son of Vennam-ayya Naga Varma.

Nagavarma's Canarese prosody online

. (page 1 of 20)
Online Libraryson of Vennam-ayya Naga VarmaNagavarma's Canarese prosody → online text (page 1 of 20)
Font size
QR-code for this ebook





Rev. F. Kittel

B. M. S.



_LONDON 1875






A. Nagavarma's Prosody".



The present edition of Nagavai-ma's Prosody is based on a collation
of the following? Manuscripts : —

B. On paper, from Bettigori near Gadftgu, to the oast of DhSravadn, — a very defoctivo

D. A lithograph from Dharavada (1865), belonging to the recension of B., but not to the

same original.
II. A copy on paper from a MS. at Hubbnlli near Dharavada. It forms a recension of

its own. It uses the old letter r (£).
K. On olo (palmyra leaf), belonging to the Raja's library at Maisflr. Mr. Raiigaoarya,

Controller of the Maharaja's palace, has been kind enough to send us its latter part for

collation. See Additions. The said part belongs to tho recension of M., but seemingly

not to the same original.
L. On ole, from the Lingaita matha at Madevapura in Kodagu (Coorg); a very incorrect

MS., with a Commentary still more so. It apparently represents the same recension as 8b.
M. On paper. It has been kindly lent to us by Mr. B. Mallappa, Head of the Canarese

department of the Maharaja's school, Maisdr. It forms a recension of its own, and

uses tho letter r.
0. One of tho numerous fragments of recent date that are called Nagavarma Chandas.

It is a collection of twenty-seven verses obtained at Madikeri (Mcrcara).
With the following four MSS. on paper B. L. Rice Esq., Bengalflr, has been kind enough
to favour us: —

Ra. It belongs to tho recension of H., and has, no doubt, been copied from tho same original.

It bears the Jaina heading "srivttaragaya namaV.
Rb. This interesting Jaina MS., though called a Prosody of Nagavarma, shows no internal

evidence whatever of belonging to him. Beginning with Pratishtho it gives 63 instances

of 22 chandas', the rules in verse, which at the same time form the instances, containing

praises of the twenty-four known Jaina Tirthaiikaras, from ijishabhasvami to <Sriv!ra.

Cf. e. g. Ind. Ant. ii., 134 seq.
Ro. A recension of its own. It begins, for instance, with v. 3 of B. or v. 6 of M., an invocation

of SarasvatI; its v. 2 is D.'s last verse, H.'s v. 3, and M.'s v. 30; H.'s v. 6 is not in

it; etc., etc.
Rd. A fragment of, or rather miscellaneous collections from Nagavarma, with a few

additions of its own; fifty-seven verses in all.
The following three ole MSS. have been kindly put at our disposal by Mr. Tirumalo
Syamanna, Muushi of the Wesleyan Missionaries at MaisAr: —

1) Here and at other places where a reference to the text is made, every number to which

the letter p. is not specially prefixed, refers to a verso.
2 —— S



Sa. An independent fragment, stating its contents as follows: Kagayarma chandas |
kovitMakshana | slintpadikrnraa | shadTrittancma (our t. 230) | Bhatpriisabage (bagc=
rtti) I shatpratyaya I ganaprastara Iganotpattisthnna | ganadcvatc | ganaprayogaphala II .

Sb. Its recension resembles that of M., (K., L.), but is not identical with it. The editor,
from fear less the olcs might bo damaged during the very damp monsoons at Mercara,
returned it to its owner before annotating nil its different readings; but nothing
essential, ho trusts, has been overlooked.

So. This the editor has used only for NSgavarma's Genealogy. From tha circumstance
of its containing the genealogy it follows that it somehow belongs to tho recension
of M. (K., L., and Sb.)

