Prof. Dr. Klaus Bruhn
The essay was published in Berliner Indologische Studien No. 13/14. 2000, pp. 273-337.
§ 3. Kubera and Ambikā (K-and-A)
As previously indicated, Kubera-and-Ambikā form, on the whole, a triple subject: both together (K-and-A), Kubera alone (“Kubera independent“), and Ambikā alone (“Ambikā independent“). We start with the emergence of Ambikā and Kubera from earlier iconographic levels.
[A: Ambikā] The derivation of K-and-A from Hindu iconography has already been touched upon (two motifs supra). This is what Debala Mitra has to say in connection with the Jaina Ambikā:
“The name Ambikā and the mount lion are clear indications of her borrowal by the Jainas from the Brahmanical pantheon. Her holding of a child also points to the conception of the Mātṛikās. At the same time, the figures of Hārītī, the yakshī of Rājagṛiha converted by Buddha, have a close iconographical resemblance with the icons of Āmrā. The Jaina conception of this divinity is, thus, an amalgam of different ideas“ (Mitra Ac: 44).
The goddess with child and lion is clearly Hindu. In Bihar she is shown
The goddess with child and lion is, however, not a type in its own right but only that part of the material (Hindu/Buddhist Devī in South Bihar) which is relevant to our discussion. Outside Bihar we find the Hindu goddess with child and lion in various forms - Joshi Mā: fig. 48 and p. 86; Williams Gu: fig. 80 and p. 73; Misra Mo: p. 100 and pl. 50.
In the present context it will suffice to say that the Jaina Ambikā (see § 6 for early and late forms) was basically derived from a Hindu goddess with child and lion.
Whether the Buddhist Hārītī exercised some general influence in this or that direction is a matter of speculation. Three points seem important, however.
If we extend the scope of the survey of the Bihar (i.e. South Bihar) Devī, we are faced with numerous possible combinations:
This scheme is no research project but a guide for studies in the iconography of the Bihar Devī, her derivatives and a wider circle of female goddesses. The goddess-with-sword-and-shield is one of the prototypes of the sixteen mahāvidyās (Chandra We: figs. 22=40, 29-30, 32; JID: fig. 63).
For the general (interreligious) archaeological evidence in Central India (Pārvatī, Saptamātṛkās and Jaina Ambikā) refer to Misra Mo: 99-104, 114-116, and 142-144. An extensive discussion on the origin of Ambikā, Iranian and Indian, has been undertaken by U.P. Shah (JRM: [215], 257-264 and Shah Or); see also Tiwari Am: 16-18.
In Jaina literature, Ambikā has a complex history or pre-history. There was a Vidyā known under different names (mainly Ambā-Kuṣmāṇḍyā), and there was a popular goddess Bahuputrikā. All this material had no influence on Jaina art. There were finally Ambikā legends, based on contemporary iconography. All the testimonia have been collected by M.N.P. Tiwari (Am: 18ff. and 31-33) and also by U.P. Shah (Am: 147-148; JRM: 246-247).
So far we have treated “the“ Jaina goddess Ambikā without reference to special developments. The discussion will be continued after [A-C]. For the archaeological material (Jaina Ambikās of every description) refer to Shah Am: passim, JRM: 247ff, Tiwari Am: passim, and Misra Mo: 142-144. [End of Ambikā.]
[B: Kubera] The case of the Jaina Kubera is not strictly analogous to that of Ambikā. There is no literary background, and independent images are rare. Moreover, Kubera is even closer to the Hindu prototypes than Ambikā.
Kubera may belong to the “real pantheon“ (§ 6) or not, the limited material makes it advisable to mention independent Kubera images only in the present section. The combinations with Kubera (Ambikā and others plus K) are treated below and in § 5 (sundry cases).
