Vedic Metre in its Historical Development c10: Chapter IV

Vedic Metre in its Historical Development c10

Language: English

c10.s1 132. Table Shewing the Growth of Sandhi Combination in the Rigveda — 10 parts

Chapter IV

115. IN classical Greek and Latin verse a final vowel is usually *elided* or omitted from calculation if the next word begins with a vowel. According to the rules of classical Sanskrit a vowel in such a position is regularly *combined* with the vowel following, so that the two syllables coalesce in one. In the Rigveda the vowels are sometimes combined, as in classical Sanskrit: but at other times each retains the character of a separate syllable, in which case *hiatus* is said to occur. Although hiatus is rarely shewn in the Saṃhitā text, it can be restored with certainty in a great number of passages by the aid of the metre.

c10.3

Thus in the verse vii 40 3d

c10.4

*ná tásya rāyáḥ pari-etá asti* (text *paryetásti*)

c10.5

it is clear from the metre that the final syllables of each part of the word *parietá* retain the character of distinct syllables.

c10.6

In the Rigveda final *-a* and *-ā* are usually combined with an initial vowel following, whatever it may be: but final *-i -ī -u -ū* are only combined when followed by ‘similar’ vowels. It is the object of the present chapter to determine the practice of the Rigveda in these points and others of the same character, the extent to which exceptions are permitted, and the historical importance of these exceptions.

c10.7

116. The discussion in this chapter is limited to the consideration of the treatment of final vowels and diphthongs, and of final syllables ending in *ḥ* and *m*, so far as syllabic combination is possible with the latter. By a final vowel is meant a vowel occurring at the end of a word or the first part of a compound, and so with the other final syllables concerned. The treatment of final consonants does not affect the metre except as included above, for there is no reason to think that any final consonant does not contribute to make position.

c10.8

The Sanskrit Grammar includes under the heading of Sandhi the treatment of the final vowels of stems: this subject is here reserved for the chapter on ‘Syllabic Restoration.’

c10.9

A final long vowel or diphthong, if not combined with a vowel following, is frequently reckoned as short: this subject is discussed in the chapter on ‘Quantitative Restoration.’

c10.10

117. The Saṃhitā text rarely admits hiatus, but follows the artificial rules of the later Sanskrit (Whitney, *Sanskrit Grammar*, § 103). As the metrical character of the Rigveda is almost entirely destroyed by this method, commentators both native and western claim a general permission to restore hiatus in all cases. The consideration of the metre however shews that the Vedic poets were guided by principles and that deviations from these principles are comparatively rare.

c10.11

Where the Saṃhitā text shews hiatus, it appears often to be guided by a true appreciation of the metre: but even so the text is practically of no value to us as evidence.

c10.12

In a few cases, which will be noticed in their order, the text uses diacritical signs to denote hiatus.

c10.13

118. There are many verses in which the restoration of hiatus is doubtful, either because the normal number of syllables can be obtained in more than one way, or because the verses in question may be interpreted as decasyllabic or defective. In all such instances it will be reasonable to give the verse that interpretation which brings it as a whole most nearly into agreement with the metrical character of the hymn to which it belongs and the general rules of Vedic prosody. Thus it is generally desirable to assign to each verse a normal number of syllables, but if in a particular hymn the number of shorter verses is considerable, as in the hymns described in § 94 iii as decasyllabic, it becomes more probable that the particular verse is also shorter. Where the required number of syllables can be obtained in more than one way, there is usually one method which is in better agreement with the rules of prosody than any other. In this way the number of cases which involve any serious doubt becomes very small: the consideration of these cases in detail is reserved for the ‘Metrical Commentary’: the interpretation implied in the lists included in the text is always that which appears to the writer on the whole the most probable.

c10.14

119. Each verse in the Rigveda is an independent metrical unit as regards Sandhi: a final vowel may therefore stand at the end of any verse, whether the next verse begins with a vowel or not, without hiatus arising.

c10.15

Within the verse hiatus is relatively more common at the caesura than elsewhere.

c10.16

The Saṃhitā text takes as its unit the *distich*, or pair of trimeter or dimeter verses. But the appearance of hiatus seems to be quite as common at the end of verses *a* and *c* as at the end of verses *b* and *d*, and very much commoner at the end of the verse than at the caesura.

c10.17

Similarly Sandhi combination is almost unknown at the end of the verse, but there are many occurrences at the caesura.

c10.18

Instances of exceptional hiatus or combination at the caesura are separately recorded in the lists that follow. Instances of combination at the caesura by ordinary rules are given in Ch. VIII below.

c10.19

The only probable case of Sandhi combination at the end of the verse is in *ix 113 7cd:

c10.20

*tásmin mám dhehi pavamān’ | āmṛte loká ákṣite*

c10.21

Here the two dimeter verses of epic Anuṣṭubh are regarded as forming one verse of 16 syllables. Somewhat similarly we find combination in viii 13 15cd (Dimeter Uṣṇih), but these verses may be regarded as forming a single trimeter verse.

