The Grammar of Jina Iconography II [Part 5]

Posted: 29.02.2012
Updated on: 02.07.2015

The essay was published in Berliner Indologische Studien No. 13/14. 2000, pp. 273-337.


§ 5. The Attendant Deities of the “Fuzzy Period“

After the K-and-A period follows a period of confusion which is the subject of the present section: confusion due to the parallelism of the old K-and-A formula and the elements of the system; confusion also in the sense of general mixture (no quasi-systematic syncretism) and the introduction of new elements. We shall first of all focus attention on the attendant deities (yakṣīs) of the Jinas as they occur in serial representations of the Jinas. Afterwards we have to deal with attendant deities on single Jina images (Deogarh, Khajuraho, etc.). A Deccan bronze will be our last subject. K-and-A are still found in some parts, but their popularity is declining. We separate the confusion in the “fuzzy period“ from the “deviating concepts“ of § 3 (deviations from K-and-A/standard).

The period under discussion is not a uniform development; we can distinguish between

  1. Western India,
  2. Northern and Central India and

And although the designation of the period was derived from the specific situation in Northern and Central India, we afterwards found a more wide-spread tendency towards diversification and confusion in the field of the attendant deities. The period has no “end“ in the strict acceptance of the term. It ends when the iconoplastic art of Jainism ends, ca. A.D. 1200-1300. As its beginning we can fix the time of Deogarh Temple No. 12 (A.D. 850-900). The earliest textual evidence (transmitted or not) probably antedates this time-span by at least a few decades. In Karnataka, Cakreśvarī (representative of the system) became popular “only in and after the 10th century“ (Settar Ca: 69), i.e. later than at Deogarh.

The serial representations, our first theme, seem to confirm the concept of the 48 YY, as analyzed by us in the previous section, but this impression (agreement between art and texts) is not correct, as will be shown below.

We find two serial representations with attendant deities in Khandagiri Caves 7 (7 Jinas) and 8 (24 Jinas). The only complete documentation is owed to D. Mitra (Mitra Śā). See for Cave 7 (“Navamuni-gumphā“) pp. 127-129 and pls. 1-2; and for Cave 8 (“Bārabhujī-gumphā“) pp. 129-133 and pls. 3-5. In both series, only the female deities are represented. Nowhere does D. Mitra say that all the goddesses (7 and 24 respectively) deviate from the textual prescriptions, but this is implied by numerous statements in her descriptions of the individual goddesses. Considering

  1. the unsatisfactory state of preservation,
  2. possible differences among the textual sources and
  3. the omnipresence of common place hand-attributes like varada and cakra,

it is impossible to say that there is nowhere agreement. There may be agreement or quasi-agreement here and there, but that would not be an effective proof for true agreement between images and texts.

A third well-known set is found in the Bhaṇḍāra Basti at Shravana Belgola, and here the Jinas are flanked by the 24 gods as well as by the 24 goddesses. There is a complete two-page documentation where the figures are depicted in small size (Settar Śr: 61, 62-63). But individual images have also been published in larger size (Settar Ka: figs. 10 and 13; JRM: figs. 61-63, 65, 67).

A series of Jinas, complete or incomplete, is found in the Suttālaya at Shravana Belgola (enclosure around the Gommaṭa image). See Settar Ka: figs. 1, 4 and 17; Settar Ca: figs. 7-8, p. 69; and JRM: figs. 59-60. U.P. Shah also mentions complete sets (with YY) at Venur and Mudbidri (JRM: 135-136; fig. 66 = Mudbidri).

It is possible that, in Karnataka, texts and images converged at some point of time, perhaps in the 12th century (see Settar Ka: 48 for the date of the Suttālaya Jinas). This would not, however, be an old tradition but a new attitude towards the texts (cf. the general situation in the medieval Śvetāmbara art of Western India) and thus a development in its own right. If a convergence took place, it should be most noticeable in the aforementioned series. The 12th century phase (whatever its precise character) was preceded by a period where special forms of K-and-A (§ 3) as well as Dharaṇendra and Padmāvatī (§4) were popular. The early images can now be studied in detail with the help of the published photos.

