Sandhi Rules

In this article, some basic sandhi-rules concerning ‘ṁ’ and ‘ḥ’ are given.

We just discover this vast subject with:

– simple answers to simple questions

– basic rules about ‘ṁ’

– basic rules about ‘ḥ’

– other questions/answers about ‘ḥ’ just to go a bit further

– explanations about the three ‘gm’-sounds: ‘gm’, ‘ge’, and ‘gge’.

 

Simple answers to simple questions

Question 1: The prayer gaṇapati prārthanā begins with ‘gaṇānāṁ tvā’. Why should we chant it gaṇānān tvā?

Answer 1: ‘ṁ’ is basically a nasal sound. When ‘gaṇānāṁ’ and ‘tvā’ are chanted in the same breath, as ‘tvā’ begins with the dental ‘t’, ‘ṁ’ is changed into ‘n’, the nasal sound of the dental series [t, th, d, dh, n].
(Please refer to the section Sounds  (Main Sounds) from the menu ‘Learning Method’).

So, to sum up: gaṇānāṁ tvāis chanted gaṇānān tvā

[In this article, it will be written:
gaṇānāṁ tvā → gaṇānān tvā ]
(गणानां त्वा → गणानान् त्वा)

 

Question 2: In the 8^th anuvāka of Namakam, we find the famous mantra to Lord Ṡhiva. During the chanting of Sri Rudram, how should it be uttered?

Answer 2: In the usual IAST transliteration (International Alphabet for Sanskrit Transliteration), this mantra is written ‘namaḥ śivāya’. The words are ‘namaḥ’ and ‘śivāya’. According to the RCCS transliteration (Roman Coloured Coding Script) used in the learning documents of this website, the words considered separately are written ‘namaḥa’ and ‘śhivāya’. When these two words are uttered in the same breath, the ‘ḥa’ becomes ‘śh’ due to the following ‘śh’ of ‘śhivāya’. Hence, the ‘śh’ is doubled and the mantra must be uttered as ‘namaśh śhivāya’.

So, to sum up:
namaḥa śhivāya → namaśh śhivāya
(नमः शिवाय → नमश् शिवाय)

 

The foregoing suggests that, as far as ‘ṁ’ and ‘ḥ’ are concerned, words may undergo some modifications (form, pronunciation) when they are associated with other words and uttered in the same breath. The involved rules are called sandhi-rules (sandhi is a Sanskrit word which refers to a point of junction). In a sandhi, when two sounds join together, the latter sound retains its character (in majority of cases), while the preceding sound adopts the characteristic of the succeeding sound.

 

Question 3: What are the most important sandhi-rules that we should know to chant the Vēdas properly?

Answer 3:
– In the usual IAST transliteration, some basic sandhi-rules are not written. Hence, the chanter has to know and apply them during his/her chanting. However, these rules are very simple and are presented in the next two paragraphs: Basic Rules about ‘ṁ’ & Basic Rules about ‘ḥ’.
– The transliteration RCCS used in the materials of this website implements all the sandhi-rules. Hence, the chanter does not need to know them: he/she has only to know how to utter each letter.

 

e.g.:

 

 

Up ↑

 

Basic Rules about ‘ṁ’

– Rule M1: When ‘ṁ’ is followed by a consonant (except the sibilant consonants ‘śh’, ‘ṣh’, ‘s’ and the aspirated consonant ‘h’) or by ‘y’, it is uttered as the nasal sound which belongs to the series of the following consonant.

Here is the recap chart concerning the consonants, with the palatal semi-consonant ‘y’ in addition:

 

Hence, we have the following examples:

– ‘k’ and ‘g’ belong to the guttural series, so:

tāṁ kuru → tāṅ kuru ; śhivāṁ giritra → śhivāṅ giritra
 तां कुरु → ताङ् कुरु ; शिवां गिरित्र → शिवाङ् गिरित्र

– ‘ch’, ‘j’ and ‘y’ belong to the palatal series, so:

 diśhāṁ cha → diśhāñ cha ; puruṣhaṁ jagat → puruṣhañ jagat
दिशां च → दिशाञ् च ; पुरुषं जगत् → पुरुषञ् जगत् 

śhaṁ yōḥo → śhañ yōḥo
शं योः → शञ् योः

– ‘ṭh’ and ‘ḍ’ belong to the cerebral series, so:

kaṁ ṭha → kaṇṭha ; paṁ ḍita → paṇḍita
कं ठ → कण्ठ ; पं डित → पण्डित

– ‘t’ and ‘dh’ belong to the dental series, so:

bahubhyāṁ tava → bahubhyān tava ; vijyaṁ dhanuḥu → vijyan dhanuḥu
बहुभ्यां तव → बहुभ्यान् तव ; विज्यं धनुः → विज्यन् धनुः

– ‘p’ and ‘m’ belong to the labial series, so:

paśhūnāṁ patayē → paśhūnām patayē ; śhalyānāṁ mukhā → śhalyānām mukhā
पशूनां पतये → पशूनाम् पतये ; शल्यानां मुखा → शल्यानाम् मुखा

 

– Rule M2: In Ṛgvēda, when ‘ṁ’ is followed by a vowel, a sibilant consonant ‘śh’, ‘ṣh’, ‘s’ or a ‘h’, it is uttered as a pure ‘m’.

