible, of an equid. At Site Q, the citadel, which occupies a hilltop position overlooking Taheri village, we made intensive surface collections. The material suggested strongly that the citadel was garrisonedin the Sasanian period, perhaps abandoned in the period of Siraf's great prosperity (c. 8oo-Iooo A.D., or later) and re-occupied in the eleventh or twelfth century.
DAVID WHITEHOUSE

TepeNush-iJan Following a fourth campaign in the summer of 1974, the excavated plan of the Median settlement includes, from east to west, the Fort (Pl. VIIIa), the Central Temple, the Columned Hall and the Western Temple, formerly called the Old Western Building. Additional structures include an ovalshaped perimeter wall, storage magazines and a rock-cut tunnel that may have led to a deep well. At the summit of the mound the last of the shale fill was removed from the stepped, triangular sanctuary of the Central Temple (P1.VIIa). Further to the west the removal of the Parthian levels has exposed all four walls of the Columned Hall. Two recessed niches in the east wall complement the opposed sets of four niches in the long north and south walls and, as our earlier reconstructionshad anticipated, three rows of four columns supported the roof. The original columns rested on flat stones set in the mud brick floor and had low, circular column bases of mud brick and plaster that were approximately 30 cm. high and I -05 m. in diameter. The wooden shafts of the columns measured 25 cm. across at the base. The excavation of the western Temple proves that this structurehad a much lower floor level than any of the other main buildings. The Temple was entered from the east via a short passage that gives onto an oblong antechamber. As in the plan of the Central Temple, the antechamber affords access
to both the main sanctuary and a steep ramp which leads to a slightly larger chamber above. The the only sanctuary itself is rectangular with a north-south axis and a maximum length of 8.85 m.; door occupies the centre of the long east wall (P1. VIIb). The room appears to have risen to the full height of the building save at the southern end where we meet an unusual feature: a relatively low vault, 2 40 m. high and 2 60 m. wide, which arches over the truncated remains of a square mud brick altar. The east facade of the Western Temple was apparently remodelled on at least two occasions. Indeed, with the construction of the Columned Hall the original low-level entrance of the building was blocked off completely and the Median architects had to cut a new, sloping passage that led down from the floor of the Columned Hall to a bend in the Temple ramp.

188

JOURNAL

OF PERSIAN

STUDIES

At the time of writing it is still not known if there was more than one point of entry through the perimeter wall. We know that a small " east gate " allowed access to the Fort towards the end of the life of that building, but it is tempting to look elsewhere for the main approach. The most likely point for a principal gate lies directly opposite the entrance to the Central Temple. Just to the east of this still unexplored point we know of the existence of four monumental arches with rounded and pointed vaults, each over 3 m. in height (P1.VIIIb), which could have served to flank one side of such a gate. Surface indications at the backs of these arches suggest that they were themselves closed, at least originally. In certain ways the ceremonial manner in which the Median settlement was closed-one might almost say entombed-is the most remarkable aspect of the site. The door to the sanctuary of the Central Temple was blocked with bricks, the main body of the Temple was filled with shale to a height of over 6 m. and the whole top of the building was capped with a mud brick platform. Outside the limits of the temple massive quantities of mud brick are known to have been introduced on all sides, while further areas of mud brick blocking have been encountered in the South Court, in South Street, in the East Court, in one of the magazines of the Fort, and in the narrow corridor that runs between the Western Temple and an adjacent part of the perimeter wall. In the Western Temple, moreover, the door to the sanctuary was blocked with mud bricksup to a height of I m.; the mud brick altar, which may have been in poor repair by this time, was largely dismantled; and the whole floor of the sanctuary was given a coating of thick white plaster. Late in the 7th century B.C. the Columned Hall was occupied by " squatters " who divided much of the floor space into a series of small rooms, each approximately 4 X 4 m. in size. Dozens of pottery vessels, several bronze fibulae, a stamp seal and a variety of bone objects all come from these rooms. The reason why the Columned Hall came to be used in this way remains uncertain; one theory is that these last inhabitants of the Median mound were members of the work force that was employed in filling other parts of the site, conceivably in preparation for the erection of an entirely new religious complex which would have stood above the original buildings. Half-way between the two temples the past season's work revealed the entrance to a monumental rock-cut tunnel, the first 18 m. of which proved to be empty. The tunnel is 1.70 m. high and I . 80 m. wide and descends towards the west at an angle of about 30 degrees. It has shallow rock-cut steps. Preliminarystudies suggest that the tunnel was cut before the construction of the Columned Hall and that its entrance was later incorporated in that building. When the Parthians reoccupied the site c. 100oo they made the best use they could of the rather B.C. awkward site available to them. In places they incorporated parts of the Median walls in their own walls and in other places they cut through the Median buildings to create level terraces. Above the Columned Hall they erected a series of small rooms many of which open off a straight, central corridor. The plan is exceptional in that we find traces of twenty or more rooms which appear to have stood under one roof rather than a series of individual dwellings. Pottery from the site includes glazed ware, clinky ware, buff-coloured common ware and painted red-on-buff festoon ware. Two months after the close of the season a large steel roof, 24 x 28 m. in area, was erected over the well preservedmud brick walls of the Fort. The cost of the roof was generously borne by the National Organization for the Restoration of Historical Monuments.
DAVID STRONACH

