Paper

The unusual layout of the text consists of two columns of oblique lines of nasta‘liq script copied onto Islamic laid paper (Figure 1). Describing the colour of the paper used in a manuscript is always problematic, but Robinson's description of the paper used for the manuscript's small folios as ‘pale cream, pink, yellow, coffee-colour, and green’ is misleading, creating the false impression of a manuscript of brightly coloured folios. In fact, the majority of the paper is off-white in colour (or, in Robinson's terminology, pale cream), some with a slight hint of tan, others with more of a yellow tinge. A much smaller group of folios are slightly pinkish. But these are all mere hints of colour that are in fact easily overlooked, especially in low light. However, there are, indeed, dark tan, coffee-coloured folios, but only three of them.

Figure 1. CBL Per 294, f. 5a.

Robinson states that the inclusion of coloured paper is ‘characteristic of anthologies of the middle Timurid period’ and, as anyone familiar with fifteenth-century Persian manuscripts knows, this is indeed true. To back up this statement, he lists three other manuscripts with coloured paper, and, even though he never actually says that all four employ the same type of coloured paper, the impression, though false, is that they do. The first manuscript Robinson cites is another Chester Beatty anthology, CBL Per 127, dated 1449 (ah 835) and measuring 21.2 x 7.7 cm.Footnote ⁴ Anthologies in the long and narrow format of this manuscript are known by the term safina; they were especially popular in the fifteenth century and are well-known for their usual inclusion of at least some coloured pages, usually in medium shades of mauve or a pinky-peach/salmon colour but in other, darker colours as well. Thus, unlike the barely coloured paper of CBL Per 294, these coloured pages are a prominent feature of any manuscript in which they occur. Moreover, typically they are decorated with stencilled patterns, left in reserve, with the coloured pigment filling the surrounding ground, though confined to the framed text area of the page, leaving the margins of the page uncoloured. The colour therefore is not part of the fabric of the paper, as are the mere hints of colour in most of the paper of CBL Per 294, but has merely been applied as surface decoration.

The third manuscript Robinson offers as a comparative example of an anthology with coloured paper is the well-known one produced in Shamakhi (Shirvan) in 1468 (ah 873), with folios that measure 23.2 x 13 cm, now in the British Library (Add 16561; see App. A.7).Footnote ⁵ It consists exclusively of bright, intensely coloured folios of mauve, pink, or yellow-green paper, which again is not in the least like the paper of CBL Per 294 described above or even that of the safina, CBL Per 127. Moreover, the paper is a distinct and highly identifiable type, generally thought to be of Chinese manufacture, though no definitive proof of its origin has as yet been found.Footnote ⁶ Surprisingly, seven folios of this same type of paper are included in CBL Per 294: ff. 72, 133–134 and 165–168 (Figure 2). All are an olive-green colour of medium intensity. Robinson seems not to have noticed the difference between these seven folios and all the others in the manuscript (or, for that matter, the difference between this type of paper and that used in the safina, CBL Per 127), having simply included ‘green’ in his list of paper colours.

Figure 2. CBL Per 294, f. 134a.

Including BL Add 16561, a total of 17 manuscripts consisting of this type of paper exclusively are currently known,Footnote ⁷ while individual sheets of ‘Chinese’ paper—74 of calligraphy and one with a drawing—are included in two albums in Istanbul.Footnote ⁸ Paper of this type may be undecorated or decorated with large flecks of gold and/or various gold designs, with both types of decoration appearing to have been added at the time the paper was manufactured, or at least at some point before it was purchased or otherwise obtained for use in any particular manuscript. In the British Library's Shamakhi Anthology, openings with facing pages decorated with large flecks of gold alternate with those that have no ‘Chinese’ decoration at all. The earliest known manuscript with paper of this type—all in shades of grey—is a Kulliyat of Sa‘di in the Bodleian Library, Oxford, dated 1437 (ah 840; BOD Pers. e. 26; App.A.1), and likewise in this manuscript openings with no gold decoration almost always alternate with decorated ones. However, here the decoration is either flecks of gold or a variety of designs meticulously painted in gold: landscapes of mountains and trees (in one case including a Chinese-style building), swirling waves, and depictions of fruit on the stem or vine, including pomegranates, small melons, and what appear to be cherries or grapes. By contrast, of the seven folios in the Chester Beatty Anthology, only one page—f. 134a—has original, ‘Chinese’ gold decoration, in this case what appears to be an immature fruit of some sort and lobed leaves on a curling vine (see Figure 2).Footnote ⁹

