History & prehistory of Sanskrit
|History & prehistory of Sanskrit
TORY AND PRE-HISTORY OF SANS BY SUKUMAR SEN, M.A., Ph.D., F.A.S., Khaira Professor of Indian Linguistics and Phonetics, and Head of the Department of Comparative Philology, University of Calcutta °rice : f Ordinary Rs. 1-25 1 Special Rs. 2-25 r MYSORE : PRINTED BY THE DEPUTY DIRECTOR, GOVERNMENT BRANCH PRE PAMPHLETS LIBRARY TY UNIVERS
'sovn G.K.V.K. Library UNIVERSITY OF MYSORE EXTENSION LECTURES HISTORY AND PRE-HISTORY OF SANSKRIT BY SUKUMAR SEN, M.A., Ph.D., F.A.S., Khalra Professor of Indian Linguistics and Phonetics, and Head of the Department of Comparative Philology, University of Calcutta De A. S. BANGALORE UNIVERSITY LIBRARY, 26 OCT 197: 64733 ACC. NO. CL NU G.K.V.K. Library CONTENTS Page INTRODUCTION ... i I. HISTORY OF SANSKRIT ... II. PRE-HISTORY OF SANSKRIT 18 III. SPOKEN SANSKRIT 36 ABBREVIATIONS 52 BIBLIOGRAPHY <3 ... INTRODUCTION THE present monograph comprises the three THE lectures delivered at the University of Mysore on September 17, 18 and 19, 1957. The First lecture describes in brief the development of Old Indo- Aryan from Early Vedic to Classical Sanskrit through its two intermediate stages Late Vedic and Early Sanskrit. The name Sanskrit was unknown to Pāṇini. I have ventured to suggest that Kālidāsa was one of the first to use the name unmistakeably for the language. It may be noted that he does not use any name for Middle Indo-Aryan or Prākąta but refers to it as “the easy” language (Kumāra- sambhava vii. 90). Professor Sivarama Sastry tells me that the name Sanskrit occurs in the Rámāyana (Sundara-kāņda) where Hanumān first approaches Sītā and thinks about the language in which be would accost her. He is doubtful whether “Samskrta” speech, fit for a Brahmin, would be opportune as Sītā might then mistake him for Rāvaņa in disguise. This reference to the Samskſta speech is of course earlier than Kālidāsa's, but it probably means the “correct diction of a Brahmin ”. Anyway.“ Sams- kệta ”here (samskytām vācam) is an adjective. As a matter of fact our ancient forefathers did not recognise any foreign language. To them language either was "purified” or “cultivated ” (samskṛta) or “off-standard” (apabhrmśa) or « rustic" was ii präkyta). Speech was to them a divine thing; it was the force that impelled the arrow of Rudra, it was the immortal vitality that sustained the gods, and all that is good and auspicious in them was contained in speech (bhadraiņām lakşmir nihitadhi vāci). They also knew that speech is like a river (Sarasvati) flowing, meandering, fertilising and sustaining. It was the chaste speech that obtained ambrosia or Sõmu for the gods and it was the unchaste speech that brought defeat to the Asuras inspite of their seniority and superiority. The Second lecture deals with the earlier stages through which the language had passed before it took shape as Early Vedic. In this pre-bistory also there had been four stages; Proto-Indo-Aryan (we may also call it, less precisely, Pre-Sanskrit), Indo+ Iranian (or Aryan), Indo-European, and Indo- Hittite. I have also given some account of the morphology of Hittite and for two reasons. It has not yet been taken up for study by linguistic scholars and students in India. This is one reason. The other is that inspite of the remoteness of connexion between Hittite and Sanskrit a comparative study of the two languages is bound to be interesting in view of the fact that Sanskrit (and Aryan) has preserved some features that are attested only in Hittite. I have mentioned Heteroclisis. Another good instance is the imperative ending -tu (as well as entu). As these endings occur only in Indo-Iranian it was formerly thought that they are Aryan innovations. But after the discovery of Hittite their remote antiquity has been admitted. 111 The Third lecture is on “Spoken Sanskrit”. By Spoken Sanskrit I mean something different from what Sanskrit scholars think. It is not Sanskrit as has been spoken from the days of Patañjali (and earlier) to our own times. It means that basic form of Old Indo-Aryan which was spoken by the people in general and which was the direct and immediate source of Middle Indo-Aryan. Mr. Walter Maurer first drew my attention to the importance of this “Spoken Sanskrit”. I dealt on the topic very briefly in my presidential address to the annual general meeting of the Linguistic Society of India held at Deccan College, Poona, in November 1956. Thanks to the University of Mysore, I have now been able to find an occasion to develop the theme. I may mention here that beside Buddhistic (Hybrid) Sanskrit this “Spoken Sanskrit” had another and a much more debased form. It was used all over India by the half-educated and by ill- baked scholars as a sort of “Dog" Sanskrit. There are few records of this speech