Takht-i Sulaiman Bei den Grabungen des Deutschen Archaiologischen Instituts auf dem Takht-i Sulaiman wurde I974 von Mitte Juli bis Mitte September an 6 Stellen mit insgesamt i60 Arbeitern gearbeitet. Ein wichtiger Teil der Arbeit galt der Vorbereitung fiir die von der persischen Altertiimerverwaltung geplanten Restaurierungen. Im Westiwan wurden die Schuttmassen bis zum Niveau des mongolischen Palastes entfernt und dann ein Stiitzgeriist an der Iwanmauer ausgefiihrt. Der noch 2 m. tiefer liegende sasanidische Fuss-

SURVEY

OF EXCAVATIONS

189

boden kann erst nach Sicherung der Nordmauer des Iwans freigelegt werden. Auf beiden Saalw~inden sind noch grosse Teile des Wandstucks mit den Abdriicken der Fliesenverkleidungmongolischer Zeit dieser Verkleidungkonnte versuchsweisewiederhergestelltwerden. In dem siidlich erhalten. Ein StUick neben dem Hauptiwan liegenden Korridor wurden viele Hunderte von Bruchstiickender Wandverkleidung der Aussenmauerndes suidlichenOktogons gefunden. In den Raiumen P und Q des Gebiiudekomplexesum E wurden Nachuntersuchungen ausgeffihrt mit dem Ziel, die Frage der urspriinglichen Benutzung dieses raumreichen Gebaiudeszu kliiren. mit T6pferbfen in In Raum P, in dessen oberen FUillschutt friiherenJahren bedeutende Werkstaitten islamischerZeit gefunden wurden, lag der sasanidischeFussboden0,40 m. tiefer als im Achtpfeilerraum E und bestand aus einer Lage Gips fiber einem Grund von grossen Steinplatten. Auf diesem Boden standen drei Einbauten aus gebrannten Ziegeln mit gestuftem Aufbau. Diese dienten offensichtlich als Plattformen ffir religi6se Objekte oder Altiire; sie erinnern an die 2-Saiulenmonumentevon Bishapur und die dreistufigen Basen der Fratadara-Tempel bei Persepolis. Eine spaiter mit Ziegeln zugesetzte Offnung in der Ostmauer von P scheint urspriinglichdurch einen Kanal mit dem Durchbruch in dem mittleren Ostpfeiler des Raumes E verbunden gewesen zu sein. Nord6stlich von P, innerhalb des bstlichen Gew6lbegangs, wurde ein sehr gut erhaltenes Gebiiude mit kreuzf6rmigen Mittelraum und Wirtschaftsraiumenund Werkstaittenin den Eckraiumenaus il-khanidischerZeit aufgedeckt. Es ist eine schon in abbassidischerZeit fibliche Hausform, die auf dem Takht-i Sulaiman bei 6 Gebaiudenangetroffenwurde und wohl auf die kreuzf6rmigenRiiume sasanidischer Zeit zurtickgeht. Im SO-Raum dieses Gebaiudes befand sich eine gemauerte Bank mit zwei Nischen, in denen zahlreiche, unfertige Fliesen, teils zum ersten Mal gebrannt, teils schon mit Glasuren versehen, gefunden wurden, die spiiter hier eingemauert waren. Zu einem der Fliesentypen wurde an anderer Stelle auch die Pressformaus rotem Sandstein gefunden. Beide Tatsachen beweisen, dass alle diese Fliesentypen in den Werkstiitten auf dem Takht-i