Convention and Reinvention: The British Library Shahnama of 1438 (Or. 1403)

ABSTRACT The geographic origin of the fifteenth-century illustrated Shahnama manuscript Or. 1403, held at the British Library, has been the subject of unresolved scholarly debate. Stepping away from the binary alternatives that have been suggested for the attribution of the manuscript in the past (Iran versus India, and Delhi versus the Deccan), this essay focuses on text and image and their potential relationships in its preface-frontispiece set and how these would have addressed the manuscript’s possible audiences. Evaluating the ways that the preface and frontispiece reimagine the established visual and textual conventions of Shahnama manuscripts in the fifteenth century, the study explores the manuscript’s engagement with a range of possible socio-historical settings, all of which reflect the complex circulation and reception of Persianate modes of culture between Iran and India in this period.

The Shahnama (Book of Kings), an epic poem composed by the poet Firdausi (d. ca. 1020) in Persian in the early eleventh century, is the most-popularly copied and circulated epic poem in Persianate societies. In it, ancient history and myth mix and converge around themes of kingship and ethics. Verses of the Shahnama and visual imaginations of many of its episodes appeared on medieval ceramics and metalwork and covered the walls of palaces in Islamicate societies from the twelfth and thirteenth centuries. From around 1300, illustrated manuscripts of the Shahnama were key items in royal and elite households. Through the processes of copying, the Shahnama was repeatedly reinterpreted and recreated as a contemporary object, laden with contemporary fashions, ideals of beauty, and socio-political circumstances. It served as a popular vehicle for the demonstration of imperial patronage and ideology that legitimised the present through identification with the past, as well as a repository and initiator of ideas about visual art and culture. 1 This article studies a fifteenth-century illustrated copy of the Shahnama kept at the British Library as Or.1403. The manuscript has a colophon recording the date of completion and the name of the scribe, but there is no attribution of the patron, nor is there any indication of where the manuscript was produced. It has been a matter of unresolved scholarly debate dominated by art historical, and predominantly formalist approaches to the question of geographical origin, having been assigned by scholars to Iran, north India, or the Deccan.
There has been a growing consensus among scholars to withdraw from attributions to a Persian court for this manuscript. This tendency is due to several "peculiarities" in the manuscript's illustrations as compared with contemporary depictions of the same scenes. Stylistic features, laboratory tests on the pigments, the existence of a glossary of Persian terms from the Shahnama, and a reference in the preface to Firdausi having stayed in India all made for a strong argument in favour of India as the place of production. 2 Within India, while Delhi has been recognised as a candidate, more recent attention to the stylistic details of the figures, animals, and architectural features, as well as broader interrelations between this manuscript and others attributed to India, have led to the suggestion of the Deccan as a place of production. 3 Instead of approaching the distinctive features of this manuscriptin both text and imageas peculiarities vis-à-vis Persian traditions, this essay aims to take up a perspective that places this manuscript at the centre of a wider field of conventions, inquiring about the transregional and/or collaborative networks of knowledge as well as the interventions and innovations that shaped them. The aim is to offer an interpretation of how this manuscript could have been read and understood as an object of "heritage" and self-imagination, weaving both a real and an imaginary "historical" past into the present, and projecting a courtly ideal of the ties between religion and state. 4 After a brief overview of the manuscript, addressing codicology, palaeography and past scholarship, the discussion will focus, in particular, on the introductory material added to the epic proper: the preface and frontispiece. It will first examine how the preface reimagines the history of the epic's production, which recounts that after the completion of the book, the author, Firdausi, journeyed to India and took up temporary residence there. It will then locate the frontispiece amidst the broader conventions for frontispieces of the Shahnama and propose a reading that does not presuppose issues of patronage and exact place of production. Throughout these two sections, my aim will be to avoid approaching the question of the manuscript's precise origin in terms of binary, mutually exclusive attributions: Persian versus Indian, and north Indian versus Deccan. Our understanding of the circumstances around manuscript production in India, and the transregional collaborations between Indian and Iranian courts prior to the sixteenth century is limited. There is, however, textual evidence that points to manuscripts produced in Iran as gifts designed for Indian courts, or perhaps even commissioned by them. Such exchanges, for instance, between the Bahmani court of the Deccan and the Timurid court in Shiraz, are evident from dedication notes and the inclusion of ex libris from the Deccani court library in manuscripts produced in Shiraz in the first half of the fifteenth century. 5 While I accept India either as the place of production or the intended audience of the manuscript, questions of precise origin and patronage are secondary to an understanding of the manuscript's visual and textual programme as well as the wider implications these could have for the reception of the Shahnama in Persianate societies. 6

Overview of the Manuscript
Or. 1403 measures about 27 by 16 centimetres, its written surfaces measuring about 19.6 by 12.7 centimetres. It contains 513 folios, 4 illuminations, and 94 paintings, none of which except for the double-page frontispiece are full page illustrations (Figure 1). Each of the illustrations occupies approximately a third of the page or less. The written surfaces are comprised of four columns in the text proper and contain 27 lines on text-only pages. Gold ruling can be seen on most folios, indicating the high status of the manuscript.
The manuscript includes a preface and glossary before the beginning of the epic itself. Its colophon (fol. 513v) records the date, 11 Ramadan 841/8 March 1438, and the scribe's name, which although not entirely legible, can be read as "Muhammad", followed by a word that starts with "shir" and ends with "ya", which is probably "Shirazi," the scribe's nisba (Figure 2). The script is a small nasta'liq written in black. The whole manuscript, including the preface, seems to be in the same hand, except for folio 291, which has been inserted, possibly to replace a damaged or missing folio, and features a much more mature and later nasta'liq ( Figure 3). The title blocks are written larger, in thuluth or naskhi script, in gold or red, and could be by a different and less skilled hand than the main text ( Figure 4). 7 The 92 illustrations in the text proper depict heroic feats, wars, enthronement scenes and a variety of other dramas involving love, lust, death, trickery and wondrous experiences. Out of these, the throne scenes are the least frequent (approximately 15). The illustrations of the Sassanian section of the Shahnama comprise about one-fifth of the total in the poetic text (18 out of 92), offering a case of relative consistency in the distribution of images across the legendary and "historical" sections of the text. 8 Several of the 16 illustrations in the historical section were not commonly depicted before their appearance in Or.1403.
The paintings have been noted for their quiet compositions, with grey or pale mauve backgrounds dotted about with thorn bushes and plants of the mountain and desert, as well as high horizons with rocky borders and fungus-type rocky hills. The picture space is shallow and figureswhich are limited in number in each paintingare short and stocky with relatively large heads, although the proportions gradually become more balanced as the manuscript progresses, especially in exterior scenes. 9 With a few exceptions, a frontal view of faces is maintained even when bodies are depicted in profile. The colour palette is restricted and marked by strong contrasts, especially in the textiles ( Figure 5). The paintings have been described as "old-fashioned"; the figures as "simple and naïve." 10 It is true that the illustrations can be minimal and formulaic. Several look like stock images, displaying only minor variations, as if they could have been used interchangeably. 11 This is especially evident in enthronement scenes, in which case the formula could have been used to communicate an idea of continuity. But scenes depicting war and heroic feats show more complex emotionswhether in the service of the drama or notand sometimes unique attention to detail and iconography, suggesting the artists' deep knowledge of the text or collaborative innovations in the workshop. Several illustrations in the manuscript offer either the earliest, or one of the earliest known examples of the scenes they depict. 12 The total number of illustrations is also noteworthy. In general, it is believed that during the Timurid period, in Shahnama manuscripts produced between the death of Ibrahim-Sultan in 1435 and the middle of the century, paintings become more simplified, narrative and repetitive, betraying a rapid production of "non-princely" and "commercial" manuscripts with a very large number of pictures. 13 These stylistic and quantitative characteristics have been identified as a fitting context for Or. 1403 in previous scholarship. 14 The manuscript has been rebound at least once before it acquired its current binding, as indicated in a note on   13 Sims, "The Illustrated Manuscripts of Firdausī's 'Shāhnāma," 54. The quality-quantity argument has also been discussed in the study of Shahrukh-era manuscripts by Golombek, "Discourses of an Imaginary Arts Council," 8-9. 14 See Sims' Appendix B on non-princely manuscript: Sims, "The Illustrated Manuscripts of Firdausī's 'Shāhnāma," 67-8.
its flyleaf (Figure 6). 15 The written or painted frames of several of the pages from the preface are re-pasted on new margins. Several illustrations look repainted, especially the skies and faces. Worms have damaged the paper on numerous folios, and stains of pigments (especially green and yellow) can be seen throughout the manuscript. 16 The provenance of the manuscript is unknown before it was acquired at some point in the nineteenth century by Jules Mohl (d. 1876) the editor and translator of the Shahnama into French. He does not mention where he acquired it. Mohl, whose annotations can be seen in the margin of the manuscript's preface, used several manuscripts from his own collection for the translation, and refers to this one as MS. no.5. 17 The flyleaf features two notes in nasta'liq script, one on the purchase of the book from a binder, and one belonging to a reader who finished reading the manu-script in 1680 (Figure 7). Under each note there is a seal: the seal underneath the purchase note belonged to a Sharaf al-Din Muhammad Ashraf, and the one under the reader's note is obliterated beyond legibility.