There are four features which connect the Jaina Kubera (JK) with the Buddhist Pāncika (P) and with the Hindu Kubera (HK):
As far as frequency values (§ 7) are concerned, we can say that the Jaina Kubera (JK) is almost always pot-bellied, almost always carries a citron and a money-bag (mongoose) and holds in rare cases a drinking cup. We ignore the deviating frequency values of the two other types (P and HK). Refer for all three types of Kubera to Misra Ya: 67-71 (general archaeological evidence) and for pot-bellied dwarfs to Misra Ya: 121-125 (archaeological evidence). Typical images of the Hindu Kubera have been published in Misra Ya: pls. 5ff. Refer for śaṅkha and padma (further symbols of richness) also to Sivaramamurti Mi: p. 11 and pl. 5, fig. 13 (left and right) and to two panels at Ellora (PJA: figs. 168 and 169).
We add a few features concerning the Jaina Kubera. The elephant vāhana is not standard, but examples are not rare (K-and-A, elephant and lion). The tree, in the first place typical of Ambikā, occurs at Gwalior (fig. 6) and Ellora (Zimmer Ar: pl. 242). In some cases Kubera has a slender body, in other words the pot belly is missing (figs. 9 and 11, JID: figs. 88, 136). In two cases, he holds a cup (Shah Ak: fig. 38a and JID: fig. 88), on one occasion he has instead of (in addition to?) the money-bag a big jar (fig. 6), on another occasion he has in place of the money-bag a big cornucopiae with two jars (fig. 9).
In Northern India, more or less identical Jaina Kuberas (independent Kubera images) and Hindu Kuberas existed side by side (Misra Ya: figs. 11-15). Truly independent images of the Jaina Kubera are rare, but the semi-independent Jaina Kuberas at Ellora and in Karnataka are de facto independent images. Refer for Ellora Kuberas in general to Pereira Mo: 120, 124, 126 (?), 130, etc. Ellora has also representations of Kubera without Ambikā (two panels supra; Kubera chaumukha [?] PJA: fig. 166).
Literary sources for Kubera are missing completely. U.P. Shah's identification of the Jaina Kubera as yakṣaḥ Sarvānubhūtir (Shah Fe: 3-4; JRM: 214) has been adopted by several authors but cannot be accepted by us. [End of Kubera.]
[C: Kubera and Ambikā] The combination “Kubera to the left, female to the right“ is not rare in early Hindu iconography. Among the sculptures published by N.P. Joshi is one where Kubera appears to the left and a goddess carrying a child to the right (Joshi Mā: drawing 7). A study of such prototypes is, however, not necessary in the present context. [End of A-C]
After the prolegomena [A-C] we proceed to the main part of our discussion. The situation is complex. We have the following forms and phases: Hindu prototypes (supra), the independent Ambikās (early and late: § 6), the independent Kuberas (supra), prototypes of K-and-A (supra), K-and-A as figures of the Jina image (infra), K-and-A as semi-independent figures (Ellora and Karnataka: infra), possible derivatives from K-and-A as part of the Karnataka material (infra), the Jaina Ambikā as part of the system (§ 4), Kubera and Ambikā in semi-systematic representations (§ 5). Monographs on (i) Kubera and (ii) Ambikā which avoid the troublesome fragmentation (but demonstrate the unity and diversity) would be a welcome addition.
Starting with what we call the standard form of K-and-A (citron, money-bag; mango-bunch, child; K and A seated), we mention the following variables: presence and absence of the vāhana.s, presence and absence of the tree above Ambikā, and “varying relations“ of the two deities to the Jina figure. By “varying relations“ we refer to three different formulas: dominating Jina figure (standard), Gwalior formula (only miniature Jinas above K-and-A), Ellora formula (only ideal relationship to a Jina in a cella). We give the following examples (all except [4] early): (1-3) early Western Indian bronzes (Shah Ak: figs. 22 and 31a, Shah Va: fig. 14), (4) later Western Indian Bronze (Shah Go: pl. 8, figs. 1-2, A.D. 1008), (5) Gwalior (fig. 6), (6-7) Deogarh (JID: figs. 24-25), (8) Ellora (Zimmer Ar: pls. 242-243 = PJA: figs. 152-153; Pereira Mo: 112 and 150).
