Pramāṇa · Search · About · API

Wilson Philological Lectures on Sanskrit and the Derived Languages c7: Lecture II. Pâli and the Dialects of the Period

Wilson Philological Lectures on Sanskrit and the Derived Languages c7

Language
English

Contents

Lecture II. Pâli and the Dialects of the Period

c7.2

In the last lecture we traced the development of the Sanskrit language until it assumed the form in which it is usually called classical Sanskrit. This form it acquired several centuries before Christ, as I shall hereafter endeavour to show, but it has since preserved it unaltered. The reason is that the number of people using it has been growing smaller and smaller, and now for a considerable period it has remained confined to a few learned men who devote long time to its study and use it only for religious and philosophical purposes. But in so far as it has been used for these purposes it has not been free from change, as was pointed out on the last occasion. The style and modes of expression used by Gadâdhara Bhaṭṭâchârya, the great Naiyâyika of modern times, are not the same as those of Kaṇâda, Gotama, and Vâtsyâyana, nor those used by Nâgojibhaṭṭa, the same as those of Kâtyâyana and Patañjali. Bnt all these changes have followed one direction. The efforts of our modern writers have been mostly, if not altogether, directed to systematizing and reducing to an accurate and definite shape the conceptions of those ancient sages, and no new ideas or methods of thought have arisen. They have thus elaborated a highly artificial style, abounding in abstract terms. In other branches of the literature also, we may discover some peculiarities and corruptions of diction and manner. But still, making due allowance for all this the generally received standard of good and correct Sanskrit is the same as it was many centuries ago. There are certain models which modern writers and speakers have to conform to. When, however, a language is living, that is, is used by people generally and in matters of every-day life, it must undergo very great changes. Men will not waste time and trouble in accurately ascertaining and imitating the pronunciation of their fathers, or reproducing exactly the several grammatical forms used by their predecessors. All that is necessary is that they should make themselves understood. Old modes of thinking give place to new ones, and new ideas spring up; and both must be represented in the language that is used. Sometimes accidents in the history of a race, such as its coming in contact or being incorporated with another race, serve to transform its language within a short time. But the alterations that occur from these and other causes obey certain laws; and hence the possibility of a Science of Language. Now, by the operation of influences such as these, Sanskrit was, in very remote times, corrupted, and new dialects arose from it. We should have known nothing of these if they had not been preserved in their literatures. Of these dialects that which approaches Sanskrit the most is the Pâli. This is the sacred language of the Buddhists of Ceylon, Siam, and Burmah, and possesses an extensive literature. I will introduce my remarks on this dialect by placing before you a specimen:—

c7.3

सावत्थियं किर अदिण्णपुब्बको नाम ब्राह्मणो अहोसि । तेन कस्सचि किंचिद्-दिण्णपुब्बं तेन तं अदिण्णपुब्बको त्वेव सं जानिसु । तस्सेकपुत्तको अहोसि पियो मनापो । .........तस्स सोलसवस्सकाले पण्डुरोगो उदपादि । माता पुत्तं ओलोकेत्वा ब्राह्मण पुत्तस्स ते रोगो उपपन्नो तिकिच्छापेहि नन्ति आह । भोति सचे वेज्जं आनेस्सामि भत्तवेतनं दातब्बं भविस्सति । त्वं मम धनच्छेदर्नं न ओलोकेसीति । अथ किं करिस्ससि ब्राह्मणा ति । यथा मे धनच्छेदो न होति तथा करिस्सामीति । सो वेज्जानं सन्तिक्ं गन्त्वा असूकरोगस्स नाम तुम्हे किं भेसज्जं करोथाति पुच्छि । अथस्स ते यं वा तं वा रुक्खतव्वादिंगं आचिक्खन्ति । सो तं आहरित्वा पुत्तस्स भेसज्जं करोति । तं करोनत्तसेवस्स रोगो बलवा अहोसि ।

c7.4

This is from the Aṭṭhakathâ or commentary by Buddhaghosha on the Dhammapada. It was written in the fifth century after Christ, long after Pâli had ceased to be a vernacular. The following is from the Dhammapada itself, which we have reason to believe is much more ancient:—

c7.5

129. सब्बे तसन्ति दण्डस्स सब्बे भायन्ति मच्चुनो ।
अत्तानं उपमं कत्वा न हनेय्या न घातये ॥
131. सुखकामानि भूतानि यो दण्डेन विहिंसति ।
अत्तनो सुखमेसानो पेच्च सो न लभते सुखं ॥
133. मा वोच फरुसं कञ्चि वुत्त पटिवदेय्यु तं ।
दुक्खा हि सारम्भकथा पटिदण्डा फुसेय्यु तं ॥
134. सचे नेरेसि अत्तानं कंसो उपहतो यथा ।
एस पत्तोसि निब्बानं सारम्भो ते न विज्जति ॥

c7.6

The Sanskrit of these passages is as follows:—

c7.7

श्रावस्त्यां किलादत्तपूर्वको नाम ब्राह्मणोभूत । तेन कस्मैचिन्किंचिददत्तपूर्वं तेन तमदत्तपूर्वस्त्वेव समज्झासत । तस्यैकपुत्रकोभूतप्रियो मनोज्ञः । तस्य षोडशवर्ष- काले पाण्डुरोग उद्पादि । माता पुत्रमवलोक्य ब्राह्मण पुत्रस्य ते रोग उत्पन्नश्चिकित्स्यैनमित्याह । भवति स चेद्वैद्यमानेष्यामि भक्तवेतनं दातव्यं भविष्यति । त्वं मम धनच्छेदं नावलोकयसि । अथ किं करिष्यसि ब्राह्मणोति । यथा मे धनच्छेदो न भवति तथा करिष्यामीति । स वैद्यानामन्तिकं गत्वामुकरोगस्य नाम यूयं किं भैषज्यं कुरुथेत्यप्राक्षीत् । अथास्य ते यद्वा तद्वा वृक्षत्वगाद्याचक्षते । स तदाहृत्य पुत्रस्य भैषज्यं करोति । तत्कुर्वत एवास्य रोगो बलवानभूत ।

c7.8

129. सर्वे त्रसन्ति दण्डस्य सर्वे बिभ्यति मृत्योः । आत्मानमुपमां कृत्वा न हन्यान्न घातयेत् ॥

