The hereti pattern

The hereti pattern

1 78 The flowing Herat style, in its most mature form, as seen in the ‘Emperor’s carpet’ in the Osterreichisches Museum fur angewandte Kunst, Vienna; 7 6 0 X 3 2 5 cm, 495,000 PK/m2 (24′ 1 i ” x io’ 8″, 320 PK/in.2).

The hereti pattern carpets

Persia as the home of oriental carpets has not always been confined to the frontiers of present-day Iran. Over the past 2,500 years, during which, at least, we know that the carpet designs of today have evolved, the Persian Empire has expanded and contracted over many parts of central and western Asia, and significant changes have occurred in the racial make-up of the population and of the rulers. As the historical maps on p. 343 show, the original Persian Empire (in the period in which the Pazyryk carpet was produced) stretched from the Mediterranean to the Hindu Kush, while centuries later Persia was for long under Turkoman and Mongol rule. In view of this, when we speak of ‘Persian’ designs, we should not think simply in terms of the boundaries of modern Iran. It is therefore quite natural for one of the most used and most widely spread of all Persian carpet designs to be named after Herat, today the second city of Afghanistan. Herat was founded, as Alexandria Areion, by Alexander the Great in 328 BC. It was already a major city on the Silk Route when Timur established his court at Samarkand in the fourteenth
century. Shah Rukh (ruled 1405-47), one of the most enlightened rulers of the Mongol Timurid dynasty, chose Herat as his capital and turned it into the most important centre of learning and culture in central Asia. The city was held briefly by the Shaibani Uzbeks, but was recovered in 1510 by the Safavid Shah Ismail and remained an important city of the Persian Empire until 1857. During the critical period of the development of Persian carpet design Herat was the capital of Khorassan province and was thus the metropolis for the whole of eastern Persia, which region was the source of many important carpets of the sixteenth and seventeenth centuries. Among these, particular attention is focussed on three types: carpets in the fully fledged Herati design; carpets in a free-flowing all-over floral
style, which seem to have been the predecessors of the first-mentioned group; and a related group in another, stiffer, all-over floral style, a group known as ‘vase carpets’.

This said, it must immediately be admitted that, despite the existence of hundreds of sixteenth- and seventeenth-century carpets of all three types, we simply do not know precisely where or when any of them were made – not even for certain whether they are all really of east Persian origin at all. As the place of origin of the vase
carpets Kerman is often suggested, for the floral ‘Herats’ Meshed is a possibility, and for the pieces in Herati pattern the villages of the Qainat in central Khorassan have a claim. But when it comes to specific pieces the experts offer widely different opinions, with discrepancies of more than a hundred years in dating and hundreds of kilometres in location. It would be beyond the scope of this book to present the various cases advanced by carpet experts. However, the reader should be aware of the historical background and of the stylistic associations of the Herati design, and it is hoped that the illustrations included here, taken together (and in conjunction with the vase designs on pp. 119 and 325-8) will contribute to an appreciation of the modern versions of Herati designs shown in this chapter.

The hereti pattern

162 The Herati pattern; this example, showing a ‘Kaimuri’ Indian carpet measuring 253 X 1 8 5 cm with 220,000 PK/m2 (8′ 3 ” x 6 ‘ i”, 142 PK/in.2) from
Khamariah (see pp. 99 and 305), is one of the latest developments in the history of a design that has spanned several centuries.

163-168 Tribal herati derivatives

These are illustrated in figs. 163-168, which, taken in order of appearance, reveal a backwards progression from the Turkoman gol a medallion or flower-motif to the Herati pattern. Any proposal that such a connection exists is anathema to the many passionate champions of the tribal rug, for there is an implied suggestion here that the Turkoman gol (cf. fig. 399), the ‘holy cow’ of the campaign for the recognition of the supremacy of nomad art, is no more than a debased derivative of a design from a Persian manufactory. To assert this might be going too far; but even if a connection exists we must also consider that when the Herati pattern first appeared it may have represented not an original idea by a skilled designer but a reflection of a design tradition already extant in
the Herat region, and that the tribal rugs of the area may embody a much more ancient inheritance of the same tradition. The tribal rugs from Khorassan illustrated in figs. 163-6 show various treatments of the Herati design. Fig. 167 shows how the zigzag edges of the diamond shape seen in fig. 166 can be opened up to produce a leaf shape; the principal features of the Herati design are thus clearly revealed. Fig. 168 shows a floral manufactured style from the village of Dorukhsh, in the heart of the Beluch area similar to that of Birjand. If A. C. Edwards is right
in believing the Herati pattern to have originated in the Qain-Dorukhsh-Birjand region, local tribesmen would have plenty of opportunity to transform the Beluch version (fig. 167) into that shown in fig. 166; the influence may also have been the other way round, but from time immemorial it has been the practice of nomads and peasants to introduce new designs by the simple expedient of copying what is selling best in the local manufactories. The conversion of the fifteenth-century geometric tribal pattern shown in fig. 398(g) into the floral version seen in fig. 398(h) shows how this kind of development can occur, and this example may indeed be one of the key pieces of surviving evidence concerning the evolution ol the Herati pattern.

The hereti pattern

163 The standard Sarukh Turkoman design: a got with zigzag edges alternating with a more rounded gol, 201 x 129 cm, 320,000 PK/m2 (6′ 7 ” X 4 ‘ 3″, 206 PK/in.2).
164 The characteristic design of the Yamut Turkomans; there is usually only one type of gol, repeated in diagonal lines; 193 x 122 cm, 196,000 PK/m2 ( 6 ‘ 4 ” x 4 ‘ o ” , 126 PK/in.2).

