Wilson Philological Lectures on Sanskrit and the Derived Languages c12: Lecture VII. Relations between Sanskrit, Pâli, the Prâkṛits and the Modern Vernaculars
Wilson Philological Lectures on Sanskrit and the Derived Languages c12
- Language
- English
Contents
Lecture VII. Relations between Sanskrit, Pâli, the Prâkṛits and the Modern Vernaculars
c12.2
Thus, gentlemen, have we surveyed the whole field of Indian Aryan speech from the Vedic times to our own, and what is it we have found? The Vedic dialect lost a great many of its words and some grammatical forms, its nouns and roots arranged themselves under definite declensions and conjugations, and thus became what I have called Middle Sanskrit. Till then there was no phonetic corruption except in isolated instances. But after that time the process went on rapidly, and words were altered in form principally by the law of the assimilation of conjunct consonants and a few vowel changes. By the law of generalization or false analogy the declensions and conjugations were reduced to the prevailing types; some new words came into use and a few old ones became obsolete, and thus the language arrived at a stage of which the Pâli, the sacred language of the Southern Buddhists, and the inscriptions of As'oka afford us specimens. Then began a general use of attributive expressions for verbal forms, which had its effect on the vernacular or derived languages also, and thus while a good many of these forms went out of use in Sanskrit, the number of tenses and moods in the derived languages was reduced to three. In the meanwhile the processes which formed the Pâli went on, other laws of phonetic change, and principally that of the elision of uninitial consonants and the reduction of aspirated ones to ह, came into operation. The use of analogy in the simplification of grammar proceeded further, a few words of a non-Sanskrit origin were adopted, and thus arose the Prâkṛits. There were several dialects of this class, but the distinction between them was not great. After a while in the Prâkṛits themselves new phonetic processes began, the chief of which were the dropping of one of the assimilated consonants and in most cases lengthening the preceding vowel, and the obviating of the hiatus caused by the vowels brought together by consonantal elision by combining them or inserting light semi-vowels between them. These were carried on further, the declensional and conjugational terminations were corrupted and mutilated, and the forms began to get confused; case relations came to be expressed by connecting independent words or pronominal forms with the genitive of the noun; and new verbal forms were constructed from the participles by appending the personal terminations on the analogy of the existing tenses, or by using the roots indicative of existence as auxiliaries. About this time the Prâkṛit speaking people came to be divided into independent communities, separated from each other by hills and rivers, or communication between them decreased; and therefore these several processes of destruction and construction were carried on at a varying rate and with a difference of detail in each; and thus grew up the modern Hindî, Mârâṭhî, Gujarâtî, Sindhî, &c. This is the whole history. The vernaculars have descended from the Prâkṛits, and the Prâkṛits from Sanskrit; and since a language is transmitted from generation to generation orally, if the modern dialects are spoken languages the Prâkṛits must have been so, and if the Prâkṛits were spoken dialects the Sanskrit must have been so.
c12.3
But these conclusions have been denied by some scholars. Professor H. H. Wilson says that the Prâkṛit of the plays is an artificial modification of the Sanskrit language devised to adapt it to peculiar branches of literature. And the reason he gives is that even in later plays composed but a few centuries ago the Prâkṛit used is not the vernacular of the day, but that which we find in the very earliest work of the species. But he is by no means certain, and leaves the question undecided. The reason advanced by him is evidently of no weight. If I choose at this day to express my thoughts in Latin and compose a treatise in that language instead of my vernacular, does that prove that that language is artificial? Latin is sometimes still used in Europe for literary purposes, though it is not the vernacular of any country. Sanskrit playwriters of later ages, and even a stray dramatist at the present day use the old Prâkṛits for their inferior characters in orthodox obedience to the rules laid down by the old authorities on the histrionic art; and since the Prâkṛits, like Sanskrit, Latin, and other languages are preserved by means of grammars and literary works, it is possible to write in them as in these languages.
c12.4
Mr. Beames at one time expressed himself decidedly on this point: "The Prâkṛit of the poets is clearly not a dialect that ever was spoken." So Dr. Pischel, a German scholar: "I agree with Mr. Beames, that none of the Prâkṛits was ever a spoken language, and that in order to learn what was the spoken language of the Aryans we must turn principally to the modern vernaculars." This, gentlemen, we have done; we have examined the vernaculars and found that there is not a rule about the changes of letters in passing from Sanskrit into the Prâkṛit given by Vararuchi or Hemachandra, of which instances are not found in the vernaculars. Of such as are general the examples are abundant; and in a great many cases even the words in which the change is of a special kind are preserved in the vernaculars in their Prâkṛit form, slightly altered according to the processes that afterwards came into operation. A good many of the vocal peculiarities of the Pâli and Prâkṛit speakers are preserved by the speakers of the Gujarâtî, Marâṭhî, Hindî, Sindhî, and Bangâlî. I have already drawn your attention to the fact that the ordinary Gujarâtî cannot pronounce ए or औ but makes ए or ओ of it, the Marâṭhî De'sastha ण of न, and the Sindhî and the Bangâlî ख्व of क्ष, and the Hindî ख्व or ख of it and स of the sibilant. The Bangâlî pronounces a conjunct ending in य as a double consonant and preserves the old Mâgadhî peculiarity of reducing all the sibilants to श, while the Hindî still manifests the Pais'âchî peculiarity of making न of ण. The case terminations in the forms they assumed in the Apabhramṣ'a have been perserved, some in this dialect, some in that. All the three Prâkṛit tenses do exist and none other; past time is expressed in just the same way as in the Prâkṛits; the participles, the causals, the passives, the denominatives, the nominal and verbal bases, and the remnants of the Sanskrit conjugations are just as they were in the Prâkṛits; and the new verbs and cases are formed out of Prâkṛit materials only. Then again a good many of the Des'ya or non-Sanskrit words that existed in the Prâkṛits are found in the vernaculars.