The vāhana.s of K-and-A are shown in (3) and (8), and the vāhana of Ambikā is shown in (5). A tree is shown above Ambikā and Kubera in (5) and (8). The Ellora Ambikās and Kuberas (8) form mostly a part of the K-and-A combination (Ellora formula). A few mutilated or indistinct hand-attributes can be restored according to the local context of our images.
Further particulars are the following. In (1-4), K-and-A are accommodated in the yakṣa slots. In (6-7), K-and-A are flanking the dharmacakra in the centre of the socle (cf. Tiwari Gu: fig. 6). A second child of Ambikā is shown in (5), clearer in Shah Ak: fig. 38b. Image (3) is one of several compositions where two mahāvidyās appear in addition to K-and-A. (5) is a multi-figured composition.
The K-and-A concept was prevalent in its standard form in the triangle formed by Gwalior, Akota and Ellora, according to a rough estimate up to ca. 900. However, the motif lingered on for some time (4).
Cases of deviation are not rare. Some are isolated, some mirror local conventions: Pal Ex: 137 (= Khandalavala Go: 31) is neither Jaina couple nor K-and-A, but closer to the former than to the latter (Varanasi). Two images from Akota (Shah Ak: figs. 27b [fig. 5] and 40/41a) show twigs in the right hands of the figures, a feature also known from the Hārītī at Ratnagiri (Mitra Ra: fig. 121a). Akota “Ambikās“ often have no child (Shah Ak: figs. 26b, 27b [fig. 5], 41a, 46a) or the child is indistinct (fig. 25). In one case at least, Ambikā is standard, while her partner holds lotus and cup (Shah Ak: figs. 38a-b [figs. 7-8]).
Deccan bronzes (ca. 9th-10th cent.) and stone sculptures from Karnataka (ca. 9th-11th cent.) have their own physiognomy. Refer for the former to the following: Shah Su: figs. 41 (= Pal Ex: 185), 42, 63, 64; Sharma Ya: fig. 6. The Deccan bronzes can be mentioned in connection with the couple and in connection with K-and-A, and can only be described as “yakṣa and yakṣī“ (P. Pal).
In Karnataka we are in the relevant cases probably always concerned with K-and-A, although Kubera and Ambikā take the form of separate images, the only published exception being presumably PJA: figs. 190-190A (K-and-A on a Jina-image). See PJA: fig. 223 (K); Dhaky Śā: pl. 14, fig. 2 (K); pls. 16-17, figs. 6-8 (K, A, A); pl. 22, figs. 19-20 (A, K); Pal Ex: 184 (K, A); Settar Ka: figs. 7-9 (K, A, K), 12 (A).
In Karnataka, the subject of Kubera and Ambikā deserves a special enquiry: Are the Kuberas and Ambikās isolated, do they form a pair, is the pair connected with an adjacent Jina? While basically sticking to the old concept, the Karnataka school obviously modified the Kubera and Ambikā types during the 9th through 10th (11th) centuries. On the one hand, the old standard iconography was no longer observed, on the other hand the new developments of the 12th century (§5) had not yet started.
A sculpture in Vallimalai (9th to 10th cent.) shows Kubera with elephant and the female counterpart with lion, but probably without child (PJA: figs. 56 = 73). This is perhaps the only published instance of K-and-A in the South and clearly irregular.
At a place like Deogarh, we can study the gradual disappearance of the K-and-A concept. The motif has become rare and disappears in the medieval period. A few figures are standing. These are the last traces: JID: figs. 24-25, 264; Bruhn An: fig. 8; JID: fig. 23 [29] and 187; JID: fig. 88 (K-and-A standing); three images in § 122 of JID (figs. 136 [standing] and 186; unpublished Image No. 134 [A seated, K standing]). Refer to Bruhn Ar: 1293-1294 for Image No. 134.