169- 173 Sixteenth century- precursors

There still exist many carpets of early times that reveal stylistic features which could well have influenced the development of the Herati pattern. The Turkoman connection strikes the observer who sees the geometric lozenge- or diamond-shaped frame as the outstanding feature of the design, but one may also ask how that frame originated. If one takes as the centre of the design not the flower which appears inside each diamond but the one which lies between the diamonds, then a quite different picture presents itself. later in the century, Islamic designers were putting motifs together in a way which, to say the least, strongly foreshadows the Herati pattern. The Persian rug shown in fig. 171 is described by C. G. Ellis as mid-sixteenth-century Kerman; it shows delicately conceived palmette and lancet-leaf motifs rather clumsily put together, but the underlying structure of the design is very interesting because it reveals a basic idea, much developed in both the Herat floral and the vase carpets, which can also be seen as an obvious potential source for the Herati pattern. This piece clearly belongs to the group known as vase carpets (see also fig. 184 and index), but is less geometric in style than others of its kind and represents a link between the vase carpets and the Herat floral pieces (cf figs. 738 and 776). The opening up of the layout to produce a structure dependent on flowing flower-stalks (called ‘islimis’) runs into difficulties in the upper portion of the rug, but provided later weavers with the key to interlinking all the elements of the design to a much greater degree than is possible with the layout of the Cairo carpet.
The rather unusual modern Bijar rug shown in fig. 200 has a Herati pattern which also clearly shows a layout based on the same principles. The underlying structure of fig. 170 can be elaborated in many different ways. The two versions shown here in figs. 172 and 173 may easily have been used to lead to the Herati pattern. There are no ‘loose ends’ in this kind of layout: the interdependence of all parts of the design will have been a valuable feature for designers producing carpets for use in mosques because it reflects an important aspect of the traditional Islamic view of the cosmos.
The most convincing expose of the background material of this era is given by Werner Grote-Hasenbalg in his book Der Orientteppich, seine Geschichte und seine Kultur, one of the great classic carpet books; published nearly sixty years ago, it is still worth reading. Although some of that author’s detailed observations have been superseded by the discoveries of later scholars, his principal line of thought is sound and is summarized below. Grote-Hasenbalg begins at the point where the flourishing edifice of Persian art was utterly destroyed by the Mongols in the
thirteenth century.
In a series of campaigns beginning in 1219, Jinghis Khan, a Mongolian chieftain, subjugated northern China, East Turkestan and Persia through to the Caucasus and beyond. Baghdad, the capital of the Seljuq Turks, fell to the Mongol Empire in 1258. Fifty years later both Persia and Turkey achieved independence, but were overrun again in the 1380s and 1390s by Timur (Tamerlane), whose ferocious Mongol and Turkoman hordes conquered almost the whole of Asia, including northern India, Persia, the Caucasus and Turkey. In 1369 he established one of the most resplendent courts of all times in Samarkand – already a great city, being the pivotal point of the Silk Route, the principal artery of trade between East and West.
Timur attached great importance to the arts and brought many Chinese artists to work in his new empire. Bokhara and Herat, both of them towns on the Silk Route, also benefited from the splendour of the Mongol court, Herat, in particular, becoming under Timur’s successors an artistic centre of the first rank and the source of regeneration for Persian art. The effect of this Mongol influence was incalculable. Whereas in western Persia the stricter older style of carpet design survived, in the east a new, Chinese-influenced style began to emerge. Chinese motifs, which were later to assume such great importance, entered the field of carpet design.

The hereti pattern

1 6 5 A Beshir Turkoman rug from Afghanistan; 162 x 1 2 3 cm, 200,000 PK/m2
(5′ 4″ x 4′ 0″, 129 PK/in.2). For a more fully developed Herati design woven by the Beshirs of Turkestan see fig. 236; the version shown here – clearly a derivative of the Herati design – is the more common of the two.
166 One of the many variants of the repeatingdiamond idea woven by the Beluch tribes of Khorassan; 1 9 0 x 9 0 cm, 195,000 PK/m2 ( 6 ‘ 3 ” X 2 ‘ i i ” , 126 PK/in.2).
167 A Beluch rug, of a type very commonly found in the area between Meshed and Herat; 2 1 8 X 128 cm, 115,000 PK/m2 (7′ 5″ x 4′ 2″, 74 PK/in.2).
168 Dorukhsh: floral manufactured style; 307,000 PK/m2 (198 PK/in.2).

174- I79 The stylistic influence of china

The cloud-band, the phoenix, the dragon and the palmette are all examples of motifs that first came to Persia with the Mongols and became absorbed into the carpet art of the fifteenth century; their significance and symbolism are set out in the captions to these illustrations. As has been pointed out in the Introduction, the palmette was a widely used decorative motif in Western Asia in ancient times.
Jessica Rawson (in Chinese ornament: the lotus and the dragon, London 1984) has demonstrated how the palmette originated in Egypt and Greece, became a key element of architectural dccoration in Central Asia after the latter’s conquest by Alexander the Great, and was carried by the spread of Buddhism from there to China.
Its use in the Islamic art of Asia remained very limited until the Mongol invasions. It may thus be said to have been re-introduced into carpet art by the Mongol/Chinese influence. Jessica Rawson documents its wide use in Persian decorative art of the fifteenth century. By the early sixteenth century the palmette was fully established as a principal element of floral patterns.
According to Grote-Hasenbalg, there is an even more important feature of the Chinese influence than the introduction of individual motifs, and that is the impulse for greater freedom of expression and a more flowing, so to speak picturesque, style.
Western Persia, he asserts, being furthest from the Chinese influence, preserved the strict Islamic style longest, while the new free style of design developed in Herat and eastern Persia. Figs. 178 and 179 illustrate the difference. The second of these, a north-west Persian medallion carpet, is one of several such pieces about which
experts differ with regard to their dates. This carpet – illustrated by Grote-Hasenbalg (his fig. 44) – is described by him as ‘around 1500’, Dimand, who shows a similar carpet (his fig. 59), suggests an even earlier date. Erdmann’s example (his fig. 99) is simply described as sixteenth century; he also illustrates a carpet (his fig. 79) with a less well developed design, which he attributes to the first half of the sixteenth century. Another piece in the McMullan Collection was attributed by Ettinghausen (Textile Museum Journal, June 1970) to Tabriz, second quarter of the sixteenth century. All these have a stiffness and angularity which Grote-Hasenbalg
sees as a residue of the strict style of the fifteenth century. A point that he does not mention is that they all also use a geometric variant of the Herati pattern in the ground, suggesting links between it and the arabesques of Islamic art of an earlier period. Dimand (in his figs. 62-4) shows how the Tabriz designers gradually overcame
this stiffness and introduced the flowing style seen in some Herat miniatures (figs. 617, 751). It is no longer possible to photograph Grote-Hasebalg’s example of the flowing Herat style of the first half of the sixteenth century, since the carpet in question was in the Leipzig Museum collection and is now lost. The famous allover
carpet in the collection of the Osterreichisches Museum fur angewandte Kunst in Vienna is illustrated here in its stead, since it is stylistically very close to the lost Leipzig piece. Its origin may be eastern or central Persia; stylistically it shows the Herat floral style at its most elaborate and fully developed.

The hereti pattern

185 Detail of a Herat floral carpet, full size 265 X 180 cm (8′ 8″ X 5′ 10″), formerly in the Museum fur Islamische Kunst, Berlin (now lost), dated by Grote-asenbalg to the first half of the sixteenth century, and by Erdmann to c. 1600.
186 Detail of a floral Herat carpet, 2 2 4 x 2 0 7 cm (7′ 4″ X 6′ 9″); this piece also has the characteristics of the sixteenth-century style. Victoria and Albert Museum, London.
187 A late sixteenth-century floral Herat carpet, 380 X 296 cm, 420,000 PK/m2
( 1 2 ‘ 5″ X 9’ 8″, 271 PK/in.2). Carpet Museum of Iran, Tehran.

