Abstract
In the history of technology, the loom has come to occupy an important place. While the horizontal handloom has a comparatively simple mechanism, this is not true of the vertical drawloom, which through centuries has developed complex forms. The question of the latter’s presence in India in early times has aroused some controversy. The case is made in this article that it arrived in the thirteenth century from Iran but failed to supplant the handloom in most areas of textile production, except for carpet weaving, mainly in Kashmir.
Both India and Iran can claim a history of textile weaving from the Neolithic times (in Iran from about 6500
In Parthian times (c. 200
During the Sasanian times (third–seventh centuries
Archaeological evidence had for long accepted a date around 100
The draw loom had one major advantage over its rivals in that while the latter limited the weaving of patterned design to geometrical figures, the complexity of which depended essentially on the number of heddles used, the drawloom alone could provide for a free-figure design: its harness permitted the control of every one of the warp threads.
The Chinese version of the modern French Jacquard loom had been truly invented in antiquity. Twills with simple geometrical patterns were woven as early as 2000
By the period of the Zhou dynasty (c. eleventh century–221
In time, there occurred the need for a silhouette pattern in a single colour against a contrasting background. 14 The two-weft system was considered more suitable for this purpose. In the beginning, the pattern was presumably darned on the warps of the second harness with one weft, the background with the other, one weft coming to the face where the second went to the back, and vice versa. This improvement was followed by some type of drawloom harness, early examples of the products of which were found in Egypt, on small patterns repeated on a diaper, covering the surface of the web. 15 It has been argued that this system of weaving arose out of the invention of tapestry weavers, from the fact that the finished bands were usually tapestry.
Subsequent to the appearance of the drawloom, different versions of it appeared at various weaving centres: cloth woven on a drawloom would have the main warp either between binders or being controlled by the drawloom harness. 16
This line of argument associates drawloom weaving with the beginning of large production of silk textiles, resulting in increasing imports of silk from China to the Near East in antiquity. The use of the drawloom in silk weaving involved complicated systems of controlling warp and weft which could be used to produce complex textile patterns. 17 This was an alien technology for the Near East, where wool weavers had a springy warp to deal with while they kept their warps under tension and their wefts loose. In addition, there was the factor of complexity of threading a warp pattern to manipulate large numbers of silk yarns per inch in place of the previous capacity limited to about 20 or 30 to the inch in the wool warps.
It has been suggested on the basis of literary sources that the Sasanian authorities shifted a number of silk weavers from western provinces (Syria) to weaving centres and factories further East. By the fifth century, the drawloom was also in use in the Byzantine empire. The Iranian weavers treated the warp differently from their Chinese counterparts. They spun silk to obtain a solid colour satin damask in order to get light reflected differently by the spun warp and unspun weft. Spinning the warps separately was not needed in the Far Eastern drawloom technique, where the wefts were locked in place by binder warps.
Silk weavers faced another practical problem. Silk yarns were qualitatively finer than wool, so these required comparatively many more warps to make a compact fabric of the same width. This required more alterations in the looms. Cloth bindings were so close together that the short floats did not reflect the light compared to the longer warp floats of Chinese silks. Of many solutions, Iranian weavers generally preferred the use of the drawloom harness with a three-harness twill loom binding. 18
Since silk yarns were fine, more than two wefts could be inserted in a single shed without making the fabric unnecessarily stiff or thick. In case several colours were used, one would come to the front of the pattern and the rest would lie at the back. This resulted in producing a fabric with a definite face and a definite back, the face showing one colour at a time and the back at least two. The other advantage of the drawloom was that the point repeat set-up produced a symmetrical pattern. The drawloom also suited plant fibres like cotton or linen, point repeats being used for cotton and comber repeats for linen. The size of the drawloom would be larger in the case of wool, especially when the same harness was used for wool patterns of much larger sizes. Such patterns, in turn, required the use of plant fibres for undyed wefts. Generally, the first woollen weft would be alternated with linen or cotton weft. The wool drawloom had the advantage of a great deal of variety in contrast to the silk one. Few silks could be woven with comber repeats, but comber and point repeats, as well as cloth and twill bindings, were common in wool fabrics. The silk warps were Z-spun and the wefts unspun, but in the case of wool, there were either S- or Z-spun single or plied wool warps and S- or Z-spun wefts. This was possibly the reason behind the general uniformity of silk designs, while with wool more variety could be obtained.
