Pramāṇa · Search · About · API

The Development of Hindu Iconography c18.s2: Śakti

The Development of Hindu Iconography c18.s2

Language
English

Contents

Śakti

c18.s2.1

The worship of the female principle can be traced in India, as in many other ancient countries of the world, to a very remote past. Many scholars have suggested that the cult of the Mother goddess existed in some form or other among the early Indus Valley people. Pottery images of the goddess have been found in course of the excavations in the sites, and Mackay is of opinion that they ‘were kept almost in every house in the ancient Indus cities, probably in a recess or on a bracket on the wall’.¹ The early Indus Valley settlers appear also to have worshipped her in her aniconic form. Many ‘ringstones’ discovered in the sites can justifiably be described as cult objects symbolising the Mother aspect of the goddess, if they are studied along with the phallic objects found there, which symbolised the father aspect of the god also worshipped in the Indus Valley. Mention has already been made in a previous chapter of this book (*supra*, pp. 170-73) of the ornamental stone rings and discs of the Maurya and Śuṅga periods, the association of which with the cult of the Mother goddess can be sufficiently demonstrated. These very ancient finds can profitably be compared with the *cakras* and *yantras* of more modern times, which were utilised by the Śāktas in the ritualistic worship of the Great Mother. The nude female figures very often shown in these ornamental ‘ringstones’ and ‘discs’ are almost invariably associated with plants and vegetation (sometimes with men and animals), and their parallels can be found among both the Indus Valley remains and the remains of the Gupta period. On one oblong terracotta sealing found at Harappa appears a nude female figure upside down with legs wide apart and with a plant issuing from her womb; an early Gupta terracotta sealing shows a goddess with her legs in much the same position, but with a lotus issuing from her neck instead of from her womb (*cf. supra*, p. 167). This idea about the association of the vegetation with the goddess is very well worked out in her Śākambharī aspect (*Mārkaṇḍeyapurāṇa*, Devīmāhātmya, 91, 48-9). This association is still emphasised in the Navapatrikā ceremony of the autumnal Durgā worship in Bengal, which shows that the Devī was in a way the personification of the vegetation spirit.¹ She is the mother *par excellence* sustaining her children, the men and animals of the universe, with food produced from her body. The nuclei of some other concepts about the goddess can be traced to the prehistoric times (*cf. supra*, pp. 166-69), and the early stages of the cult of the Śakti seem to go back to this remote past.

c18.s2.2

¹ Early Indus Civilisations, 2nd edition, p. 54.

c18.s2.3

¹ The Navapatrikā or 'nine plants' are *rambhā* (plantain tree), *kacvī* (Arum colocasia), *haridrā* (turmeric plant), *jayantī* (barley), *bel* (wood-apple), *dāḍima* (pomegranate), *aśoka* (Jonesia Aśoka), *māna* and *dhānya* (paddy). The particular forms of the goddess presiding over the individual plants are Brahmāṇī, Kālikā, Durgā, Kārttikī, Śivā, Raktadantikā, Śokarahitā, Cāmuṇḍā and Lakṣmī respectively; *Puraścaryārṇava*, Pt. III (Benares edition, 1904), pp. 1034-35. These nine forms of the Devī again can be described as comprising a variety of the Navadurgās.

c18.s2.4

Early Vedic ritualism presents to us a different picture. More prominence is given here to male deities, and the goddesses occupy a comparatively subordinate position. But the several female deities found there bring out in a striking manner the inner workings of the Vedic seers' minds. The ancient Vedic Ṛṣis assigned importance to such goddesses as Aditi, the Divine Mother, Uṣas, the goddess of dawn, Pṛthivī, the mother earth, and lastly Vāc, the goddess of speech. Sarasvatī, primarily a river goddess (it was on the banks of this river that the distinctive traits of Vedic culture were formulated), Rātri, the goddess personifying a star-lit night, Purandhi, Iḷā and Dhiṣaṇā, collectively personifying such abstract attributes as abundance and nourishment, were also sung in some hymns of the *Ṛgveda*. But in the sublime conception of Vāc outlined in the Devī-sūkta (*R. V.*, X. 125), is to be found one of the greatest and at the same time simplest expositions of the concept of divine Energy or Śakti inherent in everything,—in gods, men and animals, nay in the universe itself. This hymn as well as the hymn associated with Rātri (*R.V.*, X. 127) came to occupy a very prominent position in the Śākta ritual of subsequent times. Thus, the developed Śakti worship of the epic and Purāṇic times was not a little indebted to these goddess concepts, the very idea underlying the word śakti being based on the central theme of the Devī-sūkta. But it is also true that such names as Ambikā, Umā, Durgā, Kālī, etc., which came to designate, singly or collectively, the central figure of the Śākta cult, do not occur in the Ṛgveda. These names, however, are found in the later Vedic texts. Ambikā appears first as Rudra's sister in the Vājasaneyī Saṃhitā (III. 57) and the Taittirīya Brāhmaṇa (I. 6. 10, 4-5), and then as his consort in the Taittirīya Āraṇyaka (X. 18). The goddess is invoked in the last-mentioned work as Durgā Vairocanī, Kātyāyanī and Kanyākumārī (X. 1, 7). The Kena Upaniṣad (III. 25) refers to Umā Haimavatī as the personified Brahmavidyā (‘the knowledge about the Brahman’). Kālī and Karālī are mentioned in the Muṇḍaka Upaniṣad (I. 2, 4) as two of the seven tongues of Agni, the others being Manojavā, Sulohitā, Sudhūmravarṇā, Sphuliṅginī and Viśvarucī. The number seven is to be noted; the number of the Divine Mothers is usually the same—the Sapta-Mātṛkā. Such names of the Devī as Bhadrakālī, Bhavānī, Durgā, etc., are found in the late Vedic works like the Sāṅkhyāyana and Hiraṇyakeśin Gṛhyasūtras, and in the Taittirīya Āraṇyaka. These data prove clearly that some features of the cult, existing in a nascent stage in earlier times, were gradually taking shape and form of a type well familiar in the subsequent period.