From this list it appears that, not taking into account the fragments
and MS. Rb., at least four different recensions are represented
hy the MSS., H., which, as will be seen, is probably the oldest; then M.;
thereafter Re; and lastly B. The present edition is a collection of the
essential portion of all the MSS., with a few additions concerning the
Ragales. Of the very numerous different readings, so far as they
are quite unimportant, such have been adopted as are metrically right;
wherever necessary, different readings are adduced. As the present
edition is first of all intended for schools, some indecorous epithets of
Nagavarma's wife, to whom the instruction is given, and the superstitious
first line of v. 22 have been altered. Such alterations appear in difterent
type, as do also all editorial additions in Canai-ese'), and some spurious
verses of recension B.; other spurious verses e.g. 318, 322, 326, 330, 334
and 337, are given in the letters of the text, as they have obtained a
certain popularity, and others, as they are valuable. An alteration, which
is not marked in print, regarding the genuine terms of Naki, Nakiga
and Pinaki, and which further on will be treated of in particular, has
been introduced chiefly to avoid perverse discussions at school; Pingala
or another similar word has been put instead. It is to be regretted that,
through renumbering the verses of the MS. at the printing office, many of
the numerical references in E n g 1 i s h ha vc become incorrect ; a list
of them appears among the corrections. Thelndcx contains all the terms
of any interest.

What Dr. Weber says^) with regard to Kedara's Yrittaratnakara, a
prosody in Sanscrit, viz. that its great popularity becomes already satis-
factorily evident from the floating condition of its text, holds good also with
regard to Nagavarma's Prosody ; it is in fact t h e o n 1 y K a n n a d a Ch a n d a s.

1) Conoorning them the wording of tho last line of v. 45 and that of tho instance of the
BUa (270, 271) have been somewhat changed. 2) Ubor die Motrik dor Inder, Indisohe

Studion, viii., p. 206.

n—, 1 8s



Nagavarma's original work did not contain six Sliatpadis, but
only one, viz. the Sarashatpadi, as v. 340, in which he states that ho has
explained the chief uses of the devaks haras or of the feet formed of
syllables of the devas. This shows that he had no knowledge of the other
Shatpadis, for none of these are built on the devaksharas, whereas the
.Sarashatpadi originally was a devakshara metre. Nagavarma, on that
account, put this his Shatpadi between the Ele and Akkarike. Verses
318-338 of the text, therefore, so far as they are declared to be by NAga-
varraa, are forgeries'*. The same is to be said of verses 313 and 31G
that are introductory to them. Here recension H., by adducing v. 316
(the only verse in which it alludes to the six Shatpadis) after its ch. G in an
Appendix, clearly proves that they did not belong to the original text;
and the indisputable text of H. (chapters 1-5), on this account, must have
preceded those of the other recensions that adduce the spurious verses
in the course of their texts. The whole recension of H. (chapters 1-G and
Appendix) appears to have been formed about 1300 A. D.'-) All that is
contained in the Appendix of H., which comprises 27 miscellaneous verses (p.
130) but is not counted as a chapter, is not genuine; and H. fitly concludes
its sixth or last chapter with v. 347 of the present text. In course of
time a number of the said miscellaneous verses (and various others besides)
were introduced into the text of other recensions, e.i/. 14 of them
that chiefly bear a superstitious character, into the first chapter of M.
(p. 12, note 1; p. 130). It appears that no superstitious matter
occurred in Nagavarma's own composition, though three verses about the
subha and asubha of the syllable-feet appear in the sixth chapter of H.^'; for
obviously on account of their having never belonged to the established text,
the first of them occurs with some alterations, as M. i., 41 ; and the other
two are identical with M. i., 40, 44 (our v. 3G)'). So the 6th chapter
of H. too, as it stands, is at least pai'tly of a questionable character.