Sulaiman hergestellt worden sind. GegenUiberdieser Bank befindet sich in diesem Raum ein Herd und beiderseitsaufgemauerteDoppelschrainke, erwaihnten deren Stirnwaindemit Leisten aus Fayencemosaik verziert waren. Diese Art der FayencemosaikDekoration ist sonst auf dem Takht noch nicht angetroffenworden und dies kdnnte darauf hindeuten, dass dieses Werkstattgebiiudenoch bis zur Jahrhundertwende in Benutzung war. Unter dem Kreuzraumhaus lag ein raumreiches sansanidischesGebiiude mit mehreren Tiiren zu dem BstlichenGew6lbegang, das sehr wahrscheinlich zweigeschossiggewesen ist. Noch tiefer wurden Mauern aus Lehmziegeln auf Bruchsteinfundamentenangetroffen, wie auf der Westseite des TakhtPlateaus, so dass in friihsasanidischerZeit innerhalb der grossen Lehmziegelumwallung eine sehr ausgedehnte Bebauung mit bedeutenden Anlagen angenommen werden kann. Zwischen dem Tempeltor N und dem Tor im Gewdlbegang in der Achse der gesamten Anlage wurde die im Vorjahr untersuchte Flache nach Westen erheblich vergrissert und der Befund im spaten Siedlungen geh6rige Strassennetz mit ausgedehnten und oft erneuerten Kanalanlagen gekliart werden. N6rdlich des westlichen, zweiten Feuertempelswurde die im Vorjahre begonnene Grabungsflaiche bis zur Sinteroberfliache freigelegt. Dieses Gebiet war nie von sasanidischen Steinbauten eingenommen. Dagegen wurden unter mageren Schichten des 9.-1 i. Jahrhunderts drei Hauptphasen von Lehmziegelbauten gefunden. Zuoberst lagen die weitgehend zerst6rten Grundmauern eines Gebiaudes mit einer Nische an der inneren Stidseite. Diese Grundmauern waren auf einer Lehmziegelterrasse errichtet, die den abfallenden Grund ausglich. Gegen diese Terrasse lehnte sich ein rechteckiges Haus mit starkem Steinsockel. Als tiefste Anlage wurde ein sehr starkes Kieselsteinfundament auf dem Sinter gefunden, dessen Ziegelaufbau jedoch vollstaindig ausgeraubt war. An besonderen Funden ist eine sasanidische Tonbulle mit drei Abdruicken zu erwiahnen, die ein geflUigeltes Pferd mit Pahlevi-Inschrift und zwei verschiedene V6gel zeigen. Ein kleines Goldplittchen triigt eine schematische Figur einer stehenden Person, die die Tradition der Goldpliattchen des OxusSchatzes fortsetzt und als Votiv-Tifelchen gedeutet werden kann. In einer tiefen Schicht wurde ein Bronzeobjekt gefunden, das 3 Figtirchen zeigt, davon eines offensichtlich eine Ziege. Es scheint sich
16A

wesentlichen bestiitigt (vgl. IRAN XII, 1974, 224).

In diesem Jahr konnte vor allem das zu den

190

JOURNAL

OF PERSIAN

STUDIES

um einen unvollendeten Bronzeguss zu handeln, wodurch vielleicht Bronzewerkstaittenauf dem Takht erschlossenwerden konnen. An mehrerenschon friiherfreigelegten Gebaiudenwurden Sicherungsarbeitenund Restaurierungen vorgenommen.
RUDOLF NAUMANN