c12.5
Now if the Prâkṛits were not spoken dialects, what these scholars probably mean is that they were artificial adaptations, as Professor Wilson calls them. But we have seen that they grew up by such natural laws as guide the development of any language whatever. The development of the Prâkṛit dialects from Sanskrit is paralleled by the development of the Romance languages from Latin. If the Prâkṛit dialects are to be considered artificial, it is difficult to conceive upon what principles they could have been constructed and for what purpose. A conscious manufacture of a language would be conducted upon some general principles and would not admit of such isolated forms, not obeying any general rule, as we have noticed. The analogies would be perfect; but on the contrary we have seen that in the Pâli and the Prâkṛits a good many forms are made up according to a certain rule brought into use by analogy, as for instance making the passive by adding इज्ज or इअ, while there are others which are not thus constructed but have descended by simple phonetic corruption of Sanskrit forms, as दीसह, लम्भइ and others. In all parts of grammar there are such forms, and these have descended to the modern Vernaculars, as we have seen.
c12.6
Again, if these had been artificial languages they would not have been called after the names of provinces as we have seen they were. I have given reasons for believing that though the difference between the S'aurasenî and the Mahârâshṭrî was not great, some peculiarities of the former have been preserved by the modern Sindhî Panjâbî, Braja and Hindî, while the Marâṭhî represents the old Mahârâshṭrî, and some of the peculiarities of the Mâgadhî Prâkṛit are distinctly observable, as we have seen, in some of the inscriptions of As'oka, in which it would be absurd to suppose an artificial language was used. And how, if they were not popular dialects, could the idea of using them for women and the inferior characters in dramatic plays have in the first place arisen? That a poet should make certain persons in his work speak their peculiar dialect, especially when that is an inferior dialect and likely to create mirth, is natural, and this device is resorted to by writers in all countries. But it was probably more from considerations of propriety than liveliness that these languages began to be used by Sanskrit dramatists. For they are no less particular about such proprieties, and of even the so-called unities, than other nations. For one of the rules of the Art is that one act should not contain the events of more than a day. Similarly the minute directions about the use of certain Prâkṛit dialects in the case of certain persons are explicable only on the supposition that the original idea was to represent in the drama a state of things actually existing in the world. For these varied reasons it admits of no question whatever that the Prâkṛits of the poets and of the grammarians were really at one time spoken languages. But it is of course not meant that they were spoken just as we find them written, or that they were necessarily spoken at the time when the poets that use them flourished. The Prâkṛits became literary and dead dialects as Sanskrit itself had before them, and, as already remarked, they may be used for literary purposes even now, if one wishes to do so. The distinction between the written Prâkṛits and the corresponding spoken Prâkṛits cannot be greater than the usual distinction between the language of books and that of ordinary life; and probably the Prâkṛits of some of our early plays represent the vernaculars of the time faithfully. I must not omit to mention that Mr. Beames has considerably modified his opinion since he wrote the article from which the above extract is taken, for in the beginning of the Introduction to his Comparative Grammar he speaks of the Prâkṛits as spoken languages, and attempts to account for the use of so many in the dramatic plays instead of setting it down to a mere fancy.
c12.7
We will next proceed to consider the views that have been put forth, opposed to the conclusions we have arrived at, as regards the relation between Sanskrit and the Prâkṛits, and endeavour to determine that relation more definitely than we have done. Professor Weber believes that in the Vedic times there were several dialects, and that by a fusion of these a common language called the Sanskrit was formed, while at the same time the dialects ran on their course and formed the Prâkṛits. He says, "I incline to the opinion of those, who deny that the Sanskrit Bhâshâ, properly so called, was ever the common spoken language of the whole Aryan people, and assign it to the learned alone. Just as our modern High German, arising out of the ancient dialects of the Germans, reduced what was common to all to universal rules and laws, and by the power of analogy obliterated all recollection of varieties; and just as, on the other hand, these dialects while they gradually degenerated often preserved at the same time fuller and more ancient forms; so also the Vedic dialects became partly combined in one stream, in which their individual existence was lost, and so formed the regular Sanskrit Bhâshâ, and partly flowed on individually on their own original (Prâkṛita) irregular force, and continued to be the idioms of different provinces in the corruption of which they participated. The Sanskrit language and the Prâkṛit dialects had, therefore, a common and simultaneous origin; the latter did not spring out of the former." Professor Weber gives a few instances of Prâkṛitised Sanskrit words from the Vedic dialect, as he calls it, to show that the developments of the Prâkṛits began then, but nearly all of these are from such works as the Upanishads, which must be assigned to a very late period when the language was in that confusion which we find in the Buddhistic Gâthâs. Now if the development of Sanskrit and the Prâkṛits was contemporaneous, how comes it that almost all the words and grammatical forms in these latter dialects are manifest corruptions of Sanskrit words and forms? Professor Weber’s theory does not account for this fact, but this defect is supplied by Mr. Beames, who holds nearly the same view as the Professor. He says, "The idea of a common language is the creation of modern times, and the effect of the spread of literature." "The most probable hypothesis is, that the Aryans from the earliest times spoke many dialects, all closely akin, all having the same family likeness and tendencies common to all, perhaps in every case mutually intelligible, but still distinct and co-existent." In a note we are told that this is probable but cannot be proved. The dialect of the Vedic hymns was, he says, only one of these. Then "after a time the Brahmans consciously and intentionally set themselves to the task of constructing a sacred language, by preserving and reducing to rule the grammatical elements of this Vedic tongue." "They seized on the salient features of Âryan speech as contained in all dialects, and moulded them into one harmonious whole." They (the popular dialects) were anterior to Sanskrit, contemporary with it, and they finally survived it. Nevertheless, Sanskrit is older than the dialects. This sounds like a paradox, but it is true in two senses; first, that "as ages rolled on, the vulgar dialects went on developing into new forms, while Sanskrit remained fixed and fossilized for ever." The second is, that though Pânini, who is credited with this feat of constructing the Sanskrit language, lived when the early dialects were much changed, still among the Brâhmans there was a traditional memory of the ancient and then obsolete forms of many words . . . . . In teaching his pupils the true principles of speech, Pânini would naturally use these archaic words in preference to the corruptions current around him, and thus the language which he, to a certain extent, created, was in great part a resuscitation of antiquated terms."