A pattern achieved through a complex weave demanded a different weaving technique, namely that of the drawloom, than either the simple weave or its patola variation (warp and weft pre-dyed). The drawloom multiplies the number of sheds that are controlled by cords pulled by a person assisting the weaver. Although figured silks, cottons and brocades of the highest quality were woven in India, it is not certain whether ancient or medieval India ever had a drawloom. Vijaya Ramaswamy’s suggestion that the drawloom was known in South India in the eleventh century 19 is based on a questionable identification of the term achchutari with the drawloom, since tari means loom and achchu mould or print. She has argued that if the two words are combined it refers to a process by which the threads are tied together to form certain sequences and then heddles are lifted by hand in the weaving of patterns. However, this means extending the sense of tari much beyond the range of reasonable inference. Vijaya Ramaswamy has traced a reference to the term achchutaru to a thirteenth-century inscription of Raja Raja III, but the text does not even remotely suggest a drawloom. 20
Though there is a view that India was familiar with the drawloom or pattern loom as early as the eleventh century,
21
the evidence adduced seems quite inadequate. Fancy weaves could also have been achieved through adding more harnesses to the loom,
22
for which we have, however, no evidence. Besides, nothing is said of figured patterns. Had India of the the eleventh century been familiar with the drawloom, before the arrival of the Muslims, the weavers guild would not have decided in 1538 that ‘this mode of weaving should be done only by the Muslims’,
23
in lieu of which they were authorised ‘to collect the income from the gifted lands for their weaving’. This regulation not only forbade this practice but also prescribed fines and punishment for any violation of the same by native weavers.
24
As we have seen, the Chinese were far ahead of all others in loom construction, especially in the evolution of the drawloom. They had the essentials of the drawloom developed as early as the first century
The word atha (literally, then or now) implies that artisans from different countries brought with them the craft of silk weaving and that from them, presumably, the local artisans learnt the craft of silk weaving. 32
In the succeeding verse (No. 29), Shrivara refers to ‘special type of woollen cloths (aurna), originating in foreign countries, such as sofa, etc., which are now shining (i.e. available in Kashmir); these are suitable (ucita) and fit (samartha) for the king (nṛpa)’. 33 In the next verse, he refers to a weaving technique presumably newly introduced and thus catching the poet’s imagination. He writes, ‘The images or patterns (ākriti) of diverse figures (citra) and creepers (latā) produced (utpanna) by multi-coloured weaving [on silk clothes] are making the artists (citra-kara) silent as if they are [not animated beings, but] painted figures’. 34 In the next verse, Shrivara refers to ‘The king’s dress (veśa), called silk (kauśeyaka-ākhyāta), which is beautiful (sundara) because of innumerable threads and with a display (vicchitti) of many colours, became resplendent (babhau), so also his land’. 35
While Shrivara’s passage thus indicates that the weaving of patterned silk cloth had arrived in Kashmir, he does not necessarily attest to the presence of the drawloom in Kashmir in his day. 36
A reference to the drawloom, however, seems implied in Abū’l Faẓl’s mention of a certain weaver named Ghiyāṣ Naqshband, where ‘Naqshband’ can be taken to mean a drawloom weaver. 37 However, there is no explicit description of the drawloom. Did the drawloom, then, have only limited use in medieval India because it was more expensive in comparison to ordinary looms? The use of the drawloom (‘naqsh’ loom) survived into modern times, for we have a clear description of the drawloom in the last decade of the nineteenth century at Murshidabad. The Murshidabad loom had a draw harness that operated on the figure warp only. It consisted of a large number of vertical drawstrings. Each drawstring was connected to a horizontal gut string in a cross harness. With the lifting of one drawstring, the cross harness gut string was lifted, and with it mails attached to warp threads were raised, whereupon the pattern could be repeated. Besides, the reference to the draw boy sitting on an elevated platform manipulating the draw strings makes the identification of the drawloom beyond dispute. 38
If, then, the naqshband distinctly means a weaver working with the drawloom, it could have been introduced in India as early as the thirteenth or fourteenth century, possibly from Iran. It seems that India had obtained the Iranian version of the drawloom, as is evident from the late reproduction of the device.
39
The drawloom, as we have seen, had a possible dual origin in antiquity in China and Syria.
40
Iran witnessed a modification in both the forms during the Sasanian period.
41
Another reason behind our acceptance of Iran as the source of dissemination of drawloom technology to India is the structural similarity between the two devices, whereas the Chinese model was quite complicated, with 50–60 heddles and an identical number of treadles as early as the third century
What is surprising is the limited use to which the drawloom was put. At Dacca, J. Taylor found as late as 1800 that Indian weavers wove flowered cloth on the ordinary throw shuttle horizontal loom. 44 Two weavers put a number of threads for the desired flowers or parts of the design to be formed. They drew each of these threads between as many threads of the warp as might be equal to the breadth of the part of the flower or design desired. The weavers would laboriously count and lift together the required numbers of warp threads through a bamboo stick for the operation of the shuttle. In this way, the jamdani or flowered cloth was woven in the ordinary throw shuttle horizontal loom.