worth mention, but there is one very late record of some length. It is the very interesting book entitled Sekaśubhodayā com- piled in Bengal sometime in the fifteenth or the sixteenth century. September, 2007) SUKUMAR SEN. September 21, 1957 HISTORY OF SANSKRIT THE HE career of Sanskrit, or Indo-Aryan to be precise, started from the time of its separation from its immediate parent stock, Indo-Iranian or Aryan, which was a closely knit group of dialects spoken in Iran and in parts of Mesopotamia sometime before 1400 B.C. This may also be taken as the date of separation of Indo-Aryan from Indo-Iranian although there was not yet any question of the migration of Indo-Aryan into India. We do not know when the migration started. But there is no ground to assume that the Indo-Aryan speakers migrated from Iran in appreciable numbers before 1300 B.C. The date may be put still later, and it must be borne in mind that the migration was in waves and in successive times. The clans that migrated did not all belong to the same group nor was their language completely identical. 1400 B.C. as the date of the separation of Indo-Aryan is an assumption which is warranted on two grounds : (i) the existence of indisputable evidence of the proto-Indo-Aryan dialectal vocables occurring in some Hittite docu- ments, and (ii) comparative estimate of the language of the Ķgveda and of the Avestan Gāthās. The language of the Ķgveda takes us to circa 1000 B.C. and that of the Avestan Gáthas to circa 700 B.C. A common source would take us back to circa 1400 B.C., at a moderate estimate. 2 I do not propose to discuss here the date of the Rgveda. On linguistic and historical grounds it is safe to date it in the tenth-eleventh century B.C. It is the date accepted by linguisti. cians. Allowing 200 years for the incubation period the Indo-Aryan language may be presumed to have started on its career in the Indian soil from 1200 B.C. Before that the language was at the proto-Indo-Aryan stage. The first two hundred years (1200-1000 B.C.) may be assigned to the Early Vedic stage. By Early Vedic I mean the language of the older part of the Rgveda. During the next three hundred years (1000-700 B.C.) the language was at the Late Vedic stage. By Late Vedic I mean the language of the younger part of the Ķgueda and that of the Atharvaveda. The next three hundred years (700-400 B.C.) may be called the Early Sanskrit period (although the name Sanskrit was not given to the language till about 700 years later). Early Sanskrit includes the language of the Brāhmanis and Upanişads as well as of the Sūtras. The language described by Pāṇini was the culminating phase of this stage. The fourth stage started from 400 B.C. and it is Classical Sanskrit as we know it from Pāṇini's norm. Since then Sanskrit ceased to grow as a current language. Except in vocabulary this language which served as the literary speech par excellence through- out the length and breadth of India was virtually the same language as was used by the first students of Pāṇini. 3 If we trace the progress-by progress I mean chronological development--of the language through the four stages we notice the following tendencies gradually gathering force :- (i) The archaic (and hieratic) vocabulary is being replaced by the new (and secular) words coming from the popular speech or from other linguistic stock. At the last stage we find not a few words adopted from Middle Indo- Aryan. (ii) The accent (intonation) which was very much significant at the first stage is practically lost by the third stage. A kind of stress has come in its place but it is no longer significant. (iii) The compound.is becoming more and more popular so that at the last (i.e. classical prose) stage there is practically no limit to the number of its components. (iv) The primary affixes are losing popularity while the secondary affixes are gaining it. (v) The root nouns are gradually falling out of use, and consonantal stems are steadily changing over to vowel stems. (vi) Alternative (and dialectal) forms of the noun and the verb are being gradually eliminated. (vii) The instrumental case is fast losing its important idiom of concomitance. At the last stage an adnominal particle like saha is invariably used to signify concomitance. (viii) The injunctive mood disappears at the second stage and the subjunctive mood drops out at the third stage. The perfect loses its old value. 