c12.8.1
To be able to estimate the views of these writers at their true worth, it is necessary that we should try to fix the meaning of the word dialect, which is the source of some confusion, and endeavour to form some idea of the origin and growth of dialects generally. If little differences are to be regarded as sufficient to constitute a dialect there are as many dialects as human beings. For just as no two human faces are exactly alike, no two men speak exactly alike. Every one has his peculiarities of pronunciation, and is fond of particular words or turns of expression. The principle of individuality is strong in human nature, but its effects are counteracted by the principle of community, which guides the development of human life; and the necessity of communication compels a man to drop peculiarities in speech and to conform himself to the prevailing model. Thus then, the individual differences come to be minute and insignificant, and hence the language of a community becomes practically one language. But, coustituted as communities generally, but not necessarily, are, they are composed of classes; and there is not as much communication between separate classes as within the limits of a class, and hence there come to be class languages, or varieties of speech in the same community. There may be many such classes;
c12.8.2
but the smaller ones may be neglected, and we may distinguish between two large ones, generally called the lower and the higher, or the uneducated and educated classes. The languages of these differ, that of the lower class being more subject to those laws of development, growth, or corruption which I traced in the first lecture. Thus an uneducated Englishman uses many such forms as "I knowed," "you says," formed upon the prevailing analogies; and his pronunciation of a good many words is corrupt, as "genlman" for "gentleman," "wot" for "what," "guvner" for "governor," &c. And in the same way the language of the Marâthî, Gujarâtî or Hindî lower classes is not the same as that of the higher. But still the word dialect is not used in speaking of these two varieties, one principal reason being that the language of the lower classes is not acknowledged to have an independent existence. Again, when a community comes to have a literature, the style that is used, as a rule, differs from the ordinary conversational style. There are certain words, turns of expression, modes of construction, and even solitary forms which, though used in conversation, do not appear in books. Even these two varieties are not called dialects.
c12.8.3
The dialects of a language therefore or kindred dialects, as they may be called, are the languages of kindred communities, which, while they possess a large body of common words and forms, have also many others which are different, though in most cases the roots from which these are derived are common to all. And according to the proportion in which the similar and dissimilar elements are mixed, the dialects are more or less distant from each other. The causes that divide men into distinct communities are the causes that create dsitinct dialects. The laws of change and development are always in operation in a language, in a changing condition of society, and the processes of destruction or disappearance of old words and forms and the construction or appearance of new ones, are constantly going on. Hence, when after the separation of men into distinct communities, communication between them is lessened, these processes go on in a varied and dissimilar manner. The conditions of life in their new habitations may also be different, and when they are so they bring about a varied linguistic development. But though the processes of destruction and construction are always going on in a language, they suppose that the society, whose language it is, is moving and not stationary. The life of a community may, however, be even and unvaried, in which case there will be little change in its language. Or certain races may be very conservative, and cling to the past with tenacity.
c12.8.4
Thus the Lithuanian peasants of the present day have through a number of ages, though they possess nothing like a literature, preserved a good many old Aryan words and forms in a purer condition than all their European neighbours. So that even after men are divided into distinct communities it is by no means necessary that distinct dialects should grow up. They may or they may not. But the possibility of slight differences, such as those which exist in the speech of the different classes of a community, even when the dialect is the same, is not denied. Now, if for the various reasons above indicated dialects have come into existence, these continue until by the development of civilisation and by political events kindred communities come to form a nation; and then by increased communication and the growth of a common literature these dialects are fused together, and one harmonious language is formed; though, however, in remote quarters they may continue to have an independent existence. The writers whom I have quoted pay exclusive attention to this fact, of which there are several instances in modern European history, the most notable of which is the development of the modern German; and lay it down as a universal proposition which requires no proof, that in an early condition of society there must be a plurality of dialects, and there could be no such thing as a common language. This seems to be the opinion of Professor Max Müller also.
