Whaddaya Tawki’na’bowt? Intro to Sanskrit Noun Use

Sanskrit recognizes eight ways to use a noun. We tend to call them “cases.” In Sanskrit they are called विभक्ति (vibhakti), and are referred to by number (“first”, “second”, “third”, etc). For some reason, perhaps showing the influence of Western Indology, most people use confusing Latin names for the cases.

I will now introduce the cases, and try to make the Latin names easier to handle and remember.

Using the cases involves “declination” – which means changing the ending of the word. Exactly how you change the word’s ending depends on many things: the case you want to use; whether the word is singular, plural or dual; whether it is masculine, feminine, or neuter; and its natural ending. For the examples here, I will use simply use singular, masculine nouns ending naturally in -a.

The eight cases are:

  1. Subject (“Nominative”)
  2. Objective (“Accusative”)
  3. Method (“Instrumental”) “by”
  4. Purpose (“Dative”) “for”
  5. Origin (“Ablative”) “from”
  6. Possession (“Genitive”) “of”
  7. Location (“Locative”) “in”
  8. Evoking (“Vocative”) “hey”

Cases 2-7 show how the subject relates to the object. Let’s go case by case.

Case 1: The Subject — “Nominative”

The Latin here is easy, because “nominative” is the base of our English word “name” and that’s what the case does, it names the subject of the sentence.

कृष्णो वनं गच्छति

kṛṣṇo vanaṁ gacchati

Krishna goes to the forest

Here the word “Krishna” is in the first case (nominative), as kṛṣṇaḥ (which changes to kṛṣṇo due to sandhi). This shows Krishna is the subject of the sentence.

Case 2: The Objective / Cause of Action— “Accusative” 

The Latin here is a little weird. It sounds like “accuse.” When we accuse someone we claim that they are the cause of something (“he is accused of murder”), and that’s what this case does: it identifies the cause / objective of the subject’s action. (If you look carefully, you can see the word “cause” (cuse) in accusative.)

कृष्णो वनं गच्छति

kṛṣṇo vanaṁ gacchati

Krishna goes to the forest

Here the word “forest” (vana) is in the second case (accusative), as vanam, to show that it is the objective of Krishna’s movement. The forest is the motivator that causes  his movement.

Case 3: The Method to Accomplish the Objective — “Instrumental”

The Latin here is easy. An instrument is the tool we use to accomplish an objective, and that’s what this case is for: it indicates how the subject accomplishes its objective. In English we usually what this case accomplishes by using the words by (“I’ll go by car”) or with (“I drink with a straw”), or sometimes using on (“I will get there on a bike”).

कृष्णो पादेन वनं गच्छति

kṛṣṇo pādena vanaṁ gacchati

Krishna goes to the forest on foot.

Here the word “foot” (pāda) is in the third case (instrumental), as pādena, to show that his feet are the instrument Krishna uses to go to the forest. 

Case 4: Purpose of the Objective — “Dative”

This Latin is particularly difficult, because it sounds like “date,” which misleads me into thinking this is a case for describing time. In Latin, however, date means “what will be obtained,” and that’s what the case does – it shows what the subject hopes to gain from the objective. In English, we usually accomplish this with the words for (“I work for money”) or to (“I work to make money”).

कृष्णो वनमानन्दाय गच्छति

kṛṣṇo vanam ānandāya gacchati

Krishna goes to the forest for joy.

Here the word “joy” (ānanda) is in the fourth case (dative), as ānandāya, to show that joy is what the Krishna hopes to gain by going to the forest.

Case 5: Origin of the Objective — “Ablative”

The Latin here is also difficult. It conveys the sense of something “abating” (going away). This case lets us show where something comes from – literally or conceptually. Conceptual movement from one thing towards another is how Sanskrit shows causality and also how it makes comparisons. 

In English, we usually accomplish this using words like from (“I come from New York”), and out of (“He overeats out of stress.”). We can also use because of and due to for accomplishing the conceptual-conceptual sense of this case.