c7.9

131. सुखकामानि भूतानि यो दण्डेन विहिनस्ति । आत्मनः सुखमिच्छन्प्रेत्य स न लभते सुखम् ॥

c7.10

133. मा वोचः परुषं कंचिदुक्ताः प्रतिवदेयुस्त्वाम् । दुःखा हि संरम्भकथा प्रतिदण्डाः स्पृशेयुस्त्वाम् ॥

c7.11

134. स चेन्निरस्यत्मानं कांस्यमुपहतं यथा । एष प्राप्तोसि निर्वाणं संरम्भस्ते न विद्यते ॥

c7.12

“In Śrâvastî there was a Brahman of the name of Adattapûrvaka (‘who had not given away anything before’). He never gave anything to anybody, whence they called him Adattapûrvaka. A son was born to him, pleasing and delightful. When he was sixteen years old, he got jaundice. Seeing the son thus, the mother said, “O Brahman, your son has got a disease, put him under medical treatment.” “Lady, if I get a physician, food and salary will have to be given to him. You do not see that this will entail expense” (lit. lessen my money). “What then will you do, Brahman?” “I will do what will not involve expense” (lessen my money). He then went to physicians and asked, “What medicine do you use against such and such a disease?” They spoke to him at random of the bark of some tree. He brought it and gave it as a medicine to his son. As he went on doing so, the disease was aggravated.”

c7.13

129. “All are afraid of being hurt, all fear death. Taking one’s own self as a model, one should not kill and cause to be killed.”

c7.14

131. “He, who looking to his own happiness, inflicts injury upon creatures that equally desire their happiness, does not obtain felicity after death.”

c7.15

133. “Do not speak harshly to anybody; they may speak to you in return. For angry altercation is painful, and revenge will overtake you.”

c7.16

134. "If then, like a bottom of a shattered Matallic vessel, you will not stir yourself (utter), then you have reached Nirvâna, and you will have no angry altercation."

c7.17.1

Here you will see a great many instances of the laws we laid down in the last lecture. And first, as regards phonetic corruption, we have in these passages त for त्र as in पुत्त for पुत्र, for क्त as in भक्त for भक्त, and for प्न as in पत्न for प्राप्त, and प्प for त्प as in उप्पन्न for उत्पन्न. This change is usually called assimilation of consonants. In pronouncing a conjunct, one has to pass from one position of the vocal organs to another without letting off the breath checked or compressed in the first, an operation of great difficulty. The sound of the first member is indistinct, but when the second is a weak consonant such as a nasal or a semi-vowel, it glides into the first almost as a vowel does, and thus the sound of the first acquires greater prominence. In the first position, the vocal organs strike and press against each other strongly, and the momentum necessary for this effort is acquired by uttering forcibly and with a jerk the vowel that precedes the conjunct, *i. e.* अ for instance in the word भक्त, just as a man previous to striking and pressing anything with his fist moves the hand with force through some distance. The previous forcible vowel breath and the subsequent strong contact and pressure form the characteristic of a conjunct consonantal sound. This is the only way of rendering the two members at all distinct, for the first is thus uttered with the preceding vowel and the second with the following.

c7.17.2

Thus, भक्त if pronounced in the manner I have described becomes भक्-त, but, if the previous vowel sound is weak, it becomes भ-क्त, in which case, in consequence of the absence of momentum, the pressure cannot be strong, and the conjunct character is not fully brought out. What we find in the Pâli, therefore, is that the passage from one vocal position to another is avoided, and the more distinct sound only, whether of the first member or the second, is pronounced with this forcible previous vowel breath and the succeeding strong contact and pressure; that is we have a conjunct as before, but it is made up not of two different letters, but of two of the same kind. Hence the instances noted above and also धम्म for धर्म, जम्म for जन्म, यञ्ञ for यज्ञ, सद्द for शब्द, दुद्ध for दुग्ध, खग्ग for खड्ग, सुग्ग for सुहृ, अब्भुत for अद्भुत, लद्ध for लब्ध, उक्का for उल्का, अप्प for अल्प, तप्पण for तर्पण, अय्य for आर्य &c. But an attempt is made to pronounce the other sound also, and such of its elements as can go in with that which is uttered without involving change of position are transferred to it, that is, in effect the Pâli speaker treated a conjunct consonant as one sound possessing the characteristics of both. Thus the heavy *s'vâsa* or simple breath of स्, श, and ष in the conjuncts स्क, ष्क्, श्र, श्च, स्त, स्प and ष्प is combined with the distinct sounds क्, च्, ट्, त्, and प् and we have ख, छ, ठ, थ, and फ for those conjuncts;

c7.17.3

but when preceded by a vowel and pronounced with the usual conjunctal characteristic, they become क्ख, च्छ, ट्ठ, त्थ, and प्फ; as in सावत्थियं for श्रावस्याम् and फुसेय्यु for स्पृशेयुः in the above, and in खन्थ for स्कन्ध, पोक्खर for पुष्कर, पच्छा for पश्चात्, अच्छरिय for आश्चर्य, डिट्ठ for दिष्ट, पुट्ठ for पुष्ट, मत्थक for मस्तक, वत्थु for वस्तु, फन्दन for स्पन्दन, and पुप्फ for पुष्प. In न, भ, ण्, स्म, झ्म, and ष्म the consonantal portion of the distinct sounds, being made up of *nâda* or intonated breath, the heavy *s'vâsa* or simple breath of the sibilant is first made heavy *nâda*, that is, the sibilant is changed to the aspirate ह and we have न्ह, ण्ह or म्ह for these conjuncts; as in पन्ह for प्रश्न, उन्ह for उष्ण, अम्ह for अस्मि, अम्ह for अश्म, and गिम्ह for ग्रीष्म. These correspond to the ख, छ &c., in the above instances, and must like them be considered as simple sounds, and in our modern languages they are pronounced as such in certain places; but when preceded by a vowel they must have, as in other cases, the characteristic of conjunctal sounds, and be pronounced as न्ह, ण्ह or म्ह though they are not so written.

