<<

The Smithsonian Institution Regents of the

The Earliest Islamic Commemorative Structures, Notes and Documents Author(s): Oleg Grabar Source: Ars Orientalis, Vol. 6 (1966), pp. 7-46 Published by: Freer Gallery of Art, The Smithsonian Institution and Department of the History of Art, University of Michigan Stable URL: http://www.jstor.org/stable/4629220 Accessed: 14/12/2009 06:30

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=si.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The Smithsonian Institution and Regents of the University of Michigan are collaborating with JSTOR to digitize, preserve and extend access to Ars Orientalis.

http://www.jstor.org THE EARLIEST ISLAMIC COMMEMORATIVE STRUCTURES, NOTES AND DOCUMENTS* BY OLEG GRABAR

ONE OF THE MOST CHARACTERISTIC BUILD- The memorialbuilding is not a pecu- INGS of Islamic architectureis, without liarly Islamicphenomenon. Indian stupas doubt, the monumentaltomb. The Taj servedto honorBuddhist relics; most an- Mahalor the great Mamlukmausoleums cientNear Eastern civilizations developed in Cairoare visited by thousandsof casual varyingkinds of mausoleums;the nomadic tourists,while every traveler in North world itself, whether Semiticin Arabia or Africaor the Near Easthas seen along the Indo-Europeanand Turkic in CentralAsia, roads,on top of hills,in cemeteriesof cities usedmore or less permanentforms of com- and villages,at timeseven in fields,hun- memorative architecturalsymbols; Hel- dredsof smallshrines usually assumed to lenistic and Roman memorial structures be the restingplace of somesaint or hero, have remainedin large numbers;and the or supposedto markthe spotof somecele- was a central concept as well bratedor forgottenaction. Many terms are as form in the growth of Christianarchi- usedfor theseconstructions, whether of a tecture. roughpeasant work or exquisiteartistry. It is easy to assume that Islamic me- Theymaybecalled qubbahs and gunbadhs, morial and funeraryconstruction was but "domes,"after their prevalentform, or a continuationof the numeroustraditions turbahs,"tombs," from their most common of the pre-Islamicor non-Islamicworlds, function, or imamzadehs, "sons of an and this assumptionunderlies some of the im2m," expressingtheir religious, almost studies which have been devoted either to ecclesiastical, connotation. They may be precise Islamic examples1or to the most walis or marbats, "places of a holy per- characteristicforms and functions of such sonage,"indicating their relation to some buildings.2If suchwas the case,the problem holy hero, maqams,"places," when related for the historian of would be to the emplacementof some event, mash- simply to identify various types of com- hads, "placesof witnessing,"true martyria memorative forms used by the Muslims implyinga commemorativevalue. At times, and to study their evolution. There are, rarerterms such as qasr,"castle," or darg2bh, however, two objectionsto the hypothesis "palace," were used for them. All these that the new civilizationsimply anddirect- names-whose precise history and origins ly appropriatedthe proceduresand func- are still to be elucidated-illustrate the multiple facets of memorial construction 1 E. Diez, Persien, Berlin, I923, pp. I ff., in the mindsof the Muslims. 73 ff.; K. A. .C. Creswell, Muslim architecture of Egypt, Oxford, I952, vol. I, pp. IIO ff. Thisstudy was completed thanks to a grant 2 A. Grabar,Martyrium, Paris, 1946, vol. I, fromthe Centerfor NearEastern and North Af- pp. 85-86, I45. E. E. Smith, The dome, Princeton, ricanStudies at theUniversity of Michigan. I95O, passim and esp. pp. 4 I-44. 8 OLEG GRABAR

tionsof old. The firstobjection is theoreti- tices remained,visits to tombs,lamenta- cal. It is rareto find any branchof Islamic tions,erection of tentsover tombs,setting art merely taking over completedolder up of pillarsnearby, watering of tombs. formswithout alteration of shapeor mean- All suchpractices derived from pre-Islamic ing. Overwhelmedas they are with Sasa- habitsand were maintained in spiteof nu- nianand especially Byzantine and Roman merousefforts, at leastin thefirst centuries plans and methodsof constructionand of theHijrah, to curbthem.4 But, the early decoration,the monuments of earlyIslamic examplesof funerarypractices and the op- art can rarelybe consideredByzantine or position to them involved behavior, not Sasanian;either in theirpurpose or in the constructions.The tomb of the Prophet, relationship of their components, these whose potential significanceas a focal monumentsare new and imply new needs point for worshipwas self-evident,was or new tastes. It is usually only when a carefullykept out of themain direction of facet of the new Islamicculture developed prayerin the mosqueof Medina;it was in a mannerwhich demandedor permitted unavailableto the faithfuland enclosed in monumental expression that monuments an irregularand purposefullyungainly developedto expressit. Therefore,the fact screen; and it was only in the late I3th that there were mausoleumsin the pre- centurythat it acquireda cupola.5 Islamic world does not by itself explain If, then, early Islamwas so stronglyset the existence of Islamic mausoleums;an against the building of mausoleumsand explanation of their appearancemust be againstcultic practicesin or aroundtombs, given in the cultural terms of the time the eventual appearanceof thousands of when they appeared. mausoleumsand shrineswith a memorial The second objection to the notion of significancethroughout the Islamic world a directpassage from pre-Islamicfunctions requiresa more precise explanation than and purposes to Islamic times is raised the simple statementof a continuationof by the early Islamic view on mausoleums. pre-Islamic habits and practices. An at- It is clear that early Muslim doctrine tempt mustbe made to determineprecisely condemnedany architecturalglorification of tombs, and in fact found even most I The appropriate texts have been gathered funerary and commemorativeceremonies and discussedseveral times; J. Pedersen,art. mas- objectionable. The taswiyah al-qubur, djid, part A4, Encycl. of Islam; G.Wiet, MCIA "equalization of tombs (with the sur- Egypte 11, pp. 64 ff.; I. Goldziher, Die Heiligen- roundingground)," was felt to be the most verehrungim Islam, MuhammedanischeStudien, Halle, I890, 275 ff. appropriate expressions of the equality pp. 4 M. Canard, Vie de l'UstadhJaudhar, Alger, of all men in death, and veneration of I958, pp. i5o-i5i, and esp. note 340; Maqrizi, tombs or ceremoniesaround tombs were vol. I, pp. 3I2, 3I3, 327 are but a few examples. consideredto derive from improperChris- For an overall view of early funerary practices, tian and Jewish habits.3From the very be- as drawn from hadith literature, see I. Griitter, ginning this prohibition ran against older Arabische Bestattungsbrauchein friihislamischer Zeit, Der Islam,vol. 3I, I954, pp. I47ff. tradition of behavior, and stories abound I J. Sauvaget, La mosquzeeomeyyade de Me'- which indicate that certain funeraryprac- dine,Paris, I947, p. 44. TIHE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 9 where,when, how, and why this new and cludesa moreor lesslengthy paragraph on apparently most characteristicIslamic mashhads7;these vary from specifically monumentappeared. The documentation mentioneddomes over tombs or sanctuaries presentedhere has beengathered in an at- to simpletombs, caves, or even to ribats, temptto providean answerto theseques- militarymonasteries which often had ac- tions.My originalpurpose was merelyto quireda holy significanceby theirassocia- draw up an annotatedlist of the earliest tionwith religious leaders, but whose archi- remainingsanctuaries, but it soon became tecturalforms-at leastinsofar as we can apparentthat a simple list of standing claimto knowthem8-were quite different monumentsand of monumentsknown from those of mausoleums,even though throughinscriptions-i.e. from basically theirholy characterled peopleto be buried archaeologicalsources-was not sufficient. in or near them.9Na$ir-i Khosrowalso On the one hand,among the disusedsanc- mentionsthat many shi'ite mashhadsin tuaries,it is only accidentalthat some have Tripoli and Tyre looked like ribats.10In remainedrather than others.And on the theseinstances, it seemsfairly clearthat, other, many of the still popularshrines even though these institutionsplayed a wereso muchredone in latercenturies that functionsimilar to that of mausoleumsin archaeologicalanalysis alone cannotpro- the religiousbehavior of their time, they vide the date of foundationof the sanc- were not shrinesin the samesense as the tuary. But, as one beginsto cull literary tombof 'Ali, and their architecturaltype sourcesfor documentson the subjectof was different.In order,then, to staywith- earlyshrines, a numberof additionalprob- in our definitionof the architecturalform lemsare raisedwhich require preliminary withwhich we areinvolved, these buildings comments. 6 This difficulty with terminology has been The mostimportant of theseproblems recentlypointed out by C. Cahen in his re-edition is oneof vocabulary.If we defineour con- of J. Sauvaget,Introduction a l'histoirede l'Orient cernas being an investigationof theorigins musulman,Paris, 196I, p. I 4. of a centrally-plannedroofed building 7 Al-Muqaddasi,pp. I02-103, I30, I46, 209 erectedover a tombor a holyplace in order to 2I0, 333-334, 367, 344, etc.... 8 The questionof the ribat as an architectural to emphaziseand proclaim, as the case may form or as a function is still unsolved, at least in be, the holiness,the glory, the wealth,or the Orient. In North Africa, the works of G. Mar- the power of an individualor an event, Sais (Note sur les ribats, Melanges R. Basset [Paris, how are we to interpretthe numerousin- 1923-25], PP. 395 if.; art. ribat in Encycl. of stances,among early geographers and his- Islam) and A. Lezine (Le ribat de Sousse, Tunis, i956) have somewhat clarified the situation, but torians,of practiceswhich imply an archi- the degree to which the North African examples tecturalsetting of somekind, or to explain are applicable to the Cilician or Central Asian the exactsignificance, in theirtime, of such frontiersis still very uncertain. wordsas mashhador turbah?6A few illus- 9 For instance, Ibn al-Jawzi, Muntazam, trationsmay help to explainthe difficulty. vol. 8, pp. 2I4, 303-304, 325, lists one group of In his examples which can easily be multiplied from descriptionof the provincesof historicaland geographicsources. the Muslimworld, the i oth centurygeog- 10 Nasir-Khosrow, Safar-nameh, trans. Ch. rapher,al-Muqaddasi, almost always in- Schefer,Paris, i888, pp. 42ff. I0 OLEG GRABAR arenot includedin ourlist inasmuchas the informationis less clear,but manyof the formaland functionalproblems they pose mashhadswere pre-Islamicconstructions, arenot yet solvedand wouldhere lead us simpleoratories or even naturalfeatures. astray. At the risk of beingoverly conservative, But if the caseof the mashhad-ribatis thislist eliminatesall instancesof mashhad fairlysimple, difficulties are greater when wherethere is no clearevidence that a spe- one attemptsto evaluatethe architectural cific buildingwas erectedin Islamictimes settingsuggested by the mashbadsto Bib- for the sole purposeof commemoratinga lical prophetsand events mentionedby personor an event. Nasir-i Khosrowand Muqaddasi,by the A verysimilar problem exists in reverse thirteenshi'ite mashhads seen by the Per- for a groupof archaeologicaldocuments. sian travelerin Basrah,1"by the mashhad In southernEgypt therewere until very of thePalm (of theProphet) and the mash- recentlytwo buildingsknown as mashhads hadof Vowsin Baghdad,12or by the com- (one only still remains),"6consisting of a plex sanctuariesof the Seven Sleepers,"or hall coveredby nine or six domes,of a by numerousother examples found in his- minaret,and perhaps of a court.The date toriansor geographers.14In most instances of thesebuildings can be fixed fairly ac- thereis clearevidence of somekind of con- curately, but not their purpose or use. A struction, but these constructionsare not passagein Maqrizi may, however, help to always mausoleumsor preciselycommem- explainthem and to excludethem from our orativebuildings. Thus, the mashhad of the concern.The Cairenehistorian described a Palmin Baghdadwas originallya masjid, jami'al-fiyalah, "mosqueof the Elephants," a private mosque,in which a miracletook from Fatimidtimes, with nine domesover place; the first ascertainedconstruction, in the sanctuary, whose decoration recalled the i oth century, around the cave of the the backsof elephantsin caliphal proces- Seven Sleepersin Damascuswas also a sions."The buildingis obviously relatedin masjid."5For the Palestinianexamples our type and plan to the southern Egyptian buildings and belongs to a long series of oratories,often built in 11 Ibid., pp. 239-40. cemeteries,whose 12 G. Makdisi, The topography of eleventh- purposewas less preciselycommemorative century Baghdad, Arabica, vol. 6, I959, p. 289, than that of a mausoleum,for they were with all appropriatesources. erected for a different liturgical activity, 13 L. Massignon,Les sept DormantsdEphere, prayer.While it is true that theseconstruc- Revue des Etudes Islamiques,I 9 5 5. tions served, partly at least, to glorify a 14 For instance,in addition to the fairly easily accessibleinstances in geographicaltexts, see Ibn man or a family, they were, architecturally al-Jawzi, Muntazam,vol. 8, p. 46, for a mashhad and functionally,different both frommau- outside Kufah, which was visited by pilgrims to soleumsand from congregationalmosques. Mekkah;vol. 9, p. 104, mashhadof the cemetery That the practice was not limited to Cairo of the Quraysh;p. I39, mashhadof one Muham- mad b. Ishaq. 15 S. 16 pp. al-Munajjid,Ahl al-Kahf, Majallah al- Creswell, MAE, vol. i, I44-I45 and Mujm'ah al-'Ilmi al-'Arabi, vol. 3 I, I 9 5 6, pp. 149-I 5 2. 602ff., esp. 60o. 17 Maqrizi, vol. 2, p. 289. THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES I I is clearlyshown by a passagein Ibn al- or single burials.The second point is that Jawzirelating the burialof one 'Ubaydal- the plan of nine domesis to be relatedto lah b. al-Hasanin front of his masjidin that of the oratorieswe have just men- Baghdad.'" tioned.21Later some of theseoratories may It is in this category of cemetery have cometo be calledmashhads and the mosquesthat I proposeto put a construc- tombswhich were near them were covered tionusually assumed to bea mausoleum,the with domes,but all we can safely suggest mashhadof SharifTabataba, datable ca. is thatin the ioth centurythere were fam- 334/943. K. A. C. Creswellis responsible ily plots in the largercemeteries and that for therecovery of thiscurious structure, a oratorieswere at timesadded to themfor squarebuilding with nine domeson cross- prayerand devotions. shapedpiers and open on all sidesexcept Thepoint of theseexamples is that the the centralpart of the qiblahwall, where termmashhad, either as it is usedtoday for thereis a mibrab.Nearby, in a morerecent certainbuildings or as it was used in the building,were found inscriptions with the past,implies different things; whenever it namesof membersof theTabatabafamily.'9 is possible,each instance merits a special By relatingthis discoveryto a passagein analysisand suggestsa differentexplana- Ibn al-Zayyat describinga mashhadin tion,but it seemsthat, in earlytimes, mash- whicheleven members of the familywere had was mostcommonly used for any sa- buried,Creswell suggested that the newly cred place and does not always mean a discoveredbuilding was thatmashhad and specificconstruction over it, while, as the datedit at thetime of thedeath of themost centurieswent by, many a smallbuilding celebratedmember of theclan. One cannot with obviousreligious features, such as a quarrelwith the reconstructionor with its mibrab,acquired the name of mashhad. relationwith the Tabataba family. But two A third difficulty arisesaround such additionalremarks may be made which alter words as turbah, maqbarah,and even a little its significance.First, there is in the qabr.22In the earlyperiods the termsseem Museumof IslamicArt in Cairoan inscrip- to have been interchangeable,and the tion20which refers to the "gate(? [uncertain reading])of the maqbarah"of another 21 To the Egyptianbuildings should be added memberof the family, also mentionedby the Bib Mardiimmosque in Toledo, whose archi- Ibn al-Zayyat,who died in 348/949.The tectural type is the same, whose date is ca. iOOO, termmaqbarah (see below) simplymeans but whose local function has not yet been ex- a cemeteryor a partof a cemeteryand this plained; G. MarSais, L'architecturemusulmane d'Occident,Paris, I954, pp. I 5 I-I 52. earlypiece of information(as opposedto 22 For the latter term see the constant state- the late descriptionof Ibn al-Zayyat) ment in Ibn al-Banna'"sdiary that people were merely suggestssome kind of enclosure buried"in the tomb"(ft qabr or biqabr)of Ahmad separatinga familyplot fromother family b. Hanbal or of other religiousmen, G. Makdisi, An autographdiary, Bull. Schoolof Or. and Afr. Studies, vols. I8-I9 (I956-57), vol. 2, pp. 251, 18 Ibn al-Jawzi, Muntazam, vol. 6, p. 370. 2z5; vol. 3, p. 34. The term obviouslyrefers to a 19 Creswell,MAE, vol. i, pp. I i ff. whole area and not precisely to the sepulchre 20 Re'pertoire, No. 1495. proper. I2 OLEG GRABAR word turbabdoes not meanmore than a as it is preferredin thestate of life."26These large plot in a cemetery,which was often ceremonies,no doubt, all found certain bought in advance, in which one or more formsof architecturalexpression, but the people were buried, and which could at latterwere far fromalways involving the timesbe separatedfrom other similarplots tombsthemselves.27 by a fence, a wall, or even a portico.23Here As a consequenceof thesefacts, this list again the amountof constructionis not al- eliminatesall buildingswhich cannot be ways clear, and in a few instanceswe have clearly shown to have had, when first included turbahsbecause of the probable erected,a precisememorial connotation extent of the constructionserected around and to have been built by Muslimsover the tombs, even though no clear evidence sacredplaces and tombs.This rather strict exists about constructionsover the tombs. definitionwas dictatedby our primary As one reads the geographersof the ioth interestin investigatingthe originsof the and later centuries,it becomesquite appar- great mausoleumsof later times, but it ent that an intense life was developing in shouldnot overshadowthe fact, in itself the cemeteriesof the great cities, a life deservingstudy, that in many instancesa whichhas recentlybeen illuminatedby the whole apparatusof guesthouses,kitchens, late L. Massignon.24A great deal of atten- bakeries,enclosures, oratories, gates, even tion was given to the places where people dwellings, sprang up together with the wereburied; and it was a new form of piety commemorativebuildings and, at times, to visit cemeteriesand tombs, as, for in- even earlier. The word mashhadreferred stance, Abu Bakr Muhammadb. 'Ali, the to this whole phenomenonand a complex last of the great Madara'yun family (d. but different subjectof researchfrom ours 956-57), used to do every day in Cairo,25 would be the development of activities or as the Hanbalites did in Damascusand around holy places and the architecture Baghdad, for "the neighborhood of the these activitiescreated. saintsis preferredin the state of burial,just A last preliminaryremark concerns our exclusionof the earliestepigraphical docu- ment suggestingthe existenceof the prac- 23 Ibn al-Jawzi, Muntazam, vol. 7, pp. 207, 240 (buying of a turbahin Kufah), 289 (a dome, tice of erecting domes over tombs. It oc- which miraculouslycollapsed, was built over the curs in the comparativelymodern mauso- tomb of a man buriedin a turbah);vol. 8, p. i i 8; vol. 9, p. io (two turbahs next to each other); 26 G. Makdisi,Diary, vol. 3, p. 34. vol. io, pp. 8o (maqbarahwith apparently the 27 The examples quoted here do not pretend same meaning), 134, I42 (turbab used in the to be exhaustive;many otherscan be broughtout, plural apparently to indicate plots); Mas'iidi, including some provided by epigraphy, such as Muruij,vol. 8, p. 234. Other instancesare brought Re'pertoire,No. I8I3, by which a qabr, tomb, is out in our list when they occurredin places later built (bana) in 970 for a woman who died in 862. to be coveredwith a dome. Is this a mere stone or brickcenotaph, such as has 24 L. Massignon,La Cite'des Morts au Caire, been illustratedby D. S. Rice, 7Teoldest illustrated Bull. Inst. Fr. Arch.Or., vol. 57, I958. Arabic manuscript, Bulletin of the School of 25 H. Gottschalk,Die Madara'iyyuin,Berlin, Orientaland AfricanStudies, vol. 22, I959, pl. 1931, p. I25; Maqrlzi, vol. 2, p. i56; Massignon, VIII?-or should we interpretit as a true mauso- La Cite', pp. S-6 leum? THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES I 3 leumof Dhu'l-Nuinal-Misri in Cairoand date of the deathof the personage),3"has has beenpublished by L. Massignon28and been proved to be I 3th centuryforgery in studiedin detailby G.Wiet.29 The inscrip- a recentarticle by M. E. Masson.32 tion is presumedto have beenmade at the To these problems presented by the timeof his deathin 245/859 and includes, lack of adequatehistorical dictionariesof in additionto formulasfairly commonon Arabic and Persian, one should also add epitaphs,the followingsentence: "he re- the peculiaritiesof our informationabout quiredin his authentictestament [fi wa- the early medieval Islamic world. Thanks siyyatihal-musnadah 'anhu] that no con- to al-Maqrizi'sdescription and to the re- structionbe madeat histomb and no dome searchesof M. van Berchem,G. Wiet, and [qubbab]be raisedover it." The rather K. A. C. Creswell,our knowledgeof Egypt, neutralfirst prohibitionmay referto any especially Cairo, is quite extensive; for enclosure,while the secondone dealsvery Baghdadthe texts aremore rewarding than preciselywith a dome;this inscriptionhas the monumentsbut have not yet been fully beentaken to be an indicationof the prev- exploited; in Central Asia systematicex- alenceof domesover tombsin the middle plorations have recently brought to light of thegth century, and later evidence exists large quantities of archaeologicaldocu- that suchprohibitions were made by holy ments as yet insufficiently related to a personages.30But, with respect to Dhu ratherscanty literaryinformation. But for 'l-Nuin'sinscription, L. Massignonwas most of Iran and few authenticliter- struckby the expression'in his authentic ary or archaeologicalsources remain from testament,"and especiallyby the sugges- before the i 2th century. The extent to tion that the authenticityis shownby an which these variationsin the natureof the isnad,a chainof authoritygoing backto informationavailable have unbalancedour the mystichimself; the implicationwould results cannot now be said, but the possi- be that this is a latermedieval forgery re- bility should be kept in mind. As we have flectingopinions opposed to the prevalent attemptedto cover the whole extent of the growthof a funeraryarchitecture. The evi- Islamicworld, thereis no doubt that many dencepresented in thisstudy, as well as the examples, especially in texts, have been fact thatall otherinscriptions in the build- missed; but it is hoped that this list may ing are muchlater, confirms this interpre- serve both to suggesta few conclusionson tationof the text of the inscription.If it is the origins of a major form of Islamic ar- indeeda medievalforgery it is not unique. chitectureand to illustrate a method and Thecelebrated Central Asian tombstone of an approachwhich may be used in dealing Abu Zakariya al-Waraghshari,for in- with other types of monumentsas well.33 stance,with the date 230/844 (the actual 31 Repertoire, No. 3I0. 28 L. Massignon,Etudes arche'ologiques,Bull. 32 M. E. Masson,Sredneaziatskie namogilnye Inst. Fr. Arch. Or., Vol. 9, I9II, pp. 9I-96; Re'- kairaki, Epigrafika Vostoka, vol. II, I 9 5 6, pp. pertoire,No. 440. 8 ff. 29 G. Wiet, MCIA Egypte II, pp. 62 ff. 33 As additional information, archaeological 30 E. Dermenghem,Le culte des saints dans or textual,is acquired,it is hoped that supplements l'Islam maghre'bin,Paris, 1954, pp. 39, i i 8. and correctionsto our list will be published. I4 OLEG GRABAR

