Indian Antiquary Vol 14

Author(s): John Faithfull Fleet, Richard Carnac Temple
Publisher: Swati Publications

Page 44

36
THE INDIAN ANTIQUARY.
[FEBRUARY, 1885.
his language is without authority until tested by the actual facts of the language, as represented by the Sanskrit literature. But the principle won here is likely to prove of universal application, for we have no reason to expect to find the grammarians absolutely trustworthy in other departments of their work, when they have failed so signally in one. There can be nothing in their system that will not require to be tested by the recorded facts of the language, in order to determine its true value. How this is, we will proceed to ascertain by examining a few examples.
In the older language, but not in the oldest, (for it is wanting in the Veda), there is formed a periphrastic future tense active by compound ing a nomen agentis with an auxiliary, the present tense of the verb as 'be': thus, dátá 'smi, (literally dator sum), 'I will give,' eto. It is quite infrequent as compared with the other future, yet common enough to require to be regarded as a part of the general Sanskrit verb-system. To this active tense the grammarians give a corresponding middle, although the auxiliary in its independent use has no middle inflection. It is made with endings modified so as to stand in the usual relation of middle endings to active, and further with conversion in 1st sing. of the radical : to h-a very anomalous substitution, of which there is not, I believe, another example in the language. Now what support has this middle tense in actual use ? Only this: that in the Brahmanas occur four sporadic instances of attempts to make by analogy middle forms for this tense : (they are all reported in my Sanskrit Grammur, $ 947; further search has brought to light no additional examples). Two of them are 1st sing., one having the form se for the auxiliary, the other he, as taught in the grammar; and in the whole later literature, epic and classi. cal, I find record of the occurrence of only one further case, darśayitáhe (in Nâish. V. 71.)! Here also, the classical dialect is the true continuator of the pre-classical. It is only in the grammarians' Sanskrit that every verb conju. gated in the middle voice has also a middle periphrastic future.
• Here, as elsewhere below, my authority for the later literature is chiefly the Petersburg Lexicon (the whole older literature I have examined for myself), and my statements are, of course, always open to modification by the results of further researches. But all the best and
There is another and much more important part of verbal inflection-namely, the whole aorist-system, in all its variety-as to which the statements of the grammarians are to be received with especial distrust, for the reason that in the classical language the aorist is a decadent formation. In the older dialects, down to the last Sútra, and through the entire list of early and genuine Upanishads, the aorist has its own special office, that of designating the immediate past, and is always to be found where such designation is called for. Later, even in the epos, it is only another preterit, equivalent in use to imperfect and perfect, and hence of no value, and subsisting only in occasional use, mainly as a survival from an earlier condition of the language. Thus, for example, of the first kind of aorist, the root-aorist, forms are made in pre-classical Sanskřit from about 120 roots. Of these, 15 make forms in the later language also, mostly sporadically, (only gá, dá, dhd, pá, sthá, bhú less infrequently), and 8 more in the later language only, all in an occurrence or two, (all but one, in active precative forms, as to which see below). Again, of the fifth aorist-form, the ish-aorist, (rather the most frequent of all), forms are made in the older language from 140 roots, and later from only 18 of these(and sporadically, except in the case of grah, vad, vadh, vid), with a dozen more in the later language exclusively, all sporadic except sonk, (which is not a Vedic root). Once more, as regards the third or reduplicated aorist, the proportion is slightly different, because of the association of that aorist with the causative conjugation, and the frequency of the latter in use. Here, against about 110 roots quotable from the earlier language, 16 of them also in the later, there are about 30 found in the later alone, (nearly all of them only sporadically, and none with any frequency). And the case is not otherwise with the remaining forms. The facts being such it is easily seen that general statements made by the grammarians as to the range of occurrence of each form, and as to the occurrence of one form in the active and a certain other one in the middle from a given root, must be of very doubtful authority ; in fact, as regards the latter
most genuine part of the literature has been carefully and thoroughly excerpted for the Lexicon ; and for the Mahabharata we have now the explicit statements of Holtzmann, in his Grammatisches aus dem Mahabharata, Leipolg, 1884.