Between about 1300 and 1600, Persian painting styles had a sustained impact on the Indian art at the Sultanate and Mughal courts as well as on Hindu painting styles. Several approaches for a description of this far-reaching phenomenon would be possible. For the 14th century, we discuss the possibility that a group of so-called Small Šāh-nāmas are of Indian provenance, and trace the origin of the šāhi figure in 14th-century Jain manuscripts to a Persian origin. For the 15th and early 16th centuries, the confused tangle of undocumented and underdeveloped styles associated with Sultanate painting are examined topologically. During the last sixty years of the 16th century, Persian artists, illuminators, and calligraphers arrived in numbers due to changes in patronage at the Persian courts and political instability in Khorasan. Only the documented painters of this period are reviewed: artists from Shah Ṭahmāsb’s court, Mir Sayyed ʿAli, ʿAbd al-Ṣamad, Mir Moṣawwar, and Mowlānā Dust; the Bukharan artists Šayḵem and his father Mollā Yusof Haravi; and the Khorasani artists, Farroḵ Beg and Āqā Reżā.


The earliest dated manuscripts from the subcontinent that rely on Persian models for some of their motifs are from the late 14th century. However, several copies of Ferdowsi’s Šāh-nāma have been dated stylistically to the late 13th to early 14th century and are believed by some scholars to be of Indian origin. It is these manuscripts, all copies of Ferdowsi’s Šāh-nāma, that will be examined first.

In the catalogue of an exhibition held at the Metropolitan Museum of Art, New York, in 1985, Stuart Cary Welch attributed a Šāh-nāma, of which all known leaves are now in the Metropolitan Museum, to the Sultanate of India (Delhi or Malwa), in the first half of the 14th century (Acc. no. 1974-290; Welch, 1985, pp. 128-29, no. 73). The manuscript, called either the Schultz Šāh-nāma or the Gutman Šāh-nāma after its first-known and penultimate owners, was reattributed in 1994 by Welch’s colleague at the Metropolitan Museum, Marie Lukens Swietochowski (p. 80). Miniatures from the Schultz-Gutman Šāh-nāma, those from a Šāh-nāma represented only by a few leaves in the Diez Album (Staatsbibliothek Preussischer Kulturbesitz, Orientabteilung, Berlin, fol. 71, S. 7, 11, 29-30, 42), and dispersed illustrations to a chapter of a book on rhetoric, Moʾnes al-aḥrār fi daqāʾeq al-ašʿār (dated 741/1340-41) of Moḥammad Jājarmi, are all stylistically closely related. Alexander H. Morton (pp. 49-51) recently demonstrated that both the author and the patron of the Moʾnes al-aḥrār were associated with greater Isfahan, which leads to the conclusion that all three manuscripts must now be assigned to that city.

A more substantial attribution of illustrated manuscripts to Sultanate India during the first half of the 14th century was first suggested by Basil W. Robinson in 1976, and has been championed by him in the intervening years (Robinson, 1976a, p. 132; idem, 1976c, p. 12). There are three heavily illustrated, undated manuscripts in this group of so-called Small Šāh-nāmas, two of which have been dispersed, and a third, “larger” Small Šāh-nāma with most folios held in the Freer Gallery, Washington, D.C. The manuscripts have been carefully reconstructed by Marianna Shreve Simpson (1979, pp. 55-140, 351-82). Significant components of this idiosyncratic style, according to Robinson, include the unusual spindly outlines of the figures and motifs, unusual pastel coloring and gold-painted sky, a tendency in some, but not all, miniatures to fill up “unused” space with flowering vines and trees, and the inclusion of motifs that are appropriate to India but not to Persia. Most notable in the last category is that all representations of lions called for by the text have been drawn as tigers, as in “Bahrām Gōr kills the monkey-lion before Šangol, the Indian king” (Plate I). Also possibly indicative of an Indian provenance is the clothing of the Indian courtiers in this miniature; their heavy long shawls arranged in regular folds and beehive-shaped turbans are unlike Indian dress in other contemporary Persian miniatures (cf., “King Kayd of Hind telling his dream to Mihram” from the Great Mongol Šāh-nāma; Grabar and Blair, p. 115). Another type of hat, made of large leaves, is appropriately worn by rustics and servants (cf., Grabar and Blair, p. 127, no. 35), but it is worn in Small Šāh-nāma miniatures by courtiers in brocade robes in court settings (e.g., Simpson, 1979, pl. 22). These costume details, together with an occasional confusion of characters in a story (e.g., the reversing of the roles of Siāvoš and Afrāsiāb, q.v., in one of the Freer Šāh-nāma miniatures (29.40; see Simpson, 1979, pl. 76), illustrate the provincial nature of the miniatures. In the three Small Šāh-nāmas, soldiers and attendants often wear a type of Mongol hat ornamented with a ball of feathers. This traditional Mongol feather headdress was banned by Ḡāzān Khan (r. 1295-1304, q.v.) in 1295, but occasionally occurs later, for example in a Marzbān-nāma manuscript copied in Baghdad in 1299 and now kept in the Archeological Museum Library, Istanbul (no. 216; Simpson, 1982, p. 104, fig. 51), and on a Kāšān luster ware star tile dated 730/1338, now in the British Museum (OA+ 1123; Watson, p. 144, fig. 122). A similarly shaped hat with a modified ornament appears in two miniatures of the Great Mongol Šāh-nāma, where the hat now seems to indicate a non-Iranian protagonist (Grabar and Blair, no. 6, p. 69, and no. 13, p. 83). If the Small Šāh-nāma miniatures were made in an Indian center, as Robinson and this writer believe probable, Mongol dress codes did not need to be rigorously applied, and indeed they were not in the Mongol heartland.