c7.18

I have already observed that when a following semi-vowel or nasal glides into the sound of the first consonant, this latter acquires prominence. But there are cases in which the semi-vowel य does not so merge into the preceding; and this takes place when it is preceded by a dental mute. The palatal य and the labial व have since remote times been often pronounced in two ways, one in which the organs approach each other so closely that it is difficult to distinguish them from ज् and व्; and another in which they are kept more distant. Hence the confusion between य and ज् and व and ब that we find so often in modern pronunciation and in the modern languages. The Bengali invariably makes ब of the Sanskrit व and in most cases ज् of य and the people of Northern India follow him to a great extent. In the Hindí we have many such forms as जमुना for यमुना, बांस for वंश, बांझ for वन्ध्या.

c7.19

The heavy or close pronunciation of य, is favoured by a preceding dental. In pronouncing letters of this class the tongue is nearly horizontal. Consequently, when after the formation of the dental mute it is moved upwards to form य, the force with which it separates from the teeth carries it nearer to the palate than it should be. In the cerebral position, in which the tongue forms a curve with the concave side inwards, the force with which its tip separates acts downwards; hence if a ण् has to be pronounced afterwards, the effort is in no way aided, but, if possible, hindered. In this case, therefore, the ण् is always light; and for a similar reason it is so when preceded by a guttural. It is also light when preceded by the dental न, since a large portion of the obstructed breath passes through the nose, and the tongue is not separated from the teeth with any force. Thus in the word वैय in the passage I have placed before you, and in अय, पय, &c., the य being heavy does not glide or merge into द्, and, being the latter of two successive consonantal sounds, acquires greater distinctness; and the preceding sound must in virtue of the laws we have been examining communicate all its elements to it, except of course the organic position. The element that is communicated in this case is the complete contact, wherefore the heavy य becomes a complete ज्, and this, when pronounced like a conjunct, becomes ज्ज्. In such words as मध्य, वध्य, उपाध्याय &c., the preceding has got its heaviness of nāda or tone to communicate besides the complete contact, and thus the ज् becomes श् and thence ज्श; and so we have मज्झ, वज्झ, उपज्झाय &c. In मृत्यु, प्रेत्य, सत्य &c., to the ज् formed as above must be added the hardness or s'vāsa of त्, i. e. the letter ज् must be pronounced not with the glottis contracted but stretched, wherefore we have त् and, with the conjunctal characteristic, च्. Those words therefore become मच्चु, पेच, and सच्च. If य precedes instead of त्, the heavy s'vāsa required for its pronunciation is transferred to ज्, and so it becomes छ्, and thence च्छ; as in रच्छा for रथ्या, नेपच्छ for नेपथ्य &c. The semi-vowel व् is also heavily pronounced when it forms a conjunct with another semi-vowel, and thus we have पुब्ब for पूर्व, सब्ब for सर्व, &c. There are also instances of this pronunciation when व् is preceded by a dental mute as in उब्बेग for उद्वेग, सब्भि for सद्भिः &c. In ऊर्ध्व the preceding ध्, as in मध्य, transfers its heavy nāda or tone to व्, and makes it म्, and so we have उम्भ; but ऊर्ध्व is changed to उद्ध also.

c7.20

The conjunct क्ष or क्ष् must in some cases have been pronounced in ancient times like क्श, as it is invariably so pronounced by the lower classes of the Marâthî people at the present day. This latter sound is easier to be made than क्ष्, first because the passage from the guttural position of क् to the palatal which is nearest to it is more natural than that to the more distant cerebral position of ष्, and also because in the latter case, after the guttural contact the tongue has to be rounded and the tip brought near the cerebral position. Thus, the whole weight of that organ has to be supported; while in the case of क् when its middle approaches the palatal position in pronouncing श, the forepart falls into a natural position, and no effort is required to hold it there. Now in this क्, the sound of श, being the latter of the two, is by the general rule more distinct than the other; and the complete contact of क् being transferred to it, it becomes च्, because sibilants are produced by heavy s'vāsa as the hard aspirates are, and thence च्छ; as in तच्छक for तक्षक, छण for क्षण in the sense of ‘a festival’, छद्र for क्षुद्र, &c. In such words as वत्स and अप्सरस्, the prevailing sound is that of स which, with the complete contact of the preceding त and प्, ought to become थ्, as श becomes छ in the above case, on the theory that स is a dental. But it is not properly a dental, since it is formed not at the root of the upper teeth as त, थ, द, &c. are, but considerably above, and not by the tip of the tongue but by a part of it further inwards, which is made into a curve, having its convex side upwards, and brought near to that position. In this last respect it resembles the palatals, though the part of the tongue employed in their case is still further inwards, *i. e.* the middle. The sibilant स therefore, has no mute corresponding to it in the sounds of the Sanskrit and Pâli languages. The Marâthî dento-palatals च्, छ्, ज्, झ् answer to it completely, being formed in the same position as itself; so that स, when the organic contact is complete, should become the Marâthî छ्. But these sounds are unknown to the Pâli; the speakers of that language could not pronounce the Marâthî dento-palatals, as the Gujarâtîs and other northern nations cannot at the present day and just as these turn them into pure palatals, so did their ancestors. Hence, that sound which should properly be the dento-palatal छ् become the palatal छ, and we have वच्छ for वत्स, अच्छरा for अप्सरस् and मच्छर for मत्सर.

c7.21

The dentals following a र in a Sanskrit word are in Pâli often changed to cerebrals; as in वड्ढति for वर्धते, वट्टति for वर्तते, छड्डेति for छर्दयति अट्ठ for अर्थ &c. Here, as in the cases we have examined, the distinct sound is by the general rule the second, as we may observe even from the optional form वत्तति for वर्तते, and अत्य for अर्थ. But, as before remarked, in all these Pâli transformations of Sanskrit conjuncts we see an attempt to pronounce both the members. When the speaker, being about to put the vocal organs into the position necessary for the utterance of the first letter, which he has first heard though indistinctly, sees that the second which he has distinctly heard cannot be produced there, he gives it up, and then passes immediately to the next position. But if, at the first, he thinks he can produce something like the second sound, which he has distinctly heard, he pronounces the letter there, and does not give up that position. Thus in pronouncing तर्प्, he feels that nothing like the sound प् can be produced at the position of र्, and hence gives it up and passing on to the lips forms तप्प्; but in वर्त् the द् that he pronounces at the cerebral point sounds a great deal more like ट than प् and therefore he does not pass on to the dental position. But such of the Pâli speakers as could distinguish between त and द did not stop at the cerebral position to pronounce their conjunct, but passed on to the dental; hence the optional forms we have noticed. Such words as मित्र, सत्र, भद्र, &c., differ from वर्ध, वर्त् &c., in having the र् after the distinct sound, so that the speaker resorts first to the dental position, and uttering the distinct sound there, is not carried away by the following र् to the cerebral position, the sound of र् being weaker. But if in the same word a dental follows such a conjunct, that dental is changed to the corresponding cerebral; as in पटि for प्रति and पठम for प्रथम. The cerebral element of these words has made a distinct impression on the speaker's ear, and he is conscious that his transformation of प्र into प्, which was necessitated by his inherent inability to pronounce the two consonants together, has not brought it out. He therefore realizes it by changing the following ति and थ to टि and ठ. But if a consonant of another species follows, or if the conjunct ends the word, he cannot give effect to this impression.