The almond-shaped dedication device (fol. 9r) is placed right after the preface-glossary section and before the frontispiece (Figure 8). There is no writing on its golden surface. Decorations on the margin of the oval are in blue, red, and gold. The page has been re-margined, but traces of writingpartly cut out during the re-marginingcan be seen on the upper left side and bottom of the folio. The note on the upper left and the seal underneath are crossed out, but both record the name of a scribe, 'Ibad-Allah. The beginning of words that could possibly be majlis (gathering) and abyat (couplets) can be read at the bottom of the page.   Shahnama manuscript, British Library, MS. Or. 1403, fol. 291r. ©The British Library. 15 Or.1403, f. 2r. 16 Scott, "Laboratory Notes," 94-5. 17 Titley, Persian Miniature Painting, 167; Mohl, Le livre des rois, vol. I, p. xvi, in connection with a brief note on the prefaces of the Shahnama (see also below).

Historiography of the Manuscript
Past discussions of Or.1403 have addressed both visual and textual aspects of the manuscript, from stylistic peculiarities to the unusual subject matter of the frontispiece, the intervention in the preface, and the existence of a glossary. Based on these factors the manuscript has shifted back and forth across the obscure, binary boundaries of Persian and Sultanate, and royal and non-royal. In a gradual process, scholars have come to identify the manuscript with India, suggesting Delhi and Gulbarga as possibilities for its exact place of production. 18 Neither of the latter suggestions can be proved or disproved.
Or. 1403 displays some of the most important and unresolved questions about the mobility of objects and people, and the efficiency of style and schools associated with certain workshops as tools of identifying thein many casesdislocated and unprovenanced objects of medieval and early modern Islamicate societies found in museum collections. When considering pre-Mughal manuscripts, several factors contribute to the problem. The scarcity of manuscripts tentatively attributed to Indian sultanates means that only a few examples stand for comparison. The fifteenth-century manuscripts placed in this group so far rarely have a definitive record of place of production, at least not until the period 1490-1510, and no sultanate manuscript is believed to have survived from before the fifteenth century. 19   19 Robinson, Fifteenth-Century Persian Painting, 62. For records of patronage and place of production in Malwa manuscripts from the end of fifteenth century, see Titley,Persian Miniature Painting,170,176. To tackle the problem of pre-fifteenth century Sultanate manuscripts, earlier Jain manuscripts have also been studied in terms of the emergence of narrative painting under Persian influence: Titley, Persian Miniature Painting, 161.
Methodologically, the field has resolved these issues in a way that poses problems of its own. The relationship between these possibly-Sultanate manuscripts with the dominating works from Persian schools of painting and workshops has led Sultanate manuscripts, if at all distinguishable as such, to be viewed as offshoots of Persian painting genealogy: "outdated", "provincial", "crude" and "inferior." 20 Through this "Persian eye" that looks either for the "influence" of Persian styles on Sultanate manuscripts or "peculiarities" judged against the Persian model, the general assumption is that manuscripts made in India should be attributed either to itinerant artists or to local artists trained by them or imitating them, thus reinforcing a hierarchy that takes the Persian modelwith its own internal hierarchiesas normative. 21 Past studies of Or.1403 have been marked by such approaches.
Given Shiraz's trading and diplomatic links with India, its status as a commercial centre of manuscript production, and comparative formal analyses of Shirazi and Sultanate manuscripts, Shiraz is often considered the main influence on paintings of the Sultanate period of India in the fifteenth century. 22 Many of the visual characteristics discussed in the previous section have been connected to a conflation of Shiraz and the Muzaffarid schools of the fourteenth and fifteenth centuries: a "Muzaffarid revival" whose influence in India has been detected by several studies in the past. 23 The Muzaffarid style is known for its high horizons and rounded hills, oval egg-shaped human heads and   20 Robinson, "Areas of Controversy," 34-5. Titley, Persian Miniature Painting, 169, sees this as an opportunity to study the "originality, in execution and choice of subject, which has disappeared in the more highly-finished court styles." See also: Fraad and Ettinghausen, "Sultanate Painting," 48. For a discussion of "high" and "provincial" schools of painting, see Kadoi and Szántó, "Introduction: Why Persian Art Needs to be Studied and Collected," 13-4. For a critique of Indian painting as an offshoot of Persian painting, see Skelton, "The Ni'matnama," 44. long necks, pupils rolled into the corner of the eyes, crooked and high turbans, long moustaches, and tufts of hair below the lower lip that join up with a beard from ear to ear. 24 These features are believed to have made a resurgence in Shirazi manuscripts during the reign of Ibrahim-Sultan (r. 1415-1435), installed as the governor of Shiraz by his father Shahrukh (r. 1405-1447), Timur's son and successor in Herat, after having deposed his nephew and opponent, Iskandar-Sultan (d.1415). Shahrukh took artists and scribes back to Herat and this episode is generally connected with the simplification of paintings and a loss of "delicacy and elegance" from the time of Iskandar to Ibrahim. 25 Paintings made for Ibrahim-Sultan have been associated with shallow picture space, limited number of figures, restricted palette with strong contrasts of colours, rounded hills or fungus-type rocks, as well as other earlier Muzaffarid features such as high horizons and oval faces with squint eyes. 26 While the presence of all these qualities in Or.1403 could point to Shiraz, other features have been marshalled to argue for its production in India. The first cue was a note by Rieu in his Catalogue of the Persian Manuscripts in the British Museum, pointing out that the preface of Or. 1403 includes a reference to Firdausi having taken refuge in India after the completion of his book. 27 The 1932 scientific report at the British Museum the manuscript's home before moving to the British Libraryrevealed the use of the Indian Yellow, or peori, in the manuscript, thus strengthening the connection with India. 28 In 1967, Robinson stated that the paintings from this manuscript "may turn out to be Indian." 29 The question was explored in more detail by Fraad and Ettinghausen in a 1971 article, which looked for characteristics that differentiate their suggested group of Sultanate manuscriptsincluding Or. 1403from contemporary Timurid counterparts. 30 For Fraad and Ettinghausen, the visual features that place Or.1403 in the Sultanate group proceed entirely from the frontispiece, including the serried ranking of figures, the movement of the dancers, their loosely hanging hair, the band across the ruler's chest, and the white bars of clouds ( Figure 1). 31 To them, Rieu's point about Firdausi's journey to India strongly supports an Indian provenance for this manuscript and all others that could be connected to it on the basis of style. 32 Losty considers the manuscript a Sultanate work done in a "basic Iranian style." 33 Basing his arguments on stylistic observations, Robinson categorises the manuscript in a group of Indian manuscripts made in the Timurid period. To him the unifying feature of this group is the same "cloud convention consisting of horizontal white streaks" noted by Fraad and Ettinghausen, which he Ibid., 48. 26 Wright, The Look of the Book, 175. 27 Rieu, Catalogue of the Persian Manuscripts, 534-5. 28 Scott, "Laboratory Notes," 95. 29 Robinson, Persian Miniature Painting, 88. 30 Fraad and Ettinghausen, "Sultanate Painting," 51. 31 Ibid., 55. 32 Ibid., 64-5. 33 Losty, The Art of the Book in India, 58-9, categorises Sultanate manuscripts in three groups, Or. 1403 being in the first one together with the British Library Javami' al-hikayat (Or. 11676).