Themonuments are arranged in chron- canceas a memorialto Muslimdomination, ologicalorder with an arbitrarylimit set by its later developmentinto a shrineto aroundI I 5o, at a timewhen all provinces the Ascensionof the Prophet,and by its of the Muslimworld had acquiredlarge nameand shape-the firstqubbab of Islam numbersof mausoleums.Undated build- -this buildingfully belongsto our series, ings are set at an estimatedante quemdate, even thoughits sourcesand its apparent althoughit is clearthat someof the shi'ite lackof immediateformal posterity single sanctuaries,especially in Iraq, preceded it out as a uniquecreation. the earliestdate known for them,and that 2-Samarra, Qubbah al-Sulaybiyah, the chronologyof the numerousCentral datableon historicalgrounds in 862. A Asianmausoleums has not yet beensatis- centralsquare room covered with a dome factorilyestablished.34 A brief description followseach monument and, when it seem- 34a MainAbbreviations. ed indicated,a discussionof certainprob- Creswell,MAE: K. A. C. Creswell,Muslim archi- in- tectureof Egypt, Oxford, I952, lemsposed by individualmonuments is vol. I. cluded.The bibliography appended to each Denike: B. P. Denike, Arhitekturnyi entry is not meant to be complete but OrnamentSrednei Azij, simplyto referthe readerto the most ac- Moscow I 9 3 9. cessiblepublications.34a Ibn al-Jawzi: Ibn al-Jawzi, al Muntazamfi ta' rikh al-mulik wa-l-umam, Hayderabad, I938 ff. Iutake: TrudyIuzhno-Turkmenistanskoi LIST OF MAUSOLEUMSAND ArheologicheskoiKompleksnoi MEMORIALCONSTRUCTIONS Ekspeditzij,Moscow, 1952 if. (i i vols. to date). Le Strange, G. Le Strange,7he lands of the i-, ,dated Lands: EasternCaliphate, Cambridge, 69I A.D.Thiscelebrated building consists of I930. a centralcircular part covered with a dome Maqrizi: al-Maqrizi,K. al-Khitat, Cairo, and surroundedby two octogonalambu- I270 H, 2 vols. latories.The building has been described by MCIA: G. Wiet, Materiauxpour un Egypte II: corpusinscriptionum Arabi- K. A. C. Creswell,Early Muslim architec- carum:Egypte deuxiemepartie, ture,Oxford, I932, vol. I, pp. 42-94, and Cairo, I929-30 (vol. 52 of the its meaningwas discussedby 0. Grabar, Me'moiresde l'Institut Franfais The Umayyad Dome of the Rock, Ars d'ArcheologieOrientale au Orientalis,vol. 3, I 9 59. By its earlysignifi- Caire). Pugachenkova: G. A. Pugachenkova,Puti Raz- vitiia ArhitekturyIuzhnogo 34 There is no single recent work known to Turkmenistana,Moscow, I 958 me covering the whole of Central Asian archi- (vol. 6 of lutake). tecture;the most comprehensiveis by G. A. Puga- Repertoire: E. Combe,J. Sauvaget,G. Wiet, chenkova, Puti Razvitiia arhitektury Iuzhnogo Re'pertoirechronologique Turkmenistana,Moscow, I958, but the criteria d'epigraphiearabe, Cairo, used for the dating of the many anepigraphic I93Iff. monumentsdo not seem to me to be as yet suffi- Survey: A. U. Pope, ed., A survey of ciently preciseto justify all the proposeddates. Persianart, Oxford, I939 THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES I 5 is surroundedby an octogonalambulatory. seen, is usually difficult to interpretas re- The mainpublication is by K. A. C. Cres- ferringto a built mausoleum.It is possible, well, Early Muslimarchitecture, Oxford, of course, that funerary terminology was I940, vol. 2, pp. 283-285. The identifica- still quite vague at the time when the two tion of the buildingas themausoleum built chroniclerswrote, but, while the two points for the caliphal-Muntasir by his Christian we have made do not necessarilyinvalidate motherhas beenmade by E. Herzfeld,in the identificationsmade by Herzfeld, they E. Herzfeld and F. Sarre,Archaiologische raise the question of the origins of the Reiseim Tigris-und Euphratgebiet, Berlin, monument and point to its exceptional I9II-20, Vol. i, p. 86; and the eventual character. discoverythere of threetombs seemed to 3-Qumm, tomb of Factimah,second confirmthe statementsof thetexts that the half of the gth century(?). The history of caliphsal-Mu'tazz and al-Muhtadiwere the sanctuariesof Qumm is quite difficult buriedtogether with al-Muntasir(Tabari, to disentanglefrom the legendswhich have Ta'rikh,ed M. de Goeje and others, Leyden, surroundedthem and from the fact that i879ff., vol. 3, pp. I498-I499, I7II, and the full developmentof the city as a shi'ite I823; al-Mas'udi,Muruj al-Dhahab, ed. shrineis comparativelylate; furthermore, and tr. Barbierde Meynardand Pavet de whateverarchaeological evidence may exist Courteille,Paris, I86I-77, vol.7, p. 300). there has not been collected. For the early Two problemsare still raisedby thisiden- period our best source is the Ta'rikh-i tification.The firstone is that, sincethere Qumm, written in 988-89.36 According to was no contemporarytradition of Byzan- it (pp.2I3-2I5), the tomb of the sister of tine mausoleumsfor emperors,one may 'Ali al-Rida (who died in 8I7-I8) first wonder where the Christianmother ac- received a covering of mats and then a quiredthe idea of erectinga mausoleum qubbah.The latter was built two genera- overthe tombof herson, unless it be from tions later by Zaynab, a granddaughterof someobscure provincial tradition; the form 'Ali. The sameperson was apparentlyalso of thebuilding may very well havebeen in- responsiblefor a secondqubbah next to the spiredby the earliestIslamic commemora- first one. In thesetwo constructionssix de- tive construction,the Dome of theRock, as scendants of 'Ali al-Rida (mostly in the hasbeen suggested by Creswell.The second second generation)were buried.The exact problemis that the textsdo not mentiona date of the constructionof these family building;they simplysay that the tombof mausoleumsis difficult to establish.Zaynab the caliphwas knownand that it was the outlived Muhammadibn Musa (see below, first caliph'stomb to be known35;Tabari no. ii), who had died in 908; hence the actuallyeven refersto the place of burial qubbabscould have been built eitherin the as a maqbarah,a termwhich, as we have 36 The Ta'rikh-iQumm by Hasan b. Muham- madQummi was publishedin Tehranin I934; it 35 The same verb (azhara) is used elsewhere was used by A. Houtum-Schindler,Eastern Persian (Ibn al-Jawzi, Muntazam, vol. 6, pp. 45, I99; Iraq, London, I897, pp. 63-64, and in a compi- VOl. 9, p. 215) withoutany suggestionof mauso- lation published in 1906 by one Muhammad leum. Husayn Qummi. I6 OLEG GRABAR early i oth centuryor in the latterpart of stancesof survivalsof older decorative the gth. They were considerablyenlarged patternsin religiousarchitecture alongside in 96I-62 by a local governor,presumably new ornamentaldevelopments.37 Thus ei- becausethey had become major places of ther we have here an early mausoleum pilgrimage,even thoughother geographers fromaround goo whosepresent identifica- of the ioth century do not mention them. tion is incorrector a latermausoleum with 4-Tirmidh, mausoleumof al-Hakim archaizingdesigns. al-Tirmidhi, supposedly of the late gth 5-Najaf, mausoleumover the tomb century.No plan of the buildingis known of 'Ali.The exact date of thefirst construc- to me and the evidence for the date has tion is not certainand for a longtime even been basedon its decorativemotives: B.V. the preciseplace of the tombwas contro- Veimarn,Iskusstvo Srednei Azij, Moscow, verted(it still is in Mas'udi,Muruj, vol. 4, I940, p. 29, fig. Io; B. P. Denike, Arhitek- p. 289). A first qubbahover the spot on turnyi Ornament Srednei Azij, Moscow, which eventualagreement will be made I939, pp. 38 ff., fig. 22. Bothauthors pub- seemsto havebeen erected in 902 (L.Mas- lish the same decorative fragment which signon,Explication du plan de Kufa,Me- they appropriatelyrelate to the third Sa- langes Maspero, Cairo, I935, vol. 3, pp. marra style. There are, however, definite 3 56-3 57). Whateverthis first construction objectionsto be made to the date and to the may have been, it was replacedsome time identificationof the tomb.Recent work by before 3I7/929 by a new building spon- N. Herr on al-Hakim al-Tirmidhi(among soredbyAbu al-Hayja' 'Abdallahb. Ham- otherarticles, A Su4ipsychological treatise, dah. Ibn Hawqal's text (ed. J. H. Kramers, The MuslimWorld, vol. 5 I, I96I, p. 26) Leyden, 1938-39, p. 240) describes the shows that the date of his death is quite buildingas a high dome on columnswith a uncertain;moreover the general evidence door on eachside and superbtapestries and aboutmausoleums over tombsof holy men rugs. In spite of what has been usually as- shows that these were rare before the i ith sumed(Le Strange,Lands, p. 76; Creswell, century,and thereis, to my knowledge,no EarlyMuslim architecture, vol. 2, p. I I I), outsideevidence which would indicatethat the text does not say that the buildingwas a cult would have developed early around a square and that the dome was held on al-Hakim. As to the ornamentit need not four columns, an unlikely feature for a be of the gth centurymerely becauseit re- buildingof any size. It is muchmore likely semblesa Samarrastyle, for R. Ettinghau- that it was of circularplan with curtains sen has clearly shown that the "'beveled" between columnsand that it was open on style persisted for many centuries (The all sides. The plan has not been preserved "beveled"style in the post-SamarraPeriod, as suchfrom early Islamictimes, but it can ArchaeologicaOrientalis in MemorialEr- be presumedfor certain mosquefountains nest Herzfeld, Locust Valley, i952). It is (for instancein the mosqueof Ibn Tulun, true, of course, that the ornament of the tomb differs from the great ornamental 37 M. B. Smith, Imamzade Karrar at Biizu-n, designs known from Tirmidh in the i i th ArchEologischeMitteilungen aus Iran, vol. 7, I 93 S, and i 2th centuries,but there are other in- pp. 65 ff. THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES I7