c7.22

The conjuncts in which the second sound is weak and consequently gives way to the first are क्य्, र्य्, ग्य्, प्य्, भ्य्, र्य्, स्य्, क्र्, ग्र्, त्र्, द्र्, ध्र्, प्र्, भ्र्, श्र्, स्, क्, क्, ज्व्, त्व्, ह्, ध्व्, श्व्, &c., which have a semi-vowel for their latter member, and म्, त्म्, म्, म्, &c., which have a nasal. Thus, we have सक् for शक्य or शाक्य, अक्खान for आख्यान, योग for योग्य, कुप्प for कुप्य, अब्भन्तर for अभ्यन्तर, पस्सति for पश्यति, तस्स for तस्य, सक् for शक्र, अग्ग for अग्र, मित्त for मित्र, अद्क for आर्द्रक, गिद्ध for गृध्र, विप्प for विप्र, अन्भ for अभ्र, अस्सम for आश्रम, अस्सव for आसव, सुक्क for शुक्ल, पक्क for पक्व, पज्जळति for प्रज्वलति, नीलत्त for नीलत्व, विद्देस for विद्वेष, अद्धा for अध्वा, अस्स for अश्व, &c., and अग्ग for अग्नि, अत्ता for आत्मा, छद्द for छद्म, सोप्प for स्वप्न, &c. The semi-vowel य after न in going out turns the न into the palatal ञ, this latter representing the combined effect of both; as in अन्ञ for अन्य, राजनञ for राजन्य, &c.

c7.23

We have noticed above the change of क्ष to च्छ, but more frequently it is transformed into क्व्. This change is due to the latter व् being pronounced not incorrectly like ष् but correctly as a cerebral sound, which being hollow and undistinguishable is lost in the sound of क्. This last is distinctly heard, and the effect of ष् is only to add heavy *s'vâsa* to it and render it ख्; and thus the whole becomes क्ख् as in रुक्ख for वृक्ष, लक्खण for लक्षण, पक्ख for पक्ष, &c. Upon the principles we have laid down, it does not matter whether the Sanskrit conjunct is composed of two or three consonants, since what is done in the Pâli is to reproduce only the distinct consonantal sound with the other attendant characteristics; and thus we have ण््ह for क्ष्ण or ष्ण as in सण््ह for श्लक्ष्ण and पण््ह for पाष्णि, and म्ह for क्ष्म as in पम्ह for पक्ष्म in which words the last is the distinct sound, and क्ख for क्ष्म as in लक्खी for लक्ष्मी where the म् must have been pronounced so as to merge into the preceding क्ष्.

c7.24

A Sanskrit conjunct at the beginning of a word is reduced to a single letter; and the reason is obvious. For as I have already observed the distinctive characteristic of a conjunct consists in the previous vowel being pronounced with rapidity and force, and in the subsequent strong pressure of the vocal organs which is rendered possible by the momentum acquired by that forcible utterance. It is the reproduction of this characteristic, that makes a single consonant that is pronounced at any one of the vocal positions look like a double. When a previous vowel does not exist, that is, when a conjunct begins a word, this characteristic cannot be reproduced; hence there is no double consonant. Thus, we have खलित for स्खलित, कमेण for क्रमेण, भ्रमर for भ्रमर, फस्स for स्पर्श, &c.

c7.25

You have seen that in all the changes of conjunct consonants which we have examined, no element of the two sounds is omitted by the Pâli speakers. They had not the patience to hear and reproduce the two consonants immediately after each other, by putting their vocal organs into two different positions, or laboured under a physical inaptitude for doing so, their tongue not being sufficiently trained for the successive movements. But short of this, all the constituents of the sound, the *s'vâsa*, light and heavy, the completeness of contact, and the force of utterance, are represented in their pronunciation, and the amount of muscular exertion involved is the same as in the case of the original. This may be called the energetic mode of pronouncing conjuncts. There are, however, a few instances in which the components are separated by inserting a vowel between them. Thus Sanskrit सूक्ष्म is changed to सुखुम, औष्मिक to ओष्मिक, श्लेष्म to सिलेष्म, स्मरति to सुम्रति, श्री and ह्री to सिरी and हिरी, बैलव to बैलव and औपश्लेषिक to ओपसिलेसिक. The separating vowel is generally such as can be easily pronounced along with the previous or the following consonant, that is, belongs to the same organic position as either. Thus in the first four instances we have the labial उ in the syllables खु and owing उ to the influence of the following म्. The conjuncts र्य and ह्य are almost invariably dissolved into रिय and रह्य as in कारिय, अरिय, सुरिय, विरिय &c. and अरह, गरहा, बरिह &c. for कार्य, आर्य or आर्य, सूर्य, वीर्य &c. and अर्ह, गर्ह, बर्ह &c.; and in one or two instances we have the first change when the य is preceded by a consonant other than र, as in चैतिय for चैत्य. This change is to some extent due to the य being pronounced weak, *i. e.* almost like इ. In the same way, we have सिलोक, सिलेस, सिलाघा, किलेस, किलन्त, मिलान, मिलान, पिलुवति, सुपिन, सिनिद्ध, नहान or सिनान &c. for श्लोक, श्लेष, श्लाघा, क्लेश, क्रान्त, ग्लान, म्लान, प्लुवति, स्वप्न, स्निग्ध, स्नान &c. This mode of utterance wants the force of the one we have examined, but both the sounds of a conjunct are clearly reproduced in it. You will have observed that the second member of the conjunct dissolved in this way is what may be called an imperfect consonant, *i. e.* a semi-vowel, aspirate, or nasal, in pronouncing which the breath is not completely stopped. They therefore act like a vowel, and render the sound of the first consonant comparatively distinct and audible, but are not so weak themselves as to melt away into that sound. Hence both the sounds are audible but the Pâli speaker not being able to put his vocal organs into two consonantal positions successively, as we have seen, reproduced the two sounds by interposing a vowel between them. But when in some cases the second member was weakly pronounced it melted away into the first, and so we have the optional forms विस्सरति for विस्मरति, सेह for श्लेष्म, सोप्प for स्वप्न &c. When, however, its pronunciation became stronger than that of the first, it prevailed, and so we have कय्य for कार्य and अय्य for आर्य.