We have no firm datable evidence of the art of western Persia to confirm Grote-Hasenbalg’s contention, and Erdmann has shown that it was not until about a hundred years after the Mongol conquests that the geometric style gave way to the floral in Herat. However, it does seem that Herat led the way in carpet design as an extension of the very important school of miniature painting and bookbinding and related arts which developed there in the fifteenth century. As the Mongol power waned (Timur’s adoption of the Islamic religion of his conquered territories itself
began the process of the ‘Persianization’ of the Mongol Empire) the cultural independence of the Ottoman Turkomans in Anatolia and of the Persians began to assert itself, until, at the end of the fifteenth century, the complete political breakdown ofTimur’s inheritance occurred with the establishment of: the Safavid Empire in Persia, between 1498 and 1510; the Ottoman Empire in Turkey and the rest of Asia west of Persia – a gradual process that lasted from Timur’s death in 1405 until the accession of Suleiman the Magnificent in 1520; and the Moghul Empire –
politically Mongol and culturally Persianized in India and parts of Afghanistan, between 1525 and 1530.
Although no actual carpets from fifteenth-century Persia have been preserved, it is clear from the Herat miniatures that their designs were markedly geometric, which is sometimes claimed to be a reflection perhaps of a strict interpretation of the Islamic demand for the stylization of all living forms in art (see pp. 156-9, 270).
The floral style which developed at the end of the century marked as abrupt a change as the difference between the Gothic and the Romanesque in Western art.
The change was itself perhaps a reflection of, perhaps a contributing factor to, the resurgence of Persian political independence. In Grote-Hasenbalg’s view, the Herat free style, as exemplified in fig. r 78, was the foundation of the great flowering of the carpet art which spread throughout Persia in the sixteenth century, reaching its peak in the reigns of Shah Tahmasp (1524 76) and Shah Abbas I (1587-1629).
During this Golden Age, once the Safavid Emperors had established their power and transferred the capital, first to Ardebil, then to Tabriz, Qazvin and finally to Isfahan, it was the carpet manufactories of Tabriz, Isfahan and Kashan that were called upon to produce the monarch’s sumptuous gifts to foreign rulers and the showpieces for mosques and palaces. South-eastern and eastern Persia faded from the limelight, but continued to produce large quantities of carpets, presumably mainly for local use. They came to be called Herat carpets because it was here, in the provincial capital, that they came to the attention of the outside world. But how many, if any at all, were made in Herat is open to question. Firm evidence is practically non-existent: interest centres on the general way the designs of the area developed.
A major problem in tracing the history of a design is that, when faced with a simple version and a more complex one, some experts will call the simple version a later debased form of the more refined one, while others will call it the original from which the more complex one was developed. The chronology of the floral Herat carpets and of the vase designs (which are often attributed to Kerman) is thus much in dispute. Once again, Grote-Hasenbalg’s conclusions, with some modifications, may be cited (see notes below).

180-182 The indian dimension

Quite early in the sixteenth century eastern Persia saw the beginning of a development in which the fine naturalistic balance and flowing proportions of the original Herat style gave way to over-massive and heavily stylized flower and leaf forms joined by long meandering stalks (islimis). Many influences were at work and the
designs developed along several different lines. An important factor at this period, and one which is often neglected, was the cultural exchange between Persia and India. The last important successor ofTimur was Babur, who inherited the province of Fergana, three times conquered and lost Samarkand, established himself securely
as the ruler of Kabul and from there extended his power to found the Moghul Empire in India at the beginning of the sixteenth century.
At times there was conflict between the Moghul and Safavid empires, the worsthit town being Kandahar in south-eastern Khorassan, which changed hands several times between 1595 and 1649. At other times the two powers were allied in common cause against other forces, but above all there was a cultural interchange derived from the common heritage of the Mongol rule, the Islamic religion and a common court language, Persian. There was thus as much intercourse in the period 1500-1700 between Herat, Kerman, Lahore and Delhi (via Kandahar and the Khyber Pass) as between Herat and Isfahan (which are separated by an impenetrable desert). Carpet historians draw attention to this when writing of the development of weaving in India but usually pay scant attention to the reverse influence
in Persia. But the ‘Indo-Isfahans’, as the Herat-style carpets made south of the Khyber are called, represent an indivisible part of the development of the sixteenth-century east Persian carpet and as such had their own role to play in the crystallization of the separate designs that came to be established in the region.
Figs. 180-90 illustrate a few of the important strains of the development, which, as Grote-Hasenbalg demonstrates, led not only to the Herati pattern, but, by 1700, to the Mina Khani and Joshaqan styles as well (see figs. 754-7, 599), and which, as is amply illustrated in many parts of this book, has remained a dominant influence on the whole of Persian carpet design ever since.
Map of northern India and eastern Persia in the late sixteenth century. For most of the century there was a steady interchange of trade and ideas between Delhi, Kandahar and Herat.

183-191 The development of the herati design

Certain stylistic features not always noted by scholars seem worthy of comment. One is the relationship between the vase carpets and the floral Herat carpets. In his excellent notes on the vase design (Textile Museum Journal, December 1968) Charles Grant Ellis does not mention this point, although it was noted by Grote-Hasenbalg some fifty years earlier. A. C. Edwards points out that the floral Herat pieces could not have been woven without graph-paper loom-drawings and must therefore have been the products of a town manufactory, whereas Grote-Hasenbalg declares the vase carpets with their strict, symmetrical style to be the ‘most perfect examples of Persian folk-art’. Were they perhaps a development one from the other the semi-geometric vase designs being peasant derivatives from the floral style? Or did they develop separately from a common source? Could the geometric vase designs actually pre-date the floral ones? An interesting feature linking the two groups is the way the Chinese cloud-band motif is handled. Is it accident that the vase in fig. 183 has a form so reminiscent of the cloud-bands offigs. 186 and 187? Either way up, the cloud-band can be converted into a vase at the flick of the designer’s pen or by very elementary mis-reading of the loom-drawing by the weaver. Compare also the similarity between the cloud-bands in figs. 186 and the important lily motif at the very centre which occurs also in the diagonal axes of this and most other vase carpets.

The hereti pattern

190 Detail of a Herati carpet, full size 990 x 358 cm (32′ 6″ x 1 i ‘ 9″), known to have been made not later than 1640 (from which date its presence in the
palace of the Duke of Braganza is recorded). Thyssen-Bornemisza Collection, Lugano.

Another point rarely touched on by scholars is the importance of the introduction of leaf motifs some lime in the course of the development of both the vase and the Herat floral types. This must have been a decisive innovation seized upon by designers as a means of pushing the development along lines which might have been difficult with the purely floral and animal motifs of the early classical carpets. How or when the elongated leaf motifs, usually called lancet leaves, arose is not known, owing to the uncertain chronology of all of these carpets. These motifs are absent in figs. 739, 186 and 270, but in one of the lines of development, clearly evident in figs. 188-90, they came to occupy a dominant position in the overall design and they were the key to the conversion of the floral Herat carpets into the Herati pattern.
Whatever the origin of the Herati design, the form in which it is used in modern production is unproblematical: a flower is surrounded by a diamond-shaped framework of stalks with a large leaf set parallel to each of the four sides of the diamond. In many parts of Persia the leaf is regarded as a fish and the design is called mahi (meaning ‘fish’) or the ‘fish-in-the-pond’ design. However, this description seems to be a case of weavers – or dealers reading their own fanciful interpretation into a new design adopted from a foreign tradition, in the same way as the Indians refer to ‘parrot’ Serabends, discussed above in the context of the boteh motif.
The spread of the Herati design outside Khorassan (assuming it did originate there) is better known. There is a universally accepted but, here too, seemingly not very well documented tradition that in about 1730 or 1740 Nadir Shah transferred large numbers of weavers from Khorassan to central Persia – to the Ferahan region of present-day Arak, and the area across to Joshaqan, north of Isfahan. In the process he transferred the east Persian design concepts, paving the way for the development of the Mina Khani and Joshaqan designs (figs. 757, 599) as well as stimulating the spread of the Herati design. There remains something of a mystery about the name Ferahan (often misspelt Feraghan). Today, Upper and Lower Ferahan are administrative districts north-east of Arak, adjacent to Tafrish.
Although the famous antique Ferahan design may have originated in this area, the bulk of the production in the late nineteenth and early twentieth centuries seems to have come from a much wider area, with special emphasis on the region west of Arak towards Malayir. Whatever the explanation of this confusing situation, the Malayir-Ferahan carpets quickly became as famous as the Herati originals, and ‘Ferahan design’ is just as commonly used as a description as is ‘Herati design’.
This has become the typical traditional design in several parts of Persia, but it is in fact made almost everywhere and is found in both rectilinear and curvilinear forms, from simple peasant rugs to fine manufactured carpets. Once fully established, the Herati pattern was used as an independent design with the same degree of variation from one locality to another as is found with the boteh motif The eighteenth-century carpet from Herat (in the Victoria and Albert Museum textile study collection) shown in fig. 191 captures well the essential flavour of the design.
Here, the motifs arc drawn rather larger than is normal for the design, but what is particularly fascinating about this version is the blending of floral and geometric styles which it incorporates. The weaver has introduced a number of curves into the execution of the pattern, but without obliterating its semi-geometric peasant
flavour. This is a feature which will be noticed in many of the best versions of this design, and which no doubt contributes much to its universal appeal.

