Category Archives: Pronunciation

The Sounds of Sanskrit

Sanskrit Vowels

Knowledge of the structure of vowels is not only essential for pronunciation, it is the basis upon which many other grammar rules, such as sandhi, are built.

Basic Vowels (Svara)

There are four basic vowel sounds:

कि कु कृ
a i u

  [There is a fifth, (ḷ), but it is very rarely used.]

Which are pronounced:

(a) under

(i) if

(u) to

(ṛ) vowel “i” with tongue curved to touch the top of mouth

Elongated Basic Vowels

Each basic vowel has an elongated version, too:

आ ई ऊ ॠ

का की कू कॄ

ā ī ū ṝ

Which are pronounced:


ई (ī) letter “E”

ऊ (ū) stew

ॠ () [rarely used]

The basic vowel and its elongated version are two versions of the same sound. For example “a” and “ā” are not two different sounds, they are two variants of the same sound. So they are called savarna (“same-sound”) or ekātmikā (“of one nature”).

Extensions of the Basic Vowels(Guṇa)

“A” combines with “i”, “u”, and “ṛ” to create a third set of vowels. These are:

ए ओ अर्

के को कर्

e o ar

Which are pronounced:

(e)letter “A”[a+i]

(o) letter “O” [a+u]

अर् (ar) urgent [a+ṛ]

Matured Extensions (Vṛddhi)

Each extended vowel also has an elongated version:

ऐ औ आर्

कै कौ कार्

ai au ār

Which are pronounced:

ऐ (ai) “eye”

औ (au) ouch

आर् (ār) car

Semi Vowels

A “semi-vowel” is a consonant equivalent of a vowel, and often takes the place of a vowel when sounds blend together (by sandhi). Sanskrit recognizes four semivowels, having no consonant equivalent for the ultimate root, (a).

Semivowels are “equivalent” to a vowel because they are produced from the same area of the vocal apparatus. They are the sound that occurs when the area of the vocal apparatus engaged by the vowel becomes more clearly utilized and distinctly articulated.

The vowel “i” becomes the semi-vowel य् “y” (both are “palatal” sounds).

The vowel “u” becomes the semi-vowel व् “v” (both are “labial” sounds).

The vowel “ṛ” becomes the semi-vowel र् “r” (both are “cerebral”).

The vowel “ḷ” becomes the semi-vowel “l” (both are “dental”).


Sanskrit categorizes its phonetic components according to which part of the vocal apparatus we use to create the particualar sound. It recognizes five discrete sections of the vocal aparatus:

  1. The throat
  2. The palate (junction between the top of the throat and top or the mouth, essentially, the back of the mouth)
  3. The roof of the mouth
  4. The teeth
  5. The lips

The basic sounds made at these different points, in combination with the root vowel, (a) are

– ka (throat)

– ca (palate)

– ṭa (roof)

– ta (teeth)

– pa (lips)

If some “voice” (depth, a deeper tone) is added to the sounds we get a second group of consonants:

– ka – ga (throat)

– ca – ja (palate)

– ṭa – ḍa (roof)

– ta – da (teeth)

– pa – ba (lips)

Next, air can be added to either set, to give us two additional sets

– ka – kha – ga – gha (throat)

– ca – cha – ja – jha (palate)

– ṭa – ṭha – ḍa – ḍha (roof)

– ta – tha – da – dha (teeth)

– pa – pha – ba – bha (lips)

An additional set of consonants can be made by producing involving the nasal cavity in the production of the sound.

– ṅa (throat)

– ña (palate)

– ṇa (roof)

– na (teeth)

– ma (lips)

Another set of consonants is made by pushing a lot of air through the vocal apparatus.

– ha (throat)

– śa (palate)

– ṣa (roof)

– sa (teeth)

? – (f)?(lips)[not used]


Pronouncing / Spelling “n” in the Middle of Sanskrit Words

N will frequently show up as “Ṇ” when you don’t expect it, and it’ll seem like a misspelling.