TheHulailanSurvey Twelve years ago-in April i963-excavations were started in the Hulailan valley in Northern Luristan by an expedition from the Danish National Museum. The investigations, continued in 1964, were mainly concentrated on problems relating to the Bronze and early Iron Ages, but in addition the excavations at Tepe Guran revealed a sequence of Neolithic occupations,covering almost a thousand years from the middle of the 7th to the middle of the 6th millennium B.C.6. The Hulailan valley is about 25 kms. long. From east to west it is divided into two parts by the Saimarreh River, which receives from the north one major tributary,Jazman Rud. The central part of the valley consists of a flat alluvial plain, situated roughly 930 m. above sea level. It is bordered by heavily eroded hills, and by high folded mountains of Asmari limestone, reaching a height of I812 m. at the western end of the valley. This combination of mountains, eroded hills and an alluvial plain seems to offer an environment that would have been attractive both to Palaeolithic hunters and to early settled herders and farmers. Considering that there are few apparent morphological changes in the valley throughout the Holocene period, and that the plain by a combination of precipitation, springs and rivers is one of the best watered valleys in Luristan,-the possibility of finding a number of sites that might in a conclusive way elucidate the development of prehistoric settlement-patterns would seem very favourablejust in this area. We decided therefore,in 1973, to resume our work in the Hulailan valley, and to start with two or three seasons of intensive surveying carried out on foot in order to discover not only villages, cemetries, caves and shelters, but also minor prehistoricsettlements and open-air sites that did not show in the landscape. The first of these seasonswas accomplishedin the spring of 1973, the second season in the autumn of 1974.6 During the two seasons 161 sites were found and registeredwithin an area of approximately I40 sq. kms. Only sixteen of these sites were previously known from the work of Sir Aurel Stein (1936), the Danish Archaeological Expedition (1963-64) and Clare Goff (1963-67).7 The earliest period representedin the Hulailan valley is the Lower Palaeolithic. On a hilltop not far from the mountains borderingthe valley, and with a fine view over the SaimarrehRiver, was found a concentration of Acheulean artefacts, including a subtriangular handaxe and a great number of choppers and notched or irregularlyretouched flakes. Typologically, and by the dominance of choppers compared to handaxes, the inventory is closely related to that found by ProfessorBraidwood at three late Acheulean sites in Iraqi Kurdistan. The geological setting of one of these sites (Barda Balka) suggests a date within the later part of the Riss/Wiirm Interglacial period, i.e. approximately Ioo,ooo80,000 B.C.

The Middle Palaeolithic period is representedby Mousterian artefactsfrom four open-air sites and three shelters. The inventory from this period is typical for the Zagros Mousterian, as represented e.g. by Frank Hole's excavations at Ghar Arjeneh and Kunji Cave in the Khorramabad area.
In the Upper Palaeolithic period there is an obvious increase in the number of settlements. Seven shelters and caves, and eight open-air sites with Baradostian and Zarzian remains were found. They were all concentrated within a few square kilometres in the north-western part of the valley. In the autumn of 1974 we had an opportunity to test two of the caves by soundings. One of them, Mar Ruz,
5J. Meldgaard, P. Mortensen and H. Thrane, " Excavations at Tepe Guran, Luristan ", Acta Archaeologica XXXIV (1963), pp. 97-133, and H. Thrane, " Archaeological Investigations in Western Luristan ", Acta ArchaeologicaXXXV (1964), pp. 153-69. 6 The two seasons were carried out for the University of Aarhus (Denmark) and the Iranian Centre for Archaeological Research. They were sponsored respectively by the Carlsberg Foundation and the Centre. SA full report on the first season is now in press: P. Mortensen, " A Survey of Prehistoric Settlements in Northern Luristan ", XLV (1974). A preliminary report on the Acta Archaeologica 2nd season will appear in the Proceedings the IIIrd Annual of in on Research Iran (Tehran, I975). Symposium Archaeological

du 1974. Niveau5b: vuegne'raleen direction nord. P1. Ia. Djaffarabad

Pl. Ib. Djaffarabad 1974.

16B

Niveau 5b: ddtail d'une unitj d'habitation.

Pl. Ha. Djaffarabad I974. Niveau6: four culinaire.

I974. Niveau5: foyer. Pl. IIb. Djaffarabad

Pl. IIIa. Ganj Dareh Tepe. " Firepits " in Level E excavated into virgin soil on west side.

" showing" porthole aperture P1. IIIb. GanjDareh Tepe. Brickwall of LevelD structure, after unsealing.

kiln of P1. IVa. GanjDareh Tepe. Stone-lined from LevelD afterremoval deposits.

Pl. I"Vb.GanjDareh Tepe. Badlypreserved childbur and stonebeadsin situ. The bones neartheskul

" Pl. Va. Ghubayra.Underground chamber Room I9 " on the CitadelPlatform,after excavation. Note that the roomis lined with gypsum plaster.

Pl. Vb. Ghubayra.F in redand black, fro

to is Pl. Vc. Ghubayra. Room2 firomthesouthafterexcavation.The archin thecentre the entrance Room22, belowwhichlay Pit ZZ.

P1. Vd. G ornam

excavation. from theeast of the largebuildingat Site F, formerly" the Gatehouse before ", P1. VIa. Ghubayra.View

Pl. VIb. Ghubayra. Ffrom theeast afterexcavation. Site

shrineat Site E before excavation. viewof thehexaeonal Pl. VIc. Ghubayra.General

Pl. VIIa. TepeNush-i Jan. Part of the stepped triangular sanctuary the CentralTempleafter the reof fire movalof theshalefilling.