Maqrizi, Khitat,vol. 2, pp. 268-269) and (d.943).To my knowledge these documents in a curiousand still unclearfeature of the havenot beenpublished as such;their ear- gth centurymosque in Nishapur(al-Mu- liest mentionknown to me is in Denike, qaddasi,p. 3I9); it canalso be recognized ArhitekturnyiOrnament, p. 8; they have in the commontholoi of Gospel manu- been recalledin variousother works, in scriptsthroughout the earlyMiddle Ages, particularthe importantstudy of V. L.Vo- andespecially in Armenia.38In otherwords ronina,Kharakteristiki arhitektury Sred- and this is an importantpoint to which nei Azij epohi Samanidov,Trudy Akad. we shall returnin conclusion-the form Nauk TajikSSSR, vol. 27, I954, pp.4I ff. suggestedby the textsfor the first qubbah The exact interpretationof these docu- at Najaf was not one whichwas precisely mentsfor the datingof the monumentis identifiedwith funerarypractices. It is ap- difficult to make, inasmuchas the waqf parentlyfrom the time of the Hamdanid statementmay have referredto a turbah constructionthat we maydate the develop- (on whichsee introduction)rather than to mentof Najaf into a hallowedcemetery in a qubbah.While an early date beforethe whichprinces and simple believers came to deathof Jsma4ilin 903 is not excluded,39it be buried.The first mausoleumwas re- mayseem preferable, from a strictlymeth- placed or transformedin 366/979-80- odologicalpoint of view, to usethe earliest 369/979-80 (Ibn al-Athir,sub anno, 369, availablearchaeological document and to unlessit is a referenceto the366 reconstruc- datethe monument from the reignof Nasr tion),burnt in 443/I05i, butused again in (9I3-943). In plan the monument is a 479/Io86 (Le Strange,Lands, pp. 77ff.). simplesquare with four openingsand a 6-Bukhara, Mausoleumof the Sama- large centraldome; there are four small nids, datablebefore 943. This celebrated cupolasin the cornersof the buildingover constructionis generallycalled the tomb of a gallery.The extraordinary brickwork has Isma'il(Creswell, Early Muslim architec- oftenbeen analyzed. ture, vol. 2, pp. 367ff.; L. Rempelin Bull. 7-Salamiyah, masbhad.An undated Amer.Inst. for PersianArt and Arch., vol. inscription, assumed on epigraphical 4, I935, pp. iggff.). The cleaningof the groundsto be of the firstpart of the ioth buildingin the thirtiesand discoveriesin century,refers to a mashhadpresumably Bukharalibraries brought to light three built (the verb has disappeared)by one new documentsabout it: a waqf rescript Abu al-Faraj'Abd al-Wahhabb. ... 'Ab- copied in i568-69 relatingthat Isma'il bas b. 'Abd al-Samad(Repertoire, No. had given land for the tombof his father 949). Salamiyahis known to have been Ahmadand that several princes were buri- a major shi'ite center, but the inscription ed there;several bodies in the tombitself; a fragment of a Kufic inscription on a 39 This is the date admitted by MadamePu- woodenplaque at theeastern entrance with gachenkovain her latest study, Mazar Arab-ata the nameof Nasr ibn Ahmadibn Isma'il v 7ime, Sovetskaia Arheologia, I96I4, p. 203, whereas earlier she had been less definitive, as in the important survey she wrote with L. Rempel, 38 C. Nordenfalk, Die SpiatantikenKanon- Vydaiuchiesia Pamiatniki Arhitektury Uzbeki- tafeln, Goterborg, I938, pIs. 24, 39, and passim. stana,Tashkent, I 9 5 8, p. 6S. is OLEG GRABAR doesnot say whetherwe are dealingwith al-Saghfr;his tomb is visitedto thisday...; an actual constructionover a holy place overit a baytwas erected, which is openon or, as is perhaps more likely, with some Mondaysand Thursdays." From a further featureassociated with a hallowed area. statementby the sameauthor (vol. 7, p. 90) 8-Baghdad, tombs of 'Abbasid ca- it canbe assumedthat no suchconstruction liphs. As in the cases of Fatimid and Bu- existedin 2I2/827-28. It has been asserted wayid tombs (nos. i6 and 29), the lack of that in 270/883-84, Ahmad Ibn Tulun remainingtombs makes an interpretation built a qubbabover the tombof Mucawi- of the texts ratheruncertain. With the ex- yah.40While this is not impossiblein the ception of the three princes buried in Sa- lightof IbnTulun's activities and beliefs,4' marra(see above, no. 2), the early 'Abbasid the majortext supportingit (Ibn Taghri- caliphs and membersof their families do birdi, al-Nujum al-Zahirah, Cairo, 1929, not appear to have been buried with any vol. 3, p. 47) simply says that the Tulunid particularformalities. By the 4th century prince erected four riwaqs around it (lit. of the hijrah,however, clear evidence exists "over it," 'alayhi, which does not make of one or severalturbahs reserved for them sense architecturally)and assignedpeople in Rusafah (Ibn al-Jawzi, al-Muntazam, to read the Koranby it and to keep candles vol. 6, p. 324). Some texts dealing with al- lit. This first constructionseems, therefore, Radi, al Ta'i, his motherand son, al-Mutl', to have been an enclosure rather than a seem to imply individual turbahs(Ibn al- constructionover the tomb and it may even Jawzi, vol. 7, pp. 79, I 39; al-Suli, Akhbar, be wonderedwhether the bayt of Mas'udi tr. M.Canard,Algiers, I946, vol. I, p. 238), implies much more than what Ibn Tulun while otherssuggest some sort of collective had accomplished.An explanationfor the necropolis (Ibn al-Jawzi, vol. 8, pp. I I3 choiceof Mu'awiyah'stomb can easily be 2I7; vol. IO, p. I28). Some constructions provided by the significanceof the caliph existed in or around these turbahs,which as an anti-shi'itehero.42 were endowed, but with the single excep- io-Aleppo, mashhadof Sabykh Mu- tion of al-Radi (d. 329/940-4I), there is no basin,dated 3 5 I/962. The presentbuilding evidenceof a monumentover the tomb it- is of the I 3th century (J. Sauvaget, "Deux self. As far as al-Radi is concerned,Yaqut sanctuairesshi'ites d'Alep, Syria, vol. 9, (Buldan,sub Rusafah) mentionsthe exist- I928), but an inscription referring to it ence of a qubbah.The date of its construc- as a mashhad has been preserved by Ibn tion is not known, to my knowledge, but Shaddad(Repertoire, No. I 5 57; in the edi- even though al-Khatib, for instance, does tion by D. Sourdel,Damascus, I95 3, p. 48; not mention it, it may have been built at the time of his death. The precisearchitec- 40 'Abd al-Qadir al-Rihawi, Qubur al-'Uza- tural characteristicsof the dome are not ma' fi Dimishq, Majallah al-Majma'al-'Ilmi al- known. 'Arabi, vol. 34, I954, pp. 648-649. 9-Damascus, tomb of Mu'awiyah, 41 0. Grabar, The coinage of the Tuiunids, New York, I957, pp. 33 ff. before 332/943-44. Mas'udi (Muruj, vol. 42 Ch. Pellat, Le culte de Mu'a-wiya au 5, p. I 4) relatesthat the first Umayyad ca- IIIeme siecle de l'Hegire, Studia Islamica, vol. 6, liph "was buried in Damascusby the Bab I956, pp. 53ff. THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES I9

E. Herzfeld, Materiaux pour un Corpus a squarebuilding with a singlerichly deco- Inscr.Arab.: Alep, Cairo, I956, vol. I, pp. ratedfacade higher than the side walls and I93 ff.), and a legendhas beenpreserved of its domeis carriedon the earliestknown the miraculousdiscovery by Sayf al-Daw- instanceof a muqarnassquinch. Thus the lah of the tomb of this obscureson of Hu- buildingis an importanttransitional one sayn. The term mashhad, of course, does betweenthe Bukharamausoleum and later not guaranteethat therewas an actualcon- "Seljuq"ones. structionover the tomb. I3-Mashhad, mausoleumof 'Alial- i I-Qumm, tomb of Mubammad b. Rid-, before 375/985. This celebrated Musa, 366/976-77. This sanctuary, like imamwas buriedby al-Ma'munnear the that of Fatimahin the samecity (seeabove, tombof Harunal-Rashid (Tabari, vol. 3, no. 3), seems to have begun as a family p. 1030); both tombswere well knownin maqbarah, eventually being transformed the gth century,but the earliestreference into a mashbad (Ta'rikh-i Qumm, pp. to majormemorial construction is by al- 2I5ff.). Muhammad himself died in 908, Muqaddasi(p. 333), who wrote that the but it is apparentlyonly in 366/976-77 that amirAmid al-Dawlah Fa'iq built a fort a dome was built over his tomb (modern withhouses and bazars and a splendidora- Ta'r-kh-iQumm, p. I 3 '; Houtum-Schind- tory(masjid) near the tomb. Although the ler, pp. 63-64). As in the instance of the textdoes not precisely say so, it isprobable tomb of Fatimah, this sanctuaryseems at that some constructionwas made over the this time to have been of local and paro- tomb.The amir is knownas earlyas 354/ chial significanceonly, since it is not men- 965, butit is likelythat he did not embark tioned by the geographers. on anymajor program of constructionun- i 2-Tim, mausoleumknown as "Arab- til 372/982-83, whenhis situationseemed ata," dated in 367/977-78. This recently moresecure. Since al-Muqaddasiwrote discovered mausoleum is of considerable around375/98 5, wemay date the buildings architecturalsignificance (G. A. Pugachen- aroundthese years. The buildings were de- kova, Mazar Arab-atav Time, Sovetskaia stroyedand rebuiltseveral times in the Arheologia, i96i4, pp. Ig8ff., where all courseof theI2th century. The first archae- other studies are mentioned,the most im- ologicaldocumentation from the shrine portant one being the account by the dis- consists in an inscription dated in SI 2/I I I 8 coverer,N. I. Leonov, in SoV. Arh., I96o', (Repertoire,No. 2978). TheMongols sack- pp. I 86ff.). It is in the mountainousregions ed the city,but propably not theshrine, for aroundthe Zerafshanvalley but it has not thereare inscriptions there from 6I2/1215 yet been possibleto identify the medieval (Repertoire,Nos. 3783-84,with an im- city-apparently from the ruins near the portantbibliography). mausoleum a minor one-with any one I 4-Mosul, sanctuaryof Jonah,second city knownfrom geographical descriptions. half of the i oth century.The historyof According to local lore, the mausoleum thissanctuary is stillunknown and so is the was erectedover the tomb of an earlyArab natureof its earliestconstructions. But al- conqueror.The inscriptionhas not yet been Muqaddasimentions that the daughterof publishedin its entirety.The mausoleumis Nasir al-Dawlahbuilt it up and endowed 20 OLEG GRABAR it (p. 146). A recent archaeologicalstudy i 6-Buwayhid tombs, second half of failed, however, to discover any certain the ioth century. The major evidence for constructionsearlier that the 6th centuryof the existence of Buwayhid mausoleums the hijrah (S. al-Daywahji, Jami' al-Nabi comes from al-Muqaddasi, who relates Yunis, Sumer, vol. IO, I954, pp. 260ff.), that, at Rayy, the Daylamite princesbuilt but could not do more than survey the high and solid domesover their tombs and presentbuildings rather superficially. lower rankingprinces erected smaller ones i 5-Kerbela, mausoleumof Husayn, (Muqaddasi,p. 2IO). On the other hand, before 369/979-80. Our earliest definite under the year 388/998, Ibn al-Athir tells evidence about the existence of major that $amsam al-Dawlah was eventually constructionsat Kerbelacertainly appears buriedin the turbahbani Buwayh, where- considerablyafter the fact. Already under as Adud al-Dawlah, Baha al-Dawlah, Harun al-Rashid financial support was and several other membersof the family given by the caliph'swife to people taking were buriedin the Najaf-Kufah cemetery, care of the tomb (Tabari, vol. 3, p. 752) near the shrineof "Ali;Ibn al-Jawzi refers and in 236/85O-SI, al-Mutawakkil had to at least 'Adud al-Dawlah's sepulchreas the tomb leveled and the buildingsaround a turbah(Ibn al-Jawzi, Muntazam,vol. 7, it destroyed (Tabari, vol. 3, p. I407); these p. 149). No definite archaeologicalevi- texts do not, however, indicate the exist- dence exists concerningthese mausoleums. ence of a mausoleum.By the time of Ibn Herzfeld claimsto have seen the copy of a Hawqal (978) and from Ibn al-Athir'sac- wooden door presumablyfrom the mauso- count of the year 369/979-80, it is evident leum of Adud al-Dawlah (MCIA: Alep, that a large mashhad existed to which p. I57, n. i), but his information is not 'Adud al-Dawlah had made repairsor ad- clear; it is also possible that the wooden ditions (Ibn Hawqal, p. 243). The archae- panels with unusualinscriptions from the ological evidence(A. N6ldeke, Das Heilig- year 363/973-74 (Repertoire,Nos. I83i- tum al-Husayns zu Kerbela,Tiurkische Bi- 32) may have come from Buwayhid ceno- bliothek,vol. I I, I909; M. Streck,Kerbela, taphs, but they do not imply the existence Festschrift E. Sachau, Berlin, 1915, pp. of large constructions.Since no archaeolo- 393ff., with interestingcomments on the gical evidenceexists and since the texts do possible pre-Islamicorigins of the cult) is not fully agree, the only safe conclusions practicallynil, but it may be suggestedthat to draw would be to assume that there no major construction existed in 9 5 I, since were Daylamite princely mausoleums al-Istakhri does not mention any. It is which may indeed have reflected by their quite likely that Adud al-Dawlah was re- size the importanceof the princesburied in sponsiblefor its recreation,as is suggested them,that thesemausoleums were probably by later sources (Mustawfi, Nuzhat al- grouped together, but also that they are Qulub, London, I9I9, p. 39). There is no more clearly ascertainablefor Iran than textual proof that an actual mausoleum for Iraq, where many of the principalBu- was built over the tomb, but, if we recall wayhids were buried. that this is the time when many othershi'ite i 7-Hebron, sanctuary of Abraham, tombs acquiredone, it can be assumed. before 98gS. The early history and develop- THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 2I ment of the greatestMuslim sanctuary 20-Chehar Juy (anc. Amul on the dedicatedto Old Testamentprophets and Oxus), anonymousmausoleum, late ioth to theirwives has been greatly complicated century(?). This small (6.20 by 6.20 me- by the Crusadesand the Muslimrecon- ters)construction with a heavy low dome quest (H. Vincent and H. MacKay,He'- is remarkablein severalways, even though bron, Paris, I92I, p. i5sff.). From the of very mediocreexecution. It has three mosttrustworthy early description, by al- doors,the qiblahside being provided with Muqaddasi,we can safely assumethat in a mibrab;two large pilasterstransform the latter part of the ioth centurythere one of the sidesinto a sortof embryonnic was a stone dome built by Muslims(al- faSade(Pugachenkova, pp. I77-I78). The Muqaddasiinsists on this point) over the proposeddate in the late ioth centuryhas tomb of Abraham,but presumablynot to be acceptedwith caution. over that of other prophets.The latter 2i-Near Merv,mausoleum known as were included in the nmashhadwhich con- Kizbibi(now disappeared),late ioth cen- tained also hostels,bakeries, and various tury. This simple construction(7.5o by otherinstitutions necessary for the upkeep 7.5 meters)has a singleentrance; inside of pilgrims(al-Muqaddasi, p. I72; G. Le thereis a domedroom with four deepre- Strange,Palestine under the Moslems, Lon- cessesgiving it an almostcruciform shape. don,I 890, pp.309 ff.). A date in the ioth centuryis apparently I 8-Sarakhs, mausoleum(?) of an suggestedby thebrick technique (Pugachen- uncleof 'Ali al-Ridg,before 985. Al-Mu- kova,p. I75). qaddasiwrites (p. 333) that there was a 22-Merv, mausoleum in the Imam tombat Sarakhson whicha masbhadwas bab cemetery, late ioth century (?). This built.It is likely that it includeda dome.43 buildingalso has now beendestroyed and I9-Central Asia, mausoleumof Ab- is only knownthrough photographs (Pu- mad, late ioth century.This building,a gachenkova,pp. I79-I 8o). It is archaicin simple squarewith apparentlytwo en- thatit hasfour openings, and its bricktech- trancesfacing eachother, is only known nique(all layersseem to be one insteadof throughan unmarkedphotograph remain- twobricks thick) differentiates it somewhat ing in Leningrad(Pugachenkova, pp. I78- fromother buildings assigned to thisperiod. 179, ill. p. I79). On its facade there is an 23-Mestorian, mazarShir Kabir, late inscriptionin brick,on which the word ioth century(?). MadamePugachenkova, abmadcan be read,hence the namegiven who has studiedthe building(Pugachen- to the building.It is builtin typical"Kho- kova,pp. i 68ff.; Iutake,vol.2, pp. I 94ff.), rasani"brickwork. Its dateis impossibleto proposesan earlierdate, perhapseven in determinewith certainty,but a very late the late gth century,but its comparatively ioth centuryone is not improbable. complexcharacter would rathersuggest a somewhatlater date, following in that 43 For reasons that do not appear clearly, B. MadameKrachkovskaia's analysis of the Spuler, Iran in FriihislamischerZeit, Wiesbaden, A. 1952, p. I8I, n. ii, suggests that the Sarakhs in epigraphy(V. Krachkovskaia,Evoluzia question was near Qazvin. The point, however, is kuficheskogopisma, Epigrafika Vostoka, not important for our purposes. vol.3, I949, p. I7). As it is now the mazar 22 OLEG GRABAR