Marianna Simpson, in her 1978 dissertation, assigned the Small Šāh-nāmas to a Baghdad province, comparing them to illustrations in the Marzbān-nāma of 698/1299 in the Archeological Museum Library, Istanbul. While her detailed stylistic analysis suggests strongly that the painters of the three Small Šāh-nāmas used a manuscript of the late 13th century as a model for costumes and landscapes, provincial aspects of the Small Šāh-nāma manuscripts, as well as the quality of line, are not explained by her comparisons. Furthermore, recent research on the Manāfeʿ al-ḥayawān (Manāfeʿ-e ḥayawān; dated 697/1297-98 or 699/1299-1300) held in the Pierpont Morgan Library, New York, associates the work of the third painter on this manuscript with the frontispiece of the Rasāʾel eḵwān al-ṣafāʾ, made in Baghdad in 686/1287-88 (Suleymaniye Library, Istanbul, Esad Efendi 3638; Schmitz et al., 1997, pp. 12-16). He and a second painter working on the Morgan Library’s Manāfeʿ-e ḥayawān also worked together on a copy of ʿAjāʾeb al-maḵluqāt wa ḡarāʾeb al-mawjudāt by Zakariyāʾ b. Moḥammad Qazvini that was recently acquired by the British Library (Or.14140), Biruni’s al-Āṯār al-bāqia in Edinburgh University Library (Ms. 161), and a small Persian Kalila wa Demna in the Bibliothèque Nationale de France, Paris (Mss. or. Suppl. Persan 1965; Richard, 1997, no. 11, p. 43). For an earlier moment in this Baghdad style we have another copy of Kalila wa Demna of 1265-80, discovered in the Royal Library of Rabat (see Barrucand). The work of artists trained in the Baghdad style of the last third of the 13th century has now been established, and it bears no stylistic resemblance to the miniatures in the three Small Šāh-nāmas. As the provenances of major groups of early 14th-century Persian miniature paintings are now known (at Shiraz, Isfahan, and the Mongol capitals at Baghdad, Marāḡa, and Tabriz), it becomes even more probable that the Small Šāh-nāma manuscripts were made in a distant provincial center.

Persian motifs in Jain manuscripts. By the last quarter of the 14th century representations of Shahi (Šāhi), the king of the Sakas of Sakastān, the area of modern Sistān, and his soldiers appear in West Indian manuscripts of the Kālakācārya-kathā, a Jain text often appended to copies of the Kalpasūtra (Barrett and Gray, p. 57). The story concerns a Jain monk Kalaka, who seeks aid from the Sakas (Doshi, pp. 121-22). For illustrations of this story, new models for the foreign king—his pose of royal ease, his throne, clothing, and crown—had to be found, and the prototypes for this imagery have confused scholars. Douglas Barrett and Basil Gray, writing in 1963, argued that representation of the Shahi was probably due to the influence of Mamluk painting of Egypt and Syria, with whom Gujarat (q.v.) was closely connected by trade (p. 56). Greater publication of Mamluk miniatures in the intervening forty years makes this assertion questionable. When pictures of enthroned kings from a Bahri Mamluk manuscript of Kalila wa Demna of 1354 kept in the Bodleian Library of Oxford University (Pococke 400; Atil, all miniatures reproduced) are compared to the Saka kings shown in a palm-leaf illustration of the Kalpasūtra-Kālakakathā of 1439 V.S./1382 C.E. (Shah, color pl. H) and in a paper manuscript of this work of about 1400 in the Prince of Wales Museum, Bombay (Plate II), many significant differences become evident. Although representations of the kings are shown in profile sitting at royal ease with one pendent foot, as in the case of the Shahi illustration, similarities end here. In the Mamluk manuscript, kings wear large, wide crowns based on degenerate Sasanian winged crowns; the Shahi in the early Indian illustrations wear a very unusual crown consisting of a circular band that may either surround the head or sit on top of it and ornamented with three vertical points. The Shahi’s coat has a prominent “cloud collar,” a feature not found on Mamluk garments. The textile pattern most often found on Mamluk robes is the so-called scroll-fold, a design somewhat resembling watered-silk; this pattern never appears on the Shahi’s garments. Instead, the Shahi’s garments and throne covers have a pattern of fan-shaped flowers with seven petals encircled by vines, which again shows no relationship between Mamluk and Jain textile designs (Haldane, pp. 34-36, 79). Finally, Mamluk thrones have a flat top with projecting side finials, while the early Shahi thrones have rounded, scalloped backs that extend well above the king’s head, with paired lance points projecting from either side; and, in the Prince of Wale Museum representation, the throne is supported by gold “lions” (in fact spotted leopards).

It is believed that the narrative illustration in Jain manuscripts began only about 1300; prior to this date only iconic representations were usual, although manuscript covers (patlis) showed narrative scenes somewhat earlier (Doshi, pp. 46-48). At this same time there was a shift in style towards more precision and elegance of portrayals and an interest in illustration of new stories, such as those in the Kalpasūtra and the Kālakācārya-kathā (Doshi, p. 47). This suggests that Jain artists would be culling their iconography for late 14th-century Saka figures from roughly contemporary sources. By about 1500 the symbols associated with these foreign kings were being adjusted: in Gujarat, the crown is replaced by a turban following Timurid models, and the cloud collar disappears (Doshi, p. 146, fig. 3; pp. 50-51, fig. 7). In Mandu in central India, changes occur even earlier (Doshi, p. 51, no. 12).

Not only do the special feature of early Shahi iconography in Jain manuscripts have no counterparts in Mamluk painting, there are also no cloud collars, no similar crowns, and no similar textile designs in the 12th- to 14th-century Persian paintings, metalwork, or ceramics. Cloud collars, for example, first appear on Persian robes in Timurid painting, as a result of contacts with Ming China; Timur was in India in 1399, or well after the advent of the Shahi iconography. What then might have been the sources for the earliest Jain figures of the Saka king?