192, 193 Senjabi, bajar

The Herati design is woven throughout western Iran, from Tabriz right down to Khorramabad, each area having its own characteristic version. In Kurdish tribal rugs it appears in many guises, two of which — Senjabi and Bijar are illustrated here. Although the simple rustic execution of the motifs is common to both, the two pieces are markedly different in almost every other respect, reflecting the wide variety of types found within the Kurdish weaving area. Senjabi is the name of one of the southernmost Kurdish tribes, but it is used rather loosely in the carpet trade to describe a range of tribal or village types of the Kermanshah area. The warmth of the colouring, the light brownish shade of the red and a single-wefted construction on woollen warps are typical Senjabi features. The Bijar rug, on the other hand, has features which might remind one of the Kakaberu tribe — a coarse but tight and
heavy double-wefted structure, the kelleyi format and a rather wild interpretation of traditional Bijar motifs (for a finer version of the border see fig. 103). However, the ground of this rug is light red and the whole appearance is bright and airy, which completely excludes Kakaberu as the origin. (Kakaberu rugs are always dark; many are predominantly dark blue with little use of red at all, but where any quantity of red is used it is a dark shade, even mahogany-coloured). Most carpet experts would therefore classify the piece as a Bijar, with the rider that it is not a true Bijar in the normal sense but a coarse peasant product, probably from some village in the Bijar area.

The hereti pattern

192 Senjabi 1 9 3 X 1 2 6 cm, 90,000 TK/m2 (6′ 4 ” X 4 ‘ 1 ” , 58 TK/in.2 ).
193 Bijar 248 X 1 4 1 cm, 85,000 TK/m2 (8′ 1 ” X 4 ‘ 7″, 5 5 TK/in.2 ).
194 Hashtrud 186 X I I I cm, 106,000 TK/m2 (6′ 1 ” X 3 ‘ 8″, 68 TK/in.2 ).

The hereti pattern

195 Bijar (Kurdi) 215 x 140 cm

194 Hashtrud

Another simple village form of the Herati design, from the mountainous Hashtrud region, on the northern edge of the Persian Kurdish area, reveals stylistic influence from the neighbouring Azerbaijan region: the leaf or ‘fish’ motifs have here become completely displaced and out of order. This is a not uncommon feature of Kurdish
Herati pieces – one often encounters, for example, an all-over distribution of Herati leaves, but only a hint of the rest of the Herati pattern; and in the piece shown in fig. 532 the Herati pattern of the central medallion is quite dismembered. The Hashtrud (literally, ‘eight rivers’) region lies east of Lake Urmia. It has only a small output of rugs, which are sold in the Tabriz bazaar; they are also described as Shah Savan or Amroullah.


One of the great versions of the Herati design is that used in Bijar, the market for the finest and best Kurdish carpets. Whereas the previous illustrations have shown rugs made entirely of repeating Herati motifs, in the Bi jar carpet the Herati design provides a background dominated by a large floral medallion. This combination of a basically floral centre on an essentially geometric ground is characteristic of many Bi jar designs. Bijar is one of the fine weaves of Persia, so that technically there would be no problem at all in producing elaborate floral designs; however, the strongly geometric style of the traditional Kurdish tribal designs always makes itself felt in Bijar carpets. Note, by the way, the broad light-coloured outline to the medallion. This is a characteristic feature which is a useful clue to recognition of Bijar designs. Most of the production features a red ground, although blue and, occasionally, cream are also found. The sizes made are: pushti, zarcherek, zaronim, dozar, some kelleyis, and carpets of all sizes (though carpets under 8 m2 – approx. 85 ft2 – are rather rare). It is interesting that the centre part of the medallion in fig. 195 is used as an independent motif in the repeating-medallion Bijar piece shown in fig. 761.
From the point of view of carpet design, there is little connection with the Afshars of the Kerman region: the Bijar Afshars produce carpets fully within the Kurdish Bijar tradition, although they are usually just a little more flowery, especially in their floral medallions, an example of which is shown (fig. 197). In the Afshar carpet (fig. 196), however, we see a third use to which the Herati design is put: this remains a Herati all-over, but by using three different ground shades separated by an outline of dark blue the weaver has produced a medallion-and-corner effect. This idea is used in many places – see fig. 230 (Birjand), fig. 223 (Saruq), fig. 227 (Ardebil) – but the Bijar version is unmistakable, mainly because of its more primitive execution. Note that the edges
of the medallion are not parallel to the corners, nor are they symmetrical in themselves.
The Herati pattern itself is also distorted, and another noteworthy feature is the extraordinary shapes which the medallion can adopt. The three examples illustrated (figs. 198-200) show some of these distortions which are the hallmarks of village weaving.
At this point it is worth quoting in full the paragraph Grote-Hasenbalg (1922) devotes (p. 62) to the collapse of the great Court Manufactories of the Safavid Empire, even before the Afghan invasion of 1722: The decline of Persian art is usually explained entirely by the argument that the artisans no longer understood the artistic concepts of earlier times. I would say this was the least important reason for the decay. 1 have already drawn attention to the dangers that attend an art which is founded only upon high technical skills and precious materials and is dependent on the most extravagant wealth.
That Persian art followed this path at all no doubt has to do with the form of despotism in the Orient which drove all branches of art to extremes of showiness and to a refinement of technique down to the smallest detail, which were intended to proclaim the ruler’s splendour and his power to the whole world . . . But there came a point where increased living standards coincided with a decline in the ruler’s power and influence: there was no increased wealth to support the standards demanded and the pressure for simplification became overwhelming . . .
Very often the person commissioning the work simply could not afford the desired degree of refinement. In countries like those of western Europe, where free art has always been more highly regarded than craftsmanship, the situation was different.
The struggles to liberate the Netherlands (from Spain), which must surely have had a serious impact on the Dutch economy, were not yet at an end when the seventeenth-century school of Dutch painting led by Rembrandt reached its
highest peak of achievement. But for these works the only materials needed were paint and canvas – and the genius of the artist. The collapse of the economic substructure destroyed the art of the great city manufactories. But the folk art of the mountain villages remained untouched by all this: it neither partook of the artistic opulence of the Golden Age, nor did it suffer by the latter’s decline. It simply retained its identity and preserved its originality, the touchstone of its cultural significance.

