Pramāṇa · Search · About · API

Wilson Philological Lectures on Sanskrit and the Derived Languages c8: Lecture III. The Prâkrits and the Apabhrams'a

Wilson Philological Lectures on Sanskrit and the Derived Languages c8

Language
English

Contents

Lecture III. The Prâkrits and the Apabhrams'a

c8.2

On the last occasion we examined the language of the sacred books of the Southern Buddhists, and found that a large portion of the words it contains are pure Sanskrit and the rest are Sanskrit words corrupted or transformed according to certain laws of phonetic decay. Then by the law of false analogy the less used and less known declensional and conjugational forms have been in many cases brought over to the type of those more used in Sanskrit and consequently better known. So that in the vocabulary and the grammar the laws of growth I traced in the opening lecture are in operation, but their range is limited, and the dialect is in what may be called the first stage of departure from Sanskrit. We then examined the language of the Inscriptions of As'oka and found that it is either the same as Pâli or in the same stage of development, and that there existed in those times two or three varieties of speech slightly differing from each other. Today I propose to examine certain other dialects which exhibit a much greater departure from the parent tongue. These are the so-called Prâkrits. For a knowledge of these languages we have not to go beyond India, as in the case of the one we have examined. Prâkrit dialects possessed a literature and a portion of it has come down to us.

c8.3

There exist about six treatises on Prâkrit grammar, the most ancient of which is Vararuchi's Prâkritaprakâs'a. Next comes Hemachandra, a Jaina scholar of Gujarat, who lived in the twelfth century. His work on grammar is known by the name of Haimavyâkarana, the eighth chapter of which he devotes to the grammar of the Prâkrits. Hemachandra's treatment of these dialects is fuller than Vararuchi's; and his observation was wider. He shows a very intimate knowledge of the existing literature of these languages, both sacred and profane, Jaina or Brahmanical. His work and especially the last portion is full of quotations. He must have availed himself of the labours of former scholars, since he often mentions Pûrvâchâryas. Hemachandra also wrote a Kosha or thesaurus of the Des'î words existing in these languages. Vararuchi gives the grammar of four dialects, which he calls Mahârâshtrî, S'aurasenî, Mâgadhî, and Pais'âchî. The names of the first three themselves would show that they were the languages spoken or used in the provinces from the names of which they are derived, but doubts have been raised as to their genuineness, which will be hereafter considered. The Mahârâshtrî is called the principal Prâkrit. For instance, Daṇḍin in his Kâvyâdars'a says—

c8.4

"The language prevalent in Mahârâshṭra they regard as the Prâkṛit pre-eminent; it is the ocean of jewels in the shape of good literary works, and the Setubandha and others are written in it¹."

c8.5

¹ महाराष्ट्रयां भाषां प्रकृष्टं प्राकृतं विदुः । सागरः सूक्तिरत्नानां सेतुबन्धादि यन्मयम् ॥

c8.6

Vararuchi devotes the first nine chapters of his work to the Mahârâshṭrî, and then a chapter each to the rest. The peculiarities only of the latter dialects and their differences from the Mahârâshṭrî are given, and in other respects they are to be considered similar to the first. Hemachandra follows the same method; but he does not mention the name Mahârâshṭrî and speaks of the dialect only as the Prâkṛit. These grammarians and all others who have written on the subject treat of the grammar of the language etymologically. They take Sanskrit as the original language or Prakriti and give rules about the various phonetical and grammatical changes which have reduced Sanskrit to the Prâkṛit form. The Pâli grammarian Kachchâyana treats the dialect not as one derived from Sanskrit as these writers do, but as an independent language, though it is very probable he knew Sanskrit, since he uses Sanskrit grammatical terms, and his Sûtras greatly resemble those in the Kâtantra and even Pâṇini. Vararuchi and Hemachandra derive S'aurasenî also from the Sanskrit as they do the Mahârâshṭrî or the principal Prâkṛit, but make the S'aurasenî the Prakriti or basis of the Mâgadhî and the Pais'âchî. This appears to be the tradition; whence it would seem that older and more developed language or the language of respectable people was the S'aurasenî, and the other two were the dialects of border countries used by persons in a lower scale of society. They have some of the peculiarities of the S'aurasenî, and come nearer to it than to the Mahârâshṭrî. Hemachandra gives the grammar of two more dialects, the Chûlikâ Pais'âchî and the Apabhrams'a, the latter of which was according to Daṇḍin, the language of Âbhîras (cowherds) and others. Another grammarian of the name of Trivikrama gives in his Prâkṛitasûtravṛitti the grammar of these six dialects. He lived after Hemachandra, since he mentions him in the introduction to his work, and his book resembles Hemachandra’s a good deal. There is another work by Chandra called Shaḍbhâshâchandrikâ which is a meagre production. Another still of the same nature, the author of which is Lakshmîdhara, mentions the same six dialects; so that the expression Shaḍbhâshâ seems to have become proverbial.

c8.7

The Mahârâshṭrî derived its importance from its literature. From the manner in which Daṇḍin speaks of that literature it appears to have been very extensive and valuable. He himself mentions one work, the Setubandha, a poem attributed to Kâlidâsa but written by one Pravarasena, whose “fame,” Bâṇa says in his Harshacharita, “reached the other side of the ocean by means of the Setu.” I find in a Ms. in the colophon at the end of each âśvâsa or canto, sometimes इअ सिरीपवरसेणविरइए दहमुहवहे ( thus in the Das'amukhavadha composed by the prosperous Pravarasena ), and sometimes, इअ पवरसेणविरइए कालिदासकए दहमुहवहे ( thus in the Das'amukhavadha, the work of Kâlidâsa, composed by Pravarasena ). Some kings of Kas'mîr bore the name of Pravarasena, but there is nothing to show that any one of them was the poet who wrote this work. There is a collection of seven hundred songs, chiefly of an amorous nature, by a poet of the name of Hâla, which is called the Saptas'atî. We have an edition of this in Roman characters by Prof. Weber. Another long poem entitled the Gauḍavadhakâvya by a poet named Vâkpatirâja, who lived at the court of Yas'ovarman, king of Kanoj, in the early part of the eighth century, was discovered by Dr. Bühler about three years ago. And several other works may turn up, if diligent search is made for them. The Kâvyaprakâs'a contains about 75 Prâkṛit verses quoted to illustrate the rules laid down by the author, and S'ârṅgadhara also gives a good many in his Paddhati.

