Malayalam to English Transliteration

Manglish and transliteration are not the same thing

Manglish is Malayalam typed in English letters the way it sounds to the writer, with no fixed rules: sughaano, sukhamano and sukhamaano are all the same question written three ways. It is how people message each other, and it works because a Malayali reader can hear it.

What this tool produces is a standard romanization, where each Malayalam letter has exactly one Latin equivalent. The result looks more formal and carries marks Manglish does not use:

The difference matters in one direction more than the other. From malayāḷaṁ you can rebuild the exact Malayalam spelling, because ḷ and l are different letters and ṁ records a specific sign. From malayalam you cannot, because three Malayalam letters have collapsed into one Latin letter. That is why this tool keeps the marks.

A correction we had to make ourselves

Malayalam is the one script on this site where we do not pass the standard conversion data straight through, and it is worth saying why in public. The Unicode rules leave the two vowel signs of the "o" series as their old two-part spelling instead of romanizing them, which produces output that is not valid in any system:

Those are Kochi, Kozhikode and Kottayam, so the defect hits ordinary words and most of the state's place names. We substitute the two affected sequences and nothing else. On a sixteen word test set the unmodified data gets ten right and our output gets all sixteen. The fix is documented on the About page along with everything else we change. Checking the output of a standard against real words, rather than trusting it, is the only way these things get found.

The chillu letters

Malayalam has a set of letters that exist only at the end of a syllable, written in a shortened form called chillu. They are not the same as the ordinary consonant, and the romanization keeps them apart by dropping the vowel rather than by adding a mark:

This is why Malayalam words ending in a consonant romanize cleanly, with no trailing vowel to delete. Hindi needs that deletion and we apply it; Malayalam does not, because the script already records where the vowel stops.

The letters that need marks underneath

Malayalam separates three l-sounds, three n-sounds and two r-sounds where English has one of each. The dots and lines are not decoration: they are the only thing telling the letters apart.

ഴ is the sound in കോഴിക്കോട്, kōḻikkōṭ, the city written Kozhikode in English. The zh in that spelling is an English approximation of a sound English does not have, which is also why Kozhikode and Calicut ended up as two names for one place.

If you came here to type Malayalam, read this first

Searches for "English to Malayalam transliteration" are usually looking for a Manglish typing tool: you type namaskaram and you get നമസ്കാരം. That is the opposite direction from this page, and we have not shipped it, because the easy version of it is wrong most of the time.

Everything in the table two sections above runs backwards here. English spelling does not separate ല from ള from ഴ, or ന from ണ, or ര from റ, so a Manglish word has several valid readings and only one is the word you meant. A letter-mapping table cannot choose between them, because what you typed does not contain the information needed to choose.

The tools that do this well, Google's input method and the open-source GoVarnam among them, generate every candidate the table allows and then pick between them with a dictionary of real Malayalam words. We tested the table-only approach on Tamil, which has the same problem, and it scored three correct words out of ten. That is why this direction is missing rather than approximate.

Where the conversion data comes from

We are not linguists and we do not invent romanization rules. Malayalam conversion runs on the Unicode Common Locale Data Repository transliteration data through PHP's Transliterator class, the same ICU implementation browsers and operating systems use, following ISO 15919. The one change we make is the o-vowel correction described above, and it is the only one. Every adjustment we apply to any language is listed on the About page.

What this tool does not do

• It does not translate. നമസ്കാരം comes back as namaskāraṁ, which is how it is written. What it means is "greetings", and that is a translator's job.
• It does not type Malayalam for you. Manglish to Malayalam needs a dictionary-backed input method, not a conversion table.
• It does not produce Manglish. You get the standard romanization with its marks, which is the more precise thing but not the one people text each other in.
• The au diphthong is a known gap. The vowel signs ൌ and ൗ are lost by the conversion data in a way we cannot recover from its output. They are uncommon, and we would rather say so than quietly guess.