4 (ix) The bid of the future tense to grow into a system of conjugation proves abortive. The solitary modal form (subjunctive second person singular) in RV is karisyah. The later stages know only the conditional (i.e. the imperfect of the future): e.g. abhavisyat. i (x) The importance of the finite verb in the sentence is on the wane. The sentence tends to become nominal (i.e. without a finite verb). (xi) At the fourth stage the passive past parti- ciple (ending in ~dta) and the possessive from it (ending in -tavant) stand as rivals of the finite past tense. Thus : mayd krtam = aham akaravam. (xii) Adnominal adverbs (such as krte, artha-) are used increasingly as help words in the different oblique cases. (xiii) The prepositions tend to lose their free use and to become compounded as verbal prefixes. (xiv) The scheme of metres becomes more and more rigid but at the same time more and more prolific in variety. The language as a whole presents a homogeneous texture, and its general characteristics may be outlined as follows : — i. The phonemic system is simple. It com- prises the following speech sounds : Vowels a, i, u- short and long ; e y o— long. Diphthongs aU au — long. Sonants r— short and long; I— short. Semivowels y, v. 5 2. Consonants (stops) k, kh, g, gh; C, ch, j, jh ; t, th, d, dh ; t, th, d, dh; p, ph, b, bh. Nasals--n, ñ, n, n, m. Continuants—r, l. Fricatives š, ș, s, ḥ-unvoiced; h-voiced. All phonemes except four-n, ñ, ņ and ḥ-may occur in any position; ņ may occur in any position other than the initial and the final; ḥ may occur only after a vowel finally and also medially before an unvoiced non-dental and non-retroflex; n and ñ occur only in non-initial conjuncts. In sandhiḥ is changed into r after a vowel other than a and ã and before a vowel or a voiced consonant. Under the same conditions ḥ after à is lost and after a it combines with it into o. New words can be derived more or less freely by adding affixes to roots or to words. Some- times there is no affix but the basic vowel may undergo a change (Guņa or Vļddhi). Words may combine into compound words. 3. Flexionally words fall into three classes : the substantive, the verb, and the indeclinable. The first two are variable in flexion and the last --origi- nating mostly from the substantive -is not subject to variation in flexion. Che substantive comprises nouns, adjectives and pronouns and are declined in three numbers (singular, dual and plural), three genders (masculine, feminine and neuter) and eight cases (nominative, vocative, accusative, instrumental, dative, ablative, genitive and locative). There are several declensional patterns based on the fina phoneme or on the gender or on both. 6 The verbal forms include the non-finite and the finite. The non-finite originally belongs to the substantive but as some of them (participles) admit of tense variation (i.e , present such as gacchan, aorist such as gran, perfect such as jugmivān and future such as gamisyjan) and all of them may govern a case like the finite verb they are grouped under the verb. The finite verb presents a very large variety of forms. It distinguishes two voices (active or parasmaipada and middle or ātmane pada which includes also the passive), three numbers (singular, dual and plural), three persons (first, second and third), five tenses (present, future, imperfect, aorist and perfect) and five moods (indicative, subjunctive, optative, imperative and injunctive). There are also secondary conjugations such as the causative, the denominative, the desiderative and the frequentative. There are different temporal and modal affixes, and the personal endings are often different in the different tenses and moods. Thus : - a., - aya -, - nå -, - nu - , - ya - etc. in the present tense ; reduplication in the perfect; - S - etc. in the aorist; - a - in the subjunctive etc., etc. 4. The language being entirely flexional there is not much scope for formal syntax. Each word bears the sign of its exact relation with other words in a sentence. There are however idioms in the use of cases and moods and tenses. The general characteristics of the our stages of Old Indo-Aryan may now be discussed. 7 I. Early Vedic, as already mentioned, is represented by the earlier of the Rgvedio hymns, but as the text handed down to us has had received some modifi- cations at the time of its collection into the Ķk-sam- hitā we are sometimes obliged to resort to orthoepic restoration as warranted by internal evidence, specially by the metre, by the readings of the Pada pātha, by the observation of our ancient phone- ticians and by external evidence culled from Old Iranian (Avestan and Old Persian). To give some instances : pāvaká - is always to be read pavāká - (cf. sayaka - 'missile', nabhāka - a name); virāṣāt (also in the Pada pāțha) for vīrașāț; dukșata (i 160 3) should be read dhukșata (as in the Padapātha); etc. The vocabulary is largely peculiar, both in the substantive and in the verb. Any stanza taken at random from RV would show that not less than fifty per cent of the words are absolutely unknown in Classical Sanskrit. Sometimes the word is known only in a very different sense. I give two instances, one from the later and the other from the earlier stratum. bíranyahasto ásuraḥ sunītháḥ sumrlīkáḥ suávāṁ yātu arvā'n apsédhan raksảso yātudhānān ásthăd deváḥ pratidoşám grņā'nah. Here there are fourteen words of which seven (sunīthaḥ, sumrlīkaḥ, svavān, arvan, apasedhan, pratidoșam, grņānaḥ) are unknown in Sanskrit, and two (asuraḥ, asthāt) are used in quite different meanings. 