This graceful gold painting begins at the upper edge of the page and extends downwards for about three-quarters of the length of the page, but clearly it is only a section of a larger, more complete design. This is typical of the gold designs found in other manuscripts of ‘Chinese’ paper, wherein the design on any one page most often consists of a few, often incomplete elements surrounded either by large areas with no decoration at all or only gold sprinkles. Among the most extreme of such examples are a lone, bare branch emerging uncomfortably from the lower, gutter edge of a page (e.g. BL Add 7759, f. 30b; App.A.6) and a few short wavy lines enigmatically positioned down the long outer edge of a page (e.g. BOD Pers. e. 26, f. 179a), with the remainder of each page decorated with gold sprinkles only. Priscilla Soucek has reconstructed several of the sheets of ‘Chinese’ paper from which were cut the bifolios used in a copy of Makhzan al-asrar, dated 1478 (ah 833) and now in the New York Public Library (Persian Ms. 41; App.A.9).Footnote ¹⁰ These reconstructions make clear what is suggested by the incomplete designs on the folios of all such manuscripts, namely that, curiously, these were large and expansive designs that were never intended to be confined to a single page, or even to a single bifolio.

As for the seven folios in the Chester Beatty Anthology, it was surely mere chance that only one page is decorated with a ‘Chinese’ gold design, the other folios simply having been cut from blank areas of otherwise decorated sheets. This would have been possible especially because of the small size of the folios and bifolios.Footnote ¹¹ Indeed, the fact that so few folios are included, when all other known such manuscripts consist solely of this type of paper, suggests that these few folios were probably cut from the remains of sheets of paper used for another, larger manuscript. In other words, they were mere scraps, although expensive and very precious scraps.

Like all such paper, each of these ‘scraps’ has an exceedingly sleek and smooth, almost porcelain-like, finish. Islamic paper is typically sized with starch (often wheat starch), which, when highly burnished, results in a smooth and characteristically hard surface. However, the surface of this ‘Chinese’ paper is different, having a silky sort of feel not typical of Islamic paper and lacking its hardness. The paper is also unusually heavy. If, when turning the pages of a manuscript, a folio of typical Islamic paper—or indeed of almost any type of paper—is held upright and then released, it generally falls with, at most, only a very slight curving or bending, being capable (more or less) of supporting its own weight. This is true also of the seven olive-green folios of CBL Per 294 and those of other manuscripts of a similar small size, such as the NYPL manuscript, which consists solely of light blue folios measuring 18 x 10.5 cm, and the slightly larger BL Shamakhi Anthology. Although the unusual weight of individual folios is not evident in these small-format volumes, it is very apparent in folios of a larger size, such as those of the two companion volumes of the poems of Farid al-Din ‘Attar in Istanbul with folios in rich shades of pink, purple, orange/peach, and grey, and now measuring 35.9 x 24.3 cm (Sitta of ‘Attar, TPL A.III 3059; App.A.2) and 34.5 x 25.9 cm (Khamsa of ‘Attar, TIEM 1992; App.A.). Both volumes were made for the Timurid ruler, Shah Rukh (d. 1447/850), although only the one in the Topkapi Library is dated, to 1438 (ah 841). In turning the pages of these two manuscripts, each folio must be fully supported: if left to fall on its own, it will curve and bend dramatically, the sheet of paper being unable to support its own weight. Of course, when even a small manuscript consisting only of paper of this type is lifted, its unusual weight, as a whole, is immediately obvious. It has, in the past, been suggested that the very smooth surface finish and weight of the paper might be the result of the application of a leadFootnote ¹² or a clayFootnote ¹³ glaze, while Barbara Schmitz refers to the paper of the NYPL manuscript as ‘waxed’.Footnote ¹⁴ Recent scientific testing of the Bodleian manuscript has, however, revealed that the paper has a very high lead content.Footnote ¹⁵