The hereti pattern

196 Afshar Bijar 348 x 249 cm

The hereti pattern

198 200 Herati motifs in Bijar carpets: three examples showing typical irregularities in execution; although of fine weave, Bijar’s output is produced in villages, mainly without the controlled design work available in town manufactories.

Gogarjin, kolyai

We come now to two further Kurdish examples of the Herati pattern. Gogarjin is a village very close to Bijar, weaving a distinctive form of Bijar rug in the style illustrated; this, like all Bijars, is double-wefted. The Kolyai tribe, whose goods are marketed in Sonqur in the south of Kurdistan, weave their own version of the Bijar
design, and, like all Kolyai tribal rugs, the example illustrated is single-wefted. The broad camel-coloured outer edge to the medallion is a typical Bijar trait, but the general coarseness of the rest of the rug betrays its more humble origin.


Zenjan is a small, neat market town roughly halfway along the road from Tehran to Tabriz. Here three different carpet-weaving styles meet: the coarse double-wefted kelleyis and rugs of southern Azerbaijan (e.g. fig. 134), the fine double-wefted rugs of the Bijar area, and the coarse-to-medium single-wefted goods of the northern Hamadan region. The name Zenjan is used, confusingly, to describe some of the products of all three types. The single-wefted type is illustrated in fig. 580. The fine double-wefted category, made to the west and south-west of Zenjan, may be subdivided again into three types, two of which have distinctive names: Qoltuq (see fig. 581) and Bidgeneh (see fig. 582). There are, however, two noticeable features by which they can be distinguished. The first is a stiff gawkiness in the design: compare, for instance, the grace and elegance of the border of the Bijar in fig. 195 with the nondescript feebleness of the star-spangled lollipops in the piece shown here. Or again the sense of flowing movement in the Bijar medallion with the uninspiring drynessof the Zenjan one. The feature which always strikes one about the Zenjan ‘Bijars’ is that they look like Bijars that have not come out right, that have somehow failed to reach the required standard. The second distinctive feature is a corresponding stiffness in the weave: the structure unlike that ofBijar-is not well balanced, but over-heavy in the warp and weft and underweight in the pile; the back of the carpet is thus excessively stiff, which, of course, in the long run impairs the wearing quality.
The above comment may seem to read like an outright condemnation of the Zenjan Bijar, but that would be unfair. In the world of carpets all things are relative and a good Zenjan is always better than a bad Bijar. Also many Zenjans
have the advantage of being much cheaper – only half or two-thirds the price of a Bijar of equal fineness – which can make them an attractive buy, although quite 203 ZonJan extravagant prices are sometimes quoted for Zenjans of unusual fineness.

The hereti pattern

201 Gogarjin 162 X 105 cm


202 Kolyai


Mention has already been made of the special distinctiveness of the rugs made in the Kurdish capital of Senneh. This also applies to their use of the Herati design, as illustrated here. This rug shows the standard interpretation of the Herati motif which is used for in-filling in many Senneh designs; this particular version, with its strange dismemberment of the pattern, is not generally used elsewhere in Persia.
Note how the diamond has become a hexagon and the leaves merely rectangular shapes. All the motifs, moreover, are set out in simple vertical chains (with no feeling of the horizontal and diagonal interdependence which is so strong in most Herati patterns). For general comments on Senneh, cf. fig. 130. Note that the size of this rug, the standard Persian dozar size, is rare in Senneh; Senneh dozars are usually appreciably larger.


We now come to one of the best known of all nineteenth-century carpet designs, the Ferahan. It is no longer produced, but in its heyday it had a great reputation for quality combined with elegance and restraint. Carpets of this type were made in large numbers and many examples can be found in museums, stately homes and private collections, especially in England, where they acquired in the late nineteenth century the sort of popularity that is enjoyed, for example, by the Mir Serabend in German-speaking countries in our own times – and for the same reasons of tasteful restraint and economy of design and colour, durable construction with very good wool and suitability for use in a wide range of decorative situations. The old classical Ferahan was discovered by the European traders who began buying up Persian carpets for export from 1875 onwards. The type had two distinguishing
features apart from its design: it always featured madder red (i.e. a shade on the browny-rose side of red), either as the ground colour or as the dominant subsidiary tone on the rarer cream or blue grounds; and it was usually in the kelleyi format, i.e. rather long in relation to its width. The supply of old kelleyis did not last for
very long and towards the end of the nineteenth century, at the same time as importers based at Sultanabad (Arak) were beginning to develop the manufactured Saruq, they also began to organize the village production of Ferahans over a wide area of the Arak region. These goods were, of course, in more normal European or North American dimensions. For background notes on the Ferahan’s origin see the first part of this section. Note that many other carpets were also called Ferahan; these were made in designs and in a fine double-wefted construction which, to
avoid confusion, would be better described as Saruq. This situation has no doubt resulted from the extreme geographical vagueness of the word Ferahan, which is used very loosely to mean ‘Arak area’ in general. The Malayir-Ferahan of the kind depicted in fig. 205 was always single-wefted; the weave was rather like that of present-day Nenej or Jokar (see figs. 760, 2 1 1 ) , but the pile was always clipped very close to give greater clarity and brilliance to the design. The style of the doublewefted Saruq/Ferahans is illustrated in figs. 603, 604 and 784.
Today, the Ferahan area proper, centred around the village of Farmahin, has a significant production of single-wefted rugs and runners in the Herati design. These come within the general category of Hamadan rugs, having similarities with both their neighbours in the Hamadan region, that is, with Borchalu to the west and with Tafrish and Rudbar, which are to the north and east. However, the new Ferahan Herati design, as seen in fig. 206, is very distinctive and quite unlike the red mahi version found in so many Hamadan villages (see figs. 210, 211). To begin with, it is mostly found on a dark-blue ground (sometimes also on a muddy dark red), often with rather sombre secondary colours which may tend to make the overall effect somewhat gloomy; secondly, the motifs are rather larger than those used in other Hamadan villages; and thirdly, the Ferahan Heratis are almost all without medallion. There are several variants, some more angular than others; most sizes are made, the commonest being: mossul (200 X 100 cm; 6′ 6″ X 3′ 3″), dozar (200 X 130 cm; 6′ 6″ X 4′ 3″) and runners in sizes such as 300-350 X 100 cm (10’0″— 11′ 6″ X 3′ 3″) and longer.

The hereti pattern

203 ZonJan

The hereti pattern

204 Senneh


A stylistic neighbour of the Ferahan Heratis is illustrated here. It has the same darkblue background and the angularity of design commonly found in Ferahan, but there are subtle differences which a trained eye will notice. The overall colouring lacks the luscious warmth of the Ferahan area: the red, for example, is too hard. The
wool is also a little harsher to the touch and the weave is not so fine. The origin of this example is Kharaghan, the name of a group of villages to the north-west of the Ferahan area.