8 śámsā mitrásya váruṇasya dhā'ma súşmo rodasi badbadhe mahitvā' áyan mā'sā ayajvanām avī'rāḥ prá.yajñamanmā vșjánam tirāte. Of the sixteen words here eight (samșă, suşmaḥ, badbadhu, mahitvā, ayajvanām, yajñamanmā, vrj inam, tirāte) are quite unknown, one (rodilsī) is known but only as a borrowing from Vedic and one (pra) is unknown in such free use. 2. Intonation is significant. That is, the position of the accent often determines the semantic value and the grammatical function. Thus : rā'j uput- rah (Bahuvrihi): rājaputráḥ (Tatpuruṣa); súkrtam. “well done”: sukrtám “good deed”; dā'ta vásūni "giving riches”: dātā vásūnām "giver of riches "; brahman- (n) “prayer": brahmán- (m) one who prays”; yáśas- (n) “glory”: yaśás- (m)“glorious”. 3. e certainly and e probably are still short diphthongs, i.e. they are pronounced as ăi and ňu respectively. This pronunciation is indicated in the Prātiśākhyas by mentioning them as sandhyaksara (“ sandhi vowels”). Internal as well as external evidence supports this. In RV jyestha, and śr. ștha- are often to be pronounced as trisyllabics : jyaiştha- and śraistha -. (So in Avestan sraēsta-.) The derivation of these words warrant diphthongs. Similarly tredhā (i 154.I) is to be pronounced as traidha. The evidence for o is not clear. It is not unlikely that au had already become a monophthong. The only clear evidence in support is somya- beside saumya - . The latter (and later) form originated from the diphthongal pronunciation. 9 .. i 4. Hiatus is not avoided internally or exter- nally although the Pada text often disguises it. Thus : dáivya - (i 35 5)=dáivia- ; kvà=kúa; svá pāḥ (i 85.9)= sué pāḥ; tvám=tuám (cf. Old Persian tuvam); gām (often)=gāam<*gāvam): góḥ (often)
- gávaḥ; saptāsye (iv 51 5)=sa ptá-āsye; híranyahasto
ásu raḥ; etc. There is much freedom in Sandhi as is expected in a language not remote from the spoken tongue. 5. Cerebralisation of a dental occurs in external sandhi also; e.g. pári ștha (vii 1037); prá ņa āyur jīváse soma tārīḥ (viii 48 4). 6. There is a marked tendency of dropping the initial sibilant in an initial consonant-group; e.g., candra, paśyati, tāyu, tārā, tisthsti; but hariscandra-, spasta“, steya-, str., sthäsyathi etc. The sibilant is however retained when the conjunct contains an aspirate or a liquid. Thus : sthāna-, sphärä-, śrīmant-, slīla., etc. 7. The palatals 6, ch, j, jh, are still prepalatal stops and not affricates as later. 8. A final conjunct is simplified (unless it is protected or medialised in a phrase or word-group): gacchan (but gacchamsca), akar (for *akart), etc. But -rk, -rt, and -rt (when -g- belong to a root) are retained : vark (<vrj- ; *varkt), urk (<urj-s), amārt (<amarj-t), āvart (<vrt-t). 9. Root nouns are very common; e. g., dā-, bhid,- vrdh-, stu-t-, spas, etc. 10. Action and agent nouns in -man also are numerous : mánman-, ájman-, jániman-, vidmán-, jeman-, dáman-, etc. 10 11. Agent nouns in tr- when used participially generally govern the accusative. Thus: g.intr-, dātr- but ganty, dātr'-(with the genitive); tårutr-'winning: tarut;'. 'winner.' 12. Abstract feminine nouns in -täti and-tatare not out of use. Thus: sarvitāti, sarvátat, jyeșthátati-, devátāt. Later we have only sarvatā and sarvatva-. 13. Peculiar are the abstract neuter nouns in -tvaná (sakhitvaná- 'friendship’, janitvaná- 'wife- hood'), adjectives in -ya (paśavya- <paśu-, śaravya- <šaru-), diminutives in -la (vrşala- 'little man', śiśula-); feminine abstract nouns in -vát from prepositions (nivát-, parāvát-, udvát-), and possessive adjectives in -van, -vin, and -vant (very frequent): maghavan-, samadvan-'warlike'; aśvavant- a ſtāvant-; nȚvant-manly’; ubhayāvin-; etc. 14. The compound does not admit more than three words; usually it contains two components only. The governing compound (e.g. śikṣānara-, vidadvasu-) is a notable feature. The Babuyrīhi compounds ending in -t or ? do not add the pleonastic (samāsanta) affix -ka, but the affix -ya appears in some Bahuvrīhicompounds; e.g.dašamāsya-, madhu hastya-, viśvajanya-. 15. The dual is in regular use but it is not as fully represented in declension as the other two numbers. The dual is the number when a natural pair or a closely associated duality is to be emphasised (e.g. pitarā, mitrā - varuņā, akși, cakre), but even then we occasionally find the plural in place of the expected dual ; e.g sam añ jantu višve davāḥ sam āpo hrdyāni (for hydaye) nau x. 85 47. II nom. - 16. Alternative declensional forms are numerous e.g. devā: devena (instr. sg.); devāiḥ : devebhiḥ (instr. pl.); deva : devāu (nom. - acc. du.); vişņaụ : vişnavi (loc. sg.); rājan: ra jani (loc. sg.); yugā: yugāni (nom. acc. nt. pl.). 