R.D. MCCHESNEY

ARCHITECTURE AND NARRATIVE: THE KHWAJA ABU NASR PARSASHRINE. PART 1: CONSTRUCTING THE COMPLEX AND ITS

MEANING, 1469-1696

Amidst the mounds, ruins, and partial walls of theonce-flourishing metropolis of Balkh, stands, moreor less intact, a great mausoleum (Pers. gunbad, gun-baz) commemorating the fifteenth-century Naqshbandifigure, Khwaja Abu Nasr Parsa (fig. 1). Although thebuilding is well known through books on Islamic ar-chitectural history, 1 its narrative tradition-that is,the stories told about it and the individuals associatedwith it-remains generally unexplored. It is the ar-gument of this paper that the narrative tradition, il-luminating as it does the meanings successive gener-ations have attached to the building, not only providesclues to the architectural development of the site buthelps explain why this building, of all the great struc-tures once standing in Balkh (some built much later),alone remains.

Part 1 of this study covers the first two and a halfcenturies or so of the building's history. It begins withthe construction of the tomb in the mid-fifteenthcentury and ends in the late seventeenth century, aperiod of considerable political and social change forthe family most closely associated with the building.In part 1, I reconstruct aspects of the social contextin which the complex was erected; the role played inits erection by two personalities-a Naqshbandi shaykhnamed Khwaja Abu Nasr Parsa (d. 1460 or 1461),son and spiritual heir of Khwaja Muhammad Parsa(d. 1420), one of the principal disciples of Baha al-Din Naqshband (d. 1389); and a military figure, MirMazid (d. 1470 or 1471), the son of Sayyid Pir Mu-hammad 2 and a leading general for at least two Timu-rid princes; the long tenure as managers of the shrinecomplex of the descendants of Abu Nasr Parsa; theemergence of the Parsa'is as the leading family ofBalkh; and the consequences of their preeminencefor the survival of the complex and for its wider im-pact on the architectural development of Balkh.

This reconstruction is, of course, completely de-

pendent on written sources (assisted to some smalldegree by clues left in the early photographic record).So in addressing these issues, I will also try to con-textualize the narratives-when, by whom, for whom,and possibly why they were written-and, perhaps,tease out the interplay of narrative, architectural evo-lution, and social history.

Part 2 (to appear in the next volume of Muqarnas)will cover the last three hundred years of the shrine'sexistence. For the most recent hundred years I drewon the photographic record to demonstrate some ofthe conclusions of part 1 and to emphasize the rela-tionship of narrative, of the stories told and the truthsbelieved about the architectural complex and howthese stories and beliefs played out in changes to thefabric and form of the complex. In this later period,the boundaries of the social and political contextexpand significantly and bring in entirely new narra-tive traditions and perspectives as Balkh is drawn intothe purview of individuals and societies far removedin space from it.

THE FOUNDATION OF THE SHRINE

The Biographical Legacy of Abu Nasr Parsa

The family into which Abu Nasr was born had deeproots and enjoyed a long tradition of intellectualprominence, though not in Balkh but in Bukhara.His father Muhammad Parsa was a highly regarded'alim in addition to being the "second successor (khal-

fah-yi duwwum)"3 to Baha al-Din Naqshband, compilerof his sayings, and acknowledged as the figure whomade the order both intellectually credible and power-ful in Bukhara.4 When he died while on the hajj, hisintellectual and spiritual legacy passed to his son who,though not the scholar his father had been (we haveno known writings of his), was influential in his own

__�

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

Fig. 1. Abu Nasr Parsa's gunbad. Photo taken by C. L. Griesbach,ca. 1885. The Alkazi Collection of Photography. (Photo: cour-tesy Alkazi Collection of Photography, New York)

right. Muhammad Parsa's memory is preserved bothin his writings and in the writings of those who ad-mired his intellect. He is also remembered for tak-ing a prominent role in the politics of his day.5 Al-though his son is profiled in most of the samebiographical sources, there is no scholarship for thebiographers to refer to and he is memorialized chieflyas heir to his father's legacy. 6

However his intellectual contemporaries may haveviewed him, contemporary chroniclers indicate he hadgreat prestige on the political stage. Both 'Abd al-Razzaq Samarqandi and Mu'in al-Din Isfizari, two late-fifteenth-century historians, suggest the dimensionsof Abu Nasr's leadership role in Balkh in the middleof the century and why that role might have led to

the construction of a grand mausoleum at his gravesite. As portrayed in their historical writing, he is amediator and arbitrator, a common role for religiousfigures. In 1454, the Miranshahid Sultan-Abu Sa'idcrossed the Oxus from the.north and seized Balkhfrom Amir Pir Darwish Hazaraspi, a supporter of Abu'l-Qasim Babur, son of Shahrukh and successor to himin Herat and eastern Khurasan in 1447. Hazaraspihad been appointed governor of Balkh some fouryears before. When Abu'l-Qasim Babur heard that Sul-tan-Abu Sa'id had taken Balkh, he immediatelymarched back from Astarabad where he was campaign-ing. At his approach, Sultan-Abu Sa'id abandonedBalkh and fled back across the Oxus. When Abu'l-Qasim Babur arrived at the gates of the city, he wasgreeted, according to Samarqandi, by "Khwaja Nasiral-Din Abu Nasr Parsa." Writing ca. 1465-71, Samar-qandi also calls him "shaykh al-islam."7 Isfizari, writ-ing between 1491 and 1494, however, tells a slightlydifferent story, saying that the khwaja exerted everyeffort to make peace between the two sides but failed.8

Samarqandi does not explicitly assign Abu Nasr amediatory role, but having him greet Abu'l-QasimBabur at the gates of the city is a clear sign that Sa-marqandi believed he was the preeminent represen-tative of the residents of Balkh.

In the second recorded instance, Abu Nasr Parsareportedly interceded with Sultan-Abu Sa'id in 1455on behalf of a group of rebels in Utrar. This is men-tioned by Samarqandi in a passage which Bartholdcites but which I have been unable to find.9 Bartholdconcludes from it that the rebels had some influencewith the religious leaders of Bukhara because one ofthem was spared through Abu Nasr's intercession.

These two reports, dating to the second half ofthe fifteenth century, provide our only more or lesscontemporary historiographical portrayal of Abu Nasr'spolitical and social prestige. Another near-contem-porary, Khwandamir, writes about Abu Nasr but saysnothing of his role in mediating political disputes.Although Khwandamir is important for providing atextual link between Abu Nasr and Mir Mazid, hisview of Abu Nasr corresponds in large measure withJami's (see n. 6):

After Khwaja Muhammad Parsa's death, his son, KhwajaHafizuddin Abu-Nasr Parsa, who was known for hisknowledge and practice, took his father's place andsurpassed him in asceticism and spiritual poverty. Hisdeath occurred in 865 [1460-61], and he is buried inBalkh. 10

��___IC____s*______s_________II_�_�

95

R.D. MCCHESNEY

To this brief notice he appends an obituary chrono-gram which Fakhr al-Din "Safi", author of the Rashahat,also uses and may have written:

Khwajah-i a'zam Abft Nasr ankih shud / takyah ghashmasnad-i dar al-baqd // sirr-i chin ba khuda paywastahbuzd / z-in sabab trkh shud "sirr-i khudi"The Great Khwaja Abu Nasr is he/ whose resting place(takyah gahash) is eternity // Since his secret has alwaysbeen with God/ so his obituary date is "God's secret."[The date produced by the sum of the numeric valueof the letters in "sirr-i khudd" is 865.1 ]

Mir Mazid Arghun as Portrayed in ContemporaryNarratives

The second personality associated with the founda-tion of the shrine and the figure credited either withbuilding the mausoleum for Khwaja Abu Nasr Parsaor with building a madrasa and perhaps other un-specified buildings at the tomb, is Amir Sayyid Jalalal-Din Mazid Bahadur of the Arghun, if we accordhim his full style, or, in varying shortened forms-Amir Sayyid Mazid Arghun, Sayyid Mazid Arghun, MirMazid Arghun, or simply Mazid Arghun.l2 He is por-trayed as a prominent figure in Timurid politics,though the degree of his prominence seems to de-pend on the source. Samarqandi describes him, in aplay on the word mazid (increase, augmentation), asbeing "at the pinnacle of the amirs of the time be-cause of the augmented esteem held for him" (bi-mazid-i i'tibar sar amad-i umar-yi rftzgdr bud).13 Yet thegeneral impression one gets from Samarqandi is thatMir Mazid is only one of many prominent militaryfigures supporting Sultan-Abu Sa'id. Isfizari, on theother hand, writing nearly thirty years after Mir Mazid'sdeath, serves him up as a much more important per-sonage, often singling his name out to represent "thegreat amirs" (umar&yi kibar or umard-yi 'izam) involvedin a particular campaign or event.14

There is a considerable amount of informationabout the participation of Mir Mazid in military cam-paigns. From this and from the attribution to him ofthe mausoleum and a great chaharbdgh at Balkh, wemay assume he was a figure of power and wealth.Textual evidence of civic contributions is scarce. Bio-graphical references to non-royal figures like Mir Mazidare fairly elliptical; the real focus of the writer is usuallyon their relationship to the dynastic actors, especial-ly the aspirant to, or possessor of, nominal sovereignpower. From the chronicler/biographer's standpoint

at least, figures like Mir Mazid are worth memorial-izing to the extent that their role as backers of theclaims of one or another of the dynastic figures iscrucial to the outcome of a narrative line. We learnlittle of their backgrounds, their status within theireponymous group, in this case the Arghun, their non-military interests, their marital alliances, unless withthe royal clan, and anything about their children anddescendants unless they too enjoyed relationships withthe sources of what the writer sees as authentic pow-er. The genealogical record, so important for theTimurid royal clan, remains its preserve.' 5 On rareoccasions we find traces of such military figures be-yond the conventional narrative traditions, in the build-ings they constructed or other activities they spon-sored, in such places as building inscriptions or bookdedications. Certain inferences can be drawn abouttheir architectural predilections and the purposes theyperhaps envisioned in spending their money this way,but because of the reticence of the written sourcesand our imperfect understanding of the intellectual,spiritual, and ethical realms within which these pa-trons operated, these inferences remain largely spec-ulative.

Having said that, what is there to be learned fromthe many bits of written evidence we do have aboutMir Mazid Arghun? First, the gentilic "Arghun" in-forms the reader that he was a member of that cele-brated Mongol group and his prominent mention insource after source tells us he appears to have beenthe preeminent Arghun leader. It is, however, strangethat although Samarqandi identifies others as Arghun(actually Arghuni), he never uses the term with MirMazid's name. In the constitution of Timurid poli-tics in the middle of the fifteenth century the Ar-ghun appear as one of the most powerful groups.'6But the size, unity, and common policies, if any, ofmembers of this group are unknown.

The Arghun are often portrayed as particularly closeto the Miranshahid Sultan-Abu Sa'id. According toBartold, sometime around 1436 the Arghun, in anattempt to wrest control of Samarqand from UlughBeg, had proclaimed as king the twelve-year-old Sul-tan-Abu Sa'id, the putative son of Sultan-Muhammad,grandson of Miranshah, and great-grandson ofTimur.17 This proclamation appears to have come tonothing except to link the fortunes of the Arghunand Sultan-Abu Sa'id in later writers' minds. 18 JeanAubin, who provides a thorough, if somewhat eclec-tic, biography of Sultan-Abu Sa'id beginning with his

I� _IIC� �I�I�I�

96

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

emergence as a political force at the age of twenty-five or so (ca. 1449), says that the Arghun enjoyedpreeminence in his entourage.' 9 In a lengthy essayon Timurid political history in the Cambridge Historyof Iran, Hans Robert Roemer asserted that the Ar-ghun were Sultan-Abu Sa'id's main military support.2 0

It is true that in the accounts of Sultan-Abu Sa'id'svarious campaigns, Mir Mazid Arghun is notably theonly amir-or sometimes one of only two or three-cited by name as representative of the "great amirs"performing military service for Abu Sa'id. From this,one supposes, it is logical to infer that Mir Mazidstands for the Arghun as a whole, although there islittle if any direct, and indeed some contrary, evidenceof this. Still, we have no information on the size ofthe Arghun or whether the troops Mir Mazid com-manded were Arghun or not. (Khwandamir refers onceto Amir Khusraw Shah, who "during the time of thesultan-i sa'did (Abu Sa'id) was a soldier (nawkar) inthe employ of Amir Sayyid Mazid Arghun," 21 but givesno indication of his tribal affiliation. 2 2 Jami in a let-ter to Nawa'i also refers at one point to "the soldiers(lashkariyan) of Amir Mazid Arghun" but gives no hintof their tribal or sub-ethnic identities. 2 3

A number of other individuals identified as Arghunare mentioned in 'the contemporary and near con-temporary writings-some of the more prominentbeing Amir Sayyid Murad Arghun, Amir Asil al-DinArghuni (Mir Mazid's brother, according to Samar-qandi2 4 ) and Amir Sultan Arghuni2 5-but not alwaysas supporters of Sultan-Abu Sa'id. Amir 'Abd al-Rah-man Arghun, for example, was a prominent backerof Sultan Husayn Mirza, Sultan-Abu Sa'id's chief ri-val for Khurasan. 2 6 So there is at least some evidencethat the Arghun were not the monolithic supportersof Abu Sa'id that later historians have implied.

Mir Mazid's name is linked with a number of eventsand incidents.2 7 In the winter of 1458-59 he accom-panied Sultan-Abu Sa'id on his second campaignagainst Herat which led to its capture from the QaraQoyunlu. According to Isfizari, but not Samarqandi,he seems to have played the central role in the nego-tiations that led to Herat's fall. Samarqandi has MirMazid rescuing Sultan-Abu Sa'id in Herat after theTimurid had taken it in 1458-59.28 Later in 1459,Mir Mazid succeeded in taking the fortress at Nirah-tu from a rebel soldier,2 9 and in 1459-60 and 1460-61 he fought in Mazandaran against the 'Umar-Shaykhid Sultan-Husayn Bayqara, another of Sultan-Abu Sa'id's major rivals.3 0

In that same year (1460) both Samarqandi andIsfizari have him going to Balkh, the first indicationof his being in the same place as Abu Nasr Parsa.Isfizari says he was sent by Sultan-Abu Sa'id to Balkhfrom Samarqand in early 865 (1460) "to be adminis-trator and guardian of that region" (zabit wa mustah-fiz-i an hudud bashad),31 while Samarqandi merely hashim going there "to find out what was going on" (bi-janib-i wilayat-i Balkh raftah az an hudud bar khabarbashad ).32 In light of the building projects he under-took in Balkh, Isfizari's description, which places himthere in a more permanent capacity, would seem tobe closest to reality.

