Your name in 9 Indian scripts

What a transliteration can and cannot do

Indian scripts are not alphabets swapped one for one. They are abugidas built on the same broad inventory of sounds, but the inventories differ: Tamil does not separate the voiced and unvoiced stops the way Devanagari does, Gurmukhi has no ष and writes श with a nukta instead, and several scripts have no independent letter for a vowel another script writes plainly. A converter that pushes through those gaps produces a word that looks right to somebody who does not read the script and wrong to everybody who does.

So the engine carries an explicit rule: where a conversion leaves Devanagari characters in the output, the rendering is dropped rather than displayed. The counts in the table above are that rule, applied to our own corpus and reported. Each script's page goes further and lists the Devanagari letters that script writes with a single letter of its own, with the number of published names each merge touches.

A name is written the way its family writes it. This is a computed spelling offered as a starting point, not a correction of one.

Questions people ask

Why 9 pages for 10 languages?

Because a script is not a language. Hindi and Marathi are written in Devanagari, so those languages share a page and a rendering — the letters are the same letters. Each page is filed under the language most people search with, which is why the Devanagari page is named for a language rather than for the script.

Why is a name sometimes missing from a script?

Because the conversion could not be done honestly. Some scripts have no letter for a sound Devanagari distinguishes, and when a conversion leaves raw Devanagari characters standing in the output we drop that rendering instead of publishing a word with a foreign letter in the middle of it. A page that says “here is your name in Tamil” without saying what Tamil cannot keep is selling a certainty the engine does not have.

How is the conversion done?

The romanised spelling is read into Devanagari first, and every other script is converted out of that. Devanagari is the pivot because it is the script the transliteration rules are best defined for, and it is why the first step is the one that loses the most: several romanised spellings can land on the same Devanagari form. 536 of the 536 published names produce a Devanagari form at all, and each script page shows what its own step then keeps and loses.

Can I use these spellings on a certificate or a passport?

They are a computed rendering, not an official spelling. Whoever issues the document decides how a name is written on it, and families often have a spelling they have used for generations that a converter will not reproduce. Take these as a starting point and a check — particularly for a script nobody in the family reads — rather than as an authority.