Borehalu is the name ofa tribe centred on the villages of Khumajin and Kumbazan, a remote area east of Hamadan and to the west of Ferahan. This tribe is well known for a particularly successful version of the American Saruq design (see fig. 639), but a large part of their output is in the standard Hamadan village form derived from the nineteenth-century ‘Ferahans’ (compare fig. 205). Experienced dealers can recognize Borehalu Herati rugs fairly easily from the neat, fine weave and from certain clues in the colourings; but for most people the principal distinguishing
feature is the shape of the central medallion. Most sizes up to 4 m2 (43 ft2) are made, but especially dozars and zaronims. Most Herati Borchalus are on red grounds, but blue and some creams are also found.

The hereti pattern

205 An old Ferahan carpet in the entrance hall of Hevcr Castle, Kent.

The hereti pattern

206 Ferahan 195 X 122 cm


Next to Ferahan to the north is Rudbar whose production belongs to the Tafrish group (see fig. 555), but has its own distinctive version of the Herati design: quite like the Ferahan but more floral, and above all lighter in colour, with many cream grounds, some red and very few blue. A particular distinguishing feature of the Rudbar weave is the frequent use of pink wefts. Another, less easily recognized, feature is the soft, spongy feel to the wool frequently employed there. The lighter colours and the softer wool both link Rudbar with the Shah Savan
tribe ofSaveh (cf p. 237 and map, p. 346). A wide range of rug sizes is made.

Husseinabad, jokar

The Herati design is used in a large number of villages around Hamadan (and generally known locally as the mahi design), especially in a cluster of places to the south-east on the road to Malayir. These villages all have distinguishing features of weave and colouring which a trained eye can recognize. A Jokar piece is fine and neat but a Borehalu is finer and neater one is reminded of Hugh Johnson’s description of the individual wines of the Beaujolais region: ‘Brouilly is grapy and rich, but Cote de Brouilly is grapier and richer’. Most of the mahi rugs are sold simply as Hamadan, but the layman can tell some of the different types from differences in the design. Note that in the vast majority of the production a red ground is used; blue grounds are rare, and cream grounds even rarer, though both do exist. The biggest production comes from Husseinabad, and that name is often used to describe rugs in the mahi design made throughout the area. Husseinabad goods vary in quality from medium to very good. Fig. 210 illustrates one of the simpler weaves (better pieces will have 100,000-115,000 knots/m2; 65-75 knots/ in.2). Apart from Everu (see fig. 215), it is the only village with a significant output in the all-over Herati pattern, without a medallion (as illustrated here), but it also produces goods with its own rather small distinctive medallion (and corners). All sizes are made, from pushtis to large carpets, including narrow runners (60-80 cm wide; 2′ o”-2′ 9″) in all lengths.
Jokar also makes a range of sizes, especially large dozars (about m2 or 40 ft2). The distinctive medallion is similar to the one used in Husseinabad, but larger. The weave, too, is similar to that of Husseinabad perhaps even neater – but not as fine as in the better-grade Husseinabads. Jokar has, in effect, the same structure as Nenej, with very good wool and a thicker pile than Husseinabad. To the north of the group is the main source of the cheap, small Herati rugs, Gombad.
Gombads are mostly rather bright and crude, but above-average pieces are also to be found. Among the other sources of Husseinabad-type mahis are Ezanderian, which makes coarse cheap carpets (which are often wrongly sold as Husseinabads), and Ganjtepe, which makes fine, expensive carpets.

The hereti pattern

208 Borehalu 201 X 139 cm, 209 Rudbar 145 X 107 cm

The hereti pattern

210 Husseinabad Ezanderian area; detail of a carpet, full size 300 X 200 cm, 211 Jokar 222 X 137 cm


South of Jokar and stretching as far as Borujird (see Hamadan area map, p. 346) there is a largish area producing single-wefted rugs of varying quality in a wide range of designs. The production may be as clumsy as the piece shown in fig. 655, or as neat and finely executed as the Herati rug shown here. This piece shows the stylistic
influence of Jokar, but has the predominant dark blue which is characteristic of the Malayir and Borujird region (and also of the Nehavend and Tuisarkan group to the west, where Herati designs are normally not woven – see p. 259). The weave of this area is clearly distinguishable from that of the Jokar/Husseinabad group
owing to the use of medium-blue wefts, against which the white warp-strings show up strongly as clear white specks all over the back of the rug. Apart from Alamdar, with its own characteristic design (see fig. 217) the names of the individual villages of the area are not used and the rugs are sold simply as Borujird or Malayir.


Parallel with the unmistakable production in the American Saruq design (see fig. 638), the Armenian village of Lilihan also has a substantial output in the Herati pattern, as illustrated here. As is explained on p. 282, this is a single-wefted region, lying to the south of the Hamadan village area. The weave is medium-fine, with a
normally very sound structure and good yarn. The colours are regrettably rather variable; not only are some of the subsidiary shades very crude (greens and orange, for example), but also the dark blue, which is the most common ground shade in Hcrati-pattern Lilihans, can have a very dead appearance, with grey, black or mauvy streaks. The red shades are usually sound, however, and the open layout of the Herati motifs creates a very appealing effect if the blue is properly dyed. All rug and carpet sizes are found, including some unusual dimensions like 200 X 170 cm. Enjilas products have always been among the very best of all Hamadan rugs. In the past their workmanship, design, colours and materials were all superlative, and this day remain far above the local average. The price, of course, reflects this:even a good Husseinabad may be only a quarter of the price of a fine old Enjilas. Darklue
grounds are common; red is also found. The illustration shows a piece which may be about forty years old; the natural madder red of the ground has mellowed to a superb warm shade with a tinge of rose. The Herati design – wherever it is made is almost always successful; a true Enjilas shows us just how much better perfection is than mere success. A considerable range of sizes exists, but dozars are the most common; they are almost always without medallion, but include the small corner motifs of the rug illustrated.
The next village tojilas is Everu, which is an important centre for massproduced rugs in the Enjilas style. As often happens in the carpet trade, nomenclature is flexible — in this particular case Everu rugs are usually sold as Enjilas, but there is in fact no problem in distinguishing the genuine article from the cheaper copy. For one thing the fineness of stitch is easy to check: a good Enjilas will hav at least 160,000 knots/m2 (104 knots/in.2) while an Everu will rarely exceed 130,000- 140,000 (84-90/in.2); in addition, almost every other characteristic sets the Enjilas in a class apart. The only feature Everu has in common with its illustrious neighbour is the fine yarn used; in everything else Everu is inferior. The quality varies from quite good to very poor. The design, as shown in the example illustrated, is pretty clumsy, but the worst feature is the colouring: crude red, heavy dark blue and, above all, a vicious synthetic turquoise-green. A wide range of rug and runner sizes is made, including such things as small squares, which are rarely found in Persian goods.


To the south, outside the Hamadan region proper, there is a group of villages knownas the Kemereh, some of which are still inhabited by descendants of the Armenianstransported there by Shah Abbas I after the partition of Armenia between Persiaand Turkey in the early seventeenth century. The two best known are Lilihan (cf. fig. 638) and Reihan (cf. fig. 467), which have their own designs. The others produc single-wefted rugs in the Hamadan mahi style which the layman can perhaps bestdistinguish by the lighter shade of red used in the area, for the Kemereh lies in aregion that otherwise produces Saruq carpets, with their light madder rose, or Viss carpets, which are equally light in colour (see figs. 645, 534) There is, however, also a distinctive design feature by which one can identify Kemereh rugs – the shapeof the medallion, as illustrated here and again in fig. 585.