As far as shawl weaving is concerned, Abū’l Faẓl informs us that Kashmir had monopoly over it prior to Akbar’s times, 45 probably owing to the good quality and adequate supply of wool from both local and neighbouring sources. 46 Akbar established shawl manufacture in Lahore as well. 47 By the last decade of the sixteenth century, Lahore had more than 1,000 workshops established to weave shawls. It became famous for the weaving of a particular kind of shawl called mayan that was a mix of silk and wool. It was used for chiras (turbans), fotas (loincloths), etc. 48
It is clear that a form of the drawloom was used in shawl manufacture in Kashmir. While praising commodities of Kashmiri origin, Bernier noted,
But what commodity may be considered peculiar to Kachemire, and the staple commodity, that which particularly promotes the trade of the country and fills it with wealth, is the prodigious quantity of shawls which they manufacture, and which gives occupation even to the little children. These shawls are about an ell and a half long, and an ell broad, ornamented at both ends with a sort of embroidery made in the loom, a foot in width.
49
The italicised reference to ‘little children’ reminds us of the ‘draw boys’ required for working the classic drawloom.
A very detailed description of the working of the carpet loom in Kashmir was provided by Moorcroft and Trebeck in 1819. A summary of their description follows.
50
Once the yarn was dyed, the Nakatu adjusted the yarn for the warp and the weft. The arrangement of the warp and weft threads was in the ratio of 2:1, the warp threads being ‘cut into lengths of three gaz and a half’. Any further reduction in length was considered fraudulent. The number of the warp threads could vary from 2,000 to 3,000,
according to closeness or openness of texture proposed and the fineness or coarseness of the yarn. The weft was single untwisted yarn but a little thicker; and consequently heavier. The Nakatu used to receive yarn in hanks and convert them into balls of yarn. One Nakatu could prepare material (warp and weft) for two shawls in a day. In the next step the Pennakam (‘warp dresser’) used to stretch the yarn ‘by means of sticks’ into bands and dipped into thick boiled rice water.
Then, it was squeezed and ‘again stretched into a band, which is brushed and dried’. This process used to stiffen and keep the threads separate.
A close scrutiny of the evidence provided by Moorcroft and Trebeck reveals two interesting methods of weaving: (a) weaving designs on the drawloom; and (b) obtaining designs through chain stitch embroidery.
The operation of ‘drawing [italics ours] or of passing the yarns of the warp through the heddles’ was similar to that practised by their European counterparts. Then, the warp was placed on the loom by the shawl weaver (shawl-baf). The loom of Kashmir did not differ in principle from that of Europe. Once the warp was fixed in the loom, the pattern drawer (nakash/naqqash, the term generally associated with drawloom weavers), the tarah guru and talim guru ‘or persons who determine the proportion of yarn of different colours to be employed, are again consulted’.
51
The pattern drawer (nakash/naqqash) would draw the pattern in black and white on a piece of paper. The tarah guru would give shape to the design (tarah),
having well considered it, points out the disposition of the colours, beginning at the pattern, and calling out the colour, the number of threads to which it is to extend, that by which it is to be followed, and so on in succession, until the whole pattern has been described.
52
Thus, the tarah guru dictated to the talim guru who wrote down the particulars in some characters or shorthand and delivered copies of the documents to the weavers. 53
Next, the tarah guru guided the weavers to knot the yarn of the tuji (eyeless wooden needles with slightly charred edges) to the warp. 54
The face or the right side of the cloth, is placed next to the ground, the work being carried on at the back or reverse, on which hang the needles in a row, and differing in number from four hundred to fifteen hundred according to the lightness or heaviness of the embroidery. 55
As soon as the tarah guru was satisfied that a line or woof was completed, the comb was ‘brought down upon it with a vigour and repetition apparently very disproportionate to the delicacy of the materials’. 56 The loom under discussion was thus quite close to the drawloom in all its essential features, including the terminologies used for identification of the operations involved.
