One language, two alphabets

Punjabi in India is written in Gurmukhi. Punjabi in Pakistan is written in Shahmukhi, a Perso-Arabic script. Same language, and most speakers on either side cannot read the other half of its literature.

This is not a small problem. Waris Shah, Bulleh Shah, Shah Hussain and most of the qissa canon sit in Shahmukhi. Everything published in Indian Punjab since 1947 sits in Gurmukhi.

Automatic conversion does not solve it. Gurmukhi to Shahmukhi tools exist and produce over-vocalised text that reads oddly to a Pakistani. Shahmukhi to Gurmukhi on natural, unvocalised text largely does not work, because Shahmukhi omits short vowels and recovering them needs a language model and a lexicon rather than a character map. The one real system is Sangam, built at Punjabi University Patiala, which is a web form with no API and no published accuracy figure.

Jeevay Sanjha Punjab teaches both scripts as a workaround. That is currently the state of the art.

Gurmukhi was standardised by Guru Angad Dev Ji out of the local Landa business scripts. Shahmukhi is Perso-Arabic and is what Punjabi is written in across the border.

The result is that a very large number of speakers of one language cannot read each other. A Punjabi book published at Lahore is unreadable in Ludhiana without transliteration.

Automatic transliteration between the two exists and works reasonably well. What does not exist is an institution on either side with an interest in publishing across the divide.

What we don't know

  • Whether any project is training a proper Shahmukhi-to-Gurmukhi model. Dr Gurpreet Singh Lehal at Punjabi University would know.

Where to check this

Named but not linked. These are real, findable documents. This project has not consulted them for this page.

  • Punjabi University Patiala, Sangam transliteration system.
  • Jeevay Sanjha Punjab, script teaching programme.
  • Testing of available transliteration libraries, August 2026.

This page has not yet been read through by the editor. Treat it with more caution than the rest.

How this page was written. The text is compiled from the standard published histories and reference works, not from a document held in hand. Nobody from this project has visited the place or read the archive. The list above is where a reader can go to check it, not a record of what was consulted line by line. Dates and figures are the received account. If you hold something better, write in and it will be corrected.

This is one of 69 accounts in The Crossing, the chapter of this site covering 1947 and what it did.