Notes on some prosodical peculiarities of Chand

NOTES ON SOME PROSODICAL PECULIARITIES OF CHAND. BY A. F. RUDOLF HOERNLE, D. PH., BANARAS. As an edition of C h a n d's Epic, the Prithiraj Rasau, has been commenced by the Bengal Asiatic Society, it may be of some service to note some of the most striking peculiarities of the verse-composition of C h a n d that I have met in the course of my reading of his epic. In limine, I may remark that it is a well established rule in Hindi prosody that consonants may be doubled in order to produce a length where the word naturally has a short vowel. The observation of this rule will smooth away many seeming irregularities of metre, and it explains also the cause of a not uncommon kind of various readings of the different MSS. For different careless scribes often omit these merely metrical doublings at different places, to be supplied by the reciter: thus in Revatata Prast. Doha 14, 4, B* and T read baja bajjana vara, but A correctly bajja bajjana vara . Sometimes in two corresponding lines the doubling is made in one line, and carelessly omitted in the other, while the metre as well as the rhyme require it in both: thus in Anangpal Prast. Kavitta 7, 10, 12, A reads { dhura se bhavi trijijyai || / phiri kalahattara kijjyai || } where the correct reading clearly is lijjijyai, as C has it; while B and T, altogether incorrectly, read also kijjijyai . In many places the doubling is omitted in all MSS., and must be restored in a correct edition. It should be remembered that, on the whole, all the MSS. of C h a n d that we possess agree so closely and minutely that they must be transcripts of one original in which the metrical doublings were little attended to. In rare cases it even occurs that the doubling is made in the wrong place, as in Devagiri Prast. Kavitta 11, 3. B reads jo bhajjai graha appanai, where the correct reading is jo bhajjai graha apana, as T has it. Or in Revatata Bhujangi 61, 10, T mano damda sukki aggi vai varam ||, where B and A read correctly aggi . A nasal is doubled generally by the insertion of an anuswara; thus all MSS. alike read in Revatata Motidam 63, 15:- sura maruta phoja prasamma calai|| or Adiparva Bhujangi 5, 1:- prathamam bhujamgi sudhari grahannam|| or Anangpal Bhujangi 15, 3, 4:- taje moha maya mahamnanni kahanni|| taje bamdha putram hari citammamni|| The last example exhibits both kinds, the ordinary doubling in kahanni, and the anuswara in citammamni . Another peculiarity of C h a n d is that with him double consonants containing a (so-called) semi-vowel (ya, va, la, ra ) or ha may or may not make a preceding short vowel (positionally) long. In modern prosody, as is well known, this rule is limited to ra. Examples of the preceding vowel remaining short are-of ya in Devagiri Prast., Totaka 22, 2:- subha kau abhirama ki kamsu luyo|| Again, of va, ibidem, 22, 12:- umage taham amsua dvai nayunam|| Again, of ha, in Revatata Motidam, 63, 7:- kabadum duri kamna na pucchata nenum|| or ibidem, Kavitta 41, 10:- sahi cinhava mu ujjayau|| In the following verse the same compound ha produces the usual positional length; ibidem, Kavitta 44, 1:- utari sahi cinhara || An analogous though far more remarkable peculiarity of C h a n d is that with him even a sibilant (sa or sa-of sa I have not met with any example) in composition with a mute consonant does not always make a preceding short vowel long. The instances are very rare, though sufficiently marked; *e.g., in Adiparva* Doha 6, 1:- puththi camdu bumdaha bayanu || The first hemistich of the Doha consists of three feet of 6, 4, and 3 instants respectively-altogether 13; and the second foot may not be an amphibrach ( u - u ). If the second syllable be taken as long by position, as it would be under ordinary circumstances, we should have in the second foot an amphibrach, and should be obliged to suppress the final vowel of bayan-the first an impossibility, the second an anomaly.
