The Elegance of the Devanagari Script

The meaning of the rows and columns in the Devanagari script

The Devanagari script was not taught well in my school. I never understood why श and ष existed separately if they were pronounced the same. (The difference in pronunciation was never explained.) My 2nd std teacher explained ञ as the sound that a small baby makes when crying. And ङ is the sound of a newborn infant crying. And we continued to pronounce both of them like न. And what is ऋ? Is it pronounced “ri” or “ru”, and in any case, why does it even exist if री and रु exist?

Much later, when I understood the meaning of the rows and columns in Devanagari, everything fell into place and I saw the beauty and elegance of it all.

So let’s look at the rows and columns.

I’ve intentionally left out everything else for reasons that will become clear later.

Have you ever wondered why the alphabet is always written out as a 2-dimensional grid like this? Compare that with the English alphabet which is pretty much a one-dimensional sequence of alphabets without any organizational structure.

To get a hint, say aloud the letters in any one of the horizontal rows. Notice any similarities?

Here’s the answer:

If you read any row horizontally, you’ll notice that your lip position and tongue position remains the same, and only the method of expelling air from your voice box, nose and mouth changes. The entire क row is pronounced with the back of the tongue touching the back end of the palate (guttural/velar). The entire च row has the middle of the tongue touching the palate (palatals). The ट row has the tongue curled back and the back of the tongue touching the palate (retroflex). The त row has the tongue touching the teeth (dentals). And the प row has the lips together (labials).

Now, read any column, and notice the similarities. The table above gives away the answer of course: in any column, the way air comes out of our mouth/nose/voicebox remains the same and only the tongue/lip position changes.

Pronouncing ङ and ञ

At this point, you should be able to pronounce ङ and ञ the right way. First pronounce न and म a few times and understand where the air is coming from. They’re nasals, so the air is coming out of the nose, not the mouth. Now pronounce क ख ग घ a few times and understand the tongue position. Combine these two and you get the pronunciation of ङ. Phonetically, ङ is a guttural/velar nasal sound. Similarly for ञ which is a palatal nasal.

The Ones We Left Out

You might complain that I’ve ignored a number of other letters. These are the ones usually written after the main grid. They don’t appear to fit in the grid, and they seem like an afterthought. Basically, य र ल व श ष स ह. This is another thing that was never taught. They are really part of the same grid but never written that way.

An approximant is a sound made when the tongue is not exactly touching the palate/teeth/lips but is quite close. Similarly, a fricative is a sound when the tongue gets even closer and actually creates turbulent airflow due to friction (more accurately, frication). Now read any row again and notice how the tongue position remains almost the same. In the last two columns, the tongue has to shift a little bit so that it is not touching, but it stays in same the general area.

Now you can see how य and श are related to च, and so on. You also now know how the pronunciations of श and ष are different. In the former, the tongue is straight like in च while in the latter, the tongue is curled back like in ट.

Notice also how व is related to प and फ and you begin to understand why the Germans pronounce Volkswagen as Folkswagen (the cheap mass-produced people’s car, or car for the common folk).

By the way, ऴ should be next to र. If you don’t know how to pronounce ऴ, try to say it like ल but with the tongue curled back as in ट.

The Vowels

The vowels are feeling left out, so let’s drag them in here, shall we?

A vowel is a sound in which the tongue is not touching anywhere, and the airflow is smooth and continuous. Again, sound out the rows above and convince yourself that अ आ are related to क, इ ई are related to च, and उ ऊ are related to प.

This also allows us to extrapolate and understand ऋ. It is neither री nor रु. You’ll notice that र is a retroflex approximant, and hence the curled tongue is almost touching the palate. But ऋ is a vowel, so the tongue isn’t near the palate. The tongue is still curled up like in र and ट, but it is in the air and allowing free flow of air. This is the original Sanskrit pronunciation of ऋ but it “has been lost in the modern languages, and its pronunciation now ranges from री (Hindi) to रु (Marathi).” Basically, the original sound is no longer used by anybody and Hindi and Marathi have officially replaced it with cheap knockoffs.

Which is why you’ll never get a good answer to whether it is Krishna or Krushna, or as ISKCON prefers, Kṛṣṇa (which is technically closer to the original Sanskrit).

And I didn’t even know of the existance of the long ॠ. And I had forgotten about ऌ completely.

Finally, ए ऐ ओ औ अं and अः are troublemakers who don’t fit in this scheme, so I am going to conveniently ignore them. And best stay away from ॲ and ऑ which are abominations brought in by foreigners.

Foreign sounds: Or why Is it ghazal and not gazal?

Since you’re already here and listening to arcane details of pronunciations, let me clear up a few more things. The ग़ in ग़ज़ल is not the normal ग. By now you know that ग is pronounced with the back of the tongue touching the back of the palate. However, ग़, which was brought in by the Persians, is a different sound. Here, the tongue isn’t touching the palate at all. Instead you try to make the ग sound but by constricting your throat. This is not easy to do without training if you haven’t been doing it since childhood. In any case, try to hear how a trained singer pronounces the word ग़ज़ल and you’ll start noticing the difference.

Similarly, the words qayaamat, qaatil, qaazi are never written with a k, because that is not the क sound but the क़ sound.

Also, I never understood why my American friends would laugh so hard when I mixed up the pronunciations of “wet” (water) and “vet” (doctor). They are both वेट in Devanagari, but in English the ‘v’ and ‘w’ sounds are very different. ‘w’ is said with your lips together, like in उ (that is why it is a double-u, see?). But ‘v’ is said with your lower lip on the upper teeth. Both these sound the same to most Indians, but very different to westerners. I found that hard to believe initially, until I realized that the reverse also works. Arabic speakers can’t hear the difference between the ‘p’ and ‘b’ sounds, so they will routinely say “Bushba” instead of “Pushpa” (“ro mat Bushba…I hate tears.”)

There are many other sounds from all over the world that you can discover if you start poking around Wikipedia’s phoentic pages. Marwadi and Sindhi have some sounds whose description also I don’t understand. African languages have clicks. And much more. And who knows, you might even figure out how to pronounce the l in Tamil is pronounced, and how it is not an ‘l’ at all, which is why they spell it Tamizh. The same ‘zh’ sound is found in Kozhikode, and other Tamizh words.

And if you want to know why “Rama” is not spelt “Ram” check out my thread on Twitter which recently went viral.

Feel free to share this post with friends or on social media.


I regularly write about topics I find interesting and fascinating. It could be a serious topic like Misconceptions About the Income Distribution in India, or life advice like Winning vs Being Nice (Game Theory). Sometimes the comments on my artices are more valuable than the article itself. If you find these topics interesting subscribe to get these updates in your email.