c12.8.5
But if there was not one common original dialect, whence are the common portion of the words and forms of kindred dialects, and the common roots derived? And does not comparative philology itself, in its comparisons and classifications, go upon the assumption that there was such a language? Thus then, since the creation of dialects depends on causes and the causes may not exist, it will not do to assume that there were several dialects in Vedic times without positive proof or even positive indications. Several independent Aryan communities or tribes may not have migrated into India, but only one. Or even if more than one tribe came over they may have been, like the Lithuanian peasants, very conservative as regards their language. The circumstances we know about the language of the period do not necessitate the supposition of a plurality of dialects. Besides, if according to Professor Weber these originally existing dialects went on in their own isolated course as the idioms of different provinces, and became the Pâli and the Prâkṛits, we should expect to find much greater divergences between them, produced in the course of the many centuries that elapsed between the Vedic period and that in which they were as we now find them. But the distinction between the dialects of the inscriptions of As'oka and even between the different Prâkṛits is by no means great, and concerns chiefly the pronunciation of individual letters.
c12.8.6
The plea of linguistic conservatism which I have before used in a certain case, cannot be brought in here, since those early dialects resembling the language of the Vedas, as they must have done if they existed, were already very greatly corrupted in passing into the Pâli or the Prâkṛits. Surely such very extensive changes as from the Vedic to the Prâkṛit form, if carried on in different provinces by different communities, must occasion very great divergences, greater even than we find between the modern vernaculars. For the degree of departure from the old Prâkṛits which these last exhibit is not so great as that which the Prâkṛits do from the Vedic dialect. The divergence ought to be at least as great, but it certainly is not so.
c12.9.1
Now the way in which the supposed contemporaneous development of Sanskrit and the Prâkṛits is accounted for by Mr. Beames is, as we have seen, this:— That the Prâkṛits arose by a natural development, while Sanskrit was created by the Brâhmans, and principally by Pâṇini, out of the Vedic dialect, the salient points of Prâkṛit speech, and by a resuscitation of antiquated terms. Hence it contains the old words and forms from which those in the Prâkṛits are derived. And this also accounts for the existence of what Mr. Beames calls an Aryan but non-Sanskritic element in the Prâkṛits, by which he probably means those few forms that are not found in classical Sanskrit, but must be referred to the Vedic dialect, or others that must be Aryan, though not found either in classical or Vedic Sanskrit. Of the latter there are almost no instances which may be regarded as beyond the possibility of doubt. But such a theory is calculated to bring relief to a puzzled philologist who, not knowing to what Sanskrit word to trace any Prâkṛit or Vernacular word, will at once get out of the difficulty by putting it down as an Aryan word not found in Sanskrit. But how is it possible to create such a language as the Sanskrit out of the elements indicated by Mr. Beames? A literary style, as distinguished from a conversational style, is what one can understand, but a language which never had anything to correspond to it in ordinary vernacular speech, but is simply created, is inconceivable.
c12.9.2
And if the creation of a sacred language was the conscious and deliberate aim of the Brahmans, and Pânini resuscitated old or obsolete words, why did they or he not include in the new language all the innumerable old obsolete Vedic words and grammatical forms, which would have, from the very reverence that was paid to the Vedas, rendered their invention more sacred; and why should they, instead of doing so, have distinguished between a Chhândasa or Vedic, and Bhâshâ or popular, speech? And, if the inventors seized all the salient features of the popular dialects, would they have left a few but striking Vedic peculiarities in the popular speech, such as the absolutive in त्वान् or तून, unrepresented in their new language? And how could they from the popular dialects, in which the conjugations and declensions were reduced to fewer types, construct the innumerable forms of the Sanskrit noun or verb, the Perfects, the Aorists and the lost tenses or moods? And what must have been the basis of the incalculable Taddhita or nominal, and Krit or verbal derivatives, which have left but few traces in the Prâkrits? Even the Vedic dialect is not calculated to afford much help. For if the grammar of the later Sanskrit had been framed by the conscious inventors upon the model of the Vedic, the declensions, conjugations, and the derivatives would have been like the Vedic; whereas, as a matter of fact, the differences are great, and some of these I pointed out in the first lecture.
c12.9.3
There remains the third alternative, that there was a traditional memory of these forms. But why should they have been committed to memory when there was no motive? The sacred formulas which Mr. Beames says were orally transmitted do not, if they were the same as those that have come down to us, contain a large number of such forms; and, if they were different, they are simply imaginary. And, though the Vedas were and are committed to memory, the literally incalculable number of case forms, verbal forms, Taddhitas, and Kridants, in the absence of any general rules which were laid down afterwards, as the theory supposes, by the grammarian creators of the languages, it is thoroughly impossible to commit to memory,—impossible, as the great author of the Mahâbhâshya says, even for Indra, with Brihaspati for his teacher and a thousand celestial years during which to learn them. Of course it is a different thing when they are current in a language which one learns from his childhood, though even in this case very few know all the words and forms which are current in their own country, and nearly all only such as they are immediately concerned with. And what is the evidence for the truth of this theory? Absolutely none is given. It is simply the vague feeling of an individual or individuals, and not a conclusion arrived at after a deliberate weighing of evidence;
c12.9.4
while it sets at naught the clearest evidence available in the works of the grammarians themselves, which I shall presently adduce. The theory, therefore, is utterly untenable, and the contemporaneous development of Sanskrit and the Prâkṛits derived from it, to account for which it was invented; is an impossibility. Professor Aufrecht, though he does not believe in a plurality of Vedic dialects, derives the Prâkṛits from the Vedic language, and assigns Sanskrit to schools of the learned, but does not explain its genesis. Those who believe the Vernacular dialects to be derived directly from the Vedic must explain the origin of classical Sanskrit in some such way as Mr. Beames has done; there is no other way.