17. The stem ava- in the first personal pronoun dual is absent. The compound demonstrative tya- (syaḥ, syā, tyat, etc.) is very common. The historical acc. sg. nt. of the interrogative pronoun (kat) has not yet lost its pronominal value. 18. The injunctive mood is still living; the subjunctive mood is quite vigorous. There are all possible modal variations of the tenses (e.g. gaccha pres imp.; gahi aor. imp.; mumugdhi perf. imp. ; gacchet pres. opt. ; gamyāt aor. opt.; jagamyāt perf. opt.; muñcāt pres. subj. ; mumocati, mumocat perf. subj. ; mucti-aor. subj.) 19. The use of the three past tenses is idiomatic: the imperfect is the past tense of narration; the perfect characteristically expresses the condition attained by the subject as the result of a preceding action; the aorist is neither narrative nor durative but expresses a past action with reference to the present. The present and the present perfect of English are often the equivalents of the Early Vedic perfect and aorist tenses respectively. 20. The augment is sometimes not used (impf. and aor). 21. There is a large number of gerunds (in -tvī, -tvā, -tvāya, -ya, -ya, -tya) and infinitives (acc., dat., abl. -gen. and loc.). 12 22. The prepositions are often free, i. e., they are frequently used independently of the verb. There are also adnominal prepositions (or adverbs) that govern cases; eg. kam acchā yuñjāthe rathan (v 74 3); vrstim ava divā invantam (vii 642); jāyā patim vahati vagnunā sumat 'the wife weds the husband with a shout of joy' (x 32 3); devesu ā; etc. 23. antar precedes the word it governs ; (as in Old Persian); e.g. kadă nu antar varu ne bhuvāni (vii 86 2). At the next, i.e., Later Vedic stage, the language presents the following special features :- Old (and hieratic) words are going out of I. use. 2. some The sandhi phenomenon assumes rigidity. a is elided after -o<-as; e.g. so'pi <saḥ api. 3. e has become a monophthong. 4. There is some resistance to the cerebralisa- tion of dentals; e.g. vistara- beside earlier vișțara-; sustha- beside older sușthu, 5. Derivatives in -tat and -tati go out of use; man and -van are replaced by -mant and -vant respectively. 6. The root-noun declensions are becoming defunct (e.g. gopa->gopa-), and consonantal stems are being replaced by the vowel stems (e.g. nakt-> nakta-, spas-> spaša-,' ni->nara-). Alternative declensional forms are on their way to elimination. 7. The older patterns of i, -ī and -u declensions are dying out. That is, forms like veḥ (pom. sg.), 13 arya) (gen. sg.), pašva (instr.sg.) and vişnavi (loc.sg.) are no longer in use. The only exception is sakhi-. 8. The secondary demonstrative tya- is going out of use. The āva- forms of the first person dual are established. 9. The root kr- prefers a new conjugation (karoti in place of krņoti). 10. The conjugational variety is reduced to an appreciable extent. The injunctive mood is practically lost. The precative is fully established. There is an abortive attempt of the future tense to develop modal variation (as evidenced in Epic Sanskrit words like karisyatu fut, imp.). Some of the old gerunds and infinitives go out of use. Only -tvā, -ya, -tya (gerunds) and -tum (infinitive) remain. The infinitive in -tum rarely occurs in RV. II. The difference between the Early and the Late Vedic stage is chronological as well as dialectal while the difference between Late Vedic and Early Sanskrit is mostly chronological. The main features of Early Sanskrit are the following: 1. The alternative forms in declension are almost completely eliminated. The conjugational patterns are reduced in number and also simplified. The modal forms of the aorist are almost lost, and those of the perfect are completely lost. The subjunctive is on the way to extinction. Only a few forms of the perfect 2. 14 participle survive (cf. Paņini's sūtra : bhaṣāyam sada- vasaśruvaḥ). The periphrastic perfect (e.g., gamayām cakāra AV)and the periphrastic future (e.g., anvāgantā ya jñapatir vo atra VS, TS) are in common use. The passive past participle is established as a substitute for the finite verb, and its possessive form is coming into being (asitivant- 'having eaten’ AV ix 6.38 ; v., l. aśitā'vant-). 3. The extended primary derivatives in -anīya (two occurrences only in AV, amantranīya- and upajīvantya-) and in -tavyà (also in two occurrences only in AV, janitavyà- an't himsitavyà-) appear. 4. A large number of secondary derivatives, particularly patronymics, find currency. 5. The pleonastic affix -ka begins to be added to the Bahuvrīhi compounds ending in -1, -ū or -7.