c18.s2.5

The two Durgāstotras in the Mahābhārata (IV. 6 and VI. 23) and the Āryāstava in its supplement (Harivaṃśa, III. 3) outline the various constituent elements underlying the principal cult picture of the developed Śākta cult. The concept of the composite goddess contained in its various elements such as her ‘mother’, ‘daughter’, and ‘sister’ aspects, her Vedic Aryan element (cf. her appellations Āryā, Kauśikī, Kātyāyanī, i.e., ‘the Aryan goddess’, ‘the goddess of the Kuśika and Kātya sage clans’), and last, but not the least, the various non-Aryan strands in her character.

c18.s2.6

The Āryāstava says that ‘she was well worshipped by the Savaras, Barbaras, and the Pulindas’ (*Savarair-Barbarai-ścaiva Pulindaiśca supūjitā*). She is also described in other contexts as Aparṇā (‘not even covered with a leaf garment’, i.e., ‘nude’), Nagna-Savarī (‘the naked Savara woman’) and Parṇa-Savarī (‘the leaf-clad Savara woman’—this is the designation of a Vajrayāna goddess). The Durgāstotras also characterise the goddess as the great saviour who, being prayed to, delivers men from such terrors as captivity, wilderness, drowning, harassment by robbers, great forests, etc. It may be mentioned incidentally that the Mahāyāna Buddhist goddess Tārā is conceived as saving her votaries from ‘eight great terrors’ (*aṣṭamahābhaya*), among which those mentioned above are included (mediaeval images of Tārā from Southern and Eastern India are known in which these *mahābhayas* are illustrated in the *prabhāvali*). The Mahābhārata, thus, gives us a very interesting idea about the composite character of the cult and the cult icon. The Rāmāyaṇa is, however, less indicative of the prevalence of Śakti worship in India, but the comparative paucity of any clear mention of the worship of the Devī here does not indicate anything about the existence of the cult during its period of composition.¹

c18.s2.7

¹ The original Rāmāyaṇa does not mention the worship of the Devī by Rāma when he was in some difficulty about killing Rāvaṇa. In the 106th sarga of the Yuddhakāṇḍa (canto VI) there, the sage Agastya advises Rāma to recite the Ādityahṛdaya, a *stava* to propitiate the Sun God, and after reciting it thrice, the hero could kill the demon. It is only in the Bengali Rāmāyaṇa by Kṛttivāsa, that the worship of the Sun is replaced by the worship of Durgā.

c18.s2.8

Some of the early authoritative Purāṇas, however, fully compensate the paucity of reference to the Śākta cult in the lesser epic. The Devīmāhātmya section of the Mārkaṇḍeya-purāṇa contains the most representative and important of the Purāṇic characterisations of the cult picture. The various Devīstutis there (Brahmā-stuti, Śakrādi-stuti, Nārāyaṇī-stuti, etc.), reveal in a striking manner some of the multifarious strands that contributed to the formation of the concept about the composite cult goddess. The last couplet of the Nārāyaṇī-stuti (Ch. 91), which says that the goddess will as often incarnate herself and kill the enemies of the gods as the Dānavas will cause obstacles and distress to them,¹ distinctly reminds us of the ideology behind the divine incarnation (Avatāravāda) explained in the first few verses of the fourth chapter of the *Bhagavadgītā*. The *stutis* again express in a characteristic manner the ideas about the divine power and energy centering round the Devī, that are so beautifully emphasised in the two great hymns of the *Ṛgveda*, the Devīsūkta and the Rātrisūkta. These verses of the Purāṇas again lay stress upon the various constituent elements that collectively make up the form of the great composite goddess, the central deity of the Śakti cult. The first part of the 82nd chapter of the *Mārkaṇḍeyapurāṇa* also shows in a characteristic manner how the great goddess came out from the accumulated fury not only of Viṣṇu, Śiva and Brahmā, but also of many other gods of the Brahmanic pantheon, when the gods were defeated in the beginning by Mahiṣāsura and his retinue (1-18; the last couplet reads: *Tataḥ samastadevānāṃ tejorāśisamudbhavām | Tāṃ vilokya mudaṃ prāpuramarā Mahiṣārditāḥ*).

c18.s2.9

¹ *Itthaṃ yadā yadā bādhā dānavotthā bhaviṣyati | Tadā tadāvatīryāhaṃ kariṣyāmyarisamkṣayam*

c18.s2.10

The Mother aspect of the Devī is very clearly emphasised in the Jaganmātā or Jagadambā concept so well developed in the Purāṇic Durgāstutis, and the Vedāntins' concept of Māyā also forms a characteristic trait of the goddess as the Mahāmāyā, or ‘Great Māyā’. The Sāṃkhya theory of Puruṣa and Prakṛti is also idealised in the system of the Śāktas in which Śiva representing the former (Puruṣa) always remains passive, while the great Devī, endowed with dynamic activities, symbolises Prakṛti. All these and many other abstract ideologies gradually came to be symbolised by the Devī, the supreme deity of the Śāktas, who resides in the macrocosm as well as in the microcosm. The Śāktas believe in the Kuṇḍalinī Śakti, inherent but dormant in man, which has to be awakened through various yogic and other processes and raised by stages from Mūlādhāra, the lowest lying cakra in the human body to the highest of the cakras there, the Sahasrāra or the Ājñācakra;¹ this process of rousing the Kuṇḍalinī Śakti and carrying it to the highest cognitive centre of the human body is known to the Śāktas as Ṣaṭcakrabheda, and presupposes the difficult and sustained efforts of a Tāntric Sādhaka who, if successful in his efforts, is blessed with the beatific vision of the Devī, and attains salvation. It must be said, however, that this and other esoteric tenets of the Śāktas took time to attain full development, and none of the extant Tāntric texts that expounds these doctrines appear to go beyond the early mediaeval period.