considers "quite unknown in Persian painting, but with close parallels in later Indian works ( Figure 1)." 34 Titley also notes the clouds in the frontispiece but connects them to Muzaffarid manuscripts of the fourteenth century that she considers models for Or. 1403 and related manuscripts. 35 As additional evidence for the manuscript's Indian origin she lists the poor brittle paper, the worm damage, the colour palette (pale blue, yellow, deep crimson, pink, and a pale green), the two faces drawn in profile as opposed to the usual frontal view (Rustam in battle with Chingish: fol. 172r, and Isfandiyar killing the witch: fol. 277v), and finally the depiction of plants that resemble the lotus (fol. 368v). 36 Like Fraad and Etttinghausen, she considers the addition to the preface about Firdausi's refuge in Delhi as a unique feature that points to the manuscript's Indian provenance, and more specifically, to patronage by the Sultanate of Delhi. 37 Similarly, Brend finds the mention of Delhi in the preface of the Shahnama a "good prima facie" case for the manuscript's Sultanate origin, strengthened by the existence of the glossary at the end of the preface. She also refers to the quality of paper and the scribe's hand as reasons why the manuscript's origin could not be in Iran. 38 In addition, she singles out four features of Or.1403's paintings that she considers to be Indian: first, the depiction of elephants' movements, and secondly, an unprecedented depiction of the scene "Bazur casting a spell to defeat the Iranian army", which includes a snake that is absent from the text of the Shahnama. 39 Her third visual observation regards an unidentified painting at the end of the preface showing two men seated with a bookrest between them (fol. 8v), which she interprets as showing either the scribe at work or a teaching session, a subject she considers to be more commonly depicted in Indian paintings ( Figure 9).
Within this scene, she also remarks on the profile of the turban of the man on the left as an Indian feature. Finally, she points out several details in the frontispiece: the composition of the double-page and its emphasis on the left-hand painting; the serried ranking, barefooted figures and dancing dervisheswhich she identifies as such, moving away from the previously-held opinion that they represent dancing girlsall of which she considers to be Sultanate visual features (Figure 1). 40 Brend's argument moves the manuscript in a different direction from Titley's: to Deccan India rather than Delhi. Based on visual connections with another manuscript, the two-volume Anthology (P.124) at the Chester Beatty Library, which is dated between 1435 and 1436, Brend argues that the Or.1403's place of production must be the Deccan. The Chester Beatty manuscript has a note that shows the manuscript was in India, purchased in 920/1514 by an 'Adil Shah, who has been identified as Isma'il 'Adil Shah (r. 916-941/1511-1534), the ruler of Bijapur. Fraad and Ettinghausen had argued for a Sultanate origin for the manuscript since, in their opinion, two of the paintings in the manuscript show the artists' unfamiliarity with the correct depictions of places and costumes. 41 Brend's key argument for the relationship between the two manuscriptsthe British Library Shahnama, and the Chester Beatty Anthologyand their place of origin is based on a decorative motif of a tall stylised vase flower shared with monuments in the Deccan. A pair of such motifs appears in the painting of "Iskandar at the Ka'ba" in the Anthology (fol. 223v). The vases were recorded in the palace fort of Ahmad Shah Bahmani (r. 1422-1436) and are also shown in Or. 1403's frontispiece at the base of the walls on each side of the left-hand page ( Figure 10). 42 Brend concludes that both manuscripts were made under the Bahmanis, the first independent Muslim dynasty to rule the region to the north of the Krishna river, in the Deccan.
Although the Deccan remains a candidate for Or. 1403's place of production, the visual connections drawn in the above arguments can be both multiplied and challenged. A further example for the vase motif in Bidar, concurring with Brend's argument for the Deccan, is the tomb of the Bahmani ruler, Ahmad Shah I, where the motif is repeated eight times at the dome's zone of transition ( Figure 11). Such visual idioms, however, were not bound to certain geographies, but also circulated in regions that had nothing to do with this 34 Robinson, Fifteenth-Century Persian Painting, 64-8 and 75, places Or. 1403 in his "group four" (out of five); the other manuscripts similar in this feature are: Uppsala Nizami of 1439 (Tornberg 151), Topkapi Nizami of 1441(H.744 or K.401), and two manuscripts in Staatsbibliothek of Berlin, dated to 1416 (ms. Or.oct.2302) and 1456 (ms. Or.oct.268). 35 Titley, Persian Miniature Painting, 164-5. Most recently, Brac de la Perriere, "The Art of the Book in India," 308, mentions Or. 1403 in passing and attributes it to Muzaffarid paintings from northern Iran. 36 On frontal versus side views of the face, see Skelton, "The Ni'matnama," 47. 37 Titley, Persian Miniature Painting, 165-7. 38 Brend, "The British Library's Shahnama," 87. 39 British Library Or.1403, f. 161v: http://shahnama.caret.cam.ac.uk/new/jnama/card/ceillustration:1219690119. 40 Brend, "The British Library's Shahnama,[88][89][90] Fraad and Ettinghausen, "Sultanate Painting," 55; Wilkinson, The Chester Beatty Library, 53; Brend, Perspectives, 74. 42 Brend, "The British Library's Shahnama," 91, and Brend, "Perspectives," 78-9, suggest Gulbarga, as opposed to Bidarthe capital of the Bahmanis at the time of the completion of the manuscriptas the exact place of production. Cf. Yazdani, Bidar, pls. XXVI, XXVII, XIX. particular manuscript. The vase motif, for instance, can also be seen in Ottoman monuments of the fifteenth century, albeit at a later date. The painted decorations of the mausoleum, known as that of Cem, built in 1479 for the son of Mehmed II (r. 1444-1446), Mustafa, shows a very similar motif flanking the windows above dadoes on three walls. 43 Similarly, the depiction of the left-side figure's turban in folio 8v (Figure 9), which Brend considers Indian, bears significant resemblance to the headgear of several figures in the paintings of the surviving manuscripts of Hafiz-i Abru's Majma' al-tawarikh, produced under the Timurid ruler, Shahrukh. 44 The point is that these visual motifs can hardly be pinned down to one workshop or location, and while a preponderance of such features can point us to a certain place and time, the identification must ultimately remain speculative. Nevertheless, the work of all these scholars has advanced the field to a point today where it is possible to look at Or.1403 alongside similar objects and ask a different set of questions.
To show how deducing provenance from visual details can sometimes be misleading, Titley offers the example of the British Library Shahnama (Or. 12688), dated 1446, whose detail of an Indian mahout with an ankus on an elephant in two of the paintings could have easily supported an Indian provenance, if not for a note in the manuscript that records the place of production as Mazandaran. 45 To take her discussion in a slightly different direction, limits of textual evidence can be observed as well. While textual evidence can offer definitive answers about a manuscript's place of production, it tells us very little about the making process of the manuscript and its life after being fully or partly completedespecially since unfinished or complete manuscripts could have been completed, altered, and remade as they travelled. More important, and more relevant to the subject of this essay, is the question of how textual evidence can colour the way we look at images, closing off further investigation into objects and the cross-disciplinary and transregional connections that are embedded in them.

Prefacing the Shahnama
The continuous tradition of prefacing the Shahnama has resulted in the currently accepted classification of four different categories of prefaces to the epic, as established by Riyahi and Qazvini. 46 Although this number could certainly change with further research on the prefaces, my brief summary here relies on their work since subsequent scholarship on the Shahnama prefaces has largely been concerned with individual case studies. A broader picture of the variations and circulation of the prefaces must await a more comprehensive survey and statistical analysis of the prefaces.
