The Consonant


Consonants are any sounds that involve blocking of air before it leaves the mouth. The definition of a consonant is not very important to us however, but rather how we can represent consonants and use them to form clusters and eventually syllables in Nepali. Syllables are made by combining consonants with vowels. In Devanagari, vowels are not written separately but indicated by diacritics when they are part of a syllable.

It’s normally difficult to pronounce a consonant by itself, so we usually attach it with a vowel to make it easier to pronounce. This vowel in Nepali is ‘a‘ (which functions as a schwa). This is the form you have seen in the lesson The Devanagari Script. We will now look at consonants without this inherent vowel sound here.

Vowels also have their own form when they are their own syllables, which you have seen in The Devanagari Script. We will look at their diacritical form in the lesson Diacritics. Diacritics modify the consonant(s) to indicate a vowel sound.


In Devanagari, when we want to indicate that a consonant or a cluster has no vowel sound, we use a special diacritic called halanta. The हलन्त (halanta) looks like this and goes under the base of the character: ् . A character with a halanta carries no vowel sound.

क (ka) is the first letter of the Devanagari script. When we learn the script, we learn it with the inherent vowel sound “a“. The vowel अ (a) is special because it has no associated diacritic, and combining it with a consonant simply removes the halanta. In order to remove this inherent vowel from क (ka), we add the halanta:

क (ka) + ् = क् (k)

The logic is a bit backwards above, as the reverse process is happening; the consonant क् (k) is actually being combined with अ (a) to get क (ka). The above only serves to show how the addition of a halanta removes the inherent schwa.

क् (k) + अ (a) = क (ka)

Note | Adding the halanta removes all other vowel diacritics.


Now, it should be easy to visualize how a consonant without a vowel looks like. Basically, add the halanta under the consonant you have learnt, and now you have your ‘real’ consonant (without any vowel sounds):

न (na) + ् = न् (n)

However, things are not so easy.

This type of consonant representation is only one half of the story. Specifically, this is what I call the primary form of the consonant, and it is used when the consonant is stand-alone, or at the end of a word, but never in between. This is very important to understand.

When a consonant appears in the beginning or the middle of a word, or when combined with other consonants to form consonant clusters, they take on the secondary form. This secondary form looks like the consonant without its tail or its hook, or without whatever part the person who invented Devanagari fancied. For example, the following is an example of the secondary form of न (na):

न (na) – T (stem of the character) = न्‍ (n)

Sometimes, the primary and secondary forms look alike. In that case, we just got lucky.


क् (k)ख् (kh)ग् (g)घ् (gh)ङ् ()
च् (c)छ् (ch)ज् (j)झ् (jh)ञ् (ñ)
ट् ()ठ् (ṭh)ड् ()ढ् (ḍh)ण् ()
त् (t)थ् (th)द् (d)ध् (dh)न् (n)
प् (p)फ् (ph)ब् (b)भ् (bh)म् (m)
य् (y)र् (r)ल् (l)व् (v/w)
श् (ś)ष् ()स् (s)ह् (h)


क्‍ (k)ख्‍ (kh)ग्‍ (g)घ्‍ (gh)ङ्‍ ()
च्‍ (c)छ्‍ (ch)ज्‍ (j)झ्‍ (jh)ञ्‍ (ñ)
ट्‍ ()ठ्‍ (ṭh)ड्‍ ()ढ्‍ (ḍh)ण्‍ ()
त्‍ (t)थ्‍ (th)द्‍ (d)ध्‍ (dh)न्‍ (n)
प्‍ (p)फ्‍ (ph)ब्‍ (b)भ्‍ (bh)म्‍ (m)
य्‍ (y)र्‍ (r)ल्‍ (l)व्‍ (v/w)
श्‍ (ś)ष्‍ ()स्‍ (s)ह्‍ (h)


Now that we have seen both forms of the consonants, let’s learn how they are used. We shall not explore how they take up vowels (which we will see in Diacritics) but rather how they can combine with each other to form a more complex sound. 

Take the word ‘ask’. Notice how the sound ‘sk’ is a combination of two consonants, ‘s’ and ‘k’. The combination of two or more consonants together is known as a consonant cluster, since there is a ‘cluster’ of two or more consonants without any vowel in between. In Devanagari, such a sound is represented by concatenating two consonants to form what is known as a ligature or a conjunct consonant. A ligature is a single unit that represents two or more sounds together. An example of this is the letter ‘æ’ or the German ‘ß’. Not all clusters can be represented by a ligature, but all ligatures represent a consonant cluster.