कृष्णो ग्रामात्चरति

kṛṣṇo grāmāt carati

Krishna walks from the village.

Here the word “village” (grāma) is in the fifth case (ablative), as grāmāt, to show Krishna’s movement brings him from the village. This is an example using the case literally.

भ्रमन्ति सुखात्

bhramanti sukhāt

They wander due to happiness.

Here the word “happiness” (sukha) is in the fifth case (ablative), as sukhāt, to show that they wander as a result of their happiness. This is an example using the case conceptually, to show causality.

कृष्णो चन्द्रात्सुन्दरः

kṛṣṇo candrāt sundaraḥ

Krishna is more beautiful than the moon.

Here the word “moon” (candra) is in the fifth, ablative case (candrāt) to show that beauty moves away from the moon, towards Krishna. This conceptual movement is how Sanskrit makes a comparison expressing that Krishna is more beautiful than the Moon. 

Case 6: Possessive — “Genitive”

The Latin word for this has to do with producing (and therefore owning) something. Parents, for example, are progenitors, who give us their genes. In English we use “-’s” or “of” to accomplish what this case does.

कृष्णस्य शक्तिः

kṛṣṇasya śaktiḥ

Krishna’s potency. -or- The potency of Krishna.

Here the word kṛṣṇa is in the sixth case (genitive), as kṛṣṇasya, to show that Krishna is the producer, and thus the possessor, of the śakti.

Case 7: Location — “Locative”

The Latin here is mercifully simple. We use this case to express a position in either space or time. In English we accomplish this using words like “in”, “on”, “at”, and so on.

मनः कृष्णे निवेशयेत्

manaḥ kṛṣṇe niveśayet

Invest your mind in Krishna. -or- Set your mind on Krishna.

Here, the word kṛṣṇa is in the seventh case (locative), as kṛṣṇe. It therefore signifies the location upon which / in which the mind’s thoughts and emotions should exist.

Case 8: Evoking — “Vocative”

Here, too, the Latin is mercifully easy. We use this case to call to (“evoke”) someone, or directly address them by name. In English we just use a person’s name without modification, but we can make our evocation more explicit by by using hey, as in, “Hey John” or “Hey you!” In older English they used O, as in “O Lord.”

हरे कृष्ण

Hare Kṛṣṇa

Hey Hara! Hey Krishna!

In Sanskrit too, this case involves little change to the original names. With kṛṣna (a male noun ending in “a”) the eighth case (vocative) involves no change at all. With hara (a female noun ending in “a”) the eighth case (vocative) changes it slightly, to hare.

Advertisements

Gita 2.14 Sandhi Illustration

+ जायते + मृयते + वा + कदाचित् = न जायतेमृयते वा कदाचित्

na + jāyate + mṛyate + vā + kadācit = na jāyate mṛyate vā kadācit

No sandhi at all applies to any of these words yet, because they are all words ending in a vowel, followed by a consonant.

+ अयम + भूत्वा + भाविता + वा + + भूयः = नायंभूत्वाभावुता वा न भूयः

na + ayam + bhūtvā + bhāvitā + vā + na + bhūyaḥ = nāyaṁ bhūtvā bhāvitā vā na bhūyaḥ

The “na” at the beginning of the line will change unvoiced “t” to voiced “d” at the end of kadācit from the end of the previous line. This is because in sankrit only the midway point of the śloka is really a “new line.” The rule working there is: “Stops without voice are like me and you. We get what we lack when we meet a sādhu.”

“Na + ayam” becomes nāyam, on the “A makes everyone stronger” rule.

Nāyam + bhūtvā becomes nāyaṁ bhūtva, on the rule: “Simple M becomes a dot, way up in the air.”

There is no change for the rest of the words, for they all end in vowels, followed by consonants.

अजः + नित्यः + शाश्वतः + अयम + पुराणः = अजोनित्यः शाश्वतो ‘यं पुराणः

ajaḥ + nityaḥ + śāśvataḥ + ayam + purāṇaḥ = ajo ‘nityaḥ śāśvato ‘yaṁ purāṇaḥ

Ajaḥ + nityaḥ = ajo nityaḥ, following the rule: ““When S (Ḥ) comes with A, there’s a few things to test.  Is that consonant voiced? Then make A O and drop S(Ḥ).” Nasals are considered voiced sounds.