Marginal decoration

Throughout the manuscript, the margins of the folios are filled with an array of scenes and patterns rendered entirely in gold and apparently executed freehand.Footnote ¹⁶ Of these two basic types of decoration, the more prevalent is landscape scenes. Most consist of dense vegetation, in which trees play a prominent role, in particular the especially tall and straight trees of the long outer side margins and those that are dramatically bent to fit within the shallow height of the upper and lower margins (Figure 3). Populating these landscapes is a variety of animals and birds (Figure 4), as well as other beings, both human and celestial (angels) or mythical (simurghs, dragons, and horse-headed bipeds). Repeat patterns fill the margins of other pages, and these most often consist of blossoms and leaves on an interlaced or scrolling arabesque vine (Figure 5). Often various beings, especially birds, populate the arabesque vine, in which case the blossoms and leaves are small, merely filling the background rather than being the dominant element of the pattern (see Figure 1). In neither of these two types of decoration—landscape scenes and repeat patterns—are any of the elements and motifs outlined, and if interior parts, such as eyes, are delineated, this is usually achieved merely through the absence of pigment, with many forms therefore depicted largely in silhouette. The manipulation of the colour, density, and surface finish of gold typical of all-gold borders of Safavid and Ottoman manuscripts of later years is absent here. Human figures often have elongated torsos, as in Jalayrid paintings, and wear the tight-fitting caps (see Figure 4) seen in both Jalayrid and early Timurid paintings. Non-human figures can be slightly odd, likewise with elongated bodies (see the lower margin of Figure 3) or perhaps unusually long and thin necks, while swimming or seated ducks, as well as other birds, are often little more than a quickly executed S- or Z-shape (see the long outer margin of Figure 3). Or, animals might be correctly proportioned but awkwardly positioned. The oddities of the various elements create an immediate impression that the marginal decoration has all been hastily executed by someone not overly skilled, but closer contemplation generates a more positive view of the presumably lone artist responsible for it. Considerable care was clearly needed to execute the many minute details of the vegetation of the landscapes, and, despite the repetition of the same motifs and elements, the variation in the details, in both the landscapes and the intriguing array of patterns used for all other margins, suggests the margins are the product of a fertile and agile imagination.

Figure 3. CBL Per 294, f. 107b.

Figure 4. CBL Per 294, f. 22b.

Figure 5. CBL Per 294, f. 145b

As Robinson points out, these are the earliest known examples of a type of marginal decoration—namely, executed solely in gold—that otherwise does not appear in Persian manuscripts until the sixteenth century.Footnote ¹⁷ However, while a harbinger of what is to come, certain elements of the margins hark back to the decoration of earlier Shiraz manuscripts. The first of these is the manuscript generally referred to as the Bihbihani Anthology, after the calligrapher, Mansur ibn Muhammad Bihbihani, whose name, along with the date 1398 (ah 801), is recorded in the manuscript's colophon. Now in Istanbul (TIEM 1950), the manuscript is a product of Shiraz. Ten of its eleven paintings are landscapes devoid of human figures, which are rightly described by Bernard O'Kane as being ‘among the most intriguing in Islamic art’.Footnote ¹⁸ As O'Kane has shown, a propensity for landscapes was a feature of painting during the period of Muzaffarid rule (1353–93/754–95), both in manuscripts and architectural decoration.Footnote ¹⁹ While there are no exact parallels between the landscapes of this manuscript and those in the margins of the Chester Beatty Anthology, there are clear reminiscences of the former in the latter, the most obvious of which are the tall, straight-trunked trees that are so much a feature of the paintings of the earlier anthology. Many of the tree trunks in the paintings of the Bihbihani Anthology are entwined by what O'Kane describes as a ‘waving tendril of boa-like vegetation’Footnote ²⁰ consisting of densely spaced green leaves on a thin and barely noticeable red stem.Footnote ²¹ Wrapped around other trees, however, are longer and thicker vines, usually a deep red, that are devoid of vegetation except at their upper reaches.Footnote ²² It is trees encircled by this latter type of vine, in which the stem itself is visually prominent, that are clearly recalled in so many pages of the Chester Beatty Anthology, although there it is always an entirely leafless vine that wraps itself around straight-stemmed trees placed within the long outer margin (see Figure 3).Footnote ²³ The type of tree itself varies, but of particular note are trees with distinctive feathery leaves, of the type depicted in Figure 3, and probably intended as types of willow, which also feature prominently in the Bihbihani manuscript.Footnote ²⁴ Whether or not the artist of the Chester Beatty margins was looking directly back to the Bihbihani Anthology and other Muzaffarid manuscripts in which similar types of paintings appearFootnote ²⁵ or to now-lost intermediary models is impossible to say.Footnote ²⁶ Nevertheless, although relegated to the margins, the manuscript's focus on painted landscapes is highly unusual for its time. It is, however, about the same time as the margins were painted that landscapes populated with animals and birds—both real and mythical—began to appear stamped onto the exterior covers of bindings, as on the binding of the 1438 Sitta of ‘Attar manuscript, mentioned earlier for its use of folios of ‘Chinese’ paper.Footnote ²⁷ These landscapes in leather may also have been a source of inspiration for the Chester Beatty artist.

