The Indus Civilization (Cambridge History of India, Supplementary Volume) c17: Arts and crafts

The Indus Civilization (Cambridge History of India, Supplementary Volume) c17

Language: English

Arts and crafts

c17.1

Though the seal-intaglios of the Indus civilization are in a class of their own, the general range of Harappan artistry is not comparable with that of the contemporary civilizations of Mesopotamia and Egypt. Individual achievement, however, is of sufficient quality to suggest that our picture is still far from complete, and in particular it may be that the art of wood-carving, of which a climate less sympathetic than that of Egypt has removed all vestige, was as developed in Harappan times as it was in later India. It is fair to presume that the artists who produced the little figurine of the dancing-girl, or the vital renderings of animal-forms on the steatite seals, represent an aesthetic capacity more broadly based than the recovered examples of it alone would indicate.

c17.2

The most monumental products of the Indus civilization are the stone sculptures. Apart from two disputed statuettes from Harappā, eleven pieces of statuary have come to light, of which three represent animals. *Seriatim* they are as follows:

c17.3

1 (pl. XVI). The head and shoulders of a bearded man, the whole fragment 7 ins. high, carved in steatite.¹ It was found at Mohenjo-daro in the DK Area at a depth of only 4½ ft., and may therefore be of late Harappan date, a supposition with which its exaggerated stylization (for example, in the hair) would be consistent. The head is bearded, with the upper lip shaved; the eyes are narrowed to an extent which has been thought without much reason to indicate a state of *yogī* or mystical contemplation; the nose is (or was) long, the lips thick, the forehead subnaturally low and bound with a fillet, the ears conventionally rendered and suggesting the cross-section of a shell. A hole bored on each side of the neck may have been intended to hold a metal necklace. Across the left shoulder is a cloak carved in relief with trefoils which were originally filled with red paste. When found, one of the eyes retained its shell-inlay, and the whole work was covered with a fine smooth "slip" which will be described in connexion with the seals (p. 76).

c17.4

¹ Marshall, I, 356.

c17.5

The trefoil pattern is not uncommon in the Harappan culture, and is probably significant. It occurs on a red stone stand² and frequently on beads of steatite or steatite-paste³ where, as on the statue, the trefoils were filled and backed with red paint or paste. It is suggested that the intention was to imitate etched carnelian beads; but, though this is not impossible, hitherto no carnelian beads bearing this design have been found, and the supposition is that they were imported rarities. The trefoil pattern is found in Mesopotamia, Egypt and Crete in comparable associations, and seems likely to represent a common symbolism which may have extended to the Indus valley. The earliest occurrences appear to have been in Mesopotamia: a man-headed "bull of heaven", probably of late Akkadian period in the Louvre, is carved for trefoil incrustations,⁴ and others similarly ornamented come from Warka⁵ and from Ur.⁶ The last is of the IIIrd Dynasty, perhaps about 2200 B.C. It bears the symbols of Shamash the Sun-god, Sin the Moon-god, and Ishtar the Morning and Evening Star, together with the trefoils which probably represent stars. With similar intent trefoils appear (with quatrefoils) in Egypt on Hathor the Mother-goddess as Lady of Heaven, and are well exemplified by the Hathor cows which sustain couches in Tutankh-Amen's tomb (*c.* 1350 B.C.), and by a painted figure of the XVIIIth Dynasty from Deir el-Bahari.⁷ In Crete the symbol recurs on bull-head (or cow-head) "rhytons" of about the same period.⁸ The analogues from Egypt and Mesopotamia at least combine to suggest a religious and in particular an astral connotation for the motif, and support the conjecture that the Mohenjo-daro bust may portray a deity or perhaps a priest-king.

c17.6

² Mackay, I, 412.

c17.7

³ *Ibid.* p. 508, etc.

c17.8

⁴ G. Contenau, *Manuel d'archéologie orientale*, II (Paris, 1931), 698-9.

c17.9

⁵ *Ibid.* and A. Evans, *The Palace of Minos*, II (1928), 261.

c17.10

⁶ *The Babylonian Legends of the Creation* (Brit. Mus. 1931), p. 59; *Antiquaries Journal*, III (1923), 331.

c17.11

⁷ Evans, *op. cit.* I (1921), 513-14.

c17.12

⁸ *Ibid.* IV (1935), 315.

c17.13

2. Badly weathered limestone head, 5½ in. high. Too worn for description, though the conventional rendering of the ears and the white stone inlay of one of the eyes can still be detected. Found at a high (presumably late) level in the southern half of the citadel.

c17.14

3 (pl. XVIIIA). Limestone head, nearly 7 in. high. Closely cropped wavy hair held together by a fillet; shaven upper lip; conventional shell-shaped ears. Former inlay is missing from the eyes. The modelling of the cheeks and lips is sensitive, and the rendering of the hair schematic but expressive. The excavator remarked that “it looks as if some attempt at portraiture had been made”. Found 6 ft. 7 ins. below the surface in HR Area and ascribed to the “Late Period”.

c17.15

4. Limestone head, 7½ ins. high. The surface is worn and perhaps never finished. The hair, as on no. 3 above, is gathered in a “bun” at the back, where there are indications of three strands. The chin shows no traces of a beard; the ear is schematic as on the other examples; the eyes were formerly inlaid. The face is disproportionately large. Found 2 ft. below the surface in the southern part of the citadel, and presumably late.

c17.16

5 (pl. XVIIA). Seated alabaster male figure, 11½ ins. high. The arrangement of the clothing (which may have depended upon colour for detail) is not clear; it has been described as “a thin kilt-like garment fastened round the waist, partly covered by a shawl of thin material worn over the left shoulder and under the right arm”, but this is not certain. The left knee is raised and clasped by the left hand, which is crudely indicated. The head is missing; the back of the hair is unfinished, and is flanked by a rope-like pendant which may be hair or head-dress. As a whole, the modelling is poor. Found high up in the citadel building which produced no. 4 above.