Pl. VIIb. TepeNush-i Jan. The northend of thesanctuary Temple. A well pre(Room46) of the Western covers of themudbrick served wall niche most faces thepartlyblocked floor. doorway.A thicklayerof whiteplaster

end Pl. VIIIa. TepeNush-iJan. The western of Room23 of the Fort. Note the high windowbesidethespringing the vaultand the low windownotfar above floor level. of

Pl. VIIIb. TepeNush-iJan. Oneoffour blocked archeswhich fla wall lies infront of the archwa

SURVEY OF EXCAVATIONS

191

did not produce any Palaeolithic depositsin the cave itself. But in front of it there was a high concentration of Zarzian flints, obviously redeposited in front of the cave. At the other site, Mar Gurgalan Sarab, we found-under a rockfall-7o cms. of Zarzian habitation deposits, on top of a 50 cm. thick layer with an earlier Upper Palaeolithic flint industry (possibly Baradostian). In both caves bone was preserved. The survey has produced four small Neolithic villages, one open-air site, and three caves with Neolithic remains. It is interestingto note that the Neolithic sites are concentratedwithin the same area as the late Upper Palaeolithic sites, and that the villages are located along the rivers. In contrast to this pattern, the Susiana- and Susa A-related settlements exhibit a uniform distribution across the alluvial plain. Twenty-three villages-but no cave or open-air sites-from this period were found. The distribution pattern might seem to reflect the introduction of irrigation from southern Mesopotamia and Khuzistan to the intermontane valleys of Luristan. As in recent times the water could easily be led across the northern part of the Hulailan plain from Jazman Rud. The transitionfrom the Chalcolithic to the Uruk period is characterizedby a strange phenomenon. Towards the middle of the 4th millennium B.C.most of the villages were deserted. Only four villages seem to have been inhabited continuously into the Early Uruk period. But not a single village from the late 4th or early 3rd millennia B.c. has been located. On the other hand two cave sites found in 1973 and two open-air sites found near the Saimarreh River in 1974 can probably be referred to the Late Uruk period. The apparent depopulation of the Hulailan valley during the Susa A and Early Uruk periods seems to coincide with a drastic increase in population density in Khuzistan and Lower Mesopotamia. On the evidence available at the moment it is not possible to explain why the villages were deserted. But it seems possible that a slight fall in temperaturecombined with increasing salinisation may have complicated the irrigation agriculture on which the economy of the villages depended. The reoccupation of caves and the appearance of open-air sites in the Late Uruk period cannot be related to any permanent settlements in the Hulailan valley. If, however-as suggested by R. McC. Adams and others-there is a relationship between the development of Mesopotamian urban centres with extensive canal irrigation systems and the origin of nomadic pastoralism in the border areas between the Mesopotamian plain and the Zagros, one should expect to find the archaeological evidence of nomads in the Zagros in this period. Future excavations of Late Uruk sites in the Hulailan valley might possibly reveal such evidence. Although the primary object of the Hulailan survey is the study of prehistoric settlement patterns, we have discovered and registereda large number of sites from later periods. In 1974 special attention was paid to an interesting group of Islamic cemetries with carved tombstones and steles with reliefs. Through a detailed study of these cemetries we hope in the future to be able to give an account of the development of style and the combination of symbols on the tombstones, and also through a study of the inscriptionsto get an idea about the tribal distributionin the area throughout the latest centuries.
PEDERMORTENSEN

Kangavar Valley Survey Between September 9 and October 3o, I974 the Godin Project of the Royal Ontario Museum, Toronto, conducted an archaeological surface survey of the Kangavar drainage basin in central western Iran. Our purpose was to explore the immediate physical and historical hinterland of Seh Gabi and Godin Tepe, the sites excavated by the Project between 1965 and I973. The survey was conducted primarily by vehicle. We found, visited and mapped some I68 sites ranging in date from the 6th millennium B.c. to the recent past. Thanks to air photographs this total probably includes almost all of the mounded sites in the valley. Geographically the Kangavar valley is the area drained by the Khorram Rud and Kangavar rivers, which are the north-eastern branches of the Gamasiab river. The area divides naturally into five distinct regions. On the valley floor, moving from south-west to north-east, are Kezel-i Pain,

