________________
112
THE CANONICAL NIKSEPA
presentation of the three niksepas. The repetitive character which makes the references inevitable can be characterized in the present case (and in many other cases) by the word "application". A piece of text is repeated - in theory at least - verbatim with minor changes (e.g. NERAIYĀ replaced invariably by ASURA-KUMĀRĀ). The piece is identified (reference-technique) by "words" (e.g. jahā neraiyanam), "titles" (e.g. jaha Bhas' uddesae), or both. The ideal model is simple, but in practice the picture is blurred by three factors. In the first place, the reduction may be drastic and may go to the point of not mentioning an individual repetition-case at all (e.g. jāva VANASSAI-KĀIYĀ after the application of the NERAIYĀ text to the ASURA-KUMĀRĀ and to the PUDHAVI-KAIYA: water, fire, and wind beings not even mentioned (Prajñāpanā 28, 645]). In the second place, there may be various changes and additions (or omissions) which exceed the mere substitution of single terms. Thirdly, the extent of the piece of text which is applied repeatedly may vary (e.g. second extension of the AHĀRA niksepa not included in No. 78 37). The impact of the S... niksepa of Anuyogadvāra on the exegetical literature is perhaps the most conspicuous example for the technique of application. Application is a repetition technique affecting pieces of medium size (for example one page in the Suttāgame format). The subject-matter is dogmatical, not descriptive (as against the well-known varnakas). It is probable that in most cases the primary text (which is applied or multiplied) can easily be distinguished from the secondary texts (the "applications"). These applications may be added (appended) to the primary texts with a high degree of reduction (primary text and application(s) contiguous). They may, however, also occur separately in the 'same' work or in "other" works) with various readings etc. as independent versions and in a less reduced form). In the present case it is convenient to say that the text in question (primary text cum applications) coincides in extent roughly with our tripartite niksepa; and that Prajñāpanā 28.1 contains the primary text with contiguous applications added to it (one application of the textual matter touching the niksepa as such: No. 123), while Bhagavati 1.1.15, Bhagavati 2.1.85, and Jīvābhigama 1.13 contain separate applications of the niksepa in Pra. 28.1. Although in Prajñāpanā and Bhagavati the niksepa is embedded into related material, we have isolated it from this context. That such a procedure is not quite arbitrary is demonstrated by the fact that the version in Jīvābhigama 1.13 consists only of the AHĀRA niksepa (without further discussions on ahāra, as found in Prajñāpanā 28.1 (Ahāra-paya), preceding or following it). Below we supply a synopsis of the six niksepas appearing in the four pieces of text just quoted. The variation is always mechanical (mere substitution), but the size varies according to the strength of the reduction. Niksepas which are merely implied (not even a minimum of niksepa text supplied) cannot be considered. The quotations (in capitals) demonstrate the mechanism of substitution.
Prajñāpanā 28.1.642 (primary text). "NERAIYĀ nam, bhante, kim ĀJĀRAM ĀNĀRENTI ?" S.II, 504, line 12 - 505, line 5
No. 122