One source for all the Saka features is not known, but all of them can be demonstrated individually within the sphere of Sasanian and early Islamic art of Persia, a feature that should not surprise one, as Sakastān was a province of the Sasanian empire, ruled by princes of the royal family (Harper, p. 142). A three-pronged crown, for example, is seen on gold coins of Šāpur II (r. 309-79 C.E.); it would seem to be a disintegration of the stepped crown of earlier Sasanian rulers minus the silk hair sack (Lukonin, 1967, p. 113, pls. 106-7). The fan-shaped blossoms in encircling vines as drawn on the Shahi’s robe are present on a silver, partly gilded fifth-century Sasanian plate decorated with scroll-work and birds, held in the Hermitage Museum, St. Petersburg (Loukonine, 1967, pls. 209, 211) and were also used for 6th- to 7th-century stucco decoration at Ctesiphon (q.v.; Harper, pp. 102, 105). The elaborate gold collar of the Shahi that caps the arm socket in the Prince of Wales Museum Kālakācārya illustration (Plate II) has a parallel on a fourth-century Sasanian silver bowl in the Hermitage Museum, showing Šāpur III (r. 383-88) killing a leopard (Herrmann, p. 112, above) and in Sasanian rock reliefs from Ṭāq-e Bostān, as worn by Ahura Mazdā (q.v.) in the investiture of Ardašir II (Herrmann, p. 91, left). The symbolism of the Sasanian collar is most probably the same as that found in China and later in Persian representations of the Timurid and Safavid periods. It denoted high status and, in its earliest form, probably divinity. Both the cloud collar and the three–pronged crown are found in 9th- to 10th-century paintings from Khocho, a city on the northern silk route near Turfan, a point of artistic transfer between Chinese and Persian cultures (Bussagli, pp. 33, 105). Lastly, representations of thrones supported by lions are known from several sources, including an 11th-century Ghaznavid silver saucer in the Hermitage (Lukonin and Ivanov, 2003, p. 104, pl. 96; see also Marshak, figs. 96, 135, 170, 172, 194). Most probably a depiction of a Persian king on a metal vessel of Sasanian or early Islamic date was copied by the Jain artists.


The first “Persian” miniatures attributed to a 15th-century provenance in Sultanate India were from a Ḵamsa of Amir Ḵosrow Dehlavi (q.v.), known from some thirty-three illustrated or illuminated leaves (Brac de la Perrière, figs 1-33; Plate III). Salient Indian characteristics, described by Richard Ettinghausen (1961, pl. 1), include the presence of women wearing transparent veils, chauri bearers, thrones raised on curved brackets, handleless ewers, and Indian-type cooking pots. Ettinghausen also noted similarities between the Ḵamsa illustrations and Inju paintings of Shiraz of around 1330-41, namely the double framing lines, band-like formats extending the width of the text box, and the use of red and yellow as background colors. Other features of the Indian illustrations, such as exuberant, large floral trees filling up the background and textiles patterned with large peony blossoms or scroll-folds appear in the Inju Šāh-nāma of 731/1330 held in the Topkapi Museum Library (Binyon, Wilkinson, and Gray, ms. 23, pls. XVI-A, B, XVII-A, B) and in Farāmarz b. Ḵodādād’s Samak-e ʿayyār of about 1330-40 in the Bodleian Library (Mss. Ouseley 379-81; Robinson, 1958, pls. 2-3), but they also appear in contemporary Mamluk manuscripts.

Based on stylistic similarities it would seem that the dispersed Ḵamsa should date from the second half of the 14th century. However, scholars have given various, later dates for the script. Richard Ettinghausen believed that the rather crude writing of the Ḵamsa dates to 1450 or later (1961, pl. 1); Assadullah Souren Melikian Chirvani (1969, p. 131, n. 6) attributed the script to 1420-30, while Simon Digby (p. 48, n. 4) compared the “proto-nastaʿliq” with dated manuscripts of about 1400 and even earlier. Stylistically related later manuscripts have come to light, including a dispersed Šāh-nāma with four leaves in the Bharat Kala Bhavan, Varanasi (Khandalavala and M. Chandra, 1969, p. 47, figs. 127-28), an incomplete Ḵamsa of Neẓāmi Ganjavi, fully described in a London sales catalogue (Sam Fogg, pp. 74-79, no. 28), and another copy of the same Ḵamsa in the Biblioteca dell’Accademia Nazionale die Lincei, Rome (Caetani 36; Curatola, no. 225).

There is no indication of provenance in any of the manuscripts. Amir Ḵosrow Dehlavi was patronized by the Tughluqids (r. 1320-1414) during the last five years of his life (d. 1325), and thus their court in Delhi suggests itself as the “natural place” for prototypes of the archaic Persian imagery of the dispersed Ḵamsa of Amir Ḵosrow to have first appeared. Scholars, however, have also proposed origins in other cities of northern and western India (e.g., Brac de la Perrière, p. 4; Brend, pp. 79-89).


There are three main ways in which Persian influences on Indian painting of the pre-Mughal period show themselves. At their simplest manifestation, certain Persian motifs, not necessarily taken from paintings, are introduced into indigenous works. At another level, Persian compositions are used to “give order to” the Indian work, which usually also involves the use of select Persian motifs, including clothing. In the most complete manifestation, Indian artists simply copy Persian paintings as best they can and reuse motifs in their Persian models to illustrate other scenes in a similar Persianate style. Examples of these three types of assimilation are given below.

Dāʾud, a Muslim poet living in Avadh (q.v.) in the last quarter of the 14th century, wrote a romance Candāyana or Čandāyana (also called Laur Chanda) in Avadhi-Hindi, which is known from a dispersed copy written in Arabic script, the bulk of it kept in the Prince of Wales Museum of Western India, Mumbai (Khandalavala and Chandra, 1969, pp. 91-102). A miniature in the Bellak Collection, Philadelphia, shows the heroine Canda lying dead on a palette under a tree and being mourned by her lover Loraka (Plate IV). Persian influence is limited to the blue, gold, and silver cloud bands and incomplete geometric interlace patterns that appear on the beige hillside and on Loraka’s yellow shield and, for the cloud bands, on the blue sky. Here Persian influence need not have come from Persian paintings (and probably did not), since the cloud bands and interlace patterns are found on contemporary manuscript illumination and leather bindings, as well as other forms of Persian art.

Another type of borrowing from 15th-century Persian manuscripts is exemplified by the Neʿmat-nāma made for the Mandu ruler Ḡiāṯ-al-Din Ḵalji (d. 906/1500-01) in the British Library (Pers. Ms 149; Titley, 1983, pp. 173-74; Plate V). Here figures (often women in male clothing) wear various Persian and Indian garments and turbans, including a small, tightly wound turban worn high on the head, as featured in Turkman painting of the 1470s. Some of the male attendants and at least one female in Persian dress are in three-quarter front face, as is normally found in Persian painting, while Ḡiāṯ-al-Din and the women he is instructing in the arts of gracious living are shown in profile view, as is the Indian norm. A beige landscape covered by regularly spaced semé (clumps of grass) and flowers, or a green hillside covered with low bushes and small scallops of ground cover, the high horizon, the dark blue sky, and the white, fat clouds with curving tails, all depend on Turkman models of the last third of the 15th century (Skelton, 1959, pp. 44-50; Titley, 1983, p. 174; Barrett and Gray, p. 61; Chandra, pls. 95-96; Khandalavala, 1983, p. 25, fig. 20).