c8.8

The religious books of the Jainas form another very extensive branch of Prâkṛit literature. Prof. Weber thinks the language of these to be later than the Pâli and earlier than the Prâkṛits, so as to occupy a middle position, and calls it Jaina-Mâgadhî. But Hemachandra himself, who must have known his religious books well, and was, as I have observed, a great Prâkṛit scholar, treats it as the principal Prâkṛit or Mahârâshṭrî, and in his grammar of this he in several places gives forms of words peculiar to his sacred language, which after the example of his Brahmanic brothers he calls Ârsha Prâkṛit². In giv-

c8.9

आर्षम् । ३ । ऋषीणामिदमार्यम् । आर्षे प्राकृतं बहुलं भवति । तदपि यथास्थानं दर्शयिष्यामः । आर्षे हि सर्वे विधयो विकल्प्यन्ते । He has also said before, that the rules he

c8.10

ing his first rule about the Mâgadhî dialect, *viz.*, that the nom. sing. of Masc. nouns takes the termination ए, he says:—"As to what the fathers have said about the Ârsha (works) being composed in the Ardha-Mâgadhî dialect in such words as these: 'the ancient Sùtra is composed in the Ardha-Mâgadhî dialect', they have said so in consequence of the observance of this rule and not of those that follow³. Thus if one chooses to call the sacred language of the Jainas Ardha-Mâgadhî on account of this Mâgadhî peculiarity and a few other archaisms, he may do so; and I shall presently have to observe that the great many dialects which writers on poetics give differed from each other in such insignificant particulars only. But it is clear that Hemachandra considers the distinction to be slight, and identifies the dialect with the principal Prâkṛit; and both he and the Jaina fathers refer it to the class of the Prâkṛits of the grammarians.⁴

c8.11

³ [Footnote text missing in source]

c8.12

⁴ [Footnote text missing in source]

c8.13

is going to give even with regard to the ordinary Prâkṛit should not be considered universal.

c8.14

3. यदपि पोराणमद्धमागभ्भासानिययं हवइ सुत्तमित्यादिनार्षस्यार्धमागधभाषानियतत्वमाम्नायि वृद्धैस्तदपि प्रायस्त्यैव विधानान्न वक्ष्यमाणलक्षणस्य ।

c8.15

4. The only specific grounds one can find in Prof. Weber's book in support of the assertion that the Jaina dialect occupies a middle position are these:—1. That uninitial क्, ग्, च्, ज्, त्, द् and other consonants are dropped in the Mahârâshṭrî leaving only the vowel, and preserved or softened in the Pâli, while in the Jaina books य is substituted for them; *i. e.*, the different stages of phonetic corruption in this case are, the consonants in their original or softened forms, then य for them, and lastly their elision. 2. That न is preserved in the Pâli, and changed everywhere to ण in the Mahârâshṭrî; while in the Jaina dialect initial न remains unchanged except in enclytics. 3. That the loc. sing. of nouns in अ ends in सिं or सि which we find in the language of the column inscriptions; while it is स्मि and म्हि in the Pâli and म्मि in the Prâkṛit. Now as to the first, the य is not prior to the elision, but contemporaneous or subsequent to it, being found even in the modern vernaculars. It was introduced simply to facilitate pronunciation; *i. e.*, it is a strengthened form of the vowel. Thus the Prâkṛit of पाद foot is पाउ, but in Marâṭhî we have पाय; so राजन् Skr., राअ Pr., राय H.; सोदर Skr., सोअर Pr., सोयरा M., &c. &c. The य occurs not only in Jaina books, but everywhere in the Gaudavadhakâvya; and Hemachandra does tell us in his sûtra अवर्णो यश्रुतिः that the अ that remains after the elision of a consonant is pronounced like a soft य. With regard to the second, initial न is found unchanged in the Gaudavadha in a great many places. Thus in stanza 242 we have the negative particle न, in 241 निवइ for निपतति, in 245 नाह for नाथ, and in 251 नह नख. These instances I have found on simply opening the Ms, at random; and no great search But it is in the dramatic plays that we find these dialects principally used. Writers on Poetics prescribe that a particular dramatic person should speak a particular dialect. Sanskrit is assigned to respectable men of education, and women in holy orders; Śaurasenī, to respectable ladies in their prose speeches, and the Mahārāṣṭrī or the principal Prākṛit, in the songs or verses put into their mouths. Śaurasenī is also assigned to inferior characters; and the Māgadhī and Paiśāchī to very low persons. The general rule is that a dramatic person should speak the language of the country to which he or she is supposed to belong. Later writers give more minute rules. For

c8.16

was necessary. Hemachandra also in his sūtra वादी following another, णो णः, says that the initial न is sometimes changed to ण, sometimes not. As to the third, the termination ḟ may constitute a peculiarity of the language, but it is by no means an index to its higher antiquity, since it occurs in the pronominal locative of the principal Prākṛit. There are several peculiarities in the Jaina books, and a good many of them are noticed by our Grammarian, but they do not show an earlier stage of development.

c8.17

This continues still to be my view, notwithstanding all that has since been published on the subject. Dr. Hoernle, in the introduction to his edition of Chanda’s Prākṛitalakshana makes an elaborate attempt to prove that the dialect, the grammar of which is given in that book, is more ancient than the Prākṛit of Vararuchi and Hemachandra. But it is not at all difficult to see that he is altogether on a wrong tack. He says there is nothing in Hemachandra corresponding to the rule given by Chanda about dropping the final vowel of the first member of a compound when the initial vowel of the second is followed by a conjunt consonant, in such words as dhana+ādhya, deva+indra, &c., which in that Prākṛit have the forms dhanaddha, devinda, &c. This change, however, does come under Hemachandra’s rule I. 84, which provides for the shortening of a long vowel when followed by a conjunct consonant. The short vowels corresponding to ए and ओ are इ and उ; and among the instances given by Hemachandra, we have narindo for narendra, aharuttha for adharoshtha, Nīluppala for Nīlotpala &c. Dr. Hoernle thinks the changes of i to short e and of u to short o are later Prākṛit changes. But he will find many instances of them in the Pāli, which certainly is an older dialect than any Jaina Prākṛit. They are, he says, unknown to Chanda. Chanda’s work is a very meagre production, in which very little endeavour is made to classify facts; and thus he must be supposed to include these changes under his very general rule that one vowel takes the place of another vowel (II. 4). The instance ginhati incidentaly given by him in connection with another rule does not show that in his Prākṛit the form genha did not exist, much less that the change of i to e was unknown. Then with regard to consonants, Dr. Hoernle says there are five points in which the “older Prākṛit” of Chanda, as he calls it, differs from the Prākṛit of Vararuchi and Hemachandra. One of these is “the preservation of the instance, the author of the Sâhityadarpaṇa assigns Mâgadhî to the attendants in the royal seraglio, Ardhamâgadhî to footmen, royal children, and merchants, Prâchyâ to the Vidûshaka and others, Âvantikî to sharpers, warriors, and clever men of the world, Dâkshiṇâtyâ to gamblers, Śâkârî to Śakârs, Śakas, and others, Bâhlikâ to celestial persons, Drâviḍî to Draviḍas and others, Ábhîrî to cowherds, Châṇḍâlikî to outcastes, Ábhîrî and Śâbarî, also to those who live by selling wood and leaves, and Pais'âchî to dealers in charcoal. Handmaids, if they do not belong to a very low class, should speak Śaurasenî. Some of the modern grammarians also mention as large a number of dialects. But whether these were actually used by writers