- Thus gomātěkāḥ for the earlier gomātaraḥ.
6. The dative singular feminine is used also for the ablative and the genitive. Thus : alalyāyai jāraḥ, nadyai payaḥ. This has a parallel in Younger Avestan. 7. The perfect tense is used for the remote and the narrative past. This is no doubt an archaism. 8. The importance of the finite verb in the sentence is considerably lowered. The nominal construction is gaining in importance. An extreme development of the nominal sentence is the condensed sūtra style where the finite verb has been practically dispensed with. 15 Classical Şanskrit does not merely reflect the chronological sequence of Early Sanskrit. It is a literary language that is based on the speech of the educated men (sista) of the midland (Madhyadeśa). At the same time it contains features which really belonged to the dialect of the North-west (Udīcya), the mother-tongue of Pāņini, the codifier of Classical Sanskrit. During Pāṇini's days the language was uniform as regards the general grammatical structure but there were dialectal differences in vocabulary and syntax. Páħini refers to two broad regional dialects mainly: the Eastern (Prācām) and the North-western (Udīcām)'. Small dialect-groups are seldom men- tioned, but one rare exception is the mention of the trans-Bias dialect area (udak ca vipāśaḥ 4. 2. 74). After Pāṇini the slight dialectal differences that existed in the learned language were levelled out completely, and the old historical dual compounds like mātarapitarau which Pāṇini mentions as belong- ing to the North-west remain merely as grammarians' curiosities. Classical Sanskrit is very familiar to us all. Its main features therefore may be given in the broadest outlines:- 1. The sandhi rules are rigid. No hiatus is allowed unless it is caused by the disappearance of an Literally the words prācām and udicām mean 'according to the easterners' and according to the northerners respectively. The reference of to the grammarians of the eastern and northern regions. was course 16 intermediate sound (e.g. tāḥ etaḥ>tā etāḥ; tapah- āsanam>tapa-āsanam). 2. Vocabulary is enriched by the incorpora- tion of new material. 3. The power of combining words into a compound is almost unlimited. Alternative declensional forms are entirely eliminated. New analogical forms for the neuter are established (e.g. madhune, madhunaḥ, dātīņi etc.) 5. The subjunctive mood is lost; but its first personal forms are retained as the corresponding forms of the imperative. The precative, (originally a peculiar form of the optative aorist), is a very restricted idiom. Only the present forms of the two moods, the optative and the imperative, survive. As in the two earlier stages the injunctive survives as an alternative idiom only when the prohibitive particle mā is used, but then the imperative is an alternative idiom. Thus: mā kārșīḥ ; mā kuru. 6. The passive and the possessive past participle are accepted in the value of the perfect (nisthā). The possessive past participle does not become a popular idiom, and possibly it had really belonged to the dialect of the North-west as Khotanese (a Middle Iranian neighbouring language) indicates. The following table illustrates some points of morphological development of Old Indo-Aryan through its different stages. 17 nil : 1 webhone: -aih} }s:
nil : 1 . Ratio in Ratio in Late Ratio in Morpheme Early Vedic Sanskrit Vedic (AV) Nom-acc. du. 7:1 I: 2 nil : I -ā: -au Instr. pl. 221 : 211 I:S Pres. first per sg. -masi : -mas. 3:4 The development of the four stages in OIA is illustrated below. (A) EARLY VEDIC yas tā vi jānāt sa pituspitãsat 'He who should know them would be the father of the father' (RV i 168 16d). (B) LATE VEDIC yas täni veda pituspitāsat (AV 2 1.2d). Here we find two changes : tāni for tā and veda (unreduplicated perfect=present; cf. āha, āsa) for vi janāt (subjunctive). The second half is a repetition for RV. (C) EARLY SANSKRIT (RECONSTRUCTED) yas tāni veda sa pituh pitāsa. Here we do not expect internal sandhi between pituļ and pita and the subjunctive asat would be replaced by perfect (=present) āsa which is still in use. (D) CLASSICAL SANSKRIT (RECONSTRUCTED) yas tāni vetti sa pituḥ pitā bhavati. At this stage the few remnants of unreduplicated perfect (=present) are gone out of use, and as- has become purely a substantive verb. sa i PRE-HISTORY OF SANSKRIT Four stages backward from Early Vedic bring us to the ultimate discernible source of the language. These are Proto-Indo-Aryan, Indo-Iranian, Indo- European and Indo-Hittite. The Proto-Indo-Aryan stage continued roughly for two hundred years (1200-1400 B.C.). The Indo-Iranian stage possibly óccupied the next five hundred years (1400-1900 B.C.). For Indo-European and Indo-Hittite we may assume periods of five hundred years each: Indo-European from 1900-2400 B.C. and Indo-Hittite from 2400- 2900 B.C. For a full understanding of the Proto-Indo-Aryan stage it is necessary to have a picture of the Indo- Iranian stage. Indo-Iranian (or Aryan) as a separate branch of the Indo-European stock of languages possesses the following common characteristics. 1. The phonemic system is as follows :-} | Vowels a, i, u--short and long. Diphthongs ai, au-short; āi, au-long. Sonants ?—short and long. Consonants (stops) k, kh, g, gh: c, ch, j, jh t, th, d, dh; 7, th, b, bh. Nasals ń, ', n, m. Continuant r. Fricatives s, š, s-unvoiced; s', ze, 2- voiced; t'h, zuh-voiced and aspirated. Possibly also x, v. (18) 19 n 2. There are two sets of palatals. One is derived from IE pre-palatal stops : ś, sh, z, ;'h<IE k, kh, , gh. The other group comes from IE velars and labio-velars when followed immediately by a palatal vowel: c, ch, i, jh, <IE q (qu), qh (quh), G (Gw), Gh (Gwh). 3. The declensional and conjugational patterns are not much different from that of Old Indo-Aryan but there are larger number of alternative and archaic forms and much less uniformity. A word may end in a consonant group. The augment (in impf and aor.) is generally not used; Of the innova- tions two are of special importance. In declension the influence of the pronoun is noticeable. In conju- gation the primary endings are used also in the subjunctive mood (e.g. bhavāti beside older bhavāt). There is not yet a future tense, futurity as such being expressed by the present or by the subjunctive or implied by the context. There are two main dialect-groups in Indo- Iranian : (i) Proto-Iranian and (ii) Proto-Indo-Aryan. The earliest evidence of Proto-Indo-Aryan is to be found in the word aika-'one' (as against Proto-Iranian aiva-) and in some names of gods (e.g. Varuņa, Näsatya, Indra) occurring in some Mitannian (Hittite) documents dated in the fourteenth century B.C. od The main points of disagreement between Proto- Iranian and Proto-Indo-Aryan are as follows :- (a) The voiced aspirated are retained in PIA but become deaspirated (or very rarely) devoiced in 20 PI. Thus II* dhāta->PIA* dhāta-, PI* dāta-; II* maugha-> PIA* maugha- (OIA mogha-), PI* mauga- (Av. maoya-); II*az'ham> PIA * azham (OIH aham), PI* azam (Av. azam, OP adam); II *bhavati > PIA bhavati, PI* bavati (Av. bavaiit, OP bavatiy); II* jhanti > PIA* jhanti (OIA hanti), PI *zanti (Av. jainti); etc.' (6) The voiced sibilants (z', z',z'h, 2h including i, jh)are retained in PI but they are changed intojl<, j) or h (<z'h, jh) when intervocal in PIA. Before an unvoiced consonant i I *z, j become so in PI and ch in PIA. Thus : II *yaz'ati > PI *yaz'ati, PIA yajati ; II *jas'svati > PI *jasati (Av. jasaiti), PIA *jacchati; (for gacchati); II* iz'ta-> PI ista-, PIA işta-. (c) Before a voiced consonant II z',z", a'h become zor ó' in PIA but are lost subsequently. In conse- quence there is a mutation of the preceding short vowel and a following dental is