But whether as administrator of the region or sim-ply as fact-finder, the appointment to Balkh put himin the right place at the right time to have encoun-tered Parsa in person, although their association, ifstemming from this time, would have been of verybrief duration. The Hijri date (865) of Mir Mazid'scommission was the very year Abu Nasr's death isnormally dated. Prior to this time there is no recordof Mir Mazid's being in Balkh, although there is alittle evidence and some good reasons (family andproperty interests as well as Bartold's view that AbuNasr had close ties to Bukharan religious circles) toplace the Parsa shaykh at times in Bukhara whereMir Mazid might well have met him. In the next Hijriyear 866 (1461-62), Mir Mazid accompanied Sultan-Abu Sa'id on campaign intoJurjan (southeast of theCaspian Sea) during which he led the vanguard (man-ghalay) of the army. 3 3

His career between 866 (1462) and 873 (1468-69)is a little more obscure, although if a mid-sixteenth-century source is accurate, he was probably spend-ing time in Balkh, supervising work on the great gunbaz(mausoleum), which is dated to 867 (1462-63). Onemanuscript of the Matla' al-sa'dayn has him going toHerat in 869 (1464-65) to perform funeral rites forhis recently deceased brother.3 4 He also took part inSultan-Abu Sa'id's campaigns in Badakhshan whichhad ended by 1466-67.35 Jami, in the letter referredto above, mentions a woman's petition for the re-lease of her son taken prisoner by Mir Mazid's sol-diers "at the time of the Badakhshan campaign."3 6

The letter is undated, and it is therefore possible theincident may have occurred later, after Sultan-AbuSa'id's death, in the brief period when Mir Mazidwas in service in Badakhshan to one of Sultan-AbuSa'id's sons, Mirza Abu Bakr. It was probably also inthe period between 1461 and 1469 that the Arghun

97

R.D. MCCHESNEY

amir constructed a garden estate (chaharbagh) at Balkh,a complex of sufficiently royal proportions that whenSultan Husayn Bayqara came to Balkh in 1490-91,many years after Mir Mazid's death, he made the "Cha-harbagh-i Amir Mazid Arghun" his winter headquar-ters. 3 7

In 1469, Sultan-Abu Sa'id fought and lost a battlewith Aq Qoyunlu forces at Qarabagh-i Arran. Cap-tured by the Aq Qoyunlu leader, Uzun Hasan, theTimurid leader was put to death in February 1469,but his general, Mir Mazid, who was taken prisonerat the same time, eventually won release.3 8 Isfizariportrays him as having been in correspondence withQara Qoyunlu and Aq Qoyunlu figures, and it mayhave been because of these long-standing ties thathe was able to ransom his life.39 He made his way toBadakhshan where he was taken into service by Mir-za Abu Bakr, who had been awarded the region afterthe campaigns of the mid-1460s. But within a year offinding a new patron, Mir Mazid fell victim to thepost-Sa'idian political order.

The story of his end is somewhat murky. Accord-ing to Khwandamir, Mirza Abu Bakr greatly enhancedhis position in the Timurid political arena thanks toMir Mazid's joining him in Badakhshan with a largeforce. He was then able, as Khwandamir portrays it,to negotiate an alliance with Sultan Husayn Bayqara,the traditional enemy of his father. At a meeting ofthe two sides at Dih-i Qazi near the Amu Darya, MirMirza Abu Bakr made his peace with Sultan-HusaynBayqara in the presence of Mir Mazid Arghun andwas given a daughter of Sultan-Husayn's in marriageto seal the alliance. His amirs, as well as other wor-thies at the meeting, were honored with gifts of cere-monial robes and gold. (Khwandamir here singlesout Mir 'Ali Shir Nawa'i and Mir Mazid Arghun byname to signify the two top amirs on either side.)Later, after the Badakhshan delegation returnedhome, Khwandamir writes, "When the prince reachedBadakhshan, he suddenly found fault (bi-ranjfd) withAmir Mazid and at the urging of a group of atakchzswith whom the amir's privileges and authority (ikhtiyarwa i'tibar) did not sit well (muwdfiq-i mizdj-i shdnnabid), he was martyred."4 0

Writing between 1551 and 1555, some fifteen yearsor so after Khwandamir's death, Fakhri Harawi cor-roborates the former's account but in rather differ-ent terms. Without offering any motives for Mirza AbuBakr's actions, he writes that after the prince partedcompany from Sultan Husayn ("at night, without per-

mission" in contrast to Khwandamir's account in whichthe two parted on good terms and explicitly with SultanHusayn's permission), he returned to Transoxiana and"there had Amir Sayyid Mazid, one of his illustriousfather's amirs, put to death-because of his [Abu Bakr's]own ill-starred fortune (az bi-tli'-yi khftd).'41

The Arghun amir's downfall is difficult to explainbut seems linked to the new alliance, cemented bymarriage, between Sultan Husayn Bayqara and MirzaAbu Bakr, the son of Sultan Husayn's long-time nem-esis. For Khwandamir the alliance at the time of MirMazid's death was secure and his "martyring" led di-rectly to Abu Bakr's loss of Badakhshan. Fakhri Haravi,on the other hand, has the meeting between the twoTimurids ending in disarray. Whatever the case, theresult for Mir Mazid was the same. He had led sever-al expeditions on Sultan-Abu Sa'id's behalf againstSultan Husayn and was probably seen by the latter asan unhappy reminder of the former's victories. Hemay thus have served as a convenient scapegoat, andhis death perhaps cleared some political accounts.

Mir Mazid as Patron of Khwaja Abu Nasr

At least two texts, seemingly representing indepen-dent traditions, link the names of Khwaja Abu NasrParsa and Mir Mazid Arghun. One of these is theNMa'dsir al-mulk, another work by Khwandamir, butwritten at the very end of the fifteenth century, be-fore the final version of the better-known Habtb al-siyar. The other text is the Majma' al-ghart'ib of Sul-tan Muhammad b. Darwish Muhammad, a native ofBalkh writing in the early part of the second half ofthe sixteenth century.

It is perhaps a little strange that despite the com-paratively abundant information we have for each ofthe dramatis personae, these are the only two texts thatsuggest any connection between the two men. Andneither text offers unambiguous information that themen were personally acquainted, although their livesand their spheres of activity did, in one year at least,overlap. There is a good deal of contemporary cir-cumstantial evidence, however, to suggest that evenif Mir Mazid did not know the Naqshbandi shaykhwell, he might have wanted to align himself with theposthumous figure of Abu Nasr by becoming his com-memorating patron.

Khwandamir is the earliest to tell of a link betweenthe two men. Writing sometime just before 1497, thirty-five years after Abu Nasr's death and almost thirty

__ s�_ �II �lllrC� ��___

98

CONSTRUCTING THE KRHWAJA ABU NASR PARSA SHRINE

years after Mir Mazid's, Khwandamir says:

One of the sultan-i sa'zd's [Sultan-Abu Sa'id's] amirs,AmirJalal al-Din Mazid Arghun, built a spacious struc-ture ('imarat-i wasf') and a madrasa at the shrine (barsar-i mazar) of the Refulgent Saintly Khwaja Abu NasrParsa-May God sanctify his secret!-so that that noblesanctuary would greatly prosper and the income of itsendowments would flourish. 2

The text indicates: (1) there existed a shrine (mazar)for Khwaja Abu Nasr at Balkh; (2) the Arghun amirthen built an 'imarat-i wast', whatever that precisephrase might have meant to Khwandamir, and a ma-drasa or seminary; and (3) the purpose of theseprojects was to provide permanent endowments andthereby ensure the shrine's long-term well-being. Im-plicit, perhaps, is the modest nature of the shrinebefore the construction and endowment of these twobuildings. Khwandamir assigns no date to the work,but it was obviously done between the shaykh's deathin 1460-61 and the amir's death around 1470. Theterm 'imdrat-i wasz is somewhat vague in Khwandamir'susage and might be interpreted as referring to themausoleum itself ("a spacious structure") or perhapsto an endowment or foundation ("an extensive foun-dation") set up by Mir Mazd to support the site, ei-ther interpretation of the term 'imarat is possible atthis time. The concluding phrase in the passage ("sothat the noble sanctuary would greatly prosper andthe income of its endowments would flourish") mightbe taken as supporting the second interpretation-that it referred to an endowment made by Mir Mazidrather than an actual building-but there is no oth-er evidence that this was the case.

The second text, the Majma' al-ghara'ibfi bayan al-'aja'ib wa'l-kawakib wa'l-nawa'ib (The Compendium ofMarvels on Wonders, Celestial Bodies, and the Turn-ings of Fate) ,43 amplifies the passage in Khwandamir'swork and suggests that we should interpret 'imarat-iwasi as "spacious structure," specifically the mauso-leum building that still stands, though much altered,today. The Majma' al-ghara'ib was written in the 1560'sby a Balkh native and an official of the judiciary, SultanMuhammad (d. 10 Muharram 981/13 May 1574), theson of Darwish Muhammad. Like his father beforehim, Sultan Muhammad held the post of mufti atBalkh and as an educated man trained in the law hehad a fondness for precise data. His work includeselements of cosmography, geography, history, auto-biography, and the urban topography of Balkh and-

Bukhara and was based on sixty separate sources whichare cited in various places in the text.4 4 The worksurvives in two editions, the first of which was dedi-cated to the appanage ruler of Balkh, Pir Muham-mad Khan, aJani-Begid/Shibanid and completed by1561; the second completed sometime after 1577 anddedicated to 'Abd Allah Khan, a more famous JaniBegid/Shibanid who was by that date in control ofBalkh. 45

The work proved very popular in Central Asia, es-pecially in the nineteenth century, judging by the num-ber and dates of the manuscripts that have survived,but later copyists exercised a considerable degree oflatitude over what was included or omitted. Accord-ing to Tagirdzhanov, the contents of the manuscriptshe had seen of the second edition varied consider-ably. A later work, the late-seventeenth-century Trzkh-i Raqimi, which used Majma' as one of its sources,cites a manuscript that obviously had more materialabout Abu Nasr Parsa's shrine than the manuscriptsavailable to me. Without a comparison of all the avail-able manuscripts of the work it is not entirely cer-tain that the text which links our two principals actu-ally dates to the sixteenth century and not some latertime. However, on the grounds that the only othermodern scholar who cites this passage (A. Mukhtarov)was using a different manuscript of the work than Iwas and since the author of Tarikh-i Raqim, who waswriting around the end of the seventeenth century,also cites the passage, it is reasonable to concludethat it was included by the time the second editionof the work was compiled. Khwandamir's informa-tion also supports such a conclusion.

The passage is brief and occurs in the third chap-ter, "On regions and cities and the wonderful thingsthat happened therein." After a brief section on Bukha-ra (describing the new city walls built by 'Abd al-'AzizKhan in 951 [1544-551; Ibn Sina's birth there in 373[983-84] and an incident that happened to Ibn Sinain Isfahan), Sultan Muhammad gives a relativelylengthy (two-and-a-half page) discourse on Balkh, mostof it devoted to its buildings. Midway through it, af-ter some general remarks on the prosperity of thecity and its surrounding region, he writes:

The buildings of the illustrious shrines ('imrrdt-i mazdrit-ifayz anwiir) of that Excellency, the Lodestar of Lode-stars, Khwaja Abu Nasr Parsa, were built under thepatronage (bi-ihtimdam) of Mir Mazid Arghun in 867, thesurviving lofty gunbaz and the structure of the madrasa

__·__l_____l_____a____s___l____ 11�1�_

99

R.D. MCCHESNEY

enclosing that shrine being tended to through the pa-tronage of the Excellency, Guide of the People of theRegion, the Sublime Shaykh al-Islam, the Hazrat-i Khu-dawandi-May God extend his sublime shadow. After[that work], through the divinely bestowed good fortuneof the constantly guiding and caring Hatrat-i Khuda-wandi, a madrasa of baked brick comprising numerousrooms, large cells, and [provided with a source of] dailyincome [was built] to-the south of the Mazar Road andwest of the Khiyaban Road near the shrine and holyresting place, and he endowed that madrasa with his pri-vate property-May God bestow on it an abundance ofbenefits.4 6

The author then goes on to enumerate the dates ofa number of other buildings in Balkh-the "build-ings of Mamlakat Aghaz [sic],"47 completed in 832(1428-29), a congregational mosque erected in theinner city of Balkh by Sultan Husayn Mirza in 900(1494-95) and paid for with "40,000 Kapaki dinars"which had been presented to Sultan Husayn on avisit to Balkh by its people "as a gift" (bi-rasm-i pshkash)along with 100,000 dinars which were levied as a taxand an additional 70,000 dinars (paid out of SultanHusayn's pocket?),48 and a royal bath sponsored byKistan Qara Sultan on which construction began on12 Zu'l-Qa'da 934 (29July 1528) and was completedin 937 (1530-31). Kistan Qara Sultan gave SultanMuhammad's father, Darwish Muhammad, the hon-or of laying the first brick, according to his son's re-port.4 9

In Sultan Muhammad's description of Abu NasrParsa's mazar Khwandamir's 'imarat-i wasz' shouldprobably be identified with the "lofty mausoleum"(gunbaz-i 'alT-astr'5 0 ) of Mir Mazid Arghun, and weare given a precise year, 867 (1462-63), for its con-struction. The text also reaffirms the presence of amadrasa, called here "the surrounding or embracingmadrasa" (madrasa-i muh4t), a term suggesting a struc-ture which the earliest photographs seem to depict(figs. 1-2). The wording of the rest of the sectionyields the not surprising information that the shrinecomplex had, in the preceding century, been subjectto the ravages of time and that it was now being ren-ovated by a person referred to as "Hazrat-i Khudawan-di." In addition, by this time a new madrasa had beenadded next to the shrine precincts. The author's useof the plural "shrines" (mazarat) in referring to thesite of Abu Nasr Parsa's grave indicates that by thispoint it had become a cemetery with several sancti-fied sites worthy of a ritual visit (ziyara). In all proba-bility these included the tombs of the three or four

generations of Khwaja Abu Nasr Parsa's descendantsby now interred there as well as others for whom burialwithin the charismatic range of Abu Nasr Parsa wasdeemed desirable.

Sultan Muhammad is not the first to use the wordgunbaz (mausoleum) for the building constructed byMir Mazid. Samarqandi, also wrote of a gunbaz builtby the Arghun amir at Balkh. What is particularlyinteresting for us, though, is that Samarqandi, whoshould have known, seems to say that the the gunbazwas built as a family necropolis, and not specificallyfor Abu Nasr. At least one manuscript of Samarqan-di's chronicle says that when Mir Mazid's brother,Amir Sayyid Asil al-Din, died in Badghis in May 1465(Ramazan 869), Mir Mazid performed the funeral ritesfor him there, and then took the body to Balkh wherehe interred it in a gunbaz he had built for his fatherPir Muhammad. 5 1

Did Mir Mazid build two mausolea, one for hisfather and one for Khwaja Abu Nasr Parsa? Or werethere two narrative traditions about one mausoleum-that it was built for Mir Mazid's father and only lat-er, but no later than the 1490's, had come to be knownas the mausoleum of Abu Nasr? There is a consider-able amount of uncertainty here. The editor of theavailable printed edition of Samarqandi's Matla' al-sa'dayn, Dr. Mohammed Shafi', used six manuscriptsof the work in preparing his edition in the 1930'sand 1940's. Only one of them, a manuscript in thecollection of Eton College, contains any of this in-formation.5 2 Bregel's translated and expanded ver-sion of Storey, on the other hand, lists at least fortycopies of the second part (qism or daftar) of the work(the section at issue here) found in collections aroundthe world, and one would need to see many morecopies of the work to see if the reference in the Etonmanuscript is unique.

We can perhaps for the moment resolve the ques-tion of Mir Mazid's intention for the mausoleum ifwe think of the traditions as complementary and cor-responding to burial practices of the time. If we takeelite Timurid funerary practice as the model for MirMazid then we might reasonably conclude that hebuilt a gunbaz at the tomb of Abu Nasr Parsa andinterred his father and later his brother in it. Thiswould have conformed to Timurid funerary practiceas exemplified by the necropolis at Kash (Shahr-iSabz), the Gur-i Amir in Samarqand, Shahrukh'sensemble, Gazurgah at Herat, and the mausoleumof Gawhar Shad there. Given the desirability of hav-

I� I

100

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

Fig. 2. View of the Parsa shrine and its precincts with remains of the "encircling madrasa" visible to the left of the entry iwan.Photo taken 1924-25. Detail of plate IX(c), Centre de la ville, from A. Foucher, La Vieille Route de l'Inde de Bactres a Taxila (Paris,1942)

ing a holy man associated with the site and perhapsbecause of a fleeting personal acquaintance with AbuNasr, Mir Mazid, newly arrived in Balkh at the timeof the shaykh's death, chose the site of his grave asan appropriate place for a family necropolis. He builtthe gunbaz that still stands, though considerably al-tered, buried first his father and then his brother init, and probably also left instructions for his own burialthere. This interpretation might also account for themodern oral tradition that has Abu Nasr's gravesiteoutside the mausoleum (in front of the entryway) 5 3

for if he built it principally as a family necropolisthen placing the mausoleum in close proximity tothe head or feet of the shaykh would have probablysufficed.