It is very remarkable that of the rules on Alliterations (42, 43,
50-55, 57, 59-61, 63 and 64) that are so essential to Canarese, verse 42
is found only in Sa., and v. 43 in the spurious supplement of H. and

1) It ia strange that only three of the later Shatpadi instances (328, 335, 337) observe the
rule (313-315) that each third line has a iong syllable at its end, the others presenting at
that plaoc a short syllable that is to be looked upon us long; in these and in these alone the
rule about the end of a Hemistich (27) seems to have been put into practice. 2) This

was about the time when the later Shatpadis came into general use. 3) They are followed

by two other verses (our 236=M. i., 76, and our 230=M. i., 69) that cannot be genuine;
and then comes v. 347. 4) Regarding the state of M.'s first chapter further compare

e. g. p. 24, note 3.


in B., M., Re, whereas the other rules are only in the recensions of
B. and Re; from which circumstance it follows first that Re. is com-
paratively late, and secondly that NAgavarma did not include the subject
of alliterations in his prosody, although he always made use of the simple,
here and there of the co-ordinate ones (of. p. 21, note 2). It is worthy
of notice that the Lalitapada (v. 183) which presents the final alliter-
ation, is repeated in the supplement of H. (v. 11), and there bears the
heading "Caudanalakshana" (Ra. Caudalakshana), as if it were a Catush-
pada (cau=catush).

The recension of B. and that of the fragment Sa. are most
probably not yet one hundred years old as there occurs in it a quotation from
the Canarese Jaimini Bhai-ata (p. 125, note 1) which work belongs to about
A. D. 1760. They have apparently used the Kavijihvabandhana (on poetical
composition, etc.)'), Re.-' and M.-')

The verses (37, 38) on the Refrain (B., Sa.) and the Verse-lines
(B., M.) are certainly spurious*). The genuineness of a few other verses

1) See 34. p. 12. 42. p. 16. 65. 2) See 50-57. 59-61. 63. 64. 3) See e. g.

vs. 16-21. Vs. 286 and 316 the Kavijihvabandhana seems to have taken from one of the
recensions. A remarkable instance with regard to the carefulness bestowed on recension U. is
evinced by v. 29 being repeated in its supplement with a slight difference at its beginning, viz.
ScriOi^o instead of ^oSjOo; ScriJ^o is the reading of M. (i., G3). The insignificant scholarship
displayed in the recension of B. appears, for instance, from its verses regarding the form-
ation of the Kanda. Everybody who takes the trouble of scanning Canarese Kanda
verses, will find that the 6th foot of each Hemistich is to be either an Amphibrachys (w— w)
or a Proceleusmatious {■^^■^■J). But the said recension whilst in no less than three verses
(283, 287 and another not contained in the text) allowing the use of the Amphibrachys for the
6th place, does not even allude to the essential rule that a Proceleusmatious may be its substitute.
It is true, even Nagavarraa himself seems to have omitted to introduce this special rule, as
V. 288 which contains it, cannot bo genuine; for it occurs in the supplement of IL, from which
it has been transferred to the first chapter of M. (and Sb.; see p. 24, note 2, and corrections),
and at the same time contains a form of the Na gana that is foreign to Nagavarma, and is
probably a Telugu Indra ( n=N) gana. Under the impression that Nagavarma could not have
overlooked the rule regarding the Proceleusmatious, the editor has tried, as No. 258 shows, to
find it in v. 283, translating its conclusion as follows: "sasipura (wv^wv^, i. e. makharipu),
not being at 6 (vishayadri, i. e. if sasipura does not occur at 6), let puriiri {^—^) be (there)."
Against this translation (adri = meru), however, three objections must bo raised, viz.: 1) that
sasipura (candrapura) is an obscure word scarcely fit for expressing a name of 5iva; 2) that
NiVgavarma would have avoided the license of later writers to use the mutilated form of the
Nominative (sasipura instead of sasipura?!!); and 3) that sasipuravishnyi'idri has a different
meaning in v. 287, though there some MSS. read sasipurabanadri. The reading of the verse
ought to have been somewhat different (and it may have boon so) to allow the editor's exposition.
So the translation of the sentence as it stands is: "at sasi, pura, vishayanndadri (i. e. the odd
plaoos) not being (i. e. let it not be! But at the oven places) let purftri bo!" 4) NAgavarma

uever uses caruna; versos 318-320, 324, 327, 331, 335, wherein the word occurs, do not
belonfp to the original.


