Chapter 2 – The Consonants, Overview

1 The Consonants, Vyañjana
2 Grouped Consonants
3 Ungrouped Consonants
4 Aspiration, Prāṇa
5 Voicing, Ghoṣa
6 Subtleties of Voicing & Aspiration
7 Logical Structure of the Vargas

Lesson 1
The Consonants, Vyañjana 

Having become familiar with the thirteen vowels, let’s now delve into the consonants. The table below shows the whole collection. 

These akṣaras, numbering 33, are of two types:

1) grouped consonants (tan columns above )

2) ungrouped consonants (gray columns)

Exercise 2. 1.1 (no audio) – (Audio Coming Soon!)

Copy the above chart by hand several times on paper with all the categories and headings. (Write the akśaras in Roman script only if you do not know Devanāgari.) As a next phase, practice recreating the chart from memory. 

Lesson 2
Grouped Consonants

Grouped Consonants, Vargīya Vyañjana (25)

The grouped consonants, 25 in number, are classed as sparśas, meaning “touch” sounds. They are of two types: stop letters, and nasals.   

Stop Letters (20)

These twenty akṣaras use a full “touch” of the tongue (or lips) to stop the flow of air in the mouth as part of their pronunciation. These are the stop letters (or plosives):

ka  kha    ga   gha 1. guttural

ca  cha    ja  jha 2. palatal 

ṭa  ṭha  ḍa  ḍha 3. retroflex  

ta  tha  da  dha 4. dental  

pa  pha   ba bha 5. labial

Nasals (5)

The five nasal letters use a full “touch” of the tongue or lips to stop the flow of air in the mouth, while also redirecting the air through the nose. With the nasal letters, the closure at the place of pronunciation is not released, and hence most of the air passes through the nose. The nasal sound is formed by narrowing the air passage at the guttural area.  These are the nasal letters:

ṅa 1. guttural

ña 2. palatal  

ṇa 3. retroflex 

na 4. dental 

ma 5. labial

InDepth: How Many Akṣaras Are in the Varṇamālā?While in this course we list 49 Sanskrit basic akṣaras, you will see in various texts a count of 50 or 51— reached by the inclusion of the common compound kṣa क्ष and the obscure letter ळ, which rarely appears except in Vedika texts as a specific sandhi for ḍa ड.
Exercise 2. 2.1 (no audio) – (Audio Coming Soon!)

Copy the Grouped Consonant chart by hand several times on paper with all the categories and headings. (Write the akśaras in Roman script only if you do not know Devanāgari.) Train yourself to create the chart from memory. 

Lesson 3
Ungrouped Consonants

The ungrouped consonants (gray columns), like the nasals, are also non-touch consonants. They consist of four semivowels, three sibilants and the aspirate (ha ह). 

With these three sets of letters, the flow of air is restricted but not stopped in the course of pronunciation. In other words, the tongue does not momentarily stop the flow of air as it does with the grouped consonants, or sparśa letters. 

To experience the difference between a non-touch akṣara and a stop akṣara, compare the pronunciation of ta, which is a stop letter (the tongue touches the back of the front upper teeth), to sa, which is not a stop letter (the sound is restricted by the tongue near the teeth, but not stopped). 

The Four Semivowels 

Sanskrit has four semivowels (or approximates): ya, ra, la and va. Like the vowels, the semivowels require the use of the vocal cords. For each semivowel, the tongue (or lips) come near one of the points of pronunciation in the vocal system, restricting the flow of air, but not fully blocking it.

 य ya 2. palatal
sounded at the roof of mouth, with the tongue held down

 र  ra   3. retroflex
sounded at the roof of mouth above and behind the teeth

 ल  la   4. dental
sounded with the tip of tongue at the back of front teeth

 व  va   4/5. dental/labial
sounded at the teeth and lips

The Three Sibilants

There are three sibiliants in Sanskrit: śa, ṣa and sa. Each of these letters, like the semivowels, is pronounced by bringing the tongue near its point of pronunciation in the mouth, restricting the flow of air, but not fully blocking it. These letters are characterized by their hissing sound. The vocal cords are not used in their pronunciation. When properly pronounced, each is clearly distinguishable. 

śa   श 2. Palatal
sounded at the roof of mouth, with the tongue held down

ṣa  ष 3. Retroflex
sounded at the roof of mouth just behind the teeth

sa  स 4. Dental
sounded with the tip of tongue behind the upper front teeth

Supplementary Video

Enjoy this video from the Sanskrit Channel introducing the consonants of Sanskrit.  

Exercise 2.3.1 (no audio) – (Audio Coming Soon!)

Copy the Ungrouped Consonant chart by hand several times on paper with all the categories and headings. (Write the akśaras in Roman script only if you do not know Devanāgari.) Train yourself to create the chart from memory. 

Audio Exercise 2.3.2, with audio – (Audio Coming Soon!)

Following along with the audio, repeat the following sequence of consonants slowly, being mindful of the difference in feel of the articulation points and the way the different parts of the tongue touch the spaces inside the mouth. Notice how the ungrouped sounds feel softer in the mouth because they are non-touch consonants. Repeat these sequences over and over. Work on them again and again to learn their correct pronunciation.