In sum, Khwandamir and Sultan Muhammad pro-vide the narrative evidence linking the historical fig-ure of Mir Mazid Arghun, one of Sultan Abu Sa'id's

principal generals, with the somewhat more obscureKhwaja Abu Nasr Parsa. The latter's reputation ap-pears to have enjoyed a great post-mortem enhance-ment, undoubtedly due in large measure to the mag-nificent scale of the mausoleum built at his grave site.Mir Mazid, on the other hand, loses narrative signif-icance with the passage of time and the mausoleum'srole as a family necropolis in the Timurid style even-tually disappears from the narrative tradition, mem-ory of him, his father and brother and their asso-ciation with the mausoleum, completely eclipsed bythe legend growing around the name of Abu NasrParsa.

Despite the fact that Sultan Muhammad's workproved relatively popular down through the nineteenthcentury and was frequently copied, he is the last lo-cal source until the twentieth century to credit MirMazid with sponsoring the mausoleum construction,

�II� _·�I� 1� _____1_1__)_·___·1_____I_ ���

101

-·-':

··. ·;··-: :. ·i·i· ·-····;iw!.i ..-,,, ...·

;· - :·i··- · :'" :r;:n··l :·: :-.��;, ··?.,

R.D. MCCHESNEY

but he too already elides any role of the mausoleumas an Arghun family necropolis. The story of thebuilder continued to be preserved through repeatedcopying of the Majma' al-ghard'ib, but no other writ-er appears to have picked it up and breathed newlife into it. As far as we can tell from what is record-ed in writing of parallel oral narrative traditions, thename of Mir Mazid loses all local meaning. What-ever his plans for a family necropolis, they did notsurvive him. Even the story of those plans fades fromlocal memory. From the mid-sixteenth century on,the story of the mausoleum is exclusively Abu Nasr'sand by extension those who preserved and profitedfrom his memory.

It seems very unlikely that there were two mauso-lea built by Mir Mazid. It is more likely that two nar-rative lines about the same building converged intoa single authoritative one. The building probablybegan as a family mausoleum in the high Timuridtradition with a holy man lending it the whiff of eter-nity. As happened with the Gur-i Amir for a time,the holy man legend eventually became the build-ing's dominant meaning. In the case of the Gur-i Amir,local knowledge (not to mention the interest of Timu-rids elsewhere over time in preserving the family tra-dition) continually restoked the memory of both theTimurid and holy man ("Sayyid Barakah" of Andkhud)legends. At Balkh, the Parsa holy man tradition moreor less erased the Arghun story.

I do not intend to leave the impression that theimage of Abu Nasr Parsa as a saint was dependent onthe existence of the mausoleum. Other saintly leg-ends have survived in the soil of Balkh without sucharchitectural support. But the preservation of themausoleum and the perpetuation of its meaning un-doubtedly were major factors in determining the wayin which Abu Nasr was remembered. Had a mausole-um of such unusually large scale not been built, Ithink it more than likely that Abu Nasr's persona inlegend would have assumed quite a different charac-ter. In addition, the survival of the building may havebeen equally dependent on the perpetuation of theAbu Nasr legend.

Whether it was the building itself or the fact of asaint's presence there or a combination of both, it isclear that the site exerted a spiritual attraction fromvery early times, not, perhaps, as magnetic as theprecincts of Baha al-Din Naqshband's tomb at Qasr-i 'Arifan in the eastern suburbs of Bukhara or 'AbdAllah Ansari's at Gazurgah in the suburbs of Herat

or the shrine of 'Ali b. Abi Talib at Mazar-i Sharif.Nonetheless, it became an important place of pilgrim-age (ziyara) and a desirable place to be buried. Justhow desirable we can judge from one early and strik-ing case, that of a high religious official (sadr) at Herat,Mirak Jalal al-Din Qasim b. Mawlana Shams al-DinMuhammad Amin. In accordance with his explicitwishes, after his death in Herat on 15 Sha'ban 900(11 May 1495) his body was transported all the wayto Balkh to be buried next to Abu Nasr.5 4

THE PARSA FAMILY TRADITION AT BALKH

The complex of mausoleum and madrasa was clearlyan important element in maintaining and enhanc-ing the Abu Nasr legend, but what seems to have beencrucial to the longevity of the complex itself was theformation of a "dynastic family"55 committed to andbenefiting from the complex and the story and localmeaning of the shaykh/saint. Like the Ahrari familyat Samarqand and the Juybaris at Bukhara, the Parsa'iheirs to the shrine complex also enjoyed the right tothe title of shaykh al-Islam. This traditional associationof the family with the office lasted at least 250 years.According to the Bahr al-asrar (written ca. 1635) Par-sa family members held the office in Balkh "from thedays of Shahrukh and Ulugh Beg," 5 6 i.e., from thefirst half of the fifteenth century. The earliest confir-mation of this is found in Samarqandi, who refers toKhwaja Abu Nasr as shaykh al-Islam, though withoutany indication of when, or from whom, he receivedthe title.5 7

What the title precisely meant in terms of func-tions or duties, if any, is by no means clear. The lateShiro Ando, in a study of the title shaykh al-Islamunder the Timurids, discussed the two contexts with-in which the term usually appears-first, as a region-al or even kingdom-wide office with the bureaucraticand juridical responsibility of seeing Islamic law, orshari'a, implemented and,-second, as the title givento the head of a shrine with particular responsibilityfor its endowments. 5 8 For sixteenth- and seventeenth-century writers in post-Timurid Central Asia, the termobviously conferred or recognized high status amongreligious intellectuals. Being managers of the Abu NasrParsa shrine no doubt reinforced the family's long-term claim to the title. There is evidence that, likethe offices of qzf or mufti, some form of official rec-ognition was expected and indeed granted, so thatthe title or office had political significance as well.

_ �11�1� _� _�r�l_��_

102

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

In Samarqand, we are told by a late-seventeenth-century historian of the city, Muhammd Badi' Samar-qandi, the office of shaykh al-Islam was taken awayfrom the "Mirak sayyids" and given to the head ofthe Ahrari family.5 9 During the sixteenth and seven-teenth centuries at least, conferral of the title bypolitical authorities seems to have been a prerequi-site. There is overwhelming evidence that the headsof every generation of the Parsa'i family were locallysignificant political figures as well as shrine manag-ers, but whether the title of shaykh al-Islam recog-nized or conveyed that status is difficult to say. Cer-tainly for the recorders of the family's history overthe two and a half centuries covered here the titlewas the primary signifier of the status of Parsa lead-ers generation after generation. For about 250 yearsat least only one family in Balkh, the Parsa'i, was seenas the rightful bearer of the title.

The Second and Third Generations

Khwaja Abu Nasr's immediate successor appears tohave been a son named Khwaja 'Abd al-Malik. This isa somewhat tentative attribution, since we lack goodcontemporary information. What evidence there iscomes a generation or two after 'Abd al-Malik's death.Writing towards the end of the sixteenth century,Hafiz-i Tanish gives the genealogy of 'Abd al-Wali Par-sa, one of his own contemporaries, as "Khwaja 'Abdal-Wali, better known as 'KhwajaJan Khwaja' son ofKhwaja Abu Nasr Parsa [II], son of Khwaja 'Abd al-Malik, son of Khwaja Abu Nasr [I], son of KhwajaMuhammad Parsa." 60

Some thirty years before Hafiz-i Tanish wrote, Nith-ari, in his anthology of poets, included Khwaja 'Abdal-Hadi Parsa, who was buried at Balkh at the shrineof his ancestors.6 1 'Abd al-Hadi's lineage went back"through three [unspecified] generations" (bih sihwasitah) to Khwaja Muhammad Parsa. Nithari tells usthat 'Abd al-Hadi's father was Abu Nasr Parsa "theSecond," a figure well attested in Khwandamir andWasifi. 62 This would thus seem to make brothers ofHafiz-i Tanish's 'Abd al-Wali and Nithari's 'Abd al-Hadi. Abu Nasr Parsa "the First" would be the firstgeneration after Khwaja Muhammad Parsa, thus leav-ing us with one generation between Abu Nasr I andAbu Nasr II, a slot into which 'Abd al-Malik fits neat-ly.

Isfizari, writing'a full century before Hafiz-i Tan-ish, adds some evidence in support of the latter's gene-

alogy. In a brief reference to the three most promi-nent shrines of Balkh, Isfizari says that one of them"is the mazar of Khwaja 'Abd al-Malik Parsa."6 3 His isthe only work that I have seen to mention a shrineascribed to Khwaja 'Abd al-Malik, but given the dateof his writing it would at least make sense to see in'Abd al-Malik the link between the two Abu Nasrs.Moreover, as we will see below, 'Abd al-Hadi proba-bly died sometime between 1552 and 1566-67. 'Abdal-Malik, his grandfather, if grandfather he was, wasdeceased by 1491-92, about thirty years after Abu NasrI. The generational spread thus seems about right. 64

But I know nothing more than this about 'Abd al-Malik, and it remains to be conclusively demonstrat-ed that he was in the direct line of Khwaja Muham-mad Parsa-Abu Nasr I--['Abd al-Malik]-Abu NasrII-'Abd al-Hadi, and 'Abd al-Wali. Certainly Hafiz-iTanish accepted it as genuine at the end of the six-teenth century.

Khwaja Abu Nasr II is a somewhat clearer histori-cal figure, although it is not certain at all that weshould associate him with Balkh. The Iranian ex-patriate Khunji relates a conversation he had with aKhwaja Abu Nasr Parsa, who cannot, of course, beidentified with the first Abu Nasr. On the eve of Shiba-ni Khan's 1509 campaign against the Qazaq, Khunjitells us that he put the following question to AbuNasr Parsa: Was holy war (ghaza) against the Qazaqan obligation on every Muslim (afar 'ayn) or was itmerely a duty that the community as a whole wasobliged to fulfill (afaz kifayah)? Khunji, in his usualself-promoting fashion, first registers Abu Nasr's an-swer that it was a general obligation and could besatisfied by a sufficient number of warriors represent-ing the community and then proceeds to refute him.65

The context strongly suggests that Abu Nasr was inBukhara when the reported conversation occurred. 66

Khwandamir also mentions a "Khwaja Nasir al-DinAbu Nasr Parsa," who is probably to be identifiedwith Abu Nasr II. Khwandamir was born fifteen yearsafter Abu Nasr Parsa I died and therefore it is highlyunlikely that he was referring to him, particularly sincehe gives Abu Nasr I the laqab "Hafiz al-Din" else-where.6 7 Moreover, he portrays this Abu Nasr Parsaas a historian and authority on Mongol history, cit-ing him as the source for his own account of thehistory of Balkh at the time of Chinggis Khan's con-quests. Further, this citation suggests that Abu Nasrwas alive and related his report orally to Khwandamirabout the Chinggisid conquests.6 8 Khwandamir spent

________·_(__�__I___3_s_____________�

103

104 R.D.

some time between 1514 and 1520 at Balkh and thatseems a likely time for him to have consulted AbuNasr II, if indeed the shaykh was there at the time.6 9

There are one or two references to the Parsa shrineduring the time that Abu Nasr II would have beenshaykh al-Islam. One is in the account of the death,transportation, and burial of the Herati sadr men-tioned above.70 The second, nine years later, in 1504,comes in recounting the outcome of a complicatedconspiracy against Badi' al-Zaman Mirza, the last ofthe Timurids at Balkh, when Khwandamir, who wasprobably familiar with the mausoleum complex, of-fers a brief depiction of a bit of its architecture. Theconspirators, he reports, met in the assembly hall('ama'at-khana), while a trusted backer of Badi' al-Zaman was secreted in one of the rooms off the as-sembly hall eavesdropping on the cabal.71 SinceKhwandamir has already referred to the complex asa "spacious structure or complex ('imarat-i wasz') andmadrasa," it is quite possible that the room in whichthe spies for Badi' al-Zaman hid was part of the ma-drasa-yi muhit. The plan of the surviving gunbaz, whileshowing what could have been the "assembly room,"does not show adjacent rooms that could correspondto Khwandamir's description.

The Fourth Generation

Khwaja 'Abd al-Hadi. The activities of the family vis-i-vis the great gunbaz and the complex as a whole be-gin to come into fuller view with Khwaja 'Abd al-Hadi,first briefly profiled by Nithari in his anthology ofpoets and then mentioned indirectly by Sultan Mu-hammad b. Darwish Muhammad and after him di-rectly by the author of the TrZkh-i Rtqimi. Nitharirefers to 'Abd al-Hadi as the son of Abu Nasr II andthen says that he served as shaykh al-Islam of Balkhfor "a long time" and was highly regarded by the lead-ing citizens of the region. Nithari seems to have in-cluded him in his anthology because of a ghazal hecomposed and sent to the Shah Budaqid/Shibanid'Ubayd Allah Khan, when the latter was returning toBukhara from one of his Herat campaigns.7 Nitharireproduces the rather prosaic opening line of theghazal:

I give thanks that the khan of the age has arrived,the khan-protector of the religion, 'Ubayd Allah

Ghazi Khan, has arrived.73

Nithari leaves open the question as to both where

MCCHESNEY

and when 'Abd al-Hadi passed from the scene, say-ing only that "due to a brief illness, the bird of hissoul flew from the cage of his body and he was bur-ied in the precincts (darjiwar) of the shrine of theGreat Khwaja (khwdjah-i buzurgwar).74 The epithetkhwdjah-i buzurgwar is often applied to Baha al-DinNaqshband, and as heirs to his teachings it wouldnot, of course, have been surprising for the Parsa'isto have been buried near him in Bukhara. But theeditor of the text notes a variant reading of the pas-sage where the words, "great ancestor" (jadd-i buzurg-war) are substituted for khwdjah burzugwar. This read-ing might place him more firmly in Balkh. 75

The author of the early-eighteenth-century Tirfkh-i Raqim, relying on a version of the Majma' al-ghara'ibthat was not available to me, credits 'Abd al-Hadi withmajor renovation work at the shrine. Because of thisapparent connection between the Majma' and theTarfkh-i RaqimZ, I will consider them here as comple-menting each other. For both works, however, thereare so many manuscripts and so much variation be-tween them that any conclusions about what they sayregarding the Parsa mazar at Balkh must be some-what tentative. 76 The Trikh-i RaqimT text reads?