Alamdar, lamdar, assadabar, dabad, zagheh, Taimeh

In the Hamadan region proper there are several other villages that produce their own version of the Herati design. Typical of Alamdar are dozars and zaronims with a blue ground and the rather angular version of the design shown here. Alamdar lies to the south of Malayir in the direction of Borujird, and both the colourings and weave remind one of the latter — the warps, for example, appearing as very prominent white specks on the back of the rug.
Assadabad rugs, mainly dozars without a medallion, usually display a pleasant red, with a coarse but chunky weave. Here, too, the design is rather open (see illustration), with the peculiarity that the diamond motif is not the central feature of the design and hence does not occupy as important a position as it does in rugs made elsewhere. Assadabad lies at the top of the mountain pass on the road from Hamadan leading to the Kurdish area, and the design shows Kurdish influence, as a comparison with Kakaberu and Kolyai designs will confirm.
Just over the mountains from Assadabad, on the Hamadan side, lie Zagheh and Tajiabad and a group of villages weaving a full range of rug and runner sizes in the design illustrated in fig. 219. It is the shape of the medallion, together with the ‘teeth’ motif around it and the corners which help one remember the Zagheh
design; the ground, however, is yet another version of the Herati motif. Thequality is quite good, although the yarn is often suspect and the dyestuffs more tha suspect.
Another version of this design uses the same layout on a ground covered withbiggish boteh motifsAn example ofa Taimeh rug is included here because ofits superficial resemblance to Zagheh. In fact there is no possibility of confusing the two, Taimeh being a double-wefted rug of the Malayir/Jozan group, whereas all Hamadan rugs, including Zagheh, are single-wefted. Once again the ground is a special variant of the Herati design, although, as with Zagheh, the ground may also be filled with angular boteh motifs. For general details of the Malayir group, cf. fig. 653. Taimeh is unusual in that cream grounds, and lighter colours in general, are more common than in the other villages of the group.


Saruq is the name of a village on the edge of the Ferahan region which has given its name to the products of the whole area around Arak, the centre where the goods are marketed. The Herati is an old-established traditional design of the Saruq carpet, although today more carpets are perhaps made in other designs. Fig. 221, from the
village of Viss, illustrates a type that would normally be called a Mahal, i.e. a grade II (but still quite good) Saruq. Note the neatness and regularity, which nevertheless do not degenerate into stiffness. Fig. 222 is a really fine all-over Herati Saruq rug from Ghiassabad (which, with Mahallat village, is probably the best known of the
various villages in the region still capable today of producing very fine work): again there is exactness combined with warmth and charm. From the general appearance one may infer a relationship with the best Borchalu goods and other fine Herati rugs from the Hamadan area, but there is no possibility of confusing the two types since all Saruqs are double- and all Hamadans single-wefted.
Fig. 223 shows the best of all the new Saruqs, the Mashayekhi. Here the all-over Herati is converted into a medallion-and-corner carpet by the same expedient as is used in the Bijar carpet in fig. 196. Mashayekhi is a manufacturer who, reviving a traditional Saruq idea, has exactly captured the half-geometric, half-floral nature of the Herati design. By retaining just the right amount of the rectilinear in a stitch which could easily produce perfect curvilinear designs, he has established a standard version of the design which is widely admired and widely
copied. Copies of Mashayckhis are made in Bulgaria (fig. 226) and the author himself was responsible for starting one of the Mashayekhi copies made i n India (illustrated in fig. 162), but none of the copics have the flair of the Persian original. The prices of the originals, it must be admitted, are horrific, but the results produced rank with the best Bijar and Enjilas rugs as the modern pinnacle among goods using the Herati design. A fine village variant of the Mashayekhi style is shown in fig. 225, another Ghiassabad rug; what this example lacks is the finesse of Mashayekhi’s colour-balance, but it is a very good rug nevertheless. The Bulgarian copy illustrated
uses a good-quality yarn in a quite fine weave – although the production is fairly small, the quality is considerably above that of the standard grade of other Balkan countries. But at the heart of the success of the Bulgarian ‘Mashayekhi’ lie the same factors as inspire the original: the finely judged balance of colours and proportions of the design.


It is confusing that there are five quite different types of carpet sold as Ardebil. The easiest to deal with is what is known as the Ardebil design, the Sheikh Safi carpet, which does not come from Ardebil at all see fig. 702. Then there are the old-style kelleyis and runners from a fairly far-flung group of surrounding villages, woven in typical bold Azerbaijan designs and the heavy weave that characterizes the Persian southern Caucasus. These are illustrated in figs. 134, 508 and elsewhere (see index). The modern production of this same area apes the Russian Shirvan style: the goods are known in the trade as new-style Ardebils (see p. 225). Then there is an Ardebil town production which has the same structure as the old-style kelleyis but makes clumsily executed Tabriz designs. The clumsiness is accentuated by the fact that the designs are never unequivocally curvilinear or rectilinear. The results are rarely attractive: the colours are usually crude and the sizes often too long in relation to the width, but the weave is firm and solid and the carpets are relatively cheap. In the 1970s these carpets were displaced by an entirely new kind of Ardebil carpet based on the same design concept as is used in the Mesheriki carpets and the Bijars shown in figs. 223 and 196. This type is woven, in all carpet sizes and most rug sizes, in a wide range of qualities (in densities of less than 100,000 to over 300,000 knots/m2; 65-195 knots/in.2), the finest pieces even having a part-silk pile. Some pieces are made on woollen warps, which is most unusual for Azerbaijan. The design layout is the same in all grades, although the medallion shape varies considerably. At present (1980) only blue-ground carpets are made, often using a poor dyestuffthat tends to make the whole carpet look pinkish mauve.
The weaving pattern on the back varies enormously from one piece to another, and in many cases does not look like Ardebil at all – some even look like Birjands.
In all grades there are some very good pieces and others with a construction that is highly suspect. The worst pieces feel as thin as a rag, but the better ones are quite sound. The most attractive carpets are those where a proper deep blue is used for the ground and where the subsidiary colourings include a rich, medium leafy green and a fine madder-rose. These are the equals in appearance of good Saruqs and Bijars, and they demonstrate powerfully that the designers and colorists of Persia are still capable of producing new carpets of outstanding beauty. The fine goods of this type are called SinehbafF, i.e. ‘Senneh weave’, a rather misleading description. The name derives from the supposed resemblance of the back of the finest pieces to that of Senneh goods; it does not imply the use of the Persian (‘Senneh’) knot, since all Ardebils are woven with the Turkish knot.