becomes slightly questionable on account of some grammatical ir-
regularities'). A peculiarity is that the Rag ales (254 seq.) appear
to be misplaced in the work; for, being not bound (at least for several
centuries) to only five Mora-feet, they, as regards their form, ought
to have followed the i4ryes; their feet fall under the rules of the later
Canarese Shatpadis. As to the Ragales only v. 254 is original, and it
says very little, the words "included within certain feet" in No. 243
being an editorial addition; after it in M. there is the dry remark: "one has
to look for thcni" (in other works?). Concerning both the Ragales and the
later Shatpadis the authors of the rules have neglected to point
out the number of feet as well as their different forms; and again none
of them has called attention to the circumstance that no true Canarese
foot is to begin with an Iambus. The editor, therefore, has supplied the
necessary remarks in this respect.

According to verse 22 Niigavarma took Pingala (Piiigala) as
his guide in composing his work, calling it Ch a n d o m b u d hi (23). This
statement by itself leaves it somewhat doubtful whether he meant only
the Samskrita Pingala, or also the so-called Prakrita Pingala; but
he more than probably, to some extent, meant both. It will prove
advisable first to show his general plan as stated in his own words.
He says there are "three and a half languages (bhfishu), viz. Samskrita,
Prakrita, Apabhramsa, and Paisacika," (probably calling the last one "half
a language" as being spoken only by barbarous tribes); "the bhasha-
jatis," he proceeds, "that are born of them, are those of all the 56
countries-), e.g. (the countries called) Dravida (i.e. Tamila)^), Andhra,
and Karnataka. There (i.e. in them, the 56 jati languages) are the three
kinds of Vvittas, called sama, ardhasama and vishama; thei'e (i. e. in them)
are the 20 C'handas", called Ukte, etc." If we follow H., the oldest of our
i-ecensions, Nagavarma goes on to say (p. 23, note 2; v. 68): "Apart from
them (the twenty-six Chandas' and their Vrittas) are the Jatis (also
common to all the countries), to which e.g. the Mala vrittas, Dandaka')
Ragale, and matragananiyama Skandhaka (Kanda) belong^). Apart from

1) For instance, in v. 164 occurs ^X9rf instead of jSaV^; in v. 203 f?X^ is scanned %.< — ^
instead of ^s^^, see p. 96 ;3Xcjo='^w— ; v. 227 has SfeSjJ;d instead of 55e5jLsc.

2) Sliatpancasat sarva vishaya; for shatpanciisat, in later works, generally cappanna is
substituted. 3) Dravida (Dravida), in South India, only means Tamil; and all other
meauings given by Northern writers and their followers to the word are higlily perplexing to
a Southerner. 4) These two classes are snmavrittas. 5) Ragale (Kaghatc) occurs in
Telugu as Ragada; a Telugu Dandaka seems to belong to the Mora metres; the Kanda bears
the same appellation in that language.


them (the Chandas' and Vrittas? or the general J;'itis?) again are all
those which are born of the languages of Karaataka, Andhra, Dravida,
Varata, Lata, Malava, Gaula, Gurjara, Kaliiiga, Aiiga, Vauga, Kerala,
Bahlika, Magadha, C'cri, Vacala, Pancala, Vei'igi (diflerent from Andhra!),
Tillava (Taulava?) and other countries; and they form the (particular)
Jatis of the languages of all the countries (sarvavishayabha-
shajati), and (for Karnataka) are the following: Madanavati, Akkara,
Caupadi, Gitike, Ele, Tivadi, Utsfiha, Shatpadi (L e. .S'arashatpadi), Ak-
karike, and Chandovatawsa."'' After this the author states (v. 69, 70):
"For each kind (.jati)=), in a two-fold way, from Ukte to Utkriti, I will
give (thee) instruction. In the two languages thus mentioned') (by
me) I will hence let thee know concerning (what is common to) the
languages etc. of all the countries (sarvavishayabhashadi)."
Further, after the Kagales, in verse 281 he says: "Thus, in this order,
in all ways, I have informed thee concerning (what is common to) the
languages etc. of all the countries; henceforth learn (also) the
mode (anda) of the Kanda!" Then, after the Kanda and the other
(Sanscrit) Morar-feet metres, he begins with the prose-heading: "I will (now)
state the (particular) Jati of the Karnataka country"; and proceeds in
verse 296: "I have let (thee) know in full (what is common to) the
languages etc. of all the countries; I will (now) relate the mode
(anda) of the Karnataka language."