192

JOURNAL

OF PERSIAN

STUDIES

Afshar and Veleshgird (sometimescalled Chah-i Tuiserkhan), all lying to the south-east of the modern asphalt highway. Of these Afshar is the largest and most populous. Godin Tepe sits in a fairly narrow break in the hills on the border of the Afshar and Veleshgird regions. Seh Gabi is in the centre of the Veleshgird region. Above the valley floor is the hill country to the north-westof the town of Kangavar and the asphalt road, the upper reachesof the Kangavar river. The fifth region is defined by the narrow valleys surroundingMount Khan Gurmaz, the higher reaches of the Khorram Rud lying to the northeast of the valley floor. The archaeological data indicate settlement began first in the three lowland regions and that in earliest times only scattered settlement occurred in the hills. Then in Godin III times (2400-130oo0 B.C.) there was a considerable increase in the density of occupation in the hill country north-west of Kangavar. Finally, occupation did not become dense in the still less hospitable Khan Gurmaz valleys until Parthian times. The Shahnabad period, still known only at Seh Gabi (Godin XII, c. 5500 B.C.), appears now to be the earliest certain occupation of the area. Further study of a perhaps earlier pottery from two small sites in Afshar may lead to a somewhat different conclusion, yet it is quite clear that we have nothing in the Kangavar valley directly comparable to, or as early as, the Neolithic materials well known from further west in the Zagros (Ganj Dareh, Sarab, Guran and Jarmo). We are in no position to comment on possible Paleolithic and Mesolithic occupations in the area as we did not conduct the kind of survey likely to discover such sites. The first wide-spread occupation of the valley dates to Dalma times (Godin X), and the settlement pattern appears to have remained fairly stable from then until the end of the Hosseinabad period (Godin VII), i.e. from c. 4600 to 3500 B.C. There is an average of twenty sites per period, covering an average of 13 hectares. The situation changed markedly in Godin VI times when there were thirtynine sites with a total of 29 26 hectares occupied. Still more striking is the complete absence of any sites comparable to Godin V. At Godin Tepe itself the 1973 excavations revealed that this occupation was confined to the enclosed building complex on the summit of the Citadel Mound (see " The Merchants of Susa " in this issue of Iran). Thus the evidence from survey supports and expands that from excavation: strictly speaking Godin V is not an independent chronological period in the cultural sequence of the area, but is rather a phase of foreign influence in Period VI confined to a single site. Godin VI, from the perspective of the valley as a whole, thus dates from 3500 to 3000 B.C. We found only fifteen sites dating to Godin IV times (2950-2400 B.c.). The Godin III period, in contrast, was the third densest occupation of the valley. We found fifty-seven sites covering a total of 78-28 hectares. Of perhaps equal importance, though heretofore no site had approached Io hectares in size, two of these Godin III sites were over 15 hectares in size. While evidence for simple irrigation appears as early as Godin VI, the settlement pattern of the Godin III period, viewed against the background of the natural hydraulicsof the valley, clearly indicates that fairly large scale irrigationwas now practiced. The number of settlements and the density of occupation declined during the Iron Age: Iron Age I/II, twenty-three sites and 30-o8 hectares; Iron Age III (Godin II), thirty-three sites and 38-32 hectares. That we were unable to describe archaeologically the politically defined Achaemenid period is not surprising. Instead, we were able to isolate a corpus of pottery which, while related to Godin II at one end of a chronological continuum, probably extended at the lower end of its range into early Parthian times. This is the only period in the cultural sequence used on the survey which is not documented by excavation in the valley. We propose to call this archaeological phenomenon the Iron Age IV period. The Parthian period (recognized on the whole by the appearance of Clinky Ware in the ceramic assemblage) marks the second highest density of occupation in the valley. There are ninety-five sites covering a total of 88 - 96 hectares. One notes the considerably greater number of sites occupied at this time in comparison with the Godin III period, but also the only slight increase in the number of hectares occupied. The Parthian period is clearly characterized by a dispersed settlement pattern. Some 77 per cent of the sites are under I hectare in size; 51 per cent are smaller than 0o50 hectares. Only one site is larger than Io hectares. Other features of the period include a wider spread in occupation than ever before, and possibly the introduction of qanats for irrigation.

