Sanskrit Language and Literature

and neut. in –a and fem. in (corresponding to the –o and –CI stems of Greek and Latin), in which the accent was fixed either on the stem or on the termination throughout the paradigm without any change in the form of the stem, of the type divali =Gk. hippos, Lat. equos. There had, moreover, been a tendency in the pre-Sanskrit period to fix the form even of the variable stems, especially those consisting only of a root : thus the long vowel was generalized in vac– "voice" (nom. yak): But in the process of normalization the vowel-declension (and especially with the stems in –a –a) was predominant, and even in Sanskrit itself there was a considerable transference of con sonant-stems and root-stems to this declension : e.g., pada "foot," dvdra– "door," Banta– "tooth" replaced pad— pad–, dant dat–, dur–. This proceeded apace in the spoken tongue, for the declension of the modern languages rests exclusively on vowel-stems and of these chiefly on the –a –a stems.

As in Indo-European, the pronominal declension differed from the nominal, that of the personal pronouns radically, that of the demonstratives and others to a varying degree. In the personal pronoun each number had a different root (t [u] vdm "thou," yiiyam "you"), while that of the first person had different roots also for the nominative and the oblique cases (andm "I," mdm "me," vaydm "we," asrndn "us"). The tendency of the two declensions, the nominal and the pronominal, to influence each other is observable : e.g., the nominal ending of the neuter sing. nom. and acc. --m as in kim "what?" (=Lat. quid) replaced the–d or –t still found in other pronouns (tat "that"=Lat. is-tud). In the spoken language this mutual influence continued to react until practically all distinction between the two types of de clension was lost.

Numerals.

The numeral-system, built upon a decimal basis, is that of Indo-European. There are separate names for the numbers up to 1o; from I z to 19 compounds of the units with the word for ro partly correspond with those of Greek and Latin: e.g., d(u)vddaia "I 2"= Gk. dedeka, Lat. duadecim. Above that the tens (2o, 3o, etc.) were in origin probably compounds ex pressing a number of tens, and the intermediate numbers were formed by compounds of the units with these : e.g., paiica "s" (Gk. pente, Lat. quinque), paiicakit "so" (cf. Gk. pentekonta, Lat. quinqueiginta), paiicapalicaiat "55." There are separate words for roo and 'coo. The special word for roo,000 whence modern liikh) is post-vedic. The numerals from r to 19 are adjectives; the rest are substantives.

Gender.

Traces of the distinction of gender between animate (=masculine and feminine) and inanimate (=neuter) are seen in the use of the neut. udakdm "water as a medium of floating, etc.," and the fern. tipah "water as personified as sentient beings." But for the most part in Sanskrit gender is grammatical and largely independent of nature. except that male and female living beings are usually (though not always) respectively masculine or feminine. There were three genders—masculine, feminine and

neuter. The neuter was distinguished by its termination (or lack of termination) in the nominative and accusative, and by the fact that the form of the accusative was the same as that of the nominative. But the masculine and feminine substantives were primarily distinguished only by the form of an adjective, if there were one, in agreement with them (as often, e.g., in French or German). But there was a tendency to reserve certain types of stem for one or other gender. Thus the –a stems in Sanskrit are reserved only for masculine and neuter nouns (although in Greek they may still be feminine, e.g., he hippos "the mare") ; and the –a stems are mostly feminine. In the older language the stems might be masculine or feminine; in the later language they are almost all feminine. Similarly even in Classical Sanskrit stems are either masculine or feminine; but in the spoken language there was a growing tendency to confine them to feminine : thus words of this declension which are masculine in Sanskrit become fem inines in the modern languages : agniij m. "fire" (--=Lat. ignis) becomes Hindi ilk f.

Verbs.

In the Vedic language the verbal system is of con siderable complexity. A verb might have various stems, viz. present (sometimes more than one), aorist (three in number, root e.g., dstham=Gk. esten, strong, e.g., aricam=Gk. sigmatic, e.g., araiksam = Gk. e/eipsa) ; perfect (characterized by reduplica tion and peculiar terminations) ; future (rare in the old language). The significance of the stems lay in the mode of action they ex pressed. Very generally speaking, the present indicated con tinuous action, the aorist momentary action, the perfect a state resulting from past action. The various present stems indicated various types of present-stem action, such as intensive, repetitive, inchoative, causative, desiderative, etc. Each of the first three stems had five moods—indicative expressing fact, injunctive and subjunctive expressing will and futurity, optative expressing wish, imperative expressing command. In the indicative of the present, perfect and future stems there were two tenses, present and past; the aorist stem was in the indicative confined to the past tense. Each tense had three persons and three numbers—singular, dual and plural. Finally each tense could be conjugated in two voices with different terminations—active and middle. Among the parts of the infinite verb there was connected with each stem a participle which could be either active or middle, and inde pendent of tense stems a past participle, one or more infinitives (for the most part case-forms of verbal nouns), a gerundive and an indeclinable participle or gerund. The total number of possible forms belonging to any one verb is thus very great.