The above statement appears to say that in the SaHiskrita, Prfikrita,
Apabhrawsa and Paisacika as well as in all the 56 Jati languages (i. e.
daughter-languages) there occur the three chief kinds of Vrittas; and, as
a different class, the Maliivrittas, the Dandaka (Vrittas), the (matragana)
Ragales and the matragana Kanda, which arc Jatis (that are based on
the prosody of the mother-languages, and occur also in all the languages).
Different from these two classes are the (particular) Jatis that have
(independently) arisen in the languages of all tlie countries (and may
to some extent occur in them).

So the division of all metres into the two classes: Vritta and Jati,

I) An Ela metre, iiccorcUng to C. P. Brown's Dictionary a carol or catch, is also in
Telugu; Telugu possesses also Akkara and Uts&ha. 2) Iloro JAti as in bhiishajilti, must

(lonoto "kind", and bo the same as "anda". 3) Tho "two-fold way" and tlio "two

languages" soom to bo identical; Samskrita for tho classical language, and Prakrita (in a
general sense of tho word) for all tho languages, which, according to former Hindu notions,
have arisen from Saniskrita. NAgavarma appears to say that verses 72-295 of tho text (so
far as they are his) belong partly to the Saniskjita Pingahi, partly to tho so-called Pr&kjita


that appears in Halayudha and which had existed even for some time be-
fore him, occurs in Nagavarma; although it is not exactly based on the
difiference between metres that are formed of syllable-feet and metres that
are formed of Moras and Mora-feet (as is the case with Halayudha, etc.)'',
for also the Malavrittas and Dandaka vrittas are counted with the Jatis. In
the secondary prose-paragraph of the text under No. 55 (that with slight
differences occurs in MSS. B., D., and Re.) the term "juti'' can only denote
"kind" in a genex'al sense-').

Comparing the present text of Nagavarma's work (includ-
ing the 8 metres peculiar to Rb.: 144, 145, 146, 1G2, 169, 187, 192, 197)
with the Sawskrita Chandas of Piiigala as it lies before us in the
8th volume of Dr. Weber's excellent "Indische Studien," it appears that
Nagavarma's introductory verses 24,26,27 and 34 occur, with
some modifications, in the first chapter of Piiigala; instead of the six or
seven cases in which according to P. a syllable may be long, N. with cer-
tain later writers (e.g. Ked.lrabhatta, W. 2l5), takes only five such cases
for granted. P.'s general arrangement (with which that of Kedara-
bhatta agrees) has been quite inverted by N., P. beginning the profane
metres (laukika chandas) with the Ganachandas, and N. with the Akshara-
chandas; N. therefore brings in the Caesura (v. 39) just before the
Aksharachandas, as vs. 40-66 do not belong to the original recension.
Further, P. inti'oduces only 21 species of Samavrittas, but N. 26. P.
adduces for his species from gayatri to utkriti only 87 instances; N. for
the very same no less than 136, and including the instances for Ukte-
Supratishthe, altogether as many as 156.

The following table showing the numerical difference regarding
the Samavrittas of P. and N. may be acceptable-": —

Kuml)er of instances that occur
Species. ;,, p j.j j,- j^ ^^^^^ p_ ^nd N.

gi'iyatri 1 8 1 (9G)

ushni/i 1 11 none

anushtubh 3 G 3 (HI, 112, 113)

brihatt 3 9 2 (118, 122)

parikti 6 G 4 (12G, 127, 123, 131)

trishlubh 12 15 8 (132, 133, 135, 136, 138-140, 142)

jagat! 19 16 11 (147-152, 158-162)

1) Cf. ps. 22, 23; VTeber p3. 288, 289. 2) After Talavritta D. has still a Dindema

vritta, called Bindima tf. by Re., and Mattebha (I) by B. 3) The garva, harivara (p. 26),

ratanta (p. 27) of H., and the mahgalike (p. 48) of Rb. are not included.

g 5




Number of instances that occur

ill P.


































in both P. and N.