In 959 [1552] in Balkh, His Excellency, of guiding rank,'Abd al-Hadi Parsa, rebuilt (ta'mr namudand) the lustrousresting place of his noble forebears and distinguishedancestors. Mawlana Sultan mufti has written this qua-train:

"The date of 'Abd al-Hadi Parsa ['s work] is notrouble to obtain,

He built a shrine, his forefathers to contain,All you who circumambulate the Khwaja's crypt,

May calculate the date from the'Khwaja's crypt' script. "

The abjad (alphabetical) value of the phrase "Khwa-ja's crypt" (marqad-i khwdjah) is 959, equivalent to thecommon era date 1552. The text, though brief, isrevealing in several ways. It credits 'Abd al-Hadi withconstruction work at the shrine, the first member ofthe Parsa'i family for whom we have such a record.It tells us little about the nature of the work carriedout, but does name at least one of the sources of theTarzkh-i Raqimf's information-Sultan Muhammad-through, presumably, the Majma' al-ghara'ib. As faras I have been able to tell, neither of the Majma'manuscripts cited above mentions 'Abd al-Hadi orgives this chronogram with the date for the building.It is of course quite possible that a version available

� �-�--R�-·CIIIIC-·--�-·C- ------- ---

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

to Raqim did contain the chronogram.In Sultan Muhammad's text as it is available now,

there is considerably more information than is re-peated in the later collection of chronograms of theTdrkh-i Raqimi and the text is an extremely impor-tant one for understanding the evolution of the shrinecomplex. 77 It corroborates the information in Khwan-damir's Ma'asir al-mulbtk, without, seemingly, relyingon it, that the original construction was the work ofMir Mazid and included a madrasa (the "encirclingmadrasa"). This combination of college and mauso-leum in the mid-fifteenth century would have been acommon plan. The Gur-i Mir at Samarqand built atthe beginning of the fifteenth century and the mid-century mausoleum and madrasa at Herat of GawharShad, the wife of Shahrukh, are other examples ofthis plan, though only the mausolea in all three casessurvive to the present. The early-sixteenth-centuryMir-i 'Arab madrasa in Bukhara is a variant of thistradition in which the mausoleum is wholly incorpo-rated into the madrasa building.7 8

Khwaja 'Abd al-Wali Parsa. The Majma' text intro-duces a major problem: who was the "Hazrat-i Khuda-wandi" to whom work on the complex was attributed?Mukhtarov chose to treat the designation as a prop-er name, but it was much more likely an honorificand probably referred to 'Abd al-Hadi Parsa's broth-er 'Abd al-Wali, for the following reasons. The cru-cial identifying attribute used is shaykh al-Islam, whichallows us to rule out military figures and khans. Weknow it does not refer to 'Abd al-Hadi, who had diedby 1566 when Nithari wrote about him. Both the firstand second editions of the Majma' were composedafter 1566 and the text from the second edition usesepithets ("May God lengthen his shadow" and "MayGod bestow an abundance of favors on him") thatindicate that whoever the Hazrat-i Khudawandi was,he was still very much among the living when thesecond edition was written (i.e., sometime after 1577).

If we then accept the information as it stands fromboth the Majma' and the Tainkh-i RdqimZ (supposedlyciting the Majma'), it would mean that 'Abd al-Hadi,who must have died not long after 1552 "rebuilt" or"built" in some unspecified fashion the Parsa shrinein that year. Then his successor as shaykh al-Islamand keeper of the shrine, 'Abd al-Wali, the "Hazrat-i Khudawandi," besides refurbishing the original mau-soleum (the gunbaz-i 'dlh) and the "encircling madra-sa" added another madrasa of his own in the areanear the shrine.

'Abd al-Wali was a prominent public figure. Hewas shaykh al-Islam for some twenty years until hisdeath sometime after 1587 and in this capacity andperhaps even more in his role as civic leader in Balkhwas frequently noted in the public record of the re-gion. 79 In 1559-60, the Jani-Begid at Balkh, Pir Mu-hammad Khan (d. 1566) wanted to move his capitalfrom Balkh to Bukhara, newly acquired by his neph-ew, 'Abd Allah b. Iskandar. When the latter agreed,Pir Muhammad sent two men in advance to takecharge of the city. One of them was the Khwaja 'Abdal-Wali, who was clearly familiar with the city and itsleading citizens.8 0

'Abd al-Wali (aka Khwaja Jan Khwaja) appears re-peatedly in the pages of the chronicle written for 'AbdAllah b. Iskandar, the Sharaf-nama-i shahi by Hafiz-iTanish. His career as sketched there indicates thatthe family remained closely tied to Bukhara and mayhave divided its time between the two cities. Theauthor introduces 'Abd al-Wali early in his work, usinghim as a vehicle for forecasting the future greatnessof Hafiz-i Tanish's patron. In so doing he reveals thecontinued Parsa'i ties to Bukhara as well as the wayin which the intra-Shibanid factionalism drew in prom-inent religious figures. The author tells us that 'Abd.al-Wali himself related to him a dream in which twoof 'Abd al-Wali's ancestors appeared urging him toshow 'Abd Allah the respect due a future great leader.In the dream, the unnamed ancestors (we can imaginethat Hafiz-i Tanish's readers would immediately havethought of Khwaja Muhammad Parsa and perhapsAbu Nasr) compare 'Abd Allah to the late widely ad-mired leader of Bukhara, the Shah-Budaqid/Shibanid'Ubayd Allah b. Mahmud. Persuaded by the dream,'Abd al-Wali, who may have diedjust as Hafiz-i Tanishwas finishing his book in 1587, thereupon declaredhis backing of 'Abd Allah.

The story is told in the context of 'Abd Allah'sconquest of Bukhara from another Shibanid clan andthe growing tension afterwards between him and hisuncle Pir Muhammad who was nominal khan at thistime and was appanaged at Balkh. Hafiz-i Tanish nar-rates the Parsa'i leader's transfer of allegiance fromthe Pir Muhammad branch of the Jani-Begid Shiba-nids to the Iskandarids, 'Abd Allah's family, at a crit-ical moment. Hafiz-i Tanish also has 'Abd al-Wali inBukhara at or immediately after its conquest in 1556.In light of the connection of the family with bothBalkh and Bukhara and with their substantial inter-

� -·----�---··--·--·II�

105

106 R.D. A

ests in both places they had a stake in the outcomeof any conflicts that arose between the political lead-ers of the two places. Hafiz-i Tanish concludes thispart of the story with 'Abd al-Wali expressing contri-tion at the ill-will he had harbored towards 'Abd Al-lah (kulfata kih az an hazrat dar dil dashtam) and tell-ing theJani-Begid sultan of his dream and change ofheart.81

The structure of this story is formulaic but doesserve to cast 'Abd al-Wali as a person of considerableinfluence in both Bukhara and Balkh, at least in theeyes of 'Abd Allah's own memorializer. It is one ofseveral indications of the importance of the familyin the western regions of the Shibanid khanate, es-pecially in Bukhara, where one tends to think of an-other Naqshbandi dynastic family, theJuybaris, as be-ing preeminent at this time.

'Abd al-Wali had property interests in Bukhara (wefind him mentioned as the owner of land in the tmdnor (irrigation) district of Samjan and a caravansaraiwhich stood on the Rigistan of Bukhara sometimeafter 1533)82 and this would perhaps justify placinghim in Bukhara at the time of the narrated dream.

At the beginning of the Turki-Hijri year Pichi(n)1572 (979), 'Abd Allah b. Iskandar attempted to seizeTirmiz from his cousin Padshah Muhammad, broth-er and ally of Din Muhammad, who had taken Balkhfrom his father, Pir Muhammad. After a bitter anddestructive siege, Din Muhammad sent 'Abd al-Walito negotiate the surrender of Tirmiz, the terms forwhich 'Abd Allah accepted "because of the abundantfaith and the perfectly sincere feelings he held forthe great family and mighty clan of Khwaja Parsa." 8 3

The terms included the surrender of the defendersinto the protection of 'Abd al-Wali who was also ableto save the life of Padshah Muhammad. 'Abd Allahgave a feast in honor of 'Abd al-Wali and then per-mission to return home (presumably to Balkh). 84

We next hear of 'Abd al-Wali in October-Novem-ber 1577 (Sha'ban 985) on the occasion of the cir-cumcision at Bukhara of 'Abd Allah's son and heir-apparent 'Abd al-Mu'min. The description of thecelebration takes up six pages in the Sharafnama yetonly two of the shaykhs present, the head of the Juy-baris, Khwaja Sa'd "Khwaja Kalan Khwaja" and 'Abdal-Wali Parsa "KhwajaJan Khwaja" are mentioned byname. 85

'Abd al-Wali's most dramatic appearance in Hafiz-i Tanish's work comes with his narrative of the eventsof 987 (1579). On 11 Rabi' al-Sani 987 (Friday, 7June

MCCHESNEY

1579),86 Balkh was attacked by a coalition of exiledShibanid sultans and the Timurid ruler of Badakh-shan Shahrukh Mirza, who had come to the 'Alidshrine east of Balkh city to perform circumambulation(tawdj) of the 'Alid tomb (only a pretext, accordingto Hafiz-i Tanish). Their presence stirred enough con-cern that an army was assembled in which "the leadersand worthies of the market and the craft organizationslike the shaykh al-Islam, KhwajaJan Khwaja, his son,Dildar Khwaja, and others" played a prominent part.The Badakhshani threat was beaten back, "Chaghatay"(i.e., Timurid) heads rolled, and Balkh was saved forthe Jani-Begid cause.8 7 Besides his role in rallyingresidents to defend the city, it is interesting that 'Abdal-Wali should here be described as a leader in themarkets of Balkh, a characterization that fits with own-ership of a caravansarai in Bukhara and may suggestan official function as well. If the title of shaykh al-Islam of Balkh included supervision of the shari'a,as Ando described one possible function of the office,then supervision of the market inspector, muhtasib,would have given the shaykh al-Islam some influencein the marketplace.

In Ramazan 990 (September-October 1582), 'Abdal-Wali left Balkh to congratulate 'Abd Allah on hissuccessful campaign down the Syr Darya and his cap-ture of Yasi (Turkistan) from holdouts of the Suyun-jukid/Shibanid clan. He met the sultan at Nur Ma-hall, a camping ground a day from Karminah on theroad between Bukhara and Samarqand. He brought'Abd Allah a falcon as a welcoming gift and togetherthey went out hunting. They then rode on to Karmi-nah where they performed ziyarat (pilgrimage) cere-monies at the shrine of Hazrat-i (Qasim) 'Azizan. 'Abdal-Wali returned to Bukhara with the victoriousJani-Begid and afterwards made his way back to Balkh. 88

Two years later, in 1584, when 'Abd Allah, nowkhan in name as well as fact, brought an army southof the Amu Darya to campaign in Badakhshan againstthe Timurid Shahrukh Mirza, his own son, 'Abd al-Mu'min, who was stationed at Balkh, came to greethim accompanied by the shaykh al-Islam and by 'Abdal-Wali's two sons, Dildar Khwaja and Qasim Khwa-ja.89

Hafiz-i Tanish's final mention of 'Abd al-Wali Par-sa has him riding east to Khulm in late May 1586 towelcome 'Abd al-Mu'min back from his campaign inBadakhshan.9 0 Hafiz-i Tanish leaves little doubt thathe believed 'Abd al-Wali was the leading civilian fig-ure in Balkh during the thirty-year period before 1587,

_I I _______X______l__qR__IC___-�-ll -�---·r-�----. I� �------�-�

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

when he is mentioned for the last time (by Mahmudb. Amir Wali), or at least he wanted to memorializehim in that light.

'ABD AL-MU'MIN'S PATRONAGE OF THESHRINE

Reports of 'Abd al-Mu'min's patronage of the Parsatomb complex come from sources written severaldecades after 'Abd al-Wali's death. The earliest isMahmud b. Amir Wali who picks up 'Abd al-Wali'sstory where Hafiz-i Tanish leaves off. In 1587, 'AbdAllah Khan, having brought most of Mawarannahrunder his control, turned his ambitions towardsKhurasan. Taking advantage of a Safavid successioncrisis, he marshaled his forces to retake the strategi-cally and commercially important oasis of Herat. 'Abdal-Mu'min was summoned from Balkh to join himbut, or so Mahmud b. Amir Wali tells us, 'Abd al-Mu'min only left Balkh for Khurasan once he hadthe blessing of the leading lights of the region. Onlyfour individuals are named, the first of whom is 'Abdal-Wali, "KhwajaJan Khwaja."91

'Abd al-Wali's relations with the Shibanid princeand 'Abd Allah Khan's heir-apparent seem to havebeen particularly close. 'Abd al-Mu'min came to Balkhas a young man, probably no more than 15 years old.(He was born in Bukhara in 1568 and was sent toBalkh no later than 1583.) 'Abd al-Wali was an elderstatesman by that time and would have been a logi-cal person for the young prince to consult and con-fide in. No doubt when the appropriate moment came,it would not have been difficult to suggest using someof the funds under 'Abd al-Mu'min's control for re-furbishing the Parsa shrine.

Unfortunately, all the sources of information re-lating to 'Abd al-Mu'min's work on the shrine arelate, although some earlier ones refer to comparablekinds of renovations done by him at other sites. Theearly-eighteenth-century work, Thrikh-i Muqm Khanz,a source of modest historical value for sixteenth- andearly-seventeenth-century history, describes construc-tion of a madrasa by the late-seventeenth-century rulerof Balkh Subhan Quli, (r. 1651-81) and says that itstood "facing the main entry (taq) of the mazar ofKhwaja Abu Nasr Parsa which was built by 'Abd al-Mu'minKhan [my emphasisl."92 This passage is the earliestreference to the contributions of'Abd al-Mu'min tothe shrine and seems to indicate, when taken in con-

junction with earlier sources on building at the shrine,

that the taq (or pzshtdq) of the mausoleum, the greatentryway to the building, was his work. Writing inIndia about thirty years after the Tarikh-i Muqtm Khanthad been finished, Hajji Mir Muhammad Salim putsa slightly different twist on 'Abd al-Mu'min's addi-tions to the shrine:

Among the works created through the patronage of 'Abdal-Mu'min that remains as memorials to him are therevetting with blue tile of the taq and gunbaz at the shrineof Hazrat Khwaja Abu Nasr Parsa-may his grave besanctified-the like of which travelers to every cornerof the world have not witnessed either in terms of so-lidity or beauty.9 3

Both sources attribute shrine renovation to 'Abd al-Mu'min though the later work does not specificallysay the tq itself was the work of the Shibanid prince.Adding the entry taq and finishing it off with "Kash-an" work would have been in keeping with the kindof work more contemporary sources attribute to 'Abdal-Mu'min at Balkh. There are no major individualbuildings ascribed to him, but he seemed to haveenjoyed putting his own aesthetic stamp on existingbuildings with additions and redecorated exteriors.One of his larger projects was the addition of a domedannex to the shrine of 'Ali b. Abi Talib east of Balkhwhich doubled the size of the original fifteenth-cen-tury shrine building. 94 Another of his works, long sincevanished, was the installation of a blue-tile revetment(kash-karf) on the entryway to the Inner Citadel (theQal'ah-i Hinduwan) at Balkh.9 5 He also remodeled acovered market (chdrsf) called the BabaJanbaz mar-ket with a revetment of blue tile; constructed unspec-ified buildings ('imdrat) at the shrine of Khwaja'Akkashah, believed to be a companion of the ProphetMuhammad; 9 6 and is credited with another entry tqcalled the "Blue Gate" (darwdzah-i kabfid) and a khan-qah at the shrine of Hazrat Baba as well as an entry-way to the citadel at Andkhud (Andkhuy).97

There seems to be a definite pattern to his archi-tectural projects. He liked the effect of blue tile andwas liberal in supporting its application. He also en-joyed the effect created by the addition of massive(if the Parsa shrine is typical) entryways to shrinesand other public structures. It requires an on-siteanalysis of the Parsa shrine by an expert in buildingtechnique to see whether the fabric of the buildingcorroborates what Thrlkh-i Muqim Khang at least seemsto suggest-that is, that the great entryway which sodominates the visual impact of the building was add-

�C� ------- --- 111�--·---- --·-------- -I--·r*·r**llllrr�31Irrpasr�-·------ ------ ·- *----------

107

R.D. MCCHESNEY

ed to it at the end of the sixteenth century and wasnot part of the original plan.