c17.17

6. Much-weathered alabaster statue of a squatting man, 16½ ins. high. The right knee is raised; the hands rest on the knees, and between them the fold of a skirt-like garment is indicated. The bearded face has lost most of its detail, including the inlay for the eyes, but, as on the other Mohenjo-daro heads, the face was disproportionate to the remainder of the skull. A fillet is tied at the back of the head, and the ends hang down. Found in fragments in and about the building in HR Area noted above (p. 39), and ascribed tentatively to the “Late Period”.

c17.18

7. Fragment of a limestone figurine, formerly polished, showing a crudely indicated hand on a knee, probably similar to no. 6 above. Found 4 ft. below the surface on the citadel.

c17.19

8. Much-weathered fragment of a squatting or seated figure of limestone, now 8½ ins. high. The hand is on the knee as in no. 6 above. A series of holes drilled just above the ankles may represent affixed or inlaid anklets. Found at a high and presumably late level on the citadel near the court of the “college of priests”.

c17.20

9. Unfinished limestone figure of a squatting man, 8½ ins. high. The hands are on the knees, and there is a kilt-like garment stretched between the legs. There are indications of a fillet round the head. In pose, the figure resembles no. 6 above. From an upper level in DK Area.

c17.21

10. Fragment of a small limestone figurine of an animal, 4½ ins. high, possibly a ram. Found 2 ft. below the surface in HR Area.

c17.22

11. Limestone figure, 10 ins. high, of a composite animal; the head is badly damaged but apparently had ram’s horns and an elephant’s trunk. The body is that of a ram. Comparable animals of composite types occur on the seals. Found 3 ft. below the surface in DK Area.

c17.23

Of the eleven stone sculptures listed above,¹ it will be observed that four or five represent a stereotyped squatting figure, presumably of a god. To the same divine category may be ascribed the composite animal and, in all probability, the bust with the trefoiled garment. Two or three of the human figures are apparently unfinished. All the sculptures are derived from the higher and presumably later levels, but it must be remembered that the lower levels are much less known, so that the significance of this stratification, such as it be, cannot be computed. Five of the sculptures were found on the citadel—a significantly high proportion, having regard to the wide extent of excavation elsewhere. The special character of the building in HR Area in the vicinity of no. 6 may again be emphasized (above, p. 39).

c17.24

¹ Fragments of two tiny statuettes from Mohenjo-daro are omitted. Mackay, I, 258.

c17.25

Stylistically, these sculptures are largely *sui generis*. The rendering of the somewhat narrow (but not Mongoloid) eyes and the hair, and the extreme disharmony of the face in relation to the remainder of the head, in particular the low receding forehead,¹ are features which distinguish the series from the approximately contemporary works of Mesopotamia.² On the other hand, the notably sturdy neck and the shaven upper lip are common to the art of both countries, and the use of inlay for the eyes—a sufficiently obvious device—is familiar also in Mesopotamia and Egypt. The modelling is rudimentary, or perhaps decadent if these works do in fact belong to a late phase of the city; and the additional possibility of excessive generalization in religious sculpture conforming with a narrow hieratic tradition may be borne in mind. Certainly if two much-discussed stone statuettes from Harappā are also of the Indus period, the potentiality of the Indus sculptor is not represented by the Mohenjo-daro series.

c17.26

¹ This feature is not characteristic of known Harappan skulls.

c17.27

² The eyes of Sumerian statues are liable to be appreciably more owl-like and staring than those of the Mohenjo-daro figures, e.g. the Early Dynastic hoard from Tell Asmar, H. Frankfort, *Or. Inst. Discoveries in Iraq*, 1933-34 (Comm. Or. Inst. Chicago, no. 19), pp. 55 ff.

c17.28

These two statuettes, just under 4 ins. in height as preserved, are male torsos exhibiting a sensitiveness and vivacity of modelling entirely foreign to the works considered above.³ So outstanding are their qualities that some doubt must for the present remain as to the validity of their ascription to the Indus period. Unfortunately the technical methods employed by their finders were not such as to provide satisfactory stratigraphical evidence; and the statements that one, the dancer, was found on the granary site at Harappā and that the other was “4 ft. 10 in.” below the surface in the same general area do not in themselves preclude the possibility of intrusion. Attribution to a later period is also not free from difficulty, and doubt can only be resolved by further and more adequately documented discoveries of a comparable kind. Meanwhile it will suffice here to observe that one of these statuettes (pl. XVII c), in spite of an element of "frontality", is a realistic rendering of a somewhat adipose youth, in which the muscular forms are indicated with observation and restraint and with—be it noted—the breadth of style which is a notable feature of the engraved seals (p. 76); whilst the other, less accomplished in the rendering of detail, is nevertheless a lively figure with no affinity to the dead formalism of the Mohenjo-daro statuary. Incidentally, the figure appears to have been ithyphallic, and the suggestion that it may represent a prototype of the familiar dancing Śiva Naṭarāja is a plausible one.

c17.29

³ Marshall, I, 44 ff.; Vats, I, 22, 74.

c17.30

From the stone sculptures we turn to those of bronze. These are small "minor" works but include the most remarkable of the authenticated Indus figurines, the dancing-girl from Mohenjo-daro (pl. XVIIB).¹ Without the missing feet and ankles, this charming little statuette is 4½ ins. high; it was found 6 ft. 4 ins. below the surface in a house in HR Area and, though presumably not of the latest period, cannot be regarded as very early. The right hand rests on the hip; the left arm, covered almost entirely with bangles, hangs loosely, and the posture of the legs is easy. The head, provocatively tilted, is a skilful impressionistic rendering of a prognathic "aboriginal" type, with large eyes, flat nose and bunched curly hair; but whether, as has been suspected, a Baluch native is indicated, or whether the derivation is rather from south India, with which the Indus civilization was certainly in contact, is disputable.

c17.31

¹ Marshall, I, 44, 345; Piggott, pp. 115, 186.