----------
* B = Baidlah MS.; T = Col. Todd's MS.; A = Agra College MS.; C = Banaras MS.

Or again, ibidem, Bhujangi 5, 24:- tina ki uciti kuvicamda bhai|| So again, Revatata Kavitta 3, 7:- hoi krpala hastini|| If this line were scanned thus, hoi kripalu hastini ||, we should have an amphibrach in the second foot, which the metre does not admit. 2, there would be 12 instants altogether, while the metre has only 11 instants in the odd hemistichs. 3, the final would be a long syllable, instead of, as the metre demands, a short one. On the other hand, as I shall presently show, two short syllables at the end are often contracted by Chand into one long one; that is, hastini stands for hastiniya (Prak. hatthinia, for hastinika ); and the syllable hast is evidently treated as a short one; and thus, if we read hoi kripalu hastiniu, the line is quite regular as regards number of instants, kind of feet, and quantity of termination. Again, take Revatata Kantha-sobha 32, 15:- mukha kuddina ghumghata assu luli|| The kantha-sobha measure consists of an initial iambus and three following anapaests in each line. The initial iambus may be obtained by suppressing the final vowel a of mukha mukh, and assuming that sa + ka does not produce positional length. It appears to me that this verse possibly affords us a clue to the explanation of this strange phenomenon that sa and sa in composition with another consonant do not make positional length. The modern word for mukha is muha muh; probably Chand already spoke muha, though he continued to write mukha . Now, as has been already observed, ha, like the other semi-vowels, has not necessarily the effect of making positional length. Similarly we may suppose that also in the other case, where a sibilant in composition with a consonant apparently does not make positional length, Chand pronounced really not a sibilant, but an aspirate, which did not constitute a double or compound consonant in prosody. Thus, for hastini Chand probably read hahatini, or, what is more natural and consonant to phonetic rules, hathyani, though he continued to write hastini . Now let it be remembered that the modern Hindi is hathini, and the Prakrt haththini ; and we shall probably be correct in concluding that the principle which underlies these phenomena is simply that which is also observed in other languages, that writing does not keep pace with pronunciation, and that while the pronunciation of a word is modified, it still continues to be written exactly as it used to be written when it was pronounced differently; as, e. g., we write in English love but pronounce lov (German liebe); we write night but say nite (German nacht). We shall see, further on, that this principle affords us a solvent also of some other peculiarities of Chand. For ucita in the first-cited example Chand probably read ucchitha, a slight modification of the Prakrt ucchidr ; from the modern Hindi this Prakrtic form has disappeared, and is replaced by the Sanskrt ucchista uchchhishta ||. I now proceed to notice a few anomalies peculiar to the stanza called kavitta. This stanza is a combination of two different verses, viz. the kavya and the ullal. The verse called kavya consists of two distichs, with 24 instants in each line, and with a pause after the 11 th instant, which divides each line into two hemistichs, with 11 and 13 instants respectively. The whole line consists of five feet of 6, 4, 4, 4, 6 instants respectively. It follows that the last syllable of the odd hemistichs must be always a short syllable, and the third foot of the whole line must be either an amphibrach ( ◡ - ◡ ) or an anapaest ( ◡ ◡ - ) or a proceleusmatic ( ◡ ◡ ◡ ◡ ) ; generally it is an amphibrach. On the other hand, the second and fourth foot of the whole line may not be an amphibrach. The verse called ullal consists of one distich of 28 instants in each line, and with a pause after the 15 th instant, which divides each line into two hemistichs of 15 and 13 instants respectively. The whole line consists of 7 feet of 4, 4, 4, 3, 6, 4, 3 instants respectively. The first, the third, and the sixth foot may not be an amphibrach, the