By the end of the sixteenth century, the textualevidence paints a fairly detailed picture of the build-ing and the complex around it as well as the socialposition of the family associated with it. At the cen-ter of the complex was the gunbaz or mausoleum ofAbu Nasr Parsa I, recently refurbished by 'Abd al-Mu'min, who possibly added a massive new entryway.By now the aura of Abu Nasr Parsa had come to dom-inate completely the meaning that was attached tothe building. Any associations with the name of MirMazid's father, for whom it may also have been built,or with Mir Mazid or his brother Asil al-Din had seem-ingly been forgotten, at least in the surviving litera-ture composed in this period. The shrine itself wasflanked by a madrasa, according to Khwandamir, ora madrasa-i muhit or "encircling madrasa" accordingto Sultan Muhammad. It is possible that early photo-graphs show part of the plan of this madrasa, althoughthe actual arcades suggestive of a madrasa that arevisible in the photos could date from any period. Anarchaeological probe of the site for the original foun-dations may be the only way we will know whetherwhat those photos show is to be conclusively identi-fied with the "encircling madrasa" mentioned by SultanMuhammad. Whether that in turn is the same ma-drasa Khwandamir referred to earlier in the centuryalso remains an open question. But the similar ex-amples of Gawhar Shad's madrasa-mausoleum com-plex in Herat and the Gur-i Amir, Mirza MuhammadSultan/Amir Timur's madrasa-mausoleum complexin Samarqand, at least place such a project at Balkhwithin the architectural vernacular of the time.

POLITICS AND THE SHRINE (1598-CA. 1698)

By the end of 1598, the Shibanid khanate had cometo an end. 'Abd Allah Khan died in February of thatyear and his son and patron of the Parsa'is, 'Abd al-Mu'min, reaping the whirlwind of his father's effortsto eliminate all rival Shibanid clans, was assassinatedin August. Balkh now became one of the centers ofrival "khanates" seeking to fill the void left by thesudden collapse ofJani-Begid power. Within two years,another Chinggisid family, the Tuqay-Timurids (alsoknown as Ashtarkhanids), claiming agnatic descentthrough Chinggis Khan's eldest son Jochi, won thebacking of the Uzbek amirs and established them-selves by 1599 in Bukhara and Samarqand. In 1601

they managed first to take Balkh from aJani-Begid/Shibanid who was backed by the Safavid ruler Shah'Abbas I and then consolidated their hold by turn-ing back a Safavid army later in the same year.Throughout this three-year struggle leading to Tu-qay-Timurid ascendancy, we have no informationabout the Parsa'is or the shrine, and so there is littleto suggest how they might have been affected by theunsettled political conditions. Following the deathof the first major figure in the Tuqay-Timurid line,Baqi Muhammad b. Jani Muhammad, there was a five-or six-year period during which two of his nephews,Imam Quli and Nazr Muhammad, both sons of hislate brother Din Muhammad, were pushed by theirUzbek supporters, or willingly participated, in oppos-ing and eventually overthrowing another uncle, WaliMuhammad. From then (1611) until nearly mid-cen-tury, these two brothers shared real control (althoughImam Quli as elder bore the title "great khan") ofthe territory of the neo-Chinggisid khanate, with ImamQuli in Bukhara and Nazr Muhammad at Balkh. Thisperiod of relative political calm is covered most fullyby Mahmud b. Amir Wali, who was writing for NazrMuhammad. Because of his detailed description ofthe city of Balkh, 98 he is the main source for the familyand the shrine for the first three decades of the newcentury.

The Fifth Geeration

'Abd al-Wali (KhwajaJan Khwaja) had two sons thatwe know of, Dildar Khwaja and Qasim Khwaja. Dil-dar, the eldest, does not seem to have inherited theoffice of shaykh al-Islam, although there is a periodof time between 1587 and 1606 about which we knownothing. He may have predeceased his father. 'Abdal-Wali's exact date of death is not yet known, butoccurred sometime between 1587, when he is lastmentioned, and 1606-7, when Qasim is first men-tioned as incumbent in the office.

The earliest mention of Qasim's being shaykh al-Islam has him welcoming Nazr Muhammad to Balkhwhen the latter is sent there by his uncle Wali Mu-hammad. 99 He was thus presumably aware of andperhaps involved in the political maneuvering whichled Nazr Muhammad (pressed by his amirs) to joinforces with his brother against their uncle. Later,Qasim Khwaja also appears in a role premiered byhis father, that is, helping organize the defenses ofBalkh against attack. After the overthrow and death

__��rl� �

108

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

of Wali Muhammad in 1611, his son Rustam Muham-mad took up the gauntlet for the rights of the Walidline. But he was forced into exile from Chinggisidterritory and had to operate out of Herat under theprotection of the Safavids. In the same year that hisfather was killed he was given a force of Qizilbashtroops and led them as far as the walls of Balkh. NazrMuhammad had already departed the city to helpImam Quli fend off Qazaq attacks on Samarqand andhad left a Qunghrat amir in charge. When the threatof Rustam Sultan appeared on the horizon the amirconsulted with the notables of the city, two of whomare specifically mentioned, the shaykh al-Islam QasimKhwaja and the chief judge Qazi Tulak. These twoagreed to organize the citizens of Balkh into a laborbrigade to shore up its walls and gates. When Rus-tam Sultan set up camp a mile west of the walls, the"army" mustered to attack them included the Qungh-rat amir and his men as well as "the shaykhs, schol-ars, nobles and notables (mashayikh wa 'ulama' waashrdf wa ayan) led by Qasim Khwaja shaykh al-Is-lam, wage earners (kasabah), and craftsmen (arbb-ihira) led by Pir Muhammad Arbab and the rest ofthe [people of the] outlying districts (tfimdndt) andthe nomadic populace (qushfinat) accompanied bytheir leaders (ru'asa-yi khid)."'°° In the face of thisunified response, Rustam Sultan beat a retreat withhis Qizilbash troops.

Until his death in 1624-25, Qasim Khwaja enjoyedclose ties to Nazr Muhammad, or so Mahmud b. AmirWali tells us, and was constantly consulted by theTuqay-Timurid khan. 0l 1 Muhammad Amin Bukharistates that when construction began on Nazr Muham-mad's madrasa, Qasim Khwaja was called on to laythe ceremonial first brick. 102

THE FORMATION OF A MADRASA DISTRICT(CA. 1584-1684)

The Parsa shrine had a considerable influence onthe architectural development of Balkh. The late six-teenth and early seventeenth centuries saw the trans-formation of the area of the city surrounding the Parsacomplex with the creation of a large educational andresidential district. The renovations attributed to 'Abdal-Mu'min represent only a small part of the build-ing activity at or near the shrine that spanned theperiod from the late sixteenth to the mid-seventeenthcentury and offer tangible. evidence of the centralityof the district to the city's intellectual life.

By the middle of the sixteenth century there weretwo madrasas associated with the site, the "encirclingmadrasa" and the madrasa attributed to 'Abd al-WaliParsa the "Hazrat-i Khudawandi." By the end of theseventeenth century, at least four more large madra-sas would be built close to the shrine, only one ofwhich has left any traces at all outside textual reports.

'Abd Allah Khan Mladrasa or iVladrasa-i 'Aliyah. The firstmadrasa was built by 'Abd al-Mu'min's father 'AbdAllah Khan. We know his madrasa was near the shrine,although there is no clear indication of its exactlocation.10 3 I have found no evidence for the date ofits construction, but it was erected by February 1584.Hafiz-i Tanish, when describing 'Abd Allah Khan'scampaign against the Timurid Shahrukh in Badakh-shan, tells the reader that after settling in at the citadel"on Monday at the end of Muharram" 992 (earlyFebruary 1584), the first thing the khan did was paya visit to the "Madrasa-i 'Aliyah, which was one of hiscreations." 10 4 It is probably no coincidence that inHafiz-i Tanish's text, the reference to 'Abd AllahKhan's madrasa comes immediately after the authortells us that the Jani-Begid/Shibanid khan was greet-ed when he came to Balkh by the Parsa'i men, 'Abdal-Wali and his' sons Dildar Khwaja and Qasim Khwa-ja.'05

Mahmud b. Amir Wali reports that in the 1630's acertain Mawlana 'Abd Allah (known as Mullazadah),the son of Mawlana Jar Allah, simultaneously heldprofessorships at the 'Abd Allah Khan madrasa andthe "Khwaja Parsa madrasa" (probably 'Abd al-Wali'smadrasa), which suggests some administrative link-age and, perhaps, the proximity of the two madra-sas. 10 6

Nazr Muhammad Khan Madrasa. What corroboratesthe siting of the Madrasa-i 'Aliyah near the Parsa'ishrine complex is Mahmud b. Amir Wali's descrip-tion of another madrasa which his own patron, theTuqay-Timurid Nazr Muhammad, built sometimearound 1612.107 In the fashion of the time, it waslaid out as a mirror-image madrasa, to face the Ma-drasa-i 'Aliyah. As was also the custom, it was designedslightly to exceed its opposite in height.'0 8 Work onNazr Muhammad's madrasa began immediately af-ter work was completed on another of his vanishedbuildings, an elaborate palace complex.' 09 A late-sev-enteenth-century source, Muhammad Amin Bukhari,says the first brick was laid by "Qasim Khwaja Parsa'i,"'Abd al-Wali's second son,110 and dates the beginning

_ I_ _·I� � �C�I��

109

R.D. MCCHESNEY

of construction to "1021, the month of Sha'ban [i.e.,late September or October 1612] at the time of theconjunction of the two auspicious planets, Saturn andJupiter, when the moon was in the House of Tau-rus."111 Mahmud b. Amir Wali attributes the designand construction to three "competent engineers andarchitects" (muhandisdn wa mi'mdrdn-i kdrdan), MirQasim, Ustad Hajji, and Ustad Nur Muhammad. Themen are formulaically described as "advanced in thescience of yad-i bayiz and the secrets of i4hya al-mawta."112 The structure was comprised of arch andarcade elements (tdqdt wa riwaqdt), and it was deco-rated with lapis lazuli and revetted in blue tile (kash).Its entrance iwan was "several cubits higher" than the'Abd Allah Khan madrasa. It contained four cornerlecture rooms, a mosque, and many dormitory roomswith living allowances (hujardt wa farajat). It held atwo-thousand-volume library, donated as a waqf byNazr Muhammad, and by 1635 it supported a librar-ian, a provost (mutawallz), three professors, two ofwhom were qazis (judges), one of them the chief qaziof Balkh, Qazi Tulak. n13

From another Balkh native, Muhammad Tahir b. Abi'l-Qasim Balkhi, writing in the middle of the seventeenthcentury, we finally are given a location for this pairof madrasas.14 In the manuscript I was able to use 1 5

Muhammad Tahir speaks, however, not of madrasasbut of "two great masjids" standing on the west sideof the Parsa complex. Buri Akhmedov, using a dif-ferent manuscript, says that Muhammad Tahir refersto "two madrasas" but has incorrectly understood thisas meaning the madrasas of Nazr Muhammad and'Abd al-Wali Parsa.l 16 Mahmud b. Amir Wali is quitespecific in locating the Nazr Muhammad Madrasadirectly opposite, facing and reflecting the facade of'Abd Allah Khan's madrasa, and this fits better thearchitectural fashion of the time. Moreover, SultanMuhammad in Majma' al-ghar&'ib locates the 'Abd al-Wali madrasa, not to the west of the mazar, but "southof the Mazar Road (rah-i mazdr) and west of the Khiya-ban Road (rdh-i khiyaban) and near the mazar."117

The Mazar Road probably passed the Parsa complexon its south side and intersected the Khiyaban Road,which ran more or less north to south from the in-ner city towards Nawbahar past the entrance to theParsa shrine (fig. 3).

As was the case in Samarqand (the Tilakar madra-sa on the Rigistan, for example), it was quite com-mon to use madrasas as congregational mosques and

Hislr-i Jadid(new walls) /

(Timurid mbuilt by/

Shibanids) //

/ Hisar-inner City

Qad/ ' (old walls)

/I

/

Nazr nmadr

'Abd Allah Khan madrasa

Allh Yar divdnbegi l-madrsa ------

'Abd al-madras

Shlibanid wall

I\ Hislr-i adid

\(new walls)

Fig. 3. Balkh, ca. 1690.

the 'Aja'ib al-tabaqat's reference might have arisen fromthe use of the Nazr Muhammad and 'Abd Allah Khanmadrasas as mosques in his time. The shrine (gun-baz) itself with its large open interior and its "encir-cling madrasa" also served as a mosque. (In the nine-teenth century one of its names was "the greenmosque.") Mahmud b. Amir Wali, writing in the 1630's,mentions a man who held the post of khatzb at the"mosque of the mazar" of Khwaja Abu Nasr Parsa. 1 8

Mosque and madrasa were not exclusive categories.With Muhammad Tahir's description we are for

the first time given a sense of the extent of the Parsaprecincts as they were in the middle of the seven-teenth century and probably earlier as well. Accord-ing to him, the area covered by the shrine ('arsah-ifaizyijawanib-i mazdr) measured approximately fifteenjaribs or about seven to eight acres or three hectares,based on the approximate size of the jarib of theperiod. Panoramic photographs of Balkh taken in thenineteenth and early twentieth century show a lineof trees that may have followed the boundary of thisseven- or eight-acre site.1 19

Allah Yar Bi Qataghan Madrasa. Sometime just before1616, Allah Yar, an amir of the Qataghan tribe withthe rank of diwanbegz, sponsored the construction ofa madrasa which was erected next to the 'Abd AllahKhan Madrasa. Its endowment supported two profes-

1� �1_1�1�

110

:^z: ,

_ _

sorships, a khatfb, and an unspecified number ofQur'an reciters (huffaz).l2 0 Allah Yar was one of NazrMuhammad's senior advisers when the youth first wentto Balkh and had been a central figure in repellingRustam Sultan's efforts to retrieve the khanate fromNazr Muhammad and his brother.1 2'

Subhan Quli Khan Madrasa. By 1635 the district aroundthe shrine included at least five madrasas, the 'Abdal-Wali Parsa (or Khwaja Parsa) Madrasa to the southof the shrine; the 'Abd Allah Khan, Nazr Muham-mad Khan and Allah Yar Bi madrasas to the west ofthe mazar; and the "encircling madrasa " at the shrineitself (fig. 4). The sixth and, as far as I know, lastmadrasa to be added to the district, and the only onefor which some traces survived into the age of pho-tography, was built and endowed by Subhan Quli,one of Nazr Muhammad's sons. It was erected directlyopposite and to "mirror" the grand entry tq to theAbu Nasr Parsa Mazar built by 'Abd al-Mu'min in the1590's. Work began on the madrasa in 1660. "On the15th of Sha'ban 1070 [26 April 1660] he [SubhanQuli] assembled the worthies, the scholars, and thegreat shaykhs and intellectuals of the city" and brokeground for the building. 12 2 This time there were threehonorary initial bricklayers, but none were apparentlyParsa'is. There is little doubt, though, about the fam-

Parsa shrinearea

85

04

5.

Mazar Road

Fig. 4. Subhan Quli's madrasa. Tentative plan of surroundingsbased on the waqfnama.

