আলাপ:কেন্দ্রীয় স্টকহোম

Following the guidelines in উইকিপেডিয়া:বাংলা ভাষায় বিদেশী শব্দের প্রতিবর্ণীকরণ, I removed some odd-looking combinations of characters, and added হসন্ত as needed. In particular, I removed the use of হ্‌ও as a representation of "sk", replacing it with হো - basically, in normal Bengali spelling, it makes no sense to have a consonant followed by a হসন্ত and then a full vowel. Even if this destroys the mapping between Bengali and Swedish, it's more important to make a Bengali-friendly word, and then make a phonetically-oriented version in italics (all the weird spellings can be put in here). I also changed দীর্ঘ-ঈকার to হ্রস্ব-ইকার as this is the typical way to represent /i/ in foreign words, regardless of the actual vowel length in the foreign language (e.g. "টিচার" for "teacher", for পার্টি "party" etc.). I hope it's all clear. Let me know if you have any comments about this! --সামীরুদ্দৌলা ০৭:৫৮, ৬ নভেম্বর ২০০৬ (ইউটিসি)

I also removed the "রেফ" from "র্ট" etc. following our discussion of the coalescence of "r" with coronal consonants. --সামীরুদ্দৌলা ০৮:০০, ৬ নভেম্বর ২০০৬ (ইউটিসি)

কিন্তু Farsta'র উচ্চারণ Fashta (ফাশ্‌টা) এর মত নয়। আসলে exactly r-এর উচ্চারণ হচ্ছে না, কিন্তু একটা দীর্ঘ আ-এর মতই হচ্ছে। এইটার ব্যাপারে কি করব? --Docwho (চিনাৎসু) ১৮:১০, ৭ নভেম্বর ২০০৬ (ইউটিসি)

This is common in Germanic languages. Since we can't really represent vowel length that well in Bengali, my inclination is to just ignore this (long [a], which is sort of like what you're describing and how I've heard it as well, is not something Bengalis can reliably differentiate from short [a] anyway). If you want, you can do what I did for the Danish pronunciation of Denmark (Danmark ডান্মাআক). Or you can leave it as is. --সামীরুদ্দৌলা ০৭:৩৬, ৮ নভেম্বর ২০০৬ (ইউটিসি)

I am not sure about the র্ট -> ট. Because (I think that) when we are writing English words in Bengali such as party, sort, park, dark etc. we write পার্টি, সর্ট, পার্ক, ডার্ক etc. and not পাটি, সট, পাক, ডাক as they are pronounced. This is also written with a long [a] sound. Any comments? --Docwho (চিনাৎসু) ১৫:২৩, ১২ নভেম্বর ২০০৬ (ইউটিসি)

Maybe I'm biased for being American, but I do see the point in writing the রেফ in র্ট (পার্টি, ডার্ক, etc.) because I actually pronounce them. In the case of Swedish, the Swedish phonology article on En-Wiki indicates that the standard forms of Swedish in Central Sweden and Finland both have total [r]-[coronal stop] coalescence into retroflex sounds (while the Southern Swedish standard does not). In German as well, all dialects except in the southwest, standard or non-standard, drop the [r] in certain positions (like Berlin বেয়ালিন). Now, I realize the di(tri-?)chotomy here - for English, we choose to write everything with the র included, even though many people do not pronounce it this way, and for German we have been writing everything with র in version 1 of the Bengali pronunciation of the name (বার্লিন) but without in version 2 (বেয়ালিন), and for Swedish and Danish we have been writing everything without র for both version 1 and 2. In terms of version 2, to be consistent with En-Wiki we should leave out the /r/ when it is not pronounced in the native pronunciation (note the pronunciations given for Birmingham, Berlin and Copenhagen). In terms of version 1, however, we need to formulate a rule. Do we include the /r/ in version 1 and then show how it is not pronounced in version 2? Do we leave it out in both? I think this will unfortunately have to be language-specific. As far as I've seen, English words have always been translated with the /r/ included, so version 1 should keep the /r/ for English... and for German? We have an /r/ for বার্লিন, and many news agencies including Deutsche Welle keep it in for names like আঙ্গেলা মের্কেল/মার্কেল/ম্যার্কেল and ফ্রাংক-ভাল্টার ষ্টাইনমায়ার. So it looks like German and English would follow the same pattern. But Swedish? Do we keep it on in version 1 (translation of Swedish) and then remove it in version 2 (phonetic)? I am not against the idea, since other Germanic languages do it the same way... however Swedish is a different case. In German and English, the /r/ simply disappears, possibly lengthening the vowel (something that cannot be easily represented in Bengali letters), while in Swedish, the /r/ fuses with the following consonant to change র্ত into ট, র্দ into ড, র্ন into ণ, etc. To me, it seems like if we are already showing the coalescence by using the retroflex letters ট, ড, ণ, etc, to represent the phonetic form (as opposed to the underlying form) it's redundant and/or hypocritical to also include the র, which is not pronounced separately from those letters, and which only belong in some underlying abstract form of the pronunciation. We're getting into the foundations of theoretical phonology here, so I'm going to stop soon, but my point is that the use of ট ড etc already shows that there was an /r/ in the word - then redundantly adding the /r/ to make র্ট র্ড, etc is overkill. Either we stick to some underlying/phonemic form like র্ত, র্দ, etc (having different spellings for versions 1 and 2) or the phonetically realized form ট ড (for both versions 1 and 2). Thoughts? --সামীরুদ্দৌলা ২৩:১৩, ১২ নভেম্বর ২০০৬ (ইউটিসি)