c10.22

Other cases in which metrical difficulties would be removed by the hypothesis of Sandhi combination are i 110 9ab, *164 24cd, v 7 7cd, vi 27 5ab, ix 97 43bc.

c10.23

120. Final -a, -ā are regularly combined with an initial vowel or diphthong following: and final -i -ī -u -ū are regularly combined with similar vowels, that is -i or -ī with either -i or -ī, and -u or -ū with either -u or -ū.

c10.24

But many monosyllables are uncombined, especially ná ‘as,’ sá, and u or ū: and less regularly prá, á, smā and other monosyllables in -i or -u.

c10.25

Duals in -ā, -ī, -ū are also regularly uncombined.

c10.26

There appears to be no instance of the combination of the particle u or ū with a vowel following: the writing ūm iti in the Pada-pātha text seems to assert the principle that the long vowel is uncombinable. Similarly such forms as ó, pró, mó (where the Pada-pātha has ó iti, pró iti, mó iti) are never combined with a vowel following: it is occasionally necessary to restore for them á ū, prá ū, má ū, etc., with hiatus before ū.

c10.27

The combination of ná ‘as,’ and sá ‘he’ is comparatively rare, and the instances are therefore given as exceptions in § 122. The readings sáḥ, só, sá in the Saṃhitā text seem all to result from later theories: they are often used even when there is really combination, as in v 4 6d.

c10.28

In the cases of the other words in -a, -ā named above, and also of evá, it appears sufficient to note the instances of hiatus in the general lists, combination being on the whole more common. Hiatus is sometimes denoted in the Saṃhitā text by the sign ṁ, as saváyaṁ i 113 1c, śásadānāṁ i 123 10a, eváṁ v 6 10a.

c10.29

Duals in -ā usually appear in the text as ending in -āv, but -ā stands before u, and at the end of odd verses combination appears in the text.

c10.30

Duals in -ī, -ū are usually written in the Saṃhitā text with hiatus; in the Pada-pātha text as in the example prácī iti prácī, indicating that the final vowel is not capable of combination.

c10.31

Nevertheless there are many instances of the combination of duals in -ā, and some of duals in -ī; see § 128. Combination of duals in -ā is usually given correctly in the Saṃhitā text, but combination of duals in -ī is more often ignored, no doubt as a result of the later theory indicated by the Pada-pāṭha reading.

c10.32

Locatives in -ā are frequently found before consonants within the verse, but there are only four cases of combination with a following vowel, namely in vi 12 2b, *viii 91 1b, *x 101 11b (text -ā), and vii 39 3b (text -āv). The text also gives -ā before u- (Lanman, Noun-Inflection in the Veda, p. 386): but in these cases it would seem right to restore -āv rather than postulate hiatus.

c10.33

The compound which appears as saptarṣí in the Saṃhitā text has in the Rigveda the value saptarṣí, according to rule.

c10.34

It is evident that particles such as īm and íd are frequently used to prevent hiatus. Thus Professor H. Oldenberg points out (Prolegomena, p. 435, Note), that final vowels are combined with īm 59 times when a vowel follows, whilst there is only one occurrence in which a consonant follows. Again though evā is often used with hiatus, evā íd is never found with hiatus, but the combination evéd is found, most usually when a vowel follows. Whether these particles were so used by the poets themselves, or are due to reciters and editors, we have no means of determining, and it is therefore reasonable to accept the text as it stands.

c10.35

For the quantity of final vowels in such words as evā, adyá, see the chapter on ‘Quantitative Restoration.’

c10.36

121. Hiatus is found after -a in the following passages, combination being eight times as frequent:

c10.37

(a) in Dimeter verse: i 27 3a¹, 30 3b 9a, 40 3a² 8d³, 41 7b, 74 1b, 79 10b, 80 9d, 105 9c 10a, 120 5c², 127 3g³ 8e, 129 1c² 4d 5c 9e, 176 5c; ii 5 7c¹, 6 6a, 41 8a³; iii 13 1a, 41 8b, 51 12b², 53 13b; iv 31 13c; v 6 4b, 7 1b¹ 7a, 9 3b 5a, 16 1b, 17 1d, 18 1b, 25 8b, 33 7a, 35 8a, 50 2b¹, 51 5c, 65 5a, 67 1c 5a, 70 4a; vi 14 1d 5c, 16 3b 42b, 44 6c; vii 14 1d, 66 11b¹; viii 1 21a, 4 13a, 16 4a, 18 21a, 19 11a, 23 23a, 24 1b 25a, 27 7d⁴, 40 9d, 43 11c, 44 27c, 46 19b 28d², 52 9b, 61 3b bis 7d 14b, 62 1a², 66 9a 9b 12b, 67 4b, 68 17b⁴, 69 1b 8a² 9d 18a, 70 12a, 71 14a, 72 7a, 74 7d, 75 8b, 78 7b, 80 5a, 81 7a, 82 3a, 83 7a, 89 3b, *91 5c, 92 5a² 6b, 95 2c, 98 10a, 101 5a², 103 10b; ix *5 11b, 11 1b, 24 3b, 29 1a² 1b, 55 3b, 58 1b, 59 2b, 61 19b, 63 22a, 98 2b 5d, 101 13a; x 20 8a¹, 22 7b 8c, 26 4d¹ 8c, *59 10a, *85 41c¹, *86 2c³, *90 10b¹ *13c¹, 93 12d 13d⁴, *97 10c², *103 13a²⁴, 105 8b, 126 2b, 134 1b, *166 5e, 185 2b³. In verses of four syllables: viii 12 31d = 32d = 33d², 46 18e²; ix 102 8d².

