Spell It Out: Should English Transcription of Indian Words Be Reformed?

I remember walking down a busy street in Delhi during my first stay in India and noting the different ways in which the same words were written in Latin script. Lakshmi here, Luxmi there. Since then I have stopped paying attention to these details, accepting it as it was, as a part of contemporary India’s reality. Still, a question needs to be asked – should this rendering be reformed and unified?

The British colonial government never worked out a coherent and precise way of spelling Indian names “in English” – in the sense of representing them in the Latin alphabet with the use of English phonetics for the sake of common form (not in academic systems). While Lucknow sounds like an auspicious name, the city’s name, if we use an academic transliteration, is Lakhnaū. More importantly, certain names had been terribly maimed, such the surname of the famous Bengali poet, thinker, and novelist Rabindranath Thakur, who became Tagore.

It would seem that, by and large, the “method” was just picking what sounded like a nearest approximation for the transcription without researching the academic basis for this – and without a centralized attempt to construct a fully coherent system. Lucknow was, in perspective, not represented that badly when one remembers that the historical region of which this city was the center was often spelled as Oudh, but sometimes also as Oudhe or Oude. None of these variants is close to being correct, as the territory was called Awadh by Indians.

Naturally, the “English spelling” could not adequately represent the Indian sounds that the English language does not have, such as retroflex t and d. While two cities, Patna and Satna, are written with the same ‘t’ in English, the original sounds are in fact quite different. The name of the famous Indian company, Tata, is pronounced much differently from what this easy spelling suggests: both “a” are long and both “t” are retroflex. Similarly, both the softer “ś” sound and the harder “ṣ” [sh] must be represented by the same “sh” in English. For instance, “Shahrukh Khan” contains the first sound while “Lakshmi Narayan Mittal” contains the second — you would not know the difference in the English spelling.

Another lack of coherence appeared on the line between transliteration (representing the script) and transcription (representing the sounds). In a vast majority of cases, the non-academic English spelling of Indian words preferred the latter. There is inconsistency also on this front, however. There is a city named Gujrat in Pakistan and a state named Gujarat in India. It is actually the same name, only that the former spelling represents the pronunciation and the latter represents the way it is written (in Gujarati as well as Hindi). 

Similarly, the pronunciation of the Indian prime minister’s first name is actually Narendr, not Narendra (in Hindi, at least). This is because the short “a” at the end of Hindi words is most of the time no longer pronounced. It used to be in the ancient times of Sanskrit – and hence we should use it when we refer to those contexts. Thus, in English we should write Krishna or Rama because they are old Hindu gods and because their names come to us from Sanskrit epics. But modern Indians pronounce them as “Krishn” and “Ram.” If we would like to represent the sounds, therefore, Narendra Modi should be spelled Narendr Modi, and if we would like to represent the script, then Narendra should be written Narendra — but in that case Mohandas Gandhi should have been spelled Mohanadasa Gandhi.

What we refer to as the “English” versions of Indian words and names is, therefore, an unscientific rendering based on a chaotic, mostly colonial tradition. This spelling, it should be added, has also mostly not been corrected (except for certain city name changes) or standardized during the decades of independence. While some academic transcriptions and transliterations have been worked out, they have not gained much currency outside academia.

At the same time, English is one of the official languages of India and its usage in the country is immense – from politics through courts to academia and private schools. An Indian must therefore often write Indian terms “in English” – including his name on various documents. Thus, depending on what spelling a particular family has been using, one Indian will write his surname as “Rajput,” the other as “Rajpoot,” despite the fact it is the same surname.

Rearranging this mess into one orderly system would dispel a lot of confusion and signal a closer pronunciation – both to foreign readers as well as Indian users (for instance, for non-Hindi Indian speakers who cannot read Hindi in the Devanagari script, for Hindi speakers who cannot read or speak Bengali, etc.). Moreover, instead of inventing fire anew, authorities could pick one of the existing academic transcription and transliteration systems as the basis of their reform. The point about the “English spelling” being unable to represent certain Indian sounds cannot be regarded as an impenetrable obstacle here. Even outside academia a commonly used script can be adapted to new sounds: new letters may be added, usually through a modification of existing ones. 

When the Roman script was being adapted to represent the Polish language, for instance, a whole set of new letters was created on the basis of existing ones (e.g. ą, ć or ż). Similarly, when Persian, Arabic, and Turkish words started to appear in Indian languages, new symbols were created in the Devanagari script to represent the new sounds that came with them (for instance, k with a dot underneath was introduced to represent q). The same has not been done to represent Indian sounds in English, although it is graphically feasible.

The counterpoint to the idea of such a reform is that it would require a massive, highly centralized, and somewhat costly effort. Forget about such basic questions of whether transcription or transliteration should be used – more importantly, for all Indians to benefit from this equally, the new spelling system would have to represent all sounds and phonetic peculiarities of all Indian languages, or at least the major ones. Countless issues of various pronunciations (such as dialectal forms) would have to be resolved – or ignored, causing potential small controversies or at least retaining certain inconsistencies. Technological troubles would appear in the world of printed and digital word: for instance, English-language newspapers published in India would have to print letters with new diacritics.

And then sweeping changes would have to be made across the country – from altering the name plates or road signs to issuing new documents, such as voter cards, to citizens. The Republic of Korea proved that such a massive reform is possible when it changed the official Romanization of the Korean language, forcing itself to change names of city stations, maps, road signs, and many others. But South Korea is a much smaller, highly-organized, and rich country with one dominant language. The challenges India would have to face to achieve the same result would be far greater. 

Thus, as much as I would like to see the “English spelling” of Indian words be reformed and standardized, I am aware that New Delhi has bigger problems ahead of it now and in the coming decades. I can only hope that the matter will be taken up at some distant time in the future.