consistsof severalparts, of whichthe most I939-4I, reproduced in Melanges d'his- ancientis a singledomed hall with a cupola toire et d'archeologie,Alger, I957, vOl. I, carriedon severalrecessed squinch arches; pp. I SI ff.). the walls were originally covered with 26-Cairo, mausoleumof Abu al-Fadl stucco,of whichonly a few fragmentshave Ja'far, 39I/IOOI. The evidence gathered remained;these can properly be relatedto by G. Wiet (MCIA Egypt II, p. IoI) per- i oth century stucco in Iran, as in Nayin mitted him to concludethat a mausoleum (Survey,pIs. 265 ff.); a singleentrance was had been built over the tomb of this mem- not on the axis of the building, which is ber of the remarkablefamily of viziers, the otherwise provided with a large mibrab. Banu al-Furat. The texts usually mention Whetherthis hall was the original mauso- the word turbahand it is possiblethat we leum over some unidentified spot or a have hereanother instance of a family plot prayer hall attachedto a tomb is not very with surroundingconstructions rather than clear from the available evidence, but the a true mausoleumover the tomb. later history of the site with several addi- 27-Gunbadb-i Qabus, dated 397/ tions to the original constructionand its I007. This earliest and magnificent ex- characteras a center for pilgrimageuntil ample of tower-tombis called a qasrin the the igth century strongly suggest that it inscription(Repertoire, no. 2 II 8). Its sig- had commemorative connotations from nificanceas a work of art and as a peculiar the very beginning.The city in which it is monument, at the same time mausoleum found has been identified as the Dihistan and royal symbol, has been fully analyzed of the Hudud al-'Alam (tr. V. Minorski, by E. Diez and Max van Berchem(Chura- London, I9375 p. I33) and of al-Muqad- sanischeBaudenkmiiler, Berlin, I9I8, pp. dasi (al-Muqaddasi,pp. 3 58-3 59). 39ff. and iooff.) and later by A. Godard 24-Mizdakhaneh (Kirghizia),i oth or (Survey,pp. 967 ff.). early i i th centuries.A small mausoleumis 28-Baghdad, mausoleumof Abu Dai- mentioned there by various authors (for 'ud b. Siyamard,presumably from the time instance, E. A. Davidovich and B. A. Lit- of his death in 399/Io08-09. Ibn al-Athir vinskij, Ocherki areologij. raiona Isfara, relatesthat he was buriednear the qabral- Trudy Akad. Nauk Tajik SSR, vol. 35, nudbur(see above, p. i o) and that its qub- I955, p. I 87), but the original publication bab was well known. has not been availableto me. 29-Resget (Mazanderan),anonymous 25-Biskra, Mausoleum of Sidi Oq- mausoleum,400/IO09-IO (?). This build- bah. Neither the date of the building nor ing, whose inscription has mostly disap- the original shapeof the mausoleumto the peared, belongs to the category of tower- great conquerorof the Maghribseem to be tombs on a circularplan (A. Godard, Les clearly ascertained;the earliestremains, a tours de Lajimet de Resget, Ahtar-e Iran, wooden gate datableca. I 000, may be of a vol. I, I936, pp. io9ff.). mausoleumbut also of an enclosurewith- 30 -Cairo, saba' banat, ca. IoIO.Four out major constructionover the tomb it- squarebuildings covered with domes still self (G. Mar,ais, Le tombeaude Sidi Oq- stand and two othershave been excavated. ba, Annales Inst., Et. Or., Alger, vol. 6, They are all alike and almost of the same THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 23

sizewith a dooron eachside; the largestis considerableamount of money was put providedwith two smallmihraibs, one on into the servicingwith candlesand chant- eachside of one of the doors.All the mau- ingof thetombs of the imamsin thepalace soleumsare surroundedwith an enclosure itself (Maqrizi,vol. I, p. 408). But most of with a singledoor, whose sill hasbeen pre- these texts are not clear on what precise served (Creswell, MAE, vol. I, pp. I07ff.). type of structure,if any, was built over the The buildingshave beenidentified with the tomb proper. Ibn al-Jawzi (Muntazam, seven mausoleumssaid by Maqrizito have vol. 7, p. i 56) does mention that in 380/ been built over the graves of membersof 990-9I the great vizier Ya'qub b. Yusuf, the Maghrebifamily executed by al-Ha- was buriedin the qubbahinside the palace kim (Maqrizi, vol. 2, p. 4S9). The event whichhad been preparedfor the caliph al- took place in i OI o and suchis the date sug- 'Aziz, but, since I do not know of other gested by Creswell for the mausoleums. texts describingthe palace restingplaces as Architecturally, of course, these simple qubbabs, one may wonder whether the constructionswith their squinchesand oc- work qubbahis not meant here in its later tagonal drums with windows can well be sense of "mausoleum"rather than in its assignedto the first half of the i ith cen- precise sense of "domed chamber," or tury. Whether, on the other hand, these whether Ibn al-Jawzi's informationis al- mausoleums were likely to have been togethercorrect. One may, then, conclude erectedimmediately after the executionof that greaterimportance was given by the the Maghrebisis perhaps less certain and Fatimidsto the tombsof their caliphsthan they might have to be dated after al-Ha- by the 'Abbasids.The existenceof funerary kim's death in I02I. chambersin the palace, while obviously 3 -Cairo, mausoleumsof the Fatimid related to the common practice of burials caliphs. From the point of view of our in- in houses, had an added significancein a vestigation of monumentalmemorial con- regimewhich emphasizedlineage so much. structions,the exact position of the tombs But there is no fully documenteddescrip- of the Fatimid caliphs is difficult to ascer- tion of these rooms and their actual shape tain. The main passagesin Maqrizi which is not clear. It is quite possiblethat the Fa- deal with the subject(Maqrlzi, vol. i, pp. timids followed to a degree, intentionally 407-408; vol. 2, pp. 49 and 442-443) are or not, the Byzantine practice of one or clear enough in indicatingthat the caliphs two major mausoleumsfitted into some and membersof their families were buried other construction(the churchof the Holy in special areas, the turbah al-Za'faran, Apostles in Byzantium, the palace in which was part of the palace complex, the Cairo), in which the tombs of the caliphs extreme south of the Qarafah cemetery and of membersof their familieswere put. (Wiet, MCIA Egypte II, p. I32), or, later, As in Byzantium, the rule was probably the area to the north of the Bab al-Nasr. It not absolute (Ph. Grierson, Tombs and is also clearthat mosqueswere built nearby, obits of Byzantine emperors,Dumbarton as the magnificent mosque built in 366/ Oaks Papers, vol. i6, I962). The subject 976-77 by the mother of the caliph al- deserves a more complete textual study cAziz (Maqrizi, vol. 2, p. 3I8), and that a than has been accomplishedso far. 24 OLEG GRABAR

32MtIl-i Radkan,mausoleum of the 36-Sangbast, mausoleum of Arslan ispahbadMubammad b. WandarinBa- Jadhib, dated 4I9/I028. So far as I have wand, 407-II/I0I6-2I. This remarkable been able to gatherthere is no absolutecer- exampleof the tower-tomb,quite different tainty about the identity of the personage from the slightlyearlier Gunbadh-i Qabus, for whom this mausoleumwas built, but its is called by three different namesin its in- simple square with a dome on squinches scriptions:mashhad (Repertoire, no. 2312), and a door on each side makes the date qasr(Repertoire, no. 23 I 3),44and gunbadh plausible (Survey, pp.923, 986-988, 1275, (in the pahlevi inscriptionread by Herz- and pl. 260 B, C; Diez, ChurasanischeBau- feld). Its architecturaland symbolicmean- denkmaler,pp. 52-55; Schroeder in BAI- ings have been discussedby Diez and van PAA, vol. 4, I936, pp. I36ff.). Berchem(Chursanische Baudenkmailer, pp. 37-Yazd, Duvazdeh Imam, dated 36ff. and 87ff.). 429/IO37. This square mausoleum with a 33-Laj%m, imam-zadeh cAbdallah, single door and a mibrabis of considerable dated 4I3/Io22. It is in fact the tomb importance for being the earliest known (qabr in the inscription, Repertoire, no. building in Central Iran with a muqarnas 2331) of one Shahriyar b. al-'Abbas and squinch(Survey, PP. IOOI-Io? , Pl. 274). another instance of the circular tower Its inscriptionshave not yet been publish- mausoleum (A. Godard in Athar-e Iran, ed, and no informationexists with regard vol. I, pp. IO9ff.). to the personagein whose honor the mau- 34-Damghan, Pir-i 'Alamdar, 4I7 or soleumwas built. 4I9/Io26-27 or 1029. Called both a qub- 3 8-Baghdad, Kazimayn,mausoleums bah and a qasr, this mausoleumto a bajib, of Musaal-Kazim (d. 800) and of Mubam- Muhammadb. Ibrahim, is also a circular mad al-Taqi (d. 2i9/834). The earliestin- tower-tomb(Repertoire, no. 23 52; Survey, formation known to me about this cele- pl. 339B; cf. also Bull. Amer. Inst. for Per- brated sanctuary is that, during the riots sian Art and Arch., vol. 4, I936, pp. of 443/I05I, it was plundered of its rich I 39 ff.). treasuresand that, after the plunder, the 3 5-Ghazni, tomb of Mahmud, in or two wooden cupolas and the sarcophagi shortly after 42I/I030. The exact shape of over the tombs of the imamswere burned, the mausoleumbuilt over the tombof Mah- one of the sarcophagihaving been opened. mud by his son Mascudis not known pre- At the same time most of the surrounding cisely, but it was most probably a cupola tombs were destroyed,including tombs of over a squareplan (M.Nazim, The life and Buwayhid amirs and of membersof the timesof SultanMahmud of Ghazna,Cam- 'Abbasid family (Ibn al-Athir, sub anno bridge, I93 I, PP. I24, 167). The epigraphi- 443; Ibn al-Jawzi, Muntazam, vol. 8, p. cal evidence derives from the sarcophagus i 5o). The extent of the establishmentthus and from wooden doors later taken to destroyed suggeststhat the sanctuary,in- India (Repertoire, nos. 2377, 2379, 2380). cluding the domes, had been established some " See the modified reading by E. Herzfeld, for time. The most likely period is PostsasanidischeInschriflen, Arch'aologische Mit- the end of the i oth and the beginningof teilungenaus Iran, vol. 4, I93 3-34, pp. I40off. the i i th centuries,in all probabilitybefore THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 25 the growthof the shrineto Abu Hanifah haraand Samarkand(Le Strange,Lands, (seenext item). p. 468). The buildingis a simplesquare 39-Baghdad, mausoleumof AbuHa- (4.50 by 4.50 metersinside) with a single nifah, 43 7-3 8/I045-47. The earliest refer- entrance,a developedfacade, and no gal- ence to some constructionby the tomb is lery.The originalpublication is by V. Nil- in al-Muqaddasi (p. I 30), who relates that sen, Mavzolei.. ., Materialypo istorij i one Abu Ja'far al-Zammamerected a suf- teoriy arhitekturyUzbekistana, Moscow, fah apparently to the side of the tomb; it I950; it was discussedby V. L.Voronina, was probably some kind of platform for p. 47 and L. Rempel,Arhitekturnyi orna- prayeror gatherings.The next information ment Uzbekistana,Tashkent, I96I, p. I 52, comes from Ibn al-Jawzi who relates bothof whomtend to datethe buildingin (Muntazam,vol. 8, P. 245) that in 437 or the late ioth century(cf. Pugachenkova, 438/I 045-47 new constructions were made p. 272). The buildingis providedwith an which includea roof (saqaf) over the tomb inscriptionof which,apparently, only the of the holy man; it is interestingto note basmalahis readable(V. A. Nilsen,Monu- that the work was sponsoredby Turks. In mentalnaiaArbitektura Bukharskova Oa- 459/io67, a complete reconstruction took zisa,Tashkent, I956, pp. 37ff.). It is diffi- place which involved strong foundations, cult to assigna precisedate to this con- a brickmausoleum, a madrasah,and other struction.However, because of the exist- institutions which required considerable ence of a rather developedfacade ex- acquisitions of land and the transfer of tendingbeyond the limitsof the walls of many tombs.Whatever the later historyof the building,a later date than the one the sanctuary may have been,45it would proposedby the discoverersseems pre- seem, then, that the tomb of Abu Hanifah ferable. had acquired some kind of architectural 4I -Aswan, anonymous mausoleums, recognitionas early as in the late i oth cen- datablein thefirst half or themiddle of the tury, but it is only in the early decadesof i ith century.A full descriptionof these the i i th that we have specific documenta- 49 buildingsand of the incidentswhich led tion for a constructionover the tomb. In to the disappearanceof their inscriptions IO67 a complete institution was created andtombstones can be foundin Creswell's there,one of the earliestof its kind in Islam, summaryof and additionsto Monneretde as has been shown by G. Makdisi. Villard'soriginal publication (U. Monneret 40-Kermin, mausoleumof Mir Say- de Villard,La necropolimusulmane di As- yid Bahr, datable in the first half of the wan, Cairo, I93 I; Creswell, MAE, vol. I, i i th century.The city is probablythe Kar- pp. I 3 I ff.). There is considerablevariety miniyah of medieval texts, between Buk- in the detail of the mausoleums,but a unity of basic forms, squares covered with a 41 G. Le Strange,Baghdad under the cAbbasid dome on squinchesand on a drum; some Caliphate,Oxford, I900, pp. i59ff.; M. Jawad, buildingshave four doors, othersonly one al-Nizamiyah, Sumer, vol. 9, I953, p. 324; G. Makdisi, Muslim institutions of learning in elev- and a mibrab.The constructionmaterial is enth-century Baghdad, Bull. School of Or. and usually cheap mud brick for the walls, Afr. Studies,vol. 24, 196I, pp. i9 ff. bakedbrick for archesand vaults.The main 26 OLEG GRABAR problemposed by the Aswanmausoleums at times, mashhads,as we have seenbefore is the evaluationof theirsignificance. The (see above, p.9; numerous examples of ratherprimitive construction and the large the relationshipbetween the ribat and bur- numberof buildingssuggest that we are ials can be found in the necrologiesof Ibn not dealinghere with the sanctuariesof al-Jawzi and Ibn al-Athir). For Aswan, single saints and heroes,as in the 'Alid we may then suggestthat the mausoleums shrinesof Iraq. For some reason,consider- expressone aspect of the "frontier"spirit, able numbers of people of more modest the identificationeither of those who had origins buriedin Aswan thought it appro- died for the faith or of those who guided priate to have mausoleumserected over warriorsfor the faith. Since many of them their tombs or had memorialsbuilt by fa- were of modest origins,since many of the mily or followers. Is it possibleto identify ribiztsand other related institutionswere some peculiarity of Aswan which would private foundations,and since no cults de- explain this development?Or is it merely veloped around most of them, the simple accidental that so many qubbahs have characterof the buildings,which were mere survived from there, whose existence we memorials,can easily be explained. It re- should assumefor other cities as well? To mains to be seen whether similarphenom- the last questionthe evidencepresented so ena occurredin other frontier areas.To be far and the rather detailed account which true, the Cilician and Central Asian fron- concernat least the cities of Baghdad and tiers have been the scene of so many in- Cairo seem to indicate that there was no vasionsthat many earliermonuments have generalpractice of mausoleumbuilding. It been destroyed, but the very remarkable is, therefore,something precisely related to numberof recentlydiscovered mausoleums the city of Aswan which must explain its from Central Asia could, partly at least, memorialconstructions. The peculiarityof be attributedto the same causes. Aswan was that it was on one of the pil- 42-Near Talas (modern Janbul), grimageroutes to Mekkah and that it was mausoleumof Babaji-khatun,first half of the last major Muslim city before Nubia. the i i th century (?). This rather unusual This latter qualitymade it a thughr,a fron- building is a square(6.8o by 6.8o meters) tier area (Maqrizi, vol. I, pp. I97-I99) with a curious ribbed low dome in brick, where holy men and warriorsgathered in an inscriptionon the facade which is said order to guard the frontiers of Islam, to to contain the name Babaji-khatun,and a partakeof an intensiveif little known spir- simpledecoration of archesand medallions itual life, and perhaps to die as martyrs. (A. Margulanand others,Arhitektura Ka- There were other institutionswhich devel- zakhstana, Alma-Ata, I959, pp. 97ff.).The oped aroundthe frontier areas,in particu- date is hypothetical and especially diffi- lar the ribat, a sort of military monastery cult to establish on stylistic grounds for which had been analyzed in detail in the so remote a region. Its justification resides Maghrib. In fact Aswan was known as a ribait.46Because of the holiness attachedto 46 J. Maspero and G. Wiet, Materiaux pour them, ribats were often places in or near ... Ia geographie de l'Egypte, Cairo, 194-19, whichpeople were buriedand they became, p. I5. THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 27 mainlyin the fact that the secondhalf of 46-Baghdad, mausoleum of the the centurywitnessed elsewhere in Central shaykh Ma'ruf al-Karkhi, before 459/ Asiaconsiderable modifications in decora- I067. The earliest information available tive techniquesand estheticvalues which to me on this mausoleumis that it burned arenot apparenthere. But, of course,argu- down accidentallyin IO67,being mostly in mentsof that ordercan only be usedwith wood (Muntazam, vol. 8, p. io5; G. Mak- cautionin dealingwith provincialcenters. disi, The topography of iith century 43-Rayy, circulartower-tomb, now Baghdad, Arabica, vol. 6, I959, p. 286; esp. destroyed.The inscription known through M. Jawad, al-'Imarah. . ., Sumer, vol. 3, Coste'stranscription (P. Coste, Monuments I947, PP. 54ff.). As early as 345/956-57 modernesde la Perse, Paris, i 867, P1. (Muntazam,vol. 6, p. 382, describingthe LXIV) and a photographby Curzon(G. burial of one Muhammadb. 'Abd al-Wa- N. Curzon,Persia, London, i 892, vol. I, hid al-Zahid in the suffah which faced the P. 350) has been publishedin the Re'per- qabr of the shaykh) some sort of construc- toire (no. 3I53) as being of 546. In reality tion existed in the area of the tomb, but it the first digit is missing and, if one con- does not seemto have consistedin a mauso- siders the extreme simplicity of the con- leum. struction, 446/iO54-55 may seem prefer- 47-Imam Dur, mausoleumof Muslim able. The tomb is called a qubbahand the b. Quraysh, ca. 478/IO86. Whether the builder was one 'Abd al-Jalil b. Faris, the shrine was actually built for the prince treasurer. whose name appears on the inscription 44-Damghbin, so called ChehelDukh- (Repertoire,no. 2756) or for Muhammad taran, dated 446/iO54-55. The exact date b. Musa, a son of the fifth imam (E. Herz- of this well-known circularmausoleum is feld, Damascus I, Ars Islamica, vol. 9, not very clear (Repertoire, nos. 2572-73; I942, PP. i8ff.), or for both, is not clearly E. Herzfeld, Khorassan,Der Islam,vol. I I, established.The building itself is a square I9I8, p. I68; Survey, pl. 340a), just as the covered with the earliest instance of the name of the personageburied in it is un- spectacularmuqarnas dome which will be certain.From the inscriptionand from the so characteristicfor Iraqi and Syrianarchi- popular tradition attachedto it, it seems, tecture in the following century. however, likely that it was built for a 48-Cairo, so-called mosqueal-Juyu- whole family. shi, dated 478/IO85. The recent study of 45-Abarqub, Gunbadh-i cAli, dated this well-known monument by K. C. A. 448/Io56-57. It is an octagonal building Creswell has posed anew the question of built for a descendant of the Firuzanid its purpose. Max van Berchem,who was dynasty, Hazaraspb. Nasr, by his son (Re- the first to have brought attention to it, pertoire,no. 2582). The building (Survey, had consideredit to be a mausoleumover a P. I023, Pl.-3 3 -336; A. Godard,Abarquh, tomb with an attached small sanctuary Athar-e Iran, vol. I, I936, PP. 49ff.) is consisting of a court, a hall for prayers remarkablefor being in rubblerather than with a dome in front of the mibrab, in in the more commonbrick technique of the otherwords a typical martyrium(Max van time. Berchem,Une mosqueed'poque fatimite, 28 OLEG GRABAR