c10.38

(b) in Trimeter verse, at the caesura: i 31 8c, 33 1a 13c, 52 10a, 61 1c, 62 3a, 77 1a, 89 8b, 91 22c, 114 2a, 116 18b, 117 1a 8b 11a 21a 22a, 119 9b, 120 4b, 128 4b, 133 7f, *170 5d, 173 12a, 174 7b, 177 5a, 180 2a, 181 9a, 190 5a; ii 4 8a, 11 12a, 14 5d 6b, 18 7a, 19 1a 7a; iii 32 4c; iv 4 1b, 5 1b, 16 2a, 23 7c⁵, 28 5b, 51 3b; v 33 3a 4b, 37 2c, 54 10d; vi 3 2b, 6 6c, 12 3a, 13 5d, 15 3a 3b, 16 46b, 20 1a 4c 4d³, 21 12d, 24 6b, 31 3a 4b, 33 2a 3a 4a 5a, 36 4c, 47 4d, 48 13a; vii 1 20c, 4 1b¹ 8a, 6 3d, 14 2c, 21 4a, 22 4b, 28 1a 4a, 32 10c, *33 10c, 41 6a, 52 2c, 90 1d, 98 3c; viii 1 3c 8c, 4 13c, 10 4c, 18 15c¹, 19 26a, 22 15b, 25 3c, 96 17c; ix 76 4c, 88 4d; x *10 13a, *12 3a *6a⁵ *8c⁵, *13 5c, 32 1d, 40 11d, *42 10a, 49 2b¹, 61 3d 14a 14c, 73 9a 10a, 74 4a, *88 12c, 92 7c, 99 10d, 115 5b, *120 7b, *124 5d.

c10.39

(c) in Trimeter verse elsewhere: i 57 3c, 61 5b 8b 8d³ 15c², 62 1d 3a¹ 4b, 63 5a 6a, 79 2a, 104 4a, 115 1b, 120 3c², 122 4d 13c⁴, 127 3f³, 129 6a, 141 12a⁴, 149 3c⁴, 155 1b¹, 167 1a, 168 3a⁴ 9c, 174 2c 5b 6a⁴ 7a, 180 2d, 186 10a², 190 6a²⁴; ii 4 6c⁴, 17 8b, 19 3b², 20 4d, 27 6c, 35 6a, 38 9a 9c³; iii 5 2a², 32 16b³; iv 2 18d, 3 11a, 6 2c, 16 7c² 20b, 19 7a², 21 7d², 38 5d¹, 40 3d; v 4 4a, 12 1a², 33 1b 10c, 41 3d² 5a⁴ 9b⁴ 12d 14b¹ 14c⁴ 19d, 45 9a⁴; vi 4 4b, 10 4b⁴, 18 5c, 20 1a 6c², 21 9a², 24 9b², 34 1d⁴, 37 2a² 2a, 47 2b, 49 7a⁴ 8c⁴, 63 2a 2d³ 3d⁶, 66 3b¹, 67 8c⁴, 72 3c²; vii 1 3a²⁴ 7c 22b⁴, 3 1d⁴, 5 1a², 13 1a², 22 1b 4a, 25 1a, *33 10c, 36 1a, 40 1d, 42 †6b, 46 1b⁴, 59 1c, 61 3a², 79 4d⁶, 87 1b², 97 9b; viii 4 4a, 19 23b¹ 35c, 20 18a¹, 22 2a⁴, 23 24c, 24 15c³, 70 10c, 96 10b²; ix 81 1a, 89 1a², 91 5b²⁴, 92 5a, 96 22a²; x *10 6a *13a³, *15 13a¹, *17 9c⁴, 22 11a⁴, 23 7b¹, 29 5a², *32 7b², 40 8d⁴, 49 7b², 50 1a 1b⁴ 6d 7d, *53 4c⁴, 61 24d¹ 25a, *88 9a *14c, 93 11a, *94 5a *8c, *95 10d², 96 9d, 105 6a² 6a⁴, *108 5a⁶, 115 4a, *121 4a *4c, 133 7c⁴, 148 5b.

c10.40

¹ ca. [^2]: prá. [^3]: ná negative. [^4]: the hiatus is at the end of the prior element of a compound. [^5]: átra or yátra. [^6]: augment with hiatus.

c10.41

122. Combination of ná ‘as,’ sá ‘he’ is rare.

c10.42

ná ‘as’ is found with hiatus about 60 times: with combination only in i 104 5b, x 46 5d, 106 7c, no one of the instances being quite certain.

