Part 6a - Phonetics and Lexicography—The Varnapatala

This section relates to two Parisistas, namely, the Varnapatala (Atharvaveda-Parisistas 47) and the Kautsavya-nighantu (Atharvaveda-Parisistas 48). The Varnapatala deals with phonetics and the Kautsavyanirukta-nighantu is a small lexicographical work more or less like Yaska's famous work. As has been already noted elsewhere, the Caturadhyayika and the Atharva-Pratisakhya do not contain the treatment of the varnas which they take for granted. The purpose of the Varnapatala is to supply this deficiency. It divides the varnas according to the quality and quantity of sounds and efines their phonetic value. The Varnapatala begins by stating that it is going to plain the varnas, original (prakrta) as well as modified ikrta). It goes on to point out that there is a difference opinion regarding the organs of the production of the ds and that, therefore, it is going to explain the same. d vrtti is, according to it, used in the sense of the aranam) of production of the varnas; just as, for language samanaksa sparsas (mu ar is the organ of hearing (1.2). he phonemes (i.e. stationary sounds) in a be divided into only five kinds, namely, onophthongs), sandhyaksaras (diphthongs), antahsthas (semi-vowels), and usmans 1. Cf. Cf. Caturadhyayika 1. (London 195 W.S.Allen, Phonetics in Ancient India .23-24.

656 (fricatives) (1.3). The Varnapatala states further that there are other kinds of sounds also; but they do not exist independently. They are heard between two phonemes and hence cannot be separated. Such sounds can be grasped by the mind only and those should be understood by direct and oral instruction (1.5). Such are the small fleeting marginal sounds 3 (known as glides) like nasalisation or yamas. The yamas are not exactly nose sounds and they should be known by tradition only (1.6). The Varnapatala accepts the traditional enumeration of the varnas (given in the matrka), and goes on to divide the vowels and the consonants into their respective classes. It is stated that the first fourteen vowels are called samanaksaras (Monophthongs)" and the rest of them are sandhyaksaras (diphthongs) 5 (1.7). The vowels except a are called namin (alterant vowels, because they tend to produce 6 2. Atharvaveda-Parisistas 47.1.5: manasira napi sangrahyanupadena vartayet | 4 3. The word is explained towards the end of this section. They are so called because they consist of a single vowel which may be hrasva, dirgha or pluta. The 14 vowels are: a, a, a 3; 1, 1, 13; u, i, 3; r, f, 3; and 13. The Rk-Pratisakhya states that the first ten (not including the four pluta vowels) are samana and the remaining are samdhi. 5. They are so called because they consist of two vowels. They are 8 in mumber, namely, e, e 3; ai, ai 3; o, 03; au, au 3. 6. Cf. Caturadhyayika 2.29 and Whitney's note thereon (p.89). The Rk-Pratisakhya (1.17,20) uses the same term. The Vajasaneya Pratisakhya uses the term bhavin

657 the nati or conversion i. e. the cerebralisation of the following a to s). The phonemes beginning with k are called vyanjana (consonants). The first twenty-five consonants (from k to m) are called sparsas (1.e. mutes, as in their 7 case the tongue comes into contact with the respective organs of production). They are divided into five vargas (ka, ca, ta, ta and pa) of five consonants each (1.8). The four consonants beginning with y (i.e. y, r, 1 and v) are called antahsthas (semi-vowels) and those beginning with g, that is to say,,, s, h as well as visarga, anusvara upadhmaniya (i.e. hp) and jih vamuliya (i.e. hk) are called usman 8 (fricatives). The semi vowels and the first six fricatives (i.e. those excluding upadhmaniya and jihvamuliya) are called ayogavaha (as they are independent). Yama (twin-sound), nasikya (nasalised h sound) and abhinidhana (implosion) complete 9 the list of 65 varmas. It is said that only these varnas occur in the Vedas as also in common usage (1.8-11). 7. Cf. Atharvavediya Manduki siksa, 6.8-9: aufT-IT tu prayogesu karanam syaccaturvidham | samvrtam vivrtam caiva sprstamasprstameva || sparsini karanam sprstamantah sthanamto'nyatha | samana samvrtam prahurvivrtam ca svaromanam || 8. The Rk-Pratisakhya (1.12) agrees in stating that the anusvara is usman The Caturadhyayika ignores any such constituent of the varmas as the anusvara, acknowledging, only nasal consonants and vowels. For an exhaustive statement regarding nasal sounds, see Roth, Zur Litteratur und Geschichte des Veda, pp.68-82. 9. These are given towards the end of this sub-section.