Interestingly, another manuscript, a glossary by Moḥammad Šādiābādi entitled Meftāḥ al-fożalāʾ (British Library, Or. 3299), made for Ḡiāṯ-al-Din Ḵalji earlier in his long reign (1468-1501), has 179 small miniatures that are in true Turkman style, all of which are probably the work of one migrant painter from Shiraz (Titley, 1983, p. 72). Norah Titley believes that he was also responsible for the opening miniatures of the Neʿmat-nāma, but maintains that a Mandu artist produced the remarkable miniatures of mixed style that are usually illustrated from this manuscript. A third artist completed the illustration cycle during the reign of Ḡiāṯ-al-Din’s son Nāṣer-al-Din Ḵalji, but none of these miniatures is now reproduced. This sequence of a Persian artist and a Persianate Indian painter is, unfortunately, the only one known that is clearly dated.

A third and more elusive type of borrowing began by the second quarter of the 15th century, when Indian artists and Persian artists living in the subcontinent were ordered to copy entire Persian manuscripts or to complete the illustration cycles of books begun by Persian artists in a closely imitative style. This difficult subject was first broached by Basil W. Robinson in 1967 (pp. 85-90) and was further developed by Irma L. Fraad and Richard Ettinghausen in 1971 (pp. 48-66, figs. 133-67), with Robinson adding further manuscripts to this list in 1980 (pp. 95-115); and many additional ones have been proposed in the past twenty-five years. As enthusiastic as scholars have been in the identification of Indian manuscripts of this sort, no Sultanate place of origin has yet been established for any of them, and there is no agreement on whether a number of the manuscripts are indeed from India or are simply painted in an unidentified Persian provincial style (Adahl, pp. 88-91). A dissertation on Sultanate painting now underway at the School of Oriental and African Studies, University of London, by Emily Shovelton promises to shed some light on the subject.

Clearly, other considerations besides stylistic criteria need to be applied to this group of unknown origin if the Indian attributions are to become generally accepted. An epigraphical examination of a few miniatures undertaken by Ettinghausen in 1971 shows one path that might be followed (Fraad and Ettinghausen, pp. 56-63). Ettinghausen noted calligraphic idiosyncrasies in the headings on some illustrated leaves that are considered to be of Sultanate origin and described them by the term “rhythmic parallelism.” Here the scribe gives a parallel writing to letters of the same general shape, while nearby letters of different form can even be subsumed into this pattern. Furthermore, he showed that rhythmic parallelism, while unknown in Persia, can be seen in Indian architectural inscriptions from the late 13th century to the late 16th century.

Other avenues of investigation could include research on the paints and paper of the Indian group. Comparative material from manuscripts made in Shiraz would be necessary, as Fārs province was the main area of embarkation for the artists and manuscripts going to India. Another idea is to compile a corpus of names of patrons and calligraphers of the “Indian” manuscripts. The following represents an initial effort in this direction.

A Šāh-nāma copied in Abarquh (in northern Fārs; now held at the Raza Library, Rampur) on 6 Ramażān 840/13 March 1437 by Maḥmud b. Moḥammad b Yusof Tostari, known as Kāḡaḏi (the papermaker), was introduced recently (P.3909; Schmitz and Desai, Cat. no. IV.28, pp. 206-7, figs. 314-16). Its fifty-four miniatures are by several artists, some working close to the Shiraz Turkmen idiom of the day, others giving more dramatic rendering of well-known scenes that are notable for their strident coloring. A second manuscript by this copyist, a copy of Neẓāmi’s Ḵamsa of 849/1446, was also made in Abarquh and is in the Princeton University Library (Hitti 7; Grube, no. 44, 1 illus.). The nine miniatures in this manuscript are in a more restrained proto-Turkmen style of Shiraz. Another manuscript, held in the National Museum, New Delhi (48.6.15), is written by a calligrapher whose name has been read as Maḥmud b. Moḥammad Šuštari, known as Kāḡaḏi, but it must be by the same hand mentioned above, Tostar being the Arabic form of Šuštar, a town in Ḵuzestān. The New Delhi codex, a copy of Rumi’s Maṯnawi with seven miniatures, is dated 837/1433-34 and was assigned a Sultanate provenance by Fraad and Ettinghausen (p. 49, figs. 136-37). The question is whether we have here an example of a Persian calligrapher who emigrated to India and later returned to Persia to work in Abarquh, or a calligrapher who sold his Maṯnawi of 1433-34 before it was illustrated, and the work eventually found completion in India; furthermore, is this manuscript indeed a Sultanate production at all or does it simply originate from a provincial Persian group which now, with additional documentation on the calligrapher, can be securely attributed to south Persia?

Lastly, there is also the problem of fraudulent Sultanate manuscripts. By tweaking a date or adding miniatures to an old codex, an otherwise lackluster production can be made into a much coveted and valuable example of a new Sultanate style. Some archaistic manuscripts have already been unveiled, for example, a manuscript of selections from Ferdowsi’s Šāh-nāma from Jaunpur and dated 906/1501, in the New York Public Library (Spencer Collection, Indo-Pers. ms. 1; Schmitz, et al., 1992, Cat. no. III.25, pp. 219-221, figs. 225-26; Goswamy, pp. 122-27). Its date has been tampered with, because the rhyming reference to Jaunpur is unknown before Mughal times, and figures wear turbans that are typical of the later Mughal Homayun (see HOMĀYUN PĀDŠĀH) and the Safavid Shah Tahmasb (Ṭahmāsb) periods. Although the text may be of some age, the miniatures are a modern pastiche of styles.

Another manuscript that has drawn quizzical private comments from Persian painting scholars in recent years is a copy of Saʿdi’s Bustān that was prepared for Sultan Nāṣer Shah of Mandu (r. 1500-1510) in the National Museum, New Delhi (48.6/4); the prototype for its forty-three miniatures is said to come from the Herat school of painting of the late 15th century (Ettinghausen, 1959, pp. 42-43). A closer look at the illustrations, however, suggests that the figures and their dress are indebted more to Bukharan illustration style of the later 16th century. The coats, for example, are much longer and more voluminous than any found in Timurid Herat. The architecture, too, owes little to late 15th-century Herat; the only fully ornamented area, the spandrel of an ayvān (q.v.) on folio 134a, owes more to Bukharan painting of 1600-20 than to Heravi or Mandu illumination (Schmitz, in EIr. IV, p. 529, Plate XXIX). Inspiration for the miniatures would seem to come from paintings of various origins, principally Bukharan but also other Persian schools, all later than the date in the manuscript. An inscription on folio 1a of this modest manuscript would seem too good to be true, as it says that the book was presented to Akbar by Pazand Chand, better known as Čānd Bibi, the warrior queen of Ahmadnagar! The paintings may well be modern, and a testing of pigments and microscopic analysis should be undertaken before this manuscript is published again.


