A word about Bengali transliteration and pronunciation

A feature of Gadadhar Pran Prabhu's writing is the heavy use of Bengali or Sanskrit words. This has led to conflicts between us when I worked on his books, like the Govinda-līlāmṛta. At any rate, even if I totally surrender to his will, there is still a need for consistency, etc.

Bengali pronunciation differs from standard Sanskrit pronunciation and this is indicated in the impromptu spellings that he uses. To italicize and use diacritic marks throughout his writing becomes overly heavy, so we have decided to change our styles policy and for the most part will treat Bengali words using a natural spelling as though they were adopted English words.

Although Bengali and Sanskrit have fundamentally the same alphabet, there are numerous differences in pronunciation. In Another Side of Thakur Bhaktivinoda, Sanskrit verses, though written in Bengali script and pronounced by Bengalis according to their regional variations, are transliterated according to the Sanskrit transliteration conventions.

Some Sanskrit/Bengali words have already been adopted by devotees -- karma, yoga, bhakti, seva, Nabadwip, Vrindavan, Krishna, Radha, swami, Bhagavan, etc. etc. These will not be changed when used in an English sentence.

However, in wholesale quotations of songs, we are using the standardized scholarly system which has the advantage of being convertible to the Bengali script. The standardized scholarly system (italicization and diacritics) will therefore be used in these cases. We use the utility DiCrunch for conversion (https://scriptoq.com/tools/diCrunch/index.php) so anyone wishing to convert to Bengali script can use this app to do so.

The main characteristics of these two systems are the following;

(1) Bengali a is pronounced like a short o sound like in "box." At the end of most words it is not pronounced, except after conjunct consonants. In standardized transliteration, we use the letter a in all cases.

In some cases, however, such as at the ends of verbs, the sound is longer (ō) In the impromptu spellings, we often use o instead of a. In the standardized spellings we simply use the a. Thus,

বলিব = (standardized) baliba à (pronounced) bolibō à (impromptu) bolibo.

(2) The above can lead to another confusion in the case of long a (আ = ā), since in the impromptu system, the letter a is often used for both short and long a (অ আ).

বর্সানা à barsānā à barsana

Similarly, the impromptu forms of other short and long vowels are generally not distinguished, but they are in the standardized transliteration.

ই i ঈ ī উ u ঊ ū 

Thus কীর্তন à kīrtana à kirtan

(3) Bengali makes no distinction between b and v. So the standardized transliteration system uses only b, whereas due to the influence of other transliteration systems, impromptu will use sometimes one, sometimes the other. Thus,

বাসুদেব = bāsudeba à vasudeva

বৈষ্ণব = baiṣṇaba à vaishnava

The problem here is that in conjunct consonants b is being used where w or v are more familiar, because standardization is for consistency, even though it might not be sounded out. So, 

স্বামী à sbāmī (svāmī)  à swami 

(4) Standard y (য called antastha-ja) is pronounced three different ways in Bengali. At the beginning of a word it is always pronounced like a j. Internal to a word it is pronounced like a y and is written in Bengali with a dot under it (য়). There is a third sound, like a which comes when it follows a real long o (ও) as in words like হাওয়া (hāoẏā), which would be pronounced hāwā. So for the standardized transliteration, we are using ya at the beginning of a word, even though it is pronounced like a j, and internally y with a dot (ẏ), except where compounded.

(5) Retroflex d, ড ঢ in Bengali, like ya, has a beginning and middle form, the latter having a dot underneath ড় ঢ়. For the former, we use ḍa, ḍha, for the latter, ḏa ḏha. The difference in pronunciation is that the latter sounds like a hard ra. In the natural system these can sometimes be written either as d or as r, more often the former. Example:

দৃঢ়  à dṛḏha à dridha

(6) The letters ca and cha (চ, ছ) are pronounced cha and chha, the latter looking strange to an English speaking person. In the standardized transliteration, we use the former, in the ad hoc version we generally use the latter, but Gadadhar Pran Prabhu's preference is to use the former system, especially for names like Caitanya (rather than Chaitanya). So we have acceded to his wishes in most cases.

(7) The letters শ and ষ are śa and ṣa in transliteration, and both are sha in the impromptu version.

(8) Final a is generally dropped in the impromptu version, but always kept in the standardized transliteration.

(9) Nasals (ঙ, ঞ, ণ, ন, ম) have the standard forms ṅa, ña, ṇa, na, ma, but in the impromptu version are all reduced to na or ma. Anusvara (ং = ṁ) is usually converted to a simple n and occasionally ng, and chandra-bindu ( ঁ = ɱ) to n.

So চাঁদ  à cāɱda  à chand

I realize that all this may look complicated and difficult in the beginning, but the difficulties presented by the task make it necessary. The advantages of the two systems is that the standardized form is convertible, while the impromptu form is more easily readable for the non Bengali speaker.

Viz.

hā hā mora gaura kiśora

kabe daẏā kari śrī godruma bane, dekhā dibe mana cora

svānanda sukhada kuñjera bhitare, gadādhare bāme kari

kāñcana baraṇa cāñcara cikure, naṭana subeśa dhari

dekhite dekhite śrī rādhā mādhaba, rūpete karibe ālā

sakhī gaṇa saṅge karibe naṭana, galete mohana mālā

anaṅga mañjarī sadaẏa haiẏā, ei dāsī karete dhari

duhā nibedibe duhāra mādhurī, heriba naẏana bhari

হা হা মোর গৌর কিশোর

কবে দয়া করি শ্রী গোদ্রুম বনে, দেখা দিবে মন চোর

স্ৱানন্দ সুখদ কুঞ্জের ভিতরে, গদাধরে বামে করি

কাঞ্চন বরণ চাঞ্চর চিকুরে, নটন সুবেশ ধরি

দেখিতে দেখিতে শ্রী রাধা মাধব, রূপেতে করিবে আলা

সখী গণ সঙ্গে করিবে নটন, গলেতে মোহন মালা

অনঙ্গ মঞ্জরী সদয় হৈয়া, এই দাসী করেতে ধরি

দুহা নিবেদিবে দুহার মাধুরী, হেরিব নয়ন ভরি