c8.18

dental *n* in every case." For this statement the Doctor quotes the authority of a sûtra in which we are told by Chaṇḍa that *ṇ* and *ṇ* do not exist in the Prâkrit, as compared with another version of that sûtra which says that *ṇ*, *n*, and *ṇ* do not exist (II. 14). This last version no doubt provides for the change of *ṇ* in all cases; but the other which denies the non-existence or affirms the existence of *ṇ* cannot mean that it exists or is unchanged in *all* cases. The denial of non-existence or affirmation of existence only proves its existence or remaining unchanged in some cases. Besides we have a specific rule where we are told that a letter of the *d* class takes the place of the corresponding letter of the *t* class (III. 16), thus providing for the change of *n* to *ṇ*; and the instance given is धण्णं for धान्यम्. But this rule the Doctor thinks holds good in exceptional cases, for which however there is no authority whatever, and he gives none. Again, he says that his statement is proved by the uniform spelling of the Prâkrit examples with *n* in Mss. A and B, which, according to him, contain the older version of the work. I, however, find that the spelling in his edition, which is the spelling of A and B, follows uniformly, with one or two exceptions, in which we must suppose a mistake, the rule laid down by Hemachandra, *viz.*, that initial *n* is optionally changed to *ṇ*, while medial *n* is necessarily so changed. The instances are:—all the case-terminations which are to be traced to the Sanskrit terminations having *n* in them, such as णे, णि, and णं; सुणिणे for सुनीन् (I. 1), गयणओ for गगनतः (I. 16), मणिक्सा A. (I. 21), नाणी (I. 24) for ज्ञानी, सक्कीसाणा for शक्तेशानौ (II. 1.), काणं (II. 15), गयणं for गगनम् (II. 21), पीलणं for पीडनम् (II. 24), नाणं for ज्ञानम् (III. 6), जुव्वणं for यौवनम् (III. 15), and रअणं for रत्नं (III. 30). The second point is "the preservation of the medial single surd unaspirate consonants with the only exception of *k*." This simply means the preservation of *ch*, *t*, and *p* which according to the other grammarians are generally dropped. Chaṇḍa agrees with these in dropping not only *k*, but *g*, *j*, and *d*. Whether therefore the preservation of *ch*, *t*, and *p*, supposing that Chaṇḍa really allows it, marks off his Prâkrit as older than that of Hemachandra or Vararuchi is more than questionable. But, as a matter of fact in III. 12, he does provide for their change to *j*, *d*, and *b*; and the change of *t* to *d* is a Śaurasenî, *i. e.*, a local change, and does not indicate priority of time. Besides, even the dropping of these consonants must have been contemplated by him.

c8.19

of dramatic plays in accordance with the rules of Rhetoricians, or if they were, what constituted the exact difference between these various languages, it is impossible to determine so long as we have not got satisfactory editions of the plays. Still in those cases in which we have the assistance of the older Prâkrit grammarians, the characteristics of each can be made out with fulness and certainty. To illustrate his rules about the Mâgadhî, Hemachandra quotes from the speeches of the fisherman and the two poilcemen in Śakuntalâ, of the Kshapaṇaka from the Mudrârâkshasa, and Rudhirapriyâ from the Veṇîsamhâra. The points in which the Mâgadhî chiefly differs from the principal Prâkrit and Śaurasenî are these:—र् and स् of these are changed to ल् and श्; स्थ and थ of Sanskrit to स्त, and ट to स्ट; and स् is not assimilated as in पस्खलदि from प्रस्खलति, the nom. sing. of masc. nouns ends in ए instead of ओ, which is the Prâkrit ending; the gen.

c8.20

For in the instances given in the book, they are dropped in all the manuscripts used by Dr. Hoernle, even in A and B, which according to him contain the older version. Thus we have एइ for एति (I. 12), गयणओ-उ or गयणाओ-उ for गगनतः (I. 16), कयं for कृतम् (I. 23 and everywhere else), इहागओ-उ for इहागतः (II. 1), इच्छिअं for इच्छितम् (II. 3), कायव्वं-व्वं for कर्तव्यम्, सुइणं for सूचीनाम् (II. 4), नेउरं for नूपुरम् (II. 4), घयं for घृतम्, काउण for कृत्वा, दीसइ for दृश्यते (II. 5), गच्छइ for गच्छति, वइ for पतिः (II. 10), सरिआणं or सरियाणं for सरिताम् (II. 11), गइ-ई for गतिः, मइ-ई for मतिः (II. 17), &c. &c. It is very much to be regretted that the Doctor should in all these cases have set aside the readings of his manuscripts and invented his own with the त्, च्, and प् standing in the words, instead of being dropped. This invention or restoration, as he calls it, is based on a singular inference that he deduces from a single word, कृतं given as a Prâkrit word in the book. He says it must originally have been कतं, but the copyist, not knowing of such a word being in the later Prâkrit which he knew, took it as the Sanskrit कृतं, and wrote accordingly. If, therefore, कतं was the Prâkrit word in this case, it must have been so in all those cases in which कयं occurs in the book, and so Dr. Hoernle makes it कतं throughout. But it did not strike him that if the copyist knew Prâkrit enough to see that कतं was not a Prâkrit word, he must have seen that कृतं also was not a Prâkrit word; and could not have given it as such. Now the reason why these consonants were not admitted by Chanda, according to the manuscripts A and B, among those that are dropped, but only among those that are softened must be that all these rules are only general and not universal, and there were as many instances of softening as of dropping. Besides, I have already said that Chanda’s work is perfunctory, and does not show accuracy of observation and statement. The third point is the preservation of the medial single surd aspirate consonants with the only exception of kh; i. e., द्, थ्, फ्, are preserved. But III. 11 provides for t e change of these to द्, ध्, and भ्, if we look to the sense of the sing. of masc. and neut. nouns optionally in आह as कम्माह; and the form of the nom. sing. of the first personal pronoun is हगे. If we apply the test furnished by these rules to the several dialects used by the characters in the Mṛichchhakaṭika as it is in the existing editions, which play contains a large variety of characters, and consequently of speech, we shall find that the language of the Châṇḍâlas, the Śakâra, his servant Sthâvaraka, and even Kumbhîraka and Vardhamânaka, is Mâgadhî, though the rules about स्थ, थ, and ॡ are scarcely observed. There is hardly any dialectic difference in their speeches. But the Sâhityadarpaṇa would lead us to expect his Châṇḍâlikâ and Śâkârî respectively in the first two cases, and Ardhamâgadhî in the last three. As before observed, some admixture of Mâgadhî characteristics constitutes this last dialect. Under this view there are instances of the use of the Ardhamâgadhî, as Lassen remarks, in the Prabodhanandrodaya. The dialect used by Mâthura, the keeper of the gambling-house in the Mṛichchhakaṭika, is somewhat different. In his speeches, we sometimes find ल् and श् used for र् and स्, and sometimes not. The nom. sing. ends in ओ, as in the Mahârâshṭrî or Śaurasenî, in some cases, in others it ends in ए as in the Mâgadhî, and sometimes in अ as in the Apabhraṃśa; and the gen. sing. sometimes ends in आह as in the Mâgadhî. If the text is to be depended on, the Dâkshiṇâtya which Vis'vanâtha⁵ attributes to gamblers may be such a mixed language.