c18.s2.11

¹ The *cakras* are usually known as Mūlādhāra, Svādhiṣṭhāna, Maṇipura, Anāhata, Viśuddha, Ājñācakra or Sahasrāra located respectively in the anal region, the region just above the sex organ, the navel, the heart, the throat and the brain or the forehead of the human body.

c18.s2.12

Faint traces of Tāntricism, however, can be found in the texts and inscriptions of the Gupta period, if not earlier. The very word *tantra* occurs in an inscription of the first quarter of the fifth century A.D. found in the village of Gangdhar (Jhalwar, Madhya Bharat). In lines 22-3 of it, mention is made of the erection of ‘the very terrible abode of the (Divine Mothers), filled full of Ḍākinīs, . . . who stir up the very oceans with the mighty wind rising from the Tāntric rites of their religion.’² Reference has already been made in Chapter VI of this book to the significance of the word *maṇḍalakrama* occurring in the 58th chapter of the Bṛhatsaṃhitā, and the persons, well versed in the *maṇḍala-krama*, entitled to instal the images of the Divine Mothers, might have been Tāntric Śakti-worshippers (*supra*, p. 230). It has also been incidentally mentioned earlier in this book (*supra*, pp. 83-4) that the concept about the Śaktipīṭhas was well known in the 7th century A.D., for it did not fail to attract the notice of the keenly observant Chinese traveller Hiuen Tsang. The famous story of Dakṣa’s sacrifice as narrated in the Great Epic may not be very old, but the *pīṭha* idea which grew out of it was clearly based on the Tāntric concept of the intimate association of Śakti with Bhairava, the terrific aspect of Śiva.¹ The Tīrthayātrā section of the Mahābhārata (Vanaparva) refers to three Śakti Pīṭhas associated with the Yoni and Stana of the goddess. Kuṇḍas or sacred tanks are also their invariable adjuncts, and two Yonikuṇḍas (one situated at Bhīmāsthāna beyond Pañcanada, and the other on a hill called Udyatparvata), and one Stanakuṇḍa on a peak known as Gaurīśikhara (possibly in the Gauhati region) are mentioned there.² The Mahāmāyūrī, a Sanskrit Buddhist text composed in the early centuries of the Christian era, possibly also refers to the shrine of Bhīmā under the name of Bhīṣaṇā and to that of her consort as Śivabhadra in the extreme north-west of India (Journal Asiatique, Vol. XV, 1915, p. 370).

c18.s2.13

² *C.I.I.*, Vol. III, p. 78. The 10yal minister Mayūrakṣaka appears to have had some knowledge about the ritualism connected with the worship of the Divine Mothers, as reference to the Ḍākinīs and the terrific Tāntric rites in the inscription proves.

c18.s2.14

¹ The mythology about Satī, the daughter of Dakṣa-Prajāpati and the first wife of Rudra-Śiva, is closely connected with the ideology behind the pīṭha concept. Satī went uninvited to attend the sacrifice being performed by her father, and died there on hearing him abuse her husband. At this Śiva destroyed Dakṣa's sacrifice, severely punished him and his invited guests, and began to roam aimlessly with the corpse of his wife on his shoulders out of sheer grief for his beloved. Viṣṇu just to cure Śiva of this obsession of grief cut up Satī's body with his cakra and had the limbs scattered over different lands. The places where these severed limbs of Satī fell became Śaktipīṭhas, and such was the great love of Śiva for his dear wife that he, assuming the forms of so many Bhairavas, settled in their vicinity to keep a watch over the parts of his consort's body. The chief objects of worship in these pīṭhas were mainly aniconic.

c18.s2.15

² D. C. Sircar, ‘Śaktipīṭhas’, J.R.A.S.B., Letters, Vol. XIV, pp. 8-9.

c18.s2.16

In the mediaeval period Tāntricism associated with Brahmanical Hinduism as well as Mahāyāna or Vajrayāna forms of Buddhism attained full development in India, the cult being specially predominant in Eastern India. Śakti worship was also popular among many Hindus of the extreme south-west of India, and tradition says that Śaṃkarācārya, the greatest exponent of Advaitavāda, was at heart a worshipper of the goddess. He is said to have composed in her honour the famous Tāntric text Saundaryalaharī, and an authoritative commentary on the ‘One Thousand Names of Lalitā’ (Lalitāsahasranāma). It will not be possible here to give even a very brief account of the tenets of the developed Śakti cult which were expounded in the various extant Tāntric texts of the late mediaeval period. But a brief reference may be made here to the real nature of the Devī as explained in the Devīmāhātmya and its six limbs (ṣaḍaṅgas), the Devīkavaca, the Devīkilaka, the Argalāstotra, and especially the three Rahasyas—Prādhānika-, Vaikṛtika- and Mūrti-rahasyas. It is needless to say that all these six limbs were composed in course of time, sometime after the original Caṇḍī portion of the Mārkaṇḍeyapurāṇa had been written. It is interesting to note, however, that these texts collectively give us an idea about some of the image-types of the goddess in the Gupta and post-Gupta periods. True it is, that to a pious devotee of the Śākta cult, the Śakti is really the formless absolute principle immanent in the whole universe as the supreme and all-pervading consciousness (cf. Mārkaṇḍeyapurāṇa, Ch. 85, 34—Citirūpeṇa yā kṛtsnametad-vyāpya sthitā jagat), but the great goddess also assumes various forms of pacific and terrific character (Saumyāni yāni rūpāṇi . . . . yāni cātyantaghorāni). In the Prādhānika-rahasya of the Caṇḍī we are told that the great and primary goddess Mahālakṣmī, in whom all the three guṇas (sattva, raja, and tama) are manifest, has a four-armed concrete form, in whose hands are placed a citrus, a mace, a shield and a skull-cup, and who has a snake, liṅga and yoni on her head. Mahālakṣmī assumed in the time of dissolution the form of Mahākālī, in whom the tamoguṇa predominated; it is her four-armed secondary form in blue colour, ornamented with a skull-garland, and with a sword, a skull-cup, a severed head and a shield in her four hands. This secondary goddess came also to be known by such names as Mahāmāyā, Mahāmārī, Kṣudhā (the ‘great hunger’), Tṛṣā (‘thirst’), Nidrā or Yoganidrā, Kālarātri and others. Out of the great Mahālakṣmī again emanated the white-coloured tertiary goddess Mahāsarasvatī in whom sattvaguṇa prevailed, and who held in her four hands a rosary of beads, an elephant goad, a lyre and a manuscript. This emanation came also to be known by such names as Mahāvidyā, Mahāvāṇī, Bhāratī, Vāk, Āryā, Brāhmī, Vedagarbhā, etc. From these three forms of Śakti, one primary and the others emanatory, were evolved in turn Brahmā and Śrī, Rudra and Trayī or Vedavidyā, and Viṣṇu and Gaurī. In the Vaikṛtika- and Mūrti-rahasyas the other names and concrete forms or image-types of the great goddess are elaborately described, and these descriptions attempt in a way to outline the deep symbolism underlying her various aspects. It is also interesting to note that in the unfolding of the nature of the Devī in these supplements to the Devīmāhātmya, the other two major Brahmanical cult-gods, Viṣṇu and Śiva and the Vedic Brahmā find a well-recognised, though a subordinate, place. Some of the chapters of the original Devīmāhātmya section of the Mārkaṇḍeyapurāṇa, on the other hand, contain descriptions of the various early forms of the goddess, such as the Mahiṣāsuramardinī, the Mātṛkās, Cāmuṇḍā, and others.