SURVEY

OF EXCAVATIONS

193

A total of fifty-six sites date to the late Sasanian and early Islamic periods (the two cannot be distinguished ceramically at the moment), while 120osites fall into later Islamic times. Of these some ninety-three date to between the end of the Safavid period and the recent past, and represent the densest occupation of the valley prior to modern times. There are a total of some eighty-five modern villages and towns in the area, of which, unfortunately,only seventy-fivecould be visited on the survey.8 T. C. YOUNG, JR. The KhanaMirza Plain In the recent past a number of scholars have speculated on the highland's contribution to the development of civilization in Iraq and Iran. In order to help determine the extent and nature of the highland culture in the Late Prehistoric-EarlyHistoric periods the writer conducted a seven week survey, on foot and donkey, in the north-east Bakhtiari mountains. The area selected was the Khana The valley, whose main source of water is mountain springs generally inadequate for irrigation purposes is the home of the Janniki Sardsir, a branch of the Bakhtiariconfederation. Because of a number of environmentalconsiderationsthe area provides excellent conditions for a mixed economy of agriculture and pastoralism. Although isolated in the mountains, the area nevertheless lies near the two major pathways through this section of the Zagros: the Bakhtiari road, and the route from Behbehan to Isfahan. It was to be expected thereforethat the area would provide a meeting ground for a number of different cultures, especially those of Sialk and Fars. Results The earliest pottery is represented by straw-tempered crude softwares, with stone tools found in association with them. Perhaps the next stage, although its exact position in the chronology remains uncertain, is a small site where Sialk I related ware was recovered. The pottery however, contains many new elements including a polychrome ware. A large number of stone tools, including cores, rather small blades, burins, and worked chips were collected. The main " tradition " in the area is however buff ware. Exactly when it begins is still unclear, but it appears to have some relations to Early Susiana pottery in the Khuzistan Plain. A flange on one of the soft, chalky vessels appears to match an example from Djaffarabad level 5 (DAFI I, fig. 30; 8). The ware is frequently painted with a dull black paint. Later in time a large number of sites can be seen to be related to the late Ubaid tradition. From the Bakun A period onwards the area shows clear ties with Fars. A number of Bakun A5 sites were located. A Banesh related phase occurs (Susa 16-13), but the bevel rim bowl seems to be quite rare with only one certain example being found. A red slipped black painted ware, as well as a brown on orange-buff ware were found at a number of sites, accompanied by a crude crumbly grey ware, and large jars with horizontal ribs. A red slipped, orange-buff ware appears on the same sites. This Banesh related phase appears especially interesting as an entirely new settlement pattern occurs. One large site, at least 20 m. high, seems to have experienced its major occupation during this period. A number of cave sites was found with accompanying pottery and stone material.
The survey makes it clear that the mountains had close relations with the plains, at least from the time of the Bakun A period onwards, and that an attempt to analyse the growth of civilization in the plains, must take the mountains into consideration. The appearance of a number of large sites in the early historical period provides us, moreover, with the possibility of acquiring economic and social information about life and organization in the mountains, information, which up till now, has not been available to us. A. ZAGARELL Mirza Plain (approximately 5I 'X 31300), a small flat valley,
12 xi

8 km., surrounded by mountains.

* For myself, and on behalf of my colleagues on the survey, Mr. Claus Breede and Mr. Mehdi Rahbar, I want to thank Mr. Seif'ollah Kambaxsh, Director of the excavations at the Anahita Temple in Kangavar, and other members of his staff,

for the supportthey gave to the survey and to each of us by theirkitchen,theirknowledgeand their sharingtheir quarters, friendship.The surveycould not have been done as well or as quicklywithout their help.

NOTES ON TRANSLITERATION FOR CONTRIBUTORS TO IRAN

I, OLD AND MIDDLE PERSIAN

It is recognized that no rigid lines can be laid down here, but it is suggested that the Old Persian Texts,Lexicon,. syllabary should be transliteratedaccording to the table in Kent, OldPersian.Grammar, p. 12; that for Manichaean Middle Persian and Parthian, the transliterationsystem given in Andreasvol. III, p. 66, should be used; whilst for Pahlavi, the table of Manichaica, Henning, Mitteliranische A Manualof Pahlavi,new edition, p. 129, may be used as a reference for alphabets given in Nyberg, transcription.
II. ISLAMIC AND MODERN PERSIAN

The system used for the Cambridge Historyof Islamshould be used here as far as possible. Consonants (a) Arabic , t Sth

JZ

Ssh

i k J 1

kh
dh
r
,

y t'
.