c10.43

sá ‘he’ is found with hiatus about 150 times: with combination about 50 times, chiefly in hymns of the normal and cretic period, as follows:

c10.44

(a) in Dimeter verse: i 14 11c, 16 5a 9a, 26 1c; iii 10 3c; iv 8 4a, 37 6a; vii 15 6a; viii 43 9b; x *86 16a *16c *17a *17c, *97 23c.

c10.45

(b) in Trimeter verse: i 31 15d, 32 15c, 102 9c, 156 2d; ii 9 6a, 12 5a 5c, 13 2d 3d 4d, 15 6a, 17 6a, 18 2a, 22 1g 2g 3g, 24 1d, 35 7c 10b 13c, 37 2b; iii 4 2c 10c, 31 11a, 39 2d, 55 17b; iv 4 7a; v 4 6d; vii 1 14a 15a, 40 3a; viii 66 8c; ix 74 2c 7c; x 2 3c, *27 1a, 44 5c, *53 1a, 68 9a.

c10.46

123. Hiatus after -ā occurs in the following passages, combination being again eight times as frequent:

c10.47

(a) in Dimeter verse: i 8 1a¹, 30 17a¹ 21b¹, 38 6a², 120 6c¹, 127 6c¹ ³ 10g, 129 7e¹ 9b, 133 6c; ii 41 17b; iii 13 1b⁴; iv 52 2a, *57 7b; v 6 10a⁵, 25 9a⁵, *51 15c, 53 14b, 61 5a⁶, 75 2b, 86 6a⁵; vi 16 27a³ 42b¹¹, 46 5d¹; vii 59 5a¹; viii 1 23a¹, 5 29b, 7 33a¹, 9 13a⁷, 10 5a⁷, 16 7a, 17 1b⁴, 20 17c, 22 1a¹, 33 13a¹ 13d⁸ 15a⁷, 34 1a¹ 11b⁴, 46 10b 29b, 61 9b⁹ 17a⁷, 65 5c¹, 70 12d⁴, 98 4a¹; ix 29 6a¹, 108 14d¹; x 21 1a¹, 26 1d 9b, *60 7a, 105 4a 11b 11c, *145 3a.

c10.48

(b) in Trimeter verse, at the caesura: i 53 3c¹⁰ 5d 8b, 61 4b, 62 8c, 66 7a, 77 1b, 94 8c¹¹, 117 1b, 120 2b, 129 5a⁴, *162 13c, 167 10a⁷, 169 3c¹², 173 4a 11a¹², 174 1a, 178 2a, 180 7c¹² 8a¹², 183 2d, 184 2c³ 14, 186 6c, 190 3c; ii 20 3a 7a, 24 3b; iii 54 12b; iv 2 11d¹³, 10 1a⁷ 5a, 16 18b, 34 5a¹¹, 35 3b⁴, 40 2d; v 3 9d, 11 5b, 31 12c, 34 1b, 44 3d, 45 6b 7d, 46 4d, 51 11c, 53 9a; vi 13 3d, 20 3b, 21 9b¹³, 25 1b 7b, 36 5b⁸, 41 5a, 49 15c, 50 5b, 61 13b, 64 5a⁶, *75 18d; vii 3 7a, 21 6a, 34 5a¹¹, 60 1c, 70 4b 7a, 87 4b⁸, 93 7d 8b; viii 4 20c, 19 26c, 20 2c⁷ 24c¹¹, 35 2b, *48 14b, 96 20a 21a, *100 11c; ix 86 44b 45a, 97 4a¹¹, 107 2c, 109 1a⁴; x 1 7c¹⁴, *15 4c¹¹, 23 5b, 55 8d, 64 5a⁹, 73 7d, 74 6b, *95 16a, 99 3a, 105 4b, *117 7c, *121 2c, 132 4d 6a, 139 3b, 148 4a⁸.

c10.49

(c) in Trimeter verse elsewhere: i 48 7c 16c, 58 9c⁴, 60 4b, 61 16c¹, 62 8d, 69 9a, 77 1a⁶ 5a⁵, 87 4c, 104 1c¹⁰ 5a 6c², 108 4d¹, 113 14d¹, 120 6b, 127 6f¹², 130 1a¹, 133 6b, 140 13c, 173 4a, 174 8a, 185 4a, 186 6a, 189 4d²; ii 13 10a, 17 1b, 18 4c¹, 19 3d; iii 6 6c¹⁵; iv 2 18b⁸, 3 13b², 16 1a, 19 6c, 33 10a, 38 6a¹² 8a¹² 9a¹², 40 3a¹², 43 4b¹; v 29 15b, 41 16b³ 16c³, 45 2d; vi 4 4b, 12 5a¹², 19 9b¹, 20 8d, 21 8d¹³, 23 4a, 24 5d 9a, 48 17c⁹, 64 5a⁶, 66 3d⁶ 6c¹² 6d¹, 67 6d¹, *75 3d; vii 1 7a⁴, 4 6d², 6 7d¹, 8 1d¹, 22 6c², 32 14c, 34 1a 4a¹¹, 39 5a¹, 40 1a¹ 3d 4b 6a², 42 6a⁵, 69 3a, 71 3d¹; viii 1 16c, 70 12c⁴, 96 9d¹, 101 13c, 103 13a³; ix 71 6c¹; x 5 †5d, 30 2b¹⁴, 49 10b 11c, 50 2d⁹, 61 9c 14d 18c, 65 15b, *94 10b, *95 4a⁶, *101 3b¹¹ *7a¹¹, 115 7a⁵, *129 5d, 132 4d, 144 5b¹.