Sanskrit sounds (probably all languages too, with a few exceptions) are made in five places in the vocal structure of your mouth and throat: (1) in the throat, (2) back of the mouth, (3) top of the mouth, (4) on the teeth, (5) on the lips.

“k” is the sound made with the tongue muscles that are in the throat. “c” is the sound made with the tongue muscles at the back of the mouth. “ṭ” is made with the tongue on the roof/top of the mouth. “t” is made with the tongue on the teeth, and “p” is made with the lips.

Those are the sounds without voice or air. If you add voice and air you get a whole bunch of other letters, as described in Pronunciation of Consonants, and Sanskrit Letters.

The sounds made with the tongue on the roof of the mouth are the characteristic sounds people make to imitate the Indian accent. The technical term for it is “retroflex” (which just means “curling back”, referring to the tongue). It takes a bit more effort to curl the tongue back and stick it on the top of the mouth, compared to the effort it takes to get the tongue into other positions for pronouncing the other types of letters, sooooo there’s a special rule about these letters:

Once you get your tongue into a retroflex position, it can stay there until some other consonant forces it into a different position. This means that once you have a retroflex letter, the rest of the letters in the word will become retroflex too, unless a non-retroflex consonant comes along before the word ends.

Vowels are produced by the shape of the mouth, not the tongue, so they don’t interfere with this. So, once your tongue goes into retroflex, it can stay there through the rest of the vowels in the word, and any other consonnant that’s not in some other position.

Well, just about every other consonnant is in some other position, so basically we’re just talking about the nasal letters (produced mainly in the nose). In other words, we are talking about “n”.

Let’s look at some examples:

Aruna अरुन = Aruṇa अरुण

The word starts with a vowel, a, followed by the semivowel, “r.” Say the “r” sound out loud. Notice that your tongue gets retroflex? R is a retroflex sound. So now, at the second letter of the word, our tongue got into a retroflex position.

The third letter is a vowel, “u” so our tongue doesn’t have to change position. The fourth letter is “n”. N is a nasal, and nasal sounds can be made with the tongue in any of the five positions. Since our tongue is in retroflex, we keep it there, so the “n” becomes “ṇ” (big hint: just about any letter with a dot under it is retroflex). The word finishes with a vowel, “a” which is important. To complicated to explain why.

Arjuna अर्जुन (no change)

The r is retroflex, but it’s followed by a “j”, which is a consonant that forces the tongue to the back of the mouth, out of the retroflex position – so  by the time we get to the “n” in “Arjuna” our tongue is free to pronounce it as its originally supposed to be “n” (tongue on teeth).

Rāmāyana रामायन = Rāmāyaṇa रामायण

The first letter, “r” puts the tongue in retroflex. Then there’s a vowel – so no change to the tongue position. Then a nasal, “m” – again no change to the tongue. Then there’s another vowel. Then there’s “y” which is a semi-vowel and doesn’t change the tongue position. Then another vowel, again no change. So by the time we get to the “n” our tongue is still in retroflex, so the “n” becomes a retroflex n, “ṇ” to make pronunciation easier and more natural.

Kurvanti कुर्वन्ति (no change)

The third letter, “r” puts us in retroflex. “v” is a semi-vowel, followed by a vowel, so when we get to “n” we think we should make it “ṇ” but it’s not a real, full “n” because it has no vowel. It’s joined to “ti” to make a single letter “nti” न्ति. So it’s really a “t” which changes the tongue position to the teeth, therefore the “n” stays a dental nasal, and doesn’t become reflexive.

Brahman ब्रह्मन् (no change)

Again, the “r” sets up the retroflex, and “ahma” don’t change it (h is just air, and m is all lips), so again we expect to change the “n” to retroflex – but again, like in kurvanti, it’s not a full/real “n” because it has no vowel, it’s just the end of the previous syllable, not a syllable on its own. So it doesn’t get changed.

Brahmana ब्रह्मन = Brahmaṇa ब्रह्मण

This is the same as above, but in this word, the n is a real n, a syllable unto itself, because it has a vowel following it. So in this case the n does retroflex to ṇ.