c12.10
Dismissing feelings which we have been considering so long, we will for a time examine the reason or reasons that are given for doubting the derivation of the Pâli and the Prâkṛits from Sanskrit. Thus Professor Childers, in the preface to his Pâli Dictionary, states the relations between Sanskrit and Pâli in the following lucid manner:—
c12.11
“If we compare Pâli with classical Sanskrit, we find that about two-fifths of the vocabulary consists of words identical in form with their Sanskrit equivalents, as Nâga, Buddha, Nidâna. Nearly all the remaining words present a more or less late or corrupted form . . . . . Words of the above two classes nearly exhaust the Pâli vocabulary; but there remains a small though important residuum of forms distinctly older than Sanskrit, and found only in the oldest known Sanskrit, that of the Vedas. Nay, I do not feel sure that Pâli does not retain a few precious relics older than the most ancient Sanskrit, and only to be explained through the allied Indo-Germanic languages. It results from all this that Pâli cannot be derived from Sanskrit; both, though most intimately connected, being independent corruptions of the lost Aryan speech which is their common parent; but that Pâli is on the whole in a decidedly later stage than Sanskrit, and, to adopt a metaphor popularised by Max Müller, stands to it in the relation of a younger sister.” Then
c12.12
in a footnote he gives these Vedic forms in the Pâli: the infinitive in tave, as kátave "for doing," the absolutive or gerund in tvâna, as katvâna "having done." These two terminations, however, occur only occasionally, the usual ones are the Sanskrit tum and tvâ. The next two, imassa, genitive singular of ayam, and gonam, genitive plural of go "a cow," I consider as made up by false analogy, as I have already indicated; tinnam, genitive plural of tri, is similar. Then vidû from vid to know and meaning "one who knows;" divo, which is masculine in the Veda and in the Pâli, while it is feminine in Sanskrit; the Imperative first person plural in âmase as in yamâmase, "I shall or may restrain" kasâmase, "I shall or may plough;" the Imperfect akâ of karoti from the Vedic akat. Pâli has the Vedic ḷ, but this is a sound natural under certain circumstances, since we have it in Marâṭhî and Gujarâtî, and kuham corresponding to the Vedic kuha. The Vedic Instrumental plural termination of nouns in 𑆨𑆴 is retained in the Pâli and also corrupted to ehi. Professor Childers, however, thinks it is corrupted from the Sanskrit ablative bhyas. I agree with those who derive it from the Vedic affix. To these may be added the Prâkṛit कुणइ "he does," in which there is a remnant of the fifth conjugation, while the root belongs to the eighth class; and also the affix तण of abstract nouns, while it is त्व in ordinary Sanskrit. Prâkṛit possesses तण or ऊण corresponding to the Pâli त् वान or तून and the instrumental affix एहि.
c12.13
Then in another note Professor Childers gives the Pâli garu "heavy" or "great," while Sanskrit has गुरु, though the गर appears in the Sanskrit comparative and superlative forms गरीयस् and गरिष्ठ. The 𑆓 is seen in the Greek word Barùs and Latin gravis. The Potential अस्य "let him be," is also said not to have anything to correspond to it in Sanskrit; but it may be traced to the Vedic subjunctive or लेट्, असत्, or regarded as a corruption of अस्यात् for Sanskrit स्यात्, the 𑆱 though dropped in Sanskrit being brought over by analogy from the singular of the other tenses. He also gives इध, while Sanskrit has इह, which is doubtful, and also सब्बधि "everywhere" the ध in which corresponds to fi in some Greek words. The Pâli धि is the locative singular of the root, the instrumental singular of which is preserved in the Sanskrit धा in एकदा, पञ्चधा &c. Of these the Prâkṛits have गर in such words as गरुअ, गारव &c., and इध.