-VJiA AU NASRK ARSA SHRINE 111

ily's continuing influential position.1 2 3

It is unclear when the madrasa was completed. Theendowment deed for it, though undated, was drawnup no earlier than 1686 and so was not put into ef-fect until five years after Subhan Quli had gone toBukhara as grand khan. Coincidentally it came at atime when he was struggling to keep control of thepolitical apparatus in Balkh.

According to the deed, the two-story madrasa hadseventy-five rooms (hujrahs) on each floor. All butone of the 150 rooms had a living allowance (farjah)attached to it. (The 150th room was assigned to theprovost or chief administrator [mutawallF] and hadno separate stipend attached.) The madrasa had asalaried staff of 24 including the mutawallh, four teach-ers (mudarrisan), a prayer leader (imam) a librarian,twelve Qur'an reciters (huffz-;-four full-time and eightpart-time), a muezzin, a barber, two janitors, andanother Qur'an reciter, whose only responsibility wasteaching children to memorize the Qur'an. 2 4

This is the only building in the vicinity of the AbuNasr Parsa shrine for which the actual endowmentdeed is known to exist. In the case of some of theother madrasas, we know they had waqf endowmentsbut the deeds themselves do not seem to have sur-vived. Because the deed describes the boundaries ofthe madrasa in terms of abutting properties, somesense of land use in the area is provided. The west-ern side of the madrasa, the part that faced the shrine,is described as bounded "entirely by land belongingto the waqf and at the disposal of the mutawall ofthe madrasa and is a place where people gather andmarkets are held" (majmii'-i khaltyiq wa bazdr-ast).125

This description corresponds well with the locationof the madrasa on the north-south Khiyaban Road,one of the principal thoroughfares of the outer city.It is not hard to imagine a bustling commercial areaof temporary stalls and people with their goods laidout on the ground filling the verge between the streetand the walls of the madrasa.

The north side of the madrasa abutted a residen-tial compound (hawilf) belonging to a craftsman, UstadMiryam Zirik (or Zarak) the son of another ustdwhose name is missing from both copies of the waqfdeed. It also touched on a small through alley whichis described as a public right of way.

On the east side, the back of the madrasa, its wallsabutted three residential compounds, one of thembeing Ustad Miryam's which must have wrappedaround the northeast corner of the madrasa. The other

CNTPRT'TT"'TT_ 'T'T, . -T -A _ -. - -.. . ..

_I�P�_ _

-ot IV· Y tnE - - -

R.D. MCCHESNEY

two belonged to a Mulla 'Abd Allah and to a stone-mason named Ustad Latif b. Ustad Niyaz.

On the south side, the madrasa shared a bound-ary with a primary school (maktabkhanah), a small ware-house (timchah), and a public toilet (mustardh) all threeof which were also made part of the waqf endowment.

Until at least the end of the seventeenth centurythen, the area around the Abu Nasr Parsa complexwas a residential and madrasa district with the usualmarkets lining the main thoroughfares. At least sixlarge madrasas were in place and standing in closeproximity to each other with the Abu Nasr Parsa com-plex serving as a symbolic focal point.

The Sixth Generation

We have no reason to believe that the Parsa familydid not flourish through the seventeenth century. Thegrowth of the district may have reflected their con-tinuing social power. There is no evidence that eventhe most serious political crises adversely affected thatpower. In mid century, the relative stability of theTuqay-Timurid state during the nearly forty years inwhich Nazr Muhammad ruled the lands south of theAmu Darya from Balkh and Imam Quli, his brotherwas titular khan and ruled the lands north of theriver, came to an end. Imam Quli, afflicted by an eyedisease, elected to step down in 1641 and perform afarewell hajj. For the next ten years, from 1641 to1651, a new political equilibrium was sought, first withNazr Muhammad as khan at Bukhara (an unsuccess-ful experiment in the eyes of the king-making Uzbekamirs) and then with his son 'Abd al-'Aziz, as khanand Nazr Muhammad back at Balkh, but under pres-sure to abdicate from another son Subhan Quli. Intothis the Mughal emperor Shah Jahan sent an armyled by his son, later emperor, Awrangzib, capturedBalkh and held it, not without cost, for about a year(1646-47), briefly driving Nazr Muhammad off thethrone. 12 6 Later, when Subhan Quli did gain controlof Balkh, he and his brother 'Abd al-'Aziz at Bukhara,over the period 1651 to 1681, replicated the juris-dictional pattern established by their father and uncle,although their relations were far less amicable thanImam Quli's and Nazr Muhammad's had been.

In 1624-25, sixteen years before the crisis erupt-ed, Qasim Khwaja died and his son-in-law and neph-ew Padshah Khwaja succeeded him as shaykh al-Islam.Padshah Khwaja was a son of Dildar Khwaja who, as

far as we know, had never been shaykh al-Islam andit seems likely that Qasim Khwaja had no eligible malechildren to succeed him.

According to what Mahmud b. Amir Wali, our solesource, has to say, Padshah Khwaja was a very activeparticipant in khanate politics like his uncle and grand-father before him. In 1632, he was part of a forceunder 'Abd al-'Aziz, Nazr Muhammad's son, sent toMaymanah and Andkhud to put a stop to Qizilbashviolations of the agreement which Nazr Muhammadand the late Shah 'Abbas (d. 1629) had reached overthe borders of the two kingdoms.12 7 Three years lat-er he was part of a delegation sent from Balkh tochastise 'Abd al-'Aziz, for an unauthorized incursioninto Safavid territory and to escort him back to Balkhto face his father's wrath.12 8

The last reference to Padshah Khwaja is in con-junction with the Grand Khan Imam Quli Khan'sceremonial trip to Balkh in 1639. During his visit Mah-mud b. Amir Wali tells us the khan visited the homesof three dignitaries, two of them Uzbek military leadersand the other Padshah Khwaja, "where he distribut-ed gifts to the servants (mulaziman)" of the shaykhal-Islam."1 29

The Seventh and Eighth Generations

Padshah Khwaja was succeeded by his son, namedafter Padshah's grandfather 'Abd al-Wali. The suc-cession occurred sometime between 1639-40 and1646. In the latter year, which witnessed the invasionand brief occupation of Balkh by Mughal forces ledfirst by Prince Murad Bakhsh and later by PrinceAwrangzib, 'Abd al-Wali Khwaja, who was also knownas "Hazrat-i Ishan" was summoned by Awrangzib tohis headquarters.l3 0 We are not told what transpired,and Mughal sources do not seem to have recordedthe incident.

After the occupation ended with the withdrawalof Awrangzib and the Mughal army in late autumn1647 and Subhan Quli gained control of Balkh, 'Abdal-Wali Khwaja was admitted to the appanage khan'sinner circle. He married one of Subhan Quli's sis-ters, Shamsah Begum (alternatively, 'Ismat ShamsahBanu Khanum, Shamsiyyah Banu Khanim) and by herhad a son, Salih Khwaja. 'Abd al-Wali also probablyplayed a part in encouraging Subhan Quli to com-mit resources to building a madrasa opposite the Parsashrine.

In 1696, the story of the family takes a decidedly

_ � � �I� __ � i�__l�

112

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

new turn, one that may account for the difficulty infinding information for the line in Balkh after 1700.The shrine complex appears to have continued tobe managed by a line with the Parsa'i family namebut it may have been a different line altogether, thatis, not direct descendants of 'Abd al-Wali II.

The end of the seventeenth century saw a rise inUzbek tribal identification with particular regions ofthe Tuqay-Timurid state and the gradual erosion ofthe authority of the Tuqay-Timurid khanate. A countertrend was the repeatedly articulated need expressedby all Uzbek tribal factions for a central Chinggisidfigure to legitimize their political claims. In this po-litical calculus, Balkh had come to be seen as therightful seat of the heir-apparent (wal-'ahd, qa'l-khdn,or turah) and whoever was installed at Balkh wasunderstood to have an a priori claim to succeed tothe khanate.

When Subhan Quli moved to Bukhara to take upresidence in 1681, his sister Shamsah Begum, wife ofthe shaykh al-Islam 'Abd al-Wali II, stayed behind inBalkh to manage her household in her palace inChaqar Khwaja Parsa (the name of the district orquarter in which the shrine stood). Her householdis described by not always impartial observers as be-ing one of the centers of political intrigue in the city.When Subhan Quli wished to eliminate a rebellioustwenty-six-year-old son, Abu'l-Mansur Sultan, at leastone source has him call on his sister to help him getrid of the prince without having to send an armyfrom Bukhara.13 1 Though herself the daughter of aChinggisid, her children had no particular claim tolegitimacy, but because of her personal power andthe exigencies of the time, the combination of Parsa'iand Chinggisid ancestry represented by her son Sa-lih Khwaja must have seemed appealing, at least briefly,to the Uzbek kingmakers. In opposition to the wishesof the grand khan, the leading Balkh amir, a Qat-aghan, is reported to have simply installed Salih KhwajaParsa'i as the Chinggisid heir-apparent in 1696. His"reign" lasted only a year, after which Subhan Quliacceded to the wishes of the amirs and appointed anauthentic Chinggisid, his own grandson Muham-mad Muqim. When that happened, the same amirwho had installed Salih Khwaja sent him into exilein India.132

After Salih Khwaja, the Parsa'i family's narrativerecedes into the mists (fig. 5). The exile to India maywell have terminated the importance of the line thatfor a quarter of a millennium held the post of shaykh

Abu Nasr (I) (d. 1460-61)*

'Abd al-Malik (?) d. c. 1491

Abu Nasr (II) fl. 1514-20*

I'Abd al-Hadi*(fl. 1552)

I'Abd al-Wali (I)*

(d. post-1587)

Dildar Khwaja Qasim Khwaja*(d.1624)

Padshah Khwaja* (fl. 1639-40)

'Abd al-Wali (II)*(fl. 1680)

Salih Khwaja (exiled 1696)*

*Designated "shaykh al-Islam."

Fig. 5. Genealogy of the Parsa Shaykh al-Islams

al-Islam and supervised the shrine complex. It is prob-ably safe to assume that this did not mean the endof the Parsa'is, however. If the textual legacy is in-dicative, the family's heyday in terms of social pres-tige and political influence came in the sixteenth andseventeenth centuries, but numerous factors make mereluctant to conclude at this point that the largerParsa'i family's fortunes underwent much of any kindof change at this point. For one thing, the textualrecord for the eighteenth century still needs thor-ough examination, even though it appears significantlymore modest than that for either of the precedingcenturies. Moreover, the mere existence of the shrineand its endowments along with the power of the leg-end of Abu Nasr would have privileged anyone whocould establish Parsa'i bona fides.

In addition, language, nomenclature, and techni-cal terminology, always evolving, seem to have beenparticularly fluid in the eighteenth century. Religioustitles like akhund and damulla, for example, eitherappear for the first time or take on much greaterprominence after the end of the seventeenth centu-ry. Such is also the case with personal and relational(nisba) names, and this makes it more difficult now

1____1____·__1�_1__1_ --· 0111�- 11

113

R.D. MCCHESNEY

to trace the family's history. In part this was due tothe changing political context as first the Iranian stateof Nadir Shah Afshar in the 1730's and then its suc-cessor in the region, the Afghan state of the AhmadShah Durrani in the 1750's, invaded, looted, and de-pleted the region of its manpower and resources.Migration to India and other regions offering betterprospects may have drawn off the educated classes,too. In the eighteenth century, there was also a dra-matic increase in the Turkification of the landscapearound Balkh as Qalmaq and Qipchaq nomads in-creasingly pressed on the agricultural lands.

Finally, possibly worsening economic conditions,due in part to the effects of the theft of resources bythe grand looting invasions of the Iranians and Af-ghans and in part by a changing world economy,curtailed the kind of patronage that supported greatliterary productions like the Bahr al-asrar or grandarchitectural projects like the madrasa building thatwent on at Balkh between the 1580's and 1680's. Sowe should not be greatly surprised at the disappear-ance of the family from the narrative record. Therecord itself seems to disappear or at least be greatlyreduced. But perhaps this is merely a matter of re-search yet to be done and a reflection of the lack ofinterest shown in the eighteenth century by scholarsof the region.

Nonetheless two things speak to the continuationof the Parsa'i legacy. One is the survival of the shrinecomplex while the nearby madrasas were collapsingand disappearing. We do not know precisely whenall the other great structures of the Parsa district suc-cumbed to the ravages of wind and weather or weredismantled for use elsewhere, but the survival of theshrine complex speaks to the continuance and localpower of the Parsa'i tradition. To maintain the shrinerequired periodic infusions of capital, the usual sourcesof which were income from properties administeredby the family and offerings brought to the shrine bypetitioners seeking the saint's blessing and divine in-tervention, both of which necessitated some form ofcontinuing administration. Legitimate administrationhad always been understood to be the prerogative ofthe descendants of Abu Nasr Parsa. Survival of thebuilding meant continuation of the shrine's admin-istration and administrative continuity required aParsa'i presence. It is not surprising, therefore, thatthe second piece of evidence attesting to the survivalof the family through the eighteenth century is theappearance in the nineteenth century of a lineage

calling itself "Gawhari-Parsa'i," 13 3 one that perhapsemerges from the disruption caused by the exile ofSalih Khwaja.

As will be seen in part 2, the eighteenth and nine-teenth century brought great changes to Balkh andto the shrine complex. Persian, Afghan, British, andRussian imperial schemes incorporated the region anddirectly affected the city and the shrine. They alsobrought indirect changes, opening up the meaningof the shrine to a much wider audience, one thatwould interpret the architecture of the complex innew and influential ways. The story begins to be re-told; elements of the old remain, but significant newaspects for new audiences are introduced, fundamen-tally altering the architectural development of themausoleum site as well.

New York UniversityWashington SquareNew York City

NOTES

Author's note: Many people have contributed to this work, someperhaps unwittingly. It is often difficult to know precisely whereideas come from so to all those who spoke to me about thisproject, offered criticism and suggestions, and provided new waysto think about it I am most grateful. Special thanks are offeredto Maria E. Subtelny, Lisa Golombek, Bernard O'Kane, the lateAnnette Weiner, and the late Donald Wilber and particularly toConstance J. McChesney for a close and critical reading of thepenultimate draft.

1. Lisa Golombek and Donald Wilber, The Timurid Architectureof Iran and Turan. 2 vols. (Princeton: Princeton UniversityPress, 1988), catalogue no. 59; George Michell, ed., Archi-tecture of the Islamic World: Its History and Social Meaning (NewYork: William Morrow, 1978), p. 263; A. U. Pope, A Surveyof Persian Art from Prehistoric Times to the Present (London andNew York, 1938), pp.1136-37, pls. 422-24;Jeannine Auboyer,The Art of Afghanistan (Prague, 1968), pls. 126-28. For themost recent analysis and one that closely complements thisstudy, see Bernard O'Kane, "The Uzbek Architecture of Af-ghanistan," in Cahiers d'Asie Centrale 8 (2000): 130-47.

2. 'Abd al-Razzaq Samarqandi, Matla'-i sa'dayn wa majma'-ibahrayn (henceforth Samarqandi), ed. M. Shafi' (Lahore,1365 H.), vol. 2, pt. 3, p. 1200; see note on p. 1292, whereSayyid Pir Muhammad is mentioned.

3. Safi, Fakhr al-Din 'Ali b. Husayn al-Kashifi al-Harawi, Rasha-hat-i 'ayn al-hayat (henceforth Safi) (Kanpur, 1911), p. 57.

4. Hamid Algar, "Malamatiyya," and "Nakshbandiyya," in Ency-clopaedia of Islam, 2d ed., 6:225b and 7:934; Maria EvaSubtelny, "The Making of Bukhara-i Sharif: Scholars, Books,and Libraries in Medieval Bukhara (On the Library of Khwaja

�L�__� _�_ I_ �II_� I

114

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

Muhammad Parsa)," forthcoming in Devin DeWeese, ed. ofa festschrift for Yuri Bregel.

