[itrans_roman] Request for some extra characters on Indic Romanization from ITRANS Keyboard to cover more Indic langauges.

[milindchakraborty]

I have recently started using the ITRANS Romanisation keyboard but I realise that there is a need of a few more extra characters to correctly phonetically type quite a few Indic languages… These additions will help the keyboard support many languages like Bengali, Assamese, Tamil, Malayalam, Telugu, Marathi, Konkani, Kannada, Gujarathi, Marwari, Maithili, Bhojpuri, Punjabi, Tulu, Hindi, Sanskrit etc with greater accuracy.

If the keyboard incorporates these extra characters, quite a few of which are IPA characters, it will be easier to type not just transliterations but also Indic phonetic pronunciation in them.

1. Tulu and Telugu differentiate between æ and long ǣ sounds. But I can’t type ǣ (U+01E3) on the keyboard.

2. Tamil needs some more characters to represent certain sounds and IPA will suit. θ and ð sounds exist in Tamil and if these are incorporated it would help. The ɣ sound also exists in intermedial positions and is shown with the IPA in ISO often.

3. Malayalam and Tamil have ṟ, ṯ, ṉ which are there but also ḏ (alveolar d) which isn’t there and should be there. It is ‘naṉḏri’ in Tamil. In fact, for Assamese, both dental and retroflex set is ṯ, ṯh, ḏ, ḏh, ṉ (alveolars). Telugu even has a Unicode point associated with the alveolar d sound, ౚ.

4. Tamil Brahmin dialect and Telugu dialects render terminal anuswara as aw̃. So the w̃ letter is necessary. If the nasalised semivowel w̃ is included, then ỹ can be included too. It also happens in Marathi and Telugu when an anuswara is followed by certain consonants, for example saw̃sāra (saṁsāra).

5. The sound β in IPA is present in the ‘hv’ conjunct in Bangla, for example, biu̯βhôl (bihvala), as well as b becoming this medially in fluent speech. This also exists in Tamil medially and should be incorporated.

6. Making vowels half-length by adding - ̯ or ̆ should be incorporated.

7. ɘ, ɨ, ʉ and ɯ exist in multiple languages (as allophones too), especially in the Dravidian group, and are often represented by these IPA symbols while using ISO. They are shown as ú too in ISO sometimes, so can be included. There already is ŭ for this on the keyboard though.
   Tulu even has the long form of the sound so ɘ̄, ɨ̄, ʉ̄, ɯ̄, ú̄, ŭ̄ might be needed too.

8. It would be nice if I could use ɦ too apart from h if I wanted to show the difference comparatively that exists in multiple languages.

9. I don’t know if I can write ‘ɔ’ on the keyboard, if not, this is very necessary for Eastern Languages, as well as Konkani. The IPA one I mean.

10. Assamese ʊ is there on the keyboard but that is often written as ü/û while typing in ISO. This isn’t there on the keyboard. Long forms, ʊ̄, û̄, ǖ.

11. The difference between the three r’s in Indian languages should be available to be shown if needed, trill, tap, and approximant - r, ɾ, and ɹ.

12. Konkani has the vowel sound ɵ and also diphthongs ɵɪ, ɵu. This is probably there in Bhojpuri and Maithili too. The long form may be included too, ɵ̄.

13. Long schwa - ə̄ and long aw - ɔ̄, ô̄ and IPA long vowel marker ':' if incorporated would be nice.

14. We would need nasalised forms too æ̃, ǣ̃, ɔ̃, ɔ̄̃, ỗ, ô̄̃, ɵ̃, ɵ̄̃, ə̃, ə̄̃, ʊ̃, ü̃, û̃, ʊ̄̃, ǖ̃, û̄̃, ɘ̃, ɨ̃, ʉ̃, ɯ̃, ú̃, ŭ̃, ɘ̄̃, ɨ̄̃, ʉ̄̃, ɯ̄̃, ú̄̃, ŭ̄̃,. In fact, please incorporate - ̃ - nasalisation marker, instead of adding every nasalised form of every vowel, which could be used for the IPA vowel set already there too.

15. The sh sound is often voiced in speech to ʒ in Bangla when it precedes a voiced consonant and this voiced sh happens in a few languages, corresponding to the sound in ‘measure’. This being included in the dictionary would help.

16. Since most Indic languages have lost the distinction between palatal/post-alveolar sh and retroflex sh, post-alveolar sh is often shown as š. This is a very useful character if incorporated.

17. Some show the c sound in Marathi and Konkani as č and ċ/ĉ depending on whether it is c or ts. ch is therefore čh and ċh/ĉh.
    For j they use ǰ and ż/ẑ/ĵ based on whether it is j or dz. jh being ǰh and żh/ẑh/ĵh.

18. Also, the c being shifted to the ts sound in Telugu, ౘ, is given as ĉ in the keyboard. Similarly, when j becomes dz in Telugu, ౙ, it should be rendered as ẑ; this should be incorporated.

19. The ɸ sound exists as an allophone to ph/f in many languages, especially eastern dialects of Bangla. This IPA character can be included too.

20. Similarly, if we could use w, v and ʋ separately if we compare languages it would be helpful.

[mcdurdin]

@Shreeshrii: as this is your keyboard, would you like to respond to this feedback?

[MakaraSok]

Original post: https://community.software.sil.org/t/indic-romanization-from-itrans-keyboard-inclusion-of-extra-characters/7578