c18.s2.17.1

Gopinath Rao has collected numerous names of the various forms of the goddess and their iconographic descriptions from different Āgamas; but he could illustrate only a few of these forms by extant stone and bronze sculptures of the early and late mediaeval periods. It will be possible here neither to note the textual descriptions of the numerous images of the Devī, nor to study at some length the iconographic features of even an appreciable number of her extant image-types for exigences of space.¹ Emphasis will be mainly laid here on her Mahiṣāsuramardinī and Mātṛkā aspects and a few other allied forms. Mythologically speaking, the Mahiṣāsuramardinī form of the goddess is one of her earliest and most important forms so beautifully delineated in the Devīmāhātmya. Various texts describe different iconic types of the goddess, but the difference mainly lies in the number of arms that are attributed to the Devī. A large number of eight- or ten-armed images of the Mahiṣāsuramardinī have been discovered in Eastern India, and the ten-armed variety of such iconic types endowed with some additional features came to be the accepted model of the composite clay image in the autumnal Durgā worship in Bengal. A study of a few of the early reliefs will enable us to throw some light on the developmental aspect of the iconic motif.

c18.s2.17.2

Some very interesting miniature stone reliefs of the Gupta period depicting the two-armed figures of the Devī engaged in combat with the Buffalo Demon were unearthed by Marshall at Bhita, in which no other accessory figures are shown (A.S.I.A.R., 1911-12, p. 86. Pl. XXXI, Figs. 13 and 14). The beautiful brass image of the four-armed goddess of the time of Meruvarman, king of Chamba (c. 8th century A.D.), has an inscription which describes her as Lakṣaṇā; here the Devī is shown uplifting the hind part of the Demon in the shape of a buffalo by holding its tail with her front left hand, and piercing its neck with a triśūla by her front right, while she tramples on its neck with her right leg (her back hands hold a sword and a bell). This standing attitude of the goddess exactly corresponds to the Devīmāhātmya description of the Devī (III. 37—Evamuktvā samutpatya sārūḍhā taṃ mahāsuram | Padenākrāmya kaṇṭhe ca śūlenainamatāḍayat).¹ The goddess is shown killing the Buffalo Demon almost in the same attitude in a far earlier relief carved on the façade of the Candragupta (II) cave at Udayagiri (Bhilsa, Madhya Bharat). It is curious, however, that in a relief of so early a period the Devī is endowed with as many as twelve arms, the hands holding many weapons and attributes, two back right and left hands stretching probably an iguana (godhā; already noticed, cf. supra, p. 172).

c18.s2.17.3

The relief is very much damaged, but it still shows much animation, and when it was in a good state of preservation it must have belonged to some of the best specimens of Gupta art (Pl. XLI, Fig. 4). It seems that the Mārkaṇḍeyapurāṇa tradition about this mode of attack by the Devī was well known to the artists of Northern and Central India of the Gupta and early mediaeval periods. It is also to be noted that none of these sculptures show the lion mount of the goddess. In the Gangaikondacolapuram sculpture depicting the same theme, the lion is present on the left, but the similar standing attitude of the goddess is shown in a lifeless manner (Rao, *op. cit.*, Vol. I, Pl. CIII). Some early mediaeval sculptors of India began to change the mode of depicting the Devī's fight with the Demon, and the Mahabalipuram and Ellora panels reproduced by Rao (*op. cit.*, Pls. CIV and CV) are two of the remarkable specimens illustrative of the changed mode. The much-mutilated Ellora relief reproduced here (Pl. XLI, Fig. 2) shows the eight-armed Devī riding on her lion mount vigorously attacking Mahiṣāsura, a full-scale man of her stature with buffalo horns; other demons, some fallen and others still fighting, are shown below, and in the two uppermost rows in the panel are the divine onlookers of the fight.