U.P. Shah and M.N.P. Tiwari have published an image of Ambikā in the Allahabad Museum. The goddess is surrounded by miniature Jinas (4+5+4) and small goddesses (9+9+5) with inscriptions giving the names of the goddesses (Shah Am: fig. 28 and p. 163; Tiwari Am: 83-85). The number twenty-four (1+23) leaves no doubt that the goddesses have been understood as yakṣīs although they are represented without corresponding Jina figures. Close ups have not been published, and the entire concept was not systematic (Tiwari Am: 84). A comparison of the inscribed names with yakṣī names occurring elsewhere may nevertheless be useful. The thirteen miniature Jinas mentioned above form a commonplace motif. See also Mevissen Co: No. 118.

The last, and earliest, set to be mentioned consists of the goddesses on the exterior of Deogarh Temple No. 12 (JID: chapter 8). They are not attendant goddesses in the usual sense, accompanying a Jina (yakṣa slot, etc.), but have a slightly deviating character, as will be seen below.

They are generally described as more or less irregular Jaina yakṣīs. Each figure is attached to a slab showing

  1. the name of a Jina,
  2. a miniature Jina (all these Jinas are identical in appearance),
  3. the name of a goddess and
  4. the goddess herself.

There is no reason to doubt that sculptures and inscriptions were made simultaneously. We have reproduced two plans (figs. 13-14), the former taken from JID: fig. 342 with the names of the Jinas (standard spelling) added to the names of the yakṣīs; the second plan also taken from JID: fig. 342 but with new numbering and a few adjustments. See JID: 102-104 for the irregular spelling of the names of the Jinas as it is found in the inscriptions of No: 12. In the first plan (fig. 13), the numbering runs in the pradakinā order (clockwise) round the temple, starting from the main doorway. This numbering is no longer used by us. In the second plan (fig. 14), the numbering follows the twenty-four Jinas (true numbering), starting with no. 1 (Ṛṣabha: Cakreśvarī) and ending with no. 24 (Mahāvīra: Aparājitā). The Jina: yakṣī series thus begins and ends in the middle of the north side of the temple, and the arrangement is apradakinā. Similar unexpected arrangements can be observed in the case of South Indian Rāmāyaṇa reliefs. See especially Sanford Rā: 105/146 (apradakṣinā order; “... the story begins near the center of the temple's back [west] wall...“). The numbers in fig. 14 appear besides the yakṣī names. They refer to the Jinas (serial arrangement) and to the yakṣīs (random arrangement). The yakṣīs belong only in three cases to the respective Jinas (nos. 1, 22 and 23: infra).

The Deogarh sequence is not intact. Between Jinas nos. 1 and 8 there are disturbances, partly due to structural additions (two slabs hidden from view) and partly due to irregularities in the original arrangement. However, nos. 8-24 are intact in every respect. Between nos. 1 and 8 we find the Jinas nos. 4, 7 and 6. The complete irregular zone is /4/*/hidden/7/**/hidden/***/6/. One slab (*) has only the name of the yakṣī Jālāmālinī (the name of the Jina being covered with whitewash); one slab (**) has only a Jina figure (in the yakṣī panel) and nothing else; in the case of one slab (***) we get Jina and yakṣī, but both names are missing (JID: 107, “[no photo]“). Jinas nos. 2, 3 and 5 are not found and must have been on three of the four unknown/incomplete slabs /*/, /hidden/, /hidden/, /***/. One of these four positions is dubious (no Jina of the series, perhaps no Jina at all). The total of Jinas is thus 26 (possibly only 25).

In some cases the names of the goddesses agree with the names of the corresponding goddesses as contained in the Trilokaprajñapti. There is full agreement in three cases:

  • 1 Deogarh: Cakreśvarī / 1 Trilokaprajñapti: Cakkesarī;
  • 23 Deogarh: Padmāvatī / 23 Trilokaprajñapti: Paumā.
  • 22 Deogarh: Ambāyikā (22) must be added, because 22 Trilokaprajñapti: “Kumbhaṇḍī“ is identical with “Ambikā“ (§ 4).

Sometimes the name agrees but the number differs:

  • 9 Deogarh: Vahurupī / 21 Trilokaprajñapti: Bahurūpiī;
  • 14 Deogarh: Anantavīryā / 15 Trilokaprajñapti: Aantamadī;
  • 20 Deogarh: Sidhai / 24 Trilokaprajñapti: Siddhāyiī;
  • 24 Deogarh: Aparājitā / 20 Trilokaprajñapti: Aparājidā;
  • [*] Deogarh: Jālāmālinī/ 10 Trilokaprajñapti: Jālāmālinī.