3 (163, 1G9, 170)
3 (171, 172, 174)
2 (177, 178)

5 (188-191, 193)
2 (194, 195)
2 (199, 200)

2 (206, 207)
1 (211)
1 (215)
1 (218)

1 (221)

2 (224, 225)

Thus N.'s present edition has only 54 Samavrittas in common with P.
Of these 10 bear totally different names in P. 's work''; the names
of 7 again differ to some extent-J. A whole class of Samavrittas, the
Malavrittas, do not occur in P.^)

Thirty-one of the Samavrittas that are p e c u 1 ia r to P., are the
following: —

































^w^— — w www WW —

1) I Btato the verses, adding P.'s names: 127, rukmavatt; 131, iuddhavirAj; 142, kudmn-
ladantt; IC9, kanakaprabhii; 174, varasundnrt; 193, aritatha (kokilakn); 194, vibudhapriyA;
195, kusumitalatilrcllita; 199, vUmitA; 20U, dhritairt (saisvodana). 2) 113, 147, 158,

160, 191, 21 1, 225. 3) Tlic term of malarritta docs not occur in the St. Petersburg Lexicon.



17. Praharshant,

18. Gtturt,

19. Kutilftgnti,

20. Asambiidha,

21. Apfirnjitii,

22. Kutila,

23. Risliabhfigttjavilasita,

24. iS'ailnsikha,

25. Varayuvfttt,

26. Vai/isapatrapatita,

27. Atisayin!,

28. Vanamala,

29. Suvadana,

30. Vritta,

31. AsTalalita,

■ 1-^-1 ^ -

Eegarding the Pauses (yati) in N. it will be readily observed that
those of vs. 126, 131-133, 136, 139, 147, 149-151, 158, 159, 206 and
211 do not agree with those pointed out for them by P. Several times the
recensions of N. differ from each other with regard to the place of
the Caesura (vs. 132-134, 153, 164, 171, 174, 175, 178, 179, 189, 198,
202, 214). Besides yati, N. has usir (breath), virati, virama, visrama,
visramana, visrunta, visrama. Where the Caesura is not indicated by
him, he, like P., appears to mean that it is at the end of the line. (In Rb.
I have met no reference to Caesura).

Dr. Weber's above-mentioned volume enables us also to compare the
Samavrittas of the present text with those contained in the Sanscrit Pro-
sody, called Vrittar atnakar a, by Kedarabhatta who lived be-
fore the 1 3th century, but after H a 1 a y u d h a, who, as it appears, wrote his
commentary on Piiigala, the Mritasanjivini, towards the end of the 10th
century (W. 184, 193, 206, 417). Sixteen Vrittas that do not occur in P.,
but in Nagavarma, and the Vrittaratnakara though partly bearing dif-
ferent names, are the following: —


1. Vicitra (93, also in H.), Kcdiira's Somarajt

2. Kumuda (97, also in H.), K.'s RamanI

3. Mukula (98, also in U.)

4. Madhumati (Sulabha in M., 108; not iu H.,

but M. and Rb.)

5. Bhadraka (120, also in H.), K.'s Bhadrika

6. Maniranja (130, also in H.), K.'s Maniraga

7. C'andriko (137, also in H.), K.'s Bhadrika

8. Manikya (141, not in H., but M.), K.'s


9. Sumukhi (144, only in Rb.)

10. Sri (146, only in Rb.)

11. Nirupama (153, not in H., but M.), K.'s


12. Lalitapada (155, also in H.), K.'s Abhi-


13. Candrike (156, not in H., but M.), K.'s


14. Sukesara (Surangakesara, 181, also in H.)

15. Jagadvandita (184, also in H.), K.'s


16. Pancacamara (187, only in Rb.)




From this list it seems to become certain that there is a very close

1 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Online Libraryson of Vennam-ayya Naga VarmaNagavarma's Canarese prosody → online text (page 1 of 20)