c10.50

¹ á. [^2]: 2 má. [^3]: 3 at end of the prior element of a compound. [^4]: 4 imperative in -ā (-a). [^5]: 5 evá (evá). [^6]: 6 ká, yá, sá. [^7]: 7 adyá (adyá). [^8]: 8 plural in -ā (-a) from stems in -an, -man or -van. [^9]: 9 vā. [^10]: 10 gerund in -yā (-ya). [^11]: 11 2 pers. plural in -tā (-ta). [^12]: 12 smā (sma). [^13]: 13 aorist imperative in -svā (-sva). [^14]: 14 áchā (ácha). [^15]: 15 áthā (átha). For the true value of the final vowels of variant quantity see below, §§ 158-163.

c10.51

124. After final -i, -ī, -u, -ū before similar vowels combination and hiatus are more evenly balanced. After monosyllables hiatus is regular, particularly after *ví* as in x 32 2a, and *hí* as in i 102 5d. So *nú* *ū* should probably be restored in i 64 15a, 120 2c, iv 16 21a, v 10 6a, 16 5a, 17 5a, 52 15a, 74 6c, vi 68 8a, vii 62 6a, 93 6c, 100 1a. Combination is rare; the examples are *viva* *vii 55 2c, *x 86 7d, *nìtaḥ* *x 161 2b, *vìdám* vi 9 6b, *hím* ii 30 3c.

c10.52

On the other hand longer words, and compounds beginning with *su-*, are generally combined. The exceptions are as follows:

c10.53

(a) in Dimeter verse: iv 47 2c, 57 1d; vi *16 47c, 44 5a; viii 21 3a, 24 22a, 46 32d, 47 1f¹ 2f¹ 3f¹ (in refrain), 49 6d, 53 3d, 62 5b, *91 5b, 93 27a; ix *5 8b, 13 7a, 21 1a, 40 4b.

c10.54

(b) in Trimeter verse, at the caesura: i 36 1c¹, 53 11a, 112 1d = 2d = 3d, 122 14d, 127 6b, 169 6a, 171 6a, 173 5a; ii 34 12b; iii 57 5b; *iv 18 4d; v 58 3a, *83 7d; vi 5 6a, 10 7a, 21 8a, 47 9c; vii 28 3a, *35 10b; ix 109 18a; x 50 3d, *88 4c, 110 3d.

c10.55

(c) in Trimeter verse elsewhere: i 52 7b, 70 5b¹, 134 5b; ii 17 8c; iii 55 12c; v 29 6c, 49 5b¹; vi 9 7c 7d; vii 1 3b¹, 29 3a¹, *49 4b, 58 6b¹, 65 1a¹, 66 12b¹, 68 9a¹; viii 96 19d; ix 97 49d; x 61 26b¹, 116 7d, *169 1a.

c10.56

¹ compounds beginning with su-.

c10.57

125. Before dissimilar vowels final -i -ī -u -ū are regularly used with hiatus: but disyllabic prepositions followed by the augmented tenses of the verbs to which they belong, as ádhy atiṣṭhan, are used freely in combination from the normal period onwards. Other combinations in which a disyllable is the first element are also fairly common, so that the phonetic type, not the grammatical connection, appears to be the essential condition.

c10.58

Disyllabic prepositions are combined with an augmented form of a verb following as below:

c10.59

(a) in Dimeter verse: i 11 6b, *23 23a, 49 2b, 80 9c, 135 6d, *191 5a; *v 40 9c; viii 69 11d; ix 26 2a, 47 1b, 99 4b, *113 3c; x *9 9a, *72 3c *5c, *85 38a, *90 1d *5c, *135 2c, *159 1d.

c10.60

(b) in Trimeter verse: i 32 8c 12b, 33 10b, 113 7a, 116 15d 17c, 117 8d, 118 7d, 124 3a, 146 4c, *161 12a, *163 2b *9d, *164 25b, 168 9d, 182 7b; ii 12 1b 11b, 17 2d 4b; iii 26 8d, 51 1b; *iv 18 3a *11a, *58 4b; v 11 6b, 30 12d *15b, 31 5d; vi 38 3b, 47 25b; vii 76 4c, 78 3a, *103 3a; viii *59 6d, 96 16c, *100 5c; ix 68 8b, 69 4c, 80 2a, 86 8c 17c, 110 6b; x 4 6b, 8 8b, *13 3a *5b, 43 6a, *51 2b, 68 8a, *71 3b, *81 4d, *109 5c, *121 6b *8a, 123 2d 6b, *139 4c, *157 5b.