Little is known of Ẓahir-al-Din Moḥammad Babor’s (Bābor, q.v.) patronage of artists, and only two illustrated manuscripts and a few single paintings remain from the Homayun years, but both Mughal rulers were avid collectors of books and had scholars and poets at their courts (Babor, pp. 403, 416; Richard, 1994, pp. 37-43). The famous Moḥammad Juki Šāh-nāma in the Royal Asiatic Society, London, bears the ownership seals of Babor, Homayun, Akbar, and their successors (Ms. 239; Robinson, 1979, pp. 83-102). It was only during the long reign of Akbar that royal patronage of artists flourished; but without the fortuitous hiring of Persian painters, newly released from Shah Tahmasb’s library, the art of the Great Mughals might have taken a different direction (Soudavar, 1999, pp. 149-55).

The arrival of several of Shah Tahmasb’s painter at Kabul was the first of three major infusions of Persian art at the Mughal court. It was the result of a major patronage change in Persia itself, for at the royal court in Qazvin Shah Tahmasb’s waning interest in painting led to the release of many court painters by mid-century, and after the disgrace of his art-loving nephew Solṭān Ebrā-him Mirzā in 1565 (q.v.), a final group of royal painters was left without patrons. Similarly, at Bukhara, after the rules of Sultan ʿAbd-al-ʿAziz and Yār Moḥammad Khan, there was a crisis of patronage by the mid-1550s that led to immigration of artists to India. At Herat, artists fled to India as a result of the turmoil of the 1580s, ending in the defeat of the Šāmlu governor-general of Khorasan by the Uzbeks in 1586. After that time, the Mughal archives report no major foreign artist joining the royal Mughal ketāb-ḵāna. The contributions of representative Persian artists are reviewed below.


After his defeat by Šēr Shah Sur in 1539 and power struggle with his brothers, Homayun spent the year 1544 in Persia, first at Herat, then at Qazvin, where he became a connoisseur and collector of painting as practiced in Shah Tahmasb’s ketāb-ḵāna. At Kabul, after he subdued his brother Kāmrān Mirzā in 1545 and where he remained, on and off, for several years, he was joined in the autumn of 1549 by two of Tahmasb’s finest painters of the second generation, Mir Sayyed ʿAli (Dickson and Welch, I, pp. 178-91; A. Welch, 1989) and ʿAbd-al-Ṣamad (Soucek, pp. 162-67; Dickson and Welch, I, pp. 192-200). Mir Sayyed ʿAli had been recently released from royal service as a result of Shah Tahmasb’s considerable reduction of the activity of his royal ateliers. Mir Sayyed ʿAli had made several contributions to the royal Ḵamsa of Neẓāmi of around 1539-43 (S. C. Welch, 1979, no. 61, 66-68). No signed work by ʿAbd-al-Ṣamad (q.v.) while he was in Persia is known. He was probably a junior painter in Tahmasb’s ketāb-ḵāna. Both artists accompanied Homayun to Delhi in 1555 and became successive directors of the imperial atelier, and the most prestigious painters at court.

Other Persian painters have also been associated with the Mughals during the Homayun years. The most famous was Mir Moṣawwer, the father of Mir Sayyed ʿAli and one of the three major painters of the first generation of Shah Tahmasb’s atelier (Qāżi Aḥmad, tr., p. 185; Melikian-Chirvani, 1998, pp. 30-51). Although there is literary evidence that he joined Homayun in Kabul and died in Delhi, only one manuscript is thought to have miniatures by him from the Kabul-Delhi period, a Ḵamsa of Nezami, its thirty-four miniatures painted variously in Persian, Bukharan, and Sultanate styles, in the Kasturbhai Lalbhai Collection, Ahmedabad (Chandra, pl. 97, pp. 188-90; attribution by S. C. Welch).

The painter Mawlānā Dust, whom Stuart Welch has identified with the librarian, calligrapher, and painter Dust Moḥammad, added one of the last miniatures to Shah Tahmasb’s Šāh-nāma, illustrating the tale of Haftvād (q.v.) and the worm, about 1540. A painting in similar style in the Berlin Jahangir Album showing Homayun receiving his brother Hendāl has been attributed to Dust at Kabul. Several drawings and copies of drawings by him while working for the Mughal court are enumerated by Robert Skelton (1994, pp. 38-44), but it is unclear if Dust ever attended his Mughal patrons in Agra. Lastly the illuminator-painter Mollā Yusof Heravi, also called Mawlānā Yusof, may have previously worked with his son Shaykhem (Šayḵem) in Bukhara (see below).

The painting style preferred by Homayun until his death in January 1556 and continued by the young Akbar during the first decade of his reign was the style of the Persian court. We now know that even the very large miniatures in the first few volumes of the Ḥamza-nāma (q.v.), the fourteen books of which were compiled between 1557 and 1572, are dominated by Safavid pictorial conventions (Seyller, 2002, pp. 45, 256-61). In them, architecture acts like a screen blocking off distant vistas; landscapes with very high horizons or none at all are shown in bird’s-eye view with small Persianate figures in elevation; and there are lines of text above and below the images. During the later production of Ḥamza-nāma illustrations most of its Safavid characteristic disappeared. The most prominent reminders of the Persian heritage are the carefully rendered, colorful patterns on clothing, objects, and architecture.