The second manuscript to which the marginal decoration of CBL Per 294 appears to be greatly indebted is the well-known anthology produced in 1410–11 (ah 813–14) for the Timurid prince and then governor of Shiraz, Iskandar Sultan (BL Add 27261).Footnote ²⁸ The relative fame of Iskandar's so-called London Anthology is in large part due to its decorated margins, even though on less than six per cent of its pages (or folio sides) are the margins actually filled with decoration.Footnote ²⁹ On only two (facing) pages are the margins decorated with a treed landscape (ff. 414b–415a), and these are very different to the many landscapes that fill the margins of the Chester Beatty Anthology. In the latter, a complete landscape appears in each of the upper and lower margins of a page, with ground, sky, animals, and trees included. The tall trees of the long outer margin serve to link these seemingly separate scenes, creating what can then be imagined as a single continuous scene. Notably, the trees and plants are shown in their entirety, but on the two pages in Iskandar's anthology they are not, with the lower portion of all but a few trunks and stems being cut off and seemingly hidden behind the large block of text. It is therefore not with this but with the second type of decoration used to fill the margins of the Chester Beatty Anthology that parallels with Iskandar's anthology are found.

Of the many different patterns included in the margins of CBL Per 294, some of the most striking are arabesque vines incorporating birds, usually with exceptionally long tails. Sometimes the birds are an actual part of the vine, which enters through the bird's tail feathers and emerges from its beak. But other times, as in Figure 1, the birds instead appear to create a second, upper layer of interlacement, their crossed, elongated tails and long arched necks creating a rhythmic, wave-like movement across the width/length of the margin equal to that created by the scrolling arabesque vine. Both types of pattern are first seen in the margins of Iskandar's London Anthology, on ff. 406a and 407b, and also in the upper band of decoration on f. 543b. Related to these are two other examples from the London Anthology (ff. 407a and 408a), where the long-tailed birds are partially hidden behind large blossoms or leaves. That on f. 407a is especially interesting, as it is closely paralleled by a pattern that appears on numerous pages of the Chester Beatty Anthology, though minus the birds (Figure 5). In the Iskandar version, large blue leaves with lobed edges are the most prominent element, again creating a wave-like pattern as they force the eye to travel across, down/up, and across the margin. The artist of CBL Per 294 has created a slightly more elongated leaf with a serrated edge, but it, too, as in the Iskandar version, is an actual part of the vine, not merely attached to it, and this and its prominence creates the same sense of movement as in the earlier version. The artist has, moreover, copied the earlier artist with regard to the types of smaller blossoms chosen to fill the surrounding ground.

Two of the most appealing margins in Iskandar's anthology appear on folios 405b and 408b. On the former, split palmettes are arranged to form an enclosed space that contains either a leonine face or a large lotus blossom.Footnote ³⁰ Between these medallion-like spaces are pairs of birds with long, crossed tails, whose bodies are twisted to create other, smaller spaces, which, though not totally enclosed, nevertheless resemble medallions or shields. On folio 408b, some of the enclosed medallion-like spaces are again formed by the careful arrangement of split-palmettes, but paired fish surprisingly form the contour of alternate medallions. Roughly similar types of pattern, in numerous variations, exist in the Chester Beatty Anthology, although the main repeated unit is not the fully enclosed, and thus tightly defined, space of the Iskandar margins, but rather a collection of many, often more freely floating, motifs. Nevertheless, in both, the repeated unit is either defined by and/or encloses one or more figures. For example, on two facing pages, ff. 37b (Figure 6) and 38a, there is a repeat pattern consisting of a seated, cross-legged winged angel with outstretched arms contained within a space loosely defined by the vine and animal heads of a surrounding waqwaq (a vine populated with human and animal heads). Similarly, on two other pages, the repeated unit consists of a waqwaq surrounding a single bird with outstretched wings on one (f. 31b) and surrounding paired birds with crossed tails on the other (f. 2a).

Figure 6. CBL Per 294, f. 37b.