658 It is further stated that those phonemes which are pronounced through the mouth and the nose together, are called anunasika; and those phonemes which require equal effort of 10 the mouth are called savarna (1.12). When a is mentioned without any other syllable following it, it includes the following savarna syllable, that is, a (Similar is the case with 1, y etc.) 11. Each varga is known by its initial consonant and when that initial consonant is mentioned, it should be taken to signify all the five consonants in that varga. But when a syllable is followed by kara, it should be understood to denote that one syllable only. Thus, for instance, kakara means ka only (1.13-15). The ghosavat (voiced) consonants are the semi-vowels (y, r, 1, y), h, the latter two yamas (belonging to the sonants, which are noted below), and the three consonants at 12 the end of each varga. The remaining sounds, namely, s, §, E, the first two yamas (belonging to the surds), the first 13 and second consonants of each varga as well as anusvara, visarga, upadhmaniya and jihvamuliya are called aghoga (unvoiced) 14 (1.16-17). 10. Samanasya-prayatna ye te savarna iti smrtah. The Siddhanta-Kaumudi of Bhattoji Diksita gives (in the Samjna prakarana) a similar definition. The verse (1.12) is quoted in the comm. on Caturadhyayika 1.27. 11. 12. 13. 14. Cf. Panini’s Astadhyayi audit savarnasya ca pratyayah. 1.e. g, gh, n; j, jh, n; d, dh, n; d, dh, n; b, bh, m. These are k, kh; e, ch; , th; t, th; p, ph. Cf. Caturadhyayika 1.12-13. Cf.

659 16 Then the Varnapatala turns to an important topic, namely, the organs of production of the phonemes. It states that the throat (kantha) is the place of production for a, h and visarga. 15 It is the place of production for ai and au also, but only for their first matras. The second matras are produced from the palate (talu) and the lips (ostha) respectively. The yamas and the last consonant in each varga (i.e. the anunasikas) are produced from two places (namely, the nose and the respective organ of the varga to which they belong). The root of the tongue (jihva-mula) is the place of production for r, and kavarga (1.18-20). The palate (talu) is the place r 17, 1 of production for 1, e, cavarga, y and 18; and the cerebrum (murdhan) is the place of production for tavarga and §. 19 The 15. Cf. Rk-Pratisakhya 1.8; Vajasaneya Pratisakhya 1.71; Taittiriya Pratisakhya 2.46-48; Caturadhyayika 1.19; Pan I.1,9. 16. Cf. The comm. on Caturadhyayika 1.25: 17. Cf. 18. sandhyataresu varnesu varnantosthyamucyate | upadhmaniyakaro vah pavargastatha matah || Prat. 1.8; Vajasaneya Pratisakhya 1.65; Taittiriya Pratisakhya 2.35,44. The comm. on Caturadhyayika 1.20 quotes the verse Atharvaveda-Parisistas 47.1.20. It may be pointed out that the vowel is included among the cerebrals in the Paninian system. Cf. Prat 1.9; Vajasaneya Pratisakhya 1.66,79; Taittiriya Pratisakhya 2.36,40,44. It may be noted in this connection that the consonant ih does not occur even once in the Atharvaveda, Saunakiya recension 19. Whitney (op. cit. p.24) thinks that these syllables should be called 'linguals' and not 'cerebrals', because the word 'cerebral does injustice to the Hindu grammarians and obtrudes offensively a false and absurd theory. He has pointed out there that Max Muller proposed 'cacuminal' as a name for this class, but it did not find acceptance, perhaps as being rather cacophonus. Lanman has suggested that they should be called 'domals'. Caturadhyayika 1.23 points out that while uttering the tongue becomes retroflex or trough-shaped (dronika).