In Ṛgvēda, [10.60.12]:
ayaṁ śhivābhimarśhanaḥa → ayam śhivābhimarśhanaḥa
अयं शिवाभिमर्शनः → अयम् शिवाभिमर्शनः

Up ↑

 

Basic Rules about ‘ḥ’

– Rule H1: When ‘ḥ’ is followed by
 - ‘k’ or ‘kh’ (but not ‘kṣh’), ‘ḥ’ is uttered as the jihvāmūlīya ‘hhhk’, coded ‘[hk]’
 - ‘kṣh’, ‘ḥ’ is normally uttered, with the echo sound, followed by a pause (symbolized by a comma) and the next word.

e.g.:namaḥa kṛtsnavītāya → nama[hk] kṛstnavītāya ; duḥu kha → du[hk]kha

adhaḥa kṣhamācharāḥa → adhaḥa , kṣhamācharāḥa

 

– Rule H2: When ‘ḥ’ is followed by ‘p’ or ‘ph’, ‘ḥ’ is uttered as the upadhmānīya ‘fffp’, coded ‘[fp]’.

e.g.:iṣhavaḥa parā → iṣhava[fp] parā

śhunaṁ naḥa phālā → śhunan na[fp] phālā (using also the basic Rule M1)

 

– Rule H3: When ‘ḥ’ is followed by ‘śh’, ‘ṣh’ or ‘s’, the ‘ḥ’ is respectively replaced by ‘śh’, ‘ṣh’ or ‘s’.

e.g.:śhāntiḥi śhāntiḥi śhāntiḥi → śhāntiśh śhāntiśh śhāntiḥi
शान्तिः शान्तिः शान्तिः → शान्तिश् शान्तिश् शान्तिः

tannaḥa ṣhaṇmukhaḥa prachōdayāt → tannaṣh ṣhaṇmukha[fp] prachōdayāt
तन्नः षण्मुखः प्रचोदयात् → तन्नष् षण्मुखः प्रचोदयात्

tannaḥa sarvaḥa prachōdayāt → tannas sarva[fp] prachōdayāt
तन्नः सर्वः प्रचोदयात् → तन्नस् सर्वः प्रचोदयात्

It should be noted that these three rules do not depend on the vowel preceding the ‘ḥ’.

Up ↑

 

Other Questions/Answers about ‘ḥ’

Question 4: In the prayer sarva dēvatā gāyatrī, the second sentence of each gāyatrī begins with ‘tannas’, ‘tannaṣh’, ‘tannō’, and so on. Are they different forms of the same word? In which cases this word becomes ‘tannō’?

Answer 4: The words ‘tannas’, ‘tannaśh’, ‘tannaṣh’ and ‘tannō’ come from the same word ‘tannaḥa’. The Rule H3 explains the first three forms. Let us see when ‘aḥa’ becomes ‘ō’.

 

– Rule H4: When ‘aḥa’ is followed by ‘a’, a semi-consonant (‘y’, ‘r’, ‘l’, ‘v’), ‘h’ or any soft consonant (‘g’, ‘gh’, ‘ṅ’, ‘j’, ‘jh’, ‘ñ’, ‘ḍ’, ‘ḍh’, ‘ṇ’, ‘d’, ‘dh’, ‘n’, ‘b’, ‘bh’, ‘m’), ‘aḥa’ becomes ‘ō’.

In the following examples, vertical lines indicate that the words are considered separately.

e.g.:| tannaḥa | agniḥi | prachōdayāt | becomes tannō , agniḥi prachōdayāt which is chanted tannō , agni[fp] prachōdayāt (with the Rule H2)

Hence:

| tannaḥa | agniḥi | prachōdayāt | → tannō , agni[fp] prachōdayāt

Similarly:

| tannaḥa | viṣhṇuḥu | prachōdayāt | → tannō viṣhṇu[fp] prachōdayāt

| namaḥa | hiraṇyabāhavē | → namō hiraṇyabāhavē

| tannaḥa | dantiḥi | prachōdayāt | → tannō danti[fp] prachōdayāt

Note: Semi-consonants, ‘h’ and soft consonants are called ghōṣha-consonants (highlighted in orange in the chart below) and often appear together in the following sandhi-rules.

 

Question 5: And what about the case where ‘aḥa’ is followed by a hard consonant?

Answer 5: Let us see the following rule:

 

– Rule H5: When ‘aḥa’ is followed by:
 - a palatal hard consonant (‘ch’, ‘chh’), it becomes ‘śh’ (palatal sibilant).
 - a cerebral hard consonant (‘ṭ’, ‘ṭh’), it becomes ‘ṣh’ (cerebral sibilant)
 - a dental hard consonant (‘t’, ‘th’), it becomes ‘s’ (dental sibilant)

e.g.:
| vājaḥa | cha | mē | → vājaśhcha mē
। वाजः । च । मे । → वाजश्च मे

| namaḥa | tē | → namastē
। नमः । ते । → नमस्ते

 

Question 6: In the Sri Rudram, we often find ‘nama’. What is the corresponding rule?

