Consonant sandhi between words
Also known as: vyañjana-sandhi, hal-sandhi
Consonant sandhi is the name for sandhi changes between a consonant and some other sound. Here is a simple example of consonant sandhi:
सीता वनम् गच्छति → सीता वनं गच्छति
sītā vanam gacchati → sītā vanaṃ gacchati
Sita goes to the forest.
Consonant sandhi is complex. In this lesson, we will focus on the common sandhi changes that occur between two words. These changes also apply between the two words in a compound.
Reducing consonants
Sanskrit has many consonants. But Sanskrit words end only in specific consonant sounds. So we must first apply a few rules to make sure that our word ends in a valid consonant sound.
Before we study the specific rules, let's first see some examples of what these rules do. In the examples below, the words on the left don't follow the rules and are not valid words. After applying the rules, we get the correct results on the right:
अगच्छन्त् → अगच्छन्
agacchant → agacchan
they wentवाच् → वाक्
vāc → vāk
speechदिश् → दिक्
diś → dik
directionराज् → राट्
rāj → rāṭ
kingसमिध् → समित्
samidh → samit
(sacred) wood, kindlingनरस् → नरः
naras → naraḥ
man
Now, here are the changes that we should apply to words ending in consonants. First, a word is not usually allowed to end in multiple consonants. If a word does end in multiple consonants, we keep only the first of those consonants. You can see some examples of this below:
अगच्छन्त् → अगच्छन्
agacchant → agacchan
they wentपश्यन्त् → पश्यन्
paśyant → paśyan
while seeingप्राञ्च् → प्राञ्
prāñc → prāñ
facing, opposite
There are rare exceptions, usually if the second-to-last consonant is r:
ऊर्ज् → ऊर्ज्
ūrj → ūrj
strength, vigor (no change)
Second, consonants pronounced at the hard palate generally become k. (ñ becomes ṅ.) A very small number of words, such as rāj, use ṭ instead:
वाच् → वाक्
vāc → vāk
speechदिश् → दिक्
diś → dik
directionप्राञ् → प्राङ्
prāñ → prāṅ
facing, oppositeराज् → राट्
rāj → rāṭ
king
Third, the remaining consonant becomes unaspirated and unvoiced if it has an unaspirated and unvoiced version. In the first example below, dh has an unvoiced and unaspirated version t, so it becomes t. In the second example, m has no unvoiced or unaspirated version, so it stays the same:
समिध् → समित्
samidh → samit
(sacred) wood, kindlingवनम् → वनम्
vanam → vanam
forest (no change)
Finally, -s and -r become the visarga:
नरस् → नरः
naras → naraḥ
manद्वार् → द्वाः
dvār → dvāḥ
door
Sanskrit words do not usually end in semivowels (ya ra la va). So by the end of this process, we are left with eight possible final sounds: k, ṭ, t, p, ṅ, n, m, and the visarga.
Rules for k, ṭ, t, and p
k, ṭ, t, and p use the same voicing as the sound that follows them:
तत् वनम् → तद् वनम्
tat vanam → tad vanamतत् उदकम् → तद् उदकम्
tat udakam → tad udakamतत् फलम् → तत् फलम्
tat phalam → tat phalam
They also become nasal when the sound that follows them is nasal:
वाक् न → वाङ् न
vāk na → vāṅ naराट् न → राण् न
rāṭ na → rāṇ naतत् न → तन् न
tat na → tan naककुप् न → ककुम् न
kakup na → kakum na
If the second sound is h, then we usually get this change:
वाक् ह → वाग् घ
vāk ha → vāg ghaराट् ह → राड् ढ
rāṭ ha → rāḍ ḍhaतत् ह → तद् ध
tat ha → tad dhaककुप् ह → ककुब् भ
kakup ha → kakub bha
Some learners find it helpful to see these changes in a table:
k | ṭ | p | |
---|---|---|---|
ṅ | ṇ | m | nasal sounds |
g* | ḍ* | b* | h |
g | ḍ | b | other voiced sounds |
k | ṭ | p | unvoiced sounds |
In the table above, * means that the following h shifts its point of pronunciation to match the first sound.
Extra rules for -t
-t changes frequently. If the next consonant is pronounced at the hard palate (like ca) or the roof of the mouth (like ṭa), -t changes to a sound with the same point of pronunciation:
तत् चित्रम् → तच् चित्रम्
tat citram → tac citramतत् जलम् → तज् जलम्
tat jalam → taj jalam
If l is the second sound, it becomes l:
तत् लभस्व → तल् लभस्व
tat labhasva → tal labhasva
Obtain that.
And if the next sound is ś, we get this change:
तत् शोचन्ति → तच् छोचन्ति
tat śocanti → tac chocanti
As before, some learners find it helpful to see these changes in a table:
t | |
---|---|
n | nasal sound |
c | c, ch |
j | j, jh |
ṭ | ṭ, ṭh |
ḍ | ḍ, ḍh |
l | l |
c (ś becomes ch) | ś |
d (h becomes dh) | h |
d | other voiced sounds |
t | other sounds |
Rules for -n
Like -t, -n changes often. If the next consonant is a voiced consonant at the hard palate (like ja) or the roof of the mouth (like ṇa), -n changes to the nasal sound with the same point of pronunciation:
नरान् जयामि → नराञ् जयामि
narān jayāmi → narāñ jayāmi
If l is the second sound, it becomes a nasal l:
तान् लभन्ते → ताँल् लभन्ते
tān labhante → tā̐l labhante
In front of c/ch, ṭ/ṭh, or t/th, n becomes ṃś, ṃṣ, and ms respectively:
तान् चरन्ति → तांश् चरन्ति
tān caranti → tāṃś carantiतान् तरन्ति → तांस् तरन्ति
tān taranti → tāṃs taranti
And if the next sound is ś, two different outcomes are possible:
तान् शोचन्ति → ताञ् शोचन्ति
tān śocanti → tāñ śocanti
(option 1)तान् शोचन्ति → ताञ् छोचन्ति
tān śocanti → tāñ chocanti
(option 2)
Again, some learners find it helpful to see these changes in a table:
n | |
---|---|
ṃś | c, ch |
ñ | j, jh |
ṃṣ | ṭ, ṭh |
ṇ | ḍ, ḍh |
ṃs | t, th |
nasal l | l |
ñ (ś optionally becomes ch) | ś |
n | other sounds |
Rules for -m
-m becomes the anusvāra when consonants follow:
सीता वनम् गच्छति → सीता वनं गच्छति
sītā vanam gacchati → sītā vanaṃ gacchati
And it may optionally become the nasal sound that matches the following consonant:
फलम् चरामि → फलञ् चरामि
phalam carāmi → phalañ carāmiफलम् खादामि → फलङ् खादामि
phalam khādāmi → phalaṅ khādāmi
In modern times, this change is usually not written down. But, it is often used in spoken Sanskrit.
Rules for the visarga
We learned about visarga sandhi already. Please see the previous lesson for details.
Review
There are many small details to consonant sandhi. But this lesson is a complete summary of its most common patterns.