If we subtract obvious ad hoc names (not found in the Trilokaprajñapti: Sulocanā, Sumālinī, Surakitā and two “Śriyādevīs“) there remains a nucleus of genuine names, some of which are unica. A complete concordance with the Trilokaprajñapti is not necessary.

We now turn to the actual representations in the series, starting with three goddesses of the “real pantheon“ (see § 6). “Cakreśvarī“ (fig. 15, Ṛṣabha) is shown with four cakra.s. “Ambāyikā“ (fig. 16, Nemi) carries a child, but she is otherwise irregular. “Mayuravāhi(nī)“, the Jaina Sarasvatī, appears wrongly as the yakṣī of Supārśva (no. 7). She is represented with a peacock vāhana and a book (fig. 17). The rendering of “Padmāvatī“ (figs. 13-14, JID: fig. 57) is unexplained and has nothing to do with the Pārśva-Dharaṇendra-Padmāvatī triad of the system.

There are several goddesses who belong to the mahāvidyā circle (JID: 107-108; fig. 63, etc.). If we consider three different Jaina temples, viz. Deogarh Temple No. 12, Mālādevī Temple (Deva Mā) and Ādinātha Temple (Khajuraho: Tiwari Ād), we get the impression that the mahāvidyās (Shah Ma) exercised some influence over the Jaina art of our period in Northern and Central India, but identifications are difficult. Refer for textual descriptions of the sixteen mahāvidyās (Pratiṣṭhāsaroddhāra, Nirvāṇakalikā et alia) to § 4.

Most of the remaining goddesses are poor in their iconography but can be identified either as river goddess (one case: JID, fig. 61: goddess no. 19 in fig. 14); or as Jaina Tārās (e.g. JID, fig. 71: goddess no. 9 in fig. 14). Our Tārās are two-armed and hold only one hand-attribute, viz. an object midway between cāmara and padma (cf. also Mevissen Co: figs. 21-22 and 24). One of the Tārās (goddess no. 18 in fig. 14) is actually called “Tārādevī“.

It would appear that the tendency to produce deities with “poor“ iconography was caused by the necessity to provide all the niches of the big Jaina temples with occupants (Northern and Central India). The type of the two-armed male or female figure with snake-hoods belongs also to the category of “poor“ or “colourless“ deities. It is common on the exterior of the Mālādevī Temple (see Deva Mā: fig. 3 and p. 263), but rare at Deogarh (JID: fig. 49) and not contained in the iconographic programme of Temple No. 12. On the other hand, Tārā is not restricted to Temple No. 12 but also found elsewhere (see JID: figs. 48, 50, 51). Our expression “poor“ refers to wall-figures. The rendering of the attendant deities (infra) is very often careless, but we cannot use any general term for their characterization.

The wall-figures of Deogarh Temple No. 12 demonstrated the rise of the system already through the use of the number twenty-four. What is true of this temple is equally true of all “24 Jina representations“ mentioned by us. In addition to that, the existence of the system is always implied in a general manner by Gomukha and Cakreśvarī, whatever their exact place (series of Jinas, single Jina image, image of non-Jina).

For the general situation (dominance of “fuzzy“ iconography) we refer the reader to the yoginī material (sculptures, texts, inscriptions). See in particular Misra Mo: chapter 8.

Single Jina images with attendant deities demonstrating the “fuzzy period“ are numerous in Northern and Central India. The attendant deities are:

  1. Gomukha and Cakreśvarī combined (Jina with G and C);
  2. Kubera and Ambikā combined (Jina with K-and-A);
  3. familiar deities in unconventional combination, unfamiliar deities; and
  4. unclear figures.

A stylistic factor is the decorative medieval pattern (“grid“) of the Jina images which does not always assign much room to prominent attendant deities. The broader context of Systems A and B has already been described in § 1. - Below, we give a short survey of images belonging to the “fuzzy period“:

JID: fig. 231 shows a Jina composition with Ambikā and Cakreśvarī to the viewer's left and right, an arrangement which is against the system. The same applies to the seated Khajuraho Jina JAA: pl. 378B (described in Tiwari El: 33-35). In this case small attendant deities are depicted in the yakṣa slots and big figures of Ambikā and Cakreśvarī on the central projection of the socle. A seated Ṛṣabha from Orai demonstrates that such experiments started early. The Orai Ṛṣabha shows