second foot may not be a dactyl, but is almost universally an amphibrach; and the fourth and the seventh foot may not be a trochee. It follows, then, that the kavitta stanza consists of three distichs of 6 lines or 12 hemistichs, of which latter all even ones have 13 instants, while of the odd ones the first four have 11 instants, and the two last 15 instants. These are the ordinary rules of the kavitta, to which Chand, in the majority of cases, conforms. Not unfrequently, however, he adds 3 instants to one or several of the first four odd hemistichs, which should have only 11 instants, and thus makes anomalous, redundant hemistichs of

----------
14 instants. The three instants which are thus occasionally added are invariably a trochee (- ◡); which, therefore, together with the one terminating instant of the proper hemistich, always forms an amphibrach. It has been already observed that the third foot of the line of the kavya is, as a rule, an amphibrach; and it thus appears that the redundant hemistich owes its origin to a tendency of Chand to complete the catalectic hemistich, thus turning it into a full line, forgetting that it is merely a hemistich, and that the missing half of its last foot forms the commencement of the second hemistich; then, remembering this fact again, he commences the second hemistich again with a trochee, which in reality has already been used up by being appended to the first hemistich; for it may be noted that the second hemistich begins almost as invariably with a trochee, as the first redundant hemistich ends with it; though these redundant lines occur too frequently to allow of their being ascribed to forgetfulness-the tendency must have been a more or less conscious peculiarity of Chand. But, no doubt, in some cases, the redundancy may be got rid of by supposing the effect of another peculiarity of Chand, to be noticed presently, according to which sometimes vowels which are written long must have been pronounced by him short. Examples of redundant hemistichs, occurring within a very short space, are the following:- Revatata 44, 7 : to uccara gori narimda hama badi suratamna|| ,, 47, 1 : bara mamgala pamcami sajuddha dina sudi no prathiraja || ,, 41, 3 : alama bam alama gumana pana ujabaka niraste || Occasionally, though far less frequently, Chand gives to the even hemistichs a redundant form by prefixing two instants, always consisting of one long syllable, to their first half-foot, so that they have 15 instants instead of 13. For example:- Revatata 49, 4 : jyom suraha buddhi bala soihai || ,, 49, 6 : jyom prata ba ra baum viyogi || Deragiri 39, 4 : jo satha tari jai trayadhanni || Revatata 40, 12 : pumca vi na pimda jo upajjai to juddha hoi lajji vinam || In one or two cases I have met with an analogous kind of redundant measure in the Doha, where the even hemistich has 13 instants instead of the usual 11, the superfluous two instants, always a long syllable, being added at the beginning. Thus, Revatata 2, 4 : to bana abhuba gaja jhumda || Generally, as in the last three examples, this superfluous long syllable belongs to a word which might be omitted altogether without affecting the sense of the sentence in any essential way; though, when added, it, no doubt, adds something to the clearness of the sense. Hence we may perhaps suppose that in reciting, when the sense may be brought out into clear relief by the modulation of the voice, these redundant words were omitted; but in writing they were added to increase the clearness of the sentence. Another instance of a similar kind of redundancy is sometimes met with in the more unusual metres, as the Motidam. It consists in the prefixing to the verse one instant or a short syllable, thus, Revatata Motidam 63, 1 : rati ra ja ru jo va na ra ja ta jora || ,, ,, 63, 16 : gati lajji samkuci kamche mili ai || ,, ,, 63, 18 : pragamtai ura tuccha soi dara bhiva || Sometimes the apparent redundancy may be avoided, by assuming the suppression of a short a in recitation; thus in Revatata Motidam 63, 7 : kabahum duri kamna na pucchata naina || or ,, 63, 15 : surabharata phauja prathama calai || for there the compound i, and me, containing a semi-vowel, would not render the previous short vowel positionally long. But the legitimacy of the expedient is doubtful; suppression of a final a metrically is opposed to the genius of Hindi poetry; and at any rate the expedient would not obviate all cases of redundancy. Again, a peculiarity of Chand, foreign to