We have only a few works of Mir Sayyed ʿAli during the time he was associated with the Mughal court, and the most outstanding ones are paintings and drawings of single figures. They show the gradual adoption of greater naturalism, especially in the drawing of faces (Seyller, 2002, pp. 60-65, nos. 6a, 7, 8a) and in the more painterly depiction of horizon lines, rocks, and trees. ʿAbd-al-Ṣamad’s works are more plentiful, but they are largely from the beginning and end of his very long career. His three earliest known works are in the Golšan Album kept at Golestān Palace (qq.v.), Tehran. One of these, Akbar presenting a painting to Homayun as they sit together in a garden flanking a busy courtyard and pavilion, must date shortly before Homayun’s premature death in 1556 (Plate VI). The figures wear the distinctive headgear designed by Homayun, but the painting is otherwise indistinguishable from works executed at the Safavid court in the previous two decades. There is no characterization of faces and no shading of forms. A large and much over-painted and altered painting of the court of Homayun in the British Museum, exhaustively discussed in the papers published in Humayun’s Garden Party (ed. Sheila Canby), is probably by him too, originally painted about the same time. It gives a similar Persianate impression. By contrast, ʿAbd-al-Ṣamad’s later work, such as “King Jamšid writing on a Rock” of 996/1587-88 (Freer Gallery of Art, F1963.4; Seyller, 2002, p. 77, no. 17) is largely painted in muted shades of brown, grey, and green, as especially shown in the building up of rock forms. The small, figures are organized in a circle around an empty center that convincingly extends back into a shallow middle space; the emptiness emphasizes the splendid natural setting. This “Mughal” landscape is inhabited by small, pastel-robed figures with little dramatic interplay. As Robert Skelton has written, “Abd us-Samad ventured on a course that curiously combined innovation with conservatism. It is as though he had the vision and the power to communicate this to his pupils without the ability in his own work to arrive at the integration of idea and execution that leading spirits of the next generation triumphantly achieved” (1994, p. 37).


The finest manuscripts made for the great Bukharan patron of the book, Solṭān ʿAbd-al-ʿAziz Khan (d. 1550), also one of the last, is Rawżat al-moḥebbin (the name of the last of its three maṯnawis) of Ebn ʿEmād Širāzi in the Salar Jung Museum, Hyderabad, AP. It was completed in Bukhara by Solṭān ʿAbd-al-ʿAziz Khan’s librarian, Solṭān Mirak, at the end of the year 956/January 1550; another section dated 956/1549 was copied by the famous Heravi calligrapher Mir ʿAli Kāteb (Ashraf, 1975, no. 2281, pp. 7-23, 4 illus.; Randhawa, 10 color illustrations). Several of the ten double-page miniatures are signed by the court painters Maḥmud Moḏahheb, ʿAbd-Allāh Ḵāqāni, and Shaykhem b. Mollā Yusof Heravi; the ones without signatures can also be attributed to their hands.

Maḥmud Moḏahheb and ʿAbd-Allāh Ḵāqāni are well-known Bukharan painters (Sarkisian; Soucek, in EIr. I, pp. 193-95; Schmitz, in EIr. IV, pp. 527-30), but the third painter’s name is not familiar to art historians. According to Mirzā Moḥammad Ḥaydar Doḡlāt (q.v.), in Tāriḵ-e rašidi, Mollā Yusof Heravi, Shaykhem’s father, was a pupil of Behzād, and he refers to him as a speedy painter whose illumination (taḏhib) was superior to his painting (apud Thackston, p. 362). Mollā Yusof is probably to be identified with the Mawlānā Yusof mentioned in the memoirs of Bāyazid Beg Turkman, which is an account of the artistic activities at Homayun’s court in Kabul. An example of Mawlānā Yusof’s work was included with paintings and drawings by “Mawlānā Dust, Mawlānā ʿAbd-al-Ṣamad, and Mawlānā Darviš Moḥammad” sent by Homayun to Nawwāb Rašid Khan, the ruler of Kāšḡar (Chandra, pp. 172-73, quoting Beveridge). Possibly Mollā Yusof gave his son his unusual name to honor Shaykhem Beg Sohayli, one of Sultan Ḥosayn Mirzā’s amirs and, according to Babor, one of the three best poets at his court (Bābor, foll. 174a, 179b; tr. Beveridge, pp. 277, 286; tr. Thackston, pp. 217, 223).

In the Rawżat al-moḥebbin, Shaykhem signs and dates a royal court (darbār) scene (foll. 12b-13a); two additional double-page miniatures may be attributed to him, a mosque scene (foll. 7b-8) and a scene in the zanāna (foll. 132b-133a; Ashraf, illus. facing p. 22). The court scene is dated by an inscription over the main gate, 955/1548-49. The right-hand page is ascribed to Shaykhem Yusof (the and the ḵa are not pointed), and the left-hand, to Shaykhem b. Mollā Yusof Heravi (a name fully pointed); as the painting style is the same for both leaves and they are thematically related, they are surely by one artist. The figures wear clothing appropriate to Bukhara at this time, including turbans wound around a conical Central Asia headgear (kolāh) and fur-trimmed hats. Activities are depicted in two registers, front and back (above), that do not overlap, but figures and horses do overlap in their respective registers; the spatial organization is simple and clear. Shaykhem’s small, active figures strongly recall the Behzād idiom followed by Šayḵ-zāda. Šayḵzāda is a painter of prodigious talent, and it can only be because of the obscure publications of the Rawżat al-moḥebbin miniatures that scholars have not generally recognized his worth.

After the death of Sultan ʿAbd-al-ʿAziz, Shaykhem remained in Bukhara working for his successor Abu’l-Fatḥ Moḥammad Yār Bahādor Khan. He added a miniature of Shaikh Ṣanʿān fainting after seeing the face of a beautiful Christian maiden to Mir ʿAli-Šēr Navāʾi’s Lesān al-ṭayr (Turk. tr. of ʿAṭṭār’s Manṭeq al-ṭayr) held in the Bibliothèque Nationale (Suppl. turc 996, fol. 20a; Richard, 1997, pp. 121, 147, no. 97). The elaborately tiled interior of the Shaikh’s hostel (ḵānaqāh) includes an inscription giving the name of the patron, the date 960/1552-53, and the provenance “in Bukhara.” The faces are wide at the forehead and taper to pointed beards and have small dotted eyes; they are thrust forward and tilted upward, and are very similar to those in the Hyderabad manuscript.