Although examples of the waqwaq are known on Islamic metalwork and other media from at least the thirteenth century,Footnote ³¹ the earliest painted examples appear to be those in Iskandar's anthology, such as the one in the margins of f. 536a. Others, with animal heads only, appear elsewhere in the manuscript: in three so-called thumbpieces (on ff. 10b, 20a, and 28a) and in one of the central medallions on f. 543a, while on f. 405b a greatly abbreviated waqwaq emerges from the mouth of each leonine head. Like other motifs in the margins of Iskandar's anthology, such as paired birds with crossed tails and paired fish, the waqwaq appears in other Timurid manuscripts, but especially anthologies in the safina format.Footnote ³² Waqwaqs occur in the margins of several pages of the Chester Beatty Anthology, but most often serving only a secondary role, as in Figure 6.

Overly large blossoms on long, streamer-like stems wend their way through the margins of many pages of CBL Per 294 (Figure 7), with both the types of blossoms depicted and their precise arrangement within the margins closely mirroring the margins of f. 412a of Iskandar's anthology. Parallels with other margins within the earlier anthology are less certain. For example, the palmette-arabesque on f. 73a of the Chester Beatty Anthology is a close match for that on f. 420a of Iskandar's manuscript (with a slightly less close match on f. 411a), but palmette-arabesques are so prevalent in Islamic art that it is difficult to claim that the earlier anthology might have provided the model for the later one. Parallels with, and possible models for, many other margins in CBL Per 294 do not exist in the Bihbihani Anthology, Iskandar's anthology, or seemingly in any other known manuscript, and so these may be original creations of the Chester Beatty artist. These include cloud bands arranged in a zigzag pattern (e.g. f. 27a) and an array of arabesque patterns, some including figures such as prancing angels (e.g. ff. 41b-42a). It is, however, the myriad details included within the landscapes that are especially beguiling: benign displays of the many varied types of creatures included in the manuscript, such as harpies (f. 43a); pastoral scenes, such as that of figures herding rows of camels (ff. 60b-61a); somewhat amusing scenes, such as that of two large and angry roosters confronting one another in the upper margin of a folio (though while a hunter shoots a gazelle in the lower margin) (see Figure 4); and ferocious displays of animals, usually simurghs and dragons, attacking one another (e.g. f. 42b; see also the lower margin of Figure 3). Occasionally facing pages form pairs (though they are never precisely identical), but whether this was originally the norm has not yet been determined, many of the folios now appearing to be disordered.

Figure 7. CBL Per 294, f. 122a

The illustrations and illuminations of the Chester Beatty Anthology are in the usual Shiraz styles of the time, and even if these had been the extent of the manuscript's decoration, when new, it would have been a fine, if small, manuscript. However, before production of the manuscript actually began, decisions were clearly made to increase the visual appeal—and overall prestige—of the manuscript beyond that of most other illustrated and illuminated manuscripts of its time. Although only briefly mentioned above, the least of these decisions surely concerned the layout of the text, which is unusual. The obliquely positioned, upward- or downward-facing lines of script may impede easy reading of the text, but their arrangement is appealing and memorable precisely because it is so unusual. The two volumes of ‘Attar's poems, made for Shah Rukh, each consist exclusively of large folios of ‘Chinese’ paper. Even if the seven small folios included in CBL Per 294 are, as suggested, mere scraps, salvaged from the production of a larger, now-lost manuscript, the inclusion of such a luxurious and exotic product among the pages of an otherwise ‘ordinary’ manuscript was undoubtedly seen as a means of increasing the manuscript's overall prestige, and perhaps also—and more specifically—as a means of symbolically linking the small manuscript to royal production in Herat, the Timurid capital. But certainly the boldest and most innovative decision was that the margins of each and every page should be decorated with a vast array of scenes and patterns executed exclusively in gold, something unknown in any other contemporary surviving manuscript and something that definitely raised the manuscript above the level of ‘ordinary’. As intriguing as the mere existence of the decorated margins is the obvious response of so many of them to Iskandar's earlier, highly prestigious London Anthology. We do not know who the patron of the CBL Anthology was—nor do we know the exact date of its production—but the most obvious possible person is, of course, Iskandar's cousin, Ibrahim-Sultan ibn Shah Rukh, a patron of fine manuscripts himself, who also served as governor of Shiraz, from shortly after the end of the rebellious Iskandar's governorship in 1414 (ah 816) until his death in 1435 (ah 838).Footnote ³³