ordinary Hindi prosody, is the substitution of a long syllable for two short ones, and of two short ones for one long syllable. Examples of either case are not very common; those of the latter kind occur only in the syllabic measures (vrtta chamda ). Thus in Revatata 61-1: milecai cahuana sa campi gauri || Here we have cahuana ; in other places the anomaly is avoided by writing cohana . Again, ibidem 61, 4 : tinam amma nisamna mili addu kosam || ,, ,, 61, 18 : mare goriyam se na a ru cahuvanam || ,, ,, 61, 23 : lage dhara dhara tinai dharaha tuddai || Examples of the substitution of a long syllable for two short ones occur, I believe, only in the time-measures (matra chamda ), and of these almost exclusively in the Kavitta stanza. In the latter they are met with most commonly either at the end of the even hemistichs of the Kavya verse, or in the middle of the uneven hemistichs of the Ullal verse; e.g., in Revatata Kavitta (Ullal) 39-10 : pacchimahim samna pahana saha || ,, 41, 11 : sabha le sura samamta nrpa || Again, Devagiri Kavitta (Kavya) 52, 3-6 : naga samuha dhatta ri | dhahi devala suramga madha || thamna thamna nara u dai | camda ta sa uppama paiya || or Revatata Kavitta (Kavya) 39, 5 : cham mamgola lala ri | bisa tamki bara bamdhai || Ibidem, 40, 3 : keli dham kumja ri | saha sari dala pappara || Ibidem, 78, 7 : sabara sara ba i yau | visama da gam dhana jharai || A clue to the understanding of this apparent anomaly is, I think, afforded by the word udai (he flies, 3 rd pers. sing. pres. of udana ). Even now the word is occasionally written udai (or udaya ). If the word be so written in the verse quoted above, the metre becomes perfectly regular. Now considering that the form udai, being very nearly Prakrit uddei, is the more original of the two, it seems to me there can be no doubt that Chand must have recited udai, and that the form udai is merely a modernization of the word, probably, by subsequent scribes. Similarly bhatati, pacchimi, lalari, kumjari, baiyi are modernized spellings, adapted to the pronunciation of those words as it is now usual; in the time of Chand they must have been differently pronounced, and (if they are spelt according to the pronunciation of those times) written thus : bhatariya, pacchimiya, lalariya, kumjariya, baiyau . This is the more probable as those Chandic forms are nearer to the Prakrit bhadario, pacchimio (Skr. pascimakah ), lalario, kumjario (Skr. kumjarakah ), badiao (Skr. baditakah ); and, 2, as the modernized, contracted forms occur only exceptionally, while in most places the original uncontracted Chandic forms are preserved; *e.g., in Revatata Dandamali* 50, 25 : barabira dhara jogimda pamtiya || ibid. Doha 52, 1 : bai bisa dhumdhari pariya || ibid. Kavitta 59, 10 : darari dhamna goriya musa || ,, ,, 61, 1 : chadi addha bara ghatiya || etc. etc. In all these places we should say now in modern Hindi pamti, pari, gori, ghaड़ी, etc. Another peculiarity of Chand's Epic is that sometimes a short vowel must be read where a long vowel or a diphthong is written; *e.g.- u* for o in Revatata Bhujangi 43, 5 : dou dina dinam kudhi bamku assi || i for ai, ibidem, Dandamali 50, 20 : kai suta dhanu sasi sajhai || u for u, ibidem, Kavitta 41, 7 : himdu sena upparai || a for a, ibidem, Kavitta 41, 2 : tisa taham rati phirantai || In these verses duu must be read for dou ; ki for kai ; himdu for himdu ; taham for taham ; etc. Similarly sometimes an anunasika must be read where an anuswara is written; thus in Revatata Kav. 57, 7 : guru pamcami ravi pamcama || ibidem, Motidam 63, 2 : campyau sasura dara saisaba koi || ibidem, Doha 42, 3 : taba lagi rupi pumdira ne || In these verses pamcama must be read for pamcama ; campyau for campyau ; pumdira for pumdira, for the anuswara causes a preceding short vowel to be positionally long, while the anunasika has not that effect. Now the explanation of this peculiarity, I think, is to be found in the same principle which has