c8.21

⁵ Another gambler without a name is introduced in the same scene, whose language Prof. Lassen thinks is Dâkshiṇâtyâ and Mâthura's, Âvantikî. Very few speeches, however, are given to the former, and it is not possible to come to any

c8.22

It would thus appear that if all these inferior dialects did exist and were used by dramatic writers, they differed from each other in unimportant particulars, and that most of them belonged to the Mâgadhî species, since the Mss. have confounded them with the Mâgadhî of the grammarians. Hence we are justified in taking the real number of Prâkṛit dialects used for literary purposes to be six, viz., those mentioned by Hemachandra, Trivikrama, and Lakshmîdhara.

c8.23

sûtra and also to some of the instances that are given. The change of थ to ध, is a S'aurasenî peculiarity. Of the two remaining points one is the insertion of य to avoid the hiatus caused by the dropping of a consonant, about which I have already spoken, and the other is unimportant.

c8.24

There is, therefore, no question that the Prâkṛit, a meagre grammar of whic his given in the work edited by Dr. Hoernle, is not older than Hemachandra's.—(1887).

c8.25

Daṇḍin mentions a work of the name of Bṛhatkathâ written in the language of the ghosts, *i. e.*, in the Paiś'âchì. Dr. Bühler has recently obtained a trace of the work, and arrangements have been made for getting it copied⁶. It is traditionally ascribed to a poet of the name of Guṇâḍhya.

c8.26

Let us now examine cursorily the principal Prâkṛit and the Śaurasenî which ranks next to it in literary importance, but as the model of the inferior dialects and as the language used by the higher class of Prâkṛit-speaking dramatic persons in their prose speeches is more important than the other. Here, as before, I will place a specimen of each before you:—

c8.27

61. निय[अ]आए चिय[अ] वायाइ अत्तणो गारवं निवेसन्ता ।
जे यन्ति पसंसं चिय[अ] जयन्ति इह ते महाकइणो ॥
63. दोग्गचंमि वि सोक्खाइं ताण विहवे वि होन्ति दुक्खाइं ।
क्व्वपरमत्थरसिया[आ]इं जाण जायन्ति हियया[अअ]इं ॥
67. सोहेइ सुहावेइ य[अ] उवहज्जन्तो लवो वि लच्छीए ।
देवी सरस्सई उण असमग्गा किंपि विणडेइ ॥
99. अत्थिं निय[अ]तिय[अ]नीसेसभुवणदुरिया[आ]हिनन्दिय[अ]महिन्दो ।
सिरि जसवम्मो त्ति दिसापडिलग्गगुणो महीनाहो ॥

c8.28

Sanskrit:--

c8.29

61. निजयैव वाचयात्मनो गौरवं निवेशयन्तः ।
ये यान्ति प्रशंसामेव जयन्तीह ते महाकवयः ॥
63. दौर्गत्येपि सौख्यानि तेषां विभवपि भवन्ति दुःखानि ।
काव्यपरमार्थरसिकानि येषां जायन्ते हृदयानि ॥
67. शोभयति सुखयति चोपभुज्यमानो लवोपि लक्ष्म्याः ।
देवी सरस्वती पुनरस The definite conclusion from them; but so far as they go there is hardly any difference between his dialect and that of Mâthura. The Professor is led to attribute two languages to gamblers by the annotator on the Sâhityadarpaṇa whom he quotes, and who explains धूर्त by अक्षदेविन. But if the word is to be so understood, दिव्यताम् in the next line is not wanted, and neither योध nor नागरिक. For, supposing the warriors and clever worldly men were gamblers, gambling was not confined to them; whence there is no reason to mention them in particular.

c8.30

6. So Dr. Bühler told me at the time; and on a subsequent occasion I myself thought I had found a trace of the work. But up to this time all our search has proved fruitless. (1887).

c8.31

99. अस्ति निवर्तितनिःशेषभुवनदुरिताभिनन्दितमहेन्द्रः। श्रीयशोषर्मेति दिशाप्रतिलग्रगुणो महीनाथः॥

c8.32

61. "Victorious are the great poets who, establishing their greatness by their own words, do obtain praise only."

c8.33

63. "Those whose hearts appreciate the true beauty of poetry experience joys even in poverty and sorrows even in prosperity."

c8.34

67. "Even a small degree of Lakshmî when enjoyed adorns and delights, but the divine Sarasvatî if imperfect is an unspeakable mockery."

c8.35

99. "There lives a king named Yas'ovarman who delights Indra by removing all the distresses of the world and whose virtues have reached the ends of the quarters."