Mesheriki has manufactories in Sarab and Tabriz as well as Arak, all making the same design – and to the same high standards. The layman has no hope of telling them apart. The professional will, however, detect the difference from the fact that the Tabriz version of any design always looks too stiff and rigid. This is a feature which can be seen in the Herati illustrated, which is typical of the standard production of many weavers in Tabriz; in these the accuracy of both the designing and the weaving, clearly seen in this piece (viewed from the back), seems to make the carpets too exact and perfect; they appear somehow too set and lack the natural flow of the Saruq versions. Nevertheless large quantities are made and sold in Tabriz in the Herati design, not only in the very fine type on a dark-blue ground as illustrated in fig. 228, but in every quality and colour conceivable, both with and
without medallions. The colours can be problematical: as with other designs, Tabriz Heratis are often spoilt by poor dyestuffs and ugly colour combinations.

The hereti pattern

221 Viss 321 X 237 cm 222 Ghiassabad 205 X 126 cm


The area from which some authorities at least suspect that the Herati design came has its own considerable production in a distinctive style today. The finest carpets are from Birjand itself; a lower grade comes from the villages around Dorukhsh to the north-east and Moud to the south-east, although Moud also produces very fine pieces. The designs are all of the Herati all-over type converted into medallionand-corner by the superimposed outline of the medallion and change of colour.
In some cases the medallions are geometric, as with the Bijar carpet of fig. 196; others have curvilinear medallions. The circular medallion with the well-drawn pendants shown in fig. 229 is particularly common. All-over pieces in one ground colour are also made. Cream and blue grounds predominate; reds are also made but are rarer. Since early in the present century Birjand has had a very low reputation owing to several malpractices, including the universal use of thejufti knot (i.e. one knot on four warp strings instead of two, giving the carpet only half the weight in wool; see fig. 35). Today, however, much better carpets arc produced. There are still plenty which are badly woven, thin and scruffy, but many manufacturers have managed to raise their standards and the best pieces are as fine as Mesheriki or finer, with the proper density and weight of pile and fine attention to details of
colouring. A bad name, once acquired, is not easily shaken off, however, and even the best Birjands cost no more than half the price of Mesherikis. The design is more or less fully curvilinear. For a close-up of the weave see fig. 33. All carpet sizes are made (including squares), but rugs and runners in general are rare.

Another intriguing example of the similarity of design ideas between Birjand and Qum is illustrated here. Sometimes one must study the weave in order to tell them apart (see figs. 342, 343), but in this case there is no problem since the Qum rug shown is all silk and silks are rarely made in Birjand. Many rugs are produced in Qum in the layout illustrated – a medallion filled with Herati motifs set in a plain ground. The hexagonal geometric medallion shown is very common, but more oval or rounded medallions with the same in-filling are also found. All the typical Qum colours may be found that is almost every conceivable ground shade from bright reds and strong dark blues to pastel pink and light green – but always with an overall feeling of lightness and brightness, however strong or soft the main shades may be.

The hereti pattern

223 Mashayekhi

The hereti pattern

224 Mashayekhi Saruq Detail viewed irom ihe back.
225 Ghiassabad 200 X 129 cm, 340,000 PK/m2 (6′ 6″ X 4′ 3″, 2 1 9 PK/in.2 ).
226 Bulgaria 322 X 221 c m , 250,000 P K / m 2 (10′ 7″ X 7′ 3″, 161 PK/in.2 ).


The Mina Khani design, which is used in 90% of all Veramin carpets (cf. 757) seems to be a variant of the Herati pattern which developed in eastern Persia in conjunction with the Herati itself. Today, however, Veramin makes a particularly original version of the Herati design which is quite different from the Mina Khani, as a comparison of fig. 232 with fig. 757 shows, Although this variant of the Herati pattern is thought of as a Veramin design, it may also be found in other central and west Persian origins.

Southern Persia

The Herati design does not form part of the Kerman/Yezd design tradition, but various manufactured versions in a medium stitch have been made at various times no attempt to preserve the geometric element found elsewhere. The output in this production is mainly on cream grounds in most rug and carpet sizes up to Jezd about 8 m2 (85 ft2).

Another southern Persian area with no major tradition of using the Herati design is Fars province, with its Qashqai and Luri nomads. The Qashqai do, however, have a version which one occasionally encounters, (see fig. 234). Note the unusual feature of using a different colour for the centres of the diamonds. The- town of Abadeh, at the foot of the Zagros mountains, on the edge of the desert, has a strong element of settled Qashqais in its population. The design which is used in the vast bulk of the local production is derived from a Qashqai original (see fig. 597). The same applies to the Abadeh version of the Herati design. Older Abadeh pieces look very like the example shown in fig. 234, but in recent times the more independent version shown in fig. 235 has developed; the piece shown is a weaver’s sample, as used in several parts of Iran in place of graph-paper loom drawings. Here, too, the diamonds are in a different colour from that of the ground; note in the border a variant of the Zil-i-Sultan bird and vase (cf fig. 271).

The hereti pattern

227 Ardebil Detail of a carpet, full size approx. 350 X 250 cm, 188,000 TK/m2 (11′ 6″ X 8′ 3″, 121 TK/in.2).
228 Tabriz Detail of a carpet (viewed from the back), full size 350 X 247 cm, 250,000 TK/m2 ( 1 1 ‘ 6″ X 8’ 1″, 161 TK/in.2).


Yet another version which seems alien to the local tradition is illustrated here. It is made in Turkoman Russia and sold as Beshir. Attention has already been drawn (see p. 77) to the interpretations of the Herati design by the Beshir nomads of Afghanistan. Here we have another which is quite distinctive and unlike any other version of the Herati motif used elsewhere, yet it is clearly Turkoman in feeling. It is a good example of how designers working in a clearly defined tradition can take an established design from elsewhere and refashion it in their own mould.
The fact that such refashioning has always been done throughout the East has been illustrated in several examples in the preceding pages, but it tends to be forgotten nowadays. We are apt to say ‘this design comes from that place and if anyone else uses it that amounts to mere imitation and is therefore inferior’. This was patently never the case in the past; and with the economic pressures affecting Asia at present it must not be thought to be the case now, or the art of carpet weaving will die out. It is fascinating in this connection to read what Grote- Hasenbalg has to say about the Ferahan Heratis of the eighteenth and nineteenth centuries: to him they represent an art in decline, a design torn from its folk-art roots and merely copied in another region. But ask any dealer or collector today and one will hear the highest praise for the old Ferahans. What Grote-Hasenbalg overlooks is the regeneration, the ‘Ferahanization’ of the Herati design. Of course in composed art what we value most are the great original achievements, but in folk art there are no originals, everything is copy. To be sure, each version at any stage of the tradition, including the beginning, was created by one individual, but not as an originator, not as a free creative spirit, rather as the tool of the tradition, the mouthpiece of the people, whose culture as a whole finds expression in the art he creates. Thus, tradition means copying, and this is something we must not forget when considering the manufactured copies which account for the majority of all carpets sold today – the complete output of China, Turkoman Russia and the Caucasus, which are copies of older styles from their own areas, and
the huge output of India, Pakistan, Rumania, Bulgaria and Albania in copies of Persian designs. These modern pieces are no more copies than the old Ferahans were; what counts is not where the design originated, but what today’s weavers have made of it.

The hereti pattern2

230 Birjand Detail of a carpet, full size 370 X 260 cm, 300,000 PK/m2 (12′ 2″ X 8′ 6″, 194 PK/in.2).
231 Qum 210 X 141 c m , 510,000 PK/m2 (6′ 10″ X 4′ 7″, 330 PK/in.2).