c10.61

Words consisting of two short syllables are also found in combination as below:

c10.62

(a) in Dimeter verse: i 50 5a³ 5b³, 80 12c; v 19 1a, 52 6c, 73 3c; vii 81 1a; viii 7 24c, 43 24b¹; ix 6 5c¹, 13 1b¹, 17 3a¹, 45 4a 5b¹, *67 31a *32a, 101 16d, 106 11b¹; x *18 11d, *86 8d *11c *22c¹, *87 24a *25a, 118 8b, *119 5b, *141 1b³, *145 4a, *146 6b¹, *179 1d.

c10.63

(b) in Trimeter verse: i 34 1d¹, 79 5c¹, 116 12c², 118 3c, 120 8a, 122 12b, 139 9a², 156 2d, *161 8a, *162 2d *7c, *179 1d *3d; ii 3 1b³, 15 9a¹, 26 1b, 38 3c¹ 5d; iii 9 4c, 31 4c, 58 3c; iv 4 11b, *58 10a; v 28 1b³, 76 2d; vi 27 5b¹ 8c¹; vii *33 1a⁴, 65 3b¹, 83 8c¹, *104 2a; viii 1 34a; ix 72 3c, 80 3c³, 97 10c; x 1 4c, *27 6d, *52 3b, *53 6a, *87 15b³ *17d³, *88 13d¹ *16c³, 89 13a, *98 11d, *109 2c¹, *116 4c, 123 7b³, *124 4d, *128 1d¹ *3b *6c³, *129 7c¹, *130 7d¹, *142 1b, *157 5a³.

c10.64

¹ prior element of compound.

c10.65

² dadhyác.

c10.66

³ pratyác.

c10.67

⁴ śvityác.

c10.68

Combination of words other than those which consist of two short syllables is comparatively rare, except in the compounds ṛtvíj, gávyūti, sváhā and svid, in which it is regular.

c10.69

Combination is found exceptionally as follows:

c10.70

(a) in Dimeter verse: *i 191 3d (ní); v *51 14c and *15a (svastí); *vii 55 5d (sástu); viii 16 11b (svastí); ix 15 1a (yāti); x 134 1e 1f in refrain verses (janitrí), *146 4b (dáru).

c10.71

(b) in Trimeter verse: *ii 32 4c (śīvyatu); viii 51 4c (tú), 90 5c (haṃsi); ix 96 4b (svastáye); x *14 11d (svastí), 61 13a (nú), *95 4b (váṣṭi), 99 12c (svastím), 110 11c (pradíśirtásya, not pradíśy ṛtásya as in the text), *121 8c (déveṣu), *142 1b (asti). For i 110 9 see § 116 above.

c10.72

On the other hand suáhā occurs in iii 32 15a.

c10.73

126. Words ending in *-aḥ*, *-e*, *-o* lose their final element before an initial vowel following, and are therefore all alike treated as ending in *-a* with hiatus: and similarly words ending in *-ai* and *-au* are treated as ending in *-ā*. But duals both of nouns and of verbs ending in *-e*, and the locative *tué* (*tvé*), are unaltered.

c10.74

These rules are so far in general agreement with those of classical Sanskrit. The unalterable character of the dual endings, and of *tué* is recognised in the Pada-pāṭha text by the addition *iti*: as to *asmé yuṣmé* see § 170.

c10.75

127. According to the rules of classical Sanskrit, final *-aḥ* becomes *-o*, and final *-e* is unaltered, if the next word begins with *a-*, but the initial *a-* is lost. This result is comparatively rare in the Rigveda proper, but is occasionally found in all parts of it: in the popular Rigveda it is considerably more common. As many of the instances are found at the caesura, it seems that the loss of the initial vowel was not regarded by the poets as equivalent to combination of the syllables.

c10.76

After *-aḥ* and *-e* initial *a-* is lost as follows:

c10.77

(*a*) in Dimeter verse: i 79 11a¹, 81 1e; v 61 2a; vii 66 5c; viii 2 34b 40c, 27 22d, 50 5b; ix 9 4b, 59 4b; *x 72 4c, *85 17d, *90 3c *4b *12a, 126 3a, *127 1c, *145 6a¹, *161 5d¹, *166 3a *4d, *190 1b. At the beginning of a verse of four syllables viii 13 15d: see above § 119.

c10.78

(*b*) in Trimeter verse, at the caesura: i 59 2c, 70 4b, 118 7a¹, 168 9c, 186 8c, 190 3d; iii 59 2d; iv 1 12d; vii 1 19b¹, 61 3d; x 61 7c, 89 13d, *103 1c.

c10.79

(*c*) in Trimeter verse, elsewhere: i *24 8c¹, 30 16d, 33 13b, 51 3a 5a, 52 9d, 53 2c, 59 3b¹, 85 7a¹, 89 6c, 103 7b, 122 1a, 167 2a, 186 11a; ii 23 16b; *iii 29 3d¹; iv 16 18a, 55 1d; v 29 10b¹, 30 3d¹ 10a, 31 3d, 61 9a¹; vi 9 2b¹, 22 4d, 44 19b, 47 22b, 50 10c; vii 18 7c, *33 11b, 57 5c, 71 5a; viii *58 1c *3c¹, 96 20c; ix 86 23d; x 7 5c, 43 2c, 49 5c¹, 56 3d, 63 6c, 75 2a¹, 79 6c¹, 92 11b, *95 6d, 99 7c 7d, *108 5b, 116 6b.