5. Quoting Safi, V. V. Barthold casts Khwaja Muhammad as aleading partisan of Tamerlane's youngest son Shahrukh inhis battle against his nephew Khalil Sultan for control ofBukhara and later as an enemy of Shahrukh's son, UlughBeg (V. V. Barthold, Four Studies in the History of Central Asia,trans. V. and T. Minorsky, 4 vols. [henceforth Four Studies][Leiden: E.J. Brill, 1962], 2: 72, 115-16, apparently refer-ring to Safi, p. 61).

6. The Herat scholar and Sufi 'Abd al-RahmanJami is the prin-cipal source of biographical material for Abu Nasr. Latermemorializers more or less repeat Jami's formulation inNafahat al-Uns. Writing the work in 1476-79, Jami gives AbuNasr an entry that is brief compared with what he accordsMuhammad Parsa. "After him [i.e., Muhammad Parsa] inhis place [was] the fruit of his fine tree, Khwaja Hafiz al-Din Abu Nasr Muhammad b. Muhammad b. Muhammad b.al-Hafiz al-Bukhari--the mercy of God Almighty (be onthem)-who attributed [his knowledge of] the foundationsof the sciences of the Shari'at and the regulations (rusim)of the [Sufi] Way to his great father. He [Abu Nasr][sur]passed him in negating existence [i.e., in self-denial-dar nafy wujid] and in liberality with the existent [i.e., ingenerosity-wa baza-i mawfid]l, but they were miuch alike inconcealing and disguising their state (satr-i hail wa talbis),so it was never obvious that they had ever set foot on this[Sufi] path. He knew a bit of the sciences of this group [theSufis], in fact of all sciences, and if he were asked a ques-tion he would say, 'Let me consult a book.' He would opena book, and it would be at the very place that addressed thequestion, or one or two pages either side. One day at anassembly mention was made of Shaykh Muhyi al-Din al-'Arabi-God sanctify his secret-and his works. [Abu Nasr]quoted a saying of his father's: "The Fusfis (al-Iikam) is thesoul and the Futilhat (al-Makkiyah) is the heart" (fusfisjdn-ast wafutfthdt dil); he [Abu Nasr] also said, 'Whoever thor-oughly knows the Fusus his desire to follow [the teachingsof] the Prophet-peace be upon him-is greatly strength-ened.' He died in 865 [1460-61] and his grave (qabr) is inBalkh" (ami, Nafahat al-Uns, fol. 191a, from a late-eighteenth-century manuscript in my possession; see also the editionof Mahmud 'Abidi [Tehran, 1370/1991], p. 401). How toreadJami's intent in this entry remains something of a prob-lem for me. If read as if it were written without irony, thetext is mildly effusive about Abu Nasr's Sufi virtues (gener-osity, self-denial, devotion to the teachings of Ibn al-'Arabi,deep familiarity with the contents of books on all the 'uhlm),but if read as irony it limns a man of no particular schol-arly accomplishment, distinguished in the eyes of his intel-lectual contemporaries only by virtue of being his father'sheir. However, this is of little significance to the way in whichmemory of him developed in Balkh.

7. Samarqandi, p. 1055. In general, Abu Nasr's conventionallyaccepted laqab is "HAfiz al-Din" (according to Ghiyath al-Din Khwandamir, Habib al-siyar f akhbar afrad al-bashar, 4vols. [Tehran, 1333/1954] [henceforth Khwandamir, Habibal-siyar], 4: 5 and Safi, p. 63), or "Burhan al-Din" (accord-ing to Safi, p. 63). The laqab "Nasir al-Din" is given to his

grandson and namesake Abu Nasr Parsa b. ('Abd al-Malik).See Khwandamir, Habib al-siyar, 3: 38.

8. Mu'in al-Din Muhammad Zamchi Isfizari, Rawiat al-janndtfi awsaifmadinat-i Harat (henceforth Isfizari), 2 vols. (Tehran:Danishgah-i Tihran, 1339/1960), 2: 176.

9. Barthold, Four Studies, 2:170.10. Translation by Wheeler M. Thackston, of Khwandamir,

Habibu 's-Syar. Tome Three. The Reign of the Mongol and the Turk,2 vols. (Cambridge, Mass., 1994), p. 353 (equivalent to 4:5 of the Tehran edition).

11. Safi, p. 64.12. A manuscript of the Matla' al-sa'dayn owned by Eton Col-

lege gives the form "Farid" for Mazid (see note on p. 1230of the Lahore edition). This would have been an easy ortho-graphic error for a copyist. A. Mukhtarov, PozdnesrednevekovyiBalkh (Dushanbe, 1980), p. 56, also gives the form Farid.which he found in a manuscript of a later work, thesixteeenth-century Majma' al-ghard'ib.

13. Samarqandi, p. 1200.14. Isfizari, 2: 200, 293. It should be noted that in this edition,

the editor makes two persons out of Mir Mazid on p. 200,by inserting the conjunction wa (and) between "AmirJalalal-Din" and "Sayyid Mazid Bahadur" even though the nameis followed by a relative clause which is clearly singular (kih'umdah-i dawlat-i sultan-i sa'id bd). This edition also hasremarkably incoherent indexes.

15. See, e. g.,John E. Woods's analysis of the Timurid royal ge-nealogy, Mu'izz al-ansab, in his introduction to The TimuridDynasty (Bloomington, Ind.: Indiana University, ResearchInstitute for Inner Asian Studies, 1990). The Mu'izz givessome idea of the fairly large size of the Barlas tribe fromwhich the Timurid royal clan sprang, and perhaps one mayinfer from it similar demographics for other tribes like theArghun.

16. See Shiro Ando, Timuridische Emire nach dem Mu'izz al-Ansab:Untersuchung zur Stammesaristokratie Zentralasiens im 14. und15.Jahrhundert (Berlin: K. Schwarz, 1992), pp. 173-77.

17. Barthold, Four Studies, 2: 157.18. Barthold even raises the question whether Abu Sa'id was a

genuine Timurid or not, but later historians seem to havemore or less unanimously concluded that he was. See J.Aubin, "Abf Sa'id" in EI, 2d ed., 1:147b-48b; and Woods,Timurid Dynasty, introd.

19. Aubin, "Aba Sa'id," p. 148a.20. PeterJ.Jackson and Lawrence Lockhart, eds., Cambridge His-

tory of Iran, vol. 6, The Timurid and Safavid Periods (Cambridge:Cambridge University Press, 1986), p. 117.

21. Khwandamir, Habib al-siyar, 4: 200.22. If this is the same Khusraw Shah who figures so prominently

in the Baburnma, as appears to be the case, he is identi-fied there as a "Turkistani Qipchaq." See Zahir al-Din Mu-hammad Babur Padshah Ghazi, The Baburnama in English,trans. Annette Susannah Beveridge (1922; rpt. London:Luzac & Co., 1969), p. 49.

23. Asam Urunbaev, Pis'ma-avtografy Abdarrakhmana Dzhami iz"Al'boma Navoi" (Tashkent: Fan, 1982), p. 80, facsimile no.175.

24. Samarqandi, p. 1292.25. Ando, Timuridsiche Emire, p. 172, identifies 22 Arghun amirs

by name.

115

_ II C_ _�·_____________________

R.D. MCCHESNEY

26. Khwandamir, Habzb al-siyar, 4: 127.27. For a recent fairly complete summary of his career, see Ando,

Timuridische Emire, pp. 175-76.28. Samarqandi, p. 1190; Isfizari, 2: 199-200; Khwandamir, Habtb

al-siyar 4: 75-77.29. Samarqandi, pp. 1200-4.30. Isfizari, 2: 237; Khwandamir, Habib al-siyar 4: 89, 122;

Samarqandi, pp. 1216-21 does not mention his name here.31. Isfizari, 2: 248.32. Samarqandi, p. 1230.33. Khwandamir, Habib al-siyar, 4: 89.34. Samarqandi, p. 1292 n.35. Bartold, Four Studies, 3:29.36. Asam Urunbaev, ed., Pis'ma-avtografy Abdarrakhmana Dzhami

... 80 (facsimile no. 175).37. Khwandamir, Habzb al-siyar, 4: 190.; see also Ando,

Timuridische Emire, p. 174.38. A letter from Uzun Hasan to Sultan Bayazid reporting this

battle and its outcome has been published in L. Fekete,Einffihrung in die Persische Paliographie (Budapest: AkadhmiaiKiad6, 1977), no. 12, 145-149. Uzun Hasan names four ofAbu Sa'id's leaders at the battle including "Mazid Arghun"(146, 1. 14 of the document).

39. See e.g. Isfizari, 2: 282-83.40. Khwandamir, Habib al-siyar, 4: 157.41. Fakhri Harawi, Tazkirah-i rawiah-i salatin (henceforth Harawi)

ed. Dr. 'A. Khayyampur (Tabriz: Mu'assasah-i tarikh wafarhang-i Iran, 1345/1966), p. 31.

42. Khwandamir, Ghiyas al-Din b. Humam al-Din al-Husayni,Ma'asir al-muluik (bi-iammah-i khatimah-i Khulasat al-akhbarwa Qanfin-i Humdyfnz), (henceforth Khwandamir, Ma'asir)ed. Mir Hashim Muhaddis (Tehran: Mu'assisah-i Khidmat-i rasa, 1372/1993), p. 171. I am grateful to Professor MariaE. Subtelny for noting this reference and pointing me inits direction.

43. Sultan Muhammad, Majma' al-ghara'ib (henceforth Majma')IVAN, Uzbek Republic, no. 1494. For other mss. see C. A.Storey, Persian Literature: A Bio-bibliographical Survey (London:Luzac, 1958), II/1, pp. 135-37.

44. A.T. Tagirdzhanov, Opisanie tadzhikskikh i persidskikh rukopiseiVostochnogo Otdela LGU, vol. 1 (Leningrad: Izdatel'stvoLeningradskogo Universiteta, 1962), p. 395.

45. After the death of the Suyunjukid/Shibanid grand khan,Nawruz Ahmad (aka Baraq) Khan in the autumn of 1547,Pir Muhammad was recognized, at least by theJani-Begids,as supreme khan. His khanate lasted until 1561 when hisown son Din Muhammad ousted him. Pir Muhammad livedfor another six years, and A.T. Tagirdzhanov, who has donethe most thorough study of the work and its dedication,considers his death in 1566-67 the terminus ad quem for thefirst edition of the Majma'. But in light of the real fall frompower of Pir Muhammad in 1561 (his son Din Muhammadseems to have taken control of Balkh after his father movedto Bukhara and not to have relinquished it on his return),it is more than likely that the first edition was completedbefore that date. The second edition was by the author'sson. In the text of at least one of the manuscripts of thesecond edition, he gives the obituary of his "conscientiousand meticulous father, Mawlana Sultan Muhammad mufti

Balkhi, on the tenth of Muharram of this year [981]"(Tagirdzhanov, Opisanie, p. 396). I used a copy of the firstedition held by the Tashkent State University Library, inv.no. 09/903 (A. A. Semenov, Opisaniepersidskikh, arabskikh ituretskikh rukopisei Fundamental'noi Biblioteki SredneaziatskogoGosudarstvennogo Universiteta [Tashkent, 1935], pp. 40-41)and one of the second edition from the collection of theBiruni Institute of Oriental Studies of Uzbekistan (A. A.Semenov, Sobranie vostochnykh rukopisei [Tashkent, 1952], p.298, cat. no. 680, inv. no. 1494). Unfortunately, my notesare not sufficiently detailed to permit a useful comparisonof the contents of the two editions on the subject of the AbuNasr Parsa mausoleum. It is worth noting that most of thecopies available, whether with the Pir Muhammad or 'AbdAllah dedications, were copied in the nineteenth century.

46. Majma', fol. 16a-b. The text reads: "wa 'imarat-i mazarat-ifayz anwar-i hairat qutb al-aqtab Khwaja Abui Nasr Parsabi-ihtimam-i Mir Mazid Arghfin dar hashtsad shast haft binayaftah bdah kih gunbaz-i 'ali asar-i n 'imarat wa 'imarat-i an madrasa-i muhit-i mazar-i mazkfir-ast bi-ihtimam-i hairat-i hidayat-i ahali wilayat shaykh al-islam-i 'ali hazrat-ikhudawandi madda zillahu 'll ihtimam yftah wa ba'd azan bi-tawfiq-ha-yi hidayat-i azali wa 'inayat-i lam yazali 'alihazrat-i khudawandi mazkir barjanfib-i rah-i mazar-i mazkOrwa gharb-i rah-i khiyaban qarib bi-mazar wa manzil-imutabarrikah madrasa bi-khisht-i pukhtah mushtamil barbuyat-i kasirah wa hujarat-i kabirah wa fawa'id-i har rfazahwa amlak-ra bi-madrasa waqf namfidah-and waffaqahu'llahal-ifazah al-khayrat."

47. I.e., Malikat Agha, the widow of 'Umar Shaykh b. Timur (d.1394), later wife of Shahrukh (d. 1447); see Woods, TimuridDynasty, p. 20.) On the madrasa and her burial there in 1440-41, see Bernard O'Kane, Timurid Architecture in Khurasan(Costa Mesa, Calif.: Mazda, 1987), p. 22.

48. The passage is not entirely clear to me. It reads: "wa 'imarat-i masjid-i jami' -i darfn-i shahr bi-tawajjuh-i sharif wa [?]hazrat-i Mirza Sultan Husayn dar sal-i nuhsad az Hijrat.itmamyaftab wa chihil hazar dinar kapaki kih dar waqt-i nuzal-iSultan Husayn Mirza bi-wilayat-i Balkh mardum bar rasm-ipishkash wurdah bdand wa sad hazar dinar bar mardum-i wilayat-i mazkfir tawajjuh namudah-and bi-haftad hazardinar tamam shud."

49. Or so it appears. The text again is not wholly clear in themanuscript I used. (After the construction dates) "khisht-ian-ra bi-isharat-i Kistan Qara Sultan walid bi-raqim-i" [or"walid-i rqim-i"?] "in hurfif nihad." Sultan Muhammad alsocredits his father with composing the chronogram, hammam-i thir-i Balkhi ("the cleansing Balkhi bath"), which gives thedate 946. However, if one imagines the original chronogramas hammam-i thir-i Balkh (the "cleansing bath of Balkh"),the date 936 is produced.

50. Gunbaz, the most common contemporary term for the kindof building typified by the Abu Nasr Parsa mausoleum, sig-nifies a polygonal structure supporting a drum supportinga dome. The use of the term is also generally restricted tomausolea. (See Golombek and Wilber, Timurid Architecture,1: 467; and Robert Hillenbrand, Islamic Architecture: Form,Function, and Meaning [New York, 1994], p. 257).

51. Samarqandi, p. 1292 and note.

�--------- --- I CI ....- L��

116

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

52. Ibid.53. Recorded in Golombek and Wilber, Timurid Architecture, 1:

295, and heard again by a colleague, Amin Tarzi, who vis-ited the shrine in 1996. Perhaps Khwandamir's wording barsar-i mazar to denote the location of the 'imdrat-i wasz' cor-responds to the building's being sited beside rather thanabove the grave, but the expression is not a precise one.

54. Khwandamir, Habib al-siyar, 4: 205; O'Kane, "Uzbek Archi-tecture," p. 132.

55. On "dynastic families" in the Central Asian context, see myCentral Asia: Foundations of Change (Princeton, NJ.: DarwinPress, 1996), pp. 75 ff.