c17.32

A comparable but inferior bronze figurine found in DK Area² adds nothing to our knowledge. Of better quality is a detached bronze foot wearing an anklet,³ found in a high (late?) level; and amongst a number of bronze images of animals, a subject in which the Indus modeller was at his best, mention may be made of a buffalo and a ram or goat, also from Mohenjo-daro.⁴ The former has caught expressively the characteristic stance of the animal, with massive uplifted head and swept-back horns.

c17.33

² Mackay, I, 274.

c17.34

³ Ibid. I, 273.

c17.35

⁴ Ibid. I, 283; II, pls. LXXI, 23 and LXXIV, 18.

c17.36

From these bronzes the transition is easy to the vast number of terra-cotta figurines which are characteristic of the Indus civilization at all known periods and are, as a class, quite unlike those of Mesopotamia. Until properly stratified excavation produces a chronological series, the terra-cottas can only be considered in bulk; for it is certain that ostensibly "primitive" and "evolved" styles were often enough contemporary with each other, and that a purely stylistic classification would be merely misleading. The red colouring of the clay is or was normally heightened by a red wash or slip, occasionally polished.

c17.37

The terra-cottas may be considered in two main categories, those of human and animal figurines.⁵ Of the human figurines, one of the most remarkable is that of a man found in 1950 on the site of the granary in the citadel of Mohenjo-daro (pl. XVIII B): a flat-bodied representation seemingly of a definite (Semitic?) ethnic type, with long nose and receding, fleshy chin, beardless. The head-dress is incomplete. There is no reason to suppose that either this or other, more crudely modelled nude figures with pellet-eyes, slit or applied mouths and pinched-up noses represent religious types; but a horned figure from the DK Area¹ was presumably a deity, and a curious series of horned masks with oblique eyes, cast from moulds,² may have been suspended as apotropaic charms. A Janus-like double head, also impressed from a mould or moulds, was doubtless that of a divinity, and a squatting bearded figure³ may be significantly reminiscent of the seated stone statues which were probably those of deities. Occasionally male and female figures are deliberately grotesque,⁴ and are doubtless purely secular, though this appearance is not incompatible with a religious purpose.

c17.38

⁵ For an analysis of the principal groups, see D. H. and M. E. Gordon in *Iraq*, VII (London, 1940), 2 ff.

c17.39

¹ Mackay, II, pl. LXXII, 7.

c17.40

² Ibid. I, 267; II, pl. LXXIV, 21-2, 25-6 and pl. LXXVI, 1-4.

c17.41

³ Ibid. II, pl. LXXIV, 23-4.

c17.42

⁴ E.g. Mackay, II, pl. LXXIII, 8.

c17.43

A large number of the terra-cottas represent females (pl. XIX), and there has been perhaps an exaggerated tendency to regard these as a manifestation of the Great Mother Goddess familiar in the religions of western Asia and parts of Europe. The commonest Harappan type is a standing figure adorned with a wide girdle, often with a loin-cloth and nearly always with a necklace and an emphatic head-dress which is generally fan-shaped above, sometimes with a shell-like cup or pannier on each side. The pannier appears to have been used in some instances for burning lamp-oil or incense. The features and general modelling are of the crudest; the eyes and breasts are circular pellets, the nose beak-like, and the mouth an applied strip of clay with a horizontal gash. No special artistry went to the making of these figures. Occasionally a lump of clay is added to represent an infant at the breast or on the hip; and the general notion of fertility, whether in thanksgiving or in anticipation, is further indicated by representations of pregnancy, although there is no emphasis of the generative organs such as is normal to Mother Goddess cults.⁵ Women, with or without children, lying on beds may nevertheless be related to the idea of fecundity. Other figures are seated, or engaged upon household occupations such as kneading flour, and were doubtless toys.

c17.44

⁵ The emphatic representation of a female vulva from Periāno Ghundai in northern Baluchistan is very exceptional in the chalcolithic cultures of the Indus and its borders. A. Stein, *An Archaeological Tour in Waziristan*, etc. (1929), pl. IX, P.C. 17.

c17.45

Nearly three-quarters of the terra-cottas represent cattle, normally humped bulls although the short-horn and the buffalo also occur. Strangely, cows are *never* represented. Other animals include the dog, sheep (rarely), elephant, rhinoceros, pig, monkey, turtle, and indeterminate birds. One terra-cotta, from a late level at Mohenjo-daro, seems to represent a horse, reminding us that the jaw-bone of a horse is also recorded from the site, and that the horse was known at a considerably earlier period in northern Baluchistan (p. 60). Man-headed animals, often with beard and short horns, are not uncommon. The mould is rarely or never used for these figurines, and the modelling is generally rough and summary. A few of the bulls, however, reach a high level of excellence; pl. XX represents a boldly rendered example in which the strong neck and head and heavy dewlap show an unusual mastery. Whether these figures were secular or votive or both can only be guessed.

c17.46

Small model carts of terra-cotta with solid wheels have been noted above (p. 60) as a characteristic feature of the Indus civilization, and were doubtless in some instances associated with oxen mounted on wheels and having pivoted and movable heads, such as have been found occasionally in the two cities. Harappā has also produced a charming little copper model of an *ekka*-like cart, only two inches in height without the wheels, which are missing. It is open back and front, where the driver is seated, but is closed at the sides and has a gabled roof.¹ Two other copper toy carts were found at Chanhu-daro, one similar to the Harappā example, the other, which preserves its solid wheels, of a simpler type without cover.²

c17.47

¹ Vats, I, 99.

c17.48

² Mackay, *Chanhu-daro*, p. 164.