Bulletinde l'InstitutEgyptien, vol. 2, i88 8; and We shall proclaimto you what you Notes d'archeologiearabe, Journal Asiati- usedto do"; XLVIII, I-5: "LoWe have que, 8th series,vol. I7, I89I, pp.479-48I; giventhee a signalvictory, that God may K. A. C. Creswell,MAE, vol. i, pp. i 56ff.). forgivethee of thy sin that whichis past It is quite true that the inscriptioncalls the andthat which is to come,and may perfect building a mashhad (Re'pertoire,no. 2753; His favourunto thee and may guidethee Creswell mistakenly calls it a zawiyah), on a rightpath; and that God may help but the mausoleum,which is found near it, thee with stronghelp; He it is Who sent has been shown by Creswell (p. I 57) to be down peaceof reassuranceinto the hearts a very late addition. It is, therefore, the of the believersthat they mightadd faith mosque itself which is a memorial struc- untotheir faith; God's are the hostsof the ture; in other words an architecturalform heavensand the earth, and God is ever createdfor prayeris used with a commem- Knower,Wise; that He may bringthe be- orative functionemphasized by the found- lievingmen and the believingwomen into ing inscription.The problemis to discover Gardensunderneath which rivers flow, what was being commemorated.An inter- whereinthey will abide, and may remit esting peculiarity of the shrine is the un- from theirevil deeds;that in the signsof usual choice of Koranic quotations found Godis the supremetriumph"; and finally, in its inscriptions,as Max van Berchemhas XXXV, 39: "He it is who madeyou re- already noticed. The two passageson the gents [khala'if]on the earth;so he who dedicatory inscription on the facade disbelieveth,his disbeliefbe on his own (LXXII, i 8, and IX, IO9) are standard head;their disbelief increases for the dis- passagesfound in mosquesand refer to the believers,in their Lord'ssight, anything masaijidAllah. But the passagesfound in butabhorrence; their disbelief increases for the domedroom in front of the mibrabare, the disbelievernaught but loss." In the with one exception (XXIV, 36-37), much centerof the cupolathe namesof Muham- rarerin monumentalinscriptions: XI, 36: mad and 'Ali are intertwined.The quota- "My counsel,if I want to counselyou, will tions from the Koranall emphasizetwo not profit you, if God wants to keep you points:the doomprepared for slanderers, astray; He is your Lord and you will be disbelievers,dissenters, and rebels, and the returnedto Him"; XXIV, i I: "Thosewho victorygiven to theProphet, to thecaliphs, spreadthe slanderare a small groupamong and to the puremen for whomthe build- you; do not think it a bad thing for you; ing was erected.If one recallsthe remark- in fact it is good for you; to every man able positionof the mashhadon the edge who carried it what he has earned of the of a cliff overlookingthe whole city, it sin and to the one who had the greatest may be suggestedthat its purposewas to share an awful doom"; X, 24: "Yet when symbolizethe victoryachieved only a few He had deliveredthem [peoplein a storm] yearsearlier by Badral-Jamali in thename behold they rebeledon the earth with un- of the caliphal-Mustansir over the rebel- truth;0 men,your rebellionwill only turn lions and disorderswhich for a long time againstyou; a [brief] enjoymentof the life plagued the Fatimid empire. Like the of the world, then unto Us is your return Dome of the Rock,then, the "mosqueal- THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 29

Juyushi,"as an expiatorychapel and a governorat the time of Nafisah'sdeath symbol of victory, was a buildingcom- (206/82I-22), 'Ubaydallahb. al-Sari,had memorativeof a precisehistorical event. already built a cenotapharound the tomb As the eventreceded into timeand lost its with marble plaques, whose inscriptions pungency,the memoryof the purposeof have been assumedto have been copied on the buildingfaded away and, in orderto a later refection (Re'pertoire,no. I 62, with fit withthe more common commemorative importantcomments by G.Wiet in MCIA: constructions,it acquireda funerarysense Egypte II, pp. 33ff.). It is doubtful, how- and a smallmausoleum with a tombwas ever, whether a real construction was evenadded to it. erected at that time, not only becausethe 49-Aleppo, maqam Ibrabim, 479/ generalevidence collected here would con- io86. Only an inscriptionwith the names tradict it, but also becauseMaqrlzi trans- of Malikshahand Nizam al-Mulk(Re'per- mits the story strictly as a rumorwithout toire,no. 2760; E. Herzfeld,MCIA Alep, documentation.Furthermore, the remain- p. I77) reflectsthis early sunnite memorial ing inscription, presumably an Ayyubid constructionin Aleppo(J. Sauvaget,Alep, copy, is suspicious by its use of the word Paris, I94I, p. I07). mashhad and by its curiously antiquarian o-Salamiyah, mashbadof 'Ali b. characterin that, as G. Wiet has pointed Jarir (?), 48i/io88. This mashbad,pre- out, the Koranic quotation stops precisely sumablybuilt over the tomb of an ob- on the spot where, accordingto the legend, scureofficer, is only known throughan the holy woman had stopped when she inscription(Repertoire, no. 2772). died. There is a literary flavor to it which 5I-Tunis, mausoleumcalled "Sidi makes one question its archaeological Bou Khrissan,"486/I093. This small value. The second inscription from the squarebuilding was apparentlyopen on shrine also comes from Maqrlzi (vol. 2, all foursides and is datedby an inscription p. 442; Repertoire, no. 2776); it refers to (S. M. Zbiss, Documentsd'architecture the constructionof a door under al-Mus- fat'mited'Occident, Ars Orientalis, vol. 3, tansirand gives a precisedate. Its archaeo- I959, p.30). logical index is muchgreater than that of 52-Tunis, cupolaknown as Msidal- the first inscriptionand, while it is possi- Qubbah,undated, but typologicallyrelat- ble, even likely, that there were earlier ed to the preceding(Zbiss, Ars Orientalis, constructionsover the tomb of the holy vol.3, p. 30). woman, they cannot be dated more pre- 53-Cairo, mausoleumof Sayidah cisely. Nafisab, 482/I089-90. This celebrated 54-Ascalon, mashhadof the head of Cairenemausoleum no longerremains in Husayn, 584/Io9I-92. Nothing is known its originalshape, which was a very simple of the shape of this great Fatimid mauso- squareconstruction with a mibraband one leum and much in its purpose is still not or threedoors (D. Russell,A note on the clear. Even its exact date is a matter of cemeteryof the 'Abbasidcaliphs, Ars Is- controversy,although the very clear epi- lamica,vol. 62, I939, pp. i68ff.). Maqrizi graphicalindications of the minbarmade (vol. 2, p. 442) reportsthat the 'Abbasid for the shrine and eventually brought to 30 OLEG GRABAR