1) Semivowels: ya,  ra,  la,  va 

First, With the audio, listen and practice the correct pronunciation of each of these four letters. 

Second, compare how ya and ai, ra and ṛ, la and ḷ, va and au feel in the mouth (each semivowel paired with a vowel of the same position). 

2) Sibilants: śa,  ṣa,  sa 

Listen to the audio closely to discern each sibilant’s point of articulation: śa (palatal), ṣa (retroflex), and sa (dental).

3) Aspirant: ha ह

Note that this akṣara is sounded from the guttural position, no. 5. 

Lesson 4
Aspiration, Prāṇa

In Sanskrit, ten akṣaras are sounded with an extra force of air; for example kha. This force of air is aspiration. Thus, these letters are known as aspirates or mahāprāṇa. 

The ten consonants with a small force of air behind them (like ka) are classed as alpaprāṇa (unaspirated). These ten akṣaras are the basis for the aspirated akṣaras, as shown below.

The ten fundamental stop consonants are unaspirated (columns 2 and 4). Each of these has an aspirated version (columns 3 and 5). 

In Roman transliteration, aspiration is indicated by the letter h after the base consonant, as shown in the two tan columns above. For example, kha. In Devanāgari, as you can see in the table, the aspirated forms are entirely different glyphs.

Aspirates are pronounced with full breath.
Non-aspirates are pronounced with minimal breath. 

Exercise 4.1, with audio – (Audio Coming Soon!)

Listen closely as the teacher enunciates the words in the following list. Repeat each word after it is spoken, mimicking his pronunciation closely.

Notice where your tongue goes as you say the words. As you listen closely and practice the words, pay close attention to how the sounds, the force of air (aspiration) and positions of the tongue differ.


Lesson 5
Voicing, Ghoṣa

All of the consonants can be distinguished as either voiced or unvoiced. (Note that all vowels are voiced.) Other terms used for voicing are soft (for voiced) and hard (for unvoiced).

Unvoiced (aghoṣa) letters do not use the vocal cords; these sounds are made primarily by air simply passing through the mouth. The unvoiced letters occupy the yellow columns in the table below. The prime example of an unvoiced consonant is ka.

A letter is voiced (saghoṣa) if it requires use of the vocal cords. The voiced letters occupy the yellow columns in the table.

You can feel the difference between a voiced letter and an unvoiced letter if you hold your fingers to your Adam’s apple during pronunciation. Compare ka with ga. You can feel the vocal cords vibrate when you pronounce ga. With the sound ka there is no vocal cord resonance. To feel the difference more easily, compare ak to ag.

This distinction between voiced and unvoiced letters is an essential part of understanding the structure of the Sanskrit alphabet. Voicing plays an important role in the principle of sandhi, the way certain letters coming together in speech change one another. Though these changes occur in all languages, Sanskrit, with its fully phonetic writing system, is unique for formalizing the changes in the spelling and writing of words.

To get an idea of how the voicing of one letter can influence another, try comparing the final s sounds of the plural english words cats and dogs. Notice that the s in cats sounds like a normal s, whereas the s in dogs sounds and vibrates like a z sound. This change occurs because the s in dogs follows the  voiced/soft letter g, which activates the vocal cords. This activation of the vocal cords for the letter g affects the letter s, making it voiced as well, and changes its sound to that of z. If you try, you may be able to forcibly remove the voicing from the final s in dogs. But if you listen to yourself closely, you will find that either you have added an extra pause between the letters or you have ended up actually saying doks. In other words, to keep from adding voicing to the letter s, you have removed the voicing from the proceeding g, thereby producing its unvoiced counterpart, which is k.

To tell if any letter is voiced or unvoiced, try prolonging the initial sound of the letter (like the gggga or kkkka) before releasing the vowel after it. In the case of a voiced letter, like g, you will feel a vibrating sensation similar to what you would feel while humming. This humming sensation is the activation of the vocal cords. If the letter is unvoiced, like km you will either find it hard to prolong the initial sound of the letter or you will experience a flow of air making a hissing sound.

Exercise 2.5.1 (no audio) – (Audio Coming Soon!)

Go through the consonants below and determine whether each is voiced or unvoiced. Afterwards, check your answers by consulting the table at the beginning of this lesson.

sa   __________jha    __________ ka __________ṭa  __________
ga   __________ra    ___________kha _________ma _________ 
ca   __________ na   ___________da __________pa __________
ja    __________ śa   ___________ḍa __________ya __________

Lesson 6
Subtleties of Voicing & Aspiration

Pronouncing the unvoiced,
unaspirated consonants: ka, ca, ta, ṭa, pa

A caution to avoid aspirating these letters.

In English, if a word begins with an unvoiced consonant, meaning it does not use the vocal cords—such as k, c, t, ṭ, p—that sound naturally tends to be aspirated. Hence, to correctly pronounce the unaspirated, unvoiced Sanskrit letters, English speakers must restrain the force of air, so that ka, for example, doesn’t sound like kha

Learning the difference between
unvoiced and voiced akṣaras

In the beginning, ka correctly pronounced may almost sound like ga. Likewise, the other unaspirated unvoiced letters (ca, ta, ṭa and pa) may sound somewhat like their voiced counterparts (ja, da, ḍa and ba). It will take time to hear and understand this subtlety. With practice, the distinctions become clear.