When to worry about it

Usually word stems are already defined with this rule in mind, so generally you only have to worry about it when you’re adding the inflection (ending) to the word.

For example, take the word nara (man). If you want to use this word in Case 6, plural (“men’s”) you are supposed to add the inflection –anām. So go ahead, and you get narānām. But now notice that the “r” gets you into retroflex and there’s no vowel to get you out of retroflex before the “n” comes, so the real spelling/pronunciation is narāṇām.

Similarly if you want to use a neuter noun with a plural inflection, the ending should be –ani. For example (“leaves” – pattra), pattrāni. But the “r” in the root word retroflexes the “n” in the inflection, so the correct version is pattrāṇi.

So, really, you only have to worry about it in the inflections. And even then only in the inflections that have “n” in them, followed immediately by a vowel. So that means only the two cases shown above (as far as the cases we’ve already learned)

Pronouns as Objects (And Compound-Vowel Sandhi)

Yesterday we got our first look at Nouns as Objects.  Now lets look at how Sanskrit uses Pronouns as Objects. Here’s a table, along with a reminder about the same Pronouns as Subjects:

First Person (“I”)
मद् Singular Dual Plural
Case 1 (subject) अहम्



Case 2 (object) माम्



Second Person (“You”)
त्वद् Singular Dual Plural
Case 1 (subject) त्वम्



Case 2 (object) त्वाम्



Third Person masculine (“he”)
तद् Singular Dual Plural
Case 1 (subject) सः



Case 2 (object) तम्




So, what do you notice about all of it? Well, one thing is:

  • Just like with nouns, there’s no difference between Subject and Object inflection when the case is dual.
  • One more thing: except for third person singular and dual, all the Object pronouns have the final vowel as “ā”

Another little observation is that third person masculine object pronouns have the same exact endings as masculine object nouns.

  • gajam (“to the elephant”) | tam (“to him”)
  • gajau (“to two elephants”) | tau (“to those two guys”)
  • gajān (“to the elephants”) | tān (“to those guys”)

Compound-Vowel Sandhi

“Compound vowels” are “ai” and “au.” The rule is:

The second vowel becomes a semi-vowel (“v”, “y”, or “r”)

For example:

अश्वौ इच्छति → अश्चविच्छति

aśvau icchati → aśvav icchati {“He wants the two horses”}

That’s simple, but “e” and “o” are also compound vowels (this comes from an older form of Sanskrit, pre-classical). So let’s first get straight what vowels are what.

Vowel Treated as Example
e ai -e + i- → -ay i-
ai āi -ai + i- → -āy i-
o au -o + i- → -av i-
au āu -au + i- → -āv i-

Remember that “ai” is a stronger form of “e”, and “au” is a stronger form of “o”, and this wont seem arbitrary.

The case of “e” above has some special rule to it (because e is phonetically a very strong sound):

Like all compound vowels “e” changes to a stronger form (“ai”) before applying sandhi. But unlike other compound vowels, the second vowel just disappears rather than change into a semi-vowel.

Here is an example:

ते इच्छन्ति → त इच्छन्ति

te icchanti → ta icchanti


  1. The “e” in “te” becomes “ai” – so the sound is now “tai”
  2. The e is stronger than most vowels so it obliterates the second vowel rather than merely change it. So the “i” disappears from “tai” and you’re left with “ta”
  3. No further sandhi applies.

One more special note about this “e” business:

If the next words starts with a, that a will also disappear.

A is the most basic, subliminal sound, that’s why it’s so easy to make it disappear when next to the strongest sound, “e.”  So, for example:

ते अश्वाः = ते ऽश्वाः

te aṣvāh = te ‘śvāḥ


Pronunciation of Consonants

Consonants are the shape around the edges of a vowel. We create different consonant shapes by putting stress on different parts of our throat and mount at the beginning and end of saying the vowel. There are five places you can put stress:

Throat You put tension in the thoat when you make the sound
Back-Mouth The back of your tongue moves towards the back of your mouth
Roof of Mouth The tip of your tongue curls to lightly tough the roof of your mouth
Teeth The tip of your tongue is right on the bottom tip of your teeth
Lips Your lips touch and make sound without the tongue

Clicky / Poppy Sounds

If you make a sharp, quick sound in any of those places you get the following consonants:

Throat k
Back-Mouth c
Roof of Mouth
Teeth t
Lips p

The roof-of-mouth sound is the most unfamiliar to us English-ers. To make it you have to curl your tongue so its tip lightly touches the top of your mouth. Do that while making a “t” sound (quickly, without lots of sound) and you have pronounced “ṭ” (and also probably suddenly feel very Indian!).