c12.14
Here the question is reduced to a point. Two-fifths of the words in the Pâli are pure Sanskrit, the rest are corrupt Sanskrit, while there are about six or seven forms which are lost in classical Sanskrit but exist in Vedic Sanskrit; and there are one or two which exist in neither. The dialect then from which the Pâli was corrupted or immediately sprang, granting for a moment that it was different from the Sanskrit, had two-fifths of its words exactly like those in Sanskrit, and the other three fifths from which the Pâli corruptions were derived were also exactly like Sanskrit, but it contained these eight or nine forms which Sanskrit has lost. In other words, Pâli itself, if we take the original forms of the corrupt three-fifths of its words into consideration, is exactly like Sanskrit but for these nine forms. Is the dialect for this difference to be considered an independent dialect? Is the distinction enough to entitle the Pâli or its supposed immediate parent to be considered a dialect, that is, a language, elaborated, according to the definition given before, from the ancient speech by a community distinct from that which elaborated the Sanskrit? If one thinks it is, he may do so, and regard the Pâli as a distinct dialect; but he must for the same reason consider Surat Gujarâtî and Ahmedabad Gujarâtî; Komkaṇî Marâṭhî and Dekkanî Marâṭhî; or the Marâṭhî, Gujarâtî, English and any other language as spoken by the uneducated classes of the population, and as spoken by the educated, to be distinct dialects of the same ancient speech. In the same community, as I have observed, varieties of speech may or do prevail and must prevail, but the distinction between them is not enough to constitute them distinct dialects; wherefore Sanskrit and the immediate parent of the Pâli or the uncorrupt Pâli must be considered as one language, even though the latter contained a few more ancient forms; and if both were one, the Pâli must be regarded as a daughter of Sanskrit and not its sister. Some of its few peculiarities it may have derived from the spoken variety of Sanskrit, and the rest from that in use among the inferior classes. It often happens that the lower classes sometimes retain an old word or form after it has been given up by those above them, whence it is likely that the Pâli was corrupted from the language of those classes. And as a matter of fact it was for a long time the speech of the uneducated, as will be presently shown. These observations apply to the later Prâkṛits also; wherefore, if these derived languages were spoken, that from which they were derived, *i. e.*, Sanskrit, must have been a spoken language also.
c12.15
And, independently of this consideration, there is positive evidence that Sanskrit was a spoken language. Yâska in the Nirukta frequently refers to the Vedic dialect and to another called Bhâshâ, the peculiarities of which mentioned by him are observable in classical Sanskrit. Pânini in his Grammar gives a good many rules which are exclusively applicable to the dialect of the Vedas, to which he refers by using the words Chhandas, Nigama, Mantra, and Brâhmaṇa, and others which are applicable to the Bhâshâ alone, but by far the largest number of his sûtras have reference to both. Now, since Bhâshâ, or the ordinary Sanskrit, is thus distinguished from the dialect of the Vedas, it must be the language in use at the time when these writers lived. "Bhâshâ," as used by them, is a proper name, but in later Sanskrit it acquired a generic signification and meant language generally. The root from which the word is derived, signifies "to speak," wherefore the original sense of the word as a proper noun must have been the "speech" or "the spoken language." And, because this was its signification, it afterwards came to denote "explanation." When we ask for an explanation of something that is obscure and unintelligible, what we mean is that the sense should be expressed to us in the ordinary language of men, a language that we can understand. Thus such a sentence as स्थितप्रज्ञस्य का भाषा, means "what is the Vernacular of स्थितप्रज्ञ," an expression similar to "what is the English of it?"
c12.16
Pânini refers certain points expressly to popular usage. He says that the names of countries are conventional, and no grammatical analysis should be given of them, because it is fictitious; these should be used as we find them used. Similarly he says grammarians should not make rules to teach such things as these:— That the two words of a compound express the thing denoted by the principal word as qualified by the sense of the subordinate word; as for instance, राजपुरुष, a compound of राजन् "a king" and पुरुष "a man", does not denote "a king," but "a man," and not "man" alone but as connected with a king, i. e. a king's man or officer; and that the base and the termination express the sense of the termination as qualified by that of the base; as औपगव signifies not उपगु but a child, and not a child alone but a child as connected with उपगु i. e. Upagu's child. For the significations of words are to be learnt from usage.
c12.17
In the introduction to the Mahâbhâshya Patañjali tells us that some persons in his or Kâtyâyana's time considered the study of grammar to be unnecessary. For said they, "Vedic words or forms we know from the Veda, and those current in popular usage from that usage; grammar is useless." Now the grammar which is thus declared useless is the grammar both of the Vedic and classical Sanskrit; and the depreciators of the science profess to derive a knowledge of the first dialect from the Vedic books, and of the second not from other books but from popular usage. Hence Sanskrit must have been in the times of those two grammarians a spoken language.
c12.18
Similarly in the passage from the same work which I placed before you in the first lecture, you will remember that the objector (or *pûrvapakshin*) argues that, since usage is the authority upon which the grammarians go, certain verbal forms which are no longer used by people ought not to be taught by the grammarians, and says that instead of those verbal forms participles are employed. The principal teacher (*Siddhântin*) does not deny the facts, but refers the objector to the vast literature of the language, where he may find them used, though obsolete at the time. It is evident from the whole passage, that Sanskrit was then a spoken language though some of its verbal forms had fallen into desuetude. I have also shown that the language was considerably changed between the times of Pânini and Kâtyâyana, and called the Sanskrit that prevailed when Pânini and Yâska flourished Middle Sanskrit, and that which was current in the time of Kâtyâyana Classical Sanskrit. Now these changes from the one form to the other could not have taken place if the language had been dead or petrified into a merely literary language.
c12.19.1
I am at a loss to see why some scholars should find it so difficult to believe that Sanskrit was a vernacular. If its declensions and conjugations are considered too complicated for the language of everyday life, it must not be forgotten that such a fact did not prevent the ancient languages of Europe from becoming spoken languages. And this objection would do equally well against the Vedic dialect, which, or others like which, are regarded as the vernaculars of their times, and which are richer in inflexions than the later Sanskrit. Then it is held that the artificial regularity of Sanskrit makes it improbable that it should have been a vernacular. Where is this artificial regularity? On the contrary, it is the absence of regularity that renders its grammar so difficult and complicated. There is a freedom in the choice of words, expressions, and forms. In every department of its grammar there are innumerable optional forms; nouns and verbal roots are often declined and conjugated in several ways. One same root in a good many cases forms its special tenses in more ways than one, and in the nominal derivatives, the verbal derivatives, the formation of the feminine, and the uses of cases and tenses there is a freedom which some may consider a license.