m Sn Sh
w

-a (in construct state: -at) o (b) Persian additional and variant forms. The variant forms should generally be used for Iranian names and for Arabic words used in Persian. jz 4Sg Yp

gh

Sch ye The Persian " silent h " should be transliterateda, e.g. nima. (c) Vowels Arabic and Persian. Short: a Long: I or a -" Doubled - iyy (finalform:I) J i u , Diphthongs au
i

zh

.1V

; ai

Notes

shouldbe represented -i, or afterlong vowels,by -yi, e.g. umard-yijinki. by I. The iizdfa " 2. The Arabicdefinitearticleshouldbe writtenas al- or 1-,evenbeforethe so-called sunletters", 'Abd al-Malik,Abu 'l-Nasr. e.g. 3. The macronsof Abfi and Dhti (Zi) should be omitted beforethe definitearticle, e.g. Abu '1-Abbts (but Abi 'Ubaida). It is obviousthat for the renderingof linguisticand dialecticalmaterial,and possiblyalso for is contemporary literaryand spokenPersian,this rigoroussystemof transliteration inappropriate; contributors shoulduse theirdiscretion here.
III. GENERAL POINTS

I. 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 rigorouslytransliterated. Archaeologistsare asked to be especially careful in representingthe 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. 194

ABBREVIATIONS AASOR AfO AJA AJSL AJ AK AMI ANET AOr Arch Anz AS BA Besch BASOR Belleten BGA Bib Or BMMA BSA BSOAS CAH CIA El ESA IAE ILN Iranica JA JAOS JEA JHS JNES JRAI JRAS JRCAS KF LAAA MAOG MDAFA MDOG MDP MJ OIC OIP OS PZ RA REI SAA SAOC Sov Arkh SS Survey TT WO WVDOG ZA ZDMG Annual of American Schools of Oriental Research Archiv ffir Orientforschung American Journal of Archaeology American Journal of Semitic Languages and Literatures Antiquaries' Journal Antike Kunst E. E. Herzfeld, Archaeologische Mitteilungen aus Iran Pritchard, Ancient Near Eastern Texts Archiv Orientaln' Archiologischer Anzeiger Anatolian Studies Bulletin van de Vereeniging . .. de Antieke Beschaving, Hague Bulletin of American Schools of Oriental Research Tuirk 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 Corpus Inscriptionum Arabicorum Encyclopaedia of Islam Eurasia Septentrionalis Antiqua E. E. Herzfeld, Iran in the Ancient East (1941) Illustrated London News Iranica Antiqua Journal Asiatique Journal of American Oriental Society Journal of Egyptian 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 Kleinasiatische Forschungen Annals of Archaeology and Anthropology, Liverpool Mitteilungen der altorientalischen Gesellschaft Mdmoires de la D6l6gation Arch6ologique franqaise en Afghanistan Mitteilungen der deutschen Orientgesellschaft Meimoires de la D6l6gation en Perse Museum Journal, Philadelphia Oriental Institute, Chicago, Communications Oriental Institute, Chicago, Publications Orientalia Suecana Praehistorische Zeitschrift Revue d'Assyriologie Revue des Studes Islamiques Soviet Anthropology and Archaeology Oriental Institute, Studies in Ancient Oriental Civilisation Sovetskaya Arkheologiya Schmidt, H., Heinrich Schliemanns Sammlung trojanischer Altertumer A Survey of Persian Art from Prehistoric Times to the Present, ed. A. U. Pope, Oxford, 1938 Tuirk Tarih, Arkeologya ve Etnografya Dergisi Die Welt des Orients Wissenschaftliche Verdffentlichungen der Deutschen Orientgesellschaft Zeitschrift fir Assyriologie Zeitschrift der Deutschen Morgenlandischen Gesellschaft

'.

\
'> ab
AR BAIJAN ' . e j

250 2Altts

. "_400 C S

nketres . . ..

AM
...

J2

U.
*

S.
*

S.

R.

1,zayeh
,

*Gorgan " ?*Meshe

Damgaf L
1

Qazvio.

"

\."4 NDARAN
EA THRAN

Abbas

SA, .KDIBASan
1.

SusaY

vr N"" I Isf

RAN

AARABIAA
JA . . ?e I" s tsa ahKem. . Kr . :0..

R
*rKe Pe
a

N
rmaru
/

mom .

sepot:::
S~1L~rBan dar.bba

"

.a"..4:"

w 0" ,

...,

..

.0-0

PRINTED 109

IN GREAT BRITAIN BY HEADLEY WC2B 6PX AND KINGSWAY LONDON

LTD BROTHERS ASHFORD KENT