c18.s2.17.4

The intensity of the actual combat is very skilfully demonstrated by the artist who has also not failed to endow the relief with some *genre* interest by the introduction of the expectant divine onlookers in the upper section of the panel. If we compare this very lively panel from Aihole with two sculptures, one from Aihole and the other from Haripur (Mayurbhanj, Orissa), we are confronted with the fact how the earlier traditional mode of the combat between the Devī and the Asura was given a modified form by the sculptor of the western Calukya country on the one hand and the Orissan artist from Mayurbhanj on the other. The Aihole relief (earlier in point of date, c. 6th or 7th century A.D.) shows the eight-armed goddess piercing the upturned neck of the Buffalo Demon (no man comes out of the decapitated trunk of the animal, which is a later feature), her lion mount on the left being a silent onlooker (Pl. XLII, Fig. 3). The Haripur sculpture (a few centuries later than the Aihole one) shows the eight-armed goddess in a more aggressive pose, where the three prongs of the *śūla* pierce the upturned neck of the human Demon issuing out of the decapitated trunk of the animal, the lion also taking part in the fight (Pl. XLII, Fig. 2). The early Calukyan artist appears to lay stress on the easy and effortless grace with which the divine act of retribution was carried out, while the Orissan sculpture portrays with success the dynamic vigour underlying the act.

c18.s2.17.5

The great popularity of this theme of the goddess slaying the Buffalo Demon in distant corners of India can be demonstrated with the help of two interesting objects, one of which is being illustrated here. Plate XLII, Figure 1, shows the figure of Mahiṣāsuramardinī cast from a mould, found at Peshawar; the details are not distinct, but the turreted crown of the goddess, the face and upper part of her body, and the legs and the body of the animal are clear. The second image, unique of its kind, recovered from a North Bengal village in the district of Dinajpur, depicts in a very interesting manner the Nava- (nine) Durgā motif. The central figure of Mahiṣāsuramardinī is eighteen-armed, while eight other sixteen-armed miniature replicas of the same type are grouped round it.¹ How the same theme was given further re-orientation by the Śakti-worshippers in Bengal is illustrated by another unique sculpture found in the same district. The thirty-two arms of the goddess riding on a lion and engaged in combat with demons (not the Buffalo Demon in particular) are meant to emphasise, in however imperfect a manner, the all-powerful and all-embracing character of the Divine Śakti; on the top section of its prabhāvalī are shown the miniature figures of Gaṇapati, Sūrya, Śiva, Viṣṇu and Brahmā.²

c18.s2.18

¹ Rao, op. cit., Vol. I, pp. 338-400 and Plates; cf. also N. K. Bhattasali, op. cit., pp. 198-227 and Plates, and Dacca History of Bengal, Vol. I, pp. 449-55, and plates.

c18.s2.19

¹ J. Ph. Vogel, Antiquities of Chamba State, p. 138, Pl. VII (b).

c18.s2.20

¹ Dacca History of Bengal, Vol. I, pp. 453-4, and Pl. XIII. 35. There are different lists of the names of the Navadurgās. Monier Williams mentions Kumārikā, Trimūrti (?), Kalyāṇī, Rohiṇī, Kālī, Caṇḍikā, Sāmbhavī, Durgā and Bhadrā on the authority of some unnamed lexicon; Rao submits a list on the authority of the Āgamas, which reads Nīlakaṇṭhī, Kṣemaṅkarī, Harasiddhi, Rudrāṃśa-Durgā, Vana-Durgā, Agni-Durgā, Jaya-Durgā, Vindhyavāsi-Durgā and Ripumārī-Durgā. A list in the Devīkavaca of the Devīmāhātmya reads Śailaputrī, Brahmacāriṇī, Candraghaṇṭā, Kuṣmāṇḍā, Skandamātā, Kātyāyanī, Kālarātrī, Mahāgaurī and Siddhidātrī; while some other Purāṇas give the names thus, Ugracaṇḍā, Pracaṇḍā, Caṇḍogrā, Caṇḍanāyikā, Caṇḍā, Caṇḍavatī, Caṇḍarūpā, Aticaṇḍikā, Rudracaṇḍā.

c18.s2.21

² Dacca History of Bengal, Vol. I, p. 454, Pl. 1. 5. The five miniature figures on the top of the prabhāvalī (four representing the four Brahmanical Hindu cults of Gāṇapatya, Saura, Śaiva and Vaiṣṇava, the Śākta cult being indicated by the main image, and Brahmā standing for Vedism) remind us in a characteristic manner the display of the Pañca Dhyānī-Buddhas on the aura of many Vajrayāna images of the mediaeval period.

c18.s2.22

Mention has been made in Chapter IV of this book of the early iconic type of the Devī either accompanied by her lion mount or actually riding on it (supra, pp. 134-35).

c18.s2.23

Some mediaeval reliefs hailing from Bihar and now in the collection of the Indian Museum, show the development of this motif in a characteristic manner. One of them shows the four-armed goddess seated astride on the back of her mount; her back hands carry a sword and a shield, front left a śūla, the front right being in the varada pose. There are the usual flying Vidyādhara figures on the top section of the sparingly decorated elliptical prabhāvalī; though the artistic execution is not of a very high order, yet the relief has a character of its own (Pl. XLIII, Fig. 4). The other sculpture emphasises the mother aspect of Durgā Simhavāhinī seated in lalitākṣepa on the back of her couchant mount; her two right hands and the back left hand are shown in the same manner as in the other Bihar sculpture, while her front left hand clasps her child (probably Kārttikeya) seated on her left lap. Here the art is mainly hieratic, though the image has some grace of its own (Pl. XLII, Fig. 4). The association of the Mother Goddess with alligators (iguana, godhā) has been already commented on in Chapter V of this book (supra, pp. 171-72). One very unique bronze figure of the Devī found at Nalanda is being illustrated here to show the development of this concept about the goddess in the early mediaeval period. There are some very striking features present in this statuette, which require some detailed notice. The three-eyed and four-armed Devī stands in the samapādasthānaka pose, holding in three of her hands a rosary, a hooked staff (triśikha?) and a water-vessel (bhṛṅgāra); there is a creeping iguana near her right leg, and her lion mount and another horned animal (a buffalo?) are shown on the lower section of the pedestal. Heaped naivedya (offering) pots are shown on four corners of the pedestal, and the sun and the moon are placed on either side of the very beautiful and elaborately designed śiraścakra (Pl. XLIII, Fig. 2). The iconic motif of Pārvatī associated with iguana has been found in other parts of India, and many sculptures from Eastern India (specially Bengal) illustrate this aspect of the goddess. A type of her four-armed standing figures of the mediaeval period commonly found in Bengal shows the goddess standing in the *samapādasthānaka* pose having such attributes as a *varamudrā* or pomegranate, a Śivaliṅga, a *triśikha* in her hands, and an iguana (*godhikā*) carved on the pedestal. Different varieties of this type can be collectively described as Caṇḍī on account of the *godhikā* being given a prominent position in the story of Caṇḍī and Kālaketu current in Bengal. The goddess with iguana can have also the appellation of Śrī, one of the six varieties of Gaurī,—Umā, Pārvatī, Śrī, Rambhā, Totalā, and Tripurā,—as described in the *Rūpamaṇḍana* (*Godhāsanāśritā mūrtirgṛhe pūjyā śriye sadā*). She may also be described as Gaurī herself on the basis of another text which enjoins that Gaurī should sit (or stand) on an iguana (*Godhāsanā bhaved-Gaurī*). This motif migrated to Indonesia along with the Mahiṣāsuramardinī motif from Eastern India in early times, and mediaeval replicas of such images have been found there.¹

