Gurmukhi has nine vowel marks, called the laga matra (ਲਗਾ ਮਾਤਰਾ), which change a consonant's built-in short "a" into another vowel. There are ten vowel forms in total, but one, mukta, is simply the absence of a mark. So ਕ on its own is ka, ਕਾ is kaa and ਕਿ is ki.
I knew the 35 Gurmukhi letters for nearly a month before I could read a single real word. I could recite them like a child reciting the English ABCs, in order, smug about it, and still a road sign in Amritsar was a wall of decoration to me. What I had missed was that the letters are only half the system. The other half, the half that actually does the reading, is the vowel marks. In Punjabi they're called the laga matra (ਲਗਾ ਮਾਤਰਾ), and once I learned them, the script went from locked to legible in about a week.
This is the post I wish someone had handed me back then. If you haven't met the consonants yet, read The Gurmukhi Alphabet: All 35 Akhar first, because everything below sits on top of them.
Why the vowel marks matter more than the letters
Here's the rule that reframes the whole script: in Gurmukhi, a bare consonant is not silent. It carries a built-in short "a" sound, called the mukta. So ਕ on its own is already ka, not just k. ਘ is gha. Put two of them together, ਘਰ, and you've read ghar, "house," with no vowel marks at all.
That means the laga matra exist to do one job: to change that default "a" into any other vowel. Attach a mark, swap the vowel. No mark, and the "a" stays. Reading Gurmukhi is, at its core, just this: read the consonant, check for a mark, adjust the vowel. Everything else is detail.
What are the nine laga matra?
There are ten vowel forms in Gurmukhi, but one of them (mukta) is the absence of a mark, which leaves nine actual marks to learn. They sit above, below, beside, or wrapped around the consonant. I'll show every one on ਕ (k) so you can see exactly where it lands.
| Form on ਕ | Name | Sound | Real word |
|---|---|---|---|
| ਕ | mukta | ka (short a) | ਕਲ kal, "yesterday/tomorrow" |
| ਕਾ | kanna | kaa (long a) | ਨਾਮ naam, "name" |
| ਕਿ | sihari | ki (short i) | ਦਿਨ din, "day" |
| ਕੀ | bihari | kee (long i) | ਘੀ ghee |
| ਕੁ | aunkar | ku (short u) | ਗੁਰੂ guru |
| ਕੂ | dulainkar | koo (long u) | ਖੂਨ khoon, "blood" |
| ਕੇ | lavan | ke | ਮੇਰਾ mera, "my" |
| ਕੈ | dulavan | kai | ਪੈਰ pair, "foot" |
| ਕੋ | hora | ko | ਲੋਕ lok, "people" |
| ਕੌ | kanaura | kau | ਕੌਣ kaun, "who" |
Notice the pairs. Kanna is just a long version of the mukta a. Sihari and bihari are short and long i. Aunkar and dulainkar are short and long u. If you learn them as three families (a, i, u) plus the e/o set at the bottom, your memory has far less to carry than nine separate facts.
The sihari trick that trips up every single beginner
If you only take one thing from this post, take this. The sihari (the short-i mark, ਿ) is written before the consonant but read after it.
Take ਦਿਨ. Your eyes see the little hook sitting in front of the ਦ, so your brain wants to say i-da-na, and what comes out is gibberish. I did this for a solid week, getting furious. The fix is to stop reading by position and start reading by rule: the sihari always belongs to the consonant on its right. So ਦਿਨ is da + i = di, then na: din, "day."
Once that rule is in your hands, it never trips you again. Every beginner hits this exact wall once. Now you'll hit it knowing what it is.
Where do the vowel marks sit?
Position is part of how you read the mark, so it's worth naming:
- Above the letter: kanna sits to the right as a vertical stroke; lavan, dulavan, hora, kanaura sit on top.
- Below the letter: aunkar and dulainkar hang underneath.
- Beside / wrapped: sihari (left) and bihari (right) stand as vertical strokes flanking the consonant.
You don't need to memorise this as a chart. After fifty words it becomes muscle memory, and you'll stop seeing "a mark above the line" and start seeing "o". Not everything under a letter is a vowel, though: a few consonants also shrink into subscript conjunct forms.
Read these the slow way
Let me walk three words exactly how your brain will do it for the first few months, because this slow, deliberate decoding is the real skill:
ਪਾਣੀ: ਪ (pa) + kanna (→ paa) + ਣ (ṇa) + bihari (→ ṇee) = paaṇee, "water."
ਸੁਣੋ: ਸ (sa) + aunkar (→ su) + ਣ (ṇa) + hora (→ ṇo) = suṇo, "listen."
ਮੇਰੀ: ਮ (ma) + lavan (→ me) + ਰ (ra) + bihari (→ ree) = meree, "my" (feminine).
That's the entire method. Consonant, then mark, left to right, with one eye out for sihari hopping back a slot. Speed is just this, repeated until it's automatic. When you're ready to chain these into full words and sentences, the step-by-step reading guide takes you the rest of the way.
The part a page can't teach
I can lay out every mark on a page, but I can't hear you. The laga matra are where reading and speaking finally meet: the kanna in ਨਾਮ has to actually come out long, the aunkar in ਗੁਰੂ short, or the word doesn't land. On the page they look like decoration. In your mouth they're the difference between naam and nam.
That ear-training gap is exactly why I built Panjabi Go to pair every vowel mark with native-quality audio, so you hear the long and short vowels back to back, and it gives you pronunciation feedback when you read aloud. Short daily lessons introduce the matras roughly family by family, the way they finally stuck for me. But even just learning these nine marks well will change how the whole script looks to you. Mine went from wallpaper to words the week I learned the laga matra.