c7.26

Some conjuncts appear also unchanged in this dialect; as ब्र in the words ब्रह्म, ब्रवीति, and ब्रहेति, व्य in व्याधि, व्यग्य, व्यग्ग &c., प्ल in प्लवङ्ग, प्लवति, and प्लव, स्न in स्नेह &c. Such as have a nasal for their first member and a mute for the second are also unchanged, since the nasal, even according to the rules of Sanskrit, belongs to the same organic position as the mute.

c7.27

We will now proceed to notice the changes of single consonants. An unaspirate is changed to an aspirate when it is followed in the same word by a sibilant; as in फरसु for परशु, फरस for परष, फुस्स for पुष्य, सुकुमार for सुकुमार, थुस, for तुष &c. The heavy *s'vāsa* that forms a constituent of the sibilant sound has made such a strong impression on the ear of the hearer, that he immediately prepares himself to let it out, and thus uses it even in pronouncing the letters which precede the sibilant and do not require it. The pronunciation of a word is rendered easier by transferring one of the elements of a sound to another or others, that is, by assimilating them to each other as much as possible. In a few instances the simple breath or *s'vāsa* of a previous surd is transferred to the following, originally a sonant, as in थकेति for स्थगयति. The cerebral unaspirated sonant ड between two vowels is softened into the corresponding semi-vowel ळ, as in ताळन for ताडन, पीळन for पीडन, तळाक for तडाक &c. This peculiarity distinguished the old Vedic Sanskrit also; and the sound exists in the modern Marâṭhî and Gujarâtî, though it is unknown to the other vernaculars.

c7.28

I have already given instances in which the cerebral ड of a previous syllable in going out changes the dental of the following to a cerebral, and to these I may now add निगण्ठ for निर्ग्रन्थ, वण्ट for वृन्त, कट for कृत &c. In some instances, this change takes place without such an influencing cause, as in इसति and इंस for दशति and दंश, and डाह for दाह. The same phenomenon is observable in some conjuncts, as in विदड्ढ for विदग्ध, हेठा for अधस्तात्, and ठा or ढा for स्था in the several derivatives from that root as ठाति, ठायी, उट्ठान &c. The dental nasal न is also similarly changed to ण in a few cases, as in याण, ओणत, ओणीत, सणिकं for ज्ञान, अवनत, अवनीत, शनकैः &c. The operation of this process is very limited but still it exists and cannot be accounted for in any other way than by supposing that the vocal organs or the vocal habits of the Pâli speakers were more adapted for the production of cerebral sounds, or, in other words, the people had a natural aptitude for them.

c7.29

Sanskrit श and ष are changed to स, which is the only sibilant in the language. This change involves an economy of effort. The tongue in its several movements becomes a lever with the fulcrum at the inner end where it is fixed in the mouth, and the weight in its centre of gravity. The advantage increases with the distance from the fulcrum of the point which is moved, that is, the point where power is applied. Hence gutturals are the most disadvantageous, palatals next, and dentals, the least. The cerebrals being pronounced by rounding the tongue and raising it very high, it is a question whether this additional effort does not neutralise the advantage they possess over the palatals by their position being more distant. Thus स् is the weakest of the sibilants. But in the Pâli the change is so universal that we must suppose a special inaptitude in the speakers of that language for pronouncing श and ष, as we had to suppose in the case of the conjuncts. When, however, in a few cases the peculiarity of the palatal श was distinctly preceived, and it was felt that the substituted स् did not bring it out, that sibilant was changed to the corresponding mute of its psoition, *i. e.* to च्, which resembles श in the heaviness of the simple breath required for pronouncing it. We have thus छ्व for शव, छक् for शकृत्, and छाप for शाव. The change of षष् ‘six’ to छ is also to be similarly accounted for. I have already remarked that sometimes प् must at an early period have been pronounced very much like श्. Besides these changes there are solitary instances of others, such as the softening or toning of surds as in मृग for मृक, or the dropping of the mute element of the sonant aspirates as in होति and हेद्वा for भवति and अधस्तात्. But these will be more fully discussed in connection with the Prâkṛits and the vernaculars, in which these processes have a much wider range.

c7.30

The vowel ऋ is changed to अ as in कसि for कृषि, तण्हा for तृष्णा, गह for गृह, घत for घृत, मचु for मृत्यु &c.; to इ as in दिट्ठ for दृष्ट, तिण for तृण, मिग fcr मृग, तित्ति for तृप्ति, गिद्ध for गृध &c.; and to उ as in भुस for भ्रश, मुदु, for मृदु, भुसा for मृषा, बुस for वृष, बुट्ठि for वृष्टि &c. The last change generally takes place when the vowel is preceded by a labial; so that the vocal position of the consonant influences that of the vowel. When there is no such influence, it is changed to अ and इ. Now, the vowel ऋ is composed of a consonantal and a vowel element, the former of which is subordinated to the latter. On this account it does not give to the consonant to which it is added the character of a conjunct, and the preceding vowel is not rendered heavy³. In Pâli this consonantal element disappears in virtue of the inherent inability, which we have considered at such length, of passing from one vocal position to another without letting off the breath; but the previous consonant is not doubled because the usual character of a conjunct is absent. Or the disappearance may be accounted for by the fact that the consonantal element is so subordinated to the vowel element as to escape

c7.31

8. This fact is accounted for by the authors and the commentators of the Prâtis’âkhyas by supposing that the vowel element envelopes the र् on all sides *i.e.* precedes it and follows it. (See Vâj. Pr., Ind. St. V. 145 and Ath. Pr., Whitney, I. 37.)