c18.s2.24

¹ J.G.I.S., Vol. IV, 1937, pp. 137-47.

c18.s2.25

The Devī icons described in the preceding pages portray her independent aspect in some of which her association with Śiva is clearly emphasised. But from a very early period, she was also mythologically associated with Viṣṇu-Kṛṣṇa. The *Durgāstotras* of the *Mahābhārata* and the *Āryāstava* of the *Harivaṃśa* describe her as having been born in the womb of Yaśodā in the house of the cowherd Nanda (*Yaśodāgarbhasambhūtām . . . | Nandagopakule jātām . . .*). The *Mārkaṇḍeyapurāṇa* also gives us the same information (Ch. 91, v. 37), and in its *Nārāyaṇī-stuti* the gods characterise her as ‘the infinitely powerful Vaiṣṇavīśakti’ (Ch. 91, v. 4: *Tvaṃ Vaiṣṇavīśaktiranantavīryā viśvasya bījaṃ paramāsi māyā*). The earliest datable iconographic text (Chapter 57 of Varāhamihira’s *Bṛhatsaṃhitā*) also lays stress on this association in its account of Ekānaṃśā, as many as three varieties (two-, four- and eight-armed ones) being described (vv. 37-39). Its two-armed type should show ‘the goddess placed between Baladeva and Kṛṣṇa, her left hand resting on her hip, the other hand holding a lotus’. The Viṣṇudharmottara gives an identical description of Ekānaṃśā in this couplet: *Ekānaṃśāpi kartavyā devī padmakarā tathā | Kaṭisthavāmahastā sā madhyasthā Rāma-Kṛṣṇayoḥ* (Bk. III, Ch. 85, 71-2). Several mediaeval eastern Indian stone and bronze reliefs of Ekānaṃśā have been discovered, and the principal object of worship enshrined in the main sanctum of the temple of Ananta-Vāsudeva at Bhuvanesvara (Orissa) is nothing but this goddess with Kṛṣṇa and Balarāma on her two sides. It seems that the cult of Ekānaṃśā was of some importance in Eastern India in mediaeval times. Mention may be made in this connection of a very fine inscribed bronze image of Ekānaṃśā originally recovered from Imadpur (Bihar), and now in the collection of the King Edward VII Gallery of the British Museum. The two-armed goddess stands erect on a lotus, with her right hand stretched down in the *varadamudrā*, her left hand holding a mirror, between the four-armed figures of the snake-hooded Balarāma on her right and Kṛṣṇa on her left. It is true that the disposition of her hands does not tally with the textual description quoted above, but the composition as a whole leaves little doubt that she stands for Ekānaṃśā. The bronze image is cast with great skill, and the model from which it was cast was prepared with a fine sense of artistic balance and proportion.¹

c18.s2.26

¹ J.R.A.S.B., Letters, Vol. XVI, No. 2, 1950, pp. 247-51, Pl. XII. This image was made during the reign of the Pāla king Mahīpāla.

c18.s2.27

The number and the names of the Mātṛkās vary in different contexts. They are usually seven, though eight or more of them can also be counted. The Gangdhar stone inscription of the time of Kumāragupta I, already referred to, does not say anything about their exact number, nor does it contain their respective names. The number of the Mothers is, however, given in the preamble of the inscriptions of the early Cālukya kings who are described there as the descendants of Hāritī and nurtured by the seven Mothers (*Hāritī-putrāṇām sapta-Mātṛbhirabhivardhitānām*), though the respective names of these Mothers are not given there.