Answer 6: After ‘nama[hk]’, ‘nama[fp]’, ‘namaḥa’, ‘namaśh’, ‘namaṣh’, ‘namas’, ‘namō’, we shall now examine ‘nama’. Let us see the following rule:

 

– Rule H6: When ‘aḥa’ is followed by a vowel except ‘a’(for ‘a’, see Rule H4) , ‘aḥa’ becomes ‘a’:

e.g.:
| namaḥa | iriṇyāya | → nama , iriṇyāya
। नमः । इरिण्याय । → नम , इरिण्याय 

| namaḥa | āṣhavē | → nama , āṣhavē
। नमः । आषवे । → नम , आषवे 

 

Question 7: The Rules H4 to H6 deal with ‘aḥa’. So…, are there different rules for the cases ‘āḥa’, ‘uḥu’, ‘ūḥu’, and so on?

Answer 7: This question enables us to go deeper into the matter! We saw that the Rules H1 to H3 are the same for all cases (‘aḥa’, ‘āḥa’, ‘iḥi’…). But the other rules depend on the vowel preceding the ‘ḥ’. Let us see what happens with ‘āḥa’ (Rule H7) and the remaining cases ‘uḥu’, ‘ūḥu’,… (Rule H8).

 

– Rule H7: When ‘āḥa’ is followed by a vowel or a ghōṣha-consonant, ‘āḥa’ becomes ‘ā’.

e.g.:
| tēna | dēvāḥa | ayajanta | → tēna dēvā , ayajanta
। तेन । देवाः । अयजन्त । → तेन देवा , अयजन्त 

| dēvāḥa | bhāgam | → dēvā bhāgam
। देवाः । भागम् । → देवा भागम्

 

– Rule H8: When ‘ḥ’ is preceded by any vowel except ‘a’ and ‘ā’, then ‘ḥ’ becomes ‘r’ when it is followed by a vowel or a ghōṣha-consonant.

e.g.:
| svastiḥi | astu | → svastirastu
। स्वस्तिः । अस्तु । → स्वस्तिरस्तु

| chatuḥu | hastam | → chaturhastam
। चतुः । हस्तम् । → चतुर्हस्तम्

| bhūḥu | bhuvaḥa | → bhūr bhuvaḥa
। भूः । भुवः । → भूर् भुवः

 

Question 8: All the cases have been discussed?

Answer 8: Far from it... This field is very vast and complex, with exceptions and “exceptions to the exceptions”. For example, | saḥa | jātaḥa | shoud become sō jātaḥa, according to the Rule H4.
Actually, we have: | saḥa | jātaḥa | → sa jātaḥa. So ‘saḥa’ is a special case.

Up ↑

 

Explanations about the three ‘gm’-sounds: ‘gm’, ‘ge’, and ‘gge’

In Krishna Yajurveda [Sri Rudram, 9^th Namakam] as chanted in the Taittirīya Branch (Taittirīya Ṡhākhā), we find: | kim | śhilāya | → kigm śhilāya.

 

– Rule M3: In Kṛṣhṇa Yajurvēda Taittirīya Ṡhākhā, when ‘m’ or ‘n’ is followed by a vowel, a sibilant consonant (‘śh’, ‘ṣh’, ‘s’), ‘r’ or ‘h’, it is uttered as ‘gm’, except if ‘śh’, ‘ṣh’, ‘s’, ‘r’ or ‘h’ are followed by another consonant/semi-consonant. In the latter case, it is uttered as ‘ge’ (very brief ‘e’) if the preceding vowel is long, and ‘gge’ if the preceding vowel is short.

e.g.:
| bāṇavān | utā | → bāṇavāgm , utā
। बाणवान् । उता । → बाणवा , उता 

| tēṣhām | sahasra | → tēṣhāgm sahasra

| imām | rudrāya | → imāgm rudrāya

| gaṇapatim | havāmahē | → gaṇapatigm havāmahē (in Kṛṣhṇa Yajurvēda)
। गणपतिम् । हवामहे । → गणपति हवामहे
[Note: This mantra belongs to Ṛgvēda also.
In Ṛgvēda: | gaṇapatim | havāmahē | → gaṇapatiṁ havāmahē
। गणपतिम् । हवामहे । → गणपतिं हवामहे
and according to Rule M2, ‘ṁ’ is uttered as a pure ‘m’ (गणपतिम् हवामहे)]]

| ahīn | cha | → ahīgeśhcha
। अहीन् । च । → अहीश्च
(‘śh’ is followed by the consonant ‘ch’, and the ‘ī’ of ‘ahīn’ is long).

| priyam | śhraddhē | → priyaggeśhraddhē
। प्रियम् । श्रद्धे । → प्रियश्रद्धे
(‘śh’ is followed by the semi-consonant ‘r’, and the ‘a’ of ‘priyam’ is short).

 

Up ↑