Reviews Of Different Books

Page 14

220 REVIEWS antasya) states that yat, classed as affix, is introduced after a vowel-final root (e.g., jeya 'to be conquered' <ji). The rule is understood as stating aj-antad dhatoh paro yatpratyayah. For the other cases in question, examples of which occur in all parts of the grammar and not in one circumscribed section alone, metarules are stated. 1.1.49 tells the student of the grammar that, where there is doubt about what he is to understand, a genitive is to be interpreted uniquely as denoting a particular relation: a unit is a substituend relative to another. 1.1.66 states that, when there is doubt about what is to be understood from a locative form, one should understand that locative as denoting that before which an operation applies.32a 1.1.67 states that what is denoted by an ablative is that after which an operation takes place. In (3)6.4.34, then, one now understands sasa upadhayah sthane 'in place of the penultimate sound of sas' and an-haloh 'before an (aorist affix) and a consonant-initial unit' (e.g., asisat, sista). (4)6.1.77 is now unequivocally interpreted as providing that a vowel i, u, r, I is replaced by a corresponding semivowel before another vowel. (6)8.1.28 now unequivocally provides that a finite verb form has no high pitched vowel when it follows a nonverbal form. I think it is clear from the above that 1.1.49, 66, 67 are brought into play as interpretive rules only in cases where forms cannot be uniquely interpreted and suppletion can yield several interpretations. Therefore, the partitive genitive sasah of (3)6.4.34 is not interpreted by 1.1.49. Nor are the locative yusmadi of (2)1.4.105 and the ablative dhatoh of (5)3.1.91 interpreted by 1.1.66, 67. This is the way Paniniyas since Katyayana have understood 1.1.49, 66, 67 to operate. These metarules come into play to supply a unique interpretation where rules leave doubt at to what one is to understand. They restrict what one is to understand by supplying items from one's knowledge of normal Sanskrit syntax. 83 Note further that these are metarules (paribhasa). A metarule does not function independently. It functions in conjunction with other, operational, rules. Hence, if one ask, 'What of that which is denoted by purvasya, uttarasya in 1.1.66, 67?' the answer is self evident. This has to be an operation (karya), since these rules have to do with the correct application of operational rules. Moreover, the genitives purvasya, uttarasya cannot be interpreted here by 1.1.49. The operational rules with which 1.1.66, 67 apply show that substitution is not the only operation which is involved. The genitive ug-id-acam of 7.1.70 (see above) cannot be interpreted as denoting a substituend: it denotes units susceptible to an operation in a right context, namely the insertion of the augment n. 1.1.66 applies to interpret the locative sarvanamasthane. If purvasya of 1.1.66 were interpreted by 1.1.49, then, 7.1.70 would have to state a substitution. This is impossible. The Paniniya interpretation of 1.1.49, 66,67 is so clear and works so smoothly that I gladly dismiss Scharfe's tortuous thesis in favor of it. Scharfe's claim that understanding karyam in 1.1.66, 67 is arbitrary results from a misunderstanding. One can make this claim only if one does not take into account the way in which metarules function in the grammar. Let us now reconsider some of the problems which Scharfe has created and see how they disappear once 1.1.49, 66, 67 are understood in the Paniniya way. In 7.1.88 bhasya is a bound genitive (cf. (al), (a2), (3) above): bhasya teh 'of the ti of a bha'. Since bhasya is immediately and unequivocally interpretable as a partitive genitive, 1.1.49 does not apply to interpret it; this rule applies only to interpret teh. 1.4.105 is a restriction (niyama) on verb endings, whereby particular sets of endings (called prathama, 32. The wording of 1.1.66, 67 shows that the former applies to interpret a former denoting a contiguous locus (aupaslesikam adhikaranam). 33 Note carefully that this does not mean these rules restrict the application of syntactic rules 2.3.50 etc.