c8.36

Śaurasenî:

c8.37

कथं अणुगहीदम्हि। इअमालिलिङ्गमि। दंसणं उण पिअसहीए बाहुप्पीडेण णिरुद्धं ण लम्मीअदि। सहि कठोरकमलपम्हलो अण्णारिसो जेव्व दे अज्ज णिव्वावेदि सरीरफंसो। किं अ मउलिविनिवेसिदञ्जली मह वअणेण विण्णवेहि तं जणं ण मए मन्दभाइणीए विअसन्त-पुण्डरीअलच्छीविलासहारिणो मुहचन्दमण्डलस्स दे सच्छन्ददंसणेण संभाविदो चिरं महसवो लोअणाणं।

c8.38

Sanskrit:

c8.39

कथमनुगृहीतास्मि। इयमालिङ्गामि। दर्शनं पुनःप्रियसख्या बाष्पोत्पीडेन निरुद्धं न लभ्यते। सखि कठोरकमलपक्षमनोन्यादृश एव तव निर्वापयति शरीरस्पर्शः। किं च मौलिविनिवेशिताञ्जलिर्मम वचनेन विज्ञापय तं जनं न मया मन्दभाग्यया विकसत्पुण्डरीकलक्ष्मी-विलासहारिणो मुखचन्द्रमण्डलस्य ते स्वच्छन्ददर्शनेन संभावितश्चिरं महोत्सवो लोचनानाम्।

c8.40

"What! thou hast obliged me. Here I embrace. But I do not catch a glimpse of my dear friend, my sight being obstructed by the flow of tears. Friend, the contact of thy body, hairy like a ripened lotus, cools my body in a peculiar manner to-day. Moreover, with thy hands clasped and placed over thy head, do at my request, humbly say to that person, 'Unfortunate as I am, I have not feasted my eyes long, by looking freely at the moon of thy face which rivals the blown lotus in beauty'".

c8.41

In the Ms. of the Gauḍavadha, from which the first extract is given, अ and आ preceded by अ and इ are marked as य and या. In Mss. of other works the य does not appear; but there can be no question that it represent the later pronunciation correctly, since as already remarked in a note this य is observed in some of the modern vernnaculars. But it is to be pronounced like a strengthened अ or आ and not like the heavy semi-vowel that I mentioned in my observations on the Pâli.

c8.42

From these extracts you will observe that phonetic decay has made greater progress in these dialects than in the one we have examined. The changes in the Pâli are, with a few exceptions, such as may be attributed to the circumstances and vocal peculiarities of a foreign race. But in the Prâkṛits the usual processes of corruption have a wider range of operation; though even here we shall, as we proceed, find it necessary to ascribe a good deal to an ethnological cause. And first as regards the phonetic transformations which we have noticed in the Pâli, I have to observe that Prâkṛits also exhibit the same with but slight differences. The vowel ऋ is changed to अ as in घअ, तण, मअ, &c. for घृत, तृण, मृग, &c.; to इ as in किवा, हिअअ, किस, &c. for कृपा, हृदय, कृश, &c.; to उ as in पुहई, पाउस, बुढ, &c., for पृथिवी, प्रावृष्, वृद्ध, &c. When standing alone it is more often changed to रि than in Pâli; as in रिद्धी and रिच्छ for ऋद्धि and ऋक्ष; रिण or अण, रिजू or उज्जू, &c., for ऋण, ऋजु, &c., while the Pâli forms of these words are इद्धि, अच्छ, अण and उज्जु. The diphthongs ऐ and औ are as in Pâli changed to ए and ओ, as in केलास, वेज्ज, केदव, &c. for कैलास, वैद्य, कैटभ, &c.; and in जोव्वण, कोमुई, कोत्युह &c., for यौवन, कौमुदी, कौस्तुभ, &c.; but in a good many instances they are dissolved into their elements अइ and अउ, as in दइच्च, भइरव, दइवअ, सइर, &c., for दैत्य, भैरव, दैवत, स्वैर, &c., and पउर, कउसल, सउह, मउली, &c., for पौर, कौशल, सौध, मौलि, &c. This change resembles the dissolution of conjunct consonants into the different members, and like it is due to a weak or languid way of pronunciation. In ऐ and औ the first element or अ is rapidly pronounced, and the temporal value assigned to it by the authors of the Prâtiśâkhyas is, you will remember, half a mâtrâ, while in the Prâkṛit transformations it is one mâtrâ. The long vowels are, as in Pâli, shortened when followed by double consonants; and there is the same or even stronger evidence of the existence of short ए and ओ. In Pâli short इ and उ followed by a conjunct are in certain cases changed to ए and ओ, and from that fact we inferred that they were short. Here in a great many more cases when so followed, short इ is optionally interchangeable with ए, and short उ is necessarily replaced by ओ, as in पिण्ड or पेण्ड, पिठ् or पेठ्, निद्दा or नेद्दा, &c., and तोण्ड, मोण्ड, पोक्खर, &c., for पिण्ड, पिष्ट, निद्रा, तुण्ड, मुण्ड, पुष्कर, &c. In several cases ए not followed by a conjunct is optionally changed to इ, as in विअणा or वेअणा for वेदना, दिअर or देअर for देवर, &c.

c8.43

The u in these words must for some reason that we will hereafter consider have been pronounced short and hence interchangeable with i. Sometimes the consonant following an e or o is doubled, as in तेल्ल, पेम्म, सोत्त, जोव्वन, &c., for तैल, प्रेमन, स्रोतस्, यौवन, &c., which could only be because those vowels were pronounced short, and the loss of quantity thus occasioned made up for by rendering the pronunciation heavy and forcible. In other cases the u was so pronounced by some and not by others; and so we have एक्क or एअ for एक, सेव्वा or सेवा for सेवा, &c. The syllables अय and अव are changed to ए and ओ oftener than in Pâli, the अय of the causative and the tenth class becoming ए necessarily throughout, as in कारेइ, हासेइ, कहेइ for कारयति, हासयति, कथयति &c.