Studying the similarity in sound between the unvoiced unaspirated letters, like ka, and the voiced unaspirated letters, like ga, can also be helpful in learning the correct pronunciation. This can be done by saying ga several times and then attempting to make ka sound as close as possible to that sound. 

The reason this works is that the sound ga engages the vocal cords throughout, while ka begins with a sound made only by air (the k part) and then uses the vocal cords for the vowel. If the vocal cords activate for the vowel more quickly, it will suppress the breathy sound of the letter, giving us the unaspirated ka. So, another way to think of the unaspirated ka sound is as a very slight or short k with the vowel sound following quickly after it. This is true for all of the unvoiced unaspirated letters (ka, ca, ṭa, ta, pa).

For the examples in the following lessons on the vargas, words have, when possible, been chosen in which the unaspirated letters are preceded by another letter (for instance, skate rather than Kate). With these added letters, it is more natural to say the unvoiced unaspirated consonants in an unaspirated way.

Voiced Aspirated Consonants: gha, jha, ḍha, dha, bha

A caution to not drop the aspiration from these letters.

For the five voiced aspirated consonants (gha, jha, ḍha, dha, bha), the issue is reversed. In English, these are naturally spoken without aspiration (e.g., ga instead of gha). English speakers will have to use the combined words given (e.g., dog-house, in the lesson on Ka varga) to help them realize how to make these sounds properly. 

Exercise 2.6.1, with audio – (Audio Coming Soon!)

Listening to the recording, study the difference in the sound of each pair of unaspirated and aspirated letters below. Then listen a second time, as you make the sounds along with the teacher.

Unvoiced Voiced

ka kha ga gha

ca cha ja jha

ṭa ṭha ḍa ḍha

ta tha da dha

pa pha ba bha

Exercise 2.6.2, with audio – (Audio Coming Soon!)

In the recording for this exercise, the unaspirated and aspirated pairs above have been mixed up. Try to see if you can distinguish which one in each pair is the unaspirated letter and which one is its aspirated counterpart. Write down your findings. Afterwards check the answer key on the following page.

Exercise 2.6.3, with audio – (Audio Coming Soon!)

As in the last exercise the pairs of unaspirated and aspirated letters have been mixed, but this time each letter has been coupled with a different vowel. The goal of this exercise is to help you make sure you can clearly hear the difference in aspiration/unaspiration even when the vowels are different.

Exercise 2 Answer Key

pha pa – ka kha – ga gha – bha ba – ṭa ṭha – tha ta – ḍha ḍa – dha da – cha ca – ja jha

Exercise 3 Answer Key

kī khu – ghai gau – chū cu – ju jha – ṭā ṭha – ḍau ḍho – te thai – dhe do – bū bhī – phi pa  

Lesson 7
Logical Structure of the Vargas

You can accelerate your learning of the sounds of Sanskrit by understanding the structural uniformity built into the five vargas. 

Here we show the nature of the letters of Ka Varga. These same principles apply to all five families.  

The first letter in the Ka Varga is hard/unaspirated. The vocal cords do not vibrate when we pronounce ka

By simply adding air, we derive kha, the second letter in the family. 

If we soften ka and let the vocal cords vibrate, we derive ga, the third letter of the family. The softening of the sound through vocal resonance is known as voicing, which we will soon learn more about.

Adding air to ga, we derive gha, the fourth letter of the family. This addition of air is called aspiration. Gha is the aspirated form of ga.

By nazalizing ga, we derive ṅ, the fifth letter of the family, the nasal. We can derive the proper nasal sound for ourself, by picking a word we know that has an n before the first letter in the varga (in this case k), such as link. Practice pronouncing the ṅk again and again and then drop off the k sound to isolate to the nasal sound. That is the correct nasal sound for the Ka Varga. 

Exception: For the Pa Varga, the nasal is m, not n. 

Exercise 2.7.1, with audio – (Audio Coming Soon!)

Perform the progression described above with each varga again and again. As you do so, observe the subtleties of aspiration and voicing. Impress in your mind the pattern that persists in the relationship between each type of letter. 

ka unaspirated, hard (unvoiced)

kha aspirated, hard (unvoiced)

ga unaspirated, soft (voiced)

gha aspirated, soft (voiced)

ñ aspirated, nasalized

Exercise 2.7.2, with audio – (Audio Coming Soon!)

To gain a deeper understanding of the vargas, follow along with the audio as it goes across the five families of letters (horizontally) in the table. Study the above table as you follow along, becoming more aware of how the various features of sounds (voicing, aspiration, etc.) modify the initial sound of the varga as we go across the table. Notice also how the position of the mouth moves one step forward, from the guttural all the way to the labial, each time we recite one set of five.

Play the next audio and follow along. This audio goes down the columns of the above table instead of across the rows. The result is five sounds of separate vargas which have the same qualities (voicing, aspiration, etc.), only differing in the position of tongue and lips.