c17.49

Terra-cotta was used for a variety of objects in addition to the categories described above. Whistles made in the form of a hollow bird (hen?) with a small hole in the back or side are characteristic of the Harappan culture. Round pottery rattles with small clay pellets inside are fairly numerous. Cubical or tabular dice were of pottery, marked (save in one example from Harappā)³ not as to-day, i.e. so that the sum of two opposite sides is seven, but with 1 opposite 2, 3 opposite 4, and 5 opposite 6. A similarly marked terra-cotta die occurred at Tepe Gawra, near Mosul, in stratum VI, which ended about 2300 B.C.⁴ It may be recalled that dicing was later a favourite pursuit of Vedic India. Pottery spoons imitate the commoner shell prototypes. Abundant carrot-shaped cones of plain terra-cotta recall in some measure the coloured cones which sometimes variegated the surface of Sumerian buildings, but were more probably used in the Indus valley in spinning; there is no evidence there of an architectural use. Discoidal spindle-whirls are common. Triangular (occasionally round or squarish) cakes of baked clay, varying from 1½ to 4 ins. across, have been regarded as "model cakes" for ritual use either as offerings or as grave-goods. This interpretation is unproved and unlikely. The "cakes" are roughly made but have no determinate feature except the flat sides and rounded angles. Their great abundance, particularly in drains, would be consistent with a use in the toilet, either as flesh-rubbers¹ or as an equivalent to toilet-paper, much as lumps of earth are sometimes used by the modern peasantry. Other slabs of terra-cotta with a pricked, file-like face and smooth rounded back are more certainly recognized as flesh-rubbers and sometimes show evidence of considerable wear. Finally, reference may be made to fragments of terra-cotta cages in which insects or small animals may have been kept,² and to little terra-cotta coffers with open ends, thought to be mouse-traps.³

c17.50

³ Vats, I, 193.

c17.51

⁴ E. A. Speiser, *Excavations at Tepe Gawra* (Am. Sch. Or. Research, Philadelphia), I (1935), 82.

c17.52

¹ Soap in the modern sense did not come into use until Roman Imperial times.

c17.53

² Mackay, I, 426.

c17.54

³ Ibid. p. 427.

c17.55

Some of the most skilful models of animals are made in faïence, which was abundantly familiar to the Harappans, and was already known to pre-Dynastic Egypt and to fourth-millennium Sumer. Certainly by 3000 B.C. its manufacture was widespread in western Asia as far north as the Caucasus, and it had reached Crete by Early Minoan II (about 2800–2500 B.C.).⁴ The process is to model the object in paste, which on the Indus is sometimes composed of crushed steatite, and to coat it with a glaze which is then fused in a muffle or kiln. The resulting colour is now generally light blue or green. The objects rendered in these materials are small and may normally be classified as beads or amulets. To the latter category belong tiny figurines of sheep, monkeys, dogs and squirrels, which at their best and within obvious limitations are little masterpieces of craftsmanship. The beads will be dealt with separately below. Miniature vessels, which must mostly have been toys, were made of faïence, as of pottery and stone, and were in rare instances ornamented with paint, a procedure with analogies in Sumer, Egypt and Crete. Faïence was also employed for a number of other objects, including bracelets, finger-rings, studs, buttons, and inlays presumably for caskets and furniture.

c17.56

⁴ See generally Marshall, II, 579 ff. (but with modified chronology).

c17.57

A vitreous glaze was used in a remarkable fashion upon a certain category of pottery found at Mohenjo-daro in some of the earliest known levels.⁵ These sherds are of a light grey ware covered with a dark purplish slip which had then been carefully burnished; to this, glaze was applied but, before firing, a portion of both glaze and slip was removed with a comb to form straight or wavy lines as a decorative pattern. Nothing like this ware has yet been found in Mesopotamia, and it would appear to be a local and relatively short-lived invention, dating perhaps from the middle of the third millennium B.C.

c17.58

⁵ Marshall, II, 578, 692–3; Mackay, I, 187.

c17.59

If we pass on to the Indus pottery in general, we are at present confronted with an inchoate mass of material into which only fresh and systematic digging on modern lines can be expected to bring order. There is no doubt that the so-called uniformity of the Harappan culture in depth has been exaggerated, and is due as much to archaic methods of research as to any inherent conservatism in the ancient craftsmen. The excavations on the Mohenjo-daro citadel in 1950 showed that change and evolution are clearly recognizable in the Indus ceramic and that, in particular, there was a lowering of technical standards in the later phases. The details remain to be explored and worked out in connexion with further deep digging, and the task is well worth the considerable labour which would have to be expended upon it.

c17.60

Meanwhile, a few general points may be noted. The great bulk of the material is wheel-turned, but some hand-made pottery has been recovered from the lower levels.¹ To the later levels only belong the so-called "goblets", small pointed vessels with scored exterior and often coated with a thin cream wash (fig. 12, 3). Some of them bear a short stamped inscription (potter's name?)—the only Harappan pots so marked;² ten examples from Harappā itself bear the same stamp. Whether or no they were used once only for drinking purposes and then immediately thrown away, like the common drinking-cups of modern India, they at least occur in great quantities in late groups. For the rest, most of the pottery is of pinkish ware with a bright red slip³ and decoration, where present, in black. Occasionally three colours—buff or pink, red and black—appear, and, more rarely, white and green are used, apparently after firing. Sometimes a clay was used that burned grey, but whether the colour was natural or was darkened by the admixture of carbonaceous material with the clay has not been determined. The pots were baked in round kilns with domed tops, pierced floors and underlying fire-pits.⁴

c17.61

¹ Mackay, I, 180.

c17.62

² Many pots bear graffiti scratched *after* baking, but that is another matter.

c17.63

³ Or sometimes a white coating, possibly of gypsum, which appears to be deliberate.

c17.64

⁴ Mackay, I, 177.

c17.65

Painted decoration is of better quality in the lower levels so far explored at Mohenjo-daro, but is not entirely absent from the later. The commonest and simplest type consists of horizontal lines of various thickness. More pictorial motives include intersecting circles or derivative leaf-patterns, scales, chequers, lattice-work, "kidney-shaped" designs based upon the conch-shell sections which were frequently used for inlay, "comb"-patterns, wave-patterns variegated by cross-hatching, and semi-naturalistic forms, notably palms, pipal-trees and rosette-like floral units. Peacocks sometimes appear singly or in superimposed series (fig. 12, 10), and fish are represented, often with cross-hatched bodies. Caprids are rare,⁵ and most of the animals familiar on the Indus seals do not appear at all on the pottery.

c17.66

⁵ A notable exception is a she-goat or doe suckling her young, on a sherd from Harappā. Vats, I, 289; II, pl. LXIX, 12.