Among the four groups, the first preface starts with a section known as the Abu Mansuri preface, which predates Firdausi's Shahnama and was probably written for a prose Shahnama collected by Abu Mansur al-Mu'ammari and commissioned by the governor of Tus, Abu Mansur b. 'Abd al-Razzaq. 47 This section is dated to 346/957 in the text, and maps out the world, the creation story, and the genealogy of two men mentioned abovethe compiler and the patron of one of the prose Shahnamaswhile comparing the status of this project in its time with the translations of the Kalila va Dimna. 48 After quoting the Abu Mansuri section, the first (older) preface goes on to deal with the stories of Firdausi and Sultan Mahmud of Ghazna (r. 998-1030). 49 The second preface, which was probably written in the later twelfth or early thirteenth century is known to us from the twelfth-century Florence manuscript of the Shahnama and another sixteenth-century copy at the Topkapi in Istanbul (H.1510). 50 This does not include the Abu Mansuri introduction, and deals with the story of the writing of the Shahnama, Sultan  46 Riyahi, Sarchishma-ha-yi Firdausi-shinasi, 264-5. 47 Qazvini, "Muqaddama-yi qadim," 178, 182-3; Minorsky, "The Older Preface," 162-3. 48 Qazvini, "Muqaddama-yi qadim," 186, 190-202. 49 Ibid., 178. On the Chahar Maqala of Nizami 'Aruzi Samarqandi as a source for the life of Firdausi, see Askari, The Medieval Reception, 6-18; http://www. iranicaonline.org/articles/cahar-maqala, accessed September 2018. 50 Riyahi, Sarchishma-ha-yi Firdausi-shinasi, 264-5.
Mahmud and Firdausi, and the aftermath of their clash. 51 The third preface, known as the ausat (middle) preface integrates sections of the Abu Mansuri introduction (also a part of the first preface) and second preface with other narratives about Firdausi's life and the making of the Shahnama that were circulating at the time. It is datable approximately to the fourteenth century and has not been identified in many manuscripts so far. 52 Riyahi identifies three different beginnings in this preface: after the story of creation come the tables of the four pre-Islamic dynasties. Then there is another beginning, leading to the story of Firdausi and Sultan Mahmud. Third comes the Abu Mansuri preface, the list of dynasties, and finally a glossary. 53 The fourth preface, known as the Baysunghuri preface, starts with the praise of the Timurid prince Baysunghur b. Shahrukh (d. 837/1433) and his patronage of the Shahnama manuscript. A record in the text dates the preface to 829/1425. After this opening section, the preface continues to chart out the collection of the Shahnama's sources, the life of Firdausi, Sultan Mahmud's change of attitude towards Firdausi, the completion of the Shahnama, and the writing of the satirical verses of grievance. 54 Minorsky and Qazvini point out that most postfifteenth century copies of the Shahnama opt for the Baysunghuri preface. 55 During the fifteenth century, and after the completion of the Baysunghuri Shahnama, several manuscripts still used the first preface. Some contemporary examples include the 840/1437 Shahnama from Shiraz, now in Leiden University Library (Cod. Or. 494); the 850/1446 Shahnama at the British Library (Or. 12688); and the Ibrahim-Sultan Shahnama at the Bodleian Library (Ouseley Add. 176). 56 This suggests that the older preface remained in circulation and perhaps even remained better known than the Baysunghuri version for some time. 57 Or. 1403 records the existence of an earlier copy dating to 779/1377 in a section before the colophon. 58 The preface of our Shahnama features many elements of the third category, but the text is more organised and it does not include several opening sentences like the third preface. 59 Therefore, it should be safely categorised under the first group of prefaces.
The preface of Or. 1403 opens with the Abu Mansuri introduction, starting on folio 2v and running on until the end of folio 3v, where right before a brief mention of the four pre-Islamic dynastieswhich would have then led to the genealogies of the two men responsible for the Abu Mansuri prefacethe text is interrupted due to missing folios. 60 The next section in Or.1403's preface must have been the two versions of the story of Firdausi and Sultan Mahmud, because on the next surviving page, folio 4r, it picks up with the end of the first version of Firdausi's story, when he writes the verses of Hajvnama, and hands them to Sultan Mahmud's servant, Ayaz, before leaving for Tus. The second version of Firdausi's story, followed by the tables of the pre-Islamic dynasties, continues to folio 6v, after which there is a missing folio. This should have included the table of Kayanian dynasty, as indicated by the catchword "tabaqa-yi duvvum" (second dynasty) on folio 5v. 61 Following the tables, the glossary continues to folio 8v, ending with the aforementioned painting of the two men seated with the bookrest (8v), the empty dedication device (9r), and the doublepage painting (9v-10r), before the epic proper begins (10v) (Figure 9, Figure 8, and Figure 1).

Mapping Firdausi in the Preface
Within the outline discussed above, the second version of the story of Firdausi and Sultan Mahmud in the preface of Or.1403 (fol. 4r) bears an intervention that was briefly discussed earlier. Ibid., 364-418. On the Baysunghuri preface see also: Davidson, "Interweaving of Book and Performance," 1-11; Davidson, "Why is Baysonghori Recension a Recension?," 127-30. 55 Minorsky, "The Older Preface," 161-2; Qazvini, "Muqaddama-yi qadim," 178. 56 Sims, "Towards a study of Shirazi Illustrated Manuscripts of the 'interim period'" 615, 620. As Abdullaeva and Melville, The Persian Book of Kings, 28-30, 125, showed, the Ibrahim-Sultan Shahnama includes the old preface (as opposed to the previously-thought Baysunghuri preface). It also has elements of the "third" preface, clearly added at a later date. The Ibrahim Sultan Shahnama is fully digitised and accessible on Digital Bodleian. 57 The Older Preface, 161-2; Qazvini, "Muqaddama-yi qadim," 178. As Charles Melville points out (in his comments on this article), there are no serious studies of whether later MSS that use the Baysunghuri Preface also follow the Baysunghur "text" or edition. Paralleling the discussion of the visual role of Baysunghur's and Ibrahim-Sultan's Shahnamas in shaping later copies, it would be interesting to have a statistic analysis of the persistence of the Abu Mansuri preface as against that of Baysunghur in later periods. 58 Rieu, Catalogue of the Persian Manuscripts, 534; Titley,Persian Miniature Painting, There has been some confusion about what constitutes the third preface, and how one should distinguish between the first and third prefaces, as they both include the Abu Manusuri preface, as well as other elements. See Riyahi, Sarchishma-ha-yi Firdausi-shinasi, 189, 326. 60 Rieu, Catalogue of the Persian Manuscripts, 534, notes that there are two folios missing here. The missing folios will be discussed in my forthcoming study on Or. 1403, see footnote 4. 61 Conjectures on the missing material can be made through comparison with the Ibrahim-Sultan Shahnama (Ouseley Add. 176, Bodleian Library).
write the Shahnama in verse, they agreed on a certain amount of gold. Upon finishing the work, Sultan Mahmud was convinced by an envious courtier to reduce the amount and pay him in silver. This gave rise to a clash between the poet and the Sultan, resulting in Firdausi's being barred from the court. After a while, in despair: Firdausi sent his belongings to Khursasan and left with his brother for India. Once he had made his decision to leave, he wrote the few verses [of the Hajvnama] and left them with the Sultan's slave, Ayaz. He then left for Hindustan and reached Delhi. When the king heard his story, he invited Firdausi [to his court] and treated him with respect. Firdausi stayed there for a while. In the meantime, Sultan Mahmud was searching for him in vain. He sent someone to Khurasan to fetch him to Ghazna so that the Sultan could apologise to him and pay him his deserved gold … . The Sultan was feeling regretful for not having treated him well, not having fulfilled the desire of his heart, and was looking for him and not finding him. Having stayed with the Shah of Delhi for a while, Firdausi asked him for permission to leave. The Shah of Delhi made him a robe of honour, and gave him many horses, slaves, women, much gold and raw gems … . Firdausi retuned to Tus and died there. 62 As noted above, scholars in the past have interpreted this intervention, which takes Firdausi to India, as well as the existence of the glossary, as an indication of the manuscript's Indian origin. 63 In order to contextualise Or.1403's intervention in the story of Firdausi historically and geographically, several points about the history and circulation of the different versions of the preface and its additions should be brought into the conversation.
My first point concerns the glossary. Dictionaries and glossaries attempted to fill in for the lack of commentaries on the Shahnama in the Persianate world, especially in places were Persian was not the main or only elite language. 64 But glossaries were not specific to such places. As mentioned in the discussion of preface categories above, glossaries were a standard component in the third preface. 65 Indeed, some content in the Or.1403's glossary would seem to speak against its composition or adaptation in India: for instance, the definition of a word such as shara (sara, sari), which appears in Or.1403, as in other glossaries of the Shahnama, is a "white garment brought from India"a reference to Hindustan that implies a sense of otherness. 66 My point here is not that the appearance of these definitions rules out India as a place of production or commissioning, but that the history of the circulations of the Shahnama is too intricate to allow for direct conclusions to be drawn from any of these elements, or indeed, even certain combinations of them.