c18.s2.28

Varāhamihira also is silent about the number and names of these goddesses, and simply says that ‘Mothers are to be made with cognisances of the gods corresponding to their names’ (*Br̥hatsaṃhitā*, Ch. 57, v. 56). Utpala while commenting on this passage, names them as Brāhmī, Vaiṣṇavī, Raudrī (Māheśvarī), Kaumārī, Aindrī, Yāmī, Vāruṇī, and Kauberī in the first instance; then he says that there are other Mātṛgaṇas like Nārasiṃhī, Vārāhī, and Vaināyakī, which are to be understood (*Evamanyāsāṃ Nārasiṃhī-Vārāhī-Vaināyakīnāmapyuhyam*). It appears that side by side with the common concept about the stereotyped number of the Mothers or Śaktis of seven or eight gods there existed a belief about the Śaktis of other gods or their aspects. In the subsequent ritual literature of the Hindus, the names of as many as sixteen Mātṛkās with Gaurī in the beginning are mentioned (*Gauryādi-Ṣoḍaśa-Mātṛkā*). The Mārkaṇḍeyapurāṇa lays down in one verse that the Śaktis of Brahmā, Śiva, Kārttikeya, Viṣṇu and Indra came out of the bodies of the respective gods, and went to help the Devī in her fight with the Asuras (Ch. 88, v. 12). In the next couplet the Purāṇakāra generalises their form in this way, ‘the Śaktis of the individual gods (are each characterised by the) respective forms, ornaments and mounts of those gods (*Yasya devasya yadrūpaṃ yathā bhūṣaṇavāhanam | Tattadeva tacchaktiḥ . . . . .*). In the verses following are described successively the iconographic features of Brahmāṇī, Māheśvarī, Kaumārī, Vaiṣṇavī, Vārāhī, Nārasiṃhī and Aindrī (vv. 14-20). In verse 22 we are told that ‘from the goddess’s body there came forth Caṇḍikā’s Energy (Caṇḍikāśakti) most terrific, exceedingly fierce, howling like a hundred jackals’. The emanating goddess came to be called Śivadūtī, for she appointed Śiva himself as her messenger to the Asuras, Śumbha and Niśumbha. The number of the emanatory Śaktis in this context is eight, seven in the first lot in which Nārasiṃhī takes the place of Cāmuṇḍā, while Caṇḍikāśakti Śivadūtī is the eighth one. But we know from the preceding chapter of the Purāṇa that Kālī who came out of Ambikā’s forehead furrowed with wrath against the demons Caṇḍa and Muṇḍa, the mighty Asura generals of Sumbha, was given the name of Cāmuṇḍā by the Devī, for Kālī killed these mighty demons in battle and brought their heads to her (Ch. 87, v. 25—*Yasmāc-Caṇḍañca Muṇḍañca gṛhītvā tvamupāgatā | Cāmuṇḍeti tato loke khyātā devī bhaviṣyasi*). But the usually accepted list supported by iconographic data consists of Brahmāṇī, Māheśvarī, Kaumārī, Vaiṣṇavī, Vārāhī, Indrāṇī and Cāmuṇḍī, though there are some variants.

c18.s2.29.1

The Sapta Mātṛkās are often carved in relief on a rectangular stone slab in the order given above with the figures of Vīrabhadra and Gaṇeśa on either side. Stone and bronze sculptures are also not unknown in which the number of the ‘Mothers’ are only three, and they are usually Brahmāṇī, Kaumārī and Vaiṣṇavī. One such fine bronze composition, a companion to the Ekānaṁśā bronze just described (originally found at Imadpur and now in the King Edward VII gallery of the British Museum) shows these three goddesses seated between Vīrabhadra and Gaṇeśa.¹ The Rūpamaṇḍana while describing Vīrabhadra’s image, says that ‘he should be placed before the Mothers, the latter coming in the middle, and Gaṇeśa in the end’ (*Vīreśvaraśca . . . . . Mātṛnāmagrato bhavet | Madhye ca Mātaraḥ kāryā ante teṣām Vināyakaḥ*). Separate reliefs of the Mātṛkās and their ‘guardians’ are also common, and those that are being reproduced here, all belong to this category. The sacred tank at Puri (Orissa) known as Mārkaṇḍeya Sarovara has on one of its banks these separate images placed side by side inside a verandah, and five of the Mātṛkās of this group are now being described and illustrated. Four-faced (the fourth face on the back is not visible) and four-armed Brahmāṇī is seated in the *ardhaparyaṅka* pose with her swan mount beneath the couch, with a child on her left lap;

c18.s2.29.2

the front right hand is in the *vyākhyāna* pose, the front left holding the child (the objects in the back hands are not distinct in the plate, they may stand for a sacrificial ladle and a manuscript or a rosary). The figure does not wear any ornaments, but the sacred thread and the jaṭāmukuṭa are prominently shown (Pl. XLIII, Fig. 1). Māheśvarī seated in the same attitude with her bull mount beneath the couch wears many ornaments; her back hands are broken, the front right hand appears to have been in the abhaya pose, while the front left holds the child on her left lap, which is gone (Pl. XLIII, Fig. 5). Four-armed Kaumārī similarly seated is shown with the child on her lap and her peacock mount beneath her couch; her two right hands are broken, the back left hand holds an indistinct object, the front left clasping the child. She wears many ornaments, but the śikhaṇḍaka mode of her hair arrangement reminds one of the similar coiffure of the Puri Kārttikeya already described in Chapter IX of this book (Pl. XLIII, Fig. 3). The terrific nature of Vārāhī is emphasised by the two rows of the curled locks of hair rising upwards, the kapāla in the back left hand, and her tusks; the child on the lap (now broken) emphasises her Mātṛkā aspect, and her buffalo mount is shown below her couch (Pl. XLIV, Fig. 3).

c18.s2.29.3

A recent acquisition from a village in the district of Hooghly for the Asutosh Museum, University of Calcutta, shows one four-armed form of the Mātṛkā in question, with a fish in one of her right hands; this has not yet been published. The fish in one of the hands of Vārāhī is unique, and perhaps indicates some Tāntric trait (fish, matsya, is one of the five 'ma's, i.e., pañca-makāra). The four-armed Indrāṇī with the elephant beneath her couch is more pleasing to look at; her front right and back left hands are gone, the back right holds a vajra, the front left clasping the child on her left lap. She also wears many ornaments, and her royal head-gear and the ear-ornaments are worth noticing (Pl. XLIV, Fig. 4). Compared with these stiff hieratic sculptures from Puri, the figure of the four-armed Vaiṣṇavī with Garuḍa beneath her seat, which hails from Khiching (fixed in a niche in the Khandiya Deul there), is a real object of art. The goddess gracefully holds the *cakra* and *śaṅkha* in her back hands, has her front right hand in the *abhaya* pose and clasps the child with her front left; the several ornaments including the artistic crown are tastefully displayed, and the mellow introspective look on her face has been beautifully shown by the artist (Pl. XLIV, Fig. 1). The two reliefs that are now being reproduced illustrate in a characteristic manner the Indian idea of the terrific, weird and uncanny in the realm of iconographic art.