Although further Bukharan manuscript illustrations by Shaykhem will probably come to light, his next known work so far is from 974/1566-67, when he added six paintings, four of them ascribed to him, to a copy of Saʿdi’s Golestān in the British Library (Or. 5302; Titley, 1977, p. 147). In the miniature on folio 30a, which shows an old wrestler defeating an arrogant student, the artist has incorporated an inscription around the top of the throne, acknowledging his residence in India; and he names his patron: “It was ordered in the days of the prosperity of the great king Jalāl-al-Din Moḥammad Akbar, may Allāh perpetuate his kinship and sovereignty” (Plate VII). The king in this representation, as well as the principal figure in another miniature (fol. 91a; Martin, pl. 147), is given a swarthy complexion, and his facial features resemble other contemporary images of Akbar. The attributions to Shaykhem are in large inscriptions at the center bottom of the miniatures (the undotted tooth of the is clearly visible); but, because of an earlier misreading, the name Šahm appears in the literature (Titley, 1983, pp. 190-91). The Golestān miniatures are larger and have many more figures than those in the earlier Bukhara manuscripts, and the figures now wear the attire and turban of Akbar’s court. However, the depictions of trees and vegetation, the clarity of spatial organization, the pert, spade-shaped faces with their distinctive small eyes, and the richly patterned robes and carpets all reflect closely Shaykhem’s miniatures in the Rawżat al-moḥebbin.

Two further paintings, unsigned but in Shaykhem’s distinctive style, are in an Ḥosayn Wāʿeẓ Kāšefi’s Anwār-e sohayli (q.v.) of 978/1570-71 in the School of African and Oriental Studies, University of London (Beach, 1987, fig. 48). Shaykhem made no attempt to experiment with the naturalism of backgrounds and the plasticity of modeling for the human figure, which already in 1570 typify the new taste in Indian painting. Nevertheless, the six miniatures in the British Museum Golestān are truly the pinnacle of the mid-16th century Bukharan style.

In Mughal annals, Farrokh Beg/Farrokh Hosayn (Farroḵ Beg/Farroḵ Ḥosayn) and Aqa Reza (Āqā Reżā) are the two most important Persian artists to arrive at the Mughal court in the 1580s. Farrokh Beg arrived in Mughal territory in December 1585 after service with Mirzā Moḥammad Ḥakim, Akbar’s deceased brother and ruler of Kabul. From around 1586 to 1595 he added miniatures to several major royal manuscripts; about the year 1596, he became attached to the court of the Adelshahid Ebrāhim II at Bijapur (q.v.), where he made paintings of Ebrāhim and his courtiers and Sufis; by 19 December 1609, he was once again at the Mughal court, where the Mughal emperor Jahāngir, who described him as one of the incomparable persons of the age, presented him with 2,000 rupees (Jahāngir, p. 91). Farrokh Beg continued working at the Mughal court until at least 1615, the year of his last dated paintings (Das; Seyller, 1995; Skelton, 1957). While the outlines of his career on the subcontinent are now agreed upon, even though aspects of his oeuvre remain in dispute, the first forty years of Farrokh Beg’s life are far from clear. An initial examination of the problem has been given by Abolala Soudavar (1999, pp. 55-61), whose several hypotheses on his life and attributions of works are sure to attract discussion.

Eskandar Beg Turkamān Monši refers to a brother of the Georgian-born painter Siāvoš Beg Naqqāš, named Farrokh Beg, both of whom served Ḥamza Mirzā during the reign of Moḥammad Ḵodābanda (r. 1578-87; Eskandar Beg, p. 176, tr. Savory, p. 273; Soudavar, 1999, p. 55, n. 84). Indian sources give a different ethnic background to the painter. In the Āʾin-e akbari (Abu’l-Fażl ʿAllāmi, I, p. 114; Skelton, 1957, p. 395), he is called Farrokh the Qalmāq, indicating association with a Central Asian Turkic tribe, and a notice on a flyleaf of a copy of Badr-al-Din Helāli’s Ṣefāt al-ʿāšeqin in the Sayeedia (Saʿidiya) Library, Hyderabad A. P., refers to a sale by a Farrokh Beg Qaqshal (Qāqšāl), Qaqshal being another small Turkic group active near Kabul (see below). Yet another origin was suggested by Robert Skelton, who proposed that the painter was the son of Mawlānā Darviš Ḥosayn, a scholar and teacher of calligraphy and illumination in Shiraz, and that a well-known poet, Ẓohuri Toršizi, with whom Farrokh Beg was to serve at the Bijapur court, lived in Darviš Ḥosayn’s household for seven years (Skelton, 1957, pp. 401-2). While it is possible that friendship between the two men had been the cause of Farrokh Beg being introduced to the court of Adelshahid Ebrāhim II, no further evidence have turned up in support of Skelton’s hypothesis.

Several signed or attributed works from Farrokh Beg’s pre-Mughal days are known, but they, too, are not without controversial aspects. The best known of these is a drawing of a nobleman prisoner with his arm fettered in a wooden yoke (Plate VIII). Several version of this theme are known, but only two of them show the costume used by Farrokh Beg. The earlier, painted on silk, was once in the collection of Jacques Doucet, a Parisian couturier and collector; Fredrik Robert Martin’s suggested attribution to Kamāl-al-Din Behzād can hardly be accurate, although the painting belongs to that time period (Martin, pl. 82). A fine copy of it, in the Topkapi Saray Library, Istanbul, has an inscription, ʿamal-e Šayḵ Moḥammad (H. 2156, fol. 45a; A. Welch, 1974, pp. 462-63, fig. 1). The third one, a mirror image of the earlier two, is signed ʿamal-e Farroḵ Beg, šabih-e Bayram Oḡlān. Dickson and Welch have identified a Bayram Oḡlān, an Uzbek governor of Ḡarjestān (q.v.), who was captured by the Safavids in 1550 (I, p. 253), although there are discrepancies in the spelling of the man’s name, and this association would postdate the creation of the image by some twenty or more years.

A Ḵamsa of Amir Ḵosrow Dehlavi with colophons dated 978-79/1570-72, in Kings College, Cambridge (Ms. Pote 153), has eight miniatures all signed ʿamal-e nāder al-ʿaṣri Farroḵ Beg (and on some, the title Bahādor is misspelled; Skelton, 1957, pl. 2, fig. 4; Robinson, 1992, pl. IXb). It has not been noticed that the miniatures were drawn at least ten years after the dates in the colophons: most of the youths wear a type of bulbous, flat turban with contrasting cloth tie encircling the girth introduced during the years of disturbances in Herat in the early 1580s (Schmitz, 1986, p. 106, pl. 11a); and in one of the unpublished miniatures, a tall stalk of flowers extends above a youth’s turban, a motif introduced in the work of the painter Mahdi, a student of Moḥammadi of Herat, who was active in the 1580s (Robinson, 1992, pl. VIIIa-b). A second, unpublished manuscript, a copy of Helāli Jaḡa-tāʾi’s Ṣefāt al-ʿāšeqin of 1578 said to have miniatures in Khorasani style, was sold by a Mirzā Farrokh Beg Qaqshal in 1581 (Fischer; Skelton, 1957, p. 406; Seyller, 1997, p. 307).

