Metrical Experiments in Bengali

New Metres in Bengali

Of course, Prabodh Sen is right. I suppose what Buddhadev means is that none of the very great poets invented a metre ―they were all too lazy and preferred stealing other people's rhythms and polishing them up to perfection, just as Shakespeare stole all his plots from wherever he could find any worth stealing. But all the same, if that applies to Shakespeare, Homer, Virgil, what about Alcaeus, Sappho, Catallus, Horace? they did a good deal of inventing or of transferring ―introducing Greek metres into Latin, for example. I can't spot a precedent in modern European literature, but there must be some. And after all, hang precedents! A good thing ―I mean, combining metric invention with perfect poetry ―would be still a good thing to do, even if no one had had the good sense to do it before.       

4 November 1932

*

It is certainly not true that a good metre must necessarily be an easy metre ―easy to read or easy to write. In fact even with old established perfectly familiar metres how many of the readers of poetry have an ear which seizes the true movement and the whole subtlety and beauty of the rhythm ―it is only in the more popular kind of poems that it gets in their hearing its full value. It is all the more impossible when you bring in not only new rhythms but a new principle of rhythm ―or at least one that is not very familiar ―to expect it to be easily followed at first by the many. It is only if you are already a recognised master that by force of your reputation you can impose whatever you like on your public ―for then even if they do not catch your drift, they will still applaud you and will take some pains to learn

Page-145

the new principle. If you are imposing a principle not only of rhythm but of scansion to which the ear in spite of past attempts is not trained so as to seize the basic law of the movement in all its variations, a fair amount of incomprehension, some difficulty in knowing how to read the verse is very probable. Easier forms of a new rhythm may be caught in their movement, ―even if some will not be able to scan it; but other more difficult forms may give trouble. All that is no true objection to the attempt at something new; novelty is difficult for the human mind ―or ear ―to accept, but novelty is asked for all the same in all human activities for their growth, amplitude, richer life. As you say, the ear has to be educated ―once it is trained, familiar with the principle, what was a difficulty becomes easy, the unusual, ―first condemned as abnormal or impossible, ― becomes a normal and daily movement.

As for the charge of being cryptic, that is quite another matter. Obscurity due to inadequate expression is one thing, but the cryptic may be simply the expression of more than can be seized at first sight by the ordinary mind. It may be that the ideas are not of a domain in which that mind is accustomed to move or that there is a new turn of expression other than the kind which it has been trained to follow. Again the ordinary turn of a language, as in French or Bengali, may be lucid, direct, easy: if you bring into it a more intricate and suggestive manner in which the connections or transitions of thought are less obvious, that may create a difficulty. A poet can be too easy to read, because there is not much in what he writes and it is exhausted at the first glance, ―or too difficult because you have to burrow for the meaning. But otherwise it makes no difference to the excellence of the work, if the reader can catch its burden at the first glance or has to dwell a little on it for the full force of it to come to the surface. The feeling, the way of expression, the combinations of thought, word or image tend often to be new and unfamiliar, but that can very well be a strength and a merit, not an element of failure.

28 January 1933

*

Page-146

I am, as you know quite in agreement with you as regards the principle. At the same time there is a greater difficulty in Bengali than in Hindi and Gujerati. For in these languages the stylisation is a long-accepted fact and the ear of the writer and reader are trained to appreciate it, but in Bengali the trend has been on the contrary to more and more naturalism in metre and such stylisation as there was was not quantitative. Now the writer has the double difficulty of finding out how to stylise successfully in detail and of getting the ear of the public to train itself also. . . .

Quantitative Metre in Bengali

This question of quantity is one in which I find it difficult to arrive at a conclusion. You can prove that it can be done and has been successfully done in Bengali, and you can prove and have proved it yourself over again by writing these poems and bringing in the rhythm, the , which is absent in Satyen Datta. It is quite true also that stylisation is permissible and a recognised form of art ―I mean professed and overt stylisation and not that which hides itself under a contrary profession of naturalness or faithful following of external nature. The only question is how much of it Bengali poetry can bear. I do not think the distinction between song and poem goes at all to the root of the matter. The question is whether it is possible to have ease of movement in this kind of quantitative metre. For a few lines it can be very beautiful or for a short poem or a song; that much cannot be doubted. But can it be made a spontaneous  movement of Bengali poetry like the ordinary mātrā-vrtta or the others, in which one can walk or run at will without looking at one's steps to see that one does not stumble and without concentrating the reader's mind too much on the technique so that his attention is diverted from the sense and bhāva? If you can achieve some large and free structure in which quantity takes a recognised place as part of the foundation, ―it need not be reproduction of a Sanskrit metre, ―that would solve the problem in the affirmative.   

31 May 1932

Page-147

Quantity in Classical and Modern Languages

I can't agree with your statement about Sanskrit आ, ए, ओ, that they are long by stylisation only! In fact, I don't quite understand what this can mean; for in Sanskrit आ at least is the corresponding long to the short vowel  अ and is naturally as long as the devil ―and the other two are in fact no better. The difference between ए and  ऐ and ओ and औ  is the difference between long and ultra-long, not between short and long. Take for instance the Sanskrit phrase ; येन तेन प्रकारेण;  I can't for the life of me see how anyone can say that the ये ते  रे  or the का there are naturally short to the ear, but long by stylisation. The classical languages (Sanskrit, Greek, Latin) are perfectly logical, coherent and consistent in the matter of quantity; they had to be because quantity was the very life of their rhythm and they could not treat longs as shorts and shorts as longs as it is done, at every step, in English. Modern languages can do that because their rhythm rests on intonation and stress, quantity is only a subordinate element, a luxury, not the very basis of the rhythmic structure. In English you can writepretending that "road" is short and "runs" is long, or —where the sound  corresponding to Sanskrit ए () or that corresponding to Sanskrit ओ () is made short or long at pleasure; but to Sanskrit to the Sanskrit, Greek or Latin ear it would have sounded like a defiance of the laws of Nature. Bengali is a modern language, so there this kind of stylisation is possible, for there  ए  can be long, short or doubtful.

All this, not to write more about stylisation, but only as a protest against forcing modern ideas of language sound on an ancient language. Bengali can go on its way very freely without that, Sanskritising when it likes, refusing to Sanskritise when it doesn't like.

2 June 1932

 Aksara-vrtta and Mātrā-vrtta

I have read your account of the tridhārā and my mind is now clear about it; I have not yet read Anilbaran's contentions, so

Page-148

there I am still in the dark. But here are certain points that I want to make clear.

(1) Prabodh Sen's rule of the yaugika-vrtta does not agree . with what I was taught about the aksara-vrtta. When I first heard of Bengali metre in England, my informant was quite at sea. He confidently described Michael's blank verse as a 14 syllable line (8 + 6), but when asked to give examples we found that the lines as pronounced were of 12, 13, 14 or more syllables and when my brother Manmohan asked him to explain this discrepancy, he could merely gape ―no explanation was forthcoming! How ever, when I took up seriously the study of the literature, it was explained to me by competent people, themselves poets and litterateurs ―thus

"The line is strictly a line of 14 syllables, no more, no less (i.e. it is a true aksara-vrtta), but the aksara or syllable here is not the sonant Bengali syllable as it is actually pronounced, but the syllable as it is understood on the Sanskrit principle. In Sanskrit each consonant letter (aksara) is supposed to make a separate sound (syllable), either with the aid of other vowels or by force of the short a sound inherent in it ―except in two cases. First, if there is a conjunct consonant, e.g. gandha, the n is not sonant, not separate, but yukta to the dh, and therefore does not stand for a separate syllable; secondly, if there is a virāma-cihna as in daibāt, then also it loses its sonant force, there is no third syllable ―it is a dissyllable, not a trisyllable. Bengali has applied this rule, dropping only the last part of it, in disregard of the actual pronunciation. Thus  দান or ধন  is in Sanskrit (as in Oriya) a dissyllable, in Bengali also it is treated as such in poetry, although in fact it is a monosyllable to the ear. Externally this sounds artificial and false to fact, but rhythmically it is unexceptionable, the cadence of the voice supplying a double metre there.will be a dissyllable as in Sanskrit, because ndh is a yuktāksara. On the other handwill be a trisyllable because there is no distinction made of a virāma-cihna, no distinction therefore between and  িনধন, each is a trisyllable."

According to this explanation and the rule it supplies, it is

Page-149

true that a yugma-dhvani at the close of a word has always two mātrās, but the other part of Prabodh Sen's rule is not always true, viz. that in the middle of a word it counts only as one.  That would be invariably true of an indubitable , as in but not otherwise. On this principle there is no difficulty at all about মহাভারেতর কথা, the line is of 14 syllables and cannot be reckoned in aksara-vrtta as anything else. There is no difficulty about such lines as Michael's

10 svaras, but 14 aksaras, ―because the েমঘ , though in the middle of a word, must be two mātrās, since the ghn in Meghnad is not a compound consonant, but two separate aksaras. There is a difficulty about and , but that is because one is undecided whether to treat it as a compound and a compound or as two separate words joined together, িদক্ , being kept apart as with the t of or the k of . In the latter case and িদক্ are dissyllables, in the former, trisyllables. And so on, as regards other doubtful points like চাওয়া.

This, I say, was what I was taught and it is according to this rule that I have hitherto scanned the aksara-vrtta. I am quite prepared to adopt a new principle if it is more scientific, but I think that historically this explanation is not unsound, that it represents the idea Michael and Nabin Sen and the rest had of the basis of their verse and shows why it was considered as of a syllabic character.

(2) I did not think or hear that Tagore invented the mātrā-vrtta ―I could not, because I never heard of the mātrā-vrtta at that time. What I understood was that the svara-vrtta was . not recognised as a serious or poetic metre before Tagore, ―it was used only for nursery rhymes etc. or in some kinds of loose popular verse. Tagore did not invent, but he popularised the svara-vrtta as a vehicle for serious poetry ―it was at least professedly under his banner that a violent attack was made on the supremacy of the aksara-vrtta. I remember reading articles even in which it was reviled as a nonsensical conventional fiction: Oriya Bengali. "If you want to keep it" thundered the

Page-150

polemist, "let us all learn to read like Oriyas, `Rabana swasura mama, Meghanada swami', but let us rather be Bengalis and drop this absurd convention of a pseudo-Sanskritic past." The article amused me so much by its violence in spite of my prepossession for the aksara-vrtta that I remember it as if I had read it yesterday ―and it was only one of a numerous type. At any rate as a result of this campaign, svara-vrtta fixed itself on an  equal throne by the side of aksara-vrtta. I mention it only as a point of literary history of which I was a contemporary witness. I suppose, as usually happens, Tagore's share in the revolution was exaggerated and there were others who played a large part in its success.

(3) Mātrā-vrtta is therefore to me a new development, not as an invention perhaps, but as a clearly understood distinct principle of metre. But it exists, if I have understood your explanation, by a thorough extension of the principle which the aksara-vrtta applied only with restrictions. As the Sanskrit limitation about the virāma-cihna was swept away in the aksara-vrtta, so now in the mātrā-vrtta the limitation about conjuncts like  is swept   away and all yugma-dhvanis are reckoned as two mātrās. In that sense Anilbaran's description of it as of the aksara-vrtta would have some meaning, but at the same time it would not diminish the validity of your contention that it is a new opening with endless possibilities in a new principle of metrical rhythm. Two men may be cousins or brothers or near relatives, but one a conservative, the other a revolutionary creating a new world and a new order.

All this is no part of my final formed opinion in the matter. I have not yet gone through either Anilbaran's writing or Prabodh Sen's letter. It is only to put down my present understanding of the situation and explain what I meant in my letter.

Mātrā-vrtta

I am quite convinced of the possibilities of the mātrā-vrtta ―which would exist even if Anilbaran is right in insisting that it is the of the aksara-vrtta. Two people may be cousins and

Page-151

yet have different characters, possibilities and destinies ―and so may two metres.  

9 September 1932

*

I shall go through Prabodh Sen's letter, but it may take me some time. What is the exact scope of the discussion with Anilbaran, is it that he does not recognise the reality of the mātrā-vrtta as a separate principle of Bengali metre? That I suppose was the position before. Originally, indeed, there was only one stream recognised, ―that I remember very well, for it was the time when I was learning and assiduously reading Bengali literature; at that time what you now call svara-vrtta was regarded as mere . popular verse or an old irregular verse-form. Afterwards with the advent and development of Tagore's poetry, one began to hear of two recognised principles of Bengali metre, Swara (I was going to say Kshara) and Akshara. Is it Anilbaran's contention that only these two are real and legitimate? Whatever it be Anilbaran is ¯ a born fighter and if you tell him that all the Mahārathas are against him and his squashing defeat a foregone conclusion, he will only gallop faster towards the battle. My own difficulty is that I have not yet grasped the principle of the mātrā-vrtta― what is it that determines the long or the short mātrā in Bengali?

Mātrā-vrtta and Laghu-guru

I return you the former letter from Prabodh Sen which I managed to find time to read only today. He has a most acute, ingenious and orderly mind, and what he says is always thought-provoking and interesting; but I am not persuaded that the form of Bengali mātrā-vrtta and Sanskrit laghu-guru is really and intrinsically the same. Equivalent, no doubt, in a way, ―if we substitute Bengali metre for Sanskrit quantity; but not the same because Bengali metre and Sanskrit quantity are two quite different things. It is something like the equivoque by which one pretends that an English iambic metre or any other with a Greek name is the same as a Latin or Greek metre with that name ―an equivoque based on the fiction that a stressed and an unstressed English syllable are

Page-152

quantitatively long and short. There is a certain kind of general equivalence, but a fundamental difference ―as those who have tried to find an equivalent in the English stress system to the quantitative Latin or Greek hexameter, alcaic or sapphic metres have discovered ―they could not be transplanted, because it is only in true quantity that they can live.

23 September 1932

Laghu-guru

If you can establish laghu-guru as a recognised metrical principle in Bengali, you will fulfil one of my two previsions for the future with regard to the language. When I was first introduced to Bengali prosody, I was told that Madhusudan's blank verse was one of fourteen syllables, but to my astonishment found that sometimes ten syllables even counted as fourteen ―e.g.

Of course, it was afterwards explained to me that the syllables were counted on the Sanskrit system, and I got the real run of the rhythmic movement; but I always thought: why not have an alternative system with a true sonant syllabic basis ―and, finally, I saw the birth (I mean as a recognised serious metre) of the svara-vrtta. Afterwards I came across Hemchandra's experiments in bringing in a quantitative element and fell in love with the idea and hoped somebody would try it on a larger scale. But up till now this attempt to influence the future did not materialise. Now perhaps in your hands it will ―even apart from songs.

20 October 1932

*

It [a song composed by a disciple] is good. But there is a tendency to run into a conventional model. Originality, plasticity, vigour, a new utterance and a new music are needed to give the laghu― guru an undisputed standing equal to that of the other rhythms.

4 June 1934  

Page-153

Gadya-chanda

I can't say that I have studied or even read Bengali , so I am unable to pronounce. In fact what is ? Is it the equivalent of European free verse? But there the essence of the thing is that you model each line freely as you like ―regularity of any kind is out of court there. Is it Nishikanta's aim to create a kind of rhymed prose metre? On what principle? He seems to want a movement which will give more volume, strength and sonority than Bengali verse can succeed in creating but which is yet poetry, not prose arranged in lines and not even, at the best, poetic prose cut into lines of different lengths. All things can be tried ―the test is success, true poetic excellence. Nishikanta has sent me some of his before. It seemed to me to have much flow and energy, but there is something hanging on to it which weighs, almost drags ―is it the ghost of prose? But that is only a personal impression; as I have said, on this subject I am not a qualified judge.

29 September 1936  

Page-154