c7.32

notice, and that the letter, as a whole, is difficult to pronounce. The vowel element of ऋ is pronounced by bringing the root of the tongue in contact with the root of the palate, *i. e.* in the position of the gutturals; and is unrepresented independently in the Sanskrit or Pâli Alphabet. Hence it is pronounced either lower down, in the position of अ, or higher up in that of इ, while it is sometimes carried so far to the outside as the position of उ, especially when a labial precedes. That the cerebral or र element did exist, and was occasionally preceived by the Pâli speakers is proved by such examples as रिते for ऋते and रुक्व for वृक्ष, and also by the circumstance that in some cases, though it disappears, it changes the following dental to a cerebral, as in वड्ढ for वृद्धि, विसट for विस्मृत, कट for कृत &c. From these observations it would appear that the vowel we have been considering must have been in those days pronounced just in the manner in which Marâthî Pandits of the present day pronounce it, and not like अर्, इर्, or र् as is supposed by several European scholars. If वृत्त were really घर्त, or मृग, मिर्ग, the Pâli corruptions would be घत्त or मिग्ग; and if सुकृत were सुक्रत, or तादृश, ताद्रिश we should have सुक्रत or तद्रिश instead of सुकत and तादिस, and there is no reason why ऋतु and ऋषि should have lost their र् and become उतु and इसि if they were really pronounced like रितु or रतु and रिषि or रषि.⁹

c7.33

⁹ The explanation of this vowel souud and the others that follow is based upon the mode in which we Marâthâs pronounce them at the present day, and which is sanctioned by the Prâtis'âkhyas,

c7.34

The sounds of the Sanskrit diphthongs ए and औ are like those of अइ and अउ uttered rapidly, *i. e.* without allowing any appreciable time to elapse between the two elements. In forming अ, the tongue and the lips are in a natural position, the lips, however, not being closed but a little opened; while इ requires that the middle of the tongue should be brought close to the palate, and उ, that the lips should be completely rounded. Hence, in pronouncing ए and औ it is necessary to pass from one vocal position immediately to another, a process of which, as we have seen, the Pâli speakers were incapable. These diphthongs are therefore changed to ए and ओ which partake of the character of both the components. In the formation of ए the tongue is not horizontal as in the case of अ; its middle is raised up but not brought so close to the palate as in the case of इ. Similarly, in pronouncing ओ the lips are not so completely rounded as in the formation of उ; neither are they in a natural position as in the case of अ. These sounds, therefore, being produced in a position between those of अ and इ, and अ and उ, combine the characteristics of both, but are simple, *i. e.*, formed in one position only. Thus we have सेल for शैल, तेल for तैल, केवट for कैवर्त, खोम for क्षौम, कोसेय्य for कौशेय, मोलि for मौलि &c. Similarly अय and अव frequently become ए and ओ; as in कयेति for कथयति, पूजेति for पूजयति, नेति for नयति, ओधि for अवधि, ओकास for अवकाश, नोनीत for नवनीत &c. The sounds of these dissyllables differ from those of ऐ and औ in this respect that the अ of the former is not rapidly pronounced as is that involved in these diphthongs; or, in the words of the grammarians, the value of the first अ is one *mātrā* and of the other one-half. In the same way the अ of the two dissyllables is followed by the semi-vowel य and व, while that of ऐ and औ has the corresponding vowels इ and उ after it. When अय and अव are hastily pronounced, this distinction disappears, and these dissyllables assume the form of the diphthongs ऐ and औ, which by the rule just discussed become ए and ओ.

c7.35

A long vowel is shortened when it is followed by a double consonant; as in मग्ग for मार्ग, पत्त for प्राप्त, इक्खण for ईक्षण, कित्ति for कीर्ति, तिब्ब for तीव्र, रुक्ख for रूक्ष, मुत्त for मूत्र &c. The strong pressure that is necessary for the pronunciation of a conjunct cannot be properly exerted, as formerly observed, without momentum, for the acquisition of which the previous vowel has to be uttered with force and with a jerk. For this purpose a short vowel alone is fitted, but if a long one precedes, its utterance being by its nature slow and weak, the organs for want of momentum do not strike against each other with force, and hence the pressure they exert is weak. But the Pâli speaker, catching from his Sanskrit teacher only the generally strong nature of the pressure involved in the utterance of conjuncts, realized it to the fullest possible extent, without stopping to observe how much it was impaired by the length of the previous vowel, by pronouncing the previous vowel with force and rapidity, and thus rendering it short. And in this way the real quantity of the syllable is not diminished. Though the vowel is short the vocal organs take some time to emerge from the close contact and the strong pressure resulting from its forcible utterance, and hence the whole syllable इक्ख, for instance, in इक्खण, has the metrical value of a long vowel, *i. e.* is equal to two *mātrās* or syllabic instants. In the original ईक्षण, on the contrary, the pressure in the pronunciation of क्ष being weaker in consequence of the slowness of the previous vowel utterance occupies less time; so that in the one case the shortness of the vowel is accompanied by a longer duration of the contact, and in the other the length of the vowel is attended by a shorter duration of the contact. The change, then, we have been considering, is due to the pressure necessary for the formation of the conjunctal sounds in these words having made a strong impression on the ear of the Pâli speaker. When, however, the length of the vowel prevailed, and neutralized the pressure to such an extent as to render it incapable of attracting attention, the conjunct came to be in the condition of one standing at the beginning of a word; and like it preserved only the more distinct sound and dropped the other, leaving the long vowel unchanged. We have thus सीस for शीर्ष, दीघ for दीर्घ, सासप for सर्षप, &c. It is, however, clear that a long vowel followed by a double consonant is an impossibility in Pâli. A Sanskrit conjunct, whenever it made a distinct impression, was pronounced with the usual characteristics of that sound, viz., the previous rapid and forcible vowel utterance and the subsequent strong contact and strong pressure, the perceptible effect of which was the shortening of the previous vowel and the doubling of the following consonant.

