Reviews Of Different Books

Page 13

REVIEWS 55 hrastic Future'. p. 491: Irt 'personal endings of the Future'. let 'the personal endings of the (Vedic) Subjunctive'. lot 'the personal endings of the Imperative'. On p. 491 should be added lrn (3.3.139), the L-member introduced in deriving conditionals. All such L-members are replaced by personal endings, themselves subject to further substitutions depending on which L is at their origin (3.4.77-112). And Panini does indeed use such L's to denote endings which replace them. For example, 3.4.113 (tin sit sarva-dhatukam) assigns to the class of sarvadhatuka those post-radical affixes which are marked with s and the finite endings denoted by tin; 3.4.114 (see above, re p. 15) classes as ardhadhatuka the remaining (sesa) post-radical affixes. And 3.4.115 (lit ca) is a particular rule contravening 3.4.113; it states that lit also (ca) is ardhadhatuka. Obviously, here lit must then denote the endings which replace it; cf. Kas.: lid-adesas tin ardhadhatuka-samjno bhavati. But this is certainly not always the case. Rule 3.2.124 (latah satr-sanacau ...)11 states that lat is replaced by sats or sanac, participial affixes. Now, these are marked with s, which must have a purpose. And the only purpose possible is the classification of the affixes as sarvadhatuka by 3.4.113. But, if sats, for example, were a replacement of a finite ending such as tip, it would be marked with s merely by virtue of having replaced an ending so marked (1.1.56: sthanivad adeso'n-alvidhau); it is for this reason, for example, that Panini specifically states that the imperative ending hi replacing sip is not marked with p (3.4.87: ser hy a-pic ca). Moreover, considering sats to be a replacement of endings such as tip raises insurmountable problems regarding accentual rules. In general, the first vowel of an affix (pratyaya) is high pitched (3.1.3: ady-udattasca); thus, -at- (sats). But nominal affixes (sup) and affixes marked with p (pit) have low pitched vowels (3.1.4: anudattau suppitau); thus, ta (inst. sg. ending), tip. Now consider the derivation of a participial form such as lunata 'cutting'. If sats replaces lat directly and not an ending such as tip, lunat- is derived as follows: lu-lat + lu-at + lu-na-at (3.1.81) + lu-n-at (6.4.112). This base is then followed by the affix a, which is low pitched. Rule 6.1.173 (satur a-numo nady-aj-adi) then applies to let the ending a be high pitched: lunata. This rule provides that such an ending is high pitched if it follows a unit ending in sats which has not been augmented with num (-ant.) and whose last vowel is high pitched (antodattat 6.1.169). lu-n-at-a meets these conditions. If, on the other hand, sats replaces tip etc., we immediately run into a problem. Given lu-lat + lu-ti, we have to decide whether ti should condition the introduction of sna (as in the 3rd sg. pres. lunati) or should be replaced by sats, which, since it is also sarvadhatuka, will also condition the introduction of sna. The decision to have ti replaced by sats first must be arbitrary. Further, once we progress to the stage lu-na-at (with low pitched -at replacing tip) + lu-n-at, we must invoke an additional rule to let -at be accented as shown; 6.1.161 (anudattasya ca yatrodattalopah) provides for replacing an anudatta by an udatta if it conditions the deletion of a high pitched vowel. Though we can thus arrive at lu-n-at-, it is only at a cost; the alternative derivation involves an arbitrary decision and prolixity. But this is not all. For it is impossible to follow the alternative derivation if one is to correctly arrive at a form such as sunvata 'pressing'. Even making the arbitrary decision noted above, so that we can somehow arrive at su-nu-at, we cannot reach the required su-nv-at. For here rule 6.1.161 cannot apply. The only rule which could possibly apply is 8.2.4 (udatta-svaritayor yanah svarito'nudattasya); and this only provides for replacing, in the present instance, the low pitched vowel of -at by a circumflex vowel after the semivowel -v- which has replaced -u-. From the above it follows necessarily that lat in 3.2.124 cannot denote finite verb endings. 532: vaiyakarana- 'relating to grammar, grammatical; (m) a grammarian'. In the only rule given as reference (6.3.7) the term means 'grammarian'. It occurs as part of 11 See Lingua 25.214.(1970), where I inadvertently left out a note stating that satr, sanac were there treated as replacements of finite forms only for purposes of presentation.