My second point is that, in order to understand the Delhi insertion into the preface, it should be contextualised within the existing narrative traditions of Firdausi's travels after his conflict with Sultan Mahmud. The most frequent claim is that Firdausi returned to Tus, but there are other versions in which Firdausi's journeys take him to many other parts of the Persianate world: Isfahan, Quhistan, Tabaristan, Baghdad, as well as Delhi.
Or.1403 ends with an epilogue and two colophons. The first colophon is related to the manuscript that Or.1403 was copied from either directly or indirectly. As mentioned, the date recorded in this colophon is 779/1377. The second colophon is the one belonging to Or.1403 itself, recording the date 841/1438 and the name of the scribe. 67 In the epilogue, which, as Reiu noted, is on the completion of the Shahnama, the poet recounts how he was invited by an Ahmad b. Muhammad Abu Bakrwho governed in Isfahan and Khan Linjanto his residence in the latter. 68 Not only did Ahmad treat him well with all sorts of gifts, supply all his wants, and ignore his slanderers, but his son also saved Firdausi from drowning in an incident in Zarrinrud. 69 In a similar, though briefer narrative in the Baysunghuri preface, Firdausi is invited to Quhistan by its governor after the story of Sultan Mahmud's neglect and his vizier's cruelty had reached every corner of the world. 70 While the governors of Isfahan and Quhistan extend their invitations to Firdausi, he was treated differently 62 British Library Or.1403, f. 5r. 63 Rieu, Catalogue of the Persian Manuscripts, 534-5; Titley, Persian Miniature Painting, 165-6; Brend, "The British Library's Shahnama," 87. 64 One example is Farhang-i Qavvas, which is understood to serve as a dictionary for the Shahnama and other important works of Persian literature and was written in 716/1315 in India. See Karomat, "Turki and Hindavi in the World of Persian," 136-8, 141-2. 65 Riyahi, Sarchishma-ha-yi Firdausi-shinasi, 327. 66 Or. 1403, f. 8v. Also included in the added glossary to Ibrahim-Sultan's Shahnama: Bodleian Library Ouseley Add. 176, f. 11v. The definition of the word "shara" in the Shahnama glossary is especially interesting when compared to the one that appears in Farhang-i Qavvas (see footnote 64): "colourful cloth set around a candle so that it is not blown by the wind." This is the version of the definition that appears in the oldest manuscript of the Farhang-i Qavvas. Was it that the author found it unnecessary, perhaps too obvious, to mention the more common definition of the word, that was presumably well-known in India? Later dictionary writers did include the more common definition: "colourful garment specific to inhabitants of India." See Qavvas Ghaznavi, Farhang-i Qavvas, 153-4. 67 British Library, Or.1403, f. 513v. 68 The text says that Firdausi was invited by a governor but does not specify that this individual was the governor of Isfahan. This is rather implied by the invitation to his residence in Khan Linjan, near Isfahan. British Library, Or.1403, f. 513r. See also: Rieu, Catalogue of the Persian Manuscripts, 534-5. 69 Or.1403, f. 513r-513v; Rieu, Catalogue of the Persian Manuscripts, 534-5. 70 Riyahi, Sarchishma-ha-yi Firdausi-shinasi, 407-9.
in Mazandaran. In both the second and Baysunghuri prefaces, Firdausi decides to take refuge with the Dailami ruler of Mazandaran, one of the sons of Shams al-Ma'ali Qabus b. Vushmgir, whoas is specified in the prefacewas married to Sultan Mahmud's daughter. Unlike the above-mentioned governors of Isfahan and Quhistan, as well as the Shah of Delhi in Or.1403's preface, the governor of Mazandaran was not hailed as a hero in this story. Having arrived in Mazandaran, Firdausi wrote 500 baits in praise of the ruler, added them to the Shahnama, and sent the book to the governor. Being pleased by the book and the verses in his praise, the governor paid the poet 60,000 dinars of red gold and gave him a fine robe of honour, but being fearful of Sultan Mahmud, he asked the poet to leave Mazandaran. 71 Also noteworthy is the difference in the value and status of the gifts offered to Firdausi by the governor of Mazandaran and the Shah of Delhi (who offered slaves, robes of honour, horses, gold, and gems): this is another indication that the governor of Mazandaran, although depicted more positively, was viewed as closer to Sultan Mahmud on the spectrum of good and bad patrons.
An interesting alternative to Firdausi's treatment in both Delhi and Mazandaran is Firdausi's journey to Baghdad after his departure from Tabaristan, as recorded in the second and Baysunghuri prefaces of the Shahnama. The 'Abbasid Caliph al-Qadir (r. 991-1031), who, according to the author of the Baysunghuri preface, had made Baghdad a centre of art production, received Firdausi warmly at his court. 72 In the Baysunghuri preface, Firdausi's stay at the 'Abbasid court resulted in the addition of 1000 baits in praise of the Shahnama, as well as the story of Yusuf from the Qur'an. In return, the caliph gave him 60,000 dinars and a robe of honour and allowed him into his inner circle of courtiers. 73 But he went even further in supporting Firdausi. The second preface highlights the resentment that existed between Sultan Mahmud and the Caliph: the Sultan hoped for more status and grander titles from the Caliph, who would not grant them. Sultan Mahmud had previously sent these requests with a messenger to Baghdad, who stayed there for a long time but was eventually sent back to Ghazna empty-handed. 74 Against this background, when Sultan Mahmud learned about Firdausi's stay in Baghdad, he sent the caliph a threatening letter that was met with a firm response from the Caliph.
Mahmud's threats thus led nowhere. Firdausi then remained in Baghdad until the death of Sultan Mahmud and was treated with respect. 75 The account of Firdausi's journey to Baghdad reflects a specific historical context: the competing nodes of power and cultural production in the Islamic world at the time of the Shahnama's completion. This is attested by the pointed comments about the power dynamics between Ghazna and Baghdad, and the centrality of Baghdad in art production. The Shahnama and its poet were taken for granted in Ghazna, but were received with honour and respect in Baghdad. Although not directly stated, similar dynamics are implied in the story of Firdausi's travel to Delhi, for which a relevant political context existed as well.
This probably relates to Ghaznavid attacks on India at the time. Between 391/1001 and 417/1026 Sultan Mahmud led several campaigns into India, and the Shahnama was completed around the middle of this period. This historical background would have resonated with the makers and audience of our Shahnama in 1438, falling just 40 years after Timur's campaign in India and the conquest of Delhi in 1398-1399. 76 Several motives might be at play. First, the author of this insertion into Or.1403's preface could be using the familiar story of the completion of the Shahnama, its dismissive reception by Sultan Mahmud, and Firdausi's travels to make a point. Whereas the antagonist of the story, Sultan Mahmud, treated the poet unjustly, did not recognise the value of his work, and in addition, repeatedly assaulted India, the idealised ruler of Delhi, like those of Baghdad and Isfahan, becomes the protagonist of the story by holding the poet and his work in high esteem. Second, Firdausi's sojourn in Delhi not only creates a special sense of belonging for the Shahnama in India, on a broader level, it also resonates with India's historical portrayal as a refuge for artists, poets, and scholars from different parts of the Persianate world over the centuries, especially after the Mongol invasions in the early thirteenth century. 77 My third point proceeds from the fact that the insertion of Firdausi's stay in Delhi is not unique to Or.1403. Ibrahim-Sultan's Shahnama (Bodleian Library, Ouseley Add. 176)made in Shiraz, dedicated to Ibrahim-Sultan, and signed by the head of his atelier, Nasr al-Sultaniwhose preface was recently identified as the old preface,

71
Ibid., . This story also appears in the 'Aruzi Samarqandi, Chahar Maqala, 80-1. 72 Riyahi, Sarchishma-ha-yi  Ibid. 74 Ibid., 412-3. 75 Ibid., 285-6, 413-4. 76 Cf. the discussion of the relevance of the eleventh-century background of Mahmud of Ghazna's campaigns to Firdausi's own audience in van Zutphen, "Faramarz's expedition to Qannuj," 56-7. 77 Subrahmanyam, "Iranians Abroad," 341. also includes the reference to Delhi. 78 This shows that the narrative about Firdausi's travel to India enjoyed some circulation at the time. After all, if he could travel conventionally to Isfahan, Tabaristan, and Baghdad, why not to Delhi? It is possible that further research into prefaces of the Shahnama will bring more examples to light, and until then the significance of this narrative in Ibrahim-Sultan's Shahnama cannot be determined. Nevertheless, given the Timurids' claims on India, and especially Delhi, at the time of the manuscript's production, it makes sense that Delhi was chosen among the other nodes in the Persianate world to which Firdausi would travel. In other words, the insertion of Delhi could serve both sides of a two-way exchange: the reception of Persianate modes of culture in India, and the (re)incorporation of Delhi into the Persianate empire.