c7.36

Without the momentum acquired from the force and rapidity of the previous vowel utterance, the Pâli speaker could not form a strong contact and exert strong pressure, that is, without a short preceding vowel they could not pronounce a double consonant. If then the Pâli has such words as तेत्तिस (त्रयस्त्रिंशत्), सेट्ठि (श्रेष्ठिन्), नेत्त (नेत्र), सोत्थिय (श्रोत्रिय), योग्गा (योग्या), मोक्ख &c., it follows that the vowels ए and ओ were in such cases pronounced short. Similarly, when before doubles इ and उ are changed to them, as they often are, they must be short. This change of इ and उ to ए and ओ arises from the fact that the force and rapidity of the jerk with which the current of breath is sent up to pronounce the former vowels is apt to prevent the tongue from rising as close to the palate or the lower lip to the upper, as is necessary for the formation of इ or उ, or they are apt to be forced down by the current. And ए and ओ differ from इ and उ simply in the distance between the pronouncing organs being greater. This change, therefore, really involves an economy, since the effort to raise up the tongue and the lower lip across the strong current of breath blowing above is saved. We have thus, ओट्ठ for उष्ठ, पोक्खर for पुष्कर, पोत्थक for पुस्तक, सोण्ड for शुण्ड, नेक्ख for निष्क, नेत्तिस for निर्विंश, वेस्सभू for विश्वभू &c. In some cases both vowels are in use, as in the word नेक्ख which has another form निक्ख. Sometimes, especially before य, ए was invariably pronounced short, though no conjunct followed, and to make up for the loss of quantity thus occasioned the consonant was doubled; as in थेय्य for स्थेय, गेय्य for गेय, कोसेय्य for कौशेय &c. The vowel ओ was also similarly treated in a few cases, as in ओस्सजति for अवसृजति. Thus then not only has the Pâli a short ए and ओ, but the speakers of the language seem to have possessed a predilection for those sounds. Besides the changes we have examined, there are stray examples of others, such as उसु and उच्छु for इषु and इक्षु, in which the उ of the following syllable influences the utterance of the first vowel, सोप् for स्वप्, in which the semi-vowel व् is dissolved into the corresponding vowel उ, and thence transformed to short ओ, मोर for मयूर, in which the य being lightly pronounced loses its consonantal character and the vowel ऊ with the preceding अ forms ओ, and others.

c7.37

You will thus have observed that the phonetic changes which Sanskrit words undergo in passing into the Pâli may be brought under a few general rules. There are not such various and extensive corruptions both of vowels and consonants as we find in the later dialects and in the modern vernaculars. It has been estimated that two-fifths of the Pâli vocabulary are composed of pure Sanskrit words, and the remaining three-fifths, of words altered in one or other of the modes explained above. The conclusion to be drawn from these facts is, that when the Pâli was in use, the tradition of the original Sanskrit was not distant; the words, so to say, did not stray away long from the mother-language, so as to undergo extensive alterations. If so, how is it that some of the changes such as the transformation of ऐ and औ into ए and ओ and the assimilation of consonants are so universal, the first being without any exception, and the second with but a few unimportant ones? The principle which guides phonetic change is the economy of effort, understanding the word in its widest sense. This economy is observable in the two kinds of changes, as well as in the transformation of surds into sonants, the elision of consonants or of some of their elements, the assimilation of the vowels or the single consonants that make up a word, and in several other processes. But in the Pâli there are very few instances of some of these, and none at all of others. The language had not a sufficiently long duration of independent existence to bring them into extensive operation. If, then, the two processes we have noticed are found in full play in that dialect, the reason must be sought for in the vocal peculiarities of the people who spoke it. Though they heard conjunct consonants and the diphthongs ऐ and औ pronounced by the speakers of Sanskrit, as correctly as the other letters which they did not corrupt, their organs were not fitted to utter them. These peculiarities may have been natural or acquired. If natural, the people who first corrupted Sanskrit into the Pâli must have belonged to an alien race which came into close contact with the Âryas and learnt their language. If they were acquired, a branch of the same Âryan race must be supposed to have been isolated in some part of the country, and to have developed them, in consequence of being cut off from the main body. But this supposition must be rejected for the reasons that have been already given. Such complete isolation as could give rise to new vocal peculiarities must be expected to have occasioned greater phonetic decay in other respects than is observable in the Pâli. And our analysis of the Sanskrit conjunctal and the Pâli double sounds favours the first view. For, we have seen that these latter represent all the elements of the former, but they are combined in a sound produced in one vocal position only. The Pâli speakers endeavoured to reproduce the sound of a conjuct faithfully, their pronunciation was not weak, as is that prevailing in modern times, but energetic and correct in every other respect; but they could not combine energy of utterance with two successive movements of the vocal organs. This could only be because their organs were not sufficiently trained for the purpose; in other words, because the sounds were foreign to them. The condition of men who have to learn the language of others is similar to that of children, whose organs of speech are being exercised for the first time. Healthy children, whose utterance is energetic, pronounce the conjuncts almost in the same way as the Pâli speakers did. And there is another instance in History of an alien race having treated the sounds of the language of a civilized community in just the same way. The Barbarians who overran Italy and developed the Italian from the Latin, showed the same inability to pronounce the Latin conjuncts, and assimilated them as our Pâli ancestors did.

c7.38

If this supposition is correct, we must find other traces of the peculiarities of this alien race. And such we do find. The existence of the short ए and ओ in the Pâli, and the predilection the people showed for them, as well as the change of dentals to cerebrals without any influencing cause, are similarly to be attributed to the natural vocal tendencies of the people. These sounds must have existed and played an important part in the original language of this people, so that they were unable to shake them off entirely, even when they left their own tongue and learned that of the more civilized Âryas with whom they came to be closely incorporated. If the original Pâli speakers belonged to the same race as the Dravidians of Southern India of the present day, we have a reason to believe that their native tongue contained them; for they exist in the Dravidian languages and are very characteristic of them.