cerebralised. Thus : II* azdhi > PIA* ardhi > OIA edhi (Av. zdī); II* mazdhā> PIA *mazdhā> OIA medhā (Av. mazdā); (I* nizda-> *niz'da-> PIA niz'da->OIA nida-;' II*uz'hta->PIA* uz'dhr<OIA ūąha-. (d) Glidic (or non-glidic) s and in a consonant group is disappearing in PIA. Thus: II *badhsta- (<badh-t-1.1-)> PI basta-, PIA baddhi-; II*vidvazd- this->> PIA ridvadbhis; etc. (e) Final -50 (i.e. the palatalised s after i or u) in sandhi position before a vowel or a voiced consonant becomes voiced zu which becomes r in OIA. Thus : Il* nis-āyam 'I came out'> PI, PIA* niz'dijām Pali nedda (<nidda-?) presupposes not the less of PIA z' ittub assimilation with the following d (<d). 21 (cf. OP nijayam)> OIA nirdyam; II* du$manas> PI, PIA *duz v manas> OIA durmanah. Before r it disappears and lengthens the preceding short vowel. Thus: II *nisrauga-> PI, PIA* niz v rauga-> OIA niroga-. Dialectally in PIA this z v becomes z and is lost before other voiced consonants ; e.g. II *dus v dabha-> PIA *duz v dabha-, duz'dabha->QLA dudabha-( 7 edicy (J) The operation of Bartholomae's Law started at the Indo-Iranian stage and it is perhaps more strictly operative in Proto-Indo-Aryan. The law indicates that a consonant-group consisting of a voiced aspirate and an unvoiced aspirate changes into a group of voiced non-aspirate plus voiced aspirate. The phenomenon is a kind of mutual assimilation and meta thesis combined. Thus : II *dughta-> IA dugdhd-', II *uz'hta-> PIA
- uz'dha->01AiidhJ--,Il *basfhta-> PIA *baz'dha->
OIA badha-; II* trnaz'hti> PIA* trnaz'dhi> OIA trnedhi ; etc. (g) Proto-Indo-Aryan has, like Proto-Iranian, only r and r (for both long and short). In Old Indo-Aryan r has a dialected tendency of changing into I. (This tendency was probably already there in Proto-Indo- Aryan.) I in Sanskrit is a variant of /*, and r (long) is an analogical creation in the written language. As a matter of fact r (long) occurs only in some plural case forms of some nouns, to be exact in nom-acc, pi. nt. and in acc. and gen. pi. of nouns in -r (e.g. datr'ni, filfn,matr [ h and bhratfnam). Thenom.-acc. nt. form is very late ; it does not occur in Vedic and is not generally met with in Classical Sanskrit literature. The acc. pi. and gen. pi. forms are not historical पृष्ठम्:History & prehistory of Sanskrit.djvu/३३ पृष्ठम्:History & prehistory of Sanskrit.djvu/३४ पृष्ठम्:History & prehistory of Sanskrit.djvu/३५ पृष्ठम्:History & prehistory of Sanskrit.djvu/३६ पृष्ठम्:History & prehistory of Sanskrit.djvu/३७ पृष्ठम्:History & prehistory of Sanskrit.djvu/३८ पृष्ठम्:History & prehistory of Sanskrit.djvu/३९ पृष्ठम्:History & prehistory of Sanskrit.djvu/४० पृष्ठम्:History & prehistory of Sanskrit.djvu/४१ पृष्ठम्:History & prehistory of Sanskrit.djvu/४२ पृष्ठम्:History & prehistory of Sanskrit.djvu/४३ पृष्ठम्:History & prehistory of Sanskrit.djvu/४४ पृष्ठम्:History & prehistory of Sanskrit.djvu/४५ पृष्ठम्:History & prehistory of Sanskrit.djvu/४६ पृष्ठम्:History & prehistory of Sanskrit.djvu/४७ पृष्ठम्:History & prehistory of Sanskrit.djvu/४८ पृष्ठम्:History & prehistory of Sanskrit.djvu/४९ पृष्ठम्:History & prehistory of Sanskrit.djvu/५० पृष्ठम्:History & prehistory of Sanskrit.djvu/५१ पृष्ठम्:History & prehistory of Sanskrit.djvu/५२ पृष्ठम्:History & prehistory of Sanskrit.djvu/५३ पृष्ठम्:History & prehistory of Sanskrit.djvu/५४ पृष्ठम्:History & prehistory of Sanskrit.djvu/५५ पृष्ठम्:History & prehistory of Sanskrit.djvu/५६ पृष्ठम्:History & prehistory of Sanskrit.djvu/५७ पृष्ठम्:History & prehistory of Sanskrit.djvu/५८ पृष्ठम्:History & prehistory of Sanskrit.djvu/५९ पृष्ठम्:History & prehistory of Sanskrit.djvu/६० पृष्ठम्:History & prehistory of Sanskrit.djvu/६१ पृष्ठम्:History & prehistory of Sanskrit.djvu/६२ पृष्ठम्:History & prehistory of Sanskrit.djvu/६३ पृष्ठम्:History & prehistory of Sanskrit.djvu/६४ पृष्ठम्:History & prehistory of Sanskrit.djvu/६५ CHNo...." 64733