The corresponding sections on Firdausi's stay in Delhi in Ibrahim-Sultan's Shahnama and Or. 1403 are different in length, content and structure. First, in the preface of Add. 176, Firdausi's journey to India is presented as one of the alternative narratives in the second version of Firdausi's story, where it is stated that Firdausi "sent his possessions to Tus of Khurasan, and then, some say that he left for Tus, and some say that he left for Delhi with his brother." 79 By contrast, in Or. 1403, Delhi is the only destination mentioned in the second version of Firdausi's story. In Ibrahim-Sultan's preface, the gifts offered by the Shah of Delhi upon Firdausi's departure are only mentioned in passing, and not listed in detail. Moreover, in Or. 1403, it is made clear that it was the Shah of Delhi who heard about Firdausi and invited him to his court, thus casting him in a more active role, as a protagonist.
Finally, in Or.1403's preface, the Delhi section of the story is arranged such that we first get a glimpse of Firdausi's arrival in Delhi, the invitation from Shah of Delhi, and his reception at the court. Then we are taken back to Ghazna, where the regretful Sultan Mahmud is looking for Firdausi in vain, blaming his vizier for tarnishing his name as a result of his envy: "Firdausi has written something that for as long as the world stands, people will read and write about." After hearing more about Sultan Mahmud's remorse, we then return to Delhi to read about Firdausi's decision to return to Tus and the Shah of Delhi's generous gifts. 80 The actions of the Shah of Delhi and Sultan Mahmud are staged in contrast and thus offered up for comparison.
The example of Ibrahim-Sultan's Shahnama shows that the narratives of Firdausi's travels cannot be interpreted in direct relationship to patronage and the origin of specific copies, even though it is possible that in some cases, the workshop responsible for designing these copies would adopt and expand one or more of the existing preface narratives as especially relevant. While it is possible that the makers of Or. 1403 chose and highlighted the account of Firdausi's travel to India because the manuscript was made in, or in relation to, India, the choice of Delhi as his destination was one that historically made sense, as a competing centre being incorporated into the Islamic world at the time of the poet. If Or.1403 was indeed made within India, Delhi might well be its place of production, but any other place in India could be as well.
Rather than defining the origins of a manuscript, Firdausi's travel narratives reflect ideas about the reception of the Shahnama: its circulation and history of appreciation, and by implication, that of literature and art in Islamicate societies more broadly. These were narratives that were used to both map the popularity and reception of a work like the Shahnama and its reach to every corner of the Persianate world, and to define the territorial claims of whichever "heartland" empire was in power at the time. As a line towards the end of the Baysunghuri preface reads: "Even in regions where Parsi is spoken very little, such as Egypt and Sham, and Rum and Turkestan, the book of Shahnama is found in abundance. And in Khurasan and Fars, and 'Iraqain and Hindustan, there is no town where the book of Shahnama is not plentiful." 81 By incorporating the Delhi episode in the preface, the makers of Or. 1403 (or its model, as well as Ibrahim-Sultan's Shahnama) were laying out a full circle of India-Iran cultural exchange. The older preface, which contains the Abu Mansuri preface, starts by recognising the Panchatantra and its translations, the Kalila va Dimna, as the unparalleled project of cultural production. The ancient Sanskrit text of Panchatantra was brought from India to Iran, where it was translated into middle Persian, and only later into Arabic and Persian. This was the project that inspired Abu Mansur 'Abd al-Razzaq to patronise the compilation of the Shahnama (Siyar al-muluk) in prose. Adding the Delhi episode to the preface projects the ultimate result of Abu Mansur's project to India in return and traces the history of Shahnama's reception in India to the time of the completion of the book, even before it earned recognition from its own patron. The comparison made between the Kalila va Dimna and the Shahnama in the older preface set the stage for a reconnection between the two literary traditions: Or. 1403's preface shows how these steps were taken in fifteenth-century Iran and India.
Another narrative from the third preface of the Shahnama makes the relationship between the book and India even stronger. At the point when the Shahnama was completed and Sultan Mahmud was supposed to pay the poet, his vizier, Maimandi, used an interesting and historically specific argument to convince the sultan not to pay the poet in gold coins as was agreed. The vizier reminds the sultan that "you will take your army to a war with Hindustan. The army has expenses, and if everyone (every soldier) is not paid in gold, death will be coming upon us unexpectedly from the enemy." 82 The Shahnama's value is placed on the same level as that of Mahmud's campaign in India. Mahmud's choice to save this money for the campaign costs him a positive association with the Shahnama. To some extent, in Or. 1403's preface, Mahmud's loss becomes India's gain.
Whether located in India, or working to appeal to an Indian audience, the workshop responsible for Or. 1403 knew the circulating traditions of prefaces of the Shahnama well, and this command of textual genealogiesand the text of the Shahnama which manifests itself in some of the scenes depicted in the manuscript as well allowed them to engage in an intentional act of history writing in the preface that created a contemporary record of its reception in India. 83

Frontispiece: Visual Genealogy and Innovation
The double-page illustration on folios 9v and 10r (Figure 1), although not the first painting in the manuscript, follows the preface and acts as the frontispiece to the epic. The reverse of the right-hand side of the frontispiece (fol. 9r) is the empty dedication device (Figure 8), and on the back of the left side the epic begins with a two-page spread of illuminations (fols. 10v, 11r) ( Figure 12). Traces of both the almond-shaped dedication device and the headpiece can be seen on the frontispiece.
The left side of the frontispiece depicts an interior scene in what appears to be a domed structure, where a white-bearded figure sits kneeling on one knee on a carpet with bright colours, similar to carpets depicted in Timurid and Turkman manuscripts. 84 Engaged in a teaching session, he gestures towards a book-rest before him. The five youthful figures seated in front of him, to the right of the page, are smaller in size, indicating their lower rank. Three servants, even smaller, stand to the left of the image. On the right-hand page, another large figure (although rather smaller than the white-bearded man) sits on both knees in a respectful postureone hand on his knee, the other on his heartand gazes left. A small figure, almost identical to the three whitedressed servants on the left-hand page, stands behind him under a tree and presents a similar gesture, with his left hand on his heart. Three musicians stand behind them and play the flute, pan flute, and daf. The three figures at the bottom of the page are engaged in a dance. The eyes of many of the figures have been carefully removed at a later date. The feet of the two large figures and the students, all of whom are seated, are hidden under their garments, but all the standing figuresthe servants, musicians, and dancersare barefoot.
As noted above, Brend identified the dancerspreviously thought of as dancing girlsas dervishes. She pointed to the dancers' thin moustaches, posture, and long loose hair, as well as the matching garments they share with the musicians, to suggest that the scene depicts dervishes engaged in a sama' session. 85 The two sides of the painting might appear unrelated at the first glance, but several elements connect them. The similar skylines and matching blue skies with white streaks of cloudswhich Fraad and Ettinghausen, Robinson, and Titley had all notedas well as the rounded shapes of the dome on the left side and the tree on the right side, piercing the sky, invite the viewer to read them as a set. 86 To these should be added the depiction of water at the bottom of each page: on the right a stream in the landscape, and on the left a manmade architectural element. The sashes of the two main (larger) figures on both pages, which look like openwork showing the fabric underneath, correspond visually to other architectural elements, such as the design of the wooden door and the woodwork on the roof. Finally, the attendants (smallest figures) are depicted and dressed identically on both sides: white outfits with thin girdles tied around their waist and softly hanging down in the front, as well as white and red turbans.