Important evidence of Farrokh Beg’s style before his years in the subcontinent is offered by a recently published colored drawing from the Golšan Album, Golestān Palace Library (qq.v.), Tehran, showing Moḥammad Ḥakim Mirzā, the brother of Akbar and ruler of Kabul, with his vizier Ḥāji Yāqut, drawn in the Šahrārā Garden of Kabul in 1584 (no. 1663, fol. 47; Soudavar, 1999, p. 60, pl. 30). The work is signed Farroḵ Ḥosayn, confirming that the artist must have used both signatures throughout his life. The inscriptions are written in fine nastaʿliq script and, as Soudavar has noted, show that the “signatures” on other Persian and Indian works, written in an unpointed and awkward handwriting, are probably attributions, some possibly made by Emperor Jahāngir (Soudavar, 1999, p. 60). The drawing is highly influence by the Khorasani style of 1570-85, as are the early works by Aqa Reza at the Mughal court (see below). How, or if, this accords with the usual assumption of a Mašhad influence on the works of Farrokh Beg is not at present clear.

According to the Tozuk-e jahāngiri, Aqa Reza of Herat joined Jahāngir’s service while he was a prince (tr. Rogers, II, p. 20), before the birth of his son Abu’l-Ḥasan in about 997/1588-89 (Beach, 1978, p. 86). Although his sons used the nesba Mašhadi, the early work of the father clearly reflect training in the artistic style of Ḥerat in the 1580s. A painting of a bearded man seated on a low throne holding a lute beneath a willow tree in the Museum of Fine Arts, Boston (14.609; see Beach, 1978, p. 94), a composition notably similar to the work of 1584 by Farrokh Beg in the Golšan Album mentioned above, is inscribed with the artist’s name and that of his patron Sultan Salim. This is Aqa Reza’s earliest dateable work, executed before 1008/1599-1600, when Salim assumed the name Shah Salim. An inscription on a picture of a stocky, middle-aged man holding a spade, formerly in the Edwin Binney 3rd collection (Binney, 1973, p. 70, no. 42), mentions Aqa Reza as the true servant of Shah Salim; it was executed presumably between 1599 and 1604, when Salim succeeded Akbar, taking the name Jahāngir. Its linear style and portrait-like face are particularly close to a picture of Moḥammadi by Mahdi in the Museum of Fine Arts, Boston (14.583; Karimzāda Tabrizi, p. 1516, pl. 30).

An unsigned drawing, identified as “a portrait of ʿAbd-al-Moʾmen Khan Uzbek” in the moraqqaʿ compiled for Nāṣer-al-Din Shah in the Golestān Palace Library, may be attributed to Aqa Reza by comparison with these two paintings (Plate IX). The contours of the Khan’s robe with its attenuated curves, the treatment of the stand-away neckline of the robe revealing the inside collar, and the cone-like form of the man’s forearm have close parallels in the signed picture of the seated man with a lute in the Museum of Fine Arts, Boston. The depiction of the boots in the Golestān and Binney pictures is almost identical. All three small figure studies attempt to give a likeness of the person’s visage, but the bodies are abstracted into pleasing, curved lines. ʿAbd-al-Moʾmen Khan, the ruler of Balḵ (q.v.), was the son and successor, briefly, of ʿAbd-Allāh Khan, the Shaybanid ruler of the Uzbeks (d. 1598). He participated in the conquest of Herat in Rabiʿ II 996/February-March 1588 (McChesney, 1980, p. 54), and the likeness might have been made at that time, or perhaps Aqa Reza worked for him in Balḵ before arriving at the Mughal court.

Aqa Reza Heravi’s “portrait style” did not long find favor with his Mughal patron, for there are paintings and drawings in several disparate pens (qalam) reliably ascribed to him. These include a Persianate miniature style as seen in six paintings in a manuscript of Wāʿeẓ Kāšefi’s Anwār-e sohayli in the British Library (Add. 18579; Wilkinson, pls. 2-5, 7, 29; Beach, 1978, p. 93); a series of single figures on borders of the Golšan Album leaves copying European and Mughal works (Godard, 1936, pp. 13-14, figs. 1-8; Ātābāy, p. 353, pls. following p. 346); large drawings with color in the Mughal style of the Akbar-nāma (A. Welch and S. C. Welch, 1982, no. 60; Falk and Digby, no. 16); copies of European subjects as interpreted by Bukharan artists (Schmitz, 1997, pp. 149-52, pl. 31); and archaistic copies of Mašhad compositions of the 1580s with some portrait faces (Binyon, Wilkinson, and Gray, pl. 104-A, no. 236).

Aqa Reza’s eclectic paintings and drawings, whether their diversity was imposed by fiat of Shah Salim or, less likely, was a self-conscious exploration of artistic canons imposed by the artist on himself, needs to be studied by those who wish to know the origin of the consolidated Jahāngir style. Only in the works of his two sons does one see the chemical change that formed a united style from these disparate elements. The well-known works of Abu’l-Ḥasan (Losty, 1991, pp. 69-87; Beach, 1978, pp. 86-92) and the lesser output of his brother [Moḥammad] ʿAbid (Beach, 1978, pp. 78-80, 85; Schmitz and Desai, 2004, Album 3, fol. 17a, pl. 29) continue to reflect aspects of their Persian heritage: the bright colors of pure pigment, and a resultant hard-edge definition of figures, descend from Persian miniature paintings; and, to a certain extent, their quest for portrait likeness of the sovereign and his courtiers was rooted in Persian experiments of the second half of the 16th century.

Bibliography: See below, xxi.

(Barbara Schmitz)

Originally Published: December 15, 2004

Last Updated: March 27, 2012

This article is available in print.
Vol. XIII, Fasc. 1, pp. 65-76