c8.44

All the sibilants are reduced to स as in the Pâli, but in the Mâgadhî dialect to श; as शालश, पुलिश for सारस, पुरुष, &c. These dialects do not possess the cerebral ळ, and therefore we have ल in the place of the Pâli ळ and in some cases the original Sanskrit ड; as in तलाअ, गरुळ, कीलइ for the Pâli तळाग, गरुळ, कीळति and Sanskrit तडाग, गरुड, क्रीडति, &c., गुल or गुड, णाली or णाडी for the Pâli गुळ, नाळी, &c. and पीहेइ, नीड for the Pâli पीळेति, नीळ, &c. There are a good many more examples of the change of dentals to cerebrals than in the Pâli, both through the influence of a neighbouring र or without it. The त and द which correspond to त् and थ् are in most cases softened to ड and ढ. Thus we have पडि for the Pâli पटि, as in पडिमा, पडिहार, for प्रतिमा, प्रतिहार, &c., and पहुडि, पडण, बहेडअ, for प्रभृति, पतन, बिभीतक; डोला or दोला, डम्भ or दम्भ, डब्भ or दब्भ for दोला, दम्भ, दर्भ, &c.; पढम (Pâli पढम), सिढिल, मेढी, ओसद for प्रथम, शिथल, मेथि, औषध, &c. In Pâli the dental nasal न् is changed ण in but a few instances; but here it is so changed throughout, necessarily when uninitial, and optionally when at the beginning of a word; as कणअ, मणण, वणण, for कनक, मदन, वदन, &c., and णर or नर, णई or नई, णेइ or नेइ, for नर, नदी, नयति, &c. The opposite process is however observable in the Paiśâchî dialect, where not only have we no instances of this change but even the original Sanskrit ण is changed to न as in गुन, गन, for गुण, गण, &c. The conjunct consonants are transformed in the Prâkṛit in the same way as in the Pâli. In the former however, ज्ञ and न्य are changed to णण् and not to ञ्ञ as in the latter, as in णाण, सण्णा, अण्ण, सुण्ण, for ज्ञान, संज्ञा, अन्य, शून्य, &c. To be thus corrupted, ज्ञ must in the original Sanskrit have been pronounced as if it were composed of ज् and न्; and sometimes the latter must have been so weakly pronounced that the sound of ज् prevailed over it and the whole became ज्ञ्, as in जाण or णाण, सव्वज्ञ or सव्वण्ण, मणोज्ज्ञ or मणोण्ण, for ज्ञान, सर्वज्ञ, मनोज्ञ, &c. In the Paiśâchî and the Mâgadhî, however, the Pâli transformation is retained; as in सञ्जा, सव्वञ्, कन्ञका, अभिमञ्जू, for संज्ञा, सर्वज्ञ, कन्यका, अभिमन्यु, &c. The conjunct र्य is sometimes changed to ज्ञ् corresponding to the र्य of the older dialect, and ह्य to ज्ञ् in which case the heavy nâda of ह is transferred to the ज्ञ which takes the place of य as in the change of ह्य to ज्ञ्. In Pâli the consonants of ह्य only interchange places; i. e. it becomes यह. From this and from the change of the initial uncombined य to ज्ञ, it appears that very often the Sanskrit य, was pronounced heavily when the Prâkrits arose. The Śaurasenî and the dialects allied with it have, however, both the Pâli and the Prâkrit corruptions of र्य.

c8.45

A dental forming a conjunct with a following व is in a few cases changed to the corresponding palatal; as in भोच्चा for भुक्त्वा, चच्चर for चत्वर, पिच्छी for पृथ्वी, विज्जं for विद्वान्, बुज्झा for बुद्धा, झअ for ध्वज, झुणी for ध्वनि, सज्जस for साध्वस, &c. This seems to arise from the fact that व was pronounced so lightly that it lost its distinctive character, and the conjuncts came to be confused with those containing a dental and य which, you will remember, are changed to a double palatal. In Pâli the त्व is retained in these instances, and the द्व and ध्व are changed to द्द and द्ध according to the general rules, and the ध्व of पृथ्वी becomes थ्व. Besides the conjuncts disjoined in Pâli by the interposition of a vowel, we have र्श and र्ष also so treated in the Prâkrit, sometimes optionally and sometimes necessarily. Thus आदर्श becomes आअरिस or आअंस, (P. आदास); सुदर्शन, सुदरिसण or सुदंसण, (P. सुदस्सन); वर्ष, वरिस or वास, (P. वस्स); परामर्श, परामरिस, (P. परामास); अमर्श, अमरिस, (P. अमस्स).

c8.46

Having noticed the changes common to the Prâkrits with the Pâli, we will proceed to consider others distinctive of the former. Most of these were due to the continuous operation of processes which come into play in a living language. The Pâli exhibits but few instances of these processes. The changes observable in it are mostly to be traced to one or two vocal peculiarities of the men who spoke it. At the time when the language received the form in which we now find it, the tradition of the original Sanskrit was still distinct; the Pâli had not lived an independent life detached from its mother for a long time. But with the Prâkrits the case is different. They show a great many more instances of the usual processes, and consequently a much greater departure from the parent tongue. We will begin by noticing what may be called the softening process.

c8.47

The vowels इ and उ are softened, as we have seen, to short ए and ओ before conjuncts. These latter sounds are, as indicated in the last lecture, more open than the former; that is, do not require the tongue to be raised so high as in the case of इ and उ. They are therefore softer. But since the change principally takes place before doubles, it may, I believe, be traced to their influence, as I have already observed. In that case this would be an instance of assimilation. But the change of long ई and ऊ to long ए and ओ is due to softening alone; as in पेऊस for पीयूष, आमेल for आपीड, बहेडअ for विभीतक, केरिस for कीदृश, and एरिस for ईदृश, नेड for नीड, पेढ for पीठ, तम्बोल for ताम्बूल, तोणीर for तूणीर, थोर for स्थूल, and गलेई for गुड्ची. In the same way, इ and उ are in rare cases softened to अ, as in हलदा for हरिद्रा, पडंसुअ for प्रतिश्रुत्, and जहिढिल for युधिष्ठिर. अ requires no movement of the tongue or lips, while इ and उ do. It is therefore softer than those two vowels. Both these changes contradict another principle to be hereafter noticed, the operation of which is wide, and they must therefore be considered special or peculiar changes. The manner in which ऋ is softened has been already detailed.

c8.48

The semivowel य is often softened to इ, as in विअण for व्यजन, विलिश्र for व्यलीक, थीण for स्त्यान, उव्वीढ for उद्‌यूढ, जीआ for ज्या, &c. Here the effort necessary to bring the middle of the tongue closer to the palate is economized, while the position of the organs in other respects is the same. The य of the conjunct य is sometimes softened in this way to इ and sometimes to इअ. In the former case the resulting इ is transferred to the previous syllable and forms ए with the अ contained in it; as in पे्रन्त for पर्यन्त, अच्छेर for आश्चर्य, बम्भचेर for ब्रह्मचर्य, सुन्देर for सौन्दर्य, &c. Similarly व is changed to उ as in झुणी for ध्वनि, वीसुं for विश्वक्, गउओ for गवय, सुवइ for स्वपिति, दु for द्वि, &c.