Several of the common visual elements of this doublepage composition are repeated throughout the illustrations of the manuscript: the landscape, the threecoloured crown of trees, the carefully-drawn bricks of the architecture, the interior architectural decorationstilework, blue paintings on the whitewashed walls, and woodworkand the carpet.
In previous scholarship the frontispiece has been interpreted mostly in terms of its subject matter, which was considered odd compared with other fifteenth-century frontispieces of the Shahnama. Commentators have highlighted its "un-Persian" elements, and identified the figures depicted in relation to the origin and patronage of the manuscript. Stchoukine, for instance, had identified the largest white-bearded figure as Firdausi. Brend disagreed with this idea. Following her identification of the dancing figures as dervishes, and her suggestion of the Deccan as the manuscript's place of production, she identified the large figure on the right as the Bahmani ruler Ahmad Shah I (r. 1422-1436). To identify the white-bearded figure on the left she looked into the Sufis active at the time in the Deccan. Rejecting the Ni'matullahi Sufis, due to their ban on dancing and their fame as iconoclasts, she concluded that the dervishes shown belong to the Chishtiyya tariqa, and that the large whitebearded man holding the teaching session is Gisu Daraz, the Chishtiyya Sufi who died in 1422 before the Bahmani capital was moved from Gulbarga to Bidar. This led Brend to suggest that the manuscript was probably offered to Gisu Daraz's dargah in Gulbarga in his memory, and that Gulbarga was the exact place of production. 87 Figure 12. Sarlauh or carpet-design illumination at the beginning of epic proper. Shahnama manuscript, British Library, MS Or. 1403, fols. 10v-11r. ©The British Library. 87 Brend, "The British Library's Shahnama," 91; Brend, Perspectives, 79.
Fifteenth-century frontispieces of the Shahnama feature enthronement and battle scenes, and open-air royal audiences with the prince and courtiers enjoying sensual pleasures. None show a scene with Sufis such as that of Or. 1403. 88 In the identification of general categories of this painting I agree with Brend's suggestions on a general level. Unlike other fifteenth-century frontispieces, the leading character of this painting, the largest figure (the white-bearded man) does not seem to represent a prince or royal figure. His size and demeanour, his engagement in a teaching session, and the dervishes dancing and playing music on the right-hand side all point to a sacred figure. The second-largest figure on the right side, with his gaze towards the Sufi, his humble posture showing him respect (and his attendant following suit) must be the representation of a notable figure paying respect at the threshold of the dargah (or any other sacred or teaching space). He might be a prince or ruler.
As discussed above, I believe that the Deccan is a possible place of production for the manuscript, one that cannot be proved or disproved based on the available evidence. Rather than attempting to identify the figures depicted in this double-page painting, the following analysis turns to its composition and the innovative ways that it relates to and differentiates itself from the frontispiece traditions of the time.
The relationship between frontispieces, single-page paintings and portraiture has received attention from several scholars in the past. In the fifteenth century, the frontispiece has usually been understood to feature the idealised image of the patron of a manuscript. 89 Sims pointed to the three princely Timurid Shahnamas of Ibrahim-Sultan, Baysunghur and Muhammad Juki, which contain "at least one picture that could be interpreted as the 'portrait' of the prince who commissioned it." 90 Soucek refers to the generic nature of some of these frontispieces, but adds that "those in manuscripts produced for princely patrons meet the criteria of a true portrait, since a contemporary viewer would have recognised in them the depiction of a specific person." She calls them "contextual" or "situational" portraits since the setting and the depiction of a characteristic activity would allow the contemporary viewer to "link the image with a specific person and sometimes even with a particular event in his life." 91 Soucek's definition of context in her discussion of these manuscripts could be extended beyond questions of subject matter to include the wider context of the manuscript itself. In other words, it is not only by recognising the setting and characteristic activities performed in these paintings that they become portraits; their placement inside specific manuscripts contributed a great deal to how they were seen as well. If the frontispieces of the Ibrahim-Sultan Shahnama, for instance, were removed from the manuscript, or were placed in a manuscript with no dedication notes, would we still see them as a portrait of Ibrahim-Sultan? We could perhaps place them in their historical, geographical and stylistic context, but we would not be able to identify the human subject(s) of these portraits with certainty. In other words, while we can probably identify the qualities and general categories alluded to, the relationship between these "portraits" and their exact subjects depend on extravisual elements, such as the manuscript as a whole and sometimes other texts that accompany it.
If we compare the right-hand side of Or. 1403's frontispiece with other frontispieces, such as the right side of Leiden's Shahnama frontispiece (Or. 494), there are several elements in common: the garden setting, trees, stream of water, ruler seated under a tree, attendants and courtiers ( Figure 13). 92 In Or. 1403's frontispiece these elements are restructured: the presumed ruler is pushed towards the centre of the double page painting, differing from his position on the right edge of the painting in Or. 494. While the usual placement of the ruler towards the extreme right created much space in front of him for the display of the vast ceremonials staged in other frontispieces, his position towards the centre of the painting in Or. 1403 shifts the emphasis to the left side, while simultaneously strengthening the connection between the two sides of the double-page. The humble gesture of the presumed ruler towards the left is accentuated by the placement of his body directly on the ground, unmediated by a carpet or throne as we find in other open-air audience 88 Clinton and Simpson, "Word and Image in Illustrated Shahnama Manuscripts," 233, point out that the only scene "in the entire Shahnama corpus up to the late sixteenth century with an overt religious iconography and theme" shows Iskandar at the Ka'ba. 89 Soucek, "The Theory and Practice of Portraiture," 104-5. Sims, "Towards a Study of Shirazi Illustrated Manuscripts," 619, discussed the popularity of these socalled Timurid portraits of the patron in the Eastern Islamic world. Rizvi, "The Suggestive Portrait of Shah 'Abbas," 243, follows the tradition in the Safavid period, and beyond the frontispiece. Golombek, "Discourses of an Imaginary Arts Council," 6-7, also discussed paintings that could be interpreted as prefigurations of the image of Timur in an anthology of epics, dated 1397-98, divided between the Chester Beatty library in Dublin (Ms. 114) and the British Library (Or. 2780). Working on early modern Deccan, Hutton, Art of the Court of Bijapur, 96, argues how visual details were employed to depict qualities that belonged to the large categories of accomplished prince, sultan, mystic, loves, etc. 90 Sims, "The Illustrated Manuscripts of Firdausī's 'Shāhnāma'," 44. See also Abdullaeva and Melville,The Persian Book of Kings,22. 91 Examples that she discusses include Timur holding court, Baysunghur drinking wine, and Ibrahim-Sultan in a battle against Iskandar b. Qara Yusuf in 1429, in one of the frontispieces of Ibrahim-Sultan's Shahnama: Soucek, "The Theory and Practice of Portraiture," 104-6. 92 Accessible at http://shahnama.caret.cam.ac.uk/new/jnama/card/ceillustration:2088903695.
frontispieces. Meanwhile, the Sufi sits on a carpet on the left side of the double-page.
The devices that serve to confer respect on the prince or ruler in other frontispieces are thus made to showcase the presumed ruler's respect towards a sacred figure in this manuscript. Similarly, instead of courtiers and attendants who eat, play music and present luxurious vessels to the ruler, we have the musicians and dancers engaged in a sama' session, while the attendants in the teaching session hold vessels of food and drink to offer the teacher and students. The feast of the usual frontispiece is transformed into a ritual, foregrounding the primacy of faith and education at the court. The archetypal code of courtly behaviour is expanded to demonstrate the social role of the Sufis as royal advisorsthe culture of advice being an underlying thread in the text of the Shahnama.
