Roman Script for Indian Words

September 12, 2019

What’s wrong with English?

English is a fine language in its own context. But when it comes to writing Indian words, English phonetics fail us. Take the word India for example. It should ideally begin with a soft “thi” (as in this), but in English it’s pronounced as “di” (as in did). Over time, Indian words transliterated into English have lost their original phonetic integrity.

Even the Wikipedia page on India doesn’t mention the IAST form of the name. I’m tired of pronouncing Thamizh as Tamil. Increasingly, Indians are pronouncing words based not on native scripts, but on their English spellings—a disturbing trend for a phonetic language system like ours. As native sounds fade from memory, we risk losing the soul of our languages.

Consider the following examples:

English	IAST	Devanagari	Tamil
India	Inḍiā	इंडिया	இந்தியா
Modi	Moḍī	मोदी	மோடி
Tamil	Tamiḻ	तमिऴ	தமிழ்

International Alphabet of Sanskrit Transliteration (IAST)

Indian languages are inherently phonetic. Sounds (phones) are the soul of language, and scripts were developed to represent those sounds.

Over time, various Indian scripts—chief among them the Brahmic scripts—evolved to capture these sounds. Today, Devanagari is among the most widely used.

To bridge Indian languages and global readership, Roman transliteration became necessary. The IAST system, formalized in 1894, was created by Western scholars to provide a faithful, reversible mapping of Indian scripts using Roman characters. It was designed with scholarly precision, enabling readers to reproduce original sounds accurately.

ISO 15919

Tamil is not romanized using IAST, but using ISO 15919. So there’s a hegemony of their own arbitrary directions.

What’s wrong with IAST or ISO 15919?

While these are significantly better than standard English spellings, they are not perfect and are still not intuitive to the average Indian speaker unfamiliar with its rules.

Examples

The Tamil sound த is romaized as t but many read it as ட.
The Hindi दी in Modi becomes Mōdī in IAST and Modi in English—but Tamil speakers often read it as Mōḍi.

Until around 2015, Tamil media used to write “Dhoni” as டோணி - an example of phonetic misunderstanding.

Moreover, Sanskrit-derived scripts typically lack short vowels like e and o, which are essential for correctly writing Dravidian languages. For example:

Tamil: ondru
Kannada: ondhu
Telugu: okati

In IAST, the o in these would typically be misread as long, due to the Sanskrit-influenced vowel system.

What Should Be Fixed in IAST and ISO 15919?

Use IAST for Tamil as well. Additionally, implementing these corrections would make IAST more inclusive, phonetic, and accessible for people of all Indian origins, as well as for future generations of children growing up outside India:

Swap T and Ṭ: T should represent ट, and Ṭ should represent त.
Swap D and Ḍ: D should represent ड, and Ḍ should represent द.
Use e for short e (as in egg), and ē or ee for long e (as in ace).
Use o for short o (as in pot), and ō or oo for long o (as in door).
Always use lowercase letters when writing Indian languages in Roman script, to remove ambiguity between aspirated/unaspirated and voiced/unvoiced pairs.
All the hard consonants in Tamil க,ச,ட,த,ப must use their phonetic equivalents when they come in the middle of the words instead of blindly using the primary equivalents. This is because unlike other indian langauges Tamil uses a single character to represent multiple sounds. In that sense, it is phonetically inconsistent.
- For instance, Tamil words written in IAST are using Ṭ when ட comes in the middle - this is wrong. Eg: கொடுத்தான் is currently written as koṭuttāṉ. It should’ve been written as koḍutthān. Even that must be changed as per earlier suggestion to be written as koduṭṭāṉ.

Should India Use a Common Script?

Ideally, India should move towards a phonetically complete, script-neutral system. I am tempted to recommend a pan-Indian adoption of the Malayalam script (Chera kings were huge proponents and benefactors of Tamil language), which accommodates almost all phonemes across Indian languages. But in all honesty, the ancient origins of Tamil scripts define the history of Tamil people and culture. And due to practical and political challenges - a unified script is unlikely in the near future.

Therefore, the next best solution is to modernize IAST and make it part of our curriculum. This can help students grasp pronunciation more accurately and preserve linguistic diversity.

Some argue this might threaten native scripts. But the widespread use of English already poses a greater risk. Rather than ignoring this, we must act consciously to preserve our linguistic heritage.

For Tamil speakers: the word தமிழ் is beautifully nuanced. Calling it “Tamil” does injustice to its phonetic richness. We must raise awareness and build consensus to promote a more faithful Romanization system.

If not IAST, then we need to work together to define something better. Remain silent and watching our languages fade into approximation is the last thing we want.

Final Thoughts

Language is identity. Scripts are memory. Sounds are emotion. Let’s not lose our languages by mispronouncing them through foreign eyes. Let’s update our transliteration systems to do justice to our roots.