c17.67

c17.68

c17.69

c17.70

c17.71

c17.72

c17.73

c17.74

c17.75

c17.76

c17.77

c17.78

c17.79

c17.80

FIG. 12. Pottery from Cemetery R 37, Harappā. Scale: 2, 3-7, 11, ⅛; remainder, ⅙.

c17.81

The human form is also very exceptional. Three sherds of the same pot from Harappā show a frieze of realistic panels separated by bands of chequer-pattern or counterchanged squares; one panel bears a tree, another a doe suckling her young, with a bird on her back and a fish, reeds and other symbols in the background, and a third panel illustrates a man and a child, both with uplifted hands, with birds and fish.¹ Another sherd from the same city portrays a man carrying two fishing-nets, with parts of another human form alongside.² As a whole, these designs are without close analogy, and in the present state of knowledge the Harappan pottery helps rather to isolate the Indus civilization than to link it up with other cultures.³

c17.82

¹ Vats, I, 112; II, pl. LXIX, I, 3-4.

c17.83

² Ibid. II, pl. LXIX, 16.

c17.84

³ Links between the Harappan and Jhukar pottery on the one hand and certain Halafian wares of north-eastern Syria (Tell Halaf, Tell Brak, Arpachiyah) on the other have been proposed but are unconvincing. See D. H. Gordon, “Sialk, Giyan, Hissar and the Indo-Iranian connection”, in *Man in India*, XXVII (1947), 215.

c17.85

A few other characteristic Harappan pottery motives may be noticed before passing on. Occasional vessels bearing a knobbed decoration, from Mohenjo-daro, are comparable with sherds from Sargonid levels at Tell Asmar in Mesopotamia.⁴ The interior of certain types of dish, including occasional pedestal-dishes or “offering-dishes”, is decorated with concentric rings of incised pattern, imprinted sometimes with a reed, sometimes with a finger-nail, but sometimes certainly with a cogged wheel or roulette⁵—a remarkably early use of a device more familiar in much later Graeco-Roman pottery. It is thought that at Mohenjo-daro this decoration is early, but further evidence is required. A third type of pot worthy of note is a more or less cylindrical vessel perforated all over (fig. 12, 8), a type specially characteristic of the Indus civilization. It has been alternatively identified as a strainer (possibly for pressing curds) and as a brazier, but, though one or two examples have been found in association with ashes, traces of burning are not normally present.⁶ Another type is a bowl with an internal knob on the base, resembling a characteristic type from Jamdat Nasr in Mesopotamia, though no significance need attach to this resemblance. A series of tiny pots with narrow openings is thought to have contained an eye-powder such as antimony; certainly a number of copper or bronze rods 4½–5 ins. long resemble ancient and modern kohl-sticks in the East and doubtless indicate the practice of anointing the eyes for medicinal and decorative purposes.

c17.86

⁴ Mackay, I, 208.

c17.87

⁵ Mackay, I, 184; II, pl. LXVII, 17-20, 22, 25.

c17.88

⁶ Ibid. I, 207.

c17.89

The Harappan beads are abundant, varied in form and material, and important historically. Their materials are of gold, silver, copper, faïence, steatite, semi-precious stones, shell and pottery. The processes of sawing, flaking, grinding and boring the stone beads are well illustrated at Chanhu-daro, where a bead-maker's shop was found.¹ The boring was effected either with chert drills or with bronze tubular drills, whilst minute beads of steatite paste seem to have been formed by pressing the paste through fine-gauge bronze tubes. The stone drills were very carefully made with tiny cupped points to hold the abrasive and water that gave the drill the necessary bite. A similar drill was found at Ur,² but no site has produced so many of them as Chanhu-daro, and the possibility of an export-trade in beads from the Indus is worthy of consideration.

c17.90

¹ Mackay, *Chanhu-daro*, pp. 186, 210; and "Bead Making in Ancient Sind", *Journ. American Oriental Soc.* LVII, 1–15.

c17.91

² Mackay, *Chanhu-daro*, p. 212.

c17.92

A remarkable series of gold beads was included in an important hoard of jewellery found at Mohenjo-daro in the HR Area. It lay at a depth of only 6 ft. from the surface and was therefore presumably late, although "the rolled up condition of some of the gold ornaments" suggested to the excavator that the hoard "was the property of a goldsmith, who kept it by him until he had enough material to warrant re-melting".³ Individual beads may therefore be of appreciably earlier date. The most notable type is a flat disk with an axial tube (fig. 13, 8), a form which is natural to metal but is also copied in faïence (fig. 13, 9) and is identical with Sumerian beads of Early Dynastic III—Akkadian date (*c.* 2500–2300 B.C.).⁴ The type occurs consistently at Troy at the end of IIg, about 2300 B.C.⁵ The gold examples may be an importation into the Indus valley, but the faïence copies are perhaps more likely to be of local manufacture (fig. 13, 9).

c17.93

³ Marshall, II, 522.

c17.94

⁴ Vats, II, pl. CXXXIII, 3; D. E. McCown, *The Comparative Stratigraphy of Early Iran* (Chicago, 1942), p. 53 and Table 1; V. Gordon Childe, *New Light on the Most Ancient East* (London, 1952), pp. 162, 182.

c17.95

⁵ C. W. Blegen and others, *Troy* (Princeton, 1950), I, 367 and fig. 357, no. 37.712. Similar beads were included in Schliemann's Great Treasure A from Troy II.

c17.96

The silver beads are mostly of simple globular or barrel form and do not call for comment. Beads of copper or bronze are more common but conform with the same elementary types. Some, perhaps many, of them were originally gilt.

