Reviews Of Different Books

Page 10

REVIEWS the rule invoked to explain -yat of 1.1.21 by commentators: Kas. saptamy-arthe vatih. 96: a-mredita- 'reduplication'. This is not strictly in accord with Panini's definition. Rule 8.1.1 (sarvasya dve) is a heading which provides that, under the conditions given in subsequent rules, an entire item (sarva) is doubled. 8.1.2 (tasya param amreditam) then states that the second (param 'subsequent) of the two items thus obtained is termed amredita; e.g., in jalpati jalpati 'constantly chatters' and purusah puruso nidhanam upaiti 'all men go unto death' (see below, re p. 335) the second jalpati and purusah are amredita. 122: uttama-2 'the first person'. But prathama-2 (p. 391) is accompanied by the gloss 'the endings of the first person' and on p. 436 madhyama-2 is glossed 'the endings of the second person'. The first entry should be modified to accord with the last two: these are all terms denoting sets of verb endings. See below, re p. 148. 123: upa-pada- 'a word standing near or accompanying another to which it is subordinated'. Though 'cooccurring item' is a perfectly justifiable gloss for this item (cf., e.g., Kas. 1.3.77: samipe sruyamanam sabdantaram upapadam) and is valid in rules such as 1.3.77, 1.4.105 (see above, re p. 22), it should be noted that it is also a technical term defined by Panini. Rule 3.1.92 (tatropapadam saptami-stham) provides that what is stated in the locative (saptami-stha) in rules of the section headed by 3.1.91 (dhatoh 'after a root) is termed upa-pada. For example, 3.2.1 (karmany an) contains the locative karmani of karman 'object'; it provides that an is affixed to a root construed with an upa-pada denoting an object. And rule 2.2.19 (upa-padam a-tin) then provides for compounding such a nominal upa-pada with the syntactically related verb derivative, the compound not alternating with a sentential expression containing the same items (2.2.17: nityan). Thus, by 3.2.1, an is affixed to kr to yield kar-a'maker', which forms a compound with, e.g., kumbha + acc.: kumbha-kara- (with deleted acc., see ftn. 6) 'pot maker'; this is equivalent to kumbhan karoti 'makes pots' but does not have an alternant expression *kumbhan karah consisting of kumbha + acc. and kar-a. 135: upa-sarjana- 'subordinate, secondary'. Here again, though this gloss is perfectly justified, it should be noted that Panini both uses the term in this sense and defines a class of upa-sarjanas. The two such definition rules are 1.2.43 (prathama-nirdistam samasa upasarjanam) and 1.2.44 (eka-vibhakti capurvanipate). The first states that what is stated in the nominative (prathama-nirdista 'stated with the first triplet of nominal endings') in the section of rules for compounds (samasa) is upa-sarjana. The second rule provides for classing as upa-sarjana an item which, in the derivation of a compound, always has the same set of nominal endings (eka-vibhakti). This rule also has a qualification: such an item is upa-sarjana except with respect to occupying prior position (a-purva-nipata). Now, by 2.2.30 (upasarjanam purvam) an upa-sarjana generally occurs as prior member of a compound. For example, 2.2.8 (sasthi) provides for a nominal + genitive (sasthi) to compound with a syntactically related item; further, sasthi is a nominative form, so that the nominal + genitive is upa-sarjana, hence occurs as prior member; thus, to tad- + gen. and purusa-s (tasya purusah) corresponds the compound tat-purusa-. 2.2.18 (ku-gati-pradayah) provides for a compound from items such as nis and syntactically related items; e.g., nis-kausambi'gone from Kausambi' (kausambya niskrantah). Since kausambi- is always followed by the ablative ending in whatever sentence corresponds to the compound, it is upa-sarjanaby 1.2.44. But by the same rule it does not occupy prior position; this is preserved for nis, which is also upa-sarjana, since pradayah 'pra etc. (including nis)' is stated in the nominative. The reason for classing kausambi- is to allow -i to be replaced by the short vowel -i: nis-kausambih; see Pratidanam, p. 451. Now consider the derivation of compounds of the dvandva type. 2.2.29 (carthe dvandvah [anekam 28]) provides that multiple syntactically related items may be compounded to form a derivative in the meaning '... and ...', such a compound being termed dvandva. Thus is derived, for