The teeth sound is a lot like a normal English “t” but try to force your tongue lower than normal, right onto your teeth, and don’t put much stress on the sound at all, say it lightly.

Those are the first five consonants. The next five are the strong versions of the same sounds. The pronunciation is the same, they just have more air behind them. They are a lot of fun to pronounce correctly.

Throat k kh
Back-Mouth c ch
Roof of Mouth ṭh
Teeth t th
Lips p ph

Some pointers:

  • c is different from ch. It’s pretty easy to do the “ch” sound. Concentrate on making the “c” different from it, by making it very light and gentle.
  • Don’t ever let yourself say “th” like windy sound in “the”.
  • Don’t ever pronounce “ph” like the weird English variant where it sounds like “f”.

Sounds with “Voice”

So now you have the first ten Sanskrit consonants! Congrats! The next 10 are exactly the same as these, they just have more “voice.” That means you use a little tone to sound them out, not just the click and pop of the first group of ten.

Throat k kh g gh
Back-Mouth c ch j jh
Roof of Mouth ṭh ḍh
Teeth t th d dh
Lips p ph b bh

The difference between the “ḍ” in row three and the “d” in row four is the same as the difference between the “ṭ” and “t”: like all letters with dots under them, you need to curl your tongue and touch the roof of your mouth to pronounce them. All teeth letters need your tongue-tip touching your teeth (and pick a pack of pickled peppers, while you are at it!).


“Nasals” are sounds you make by forcing the air in your nasal passages to vibrate. These are all the “n” type sounds, basically. Sanskrit recognizes five:

Throat k kh g gh
Back-Mouth c ch j jh ñ
Roof of Mouth ṭh ḍh
Teeth t th d dh n
Lips p ph b bh m

The “ṅ” is a very difficult sound, at least for me, to pronounce. It’s like saying a “g” while holding your nose – that’s the best I can do to describe it. Remember the tension is in the throat, yet its a nasal sound. Think of it as an “n” with some “g-ish-ness.”

The “ñ” is a more familiar sound, because its common in Languages like Spanish. The tongue is pushing backwards, towards the back of the mouth – and the result is that there’s a slight “y-ish-ness” to the “n.”

The “ṇ” is fun and not too hard to say. Just curl your tongue and touch the roof of your mouth while making the “n” sound.

The naked “n” is close enough to the normal English “n.” And the nasal sound that you make on your lips is what we write in Enlgish as an “m.”


It’s not so important, but the consonants called semi-vowels get that name because you can sort of sound out all by themselves (vowels are “vowels” because they can sound out all by themselves). Sanskrit, at least as we currently have record of it, does’t recognize a semi-vowel sound from the throat, only in the four other positions: back of the mouth, roof of the mouth, on the teeth and on the lips.

Throat k kh g gh
Back-Mouth c ch j jh ñ y
Roof of Mouth ṭh ḍh r
Teeth t th d dh n l
Lips p ph b bh m v

These are all straightforward, familiar sounds to English-speakers. A technical note is that “r” and “l” with dots under them are slightly different and are thought of as full vowels. But this is purely scholastic, if you ask me.

How Many Ways Can You Pronounce “s”?

Sanskrit has three kinds of “s” sounds. This is really very important to get right, early on, or your pronunciation will be way off. Pronounce “ś” with the body of your tongue reaching towards the back of your mouth. Try to make it a relatively high-pitched “shh” sound.