c17.97

One of the most significant types amongst the fairly numerous faïence beads has already been noticed. Another is the so-called "segmented" bead, of which about thirty examples have been found at Harappā, some at least in late levels, and a number at Mohenjo-daro and Chanhu-daro (fig. 13, 10–11).⁶ This type is familiar widely in space and time, from Tell Brak in northern Syria in the Jamdat Nasr period (about 3000 B.C.) to Crete and Egypt in Middle Minoan III and the XVIIIth Dynasty.⁷ It is even found in barrows in Wiltshire, where it is regarded as a fixed chronological point in our Middle Bronze Age.¹ Dr P. D. Ritchie has shown by spectrographic analysis that two segmented beads respectively from Knossos and Harappā are absolutely identical in composition, and it is to be presumed therefore that they were derived from the same source approximately at the same time, i.e. about 1600 B.C. But what that source was remains to be discovered.

c17.98

⁶ Mackay I, 511; Mackay, *Chanhu-daro*, p. 205; H. C. Beck in Vats, I, 406; and especially J. F. S. Stone in *Antiquity*, XXIII (1949), 201–5.

c17.99

⁷ M. E. L. Mallowan in *Iraq*, IX, 254 f.; and Beck and Stone as cited. For the spectrographic analysis, see Stone.

c17.100

¹ Beck and Stone, "Faïence Beads of the British Bronze Age", *Archaeologia*, LXXXV (1935), 203.

c17.101

By far the commonest material is steatite, or a paste made of ground-up steatite. Of the latter a noteworthy series is barrel-shaped or a convex bicone and carved with a trefoil-pattern (fig. 13, 4), which is cut with a drill; the background is also cut away, and the recessed surfaces were filled with red (occasionally black) paste, leaving the design in white outline as on an etched carnelian bead.²

c17.102

² Mackay, I, 508; II, pl. cxxxvii, 94-8; Vats, I, 435-6; II, pls. cxxviii, 5 and cxxxiii, 2.

c17.103

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

c17.104

FIG. 13. Beads from Mohenjo-daro and Harappā. ½.

c17.105

Occasionally the pattern is rendered by a background of red paint without cutting (fig. 13, 1-3). For the wider connexions of the trefoil-pattern, see above, p. 64; it was in use in and after the Sargonid period (about 2300 B.C.) in Mesopotamia and nearly a thousand years later in Egypt but is not known to occur on beads outside the Indus valley, and the presumed carnelian prototypes have yet to be found.

c17.106

Decorated carnelian beads, though not numerous, occur at all three excavated Indus sites and have close counterparts in Mesopotamia. They have been classified into two main groups: I, white on red, and II, black on a white base (very rarely, black on red).³ Beads of type I are the more common and are made by drawing a pattern on the stone with a solution of alkali (generally soda), and then heating the stone until the alkali enters into it, thus making a permanent white design (fig. 13, 5). In type II the stone is flooded with the alkali and a black pattern is drawn on top of the white, probably with a solution of copper nitrate. Scarcely more than half-a-dozen examples of type II have come from the Indus, but the technique is known from Mesopotamia and as far afield as Damascus. Of type I, “eye” beads and beads decorated with figure-of-eight circles and rectilinear lozenge-patterns are identical at Mohenjo-daro, Chanhu-daro, Ur (“Royal Tombs”), Kish, and Tell Asmar (Sargonid period),¹ and must derive from a common source.

c17.107

³ Beck, "Etched Carnelian Beads", *Ant. Journ.* XIII (1933), 384-98.

c17.108

¹ H. Frankfort, *Tell Asmar, Khafaje and Khorsabad* (Or. Inst. Chicago Communications, no. 16, 1933), p. 48; and Mackay in *Antiquity*, v (1931), 459–61.

c17.109

Bracelets, rings, gamesmen, and a multitude of other objects come properly within the compass of this section, but, reserving certain types of pin for a later section, we may conclude with some account of the seals which are the outstanding contribution of the Indus civilization to ancient craftsmanship. The fact that over 1200 of them have been found at Mohenjo-daro alone indicates their ancient popularity and, although there is considerable variation in the quality of their cutting, their average attainment is exceedingly high for what must have been an almost mass-produced commodity. At their best, it would be no exaggeration to describe them as little masterpieces of controlled realism, with a monumental strength in one sense out of all proportion to their size and in another entirely related to it. The normal seal was of steatite and square in shape with sides from ¾ in. to 1¼ ins. in length, and with a perforated boss at the back for handling and suspension. Occasionally the boss is absent; sometimes the seal is round, with or without a boss; and there are a few cylinder seals. But these variants are very exceptional and may in most instances be ascribed to external influences. In manufacture, the stone was cut with a saw and finished with a knife and an abrasive, the carving being done with a small chisel and a drill. Finally, the whole stone was coated with an alkali and heated, so as to produce a white lustrous surface which has sometimes been mistaken for a steatite slip, a process reminiscent of, but perhaps not identical technically with, the “glazing” of steatite in the West as early as the Jamdat Nasr period (about 3000 B.C.), for example at Tell Brak in northern Syria.²

c17.110

² M. E. L. Mallowan in *Iraq*, ix (1947), 254.

c17.111

The intaglio designs on the seals include a wide range of animals associated in almost every case with groups of signs in a semi-pictographic script (below, p. 81). Some seals, however, bear script only, and some, which will be reserved for later consideration, bear human or semi-human forms. There are likewise purely linear designs, notably the swastika, but also multiple squares set concentrically, a criss-cross pattern, and a plain multiple cross. The animal most frequently represented is an ox-like beast seemingly with a single horn and nicknamed therefore the "unicorn"; it may be supposed that two horns are in fact intended, one behind the other, but it has been recalled that both Ktesias and Aristotle ascribed the unicorn to India and called it the Indian ass. In front of the beast is always a curious object which occurs in association with no other animal: a "standard" consisting of a bowl or table-top(?) on a central post, carrying a cage-like object under the nose of the animal. The significance of this object is unknown. It has been suggested that the cage-like object was in fact a bird-cage, but it may be doubted whether, if so, the ancient artist could have refrained from indicating the bird within. It has been designated a "sacred manger" or "sacred brazier"; an incense-holder may in fact have been intended. Whatever be the explanation, the ritual character of the scene is emphasized by a remarkable seal-impression from Mohenjo-daro showing a figure of a "unicorn" being carried in procession between two other objects, one of which was evidently a "standard" of the type under discussion.¹ On the other hand, under the nose of a "unicorn" on a cylinder-seal from Ur, either Indian or made under Indian influence, the "standard" is replaced by the "fish"-sign from the Indus script.² The "standard" itself scarcely occurs outside the Indus civilization, but may be recognized on a potsherd from Mehi in southern Baluchistan showing a typical Kulli-Mehi bull tethered to one.³

c17.112

¹ Marshall, III, pl. cxvii, 9.