c12.19.2
The only difference is, that Sanskrit has had the most perfect grammarians in the world, who observed all the facts of their language and laid them down as unchangeable facts, and it is this which gives that language a stiffened appearance. Then the Saṃdhis or euphonic combinations of letters, which are necessary in Sanskrit, are regarded as inconsistent with the character of a spoken language. It is, however, not denied that such combinations are observable in all languages, and paticularly so in Latin and Greek; but it is urged that in Sanskrit there is a regularity or universality about them which is not found anywhere else. It should not, however, be forgotten that Saṃdhi in the same word and the same compound, and of a preposition with a root is alone necessary. Between different words it is optional, which means that it was on occasions neglected. Now Saṃdhi in the same word is necessary by a law of nature. The Sanskrit does not allow a hiatus; and this is a characteristic of most of the modern idioms also. Some languages, such as our Prâkṛits and the old languages of Europe, tolerate it. But the euphonic combination of consonants in the same word is necessary even in Latin; as in rex = reg-s, scriptus = scrib-tus, cinctum=cing-tum, lectum = leg-tum, tractum from traho, though the h like the Sanskrit ह stands for an original gh; d and t combine to form an s, as defendo, defensum; sedeo, sessum; claudo, clausum; &c.
c12.19.3
Prepositions are really parts of words, and hence by the same law they also must form one harmonious sound with the initial letter of the word to which they are attached; and modern vernaculars have got corruptions of the combined words, which shows that they must have been used in those forms in the colloquial Sanskrit. Sk. पर्यस्त, Pr. पल्लट्ट, M. पालट; Sk. प्रत्यभिजाना, Pr. पच्चहिआण, H. पहिचान; Sk. अत्यर्थ, Pr. अच्चह, M. अचाट, &c. Now, as to compounds, this peculiarity of the Sanskrit has been carried to an extravagant extent by later writers, but Pâṇini allows only certain formations of this nature. These grew up as independent words in the language, and hence in the matter of Saṃdhi were treated like other words. In the spoken language the euphonic combinations we have been considering were not consciously made, but the words themselves acquired those forms by habitual use in the same manner as in other tongues. The grammarians, however, discovered and laid down these rules; and the practice of using them in books even in combining different words gained ground, though, however, many instances in which there is no such Saṃdhi are found in the Itihāsas and the Purāṇas. But if in colloquial speech such a combination was not possible, the grammarians do not enjoin that it should be made; and very probably it was not made.
c12.20
And traces of many expressions, which only a colloquial use of language can generate, have been preserved, not so much in the literature, as by the grammarians. Such is one expressive of an intensive or excessive action, composed of the Imperative second person form of a root repeated, followed by a verbal form of the same in any tense of the Indicative and in any person or number; as खाद खादेति खादति, lit. "eat, eat, he eats," *i. e.* eats much, कुरु कुर्वीति करोति "do, do, he does," *i. e.* does much. This expression exists in Marāṭhî and is considered so colloquial that no Marāṭhî grammarian has noticed it, as खा खा खातो, कर कर करतो, in which, as in Sanskrit, खा and कर are forms of the Imperative second person singular. A similar expression is used when several actions are attributed to the same agent; as ओदनं भुङ्क्ष्व सकृत् पिब धानाः खादेत्येवायमभ्यवहरति "eat rice, drink barley water, devour fried grain, in this way he fills his stomach"; which in Marāṭhî is भात जेव पेज पी लाखा खा असें पोट भरतो. In this case the Indicative should signify a general action of which the Imperatives denote the species, and we may have here करोति "does" instead of अभ्यवहरति. Similarly those innumerable expressions consisting of a form made up by adding अम्, technically णमुल्, to a root, preceded by a noun, and of a verbal form of the same root must be colloquial; as हस्तग्राहं गृह्णाति "holds by the hand," जीवनाशं नश्यति "he perishes," उदरपूरं भुङ्क्ते "eats a stomachful," यथाकारमहं भोक्ष्ये तथाकारं भोक्ष्ये किं तवानेन "I will eat as I eat (as I like), what have you to do with it?" &c. Etymologically हस्तग्राहम्, जीवनाशम् &c. are accusatives, and they may in these cases be called cognate accusatives, and the expressions somewhat resemble such ones in English, as "run a race," "walk a walk," "die a death," &c. The compounds दण्डादण्डि, केशाकेशि &c. meaning "a scuffle in which there is a brandishing of sticks and seizing each other by the hair;" and a great many others made up of Imperative forms of verbs, or of a verb and its object, which are used as nouns, are of a nature suited for the purposes of a light conversation; as अत्र खादतमोदता वर्तते "'eat and enjoy' is the rule here," अत्राभीतपिबता वर्तते "'eat and drink' is the rule here," उद्धरोत्सृजा तस्य दानशौण्डीरस्य गृहे "'take out and give' is what takes place in the house of that bountiful man," जहिस्तम्ब्रोयम् "he is one who constantly says, ‘strike the sheafs of corn’ ”, एहिस्वागता वर्तते “ ‘come, welcome to thee’ is the practice”, &c.