Hebronmake it certainthat the building wall whichmay have been higher than the existedin I09I-92 (Repertoire,nos. 2790- dome,and an extraordinarydecoration of 9I; M.V. Berchem,La chairede la mosquee terracottafragments in whichornamental d'H'ebron,Festschrift E. Sachau,Berlin, principlesof Samanidtimes appear to have I9I5, pp. 298ff.; G.Wiet, Notes d'epigra- beenexecuted in a new decorativetechni- phie, Syria, vol. 5, I924, pp. 2I7 ff.). que (A.Margulan and others,Arhitektura 55-Mehmandust (nearDamghan); a Kazakhstana,pp. 99ff.; Denike,pp.98 ff.). ten-sidedtower-tomb was found there with Likethe groupof buildingswhich follows, a date whichwas readas 49 x by Schroe- this mausoleumis difficult to date and der (E. Schroeder,BAIPAA, vol. 4, I936, couldin fact be put anywherein the i ith p. I35, fig. 4) and publishedas 490 by century. Godard(A. Godard,Athar-e Iran, vol. 4, 58-Sayat (on the right bank of the I949, p. 259, fig. 2IO). Thebuilding is thus lower part of the Kafirniganvalley in to be dated ca. i096. Tajikistan). Two mausoleums (ca. IO.50 56-Iarti-Gunbadh (near Sarakhs), metersto the side) are preservedthere and dated49I/I098. Thiscomparatively large are connectedwith eachother by a vaulted (I2 by I 2 meters)mausoleum is remarkable passageway.The interest of the building in manyways. First it is one of the few resides in its brickwork,especially in the early CentralAsian mausoleumsto pro- squinches,and in a i 6-sidedzone above the vide us with a dated inscription. Second, squinches.Parts of its faSadewere provided its zone of transition from square to dome with a decorationwhich is certainly later. hasa rathersubtle variation on themuqar- The mausoleumsare not dated (A. M. Be- nas type first seen at Tim and Yazd (see lenitzkij, Mavzolei u seleniiaSaiat, Trudy above, nos. I2 and 37) and later through- Sogdiisko-TadjikskoiArheologicheskoiEk- out Central Iran. Third, it has elaborate speditzij,vol. I, Moscow, I 9 5 0, pp. 207 ff.; decorative compositions on all four sides, also in KSIIMK, no. 33, I950, unavailable even though there was only one entrance; to me). the type of composition (cornercolumns, 59-Sarakhs, mausoleumof Abu Fadl. pilasters framing the central part, etc.) This very monumental construction,de- relate the monument to the ioth century scribed in detail by Madame Pribytkova Samanid monuments rather than to the (A. M. Pribytkova, Pamiatniki Arhitek- tradition of a single faSade typical of i i th tury XI veka v Turkmenij,Moscow, I 9 5 5 century monuments in the area. Fourth, pp. 6ff.), consistsof a squareof heavy brick for unexplained reasons, its eastern faSade masonry carrying a gallery and a superb was more elaborate than the southern one dome; while its sides are decorated with on which the entrance is found (G. A. Pu- blind arcades,its faSade is of a new type gachenkova, Iarty-Gumbez, Epigrafika with projecting towers making a sort of Vostoka, vol. I4, I96I, pp. I2ff.). entrance eywan. It has been dated by 57-Aysha-bibi mausoleum,near Ta- MadamePribytkova and otherswho have las. This heavily decorated brick mauso- followed her in the second quarterof the leum is comparatively small (8 by 8 meters) i ith century. The personagefor whom it but remarkable for two features: a facade is supposedto have been built (see refer- THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 3I encesin reviewby 0. Grabar,Ars Orienta- count of the Tirmidharchaeological ex- Us, Vol. 2, I957, p. 545) died in i023, but, pedition (in Trudy Akad. Nauk Uzbek so long as there is no epigraphical or tex- SSR, ser. I, Tashkent, I945, pp. ig6ff.). tual justification for its attribution, the From this account it would appear that identification, which is of popular origin, several mausoleumsmay, on architectural is hardly more than a terminus post quem. grounds,be dated fairly early, but the only On the other hand, the tremendous devel- one which has been discussedmore fully opment of the faSade relates this building (Denike, pp. I2-I4) shows close similari- to definitely dated 12th century examples. ties to the preceding monuments and It is, therefore, mostly because of the so- should be dated in the late IIth century. briety of the ornamentation and of the 63-Mausoleum at Baba-Gamber decoration of arcades on the sides of the (small town on the river Mug). Madame building that a late Iith century date is Pugachenkova(pp. 273 ff.) has discovered tentatively proposed. a photograph of a mausoleum now dis- 6o-Mihnah (near mod. Mean; Le appeared;it is a rathersimple construction Strange, Lands, p. 394), mausoleum of Abu with a high facade which can be dated in Sa'id. In plan this other remarkable mau- the I Ith century. soleum is similar to the preceding one (Pri- 64- Uzgend, so-called middle mauso- bytkova, pp. 20ff.; Pugachenkova, pp. leum.This simplesquare building is squeez- 272 ff.). Some of its proportions are differ- ed betweentwo later and dated buildings. ent and its faSade and side decorations are Cohn-Wiener (E. Cohn-Wiener, Turan, quite different. Its most remarkable char- Berlin, I930, pp. I8, 35) had thought that acteristic, however, is that it has a double it might be of the early I Ith century,per- brick dome. As in the preceding example haps the tomb of one of the first Karakha- and for somewhat the same reasons, a late nids (cf. also Denike, pp. I4ff.). Its large II th century date seems preferable to the gate and extensivedecoration make it more earlier date proposed by some of the Rus- likely to be a somewhat later monument. sian scholars who have written about the 65-Vekil-Bazar, in Turkmenistan, building. mausoleumof 'Abdallzhibn Burayda,late 6i-Astan-baba near Karkhi (Le II th century. This small mausoleum with Strange, Lands, p. 404, so-called mauso- two Kufic inscriptions,one in brick, the leum of Alamberdar. This is another in- other in stucco, is not dated and is only stance of an anonymous mausoleum of known to me through a referencein Epi- heavy proportions with a slightly devel- grafika Vostoka, vol. i6, I963, p. I49. oped faSade, a low and wide dome, and 66-So-called Shaburgan-ata(in the most original brick squinches (Pribytkova, area of Bukhara).This very curiousbuild- pp. 77 ff.; Pugachenkova, pp. 268 ff.). ing is octagonal with a large protecting 62-Tirmidh, mausoleum in the ceme- fagade (Nilsen, pp. Ss fI.). It is dated tery known as Sultan-Sa'dah. The only around i 00 becauseof the simplicity of publication available to me that deals with its decorationand of its methods of con- the religious buildings of Tirmidh is the struction.Nilsen mentionsalso, p. 6I, two poorly printed second volume of the ac- other similar buildings but without any 32 OLEG GRABAR reference.The plan is curiouslyclose to kova, pp. 292ff.). These mausoleums, that of the Jabal-i Sang near Kerman whichmay have fulfilled a functionsimilar (Survey,pl. 28 I). to thatof theAswan examples, have rather 67-Asterabad (Mazanderan),mauso- curiousforms. In plan they arecircular or leum.This buildingknown from an illus- octagonal,with, in a few instances,thin tration in Diez, Persien,Islamische Bau- half-towersor star-likepoints. In elevation kunst in Churasan,Berlin, I923, pl. 8, is a they arevery squatand curiously tapered, low octagonalconstruction and has been somewherebetween the tower and the relatedby E. Schroederto ourno. 55, butit polygoncovered with a dome. is possiblethat it is later. 7 i-Ferav (southernTurkemistan). 68-So-called KhojaChisht (near He- Two smallmausoleums were found, one a rat). Two squaremausoleums are found squarewith projectingentrance, the other there, apparentlywith two entranceson also a squarewith large faSadebut pro- one axis, a double dome, and a very good videdinside with a curiousseries of recesses decorativedesign of bricks.The fragments on fourdifferent axes. (Pugachenkova, pp. of inscriptionswhich have been published 299 ff.). are not easily legible and do not seemto be 72-Kufan (LeStrange, Lands, p. 394). of a historical character (0. V. Nieder- The ruinsremain of a smallsquare mauso- mayer, Afghanistan, Leipzig, I924, pp. leumwith a longentrance iwazn (Pugachen- 62-63, pls. I 82-I 84). kova, pp. 30I-303). The mausoleumwas 69-Mausoleum of MuhammadBash- redonein the secondhalf of the I 3th cen- shar, near Kolohozhion (in Tadjikistan). tury, as is attested by inscriptions-to my This complex buildingis in fact a mosque- knowledge unpublished-on its stuccoes, mausoleum.It is divided into three parts: but the originalconstruction is supposedto a large central domical room with a later be earlier. portal and a mibrabto the right of the en- 73-Talas, so-called mausoleum of tranceand two side areaswhich are partly Karakhan.This mausoleum,popularly at- covered with domes, one of which con- tributedto the founderof the Karakhanid tained two tombs. The zone of transition dynasty (cf. above, no. 64), was destroyed has a sort of pendentive in brick which is in the early part of this century. On the comparativelyrare in Central Asia (L. S. basisof remainingphotographs it is rather Bretanitski, Ob odnom maloizvestnom difficult to decide on its date, but its deco- pamiatnike, Materialy i Issledovaniia po rative designs(Denike, p. I 5) are certainly arheologij SSR, vol. 66, I953, pp. 325 ff). later than the beginning of the dynasty, The date is not given but around i ioo and, while an early core to the building is is suggested as plausible on comparative conceivable,the visible parts can hardly be grounds. datedbefore i I 00. 7o-Dihistan, mashhad. The name 74-Astan-baba (cf. above, no. 6i). A mashhadis given to a large cemeterynear shrine stands there consisting of four Dihistan, where seven mausoleumsare still domedrooms of differentsizes fitted with- standing,of which five are datablearound in an irregularrectangle (Pugachenkova, I IOO on architecturalgrounds (Pugachen- pp. 286ff.). Various added constructions THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 33 suggestthat this was a mashhadof some squaremausoleum of mediocreconstruc- significance.The various legends which tion was found whose systemof squinch exist with respect to the identity of the archesrecalls that of the latermausoleum peopleburied there do not helpin explain- of Sanjarin Merv, hence the suggested ingits originalfunction or evenshape. One date at the beginningof the I 2th century of the domesis set on a very developed (E. A. Davidovich and B. A. Litvinskij, type of bricksquinch. The mausoleumhas Ocherkiarheologij raiona Isfara, Trudy beendated in the ioth century,but a date Akad Mauk Tajik SSSR, vol. 35, 1955, aroundII00 seemspreferable because of pp. 185ff.). the extentof the wholecomplex (B. A. Lit- 78-Mausoleum also known as Khoja vinskij, Arhitekturnyikompleks, Trudy Rushnai, in the southernpart of the Sur- Akad.Nauk TajikSSR, vol. 17, I953, pp. khan valley, not far from Tirmidh. It is a 3 I ff.). square building with a projecting faSade 75-Irayadh, a village in Kuhistan. and two side entrances; the dome is on Accordingto Yaqut(sub voco), a khanqah squinchesconsisting of a seriesof recessed for sufis was there and by it a mashhad arches.On comparative grounds the dis- with a qubbahover the tomb of shaykh coverer of the building has suggestedthe Abu Nasr al-Zahidal-Irayadhi, who died late I I th centuryas a probabledate (V. A. after 500/I006-07 (cf. E. Herzfeld,Kho- Nilsen, Nekotorye syrtzovye... postroiri, rasan, Der Islam, vol. ii, I918, p. i68). Istoria Materalnoi Kultury Uzbekistana, Yaqut does not give us the date of the no. 3, I962, pp. I02ff.). buildingand it is only a suggestionthat it 79-Turtas gunbadh,in the same area was erectedin the first decadeof the ioth and of a similartype (ibid., pp. I o7 ff.). century. It is interestingin illustrating a 8o-Uiuk-gunbadh, also in the same phenomenonfor whichthere probably are area,but differentfrom the precedingones otherexamples in texts:the honoringof a in that it has no axial entrancebut a side local holy manthrough a mausoleumand one; on the other hand, its squinchesare of throughsome philanthropicor pious or- a rather peculiar type, as though some ganization.Conversely many of the anon- local craftsman was trying to copy the ymous buildingswe have listed may be newly developed Iranian muqarnaswith- explainedin similarterms. out quite understanding its significance 76-Paahtabad (inTajikistan).Of two (ibid.,pp. I o9 ff.). It may be that this build- mausoleumsdiscovered there, one, a simple ing is a little earlierthan thepreceding two. square construction with an embryonic 8I -Alasher-khaneh, in the Kengir portal,has beendated in the first decade valley of Kazakhstan.Built in bakedbricks, of the I 2th century (B. A. Litvinskijin it is a squarebuilding with heavy "pylons" Trudy Akad. Nauk TajikSSR, vol. I7, on the main facade. Its walls are heavily 1953)) decorated in "brick-style,"and its exact 77-So-called KhojaRoshnay, in the date is uncertain (A. N. Margulan, Arhi- area of Isfara (V. Barthold, Thrkestan tekturnye pamiatniki raiona r. Kengir, down to the Mongolinvasions, 2nd. ed., Krat. Soob. Inst. Ist. i Mat. Kult., vol. 28, London, I958, pp. i6o-i6i). A simple I949,Ppp.4546). 34 OLEG GRABAR

82-Cairo, Khadra Sharifah, 5OI/ was involved,but the combinationmauso- II 07. This buildingwith an entrancecom- leum-madrasahbecomes so commonduring plex, a courtwithout porticoes,and a sanc- the I 2th centurythat it is possibleto inter- tuary with threehalls, the centralone being pret the Bu$rastructure as an early ex- covered with a dome, poses a problem. ample. When it was cleared, tombs were discov- 84-Aswan, mashhad,ca. I Io.i This ered in the rooms to the side of the dome remarkablestructure bears a close resem- (Creswell,MAE, vol. I, pp. 224-226); the blancein plan to the Juyushimosque (see whole constructionshould then be inter- above, no. 48). Since it has preserveda preted as a funerary compound. On the cenotaph,it mustbe assumedto havebeen other hand, Maqrizi'stext used to identify a funerarybuilding, although the identity the buildingand to date it (MaqrIzi,vol. 2, of the personageinvolved is not known P. 452) refers to this and other similar and thereis no absolutecertainty that the structures as masjids. If the tombs are cenotaphis contemporarywith the build- indeed comtemporarywith the building, ing (Creswell,MAE, vol. I, pp. 222-224). the use of the word masjidwould indicate The problemhere is in realitythe sameas an interestingextension of the meaningof for no. 78. the word or a changein the understanding 8 - Damascus,mausoleums of Safwah of the building between the time of its al-Mulkand of Duqaq,504/I I IO-I I. The foundation and the I5th century. If the mausoleumof Safwah al-Mulk, now dis- tombs are not contemporary,the building appeared,consisted in a centraldome sup- shouldbe relatedto the categoryof orato- ported by two half-domes (J. Sauvaget, ries built in cemeteriesbut not primarily Les monumentsayyoubides de Damas,vol. for memorialpurposes, which I have dis- I, Paris, I938, pp. I ff.). Its inscriptionis cussed in my introduction. By Maqrfzi's of greatinterest for the historyof funerary time, of course,many of these distinctions and commemorativeterminology; it begins had become somewhat blurred and it is as follows: "Has orderedthe construction possible indeed that it is in the later Fa- of his mashhadand of the turbahwhich is timid period that the tremendousgrowth in it..." (Repertoire, no. 2942, with the of an architecturein cemeteries(Creswell, correctionalready made by Sauvaget,p. 8, p. 226) led to the convergence of two n. 8). As Sauvaget has reconstructedit on originally separatearchitectural and func- the basis of literary and topographical tional traditions,the funeraryoratory and evidence, the constructionof Safwah al- the mausoleum. The Khadra Sharifah Mulk, made in her life-time (she died in would then be one of the earliestexamples i i I 9), was fitted into an existing complex of a new type of building. of buildingswhich included the mausoleum 8B3Busra, mausoleum(?) of shaykh of her son Duqaq (who died in II 04), an $afi al-Din, before 503/I0og-io. An in- oratory, and perhapsa khanqah for sufis; scription on a sarcophagusrefers to the all of them made up the mashhadand it is waqf of a madrasahand to the renovation likely that the introductioninto it of a new of a turbah(Re6pertoire, no. 2932). Thereis mausoleum required certain refections no absolute certainty that a mausoleum (hencethe word 'imnirahin the inscription, THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 35 whichcan referto a new buildingas well the halls was an oratory.Its walls were as to a reconstruction).The word turbab paintedand the remainsare acceptedas appearsto referto the mausoleumSafwah beingof theearly I 2thcentury. The faSade, al-Mulkbuilt for herself;if so, it wouldbe as reconstructedby MadamePugachen- one of the earliestinstances of the clearly kova, had a compositionbased on three documenteduse of the word for a domed arches,a comparativelyuncommon motive building over a tomb. It is possible, on the in Iranianarchitecture of thattime. other hand, that the mashhad refers to the 87-Aleppo, mausoleumof the qadi mausoleum and the turbab to a cenotaph.47 Abu-al-Hasan, 508/I I I4-I 5. Known only 86-Merv, mausoleumof Mubammad throughIbn Shaddad(ed. D. Sourdel,Da- b. Zayd b. 'Ali Zayn al-'Abidin, 506/1 II2 mascus, I953, P. 35), it is called a turbah tO I 3. It is only recently that the inscrip- and the text insistson the remarkablequal- tion of this long known mausoleum has ity of its stones. There is no absolute cer- been entirely deciphered, but, to my know- tainty that it was domed. ledge, M. E. Masson's reading has never 88-Damascus, cemetery Dabdah, been published (it is paraphrased with a 5I4/II20. An inscription mentions a qub- discussion of the building in Pugachen- bah built by the mother of Fakhr ak-Din kova, pp. 304ff.). The mausoleum was Buri (Repertoire, no. 298I). built by the local governor Sharaf al-Din 89-Cairo, mausoleum of Sayidah Abu Tahir b. Sa'ld and is referred to in the 'Atikah, ca. II20. This small squaremau- inscription as a mashbad. Like the Astan- soleumis to an aunt of the Prophet (Cres- baba complex (see above, no. 74) it is a well, MAE,vol. I, pp. 229 ff.). couriously composed construction of four go-Cairo, mashhad of Umm Kult- domical rooms of varying sizes-the largest hum, before 5i6/1i22. Only a mibrab re- of which was over the tomb of the 'Alid- mainsof this building,and the date as well arranged in an irregular rectangle. One of as the identification is hypothetical but plausible (Creswell, MAE, vol. I, pp. 239 47 Theremay be yet anotherpossibility. The to 240). The textual referencewhich per- text of the inscription(now gone),as it was pub- mitted the identification (Ibn Doukmak, lishedin theRe'pertoire from van Berchem'snotes, begins:amara bi-'imarah hadha al-masbhad wa Description de l'Egypte, ed. K. Vollers, al-turbahqubbah(?).... This is obviouslyim- Cairo, I893, vol.4, p. I2I; Maqrizi,vol. 2, possibleand Sauvagethad alreadyproposed the p. 442) relates that seven mausoleums were obviouscorrection of qubbahinto fihi, following redone at that time; the other six have in this a note left by van Berchem.If the inscrip- probablydisappeared. tion had been perfectlyclear, as experiencedan epigraphistas Max van Berchemcould not have 9 -Qus, mausoleum,ca. II25. The been misledinto the inconsistencyshown in his date of this simple qubbahis based on its originalreading. It may be wonderedwhether architectural characteristics (Creswell, al-qubbahshould not be readinstead of al-turbab, MAE, vol. I, pp. 236-238). a possiblesuggestion if one simplyassumes that 92-Cairo, mausoleumknown as ikh- the beginningof the nounwas not clear.Were the inscriptionto read... al-masbhadwa al-qubbah wat Yusuf, ca. I I 25. Similarto the preced- fihi, the difficultyin understandingturbah as ing and dated at the same time (Creswell, referringto a domedmausoleum would be gone. MAE, vol. I, 234-236). 36 OLEG GRABAR