However, there was another method of obtaining designs on worked shawls (doshali amli) through the looms. Here, the whole of the embroidery was worked on the cloth, with needles having eyes, and with a particular kind of woollen thread, instead of the silk employed in the usual embroidered work (i.e., in the drawloom). In weaving doshali amli shawls, the pattern was delineated on the already woven cloth. 57 The chain stitch form of this art of embroidery was thus practised on ‘a design transferred from a wooden block to the cloth’. 58
The drawloom described in such detail by Moorcroft and Trebeck was one specifically applicable to silk yarn. It enabled the weavers to produce distinct elaborate designs, though evidently such weaving was a slow and diligent process involving the use of a draw boy to select and lift the correct warp threads for each insertion of weft. The draw boy would pull one of such specific cords with the corresponding heddle eyes, and the desired warp threads would be lifted accordingly. 59
The use of the drawloom by shawl weavers in Kashmir seems to have spawned an industry with considerable employment. Rafail Davibegow reported that ‘in the vicinity of the town and in the town itself there are upto 24000 looms, many of which specialize in the weaving of shawls’. 60 W. Moorcroft and G. Trebeck reported (early nineteenth century) that about 120,000 out of the 800,000 people inhabiting Kashmir were reputedly employed in ‘shawl manufacture’. In other words, by the end of the eighteenth century, about 15 per cent of Kashmir’s population was engaged in shawl weaving. Master manufacturers (ustād) owned 3–300 looms each for weaving shawls, which were ‘generally crowded together in long low apartments’. 61
Carpet weaving was an important industrial craft in medieval India, apparently much influenced in its technique by Iran, where carpet weaving was practised as early as sixth–seventh century The Loom is stretched right up and down, made of cotton thread, and the Carpet wrought upon them with the woollen yarn of several colours by young boys of 8 to 12 years old, a man with the pattern of the work drawn upon paper, standing at the backside of the carpet, and directing the Boys that work it how much of each colour of yarn should be brought in. On every thread being wrought, they sheer it with a pair of sizers, and then proceed to the next.
74
His noteworthy observation is that ‘every thread being wrought, they share it with a pair of sizers, and then proceed to the next’. In pile carpet weaving, this technique of cutting the thread end of the tightened knot with a ‘pair of sizers’ is apparently very close to the Persian method of weaving pile carpets, though it does not involve the employment of the proper drawloom. 75
Footnotes
Multi-colour Silk-weaving in Kashmir: Note by Professor S.R. Sarma on Passage in Jaina-Rājataraṅgiṇī
Silk weaving in Kashmir, Jaina-Rājataraṅgiṇī, Book I, Chapter 6.
Verse 27: Dhar’s translation of this verse is all right. In stead of ‘with the idea of introducing crafts [here]’, literal translation would be ‘their imagination filled with [the idea] of craft’.
28: ‘Then (atha), the people of Kashmir, having practised the skill (cāturī) [of employing] the shuttle (turī), loom (veman) etc., are weaving (vayani) now very valuable (bahu-mūlya) and attractive (manohara) silk cloth (kauśeyaka)’.
The word atha (literally: then, or now) implies that artisans from different countries brought with them the craft of silk weaving and that from them the local artisans of Kashmir learnt the craft of silk weaving.
29: ‘Special type of woollen cloths (aurṇa), originating in foreign countries, such as sofā, etc., are now shining (i.e., available in Kashmir); these are suitable (ucita) and fit (samartha) for the king (nṛpa)’. The word sofa can also be read as sūf. But there is no mention of white colour or of pashmina wool.
30: ‘The images or patterns (ākṛti) of diverse figures (citra) and creepers (latā) produced (utpanna) by multi-coloured weaving [on the silk clothes] are making the artists (citra-kara) silent as if they are [not animated beings, but] painted figures’.
It is a bit convoluted way of saying. A person looks as if he is painted in a picture, which means that he is so amazed that he cannot say anything.
There is no word in the text for embroidery. The verse clearly says that these colourful patterns are produced by weaving (vayana).
Dhar is wrong is saying that the art of drawing became dumb-founded. The word used is citra-kara, artist.
31: ‘The king’s dress (veśa), called silk (kauśeyaka-ākhyāta), which is beautiful (sundara) because of innumerable threads and with the display (vicchitti) of many colours, became resplendent (babhau), so also his land’.
Dhar likes to read ‘the country named Kashgar’ in the expression kauśeyaka-ākhyāta deśa. It is a bit forced interpretation. Actually the word kauśeyaka (meaning that which is obtained a cocoon kośa) is attested even in the Mahābhārata.
32: There is something wrong with this verse, I do not see any verb here. Apparently the verse is reproduced incorrectly. I shall try to make a rough translation:
The first line describes the silk cloth as made beautiful by many special colours, figures (citra), circles (kaṭaka) and other decorations (alaṃkāra).
Dhar’s rendering of the verse as bracelets and other ornaments being printed upon silk robes is silly.
The second line is difficult. Vidyā-mānava cannot mean ‘the well read class’. Kauśeyatā khyātimān may roughly mean “one who became famous because of having silk”, referring to the king.
The third line describes the king: Illustrious, ever resplendent, with incomparable qualities, bearing the wealth of excellent tantra.
The king’s attribute sat-tantra-sampatti-bhṛt is imaginatively chosen, because tantra has meanings, all of which are appropriate here: thread, loom, warp, army, rule, administration.
The last line reads: By that king, through his own intelligence, his dress as well as his country. Here there must be some verb like [was made famous]. Probably the verse is reproduced incorrectly.
I do not see any word for printing in the text.