c8.49

The surds are softened into sonants. The pronunciation of these requires, as you will remember, less exertion than that of the former. For sonants such as ग्, घ्, द्, ध्, &c., are pronounced by means of nāda, or vocal sound which is produced when the glottis is in its natural condition and the chords vibrate, and the surds क्, ख्, त्, थ्, &c. are uttered by sending forth simple breath or śvāsa, to produce which it is necessary to stretch the glottis. This effort is saved by changing the surds to sonants. Thus क् is changed to ग् as in मरगअ for मरकत and गेन्दुअ for कन्दुक; द् to द् as in नड, भड, and घड, for नट, भट, and घट, and in पडि for प्रति and the other instances given above; द् to द् as in मढ, सढ, कमढ, कुढार, and पढ, for मठ, शठ, कमठ, कुठार, and पठ; and त and थ to द and ध as in एद for एत, नाथ for नाथ &c. (in Śaur.). The cerebral, and द् resulting from the original dentals त and थ are also similarly softened to ड् and द् as in पडि for प्रति (पटि in Pâli), पहुडि for प्रभृति, बहेडअ for बिभीतक, हरडई for हरीतकी, मडअ for मृतक, &c., and सिदिल for शिथिल, पुढवी for पृथिवी. पढम for प्रथम, &c. The sonant answering to प् is व् but this is further softened to व्, as in कूव, कवाल, कुणव, तवइ, &c., for कूप कपाल, कुणप, तपति, &c.

c8.50

An original ड् is softened to ळ्, in which case there is a saving of two efforts. The complete contact necessary for the pronunciation of ड् being avoided we have ळ्. This sound is similar to ळ् in the fact that the breath before the break of the contact or close approach is allowed to pass out by the sides of the tongue, but is more difficult or harder because its position being higher up, the tongue has to be raised higher than in pronouncing ळ्, the position of which is near the root of the upper teeth. This additional exertion is saved by proceeding from ड्, at once to ळ्. We have thus गरुळ, तलाअ, कीळइ, दालिम, गुल, नाळी, for गरुड, तडाग, क्रीडति, दाडिम, गुड, नाडी, &c. In the Pâli, however, you will remember and probably in some unwritten Prâkṛits also, the cerebral character of the letter was preserved and we have ळ् instead of ळ्. Sometimes ड् is softened to र् only. The position of this is lower than that of ड् and the contact is incomplete. It is therefore softer than ड् but harder than ळ्. This last sound is produced lower still, but a light close approach is formed, and the breath is allowed to escape by the sides of the tongue; while in uttering र् the tongue is kept in its position with greater force, and the breath is allowed to vibrate. This change is principally to be met with in the case of the द् resulting from an original द् in such words as एआरह, बारह, तेरह, &c., for एकादश, द्वादश, त्रयोदश, &c., the द् of which must first have become ड्. In कलम्ब, दोहल, and पलित, for कदम्ब, दोहद, and प्रदीप, we have instances of an original द् passing into ळ्. An original र् is changed to ळ् which, as just observed, is softer than र्, as in हल्दी, दलिद, सुहल, इङ्गल, &c., for हरिद्रा, दरिद्र, सुखद, अङ्गार, &c. When the र् forming the first member of a conjunct is softened in this way, and the following member is weak, or is itself capable of becoming ळ् the whole conjunct becomes ळ्; as पर्यस्त becomes पल्लु-त्थ; पर्याण, पल्लाण; आर्द्र, अल्ल or ओल्ल; मर्द, मल्ल, &c. The sibilant स्, which is the only one we have in the Mahârâshṭrî and Śaurasenî, is sometimes softened to ह्, that is, the simple heavy breath somewhat compressed at the dental position is made uncompressed heavy vocal sound; as in दिवह, पाहाण, दह, एआरह, बारह &c., for दिवस, पाषाण, दश, एकादश द्वादश, &c.

c8.51.1

Another phenomenon characteristic of the Prâkrits is the dropping away of single uninitial consonants. Thus क् is dropped, as in सअद, कुम्भआर, for शकट, कुम्भकार, &c.; ग् as in नअ, नअर, साअर for नग, नगर, सागर, &c.; च् as in वअण, सूई, for वचन, सूचि, &c.; ज् as in रअअ, राआ, गअ, for रजत, राजा, गज &c.; त् as in विआण, पआवाई, गअ, for वितान, प्रजापति, गत, &c.; द् as in मअण, पाअ, कअल, for मदन, पाद, कदल, &c.; प् followed by any vowel except अ, as in रिऊ, कई, विउल, for रिपु, कपि, विपुल, &c.; यू as in वाउणा, णअण, for वायुना, नयन, &c.; व् as in जीअ, दिअह, for जीव, दिवस, &c. The vowel अ or आ that remains after the elision is when preceded by अ or आ pronounced like a light य. We find it written in Mss. also, as in सयढ, तित्थयरो, नयर, &c. In the same way, the mute element of the aspirates is dropped, except in the case of the palatals and of फ्, the हू or the heavy nâda or vocal sound being alone preserved. Thus we have मुह, लिह, सही, for मुख, लिख, सखी, &c., मेह, माह पाहुण, for मेघ, माघ, प्राघुण, &c.; नाह, मिहुण, कहा, for नाथ, मिथुन, कथा, &c.; बहिर, वहू, साहु, for बधिर, वधू, साधु, &c.; लाह, सोहग्ग, सोहा, for लाभ, सौभाग्य, शोभा, &c. The range of the operation of these processes is very wide; and the number of words transformed in accordance with them is very large. Now, the principle which is at the bottom of this elision of consonants is the same as that which brings about softening, viz., economy of exertion.

c8.51.2

But in a growing language which knows of no accidents in the course of its history, there is another principle which counteracts this, viz., the necessity of pronouncing words in a manner to be understood by others. Hence a wholesale elision is impossible; and the natural course is to soften sounds away, until finally they are elided. This of course must be a very slow process. But in the Prâkrit, and especially the Mahârâshtrî, we do not find it to be slow. Elision seems to be a distinguishing characteristic which strikes one very forcibly on reading a Prâkrit passage. It is too regular, systematic, and far-reaching, to have been the product of a long course of softening. In fact, because it is so systematic and general, some writers have doubted whether the Prâkrits were genuine dialects, and have expressed their belief that they are the creation of Pandits. But we shall find that the modern vernaculars retain the words shorn of their elements by the Prâkrits in this way; and that these latter dialects were the immediate parents of those we now speak in Northern India. If this general mutilation of words was brought about by a natural decay, we must suppose the process to have gone on for a great many centuries. The Prâkṛit vocables that have descended to the modern vernaculars have not, since the period when the Prâkṛits arose about fourteen centuries ago, suffered at all so much as Sanskrit words in passing into the Prâkṛits.