  1. Kubera and Cakreśvarī in the yakṣa slots, and
  2. an unidentified goddess and Ambikā flanking the dharmacakra.

See Tiwari El: fig. 3 (and pp. 45-47), Bruhn Gr I: fig. 16, Shah Ca: fig. 27. An interesting piece is the elaborate seated Pārśva image (Mevissen Co: fig. 7) of which only the lower part remains. It shows numerous figures (dii minores and others), some conventional and some unconventional; Pārśva is flanked by two standing Jinas, but the piece was not a tritīrthika in the strict acceptance of the term. Two deities (female attendant figure and Kubera) are depicted to the proper left of a standing Pārśva at Khajuraho (Mevissen Co: fig. 8), originally part of a tritīrthika. The standing Jinas of figs. 18-19 (Vidisha [former Bhilsa] and Udayagiri) belonged to two different but similar tritīrthika.s. One Jina shows a four-armed Ambikā and the other an unidentified four-armed goddess on the socle. Refer also to the unidentified socle deities of the tritīrthika (standing Jinas Pārśva and Supārśva, central Jina gone) in JID: fig. 264.

Incomplete standing Jina images can also belong to dvitīrthika.s (compositions depicting two standing Jinas), but that type of composition is rare among elaborate medieval Jina images in Northern and Central India. We therefore considered such pieces parts of tritīrthika.s.

More conventional are images (Deogarh) showing male and female deities in the yakṣa slots and no additional deities. Thus, JID: fig. 221 combines Gomukha with an unidentified four-armed goddess. In JID: fig. 220, stylistically closely related to the previous image, the socle demonstrates in every respect the system (Gomukha and Cakreśvarī). The system was practised, no doubt, but at random and not as a rule. In addition to JID: fig. 221, we mention here two further irregular Jina images from Deogarh, namely JID: figs. 195 and 242 (see § 1: Systems A and B). Refer for the K-and-A formula at Deogarh to § 3 (last paragraph).

A veritable explosion of indistinct and undefined attendant deities can be observed at Khajuraho (Tiwari Kh, Tiwari Ma). There are many Kuberas or rather attendant deities with a money-bag, but neither Kubera on an elephant nor K-and-A (Tiwari Ma: 224). According to M.N.P. Tiwaris's estimate, there are at Khajuraho “over four hundred various Jaina images excluding those represented on the doorways and fragmentary door-lintels“ (Tiwari Ma: 217).

A socle from Ladol in North Gujarat (Jina: Abhinandana) is not very remarkable in itself. The fragment, however, and the description given by U.P. Shah (Shah La: fig. 28, pp. 71-72, see also p. 69) demonstrate the peculiarities of the “fuzzy period“ and the problems of adequate evaluation.

The general predominance of female figures among the various Jaina deities is too obvious to require a special discussion. Besides the Jina, Ambikā is so to speak a separate centre of gravity; her images show all sorts of parikara figures. Here we mention only JID: fig. 208, where the yakṣa slots with attendant goddesses are reminiscent of the pattern of Jina images.

An elaborate bronze Jina from the Deccan (Barrett De) is an iconographic concept in its own right, but it demonstrates also further tendencies of the “fuzzy period“. The Jina (ht. of the image 23'') has three lions, but his identity is uncertain. Of three lions on the socle of a Jina image (Tamil Nadu, Deccan, etc.) the central one is normally part and parcel of the three-lion motif (Bruhn Gr I: 258), but it could be used as cihna of Mahāvīra if the three-lion formula was otherwise not used in the region. The attendant god of Barrett's Jina is Kubera, while the attendant goddess has the attributes axe / lotus (upper hands) and citron / flower (lower hands). The Jina may be Mahāvīra or not, the present combination of attendant deities has no place in the texts. The citron (a small coin in Indian iconography) is not only found with the two attendant deities but also with several other subsidiary figures of the bronze Jina. The two cāmara-bearers hold citrons in their inner hands, and a citron is also seen in the hand of two small figures seated on decorative pillars (architectural members flanking the composition at the height of the central Jina). Four more small figures are found in the upper section of the image, surrounding as it were the triple parasol (“flying dwarfs carrying offerings“: Barrett De, p. 165b).

The main lesson to be learned is that one has to study the attendant figures (and the dii minores generally speaking) not one by one, but as part of the archaeological ensemble of the region: the “fuzzy“ figures appearing on Jina or Jaina images (slabs) in stone are a peculiarity of Northern and Central India.



Share this page on: