I Won't Let You Go (8 page)

Centenary edition of Tagore’s collected works (1961). I also
consulted
the earlier Visvabharati edition of the collected works on many occasions, whenever I had any doubts about the
interpretation
of a particular line or phrase or about any punctuation, and especially for compiling the notes, but nowhere have I noticed any significant discrepancies. The books from which the poems are taken are arranged chronologically, as in the collected works; even when two books have been published in the same year, we know which comes first because the month of publication is known. The sequence of poems from a particular book corresponds to Tagore’s own intentions.

Tagore in January 1940, a year and a half before his death.

A note on one of the translations may be of interest to readers. During my student days at Oxford in the early sixties, friends sometimes urged me to translate Tagore. Looking back, I think their curiosity might have been provoked by media references to the centenary of his birth in 1961. I did translate a few things, one of which, through the assiduity of an enterprising friend, eventually found its way to the Dublin magazine
Poetry Ireland,
where it saw the light of day in 1964. ‘The Victorious Woman’, included in this anthology, is a revised version of that early effort. Richly descriptive and sonorous, it had been an ambitious choice for those days. I am sure I had been unconsciously goaded to show off the pedigree of my native tongue, for two of the most irksome questions that I was frequently asked were: first, whether, seeing that I was studying English literature and spoke English with
reasonable
fluency, I had not really always spoken English at home in my childhood, and secondly, whether Bengali could as yet be
called
a language, properly speaking, or was it not still in the stage of a dialect. (Dialects had not acquired class in those days. And Macaulay’s ghost was tenacious.)

I have also ventured to include some songs, all taken from
Gitabitan
, the standard collection of Tagore’s songs, which is also included in the collected works. The songs have been arranged chronologically, following the chronology given in a standard work of reference,
11
which has been kindly re-checked by Professor Sankha Ghosh. Some of the songs also occur in other books, verse-collections or plays. Such information will be found at the end of each song. Thus those who may wonder why I have not translated anything from the famous
Gitanjali
will find their
answer
in the section of songs: two songs from the Bengali
Gitanjali
(1910) have in fact been translated. Songs have also come, via
Gitabitan
, from
Gitimalya
(1914) and
Gitali
(1914). But I must
emphasise that Tagore’s songs have very strong independent lives as songs. That is how most Bengalis approach
Gitabitan
– as an anthology in its own right, in which each piece happens to be a song. All the songs have therefore been translated from
Gitabitan
, as individual songs and discrete pieces. Each song has been
treated
as a text in its own right.

I am aware of the opinion expressed by Radice in his
Selected Poems
of Tagore that songs cannot be translated. In so far as this means the obvious, that a song is made up of its words and its melody, and the melody cannot be translated, one necessarily agrees, but granting that the translation of all poetic material is a difficult and delicate task, the translation of the lyric of a song is no more difficult than that of a poem which has not been set to music. Indeed, because of its structure, the lyric of a song may be much easier to translate than a complex poem. A rigid division between poems and songs cannot be maintained, certainly not in the Indian context, where a considerable overlap between the two modalities has traditionally been taken for granted by artists as well as audiences. The
bhajans
of Mirabai are both poems and songs. The Baishnab lyrics from the “medieval” period of Bengali literature are both poems and songs. It is clear from Tagore’s own use of words that he often thought of the roles of the poet and the singer/songmaker as interchangeable, and surely this is true of many other cultures as well, much of folk poetry being also folk song, and vice versa.

I have always been attracted to songs in different languages, and I believe that like poems, they can be marvellous
introductions
to a new language and genuine aids to language-learning, making it a pleasurable experience, a fact which does not seem to be much appreciated by modern language teachers in British schools. Being simple-structured poems, songs exhibit the “works” of the language like a device in a crystal jar. I remember the
pleasure
I used to get as a young girl from the 78 r.p.m. records of French songs my father used to borrow from the Alliance Française de Calcutta, and in later life I learned to follow Spanish by comparing the Spanish texts of songs with their English
translations
on the sleeves of records. Indian children who are not born to Hindi or Urdu learn them from Bombay film songs, and all over the world youngsters today pick up lines of English from pop songs. I have also enjoyed translating songs from other
languages
into Bengali, for instance, Shakespeare’s ‘When icicles hang by the wall’ and the folk songs of the Judaeo-Spanish
Ladino tradition, and the results have been appreciated. These experiences have given me the confidence to attempt the
translation
of a few of Tagore’s songs, without which, to my mind, a selection such as this would not be complete, because it is through his songs that Tagore speaks,
as a poet,
to his widest audience in Bengali, and no Bengali party is really ever complete until some Tagore songs have been sung. I have kept the songs together in one section so that readers do not forget that the originals do have melodies and are meant to be sung, but I hope my renderings have also captured something of their swing and lilt.

Capturing the form and the content together when one is
trying
to translate poetry into poetry (and not into prose) is
admittedly
the hardest task. By allowing my selection to proceed along the free-flowing route outlined above, I have spared myself
unnecessary
struggles with structures that refuse reincarnation. Some poems are simply untranslatable. There is no need to wrestle with these when one is trying to introduce the poet to a new audience, because there is a whole range of excellent poems which can, with a little sensitive effort, be given an English form. Given the
structure
of a poem in the original language, the aim is to create a
parallel
or corresponding poetic structure in the language into which one is translating, using the various sonic devices available in the latter. The two structures will not be ‘equal in all respects’ like the congruent triangles we cut out of paper in school geometry lessons, but there should be a certain resemblance or
correspondence
between the two. It is an act of approximation, a dance of interpretation, making good use of the area of overlap between the two languages in meanings, sounds, moods, suggestions, and so forth. A good rule of thumb in practice is to adjust to the natural rhythm and cadence of the language into which one is translating by means of many micro-decisions, while
stretching
the capacity of that language by allowing it to mirror slightly alien patterns of thought. The meandering free verse or prose poetry of Tagore’s later years is naturally fun to recreate, but I have also attempted a fair number of poems with much tighter structures. The language of Tagore’s poetry is exceptionally rich and musical; I hope I have succeeded in conveying something of these qualities in the
translations
.

A fact of which readers should be aware is that Bengali does not distinguish between masculine and feminine forms in its third person pronouns, between
he
and
she
, between
him
and
her
, between
his
and
her
. Naturally, most of the time the context tells
us quite unambiguously how to interpret the signs, but sometimes no such help is available. Nor does Bengali have gender-markers in verbs, as some other Indian languages do. In Tagore’s lyric poetry and songs one encounters certain instances of gender-ambiguity which make one suspect that he has in fact used the “unisex” third person pronoun in a deliberate manner to create an
atmosphere
of refined poetic ambiguity. It is as if he is asking us to forget the
he
and
she
and to concentrate on the essence of the human situation. It is a great pity that this androgyny, which
contributes
substantially to the subtlety of the original pieces in which it occurs, cannot be captured in translation. A translator has no option but to reach out for either a
he
or a
she
. I have tried to do my best, using my instincts as a Bengali and looking
carefully
for cultural clues in the inner landscape of each text. But what this means in effect is tuning oneself to respond to certain conventional cues, ignoring the rich possibilities of an
unconventional
interpretation. And as soon as a choice is made, which can sometimes be alarmingly automatic (so powerful is the hold of gender-stereotypes on our minds), the possibility of the alternative interpretation is blocked and the androgynous power of the
original
inevitably destroyed. This is an intractable problem. No
matter
how carefully we proceed, taking all kinds of other factors into account, such as connections a particular text may have with other texts, the fact remains that an ambiguity within a particular text is still an ambiguity within that particular text. It is embedded there, was perhaps consciously put there by the poet, and is an integral part of the poetic gestalt of
that text
. A poetic ambiguity
destroyed
is a poetic ambiguity destroyed, and there is really no way in which one can compensate for it. Perhaps some readers will enjoy the intellectual exercise of spotting for themselves the pieces where this specific form of ambiguity occurs in the originals.

In the transliteration of Bengali names and words I have
adopted
a working compromise. Without considerable modification, the Roman script is not suitable for the representation of most Indian languages, because it simply does not have enough characters to cope with the needs of these languages. To mention just a few problems, it cannot readily distinguish between the short
a
and the long
a
, it cannot aspirate except by the addition of a whole
character
(
h
), and it cannot distinguish between dentals and cerebrals. Thus Bengali has a separate character for each of the following: dental
t
and
d
, cerebral
t
and
d
, and an aspirated version of each of these, eight characters in all. If we use the resources of the
Roman script in the way as it is used in English, we have just
t, d, th, dh,
and no way to distinguish between the dentals and
cerebrals
at all, a distinction which is vital to Bengali and other Indian languages. To counteract these and other problems, an elaborate system with many additional diacritical marks has been developed by Sanskritists for the romanisation of Sanskrit. This system, however, is not really very appropriate for Bengali, where
orthography
and pronunciation are not in the same accord as in Sanskrit, and apart from the inconvenience that non-specialists do not know how to interpret its signals, the system constantly gives rise to distorted ideas about the pronunciation of Bengali names and words. One could develop a new system altogether, which would do more justice to the sound of Bengali, but it would leave us with the following problem. A certain convention about the romanisation of Indian names and words has already grown up over the years, not the complicated system used by Sanskritists, but for general use, without diacritical marks, but still rather spelling-oriented. Introducing a new system for Bengali all of a sudden would be a source of confusion, disorienting readers, both Westerners and non-Bengali Indians, who have become familiar with certain transliterations. Drastic changes in the spellings of names of authors or book-titles would present problems of
recognition
in bibliography. So unless Bengali studies came into vogue in English in a big way, a new system would be a pointless
exercise
. I have therefore adopted the system in general use as the basis of my transliterations, adding slight tilts, wherever I have felt it to be necessary, to the real Bengali sound-values. This means that there are inconsistencies in the way I have done things, but I reckon it is the lesser evil. I have tended to leave certain conventional spellings well alone, to avoid problems of recognition, but have felt freer to adapt to Bengali pronunciation in other cases, such as the names of trees and flowers within poems. I have also refrained from italicising such names in the poems, to avoid giving them a false exotic air. They are not aliens in the contexts of the poems, so should not stand out when one is reading the poems. The place where they should stand out is the Glossary. I would like to point out to readers the Bengali
pronunciation
of two important names: Rabindranath is pronounced ‘Robindronath’, with the
d
soft, as in French and Spanish, and Santiniketan is pronounced ‘Shantiniketon’, with the
t
soft, as in French and Spanish.