c17.113

² C. J. Gadd, "Seals of Ancient Indian Style found at Ur", *Proc. Brit. Academy*, XVIII (1932), 8 (seal no. 7).

c17.114

³ A. Stein, *An Archaeological Tour in Gedrosia* (Arch. Surv. of India, 1921), pl. xxx, Mehi, ii, 4-5; and Piggott in *Antiquity*, XVII (1943), 17.

c17.115

Next in popularity is the short-horned bull, probably the Indian bison or gaur, with wrinkled neck and lowered head twisted slightly towards the spectator. Beneath the nose is an object suggesting a manger. An unstratified square seal bearing this type but with a cuneiform inscription which has not been interpreted was found at Ur,⁴ and four or five circular seals bearing the same device with Indus script have come from the same site, whilst yet another, in the British Museum, is from an unrecorded site also in Babylonia.⁵ These Mesopotamian seals will be considered later (p. 84).

c17.116

⁴ C. L. Woolley in *Ant. Journ.* VIII (1928), 26; S. Smith, *Early History of Assyria* (1928), p. 50; C. J. Gadd in *Proc. Brit. Academy*, XVIII, 5.

c17.117

⁵ Gadd, as cited.

c17.118

The buffalo, with its large swept-back horns, is rarely represented, but the Brahmani bull or zebu, with hump and heavy dewlap, occurs fairly abundantly, and its pronounced muscularity and dignified stance inspired the stone-cutter to his most masterly efforts. The one-horned rhinoceros is not a common type, but its angry, beady eye and hide are rendered with an observation and actuality that remind us of its physical survival in the Himalayan foothills at least until the sixteenth century A.D. Curiously, it is represented with a “manger” similar to that associated with the short-horned bull, presumably implying veneration. The tiger, too, is represented with a “manger”; the emphatic stripes and lowering head are again based on direct knowledge, as well it might, for the tiger survived in Sind into the nineteenth century. Four seals show the tiger looking backwards and upwards at a man sitting in a tree, which he holds with one hand whilst he extends the other. There is no indication that this is a hunting-scene, and a religious interpretation is more probable, though in what sense can only be surmised. On one seal at Mohenjo-daro the “manger” is placed in front of an elephant, which appears to be feeding out of it; otherwise, this animal is represented without adjuncts, and differs from most other seal-animals in being shown in a walking attitude. The species is probably that of the existing Indian elephant, though certain discrepancies in detail have been noted.¹ Two or three seals represent an antelope; one from Mohenjo-daro, bearing two admirably rendered crouching animals, is probably an intruder from Elam or Mesopotamia.² The fish-eating crocodile or gharial occurs on a number of seals, its scaly hide represented by hatching or dots. A seal of Harappan type from Ur shows a scorpion, but this animal has not with certainty been found on seals from the Indus valley itself. A fragmentary seal from Harappā appears to represent a hare. Finally, a double-sided lozenge-shaped seal with stepped edges from Harappā³ bears on one side a cross and on the other a splayed eagle with the head turned to the left and seemingly a snake above each wing. The motif is reminiscent of spread-eagles found in the environs of Mesopotamia; thus it occurs at Susa on the one side and at Tell Brak in Syria on the other. At Tell Brak a bronze example is dated to *c.* 2100 B.C.⁴ At Susa it may have been the symbol of Nin-Gir-Son, one of the forms of Nin-Ip, the divine hunter.⁵ In an Indian context, it was perhaps a prototype of Garuḍa who, as the vehicle of Viṣṇu, is represented flying with a snake in his beak.

c17.119

¹ Marshall, II, 388.

c17.120

² Mackay, II, pl. C, B.

c17.121

³ Vats, I, 324; II, pl. XCI, 255. Compare a pottery amulet from Mohenjo-daro (Mackay, I, 363; II, pl. CII, 15), and a devolved eagle-like form on a circular seal of the post-Harappan Jhukar culture at Chanhu-daro (Mackay, *Chanhu-daro*, pl. L, 15a).

c17.122

⁴ M. E. L. Mallowan in *Iraq*, IX (1947), 171 and pl. XXXII, 5. I am greatly indebted to Professor Mallowan for drawing my attention to this reference.

c17.123

⁵ *Délégation en Perse, Mémoires*, XII (1911), 138-9.

c17.124

The likelihood that the seal-animals are in most or all instances religious devices is thus suggested by the character of the “unicorn” and its accompanying “standard”, by the offering of food or incense to the bison, elephant, rhinoceros and tiger, and possibly by the splayed eagle. A series of composite animals emphasizes this inference. A recurrent monster has the face of a man, the trunk and tusks of an elephant, the horns of a bull, the forepart of a ram, and the hind quarters of a tiger with erect tail which is in one instance armed with claws.¹ On one seal the beast appears to have three ornamental collars. A three-headed animal on a seal from Mohenjo-daro has the heads of antelopes and the body of a "unicorn". Another shows six animal heads—"unicorn", bison, antelope, tiger, the remaining two broken—radiating from a ring, and recalling a whorl on another seal from the same site with a single "unicorn" and five featureless lobes. Another represents three animals, probably tigers, centrally superimposed rather than composite. On yet another seal, two "unicorns'" heads branch symmetrically from the base of a pipal tree. These various monstrosities sufficiently indicate the range of the series; of them all, the first is the commonest, and is probably represented also by a fragmentary animal-sculpture mentioned above (p. 65).

c17.125

¹ An archaic Sumerian seal bearing a bull with an elephant's trunk is apparently out of context in Mesopotamia but its relationship, if any, with the Indus series cannot be conjectured. H. Frankfort, *Cylinder Seals*, p. 307.