c7.39

We shall also find in the Pâli, and even in the passage placed before you, examples of another phenomenon presented by a growing language. Several new words, unknown to Sanskrit, but formed from Sanskrit roots, have come into use. Such is मनापो, derived very likely from मनस् ‘mind’ or ‘heart’, and आप् ‘to obtain’ ‘to meet,’ so that the word signifies ‘something that comes up to the wishes of the heart’, ‘pleasing’. The word सच्चे is from चेत् with स prefixed, which स seems to be the same as the nominative singular of the masculine of तद्. For, it is so used in Sanskrit before चेत्, though it has there an independent sense; as in स चेन्निदौहित्रस्तलक्षणोपपन्नः (Śâkuntala). Often used together in this way, the two words formed a compound expression, and, the independent character of स being forgotten, it came to be looked upon as one word. In the same way, सन्निक must have come into existence from the frequent use of such expressions as तस्सन्निकं, ब्राह्मणसन्निकं &c. made up of the genitive singular of a noun or pronoun ending in अ and अन्तिक. The portion सन्निक then came to be regarded as an independent word, and was used as such. Another new word is फासु or फासुक ‘agreeable’, ‘pleasant’, corresponding to such a Sanskrit word as स्पर्श or स्पर्शुक formed from the analogy of such verbal derivatives as भिक्षु, इच्छु, or लाभुक, पाबुक &c., and meaning ‘that which is or deserves to be touched’, or ‘pleasant to the touch’. Other instances are पिलन्धन ‘an ornament’, पेय्याल ‘plentiful’ &c.

c7.40

We will now proceed to the examination of Pâli grammar. When after years of successive creative efforts, the language of our Âryan ancestors came to be so rich in all kinds of grammatical forms, as the Vedic or middle Sanskrit is, it became cumbrous, and the tendency set in, as we have seen, of dropping away some of them and rendering the grammar simpler. The duals of both nouns and verbs are unnecessary; the occasions for using them do not often present themselves. Hence, even in Sanskrit, their use must have been rare, and the Pâli, which, in its original form at least, must be taken to represent the current usage, has dropped them away altogether. But the manner in which the process of simplification is principally carried on is by the use of false analogies. Thus in Sanskrit, nouns ending in इ and उ of the neuter gender prefix न to the terminations of the vowel cases. But in Pâli it is added to those of the corresponding cases of masculine nouns also, as in मचुनो for मृत्योः in the passage before us. Similarly, from the analogy of neuter nouns in अ which form their nominative and accusative singular by adding म, the pronouns यद् and तद् come to have यं and तं for the corresponding cases, instead of the Sanskrit यद् and तद्. In Sanskrit, the conjugation that is very often used is the first in which अ is tacked on to the root in the special tenses, and it embraces a large number of the most ordinary roots. Here, in our passage, we find the analogy extended to the root हिंस् the present tense of which is in the mother dialect ordinarily formed by inserting न between the two letters, and adding the termination to the final (हिनस्ति 3rd person singular), and we have हिंसति. The root ज्ञा takes the form of जा in the present and other special tenses, and to it is added the conjugational sign ना, so that it becomes जाना. Now, this special form is generalized, and used in other tenses also, such as the Aorist and the Future. Thus, we have संजानिंसु, the Aorist 3rd person plural of ज्ञा, and in other places we find जानिस्सति as one of the forms of the Future. The special forms इच्छ, पश्य, शृण, बुध्य, and प्य are similarly generalized, the last four being phonetically changed to पस्स, सुण, बुज्झ, and पज्ज. In the same way in forming the causative of a root प or आप is added in Sanskrit to roots ending in आ and to a few others. It is extended to all roots in the Pâli, and thus we have तिकिच्छापेहि for चिकित्सय in the above. The termination त्वा of the absolutive is, in Sanskrit, replaced by य when a root has a preposition prefixed to it. But here no such distinction is observed, and त्वा is used in all cases. The operation of this law of false analogies is very extensive in the grammar of the Pâli and the later Prâkrits; and by its means new forms have been made up instead of those current in the primitive language. But side by side with these, we often find the latter also in corrupted by the usual phonetic laws, and having a sort of isolated existence, since they transgress the general rule that has newly come into operation. Thus we have जानिस्सति, पस्सिस्सति and लभिस्सति according to the law of analogy; but अस्सति from ज्ञास्यति, दक्खति from द्रक्ष्यति, and लच्छति from लप्स्यते are also in use. These three processes then, viz, the dropping away of forms not required for the expression of the current ordinary thought, the formation of new ones on the principle of analogy, and the preservation of the old ones in what may be called an isolated or petrified condition, have contributed to the formation of the grammar of this and the succeeding languages. With these preliminary observations, I will now hastily pass under review the principal points in the grammar of this dialect.¹⁰

c7.41

¹⁰ My authority in this portion of my subject is Kachchâyana, as edited by M. Senart.

c7.42

The Pâli has lost the dual, and also the dative and ablative cases, except of masculine and neuter nouns ending in अ.∗ The only occasion when the first is ordinarily used is when ‘giving’ is expressed, but the genitive case has such a comprehensive signification in Sanskrit that it denotes the dative relation also, and is often used in that sense. Most of the relations expressed by the ablative are denoted by the instrumental, and for the expression of the peculiar ablative sense, viz., the separation of one thing from another, the particle तम् had come into very general use, even in the parent language. The genitive and instrumental thus took up the place of those two cases. But people do not forget what they have frequent occasion to use. The great majority of nouns in Sanskrit end in अ, and of these the singular is oftener used than the plural. Hence the singulars of the dative and ablative of these nouns are preserved in the Pâli, notwithstanding the operation of the causes that drove away these cases from other places. The म् of the termination of the instrumental plural is optionally changed to ह. The locative singular of masculine and neuter nouns ending in a vowel is formed by adding the pronominal termination स्मिन्, phonetically changed to स्मि and म्हि; and the ablative स्मात् in the form of स्मा or म्हा is used optionally in the case of nouns in अ. The termination स्य of the genitive singular of this class of nouns is generalized, and in the form ofस्स applied to all nouns of the masculine and neuter genders. It should be remembered that in accordance with the general rule, the Sanskrit grammatical forms drop the final consonant, including a visarga, in passing into the Pâli. The nasal म् is changed to an anusvâra, and since this, like a conjunct consonant, necessitates the rapid utterance of the preceding vowel and renders it heavy, the vowel is shortened. We have thus नं for नाम्, सं for स्याम्, यं for याम्, &c. The न् of स्मिन् and of the syllable अन् occurring in some of the cases is also changed to an anusvâra.

First 43 of 107 passages. Read on

Linked texts (25)