Here is the recap for the two forms and how they are used:

  • The primary form is used when the consonant is stand-alone, or when it appears at the end of a word.
  • The secondary form is used when the consonant appears in the beginning or in the middle of a word, and never when the primary form is preferred.

With the two rules above, we can finally start building up clusters.

Example | Take the cluster ‘sk’. Since ‘s’ precedes ‘k’, we apply the secondary form to ‘s’. Since ‘k’ succeeds ‘s’, the primary form is used. This results in the cluster स्क् (sk).

We can represent even more complex consonants this way by simply applying these two rules.

Example | Take the consonant cluster ‘nkys’. The consonant ‘s’ is at the end of the cluster, so it appears in its primary form. This also means that ‘n’, ‘k’ and ‘y’ all take the secondary form. This results in the ligature न्क्य्स् (nkys).

For a stand-alone consonant, we simply use the primary form.

Example | प् (p) uses the primary form, since it is stand-alone.


Some clusters appear with such frequency that they are represented by their own custom ligature. Examples include tt and kt. We explore these clusters in Special Characters And Variations. However, they can also be expressed without using special characters. Just like how ‘æ’ can be represented by ‘ae’, we can also simply use the consonant combination for that cluster. Some devices display special ligatures while others do not. For example, my PC does not display the kt ligature but my mobile phone does. This is simply a matter of software.

However, there are some notable exceptions. When we form clusters with the consonant ‘r’, a special form of diacritic is used, which we will call the r-diacritic. This diacritic has two names depending on the position of ‘r‘: ‘reph’ and ‘rakār’. When the consonant cluster contains an ‘r‘ in the beginning, we call it a reph. Reph looks like a right-facing sickle and appears above the letter. When the consonant cluster contains an ‘r‘ in the middle or at the end of a cluster, we call it a rakār. Rakār looks like a left-pointing diagonal dash (two dashes for retroflex consonants) and is marked at the bottom stem of the consonant. They are highlighted in red below:

The 'Reph' and both forms of the 'Rakār' (r-diacritic)

reph goes above the letter while a rakār goes beneath. In the above picture, an example of rakār is given in the first character, while the other two characters are examples of reph.

Example | Take the cluster ‘rj’. Since ‘r’ precedes ‘j’, the resulting ligature will be र्ज्.

Example | Take the cluster ‘jr’. Since ‘r’ succeeds ‘j’, the resulting form will be ज्र्.

When the rakār is used for the retroflex consonants ट (ṭa), ठ (ṭha), ड (ḍa) and ढ (ḍha), instead of one dash, two dashes are used. This excludes the nasal retroflex ण (ṇa), which only takes one stroke.

Example | Take the cluster ‘ṭhr’. The resulting form will have two dashes instead of one, ie. ठ्र्.

The following is a table for the retroflex consonants with the r-diacritic. Note that the schwa has been retained for clarity.

ट्र (ṭra)ठ्र (ṭhra)ड्र (ḍra)ढ्र (ḍhra)

Of course, you can just ignore the r-diacritic and just use the half-form instead. The two forms are equivalent:

र्‍प् (rp) ↔ र्प् (rp)

The position of reph is also a noteworthy point, even though r-having multi-consonant clusters are not very common in Nepali. The rule is that the reph is placed on the last character of the consonant. Some fonts may parse this differently, so the reph is sometimes written over the first consonant that looks like its primary form. This basically means that reph is placed over the first consonant in the cluster that uses the halanta. Of course, it doesn’t make a difference because the r is pronounced first anyway.

र्क्न्स्त् | rknst
Reph on top of the last member of the consonant cluster

र्क्न्स्ट्स् | rknsṭs
In some fonts, the reph goes over because both the primary form and the secondary form of are identical



1. pch
2. kst
3. rml
4. śs
5. ṭphr


1. ख्ग्
2. र्ङ्ल्द् 
3. च्फ्
4. झ्य्ल् 
5. ञ्र् 


A. 1. प्छ्
A. 2. क्स्त्
A. 3. र्म्ल्
A. 4. श्स् 
A. 5. ट्फ्र्

B. 1. khg
B. 2. rṅld
B. 3. cph
B. 4. jhyl
B. 5. ñr