c12.21
Sanskrit was not the only language spoken in the times of Kâtyâyana and Patañjali. In the Mahâbhâshya there are several passages which contain allusions to a dialect arising from a corruption of Sanskrit. Thus, in the comment on the Vârtika *Siddhe S’abdârthasambandhe*, we are told, with reference to the question, whether words are created or exist of themselves, that Pâṇini’s rules suppose that they are not created but exist of themselves; and so is the relation between them and the things they denote, *i. e.* their power of expression, uncreated and self-existent; and, according to another interpretation of the Vârtika, the things also which words denote are so. How is it known that all these are self-existent? Because in the affairs of the world or in ordinary life men think of certain things and use words to express them; they do not set about producing words before doing so. But this is not the case with those things that are created and not self-existent. If a man wishes to do something with an earthen pot, he goes to the house of a potter and says, “Make a pot for me, I have to do something by its means.” Now, if he has to use words, he does not in the same way go to the house of a grammarian and say, “Make words, I want to use them”; but immediately that he thinks of this thing and that he uses words [for expressing them]. Well then, if words are to be taken from ordinary life [and are not made by grammarians], what is it that the Sâstra [grammar] does? “The Sâstra lays down a restriction by observing which a man may attain religious merit. It does what other Sâstras in ordinary life do. Thus, while it is possible to satisfy hunger by eating anything whatever, it is enjoined that one shall not eat a domesticated fowl or pig; and the object is, that he may by regulating his conduct thus attain religious merit. In the same way this Sâstra (grammar) tells us that, while it is possible to express one’s meaning by using correct words or incorrect words, correct ones alone which it teaches should be used to secure the religious merit arising therefrom.” After this follows the discussion translated in the first lecture; and then we have another of which the following is a portion:—
c12.22
PÛRV. Does religious merit arise from a knowledge of correct words or from their use?
c12.23
SID. What is the difference?
c12.24
PÛRV. If you say religious merit arises from their knowledge, religious demerit also must arise. For he, who knows what words are correct, also knows what words are incorrect. If merit results from the knowledge of those that are correct, demerit must result from the knowledge of those that are incorrect. Or greater demerit must arise [from their knowledge], as the number of incorrect words is larger, and that of correct words smaller. For the corruption of one correct word are manifold; as, for instance, the corruptions of the correct word गो are गावी, गोणी, गोता, गोपोततिका &c. And the Rishi also indicates (in a passage which is quoted) that the restriction as to correct words concerns their use [and not knowledge].
c12.25
SID. Well, then, let it be that religious merit arises from their use [and not from knowledge].
c12.26
PÛRV. If from their use the whole world would obtain heavenly felicity.
c12.27
SID. And now why should you be jealous if the whole world obtain heavenly felicity?
c12.28
PÛRV. No, certainly, I am not jealous. But what I mean is, that it thus becomes useless to make any effort; such effort only as is fruitful should be made. If you get the fruit without any effort, the effort is useless. [The effort meant is that involved in the study of grammar, i. e. of correct words. People use some correct words at least without studying grammar, and, if eternal felicity results from the use of such words, they get it without making the effort of studying the subject.]
c12.29
SID. Why, verily those who make the effort will largely use correct words, and will obtain a large amount of heavenly felicity.
c12.30
PÛRV. That the fruit does not follow the effort is also an observed fact. For there are persons who though they have made the effort are seen to be incompetent [in the use of correct words], while others, who have not, are competent; wherefore it is possible, the fruit, i. e. heavenly felicity, may not follow.
c12.31
SID. Well then, religious merit arises not from knowledge alone, neither from use alone.
c12.32
PÛRV. From what then?
c12.33
SID. Heavenly felicity arises from the use of correct words when it is accompanied by the knowledge, that they are correct, derived from a study of Śâstra.
c12.34
And thus it goes on.
c12.35
Now it is clear from all this that correct words, *i. e.* Sanskrit, was spoken in those days by all, but that incorrect words had got mixed up with it, and the object of grammar was to teach how to avoid incorrect words or corruptions, though there were men who could speak correctly without knowing grammar. And this is the state in which more or less all languages are at all times; and even at the present day the purpose of grammar is considered to be to teach how to speak correctly. By the way, it will be seen how Sanskrit grammarians distinctly declare that they teach nothing that does not exist, they do not create words, but separate the correct ones from such as are incorrect. But what did Patañjali consider to be the standard of correct Sanskrit, who was it that spoke the language correctly, and in whose speech were corruptions observable? This is clear from another passage at the beginning of the third pâda of the first chapter of the Mahâbhâshya. Patañjali interprets the sûtra भूवादयो धातव: (I, iii, 1.) in a manner to yield such a connotative definition of a dhâtu or root as this, that a dhâtu is that which denotes action or being. Then a question is raised, if this is the way to distinguish a root why should a list of roots be given; in other words, if we have got a connotative definition, a definition by enumeration is not wanted. In this case there is a difference of opinion between Kâtyâyana and Patañjali.
First 44 of 93 passages. Read on
Cites (2)
- Aṣṭādhyāyī — Pāṇini, Sanskrit
- Aśoka