c18.s2.29.4

Both are from Jajpur (Orissa) which is regarded as the Virajākṣetra, an old seat of Tāntric worship. The first of these reliefs depicts the most fearful goddess Cāmuṇḍā. She has four arms, emaciated body and shrunken belly showing the protruding ribs and veins, skull-garland (*muṇḍamālā*), her corpse seat (*pretāsana*, but the dead body has its hand in the *añjali* pose), bare teeth and sunken eyes with round projecting eye-balls, bald head with flames issuing from it. She holds in her back hands a *kartṛ* (chopper) and a *śūla*, while her front right and left ones hold a *kapāla* and a *muṇḍa* (human head) respectively; the skull on the armlet on her right hand has a grinning smile on its face, while the severed head in her left hand has a life-like expression. The Orissan artist has skilfully produced one of the most terror-striking images, not a lifeless fetish of an uncultured people, but a concrete representation of the esoteric symbolism underlying one aspect of the Tāntric faith (Pl. XLIV, Fig. 5). The second sculpture depicts the Danturā form of the dire goddess Cāmuṇḍā (it does not represent Śivadūtī as it has been described in the records of the Indian Museum) and shows the two-armed emaciated Devī sitting on her haunches with long distended ears, lean pendulous breasts and projecting ribs, an evil cruel smile lurking in her broad bare face; the mocking and ghastly expression of the whole face is further emphasised by the way in which the eyes are shown.

c18.s2.29.5

All these features endow the sculpture with a character, and prove that the artist has been able to portray in a remarkable manner the weird and the uncanny in Indian sculptural art (Pl. XLV, Fig. 1). Some ugra types of Devī images are in the Varendra Research Society's Museum at Rajshahi, but the one which comes very close to the Jajpur sculpture is in the collection of the Bangiya Sahitya Parisad Museum, Calcutta. This image was originally recovered from Aṭṭahāsa (Burdwan, Bengal), one of the fifty-one Śaktipīṭhas in India according to one enumeration of these pīṭhas (Dacca History of Bengal, Vol. I, p. 455, Pl. XIV, Fig. 36). A third relief originally hailing from Sutna (Madhya Bharat) and now in the collection of the Indian Museum, Calcutta, depicts an eight-armed (most of the hands are gone) image of Nārasimhī seated in the ardhaparyaṅka pose on the back of a stylised lion. The goggle-eyed lion-face of the goddess and the face of the lion mount opened wide as it were for a mighty roar endow the composition with a grotesque character, this grotesqueness being partially relieved by the modelling of the body of the main image (Pl. XLIV, Fig. 2). It may be mentioned in this connection that most of the Śakti images that have been described above have corresponding texts generally supporting their iconographic features.

c18.s2.30

¹ There is little doubt that these two Imadpur bronzes were cast by the same artist, for the fine technique of modelling and casting in both of them is identical: J.R.A.S.B., Letters, Vol. XVI, No. 2, p. 251, Pl. XIII.

c18.s2.31

It is now a relief to turn to the study of a very remarkable sculpture created by the genius of an unknown Bengali artist of mediaeval times, to satisfy the religious needs of a pious worshipper of Śiva and Śakti in a corner of Bengal. Bhaṭṭasali says, ‘The unique image was discovered in the ruins of Vikrampur, within the limits of the ancient capital of the Senas and their predecessors, in the quarter known as Kagajipara’ (op. cit., p. 192). Four feet in height it shows in its lower part a well-carved Sivaliṅga, from the top of which emerges the half-length figure of a four-armed goddess with her front hands in the dhyāna-mudrā, the back hands carrying a rosary and a manuscript. The Devī is profusely ornamented, and her beautifully carved youthful face with three eyes has a serene meditative expression which is lacking in most of the Devī images described above (Pl. XLV, Fig. 2). There has been some difficulty about the correct identification of the image, but whatever may be its real character, it cannot be interpreted in the way in which it has been done by R. D. Banerji. He finds in it an evidence of gradual blending of ‘Buddhist Tāntrism with Hindu Tāntrism,’ and he suggests that ‘the goddess (evidently Pārvatī) is in co-itus with the phallus’. Bhattasali’s tentative explanation of this sculpture as Mahāmāyā on the authority of the Prādhānika-rahasya of the Mārkaṇḍeyapurāṇa and a Kālikāpurāṇa passage (Ch. 76, 83-93) is far more acceptable. The Purāṇic descriptions of Mahāmāyā, it is true, do not fully tally with this unique composite icon, but one main peculiarity of the goddess as described in the latter text that she ‘rent open the Śivaliṅga and came out’ agrees with the general outline of the sculpture. The ideology about Ādyāśakti, the source of all the manifestations of the universe, coming forth from the eternal Puruṣa, can be traced to a Bhagavadgītā passage (XV, 4: Tameva cādyam puruṣam prapadye yataḥ pravṛtti prasṛtā purāṇī). The goddess Tripurā-Bhairavī, as we know from another passage in the Kālikāpurāṇa, is the same as Mahāmāyā, and she holds a rosary and a book according to the description given here, and this also appears to support Bhattasali’s identification.¹ The esoteric symbolism underlying this fine sculpture is of a sublime character, and this is a unique mode of representing Śiva and Śakti together, as the Elephanta sculpture already noted and the Ardhanārīśvara motif to be noticed in Section II of the next chapter are other ways of combining the primaeval parents of the Universe, Pārvatī and Parameśvara, Śiva and Śakti, in one composite form.

c18.s2.32

¹ For R. D. Banerji’s interpretation, cf. A.S.I.A.R., 1924-25, p. 155, E.I.S.M.S., p. 100. For Bhattasali’s explanation, cf. Iconography of Buddhist and Brahmanical Sculptures in the Dacca Museum, pp. 192-94.

Cites (23)