been already noticed as explaining another difficulty, namely, that writing does not generally keep pace with pronunciation; the former often exhibits an earlier phase of language, where pronunciation shows it in a more recent one. Thus against the old Hindi doda we have the modern low-Hindi dui two; for the old Hindi kai modern Hindi has ki . In the two words himdu and taham, indeed, the incongruity of writing and pronunciation is preserved even in modern Hindi; for though both words are still written in the same way as in C h a n d, practically they are now pronounced as C h a n d must have done in those verses, viz. himdu and taham . And as regards the anunasika, the change of the anuswara to the anunasika in modern Hindi (generally, though not universally, with the effect of lengthening the preceding vowel) is one of its distinguishing features; thus Prak. pamca is Hindi pamca, Prak. campana is Hindi campana, etc. It may be noted, en passant, that the MS. A actually reads himdu, not himdu like B and T, in the above-quoted example. Further, that long vowels or diphthongs may be occasionally read as short vowels is shown by the fact that in some cases the short vowel is actually substituted for the long one: e.g., in Revatata Dandamali 50, 25: bara bira dhara jogimda pamtiya jugimda is to be read for jogimda ; and the word is actually so spelt in Revatata Kavitta 78, 1- jodhari jogi jugimda kahyau kaharau|| As regards the kinds of metres employed by C h a n d, I have only met with one kind which, as far as I am aware, is altogether peculiar to C h a n d. All other metres used by him are found in native treatises on prosody, and are the common property of native poets. But in one place C h a n d uses a metre which, from the fact that he particularly explains its properties (which he never does in the case of any other of the established metres, however uncommon it be), I am inclined to conclude was his own invention. It occurs in the Revatata Prastava and is numbered 35. In the preceding doha, where its rules and name are stated, it is introduced thus: ● gyaraha aksara pamca sata | laghu guru hoi samana || kamthasobha vara chamda kau | nama kahyau paravana || *i. e.*, "let there be eleven syllables, made up of parts of five and six (in each line), and let the long and the short ones be placed alike (in them); Kantha-sobha, truly, is the name of this noble metre." As an illustration I may quote the first two verses in this metre: phire haya sassara sassara se || manau phiri imdua pamtha kuse || that is, each verse or line consists of an iambus and three anapaests. It belongs to the syllabic metres (varna chamda ). All the other metres of C h a n d are established ones; though several of them are habitually called by names by which they are not usually known; and under this guise they are at first apt to pass unrecognized. Thus the metre always called sataka or sattaka by C h a n d is nothing else but the well-known Sanskrit metre sardulavikridita. Witness, for example, in Revatata Sataka 15, 1-4:- srotam bhupaya goriyam vara bharam vajjadra sajjadra ne || sa sena caturamा bamdhi ulalam tantara maruphayam || tujjhi sarasa upparatra sarasi prajnanayam pamnayam || ekam jiva sahava sahi nutrayam viyam svayam senayam || Thus the metre called Dandamali by C h a n d is identical with the Harigita or Mahishari; the Kavitta of C h a n d is the same as the Chhappai, etc. In conclusion, I may notice a peculiarity of C h a n d which is merely one of spelling, and in no way connected with prosody. Guttural, and dental aspirate consonants are, as a rule, reduplicated by means of an aspirate; a double sa (*i. e.* kha) and tha are always sasa and thatha ; a double gha and dha sometimes ghagha and dhadha ; but double cha, jha, tha, pha, are always regularly ccha, jjha, ttha, ppha . Again, the cerebral and labial sonant aspirates are by preference reduplicated by their respective surd aspirates: thus double dha is ttha, and double bha is pbha .