93-Cairo, anonymous mausoleum, culiaritiesand obviously legendaryele- first third of the 12th century. This small mentswhich have beendiscussed in detail square constructionin front of the khan- by G. Wiet. If we havepreferred the date qizh of Baybars (Creswell, MAE, vol. I, of 525 to 549 it is merelythat the mira- pp. 227-228) belongsto thesame group. culousinvention of the head of an ima-m 94-Cairo, mausoleumof Mubammad is likely to have been followed by the al-Ja'fari, before II25. Another instance buildingof a shrineand that a waiting of the same type (Cresswell,MAE, vol. I, period of almosta quarterof a century pp. 232-234). seemsrather meaningless. The monument 95-Cairo, mausoleum known as containedan oratorywith court,an eywan, shaykh Yunis, before II25. This mauso- and a domedmausoleum, but mostof the leum, typologically related to the preced- presentstructure is modern. ing ones, had, until Creswell'sstudies, been 99-Cairo, mashhadof SayidahRuq- thought to be the mausoleumof Badr al- qayah,527/II33. This mausoleumwith a Jamali(Creswell, MAE, vol. I, pp. 232 to central domed hall, two side halls, and a 234). porticoedante-hall, has often been studied 96-Huday-Hazar, some28 kms.from (Creswell,MAE, vol. I, pp. 247ff.; and esp. Bayram 'Ali in Turkmenistan,near the G. Wiet, MCIA: Egypte II, pp. i95 ff.), ruins of a medieval village. This simple and G. Wiet, in particular,has broughtout squarebuilding is remarkablefor its large all the complex features of Fatimid piety facade and for its sober decoration(Puga- whichare involved in it. chenkova,pp. 3Ioff.). i oo-Sohag, 529/I 1 34. An inscription 97-Aleppo, mausoleumbuilt by Amr publishedin the Repertoire(no. 307I) re- b. abi al-Fadl, 5 22/I I28. The curiosity of fers to some construction('imarah) which this mausoleum,known only through an served to commemoratea victory of the inscription (Re'pertoire, no. 3027), is that army commandedby the son of the caliph it is called an eywan. al-Hafiz. Althoughthe eventswhich are so 98-Cairo, mausoleumof the head of commemoratedare not recorded and the Zayn al-AbHdin, 525/I130-3I. A rather nature of the constructionunknown, we complex set of problemsis connectedwith have here another instance of a rare type this mausoleum(Maqrizi, vol. 2, pp. 436, of memorialconstruction, the victory mo- 440; G. Wiet, MCIA: Egypte II, pp. nument. 21 3 ff.). First, as Maqrizi already pointed ioi-Madinah, mausoleumof 'Abbas out, the mausoleumwas built in fact for b. 'Abd al-Muttalib and of al-Hasan b. Zayn's son, Zayd. Second, the date 525 is 'Ali,before 5 29/I I 34-3 5. Ourmajor source the date of the invention of the head and for the monumentsof Madinahis Samhudi, the earliest referenceto a constructionis who relatesthat there is some argumentas in a bad Ottoman copy of a Fatimid in- to the time when this mausoleumwas built scription giving the date 549/I I 54 (Re6per- but opts for 529 and the caliphate of al- toire, no. 3I63). Third, the historical cir- Murtashid (al-Samhudi, Kitab wafa' al- cumstancesof the invention, as they are wafa', 2 vols., Cairo, I326 H., vol. 2, p. told by Maqrizi, contain a numberof pe- ic i). The building had a high dome and THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 37 two doors,one of whichwas open all day.48 with a dome on muqarnasand a superb So far as I have been able to discover,this fagade,is empty, but its inscriptionhas is the earliestcertain instanceof a mauso- preservedthe namesof the patronand of leum in Madinah, but it is possible that a the builder (Repertoire,nos. 3I35 and numberof Alid mausoleumswere destroy- 3I36; A. Godard,Les monuments de Ma- ed in earliertimes. ritgha, Paris, I937, p. 4). I02-Mazar-i Sharif, mausoleum to io6-Turan-pusht (betweenAbarquh 'Ali, 53 0/I I36. Although no part of this ce- and Yazd), Gunbadh-ishaykh Junayd, lebratedTimurid building seems to be asear- 543/II48-49. An octagonal tomb. The ly as the I2th century,a sanctuarywas ap- planand photographs have been published parently createdthere as early as I I 36 (0. by A. Godard (Les coupolesiraniennes, von Niedermeyer,Afghanistan, pp. 65 ff). Athar-6Iran, vol. 4, I949, fig. 2I3 and I 0 3-Shahristan (nearIsfahan), Imam- 258). z7-dehSehzadeh Husayn and Ibrahim, ca. I07-Madinab, mashhadof Isma'ilb. 532/II37-38. This mausoleum has been Ja'faral-Sadiq, 546/1 I I5-S2. As described assumedby A. Godard to be that of the by Samhudi(p. I04), thiswas a complex caliph al-Rashid, who was indeed killed buildingwith a courtand a well in addi- near Isfahan (A. Godard, Les anciennes tion to the qubbahitself. The dateis given mosquees, Arts Asiatiques, vol. 3, i956, by an inscription. p. S) but the evidence of contemporary io8-Cairo, mashhadof Sayidnaal chroniclesdoes not justify this conclusion, Husayn, 549/I II54-55. In II54-55, the although, of course, this simple octagon headof Husaynwas movedfrom Ascalon (Survey,p. I023 and pl. 334A) could well to Cairo,where a new sanctuary,included be dated at this time. in the palace area, was built (Creswell, I 04-Rayy, so-calledtomb of Tughril, MAE,vol. I, pp. 27I-273; Maqrizi,vol. I, 534/II39-40. This much reconstructed pp. 427ff.). Mostof the presentbuilding is tomb is still pretty muchof a mystery.The modern. evidencefor the date is basedon aninscrip- I09-Cairo, mausoleumof Muham- tion, now in the Museum of Art at the madal-Hasawati, ca. II25-55. The main University of Michigan (formerly in the curiosityof thismausoleum is thatit seems possessionof E. Herzfeld), which is said to to have revertedto earlierpatterns by have come from the tomb (E. Herzfeld, beingopen on threesides (Creswell, MAE, imamzadehKurrar, Arch. Mitt. aus Iran, vol. I, pp. 2 59-260). vol. 7, I935, p. 8o); cf. Appendix. II0o-Cairo, mausoleumof Yabyaal- I05 Maraghah, so-called red mauso- Shabih,ca. i I 50. Thisis a morecomplex leum, 542/II48. This square mausoleum, structurewith a courtand an oratory,in additionto the domedroom over the tomb 48 It is conceivablethat a few otherMedinese (Creswell,vol. i, pp. 264 ff.). mausoleums,for whichno date is given in Sam- I I i -Cairo, mausoleumof Qasimabu hiudi,are as early as the first part of the i2th century (in addition to our no. 107); only a Tayyib, ca. i I 5o. Here also we are pro- thoroughstudy of the personagesinvolved could bably dealing with a complex building providea solution. (Creswell,MAE, vol. I, pp. 269-270). 38 OLEG GRABAR

I I 2--Tus, anonymous mausoleum. with evidenceprovided by thegeographers This well-knownlarge mausoleum with a of the time.The writers of the 3rdcentury superbfacade has not yet beensatisfacto- of thehijrah (gth century), like IbnRostah rily dated (E.Diez, Khurasan,pp. 57-62; or Ya'qubi, not only do not mention Survey, pp. Io72ff.), but the evidenceof monumentsbuilt over tombs, but also otherCentral Asian mausoleums makes it rarelylist namesof personagesburied in quite likely that the middle of the I2th givenplaces; on theother hand, the writers centuryis a possibledate (Pribytkova,pp. of the4th century(ioth centuryA.D.), like I 9 ff.). al-Muqaddasior Ibn Hawqal,almost sys- tematicallyprovide the readerwith such CONCLUSIONS lists. By the time of Ibn Jubayr,in the 6th century(I2th A.D.), placesof burialand of Thislist of I I 2 itemscorresponding to pilgrimageand mausoleumsare standard over i 6o mausoleumserected before the fare,and guide-booksappear for pilgrims.50 middleof the I 2th centurysuggests a num- The comparisonbetween monumentsand of conclusionsand raisesa seriesof prob- texts leads to a furtherremark: the interest lems. in places of burialof holy men, their more The first point to be noted is that the or less systematic cataloguing, and cultic two earliestmonuments in the list-the and social functions around tombs and in Dome of the Rock in Jerusalemand the cemeteries-as they appear in the studies Qubbahal-$ulaybiyah in Samarra-are of L. Massignonand G. Makdisi on Cairo extraordinarymonuments, which, in their and Baghdad-seem to have precededthe functionand in the circumstanceswhich actual appearanceof an architecturalform surroundtheir construction,stand out as to single out and to honor tombs and holy exceptional creations explicable only places. It is difficult, if not impossible,to throughprecise events of theirtime. It is, determine the time which separated the therefore,in the late gth, but especiallyin the i oth and i i th centuriesthat we can ascertainthe growth of commemorative towers of victory (E. Diez, Persien,pp. 73 ff.), but buildings. Their function was almost al- they obviously could not all fulfill that purpose. Some of them certainly had a commemorative ways funerary;there are only a few in- aim, which explains their use as tombs. But, ty- stances(nos. 48, i0o) in whichother aims pologically and genetically, they belong to a dif- areclearly indicated.49 This result coincides ferent group and their understandingmust await more complete archaeological,epigraphical, and literary documentation than has so far been 49 We have entirely left out of the discussion available. The importanceof suchprecise studies a whole group of Iranianmonuments whose com- is fully shown by the many surprisingconclusions memorativesignificance is still a matter of uncer- reached by Madame Sourdel-Thomine in her tainty, the manar. There is little doubt that the reading of the inscriptions on two such manirs tall "minarets"which are so numerousin cities, (Deux minaretsseldjoukides d'Afghanistan, Syria, the countryside,or remotevalleys, like the extraor- vol. 30, I953). dinary example at Jam (A. Maricq et G. Wiet, 50 For instance, al-Harawi, Guide des lieux Le minaret de Djam, Paris, i959), did not serve de pe'lerinage,tr. J. Sourdel-Thomine,Damascus, merelyfor the call to prayer.They have beencalled I957. THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 39 growth of the two phenomena,but it of the conqueringTurks was to createa shouldbe pointed out that,both during the sanctuaryto Abrahamin orderto counter- periodwith whichwe have dealt and in balancethe earliershi'ite shrine.51 In their latercenturies, several important branches searchfor personagesaround whom cults of Islam, banbalismbeing a conspicuous and ceremonieswere to be developed,the example,consistently avoided the use of sunnistended either to use scholars,Com- architecturalsymbols for theirheroes and panionsof the Prophetand early conquer- greatevents. ors,or Old TestamentProphets, especially As far as the purposesof the mauso- Abraham and Solomon, whose Islamic as- leumsare concerned, the fact emergesthat sociations are particularly strong.52It is, of the overwhelmingmajority of earlymau- course, unlikely that all sunnite mauso- soleumsserved either to emphasizeshi'ite leums and mashhads were built as a reac- holy places or to glorify princes from tion to shi'ism, and the many early con- smallerdynasties, usually heterodox. This structions of Egypt and Central Asia in is not surprising,for the very basicshi'ite particular, if properly dated, pose a major emphasison descentfrom the Prophetand problem. In some instances(no. 17), older the mysticalsignificance of the succession sanctuaries were taken over. Elsewhere of imaimsmight naturally result in thedesire (no. 48) attention has been brought to one to transforminto placesof venerationthe aspect of sunnism which may have affected real or allegedplaces where the members the growth of shrines, even though it is still of the holy family were buriedor lived. comparatively little known: the frontier Princely mausoleumsare, in earlier in- spirit of the ghazw, which often involved stances,particularly characteristic of the allegiance to a man or to the place from new dynastiesof Iranand it is interesting which the jihad originated. To perpetuate to contrastthe funerarypractices of these the memory of the founder or leader of princeswith the comparativesimplicity of such a group through a mausoleum is a the burialsof the 'Abbasidand even Fati- logical enough procedure, and it could well mid caliphs.Many moreprincely mauso- explain the monuments at Aswan as well leums probablyexisted than have been as many Central Asian ones. Altogether, identified,since so few local chroniclesare however, the evidence of the buildings knownand since the meaning of the build- seems to indicate that it is not before the ing fadedaway with the disappearanceof ubiquitous impact of shi'ism in what Mas- the dynastieswhich built them. Some of signon had called the "Qarmatian cen- the remaininganonymous mausoleums of tury" that either branch of Islam began CentralAsia, for instance,could originally systematically to transform its tombs and very well have beenprincely mausoleums historically meaningful places into sanctu- to whichat a later date a holy man or a 51 holyevent was attached. Many of theearly J. Sauvaget, Alep, Paris, I94I, pp. I24 ff., sunni shrines(nos. 9, 49) were probably with a discussion of the further development of built in answerto the growth of shi'ite these shrines in the second half of the I 2th century. placesof veneration;it is 52 J. Sourdel-Thomine, Les anciens lieux de known,for in- pe'lerinage, Bulletin d'Etudes Orientales, vol. I4, stance,that in Aleppoone of the first acts I951-54, pp. 65 ff. 40 OLEG GRABAR aries and to build shrinesover them. It beforethe middleof the I2th century,but mightalso be notedthat a significantnum- they exist and pave the way for the great berof earlymausoleums served for several monumentalensembles consisting of a personages,one of which,usually, was of tomband of piousor philanthropicinsti- greaterholiness or qualitythan the others. tutionswhich will appearin the second Whetherthis featureis merelyaccidental half of the century.Yet the singlesimple or whetherit reflectsthe fairlyearly iden- "canopy"is neverabandoned. The growth tificationof certainparts of cemeteriesfor of the pishtuq,the high gate almostsepa- specificindividuals or families(cf. intro- ratedfrom the mausoleumitself, in north- duction)is still unresolved. easternIran is moredifficult to date;and The third conclusionto emergefrom I havepreferred here a late i i th century our list is that while early shrinesexisted date to the early i i th centurydate pro- throughoutthe territoriesof Islam and posedby mostRussian archaeologists, even while someof the earliestsanctuaries ap- thoughthe apparentlyclearly dated ex- pearedin Iraq, where 'Ali and Husayn ampleat Tim(no. I2) hascomplicated the had died,the two areaswhich show by far problemconsiderably, just as it hasput the the greatestconcentration of earlymauso- questionof the muqarnason a new basis. leumsare the northeasternIranian prov- The interestof the many instancesof the incesand Egypt. North Africaand Spain, pishtaqwhich have beenlisted is in show- from whichconsiderable literary and ar- ing many variants,often embryos,of a chaeologicalremains are available, seem to featurewhich was destinedto becomeone havebeen least affected at thistime by the of the most spectacularcharacteristics of developmentof mausoleums. laterIranian architecture, and whosepur- A last conclusionconcerns the architec- pose and evolution require a separate tural types which are involved. On this study.The point whichmay be of greater subjectour list did not lead to any major significance,as one dealswith the physical newconclusion, at leastinsofar as themau- appearanceof the mausoleums,is that, soleumitself is concerned.The tower-tomb even thoughthere is clear evidencethat is peculiarto the mountainsand plains quite early thereexisted constructions in aroundthe Caspian sea and to thenorthern cemeteries(cf. the introduction),it is only edge of the Iranianplateau; it affecteda slowly that a compositebuilding appears few monumentsin CentralIran (no. 45 for whichunited in a singlearchitectural en- instance).The squareor polygonal "cano- semblethe manifoldfunctions of a shrine. py" type of tomb coveredwith a dome ex- Whilethese conclusions appear to fol- isted everywhere.In the earliest instances low the evidenceprovided by the docu- it was open on all sides and then, little by ments,they alsoraise problems, which can, little, was transformedinto a hall with a at this stage, only be defined, for their single door-or sometimes two doors- solutiondemands investigations which go and a mibrab. Unusual plans, accessory muchbeyond my originalaim. buildings,crypts, combinationsor several The first of these problemsconcerns squarehalls into one ensemble,and mosque- certainaspects of the purposeof the mau- mausoleumsappear only rarely, if at all, soleums.It has beennoted that two areas THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 4' of the Islamicworld-Egypt and Central even merchants.54The socialbasis for the Asia-seem to havedeveloped mausoleums sponsorshipof a monumentalarchitecture in the i oth and i ith centuriesto a much widenedremarkably. The resultis evident greaterdegree than other provinces,and in thenature of thebuildings: from the few that many of thesecannot readily be ex- largeearly mosques (al-Azhar, al-Hakim) plained as obviouslyreligious shrines or and palaces, we move to many small ora- princely memorials.The problem is to tories (of which only a handful remain)"5 evaluatethe significanceof thesefacts and and to many pavilions, private housesand to find out whetherthere were certain pe- apartmentsthroughout the city.56.Within culiaritiescommon to thehistory and civi- this context, if the conclusionis acceptable lizationof thesetwo areaswhich could ex- that the main original impetusfor mauso- plainthe phenomenon, or whetherentirely leums derived from religious shi'ite and differentcauses led to the sameresults in dynastic royal needs, it may be suggested both areas.The difficultyof the problem that quite rapidly, at least in Cairo, a sort is complicatedby the fact thatmuch of our of ""democratization"of the mausoleum view of early Islamichistory is basedon took place to include all who could afford chronicleswhose concern was narrowgeo- and desiredto have one built. In addition graphicallyand limited in content.Fur- to its religiousand princely connotations, thermore,the concernof manyhistorians, the mausoleumbecame a symbol of con- untilvery recently,has been in the secure- spicuous consumption. In many instances ment of a fundamentalchronology of suchmausoleums were to be relatedto phil- eventsand in problemsof doctrinerather anthropic (schools,hospitals, guesthouses) than in the establishmentof the cultural, or devotional (monasteries,ribats, orato- social,and economic factors which formed ries) institutions. Whether this occurred the civilizationof the differentprovinces commonly before ii5o in Egypt is still of the Islamicworld. difficult to say, except in the one instance Thanks to Maqrizi's Khitat, to the ex- of the oratory (masjid), which may well istence of good epigraphical surveys, and have precededthe mausoleumin cemeteries to the still comparatively few published (see above p. io). Outside of Cairo, the Geniza fragments, the Egyptian situation only instanceof mausoleumson any large can be better understood than the Eastern scale is provided by the Aswan construc- Iranian one. Even a rapid perusal of Maq- tions. The explanationfor them is suggest- rizi's description of Cairene monuments53 ed by their location in a frontiercity, with clearly shows that the i oth and i i th cen- turies are characterized by the building of many religious monuments by private in- 54 See the very remarkableinscription of a merchantwho built a mosqueand planted a palm dividuals, at times members of the ruling tree for the benefit of all Muslims (Repertoire, families acting as private believers (this is no. 2173), datedIOII-I2. particularly true of women), more often 55 Mosques al-Aqmar and al-Salih Tala'i, officers, administrators, dignitaries, and Creswell,MAE, vol. I, pp. 242ff. and 275 ff. 56 The admirativedescription of Nasir-i Khosrowis, in this respect,particularly remark- 53 Maqrlzi,mainly vol. 2, esp. pp. 244 ff. able. 42 OLEG GRABAR