660 teeth (danta) constitute the place of production for tavarga, 120 and 3; and the lips (ostha) form the place of production for u, o, pavarga, y and upadhmaniya. 21 The nose (nasika) is the place of production for nasikya, anusvara, yamas and 22 anunasikas; and the root of the (upper) teeth (dantamula) 23 is the place of production for g or it may be taken to be slightly back (to a part which is called Alveolum). Other opinions regarding the place of production of gare given: According to some, it is produced from the root of the jaw (hamumula), 24 while others hold that it is produced from the cerebrum 25 (2.1-5). The Varmapatala then turns to another important topic, namely, that relating to the manner in which sounds are produced. It says: "The impulse (ghosa) produced from the chest is struck in the throat and the sound (nada) 26 produced therefrom is modified (in the various parts of the mouth) and is called a syllable (aksara)." This is quite a scientific explanation as it denotes a change from ghosa to nada and from 27 nada to varna. The Parisista adds some more details when 20. CF. Rk-Pratisakhya 1.9-10; Taittiriya Pratisakhya 2.38,42.44. 21. Cf. Rk-Pratisakhya 1.10; Vajasaneya Pratisakhya 1.70,80,81; Taittiriya Pratisakhya 2.43. 22. Cf. Rk-Pratisakhya 1.10; Vajasaneya Pratisakhya 1.74,80,82; Taittiriya Pratisakhya 2.49-51. 23. Cf. Rk-Pratisakhya 1.9-10; Vajasaneya Pratisakhya 1.68,77; Taittiriya Pratisakhya 2.38,41,42. 24 It may be pointed out here that the pronunciation of one of the r sounds in the Dravidian system is made from the hanumula. 25. The consonant is included among the cerebrals in the Paninian system. 26. Cf. Allen, op. cit., p.38. 27. ibid pp. 13-16.

661 it suggests that each syllable contains components such as the fundamental note, the overtone and the like. It points out that the first little component (anu-matra) of every syllable passes through the throat and then the niceties are added to it in the mouth. These niceties are called mukhavikrtis. Scientifically speaking, the syllables are, thus, a composite production from more than one part of the mouth, but it is for the sake of convenient common usage that we speak of them as being produced from one part of the mouth or the other. Then it is stated that when consonants come together without being separated (by vowels), they form a conjunct (samyoga). 28 It is pointed out that the second syllable from each varga is to be understood as usman (aspirate), the third as ghosa (sonant) and the fourth as ghososma (sonant aspirate). The fifth varna i.e. anunasika (nasal), the nasikya and the antahsthas (semi-vowels) are to be understood as ghosa and h is to be known as ghososma (2.9-10). As regards the vowel r it is said that there is a doubt regarding its component parts. The Varnapatala points out that one part of it is a and the second part, which precedes or follows it, is 29 (3.1-2).30 The vowels a, i, u and r 28. Cf. Rk-Pratisakhya 1.7; Vajasaneya Pratisakhya 1.48; Caturadhyayika 1.49,98. 29. Cf. Prat. 13.4; Vajasaneya Pratisakhya 4.145; Taittiriya Pratisakhya 2.18; Caturadhyayika 1.37. The comm. on Caturadhyayika 1.37 quotes the following verse:

662 are all hrasva, dirgha and pluta; whereas the vowel is only hrasva and pluta but not dirgha. The diphthongs e, ai, Q and au are only dirgha and pluta but not hrasva. It is also stateed that the udattas, anudattas, svaritas and kampitas are both amunasika as well as suddha (3.3-5). At the end, the Varnapatala states that the total number of varnas is 65. They include 22 vowels, consisting of 14 monophthongs and 8 diphthongs, out of which 19 (that is, all excepting a, a and 3) are alternant vowels; and 43 consonants, out of which 25 are mutes, 4 are semi-vowels, 4 and yamas, 8 usmans and the remaining two are nasikya 31 abhinidhana. 32 It is pointed out that the ayogavahas are ten in number (namely, semi-vowels and six fricatives excepting upadhmaniya and jihvamuliya). 33 The contents of the Varnapatala, as stated above, will (f.n. 29 continued) *varne svaramatra ya tasya madhye'rdhamatrya | repheा bhavati samsprsta yathavyanam tatha || r Thus a Prat (13.14) points out that in the case of Cf. Caturadhyayika 1.38. Caturadhyayika 1.39. and ris found only in the first mora. Similar is the case with 1 and 13. Cf. 30. The verse 3.1 which is mutitated and incomplete appears to suggest that the svarita accent consists of two components, namely, anudatta and udatta. 31. Cf. Allen, op. cit., pp.39,77. 32. Cf. Caturadhyayika 1.43,44. 33. Cf. Mahabhagya I.1.2; Allen, op. cit., pp. 16-17.