c10.80

¹ forms in *-e*.

c10.81

The Saṃhitā text usually shews this combination correctly where it occurs, and also incorrectly at the end of odd verses.

c10.82

128. Duals in *-ā*, *-ī*, *-e* are very frequently combined with following *iva*: the instances are given in the next section. In addition the following examples of combination of duals are found, chiefly in the normal and cretic periods:

c10.83

(i) Combination of duals in *-ā*:

c10.84

(*a*) in Dimeter verse: i 47 9a; iii 12 4b; v 73 6d; viii 5 2c 16c, 8 6c, 22 5d, 31 9b, 35 22c, 38 2b; *x 85 9b, *141 4b.

c10.85

(b) in Trimeter verse: i 34 9d, 102 2c, 108 3d, 110 8d, 112 1d, 116 3a 8c 9a 10a 14b 19c, 117 1d 11d 12d 13d 18b, 118 9b, 119 2d, 152 3b, *161 7b, 182 4d, 183 2b 3c 5d; ii 16 5c, 31 1a, 39 8c; iv 2 4b, 21 9a, 44 3d 4b; v 43 8c, 46 3a, 62 5c 5d bis 6b, 63 1c, 76 1c 2b bis 3d 4b; vi 49 5c, 68 11b; vii 9 5c, 44 1a, 63 5d, 64 2c, 71 4c, 72 5a, 83 1d 2d 3d; viii 22 10c 12a, 57 4b, 86 3a; x 22 5a, 39 3b 10a, 41 2a, *52 2c, 64 5b, 73 3d, *87 3a, 106 11b, *125 1c *1d, *128 7c, 131 5a.

c10.86

(ii) Combination of duals in -ī with following i- is found in i 121 8a, 177 4d, iv 56 1a, vii 87 2c, 90 3a, the combination being correctly given in the text in the last two instances. Combination with dissimilar vowels following appears to be found also, namely prācī ’dhvará in iii 6 10c, and ródasī ’bhé (for ródasī ubhé) in i 33 9a, iii 34 1d, v 31 6c, vi 30 1d. There is no instance of the combination of a dual in -ū.

c10.87

(iii) Combination of duals in -e is very rare; there is perhaps an instance in vii 72 3c.

c10.88

(iv) In the dvandva duals as indrāgnī, náktoṣásā, combination is the more usual: but the following examples of hiatus are found:

c10.89

(a) in Dimeter verse: v 86 1a 4b 6a, vi 59 2b 5a 6a 8a 9a, 60 4c 5b 7a, viii 40 1a 3b.

c10.90

(b) in Trimeter verse, at the caesura: vi 60 13a.

c10.91

c10.92

129. The particle iva is so frequently found in combination with preceding words, without regard to the form of the ending, that the simplest explanation is the presence of the parallel form va, regularly used in Pali and in the verses of Buddhist Sanskrit. As such, the occurrences have already been taken account of amongst the forms which indicate late date. As however they may also be regarded as examples of a peculiar form of Sandhi, it will be well to give the list here.

c10.93

(i) Duals in -ā -ī -e are combined with iva following:

c10.94

(a) in Dimeter verse: i 28 7c¹; iv 32 23a²; v 64 1c, 74 9d, 86 5d; viii 5 21c, 34 9b, 73 4b.

c10.95

(b) in Trimeter verse: i 166 1c² 1d², 173 4d, 184 3a² 3d, 185 1d; ii 3 6b², 39 21 times; iii 6 10c, 30 4d², 58 2b; iv 41 5b 5c 7d; vi 8 3c¹; vii 69 6a, *104 6d¹; viii 26 15c, 35 7a 7b; ix 88 3b; x 4 6a, 40 2c 3a 3d 4a, *59 1b, 106 27 times.

c10.96

¹ duals in -ī.

c10.97

² duals in -e.

c10.98

(ii) Words ending in -aḥ, -āḥ, -iḥ, -īḥ, -uḥ, -eḥ, -aiḥ appear to be combined with iva following in i 141 11c, 175 6b; ii 6 7c, *43 2b; iii 36 6b; v 7 8b, 54 6b 6d; vi 46 4a, 48 18a, *75 14a; vii 56 8b; viii 19 14d, 35 5a, *91 3c; ix 96 15d; x 62 9b, 69 5c 6c, *84 2a, *97 10b, *146 2c, *149 4a *4c, *166 2b, *173 2b *2c.

c10.99

Words ending in -am, -im, -um appear to be combined with iva following in *i 97 8a; ii 13 4b; *iv 18 5a; vi 19 3c; vii 41 6d, *59 12c; viii 49 4d; x 49 6b, *149 1c.