a specialveneration of martyrsand the seemsto indicate that there was a shift or leadersof the variousrib?its organized for power from landowning dihqans to the the defenseof Islam. cities with their military rulers and their But the historical circumstancesof numeroussocial and professionalorgani- Egyptwere not duplicatedin totoin north- zations. The passagefrom one to the other eastern Iran. The extraordinary variety of probably took place around the time of culturesand influenceswhich created Is- the fall of the Samanids(i.e. late ioth and lamicCentral Asia is only now beginning early i ith centuries).It may be suggested to emergefrom obscurity, and the process that, in northeasternIran, the originalim- of Islamizationof Iranianand Turkish petusfor mausoleumsderived from prince- groupsthere is still far fromclear. Of the ly constructions,but that, just as in Egypt, four causeswhich can be proposedto ex- the wideningof the patronageand changes plain the growth and developmentof in religiousand cultic habits (here related Egyptian mausoleums-shi'ism,princely more precisely to the importanceof semi- symbols, wider social patronage, the fron- religiousorders guardingthe frontier and tier-two are clearly present in Central of social organizationswith mysticalover- Asia. It is there that we find most of the tones) led to a wider use of the monumen- earlier instances of purely secular tombs tal tomb. This explanation,however, can- (nos.6, i6, 27, 32), and there is little doubt not be more than a hypothesisso long as that all the symbols, institutions, and alle- our informationis so limited on the indi- giances which developed on the frontiers viduals who were buried in the mauso- of the Muslim world were particularly leums or the events they commemorated; acute in Central Asia.57 Many ribats ex- and the uncertainchronology of the tombs isted there; many were sponsored pri- adds to the difficulty. vately, and in or near some of them found- Thereis, however,a furtherpoint which ers or their descendants were buried.58The emerges from these remarks. Just as we social history of the i oth and i i th cen- have seen that funerarycults precededthe turies has not yet been entirely worked out, buildingof mausoleums,so it appearsthat but the evidence provided by archaeology59 the mausoleumis closely relatedin another way to a whole group of social and pious 57 In summarizedform the point comes out institutionsthat appearedbefore it, espe- in V. V. Barthold, Four studies on the history of cially the ribat and the private oratory. Central Asia, vol. i, Leiden,i956, pp. i 8ff. and The central feature of the latter is that it 7c; and it appears quite clearly as one glances was heavily throughal-Narshaki's, History of Bukhara,tr. R. endowed and sponsoredby a N. Frye, Cambridge,I954. wide variety of people. As suchthe orato- 58 See above p. 9. ries reflected more than piety; they were 59 The best, even though not always fully also forms of investmentand of self-glori- documented and at times schematic,description fication. These secularideals were carried for the southern part of Central Asia is in Puga- over to the mausoleums.Theyexplain why, chenkova,pp. I I8 f., especially the introductions and conclusions of chapters. For the northern in later centuries,with only few exceptions areas one should consult the syntheses of S. P. (for instancethe mausoleumsto al-Shafi'i Tolstov on Khorezm. in Cairo), next to popular, almost pagan, THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 43 memorialsto localsaints, the characteristic A problemof comparablecomplexity domicalmemorial will be sponsoredby confrontsthe arthistorian when he triesto and built for princes,members of their explainthe originsof the formsused for families,dignitaries, and officers.The Taj the mausoleums.The squarecovered with Mahal is the supremeillustration of the a dome and the simple canopy are, of fact thatby its origins,its sponsorship,and course, not new architecturalforms: they its growth, the Islamic mausoleumre- were used in the funerary architectureof flectedsecular conceptions. Next to these antiquity.The problemis that thereis little memorialsto rulersare the great shi'ite evidence that these forms had been used sanctuariesto which constant attention for these purposes in the centurieswhich was given throughthe centuries.Sunnite precededor accompaniedtheir appearance shrines,however, remained few in numbers. in the Islamicworld. In the Christianem- Mostof themwere createdin the histori- pire of Byzantium, imperial sarcophagi cally and culturally tumultuousperiod were put in Constantinopolitanchurches, whichextended from the time of break- but, with a few exceptions in Armenia, downof the'Abbasid caliphate in theearly specific funerary architecturefor individ- ioth centuryto the Mongolconquest in uals does not seem to have developed. In the middleof the I3th. Few of them re- early Christiantimes many instancesexist ceived,in the followingcenturies, the care of mausoleumsfor saints and heroeswhich and attentionwhich was given either to are clearly related to pagan types, but 'Alid shrinesor to the new secularones. these antique funerary cults and forms The two factors,thus, whichfirst caused were soon sublimated in the martyrium the growth of mausoleums-sh'ismand church,which by the 6th century had de- secularglorification-remained through- veloped a large repertory of shapes and out as the main sourcesof memorialcon- forms based on the central plan and had structions." almost abandonedthe simple types which Islam used in the i oth century;it is inter- 60 On a lower social and artistic level, the esting to note, however, that some of the appearance of local shrines, along roads or on Egyptian examples preserved the early elevations of the ground, dedicated to local holy men and often with pagan overtones, cannot be types much longer.6"In Iran the ancient dated precisely. None of the ones known to me tombs of the Achaemenidsand of the Par- are earlier than the I2th century. Even though thians were not continued in Sasanian many reflect cults of pre-Islamic origin, the evi- times.62Towers of silence and ossuaries dence seems to indicate that the construction of popular qubbabs followed the growth of official are published; some indications exist in Sarre- shrines rather than continued a pre-Islamic tradi- Herzfeld,Archaeologische Reise, im Eupharat-und tion of small sanctuaries and mausoleums of popu- Tigris-Gebiet,Berlin, I9II-20, and in T. Canaan, lar origin. The facts that Islam lacked the eccle- Mohammedansaints and sanctuariesin Palestine, siastical organization of Christianity and that its Jerusalem, I927. major developments were urban rather than 61 A. Grabar, Martyrium, vol. I, pp. 47 ff., agricultural also explain why these symbols of I44, and passim. popular and peasant beliefs were not incorporated 62 See, however, some curious paintings in into the mainstream of the faith, as they were in Panjikent, Zhivopis Drevnego Pianjikenta, Mos- Christianity. Descriptions of few of these buildings COW, I954, pl. XX. 44 OLEG GRABAR were presumablythe only monumental mosquesin front of the mibrabfrom the formsassociated with funerarypractices.63 momentthe mibrizbwas introducedin Me- And one cannotexplain all typesof mon- dina64; it was a commonfeature of throne umentaltombs from the impactof the In- rooms(Mshatta, Baghdad, Samarra), gates dian stupa or of nomadic mounds form (KhirbatMinyah, Baghdad), and pavilions CentralAsia, while an influenceof pre- (betterknown throughdescriptions as in Islamic Arabian practicescould hardly Egypt or Iran) in the architectureof pa- have affectedeastern Iran or even Egypt laces;it was found over fountains in palaces to the exclusionof closerprovinces. Under (Khirbatal-Mafjar) or in mosques(mosque thesecircumstances the assumptionby the of IbnTulun in Cairo).The domical struc- Muslimsaround the beginningof the i oth ture was adopted as the most common centuryof an ancientfunerary form which formto be usedover holy spotsand tombs had not beenused for severalcenturies for not becauseof its precisefunerary attribute thesepurposes poses an importantproblem. butbecause of its generalmeaning as a sign Severalanswers might be suggested.It of veneration. may be that in partsof the Islamicworld, This last point may also serve to ex- older martyriaof simplerform were still plain the cloudy originsof the northern used,as for instancein the biblicalsanctu- Iraniantower-tomb. There a differenttype aries of Syria and Palestine, and in Egypt, of structure,whose symbolic meaning as a and that these modest symbols served as memorial,a tower of victory, or a beacon models for the Muslims. Or perhaps the has often been discussed,65was used for Islamic world revived for its own purposes funeraryarchitecture because of its abstract architectural formulas which had existed significanceas a symbol of power or holi- for these very purposes several centuries nessrather than preciselyfor any funerary earlier. It may be asked, however, to what reason. In the case of both forms, then, it extent the medieval Islamic world was con- would appear, in this hypothesis,that, in scious of the significance of the early mon- its searchfor a monumentalexpression of uments which were standing in the areas it certain spiritual and cultural needs, the had conquered several centuries after the Islamic world did not always invent new conquest. Moreover, examples taken from forms or borrow them directly from other the former Christian Near East cannot be culturaland artistictraditions, but at times used to explain the Central Asian buildings. adapted an existing architecturalvocabu- A more likely explanation of the use lary of its own to new needs.If it happened of the dome over a square or polygon is that the final result bears resemblanceto not its association with funerary architec- earlier,pre-Islamic monuments erected for ture in the past, but its significance in other similarpurposes, it is not so muchthat they aspects of Islamic architecture as a sign of were borroweddirectly one from theother, honor and veneration. It was common in 64 J. Sauvaget, La mosquee omeyyade de 63 R. Ghirshman,Etudes IraniennesII, Arti- Medine,fig. 5 on p. g I. bus Asiae, vol. 2, 1948, pp. 292 ff., with an im- 65 See the references to the works of Diez portant bibliography. undernos. 27 and 66; cf. also note 49. THE EARLIEST ISLAMIC COMMEMORATIVESTRUCTURES 45 but that both derivedtheir architectural theinscription to my attentionI tooksome formsfrom the wide repertoryof moreor briefnotes on what he told me of the cir- less abstractformulas which the Mediter- cumstancesin which he acquired the plaque. raneanand the Near Easthad in common It appearsthat during his residencein Iran sinceantiquity.66 he foundand purchasedthe inscriptionin Theseconsiderations on the sponsorship one of the Teheranbazaars. Some time andpurpose of theIslamic commemorative laterwhile examining a notebookof a Per- buildingand on the formsit took go much sian antiquarianfriend of his he saw a beyondour originalaim of gatheringto- drawing of this same inscription in situ gethera seriesof documentson a certain overthe door of thethen unrestored "Tower type of building.They serve,however, to of Toghrul"at Rayy. Herzfeldlater veri- illustratesome of the problemswhich are fied thisprovenance to hisown satisfaction posedin any investigationof the manner by measurementsand an examinationof in which Islamiccivilization created the therestored brickwork where the tablet had artistic forms which eventually became formerlybeen fixed in the tower.Unfor- identifiedwith it. It is only whendetailed tunatelyI do not now recallwhat he told investigationsof other architecturalenti- meof its exactposition in thedoorway.1 ties-the masjid,the ribat, the khanqah, The inscriptionin questionis cast on a the khan-and of the culturaland social single sheet of iron (about 2 mm. in thick- phenomenawhich led to theirgrowth have ness), measuring80.5 cm. in length and beenmade that it will fully be possibleto from I 2 to I 3 cm.in height.Notches at the understandthe way in whichIslamic archi- top and bottomof the righthandend give tecture was formed and to explain the it a tabulaansata form,2 at leastat thisend. formsit took. The plaqueis piercedby eleven circular holes, approximately9 mm. in diameter, throughwhich nails or bolts were driven APPENDIX to attach the sheet to the architectural BY C. G. MILES memberon whichit was mounted.There arefive of theseholes above the upperline In I940 the lateErnst Herzfeld, then at the Institutefor AdvancedStudy in 1 The existence of this dated inscriptionhas Princeton,showed me an iron plaquein been noted several times, but never with any ex- his possessionbearing an inscriptiondated planatory details. In Arch. Mitt. Iran, vol. 7, S34 H. Afterhis deaththis inscription be- 1935, p. 8o, Herzfeld's list of early Kufic and Naskhi inscriptions in Iran included the entry camethe propertyof his sister,Mrs. Char- "534, Rayy, sog. Turm des Toghrul, Eisenthiir." lotte M. Bradford,who entrustedit to me This is the basis of the bare mention of the date for safekeeping.It is now in the Museum of the tower in A. V. Pope, Survey of PersianArt, of Art at the Universityof Michiganin London,I938, vol. 2, p. Io22, and vol. 4, plate Ann Arbor.When Herzfeld first brought 346; and of my entry, 534, Rayy, tower, in Ars Islamica,vol. 8, I94I, p. io8. 2 See Ernst Herzfeld, Die Tabula ansata in 66 Cf. the analysesof the "basilicalhall" made der islamischen Epigraphik und Ornamentik, Der by Sauvaget,op. cit., pp. I 5 8 ff. Islam, vol. 6, I9I6, pp. I89-I99. 46 OLEG GRABAR of inscriptionand six below the lower line. debtedto ProfessorV. Minorskyfor draw- Thetext readsas follows: ing my attentionto analogousnames com- poundedwith -avar, such as Namavar,3 Bakhtavar,4and of course Dinavar.5The Fecit cAbdal-Wahhab al-Qazvini b. Fakhra- order of the names is curious:one would var at the end of Rajab of the year 534 [March, expect al-Qazvini to follow the father's II40 A.D.] name.The unsophisticatedcharacter of the The rathernaive, quasi-Kuficcharac- inscriptionsuggests the possibilitythat the ters are completelyunadorned. Several addition of the father's name,6who may peculiarities suggest that the craftsman have been a Daylamite, with whom the who designedthe letters was not really inhabitantsof Qazvin had frequent, not versedin the Kufic alphabetbut that he always hostile, intercourse,may have been attemptedto give the inscriptiona Kufic an afterthought. appearance.For example, the lam of Just what it was that 'Abd al-Wahhab 'amaladoes not descendbelow the base "made"('amala) cannot now be deter- line, nor does the nun of thalathinor the mined. It is unlikely that he was the archi- kha of salkh;the dal of 'abdis not Kufic, tect or the builderof the tower itself; such thesin of khams,being in themedial form, an individual would doubtless have re- shouldjoin to the followingmim, but it corded his achievementin a more impres- doesnot; the final ya's of Qazviniand fi sive inscription.Herzfeld's implication is are curiouslyabbreviated. Despite these that the plaquewas associatedwith an iron errorsand the imperfectbalance of the door. In any case there can be little doubt two lines, some thoughtand planningin that the inscription provides us with a the layingout of the inscriptionis evident specific date for at least a part of "Togh- in the spacingof lettersin the upperline rul's Tower" at Rayy, a date shortly after thatpermits the topsof severaltall charac- the death of Toghrul I, sultan of the 'Iraq ters in the lower line to extend into the branchof the Seljuqs. upperregister. Withregard to thereading and content I E. g., Hasan b. Namavar, one of the last of the inscriptionthere can be no doubt survivorsof the Buyids (cf. Juvaini,7he history of the world-conqueror,tr. John A. Boyle, Manches- aboutthe date and the name 'Abd al-Wah- ter, 9 58, vol. 2, p. 697). hab, and very little doubtabout the cor- 4 Cf. F. Justi, IranischesNamenbuch, Mar- rectnessof the nisbabal-Qazvini. The only burg, i895, p. 6i. problemis that of the correctreading of S Ibid., P. 85. the propername immediatelypreceding 6 The first word in the secondline is undoubt- edly bin and not [a]bu7;all the waw's in the in- the date. Fakhravar,"Bringer of Glory," scription have clear open heads. Also a close ex- althoughevidently previously unrecorded, aminationof the plaque itself shows that no letter is a quitepossible Persian name. I am in- is missingat the beginningof the secondline.