Prof. Dr. Klaus Bruhn
The essay was published in Berliner Indologische Studien No. 2. 1986, pp. 133-174.
§ 11. Statistics (introduction)
As we mentioned already in § 7, we are here concerned with the standing non-Pārśvas of the Early-Medieval Style at Deogarh. The idea of “statistics“ (i.e. motif-statistics) appears so simple that no justicifation or explication seems to be called for. We want to know whether the relevant Jinas are shown “with or without bhāmaṇḍala“, “with our without Navagrahas“, and so on. However, a difficulty arises as soon as we postulate that statistics are the complement of image-description (§ 4). In order to make this statement plausible, we have referred the reader to the phenomenon of “changing combinations“.
In this connection, we have to make some further observations. A stylistically related group of ten Jina images within a larger corpus may ex hypothesi consist of five images with and five images without Navagrahas. Here the description of the group will not say anything about the role played by the Navagrahas within the entire corpus. The question remains, is such a ratio restricted to one single group or do we find everywhere images with and without Navagrahas side by side? Even if the statistical survey is not a very detailed one it will give us some general idea of the situation. By contrast, statistics alone would never help us to isolate stylistically related images. Our hypothetical group of the ten related images may be iconographically split in more than one respect and with different dividing lines (images with and without Navagrahas, with and without bhāmaṇḍala, and so on) but it may nevertheless show some sort of family likeness (being identical in “style and stone“, bearing the “same stamp“).
In this case, one will say that the isolation of related groups (image-description) is clearly more important than statistics. The full impact of statistics is felt in the case of very common and very rare developments: almost all images shown with Navagrahas, only one image shown without Navagrahas - and vice versa.
For the above argument a very general concept of changing combinations was sufficient and we could leave matters at this point. But let us imagine for a moment, that within the corpus as a whole only 30% of images are without bhāmaṇḍala, whereas almost all the images with Navagrahas have no bhāmaṇḍala. Such a fact will not appear in the image-description, which is only concerned with individual groups, and it will also not appear in the statistics (previous para) unless we decide to include such detailed observations in our statistical survey. Here we suggest a twofold compromise. First of all, on the theoretical level: we cannot devise terms for all the possible cases of changing combinations (in our case, mutual exclusion of two features within a corpus). This point has already been touched upon in § 5. Secondly, on the practical level, we prefer an approach where things are mentioned as far as they are noticed by us and as far as they are simple enough to be explained in a few words.
In other words, the statistical survey can be rendered still more concrete by occasional references to unmistakable distribution data which do not catch the eye at once. In certain cases, e.g. the parasol-unit of § 16, this is even necessary if we want to obtain a transparent and plausible description. We can add in this connection that statistics are not only concerned with the individual motifs appearing on the images. It is often necessary to study pairs of motifs (motif to the left, motif to the right), clusters of motifs, or even entire zones of the composition.
It might be tempting to demonstrate that statistics help to define local varieties (e.g. one or more “Western Indian varieties“), but even a preliminary discussion of it would of course go well beyond the scope of this essay.
What is more important at this stage is the connection between statistics and reduction. Statistics tell us at a glance what a set of images (normally a “corpus“) looks like. In our case (§§ 12 ff.), the reader understands at once that the treatment of the pedestal is not very elaborate and that the lion-throne is conspicuous by its absence. Diversity is thus reduced to a short description of a corpus (in the present case: ca. 137 numbered plus 25-50 unnumbered images). We can also say that statistics show the physiognomy of the corpus. Naturally, different areas of Indian art require different types of reduction. A. Foucher observes in connection with Gandhāran art: “II n'y aurait aucune exagération à prétendre qu'une centaine de reproductions bien classées peut donner une idée assez complète de l'art du Gandhâra.“ [1]
Apart from the question of reduction we can mention again that statistics may have - and normally does have - an opposite effect as well: It guarantees that the variety inherent in the material is fully explored (see § 7 on variety-taxonomy).
To conclude this introduction, we need to mention that in §§ 12 ff. the late images within the corpus (mainly “Section of the Late Images“) have received less attention than the bulk of the material: sometimes images with special features have not been mentioned at all, sometimes the relevant images have been mentioned but the features have not been specified. It is possible that here and there an interesting feature restricted to one or two late images was overlooked but such omissions if made at all are rare.
§ 12. The pedestal (squares 39-50)
The iconographic inventary of the pedestal of standing images is rather poor. Before discussing its motifs, we have to say a few words about the pedestal as such.
The front of the pedestal is continuous (JID 78), or tripartite with projecting middle portion and separate lateral portions (JID 83b). Deviations from the standing form do occur but they are more common with later images than with the bulk of the material. The central projection can be stepped: JID 153 (see Neg. 934 for a full view) and JID 154. The pedestal can be very high: No. 124 (Neg. 1860) and No.125 (Neg. 1881). The Jina may stand on a tiny socle placed on the pedestal: No. 124 (Neg. 1860) and No. 101 (Pārśva image, Neg. 1447). The pedestal may even protrude upwards in its central portion: fragment lying to the west of the Rampart (Neg. 1976).
The majority of the images show on their faces two frequent but optional motifs and nothing else:
According to the contemporary conventions one might expect the following motifs:
But the existing carvings are only a very faint reflection of the available iconographic vocabulary. One image shows the rare combination lion-and-antilope (JID 114, see Neg. 1309 for a full view of the pedestal). Image No.167 (AJI 3) is the only one to have a true cihna (quadruped: bull). [2] In the case of AJI 9 we notice a rich lotus motif on the projecting middle portion, and Kubera-and-Ambikā to the left and to the right. Extended forms of the foot-lotus are found with two other images (JID 128 and 156a). The image No. 225 (JID 190) which has a lion-throne on its projecting central portion is a transition between the Early-Medieval and the Medieval Style. A small double-image in Wall-Section VIII (partly damaged and of poor quality) shows on the face of its pedestal a horizontal creeper motif (Neg. 1747).
A standing medieval Jina before the Mālādevi Temple at Gyaraspur has an early-medieval motif on its pedestal. The cihna in the centre (quadruped) is flanked by two lions known from Jina images in Orissa (e.g. Mohapatra, Or.116). See Neg. Gyaraspur 2924b. - There are no inscriptions mentioning the names of the relevant Jinas.
According to the bull-cihna, the figure must be identified as Ṛṣabha. On the other hand, the Jina has no lateral strands but merely “strands on the head“ (same formula as in JID 151). Accordingly, the image should show a “non-Ṛṣabha“ (IJI § 5B). However, the incised cihna which does not even show the animal (bull) in a clear manner can hardly supply an argument against our thesis in IJI § 5B.