c8.51.3

We have not dropped away any of the elements of Prâkṛit vocables, though we may have re-arranged them in some cases. If, then, fourteen centuries have not been sufficient to make words lose any of their elements by the action of natural causes, the process by which Sanskrit words were mutilated and became Prâkṛit must, if it was natural and ordinary, have been in operation for a considerably longer period, a supposition which appears extremely improbable. And it is questionable whether a language which has not been exposed to accidents in the course of its history suffers so much, even after the lapse of any conceivable time. The Teutonic languages, though they have been going through an independent course of development since the period when the ancestors of the modern English and Germans separated from the ancestors of the Hindus in pre-historic ages, have preserved the elements of old Aryan vocables which the Prâkṛits lost. They may have transformed the words, but have not mutilated them to any extent. The English words *father*, *mother*, *brother*, *foot*, and others, have all the elements of the old पितृ, मातृ, भ्रातृ, पद् (पाद), and others; while the Prâkṛit पीआ or पीउ, माआ, माइ or माउ, भाआ or भाउ, and पाअ, and the modern vernacular पिउ, माउ, मा, मा, भाई, भाउ, and पा or पाय, have lost all except the first consonantal sound.

c8.51.4

The elision of consonants on a large scale in the Prâkṛits is therefore to be accounted for in another way than by attributing it to a natural process of decay. The Prâkṛit words must be taken to represent the pronunciation of the corresponding Sanskrit words by an alien race. The vocal organs of the people of that race were unused to the utterance of Sanskrit sounds, and in this respect they were in the condition of children making their first attempts at articulate speech, and elided uninitial consonantal sounds and assimilated conjunct consonants as these do. When a child or a barbarous foreigner listens to an elderly or more civilized person, he has not the patience to attend to all the sounds composing a word and to reproduce them carefully, or has not acquired the habit of doing so. The first letter only makes a strong impression on his ear, and this he faithfully reproduces; and as to the rest, he realizes their quantity by pronouncing the vowels, but his tongue being untrained, the peculiar movements necessary for uttering the consonantal sounds he cannot go through, and avoids. In going over the Pâli we had to attribute the assimilation of conjuncts and other peculiarities observable in that dialect to such a cause; and now the elision in the Prâkrits that we have been considering must, I believe, be accounted for in the same way.

c8.52

But what could be the historical events which led to the wholesale assimilation of consonants in the first place and to the similar elision of uninitial consonants in addition to the assimilation, in the next? I would put forward the following hypothesis. It would appear that when the Aryas moved from the land of the five rivers and settled in the country known afterwards as Brahmâvarta and Kurukshetra that is the country about Thânes'var, they formed a consolidated community in which an aboriginal or alien race was incorporated and the language represented by the Pâli was the language of this race. When, however, the consolidated community spread eastwards and southwards, they met other alien races and the dialect spoken by the lowest class of the consolidated community was learnt by these new alien races and the words which were in the Pâli stage were further corrupted into the forms which we find in the Prâkrits. The dialects of the As'oka edicts do not show instances of the Prâkrit rule of elision, though some peculiarities of one or two later dialects are discernable in them. Does this necessarily show that the Prâkrits had not developed in the time of As'oka? I think not. The dialects of the edicts represent the spoken language of the more cultured portion of the consolidated community and not that used by the races which were newly incorporated with the community and formed a class lower in the scale, though the Prâkrits must have begun to be developed about the time of As'oka. In the Nâsik cave inscriptions of about the second century of the Christian era we observe some instances of the Prâkrit peculiarity of elision, whence it would appear that the Prâkrits had then come into existence though they were still regarded as vulgar and were neglected, the principal inscriptions being in the Pâli which had then become a learned language. The conclusion we came to in the last lecture, that the Pâli represents middle Sanskrit in a vernacular form while the Prâkrits correspond to the last or classical stage, is in harmony with this view.

c8.53

Another process which transforms the words of a language is assimilation. We have already discussed the assimilation of the members of a conjunct and of the diphthongal sounds ऐ and औ, and will now consider the application of the process to the different syllable, composing a word. The अ of the second syllable of ईषत्, वेतस्, and कृपण, is changed to इ, because the vowel in the first is इ or ए, the ऋ of कृपण being changed to इ; and we have ईसि, वेडिस, and किविण. Similarly, the अ of मरिच and वल्ली, and the इ of इक्षु are changed to इ, ए, and उ through the influence of the following इ, ई, and उ; and the forms are मिरिअ, वेल्ली, and उच्छु. Sometimes, when य as a member of a conjunct is changed to a palatal, it influences the preceding vowel and brings it over to its own position, as in सेज्जा for शय्या, मज्झिम for मध्यम, गेज्झ for ग्राह्य, &c. An original palatal also produces similar effect, as in जिमइ or जेमइ for जमति. In the words पुढम for प्रथम, पोम्म for पद्म, and ओप्पेइ for अर्पयति, the adjoining vowel is changed to उ to make it go easily with the labial प. You will observe that it is अ that yields so readily to the influence of an adjoining letter, because in pronouncing it no movement of the tongue is necessary. It is sounded by letting off náda or intonated breath, without offering any obstruction to it; and this current gives rise to the different vowels when it is obstructed in various degrees by various movements of the tongue and the lips; so that if the organic position of the previous consonant remains a little longer, or if the following is begun a little earlier, the अ current assumes the form that that position gives it. An aspirate or an aspirated mute existing in a word influences an adjoining vowel or a consonant; that is, the heavy breath that is necessary for the utterance of the aspirate mingles itself with the other sounds also. Thus कासित becomes खासिअ ; परुष, फरुस ; परिघ, फलिह ; परिखा, फलिहा ; पनस, फणस ; and बिसिनी, भिसिणी, in which instances क्, प् and ब् are changed to ख्, फ्, and भ्. In other cases the heavy breath is transferred to an adjoining vowel which becomes ह्; as in फलिह for स्फटिक, निहस for निकष, विहत्थी for वितस्ति, वसही for वसति, भरह for भरत, in which cases by the dropping of the mute we should have, but for this heavy breath, फलिअ, निअस, विअत्थी, वसई, and भरअ. The words सटा and शकट become सढा and सअढ, the द् resulting from the transference of the heavy breath being softened to ढ.

First 55 of 147 passages. Read on

Linked texts (33)