In this light, the artist(s) of this double-page spread in Or. 1403 appear to presume the viewer's familiarity with common frontispieces of the time. They use this familiarity to signal that the key figure on the right side is of courtly status, but also to displace his centrality through juxtaposition with the garden composition and teaching session. This juxtaposition introduces the new key figure of the painting, demonstrates the prince's humility towards him, and effectively transfers the spiritual authority of the Sufi onto the prince. In the Or. 1403 frontispiece, the contrast between the simplicity of the landscape setting on the right side and the intricacies of the painting on the left side (the bricks, woodwork, and carpet), as well as the dichotomy inherent in the interior and exterior settings, convey the idea of both a separation and marriage of two realms: that of material and spiritual power. 93 Figure 13. Double-page frontispiece. Shahnama manuscript, Leiden University Library, MS Or. 494, fols. 7v-8r. ©Leiden University Library. 93 The paradigms of sacred kingship that Muslim dynastiesincluding those in Indiawere experimenting with in this period regarded kings as both the ruler of the realm and the upholder of religion. Among the strategies that were employed in these processes, the establishment and demonstration of relationships between courts and Sufis were common in the fifteenth century. The depiction of Sufis in a Shahnama frontispiece has upset our understanding of frontispiece conventions in the period due to a presumed separation between the court and Sufi networks. The frontispiece was indeed supposed to reflect an aspect of the authority and glamour of courtly life, but Sufis had already been integrated conspicuously into court ceremonials and held high profile positions in court events. At the Bahmani court for instance, Sufis from different lineages were to stand on the two sides of the king at coronations. Against this background, Or. 1403's frontispiece should not be seen to mark a definitive break from the norms of other frontispieces in the period; rather, it simply chooses to show a different side of courtly life than those depicted in conventional court scenes.
That difference is also displayed in the representation of hierarchies and ethnicities. Elements such as scale and costume serve to establish categories and hierarchies between figures in the Or.1403 frontispiece. As mentioned above, the smallest figures, the servants, wear identical outfits. Dancers and musicians all wear the exact same type of outfit, in a variety of colours, which features an open cut at the bottom that reveals the long dress worn underneath. The students, the royal figure, and the Sufi master, although shown in different sizes, wear more or less the same type of clothes. The only differences are the sashes worn by the presumed royal figure and Sufi master across their upper body, and the variety of headwear: the students' headwear shows traces of red in their dominantly white turbans, while that of the royal figure has traces of indigo blue, and that of the Sufi master seems to be completely white, intensifying his fair skin and white beard.
Another way that groups of figures in the painting are differentiated from each other is their skin tone. The royal figure is depicted with dark skin that stands in sharp contrast to the master on the left who has the fairest complexion in the painting. The students seem to have a similar or only slightly lighter skin tone than the royal figure, although damage done to their faces make this observation difficult to confirm. The Sufis, musicians and servants bear a mid-tone skin colour that falls between the above colours: mostly pink for the dancers and musicians and mostly brown for the servants.
The frontispiece shows the most striking variety of skin colours in the manuscript: the few other paintings that show any variety are restricted to contrasts between white and pink skin tones. 94 One of the Shahnama scenes that regularly shows a striking difference in skin colours (although not illustrated in Or. 1403) is "Tahmina visits Rustam's chamber." In the poem, Tahmina is led to Rustam's chamber by a slave (banda) holding a candle. In several fifteenth-century manuscripts, the artists imagined the slave as black. 95 Outside the Shahnama tradition, another near-contemporary scene that depicts racial differences is "Iskandar's battle with the army of Zangis" from the Sharafnama of Nizami. Iskandar fights the zangis (lit. of Zang and Zangbar or Zanzibar, generally from East Africa, or Africa) in defense of the Egyptians. One example is the illustration of this scene in a manuscript of the Sharafnama at the British Library (Or. 13836), dated 938/1531, dedicated to the Hussain Shahi ruler of Bengal, Nasir al-Din Nusrat Shah (r. 1519-1533). The painting (fol. 14v) shows the armies of Iskandar and the Zangis arranged on the two sides of the painting, separated both spatially and by the colour of their skin: the army of Iskandar is shown with light pink complexion, while that of the Zangis is blue. 96 I draw these comparisons to make the point that the depiction of skin colour in the Or.1403 frontispiece is more nuanced than the use of race to divide the good from the bad, as in the Sharafnama scene, or the association of black skin with slaves. The frontispiece seems to be invoking racial distinctions, but does not necessarily associate them strictly with status: the attendants are fair-skinned, while the ruler is dark.
Understanding the portrayal of race in this painting, and more generally in the fifteenth century, will certainly depend on further research: my remarks on this matter here are merely preliminary. 97  See also Goswamy's discussion of skin colours (in particular copper-red and pink) in the so-called "Jainesque" Sultanate Shahnama: Goswamy, A Jainesque Sultanate Shahnama,24. 97 On this subject see Hillenbrand, "The Image of the Black in Islamic Art," 215-53; McLeod and Robbins, "The Image of the Black in India," 254-94. While nuanced variations in skin colour are frequent in Mughal paintings, a potential area of further research on this question in the fifteenth century are the illustrations of the Jami' al-tawarikh and the Majma' al-tawarikh, especially the sections on Hind and Sind, and the life of Buddha. It seems that while the Ilkhanid depictions do not noteworthy intervention in the preface of this Shahnama that could potentially relate to race, one that likewise poses more questions than can be answered at present. As mentioned in the foregoing discussion of the Shahnama's preface, the first version of Firdausi's story included verses of the Hajvnama, the poem that Firdausi was thought to have written as a reproach to Sultan Mahmud. A few lines of the poem that allude to Sultan Mahmud convey the idea that expecting positive change from those who are impure and evil at heart is a hopeless delusion. In the standard versions of the Hajvnama, one line in particular reads: "do not place any hope in those born impure, for the Zangi will not turn white by washing." As noted above, the term Zangi means "from Zangbar" (Zanzibar) and is used to refer to people from East Africa, but it could also be used more broadly to mean "black-skinned". It is commonly used in Persian literature as a racial slur to refer to anyone with dark skin, ignorant people or savages. 98 In the preface of Or. 1403, the word Zangi is replaced by the word "Hindu", which could refer both to people from Hindustan and practitioners of Hinduism. 99 Which of these meanings are intended here, and how does it relate to the depiction of skin colour in the frontispiece? If the word should be read as "Indians", how should this manuscript's association with India be rethought in light of this seemingly contradictory reference? If it is meant to refer to "Hinduism", should we understand it as an anti-Hindu expression voiced by Muslims/non-Hindus? Is the writer turning the racial slur into a religious slur? Is this merely meant to be against Hindus (or Indians), or also in defense of Zangis in India at the time? If the latter, how could this manuscript's production be related to them? If it is transferred from race to religion, would it represent an attempt to remove racial insult, in favour of a multi-ethnic regime defined by a majority religion?
Pursuing answers to these questions will require more information about the socio-political circumstances in which this manuscript was produced. But whatever those circumstances were, the relatively non-hierarchical display of skin colours in the frontispiece seems to be an attempt to show a non-homogeneous, multi-racial, and multi-ethnic vision of the courtly environment and perhaps, by implication, society at large. * Extending Hutton's discussion of Deccani paintings to Sultanate manuscripts, minimal inscriptional and historical evidence has resulted in the shaping of a sub-field that has mostly focused on issues of identification, attribution, and dating. 100 My understanding of Or.1403 is that whether it was made in one single place or between places, whether by local artists in Iran or India or through collaboration between itinerant and local artists, it is very likely that an Indian destination and audience was kept in mind in the production process. 101 While the existing scenarios for the origin of this manuscript remain speculative, putting aside the thorny question of attribution opens up a space for inquiry into the textual and visual programme of the Shahnama Or.1403 and the collaborations that brought it into being. Its preface, frontispiece and illustrations demonstrate considerable command of the text, its history, narrative traditions regarding the life of the poet, and visual strategies of prefacing the text in the fifteenth century. The preface and the frontispiece constitute a unified set that betray a sensitivity towards the manuscript's audience, and an agenda of historicization: one that used the text-image continuum to map the reception of the Shahnama as a literary text that moved across geographical boundariesparticularly in Indiaand thereby reflects the socio-political dynamics of the milieu in which this specific manuscript was produced, or for which it was intended. Moreover, text and image deploy similar strategies to achieve these goals: they both engage with received conventionshistorical and period prefaces and frontispiecesbut twist them to chart out a new history. The epic proper being somewhat immune to substantial changes, the preface and frontispiece provided an opportunity for insertions and alterations. To the informed viewer and reader, the manuscript and more particularly, the preface-frontispiece set, would have presented fifteenth-century India as an ancient and living crossroads in broader networks of cultural exchange that spanned the Persianate and Islamicate worlds.