c17.126

Human figures, whose summary depiction on the seals in comparison with the skilful animal-forms recalls a similar disparity in the Cave Art of Western Europe, are evidently in most cases either divine or engaged in religious ritual, though in rare instances the intent may be purely secular. Into the last category perhaps fall a linear representation from Mohenjo-daro apparently representing a man working a shadoof or water-raiser, and more doubtfully the man (hunter?) in a tree above a tiger (p. 78). Scenes in which a buffalo is confronting half a dozen prostrate human figures, and another in which a man appears to be vaulting, somewhat in Minoan fashion, over a bull,² may represent hunting scenes or may have a more symbolic significance. The former scene has been compared with one in which, on Ist Dynasty slate palettes in Egypt, the king as Strong Bull gores a prostrate enemy. Another seal from Mohenjo-daro also shows a man (or god) attacking a buffalo with a barbed spear,³ a scene which recalls the attack on Dundubhi by Śiva and other gods with a trident.

c17.127

² Mackay, I, 336, 337, 361; II, pl. XCVI, 510 and CIII, 8.

c17.128

³ Ibid. I, 336; II, pl. LXXXVIII, 279.

c17.129

But no doubt arises as to the divinity of a remarkable figure on three seals from the same site.⁴ The figure is represented as seated either on the ground or on a low stool. In two instances the head is three-faced, and in all it bears a horned head-dress with a vertical central feature. The arms are laden with bangles from wrist to shoulder, after the fashion of the left arm of the dancing-girl (above, p. 67), and there is a girdle or waist-cloth. On one of the seals, the figure is flanked on its right by an elephant and a tiger and on its left by a rhinoceros and a buffalo, whilst below the stool are two antelopes or goats( pl. XXIII). Marshall recognizes in the figure a prototype of Śiva in his aspect as Paśupati, Lord of Beasts.

c17.130

⁴ Marshall, I, 53; III, pl. XII, 17; Mackay, I, 335; II, pls. LXXXVII, 222, 235 and XCIV, 420.

c17.131

Of other figure-seals, the most elaborate is again from Mohenjo-daro. It shows a deity (god or goddess?), with flowing hair and horns flanking a central feature as on the “Śiva” seals just mentioned, standing nude between the branches of a pipal tree, before which kneels a worshipper apparently with similar hair and head-dress. Behind the worshipper stands a human-faced goat, of a type occasionally seen on seals, and below are seven clothed ministrants or votaries (?) with long pigtail and tall head-dress, perhaps engaged in a ritual dance. The whole scene is repeated, less clearly, on another seal from the site,¹ and a part of it on a seal from Harappā,² whilst the seven “votaries” occur on another broken seal from the same site. Another repeated scene on Mohenjo-daro seals shows a standing human figure with knobbed hair and outstretched arms holding back two rearing tigers:³ a composition recalling one characteristic of the Sumerian Gilgamesh and his lions, with which it is doubtless related. A tiger, with the addition of horns, appears on another seal in a mythological scene where the animal is being attacked by a “minotaur” or bull-man reminiscent of the Sumerian Eabani or Enkidu whom the goddess Aruru created to combat Gilgamesh.⁴ The semi-bovine monster or god occurs also on other seals and may be related to the horned deity already mentioned.

c17.132

¹ Marshall, III, pl. CXVIII, 7.

c17.133

² Vats, II, pl. XCIII, 316; cf. pl. XCI, 251.

c17.134

³ Mackay, I, 337; II, pls. LXXXIV, 75, 86, LXXXV, 122 and XCV, 454.

c17.135

⁴ Marshall, I, 67; II, pl. CXI, 357.

c17.136

Two other crude figure-sealings may be added, both from Harappā. One shows on one side a central squatting “Śiva” with a blurred group of animals on his left and the motif, already described, of a tree above a tiger on his right; on the other side a bull and a standing figure in front of a wooden structure, possibly with a second figure seated at its entrance. The second sealing bears on one side a central group of pictographs with two rearing and confronting animals (probably tigers) on one flank and, on the other, a nude woman upside down giving birth to what has been interpreted as a plant but may equally be a scorpion or even a crocodile; whilst on the other are a repetition of the pictographs and a scene representing a man with a curved knife in one hand and an uncertain object in the other, approaching a woman seated on the ground with upraised arms and dishevelled hair, possibly, as has been suggested, a scene of human sacrifice.⁵ Once more, the inadequacy of the seal-cutters in the representation of the human form is very noticeable; the interest of these figure-seals lies in their obscure subject-matter rather than their ingenuous artistry.

c17.137

⁵ Marshall, I, 52, pl. XII, 12; Vats, I, 42, 129; II, pl. XCIII, 303-4.

c17.138

It would be of interest to know a great deal more than we know at present about the chronology of the Indus seals. Mesopotamian contacts will be discussed later (p. 84). Meanwhile, it is only at Harappā that some hint of a sequence has been recovered in the Indus valley itself. There, in Mound F, the excavator observed that in the lower levels seals of what may be described as the normal type gave place to miniature seals measuring from 0·7 to 0·36 in. in length, from 0·6 to 0·2 in. in width, and from 0·13 to 0·05 in. in thickness. At the lowest level reached (the "sixth stratum") seals of this category were the only ones found.¹

c17.139

¹ Vats, I, 324.

c17.140

These tiny seals have no knob or hole, and do not bear the "unicorn" or other major animals of the larger series. Most of them have a line of roughly scratched pictograms on one side and a symbol resembling VII, VIII, VIIII, IIV, IIIV, or IIIIV on the other. Occasionally a crocodile or a fish is shown, and, more rarely still, a goat or a hare, whilst four examples bear a "standard" or incense-burner of the kind associated with the "unicorn". Whether these distinctive little seals were related to other cultural variations is not recorded but is worth further investigation by careful digging.

First 140 of 143 passages. Read on

The Indus Civilization (Cambridge History of India, Supplementary Volume) c17: Arts and crafts

Arts and crafts

Linked texts (6)