Pronounce “ṣ” with a curled tongue, the tip touching the roof of your mouth. It should make a rich, deep, lower-pitched “shh” sound.

Every English speaker knows how to pronounce the unadorned “s.” It is the familiar, whistle-loving s with the tongue on the teeth.

Here are the three S’ in their correct positions on the table:

Throat k kh g gh
Back-Mouth c ch j jh ñ y  ś
Roof of Mouth ṭh ḍh r  
Teeth t th d dh n l  s
Lips p ph b bh m v

Just Three More to Go!

OK! You got through learning 32 Sanskrit consonants, awesome! There are just three more left.

The simplest is “h” – although ancient Sanskrit probably pronounced it differently (with actual voice, not just air), from the classics onward it’s close enough to the same “h” we have in English. It’s never “silent” – there are no “silent letters” in Sanskrit spelling.

Then there is a dotted h: “ḥ” – it always follows a vowel (almost always at the end of a word, too). Its just supposed to be air that comes after a vowel. But most people enunciate this air by faintly echoing the preceding vowel.

Finally there is a dotted m: “ṁ” (sometimes written as “ṃ” – I don’t like it that way, though). It also always follows a vowel and is almost always at the end of a word. Pronounce it by nasalizing the conclusion of whatever vowel came before it.

How to Practice

Just wait a bit, as we start to look at Sanskrit words, in the next posts, you’ll get more than enough chance to practice all this stuff.

Pronunciation of Vowels

Sanskrit pronunciation is important not just for the aesthetic effect, but also because it will help you spell correctly, which is really import when trying to find words in dictionaries.

First of all, like just about every language, Sanskrit has the basic five vowels, “a, e, i, o, u.”

vowel pronunciation
a like the vowel in car, far, star, jar.
e like the vowel in say, pay, day, may.
i like pick, quick, thick, trick.
o like go, no, oh, thow
u like book, took, look, hook

Then Sanskrit has five more vowels, that are the strengthened and elongated forms of the original five.

a ā like the vowel in cat, fat, sat.
e ai like high, bye, why, try.
i ī like we, me, three, tree.
o au like ow!, how, now, cow
u ū like “ooh” in “oohs and aahs”, knew, threw, true

Sanskrit Rap – Meter

Sanksrit is usually spoken with meter a.k.a cadance a.k.a. rhythm. Syllables are either “1-beat” or “2-beats” long. Keep this in mind and you will really enjoy reciting Sanskrit, because it will become song-like. And you will get into the right accent.

All the vowels in the second group (“strengthened”) are long. And from the first group, e and o are also long (aka, they last for 2-beats, twice as long as short vowels). That means that only a, i, and u are short. But you get “a” a lot, so it balances out.

Sometimes you stretch the duration of a, i, and u, too; if they are followed by more than one consonant.

Sanskrit is full of beautiful meter-templates, maybe in a future post I will get into showing you some of them!

The Elusive ṛ

Sanskrit has a strange vowel that we don’t have in English: “ṛ”. Being absent from most modern languages, this sound has become confusing even to Indians, and there are now many different interpretations of how to pronounce it. The purely Sanskrit way is:

  1. Curl up your tongue so that the tip touches the middle of the roof of your mouth.
  2. With your tongue in that position, try to say the vowel “i” (the vowel in pick, thick, etc).

That’s the sound.

Technically speaking, there is also a strengthened form of this vowel, “ṝ”.  But you’ll almost never see it used if you read classical or post-classical Sanskrit. In the same boat is another vowel, “ḷ” and its strengthened form, “ḹ”. You’ll almost never encounter these in normal Sanskrit, so concentrate on the 10 main vowels and the one strange one that does come up often: ṛ.

What Are Vowels?

Vowels are the basis and foundation of sound. Every syllable is a vowel. Vowels are the way your mouth and throat make sound when you push air through them while holding them in different shapes.

Consonants are sounds (“sonance”) that work with (“con-“) vowels. They shape the beginning and end of the vowel sound. Next post we’ll talk about the beautifully rational structure of pronouncing Sanskrit consonants.

Thank you,

Vic DiCara