663 show that the work was written at a time when the phonetical science had developed considerably. They also show that this science, together with additional information, was transmitted through oral tradition, as is seen from the repeated reference, namely, upadesena vartayet (1.5,6, 2.10). The Varnapatala takes delight in the subtleties of phonetic analysis, such as yamas (twin sounds), abhinidhana (separated mutes) and nasikya (nasalised h sound). Yamas are produced when a nonnasal mute is folled by a nasal. 34 They are essentially 35 nose-sound st which are the counterparts of the previous mutes. Thus, for example, in a word like grbhnami (Atharvaveda, Saunakiya recension III. 8.6), the latter part of the consonant bh gets nasalised before the following na is pronounced. This twin sound is called yama. The abhinidhana is produced when one mute is followed by another mute. 36 In such a case, there is a brief emission of unintonated breath which forms a sort of barrier between the two sounds and thus separates them. Thus, in a word like marudbhih (Atharvaveda, Saunakiya recension II.29.4), the consonant d is produced separated 34. Cf. Caturadhyayika 1.99. The Manduki Siksa states: 35. sparsanamuttamaih syah samyamacedanukramat | anupurvya bharata janiyaccaturastatha || 11 .2 sakmeti prathamam vidyannrtetyaparam viduh | trtiyam padmamityahuh sankhadhmamiti vattamam || 19 .2 varsanta yatra drsyante tapaseh saha samyutah | yamastatra nivartante ramasanadiva bandhavah || 11 .5 Cf. Taittiriya Pratisakhya 21.12-13; Vajasaneya Pratisakhya 1.74,82. 36. Cf. Rk-Pratisakhya 6.5; Caturadhyayika 1.43-44.

664 from the following bh by a short gap. This is styled as abhinidhana. The nasikya is produced when h is followed by a nasal mute. 37 This in a word like brahma (Atharvaveda, Saunakiya recension I.10.4), a part of the consonont h takes the position of m, thus forming some kind of a grunt. This is treated as a separate element by the Hindu phonetics and is given the name nasikya. As regards the number of yamas, there is a difference of opinion. The Rk-Pratisakhya (6.8-10) seems to hold the number to be 20, one for each non-nasal mute, but its commentary says that there are only four yamas - one for all the first mutes of the vargas, one for all the seconds, one for the thirds, and one for the fourths. The commentary to Taittiriya Pratisakhya 21.12 supports the same view and the Varnapatala (2.6) also has clearly stated that the yamas are four in number. But Whitney 38 feels that this is indefensible. He remarks: "I cannot at all see how the nasal counterpart of the tenues of the five mute series should be identical with one another; nor, on the other hand, how they should be physically different from the yamas of the following mutes of each series respectively; although it might well enough be loosely said, considering their title of "twins", that there are as many of them as of the sounds to which they sustain that relation." On the whole it can be said that the Varnapatala gives, in a nutshell, the classification of sounds, the organs of 37. Cf. Rk-Pratisakhya 1.10; Vajasaneya Pratisakhya 1.74, 80; 8.28; Taittiriya Pratisakhya 21.14; Caturadhyayika 1.100. 38. The Saunakiva Caturadhyayika (1.99), p.65. 130

665 their production, and their total number. As has been already noted with regard to the places of production of I and k, the Varnapatala is nearer to the Pratisakhyas than to Panini. The statement relating to the production of sound, namely, that ghosa is changed into nada in the throat and nada is modified into a varna in the mouth, as also the statement that a sound is a composite function, are quite scientific. Though the Varnapatala is thus generally scientific in its contents and outlook, some of the statements made in it are not quite clear. Often they are clumsy, curt, confusing and loosely worded. Neither the eight usmans nor the ten ayogavahas are clearly enumerated. A number of terms like yama, abhinidhana and nasikya are taken for granted, as being already known from the Caturadhyayika , and several other important points have been only briefly referred to in this Parisista. Incidentally it may be pointed out that the verse 1.4 is corrupt and is clearly an interpolation.