c10.100

130. The occurrence of Sandhi combination in the Rigveda in positions in which it is not admissible by the rules of classical Sanskrit is proved in the case of duals (§ 128 above). In other cases in which it is suggested by the metre it is not always possible to determine whether the irregularity is in the metre or in the combination.

c10.101

Probable instances of irregular combination are:

c10.102

(i) -ah, -e with following vowel other than a: ii 20 8d, iii 43 5c, v 52 14c, vii 86 4d, viii 1 26c, 46 28c, x *85 40b, *87 15c, *90 4a, 93 10d, *166 5d.

c10.103

(ii) final -āh: i 177 4a, ii 20 2b, iii 30 21c, iv 34 3d, x *51 9a, 75 3a.

c10.104

(iii) final -ai, au: iii 32 6b, v 17 3a, vii 39 3b.

c10.105

(iv) final -am, -ām, -im: i 14 3b, *161 8a, iv *18 2a, v 46 2b, *51 15a, vi 17 7b, 48 4d, vii 19 5c, x 20 2a, 49 6d, *114 4a.

c10.106

131. The distribution of the principal variations from the Vedic rules of Sandhi combination is shewn in the table following. The whole number of variations is about 1200: the instances in which the rules are observed may be estimated at ten times that number. About one half of the variations are in favour of hiatus, and one half in favour of combination. Of those in favour of hiatus two-thirds are found in the bardic period; of those in favour of combination only a quarter are so found. This result is a substantial confirmation of the view that the hymns of the bardic period are earlier in date than the rest of the Rigveda.

c10.107

The history can be pursued in more detail by considering the number of occurrences of each kind in relation to the bulk of the groups: except that Group V contains so little trimeter verse, and Groups IV, VIII, and IX so little dimeter verse that no comparisons can usefully be made in these particulars. It then appears that hiatus is far more common in Groups I—III than elsewhere. In Groups IV and V hiatus is only half as common as before, but combination is unaltered. In Group IX irregular hiatus hardly exists; in 3000 trimeter verses there are only 20 examples, and of these 13 are at the caesura: combination of all kinds has greatly increased. In the popular Rigveda some reaction is to be noticed: hiatus is more common, and combination of sá, of duals, and with iva is rarer; but combination of final -i, -ī, -u, -ū, -ah, -e is much more common than in Group IX, and a close approximation to the standard of classical Sanskrit is reached.

c10.108

In trimeter verse the occurrences of hiatus at the caesura are equally common with the occurrences in all other positions in Groups I to VI: in Groups VII to IX they are twice as common.

132. Table Shewing the Growth of Sandhi Combination in the Rigveda

c10.s1.1

| Group | I | II | III | IV | V | VI | VII | VIII | IX | X |
| Bulk¹ | 2·3 | 3·5 | 3·8 | 2·7 | 1·7 | 6 | 3·3 | 2·8 | 3 | 4·7 |
| Exceptional hiatus | 76 | 143 | 157 | 72 | 34 | 84 | 28 | 33 | 22 | 63 |
| ,, combination | 16 | 38 | 41 | 27 | 15 | 78 | 43 | 30 | 113 | 134 |
| Per cent.² | 17 | 21 | 21 | 27 | 31 | 48 | 59 | 48 | 85 | 68 |
| Hiatus after -a³: | | | | | | | | | | |
| dimeter verse | 18 | 28 | 17 | — | 9 | 7 | 5 | — | — | 8 |
| trimeter at caesura | 9 | 10 | 11 | 5 | — | 2 | 4 | 2 | 3 | 3 |
| ,, elsewhere | 11 | 18 | 14 | 7 | — | 5 | 1 | 1·5 | 1·5 | 4 |
| Hiatus after ā⁴: | | | | | | | | | | |
| dimeter verse | 9 | 9 | 4 | — | 9 | 3 | 3 | — | — | 3 |
| trimeter at caesura | 7 | 10 | 9 | 5 | — | 4 | 2 | 2 | 1 | 2 |
| ,, elsewhere | 8 | 7 | 11 | 6 | — | 3 | 1 | 3 | ·5 | 2 |
| Hiatus after -i, etc.⁵ | 3 | 1 | 2 | 3 | 2 | 2 | 1 | 2 | 2 | 2 |
| Combination: | | | | | | | | | | |
| sá | 5 | 1 | 0 | 1 | 1 | 2·5 | 2 | 1 | 4 | 2 |
| -i, etc.⁶ | 3 | 3 | 2 | 3 | 2 | 5 | 4 | 2 | 7 | 18 |
| -ah, -e⁷ | 1 | 2 | 3·5 | 2 | 2 | 1·5 | 2 | 3 | 3 | 5 |
| duals | 2 | 2 | 1·5 | 3 | 3 | 2·5 | 4 | 4 | 4 | 2 |
| iva⁸ | 1 | 2 | 3 | 1 | 2 | 1 | 1 | 2 | 20 | 4 |

Linked texts (1)

Ṛgvedasaṃhitā — Sanskrit

Vedic Metre in its Historical Development c10: Chapter IV

Contents

Chapter IV

132. Table Shewing the Growth of Sandhi Combination in the Rigveda

Linked texts (1)