56. Mahmud b. Amir Wali, Bahr al-asrrfi mandqib al-akhydr(henceforth Bahr), London, India Office Library, ms. no.575, fl. 286a.

57. Samarqandi, p. 105558. Shiro Ando, "The Shaykh al-Islam as a Timurid Office: A Pre-

liminary Study," Islamic Studies 33 (1994): 253-80.59. Muhammad Badi' Samarqandi, Muzakkir al-ashdb, Tashkent,

Biruni Institute of Oriental Studies, ms. no. 58, fols. 248a-b. The story of the transfer of the title centers on an inci-dent that took place during a court session presided overby Khan 'Abd Allah b. Iskandar (d. 1598). According toMuhammad Badi', the Mirak sayyids had held the title "sinceChaghatay times." One prerogative of the Mirak sayyids wasto sit on a separate platform or tier (masnad) when the khanwas holding court. Another was their right to their own towelafter hand-washing. Angered when the leader of the Ahrarifamily used the towel, the shaykh al-islam stormed out of courtand 'Abd Allah Khan then transferred the title to the headof the Ahraris. This story has several levels of meaning, oneof which is to show how the office was seen to be embed-ded in political life.

60. Hafiz-i Tanish, Sharafndma-e shah, ed. and trans. into Rus-sian as Kniga shakhskoi slavy by M.A. Salakhetdinova (hence-forth Sharaf-namah [Salakhetdinova]) (Moscow: Nauka,1983),) pt. 1, p. 184 (translation); fol. 81b (facsimile text).To date, Professor Salakhetdinova has published two of thefour projected volumes. The first volume goes to the middleof 1567 (end of 974) and the second to the beginning of1579 (beginning of 987). For the subsequent period, I haveused the India Office Library ms. no, 574 designated hereas Sharaf-namah (IOL).

61. Khwaja Baha al-Din Hasan Nithari Bukhari, Mudhakkir al-ahbdb (henceforth Nithari), edited with critical notes andintroduction by Syed Muhammad Fazlullah (New Delhi: Min-istry of Education, 1969), p. 319.

62. See below, nn. 66, 67.63. Isfizari, 1: 156. "The prospering shrines (mazardt-i ma'mr)

with abundant incomes and endowments (ba madakhil wamawql~uat-i mawffir) of this country [Balkh] their like areperhaps not to be found anywhere else and [they include]the Khwaja 'Akkashah, Khwajah 'Abd al-Malik Parsa, [and]Khwaja Abo Nasr Parsa among others."

64. Compare this with Arends's note dating Muhammad Parsa'sdeath to 1441 (instead of 1422) and the death of Abu Nasrto 1537. He cites Safi for the former, but seems to haveconfused the two Abu Nasrs besides giving a non-standarddate for Muhammad Parsa. Moreover the 96-year spread

between father and son is possible but unlikely. See Fazlullahb. Ruzbihan Khunji, Mihmn-ndmah-i Bukhard (ZapiskiBukharskogo Gostia) (Moscow: Nauka, 1976), p. 174, n. 31.

65. Ibid., fol. 24a (p. 63 of the Russian translation).66. Zayn al-Din Wasifi, Bada'i' al-waqd'i', ed. B. A. Boldyrev, 2

vols. (Moscow: Nauka, 1961), vol. 1: 230-31, reports a moveby 'Ubayd Allah b. Mahmud Sultan (the Shibanid appanageholder at Bukhara, 1512-1540 and grand khan, 1533-40)from Qarshi where he was hunting to Bukhara to attendfuneral rites (ta'ziyah) for "Khwaja Parsa" who is only de-scribed as "one of the great descendants (az awlad-i amjad)of Khwaja Muhammad Parsa." No indication of the date isgiven, but the passage could refer to Abu Nasr II who flour-ished 1514-20.

67. Khwandamir, Habzb al-siyar, 4: 5.68. Ibid., 3: 38.69. H. Beveridge and J. T. P. de Bruijn, "KhWandamir," E12, 4:

1021a.70. Khwandamir, Habtb al-siyar 4: 205; O'Kane, "Uzbek Archi-

tecture," p. 132.71. Khwandimir, Habib al-siyar, p. 295. The plan of the shrine

building in Golombek and Wilber, Timurid Architecture, vol.2, plans 65-66 (cat. no. 59), does not immediately suggestwhere one could eavesdrop on a conversation in the assemblyhall.

72. The many campaigns of 'Ubayd Allah Khan to Herat arethoroughly documented in Martin Bernard Dickson, "ShahTahmasb and the Ozbeks," Ph.D. diss., Princeton Univer-sity, 1958. There are few clues as to when the ghazal wascomposed. 'Ubayd Allah Khan may be said to have returnedfrom Herat" any number of times in the period between 1510and 1537. 'Ubayd Allah did not become titular grand khanuntil 1533, but the title was often applied in literary con-texts to appanage rulers as well as the reigning khans.

73. Nithari, p. 320.74. Ibid., pp. 320-21.75. Ibid.76. For the work of Mir Sayyid Sharif "Raqim," Tdrikh-i Raqim

(var. Raqimi or Tagkirah-yi Raqim). I used the Royal AsiaticSociety, London ms. no. 162 and the Edinburgh UniversityLibrary ms. no. 246, a copy dated 1732.

77. To my knowledge only Akhror Mukhtarov has made use ofthe Majma' text (in Pozdnesrednevekovyi Balkh [Dushanbe:Donish, 1980], pp. 57- 58 (English trans. R. D. McChesney,Nadia Jamal, and Michael Lustig, Balkh in the Late MiddleAges [Bloomington, Ind.: Research Institute on Inner AsianStudies, Papers on Inner Asia, No. 24, 1993], p. 45).

78. O'Kane, Timurid Architecture in Khurasan, pp. 21-26, 106 andGolombek and Wilber, Timurid Architecture, 1: 47-56, pro-vide a good introduction to the functions and cross-influ-ences of the madrasa, mausoleum, and madrasa/mausoleum.

79. Bahr (I.O.L. ms. no. 575), fol. 332a for the date of his death.80. Sharaf-ndmah (Salakhetdinova), 1: 241 (text fol. Illb).81. Sharaf-ndmah (Salakhetdinova), 1: 183 (text fol. 81b-82a).82. See the affidavit (iqrnr) dated 19 F-ebruary 1572 (24 Ramazan

978) for "all the lands of the locale (mawi') of Biyan Tughanand Khuwaysta" purchased by Khwaja Sa'dJuybari. One ofits borders was the village of Qasr-i Mansur which was "theprivate property of Khwaja 'Abd al-Wali Parsa, shaykh al-islam,

117

_�1_1 _·�_I�

R.D. MCCHESNEY

the son of (after honorifics) Khwaja Abu Nasr Parsa" inE. E. Bertel's [originally prepared by F. B. Rostopchin], Izarkhiva sheikhov Dzhuibari (Leningrad, 1938), p. 421 (doc.338). A sale document summarized by A. A. Egani andO. D. Chekhovich, "Regesty Sredneaziatskikh aktov: Mate-rialy k svodnomu katalogu aktovykh istochnikov v sobra-niiakh SSSR," Pismennyepamiatniki Vostoka 1976-1977 (Mos-cow: Nauka, 1984), p. 105, mentions a caravansarai belong-ing to "KhwajaJan Khwaja son of thelate Khwaja Abu Nasr."

83. Sharaf-ndmah (Salakhetdinova), 2: 102 (facsimile of the textfol. 161b).

84. Ibid.85. Ibid., fols. 206b, 186 (Salakhetdinova omits the first

"Khwaja" in 'Abd al-Wali's name and enters him in the in-dex of proper names under "Dzhan").

86. June 7 was a Saturday, according to the Wiistenfeld-Mahler'scheVergleichungstabellen.

87. Sharaf-ndmah (IOL), fol. 261b.88. Ibid. (IOL), fol. 370a.89. Ibid. (IOL), fol. 411b.90. Ibid. (IOL), fol. 453b91. Bahr, Tashkent, IVAN, ms. no. 1375, fol. 248b.92. Muhammad Yusuf b. Khwaja Baqa Balkhi, Tdrikh-i Muqlm

Khani, Royal Asiatic Society, ms. no. 161, fol. 126a. The textreads (referring to Subhan Quli Khan's madrasa) azmustahdast dar wiliyat-i Balkh bi-ma4z&i t&q-i mazar-i KhwajahAbi Nasr Prsd (kih) bind-yi 'Abd al-Mu'min Khan madrasa tarttbydftah. A. A. Semenov in translating the text (Mukim-khans-kaya istoriia [Tashkent, 1956], p. 184) either had a manu-script containing the missing relative pronoun kih or un-derstood the necessity of it.

93. Hajji Mir Muhammad Salim, Silsilat al-saltitn (or Tawdrikh-i Badz'a) (henceforth Silsilat), Oxford, Bodleian Library,Ouseley 269, fol. 155a-b. The text reads: "wa minjumlah-imustahdasat-i himmat-i 'Abd al-Mu'min Khan kih yadgarmand-ast taq wa gunbaz kashi mazar-i hazrat-i Khwjah AbuNasr Parsa-qaddasa sirrahu-kih sayyahn-i rub'-i maskanbi-matanat wa khabi an darja-yi digar nishan na-midahand."

94. Bahr (IOL), fol. 318b. O'Kane, Timurid Architecture of Khura-san, p. 256, calls the building a dar al-huffaz because of itssimilarity to Timurid structures by that name at Mashhad.But this is not a Timurid building. Mahmud b. Amir Walisays in his time (the 1630's) it was called the jami'-i dsitanah,i.e., the "Friday mosque of the shrine.'

95. Silsilat, fol. 155b.96. Ibid.97. Mir Rabi' b. Mir Niyaz, 'Umdat al-tawrikh-i khdqann, Insti-

tute of Oriental Studies, Dushanbe, Tajikistan, ms. no. 2030,fol. 41a. The shrine of Hazrat Baba being referred to heremay be the shrine of Baba Sangu at Andkhud, called the"ziarat-i Baba Wali" by C. E. Yate. See his Northern Afghani-stan, p. 347. Yate was unaware of any story about work by'Abd al-Mu'min on the shrine.

98. Bahr (IOL), fols. 212b-216a, on the architectural and hy-draulic projects of the author's patron, Nazr Muhammad,and fols. 309b-374b on the history, topography, shrines,and notables of Balkh.

99. Ibid., fol. 286a.100. Ibid., fol. 191a.

101. Ibid., fol. 286b.102. Muhammad Amin Bukhari, Muhit al-tawdrikh (henceforth

Muhit), Paris, Bibliotheque Nationale, ms. 835, fol. 95b.103. It is called both Madrasa-i 'Ali and Madrasa-i 'Ahiyah in Hafiz-

i Tanish, Sharaf-namah (IOL), fol. 41 Ib; Bahr (IOL), fols.348a, 350a-b, 364b, and Silsilat, fol. 172a, simply refer toit as the 'Abd Allah Khan madrasa.

104. Sharaf-ndmah (IOL), fol. 411b.105. Ibid.106. Bahr (IOL), fol. 364a-b.107. Ibid., fol. 214a-b.108. Surviving madrasa complexes'of this type are found in

Samarqand (at the Rigistan) and Bukhara (the two Khiyabanmadrasas of 'Abd Allah Khan and the Ulugh Beg and 'Abdal-'Aziz II madrasas near the Taq-i Zargaran.

109. For a lengthy description of this "twelve-darband" dawlat-khanah, designed by the architect MirJan Kildi, see Mahmudb. Amir Wali, Bahr al-asrar (IOL), fols. 213a-214a. Theexcavations in the Balkh citadel carried out by A. Foucherand the Dil6gation Arch6ologique Francaise en Afghani-stan in the spring of 1924 revealed in the upper levels a"serai" which may correspond with the dawlatkhanah de-scribed by Mahmud b. Amir Wali (see A. Foucher, La VieilleRoute de 'lndede Bactres a Taxila ([Paris, 1942], 1: 102-10).

110. Muhft, fol. 95b.111. Ibid.112. The yad-i bayiz, literally the "white hand" or the "miracle

hand", a Qur'anic reference to Moses being granted a "whitehand" as a divine sign of his mission (see D. B. MacDonald,"Musa," El, 2d ed., 7: 638.) Ihyi al-mawta, raising the dead,is one of the miraculous signs ofJesus' prophethood (seeG. C. Anawati, "isa," EI, 2d ed., 4: 82). Both phrases areused repeatedly as cliche descriptors for builders and ar-chitects.

113. Muhit, fol. 214b.114. Muhammad Tahir Balkhi, 'Aja'ib al-tabaqdt, Royal Asiatic

Society, ms. no. 79, fol. 18a.115. Ibid.116. B. A. Akhmedov, Istoriia Balkha (XVI-pervaia polovina XVIII

v.) (Tashkent: Fan, 1982), p. 27. However, on page 29,Akhmedov describes the Nazr Muhammad Madrasa as stand-ing opposite the 'Abd al-Allah Khan Madrasa (citing theBahr al-asrar).

117. Sultan Muhammad, Majma' al-ghard'ib, fol. 16b.118. Ibid., fol. 374a119. Foucher, La vieille route, pl. viii b, "Centre de la Ville."120. Bahr (IOL), fols. 278b, 349a-350a, 365b.121. Ibid., 90a, 101b, 169a-b, 179a, 180b, 185b, 187a, 278b.122. Silsilat, fol. 269b. The anonymous author of another early-

eighteenth-century chronicle (Tarkh-i Shibani Khan wamu' malat ba awldd-i Amir Timur), Tashkent, Institute ofOriental Studies, no. 4468/II, fol. 115b, likewise gives thedate 1070 (1660) for the building.

123. Silsilat, fol. 269b. The three named are Khwaja 'Abd al-Ghaffar (earlier referred to as hazrat-i shan and the ptr ofthe "rizwan makanf," here perhaps a reference to the lateNazr Muhammad Khan (d. 29Jumada II 1061/20 May 1651),although the epithet seems to have been more commonlyapplied (later) to his eldest son, 'Abd al-'Aziz (d. 14 Sha'ban

�___� ._ 1�1��_��_��_

118

CONSTRUCTING THE KHWAJA ABU NASR PARSA SHRINE

1094/9 August 1683). The second and third dignitaries wereMir Mahmud 'Ulyabadi and Ata Khan 'Azizan.

124. The endowment deed was published by A. D. Davydov,"Imeniia medrese Subkhankuli-khana v Balkhe (po vakfnoigramote XVII v.)," Kratkie soobshcheniia InstitutaVostokovedeniia 37 (1960): 82-128. I discuss the deed in myWaqf in Central Asia (Princeton, N.J., 1991), pp. 131-32.

125. A. D. Davydov, "Imeniia medrese." All four boundaries aredescribed in lines 54-56 and 110-113 of the two copies ofthe document; line 112 of the Russian translation.

126. For a recent discussion of the Mughal occupation, see Ri-chard Foltz, "The Mughal Occupation of Balkh," Journal

of Islamic Studies 7 (1996): 49-61.127. Bahr (IOL), fol. 207b.128. Ibid., fol. 226b.129. Ibid., fol. 269b130. Silsilat, fol. 253a.131. Ibid., fols. 299b-300a.132. See my Waqf in Central Asia, pp. 158-59, for more details

of this episode. The accounts vary considerably in termsof detail and emphasis.

133. SeeJonathan Lee, The "Ancient Supremacy " Bukhara, Afghani-stan, and the Battle for Balkh, 1731-1901 (Leiden: E.J. Brill,1996), Appendix IV, vii.

119

·1____1____�_�___1__l_________�_l_ _ _