________________
Vol. xx, 1996
REVIEWS
193
Basically, an annotated bibliography of text editions will always be midway between a catalogue of manuscripts and a history of literature. We mention information about the extent of the text, about the traditional sub-divisions, about minor elements (mangala, colopliones, prasasti), about the form (prose/metre), about the language (Sanskrit/Prakrit), about the text-commentary relation (if a text edition contains both), about authors, and about substantial textual parallels. Basically, a conspectus is produced by a continuous segmentation of a given text, extending from the beginning to the end. More often than not a transformation of the continuum of the text into a string of segments is a prerequisite for a quick comprehension of the structure and the contents of a composition. In princile, segmentation will be unilinear, so that frequent sub-segmentacion is avoided. The contents of each segment must be mentioned, and surrogates must be used where a proper title (epitome) cannot be given. In many cases it will be useful to isolate specific units which are known a priori and which do not participate in the intricasies of the text to the same extent as other portions. Attention must be paid to explicit traditional divisions. In metrical works verses are as a rule counted within such divisions. A conspectus is descriptive and thus basically synchronous. There may be different historcal starata but their spearation is not a part of the conspectus project. But, no conspectus of the Vasudevahindi can ignore the fundamental difference between Brhatkathā elements and the elements belonging to the universal history of the Jainas (24 tīrthankaras, etc.). A labyrinthine structure of a text may be clarified by isolating certain "blocks" such as the Vetālapancavinsatikā in the Brhatkathă versions of Somadeva and Ksemendra. Conspectuses of a tabular type have their main habitat in studies concerned with the Niryukti-Bhāsya stratum, the metrical section of the "exegetical" literature. Thus, the conspectuses by Leumann, N. Balbir, Oberlies, Bruhn, Webor, Delue, A.N. Upadhye deal with dogmatic material and with stories only as far as stories partake in the technical character of non-narrative texts or are not easily spearated from their dogmatic context. In some cases the narrative material is either absolutely autonoinous or merely loosely connected with basically dogmatic texts. It is a different matter if stories are units sui generis but nevertheless are embedded in, or connected with dogmatic works for purposes of demonstration. Abstracts of stories or story collections prepared by H. Jacobi, M. Bloomfield, A.N. Upadhye, Mette, Balbir/Oberlies and K.Bruhn can be cited in this field. Besides the problem of the breadth of Indian (and Jaina) literature, there exists the universal problem of the enormous increase in the number of niodern publications. It has thus become necessary to find ways of making the growing corpus of information manageable. Here Bruhn suggests that bibliographies as used in research should not converge in a unified bibliographical system, but that we need as many different types of bibliographies as possible and consequently of bibliographies of different proportions. And, as to the "responsibility of the author, he admits that summaries (mostly in English) and other devices have come to stay and have made the use of books and articles easier than in the