Nityaḥ + śāśvataḥ has no change. It doesn’t follow the above rule because sibilants (ś in this case) are not voiced sounds.

Śāśvataḥ + ayaṁ = śāśvato ‘yam. If follows the “make A O and drop Ḥ” rule (as above), and since the voiced sound is “a”, “A too disappears.”

‘yam + purānaḥ = ‘yaṁ purānaḥ, following the rule, “Simple M becomes a dot, way up in the air.”

न हन्यतेहन्यमानेशरीरे

na hanyate hanyamāne śarīre

There is no sandhi on this line at all, because all the words end in vowels, followed by consonants. However, the “na” at the beginning of the line causes purāṇaḥ at the end of the previous line to become purāno – following the “make A O and drop Ḥ” rule, since nasals are voiced.

Gita 2.13 – Sanskrit Sandhi Illustration

With Sandhi

देहिनोस्मिन्यथादेहे कौमारां यौवनं जरा ।
तथा देहान्तरप्राप्तिर्धीरस्तत्र न मुह्यति ।।

dehino ’smin yathā dehe kaumāraṁ yauvanaṁ jarā
tathā dehāntara-prāptir dhīras tatra na muhyati

Without Sandhi

देहिनः अस्मिन् यथा देहे कौमारम् यौवनम् जरा ।
तथा देह अन्तर प्राप्तिः धीरः तत्र न मूह्यति ।।

dehinaḥ asmin yathā dehe kaumāram yauvanam jara
tathā deha antara prāptiḥ dhīraḥ tatra na muhyati

The Sandhi Rules Used

देहिनोस्मिन् = देहिनःअस्मिन्

Dehino ‘smin = dehinaḥ + asmin

When S (Ḥ) comes with A, there’s a few things to test.
Is that consonant voiced? Then make A O and drop S(Ḥ).
With vowels, if A is the voice than it too
disappears. If not, then only the S (Ḥ) has to shoo”

“Dehinaḥ” is an example of “aḥ” ending a word. The next sound is a vowel. All vowels are voiced, so it triggers the voicing rule, converting the “A” to an “O” (dehinoḥ) and dropping the “S” (dehino). Note: the final “s” has already converted to “Ḥ”. Finally, since the vowel is an “A” then “it too disappears” (dehino ’smin)

कौमारं यौवनं जरा = कौमारम् यौवनम् जरा

Kaumāraṁ yauvanaṁ jarā = kaumāram + yauvanam + jarā

“M becomes a simple dot, way up in the air.”

Coming before a voiced consonant, M comverts to Ṁ.

देहान्तर = देह अन्तर

Dehāntara = deha + antara

“A makes everyone stronger.”

प्राप्तिर्धीरस्तत्र = प्राप्तिः धीरः तत्र

Prāptir dhīras tatra = prāptiḥ + dhīraḥ + tatra

Dhīraḥ: “When at the end, S becomes the visarga
But if met without voice, it converts to that varga”

The “ḥ” at the end of dhīraḥ is an “s” in disguise, it has “become the visarga.” It is met by a sound that has no voice, “t” so it converts to the sibilant in “t”’s varga (dental), becoming “S” (dhīras)

Prāptiḥ: “When met with a voice, then the S becomes R.”

Again, the “ḥ” at the end of prāptiḥ is an “s” in disguise,. It is met by a sound that has voice, “dh” so it becomes “r” (prāptir)

Sanskrit Sandhi Rap

Here are the rules of Sanskrit set to rap!

Sandhi’s about how sounds combine
So bring in the beat and let the
vowels rhyme

A makes everyone stronger

The rest only make themselves longer.

The others, well they merely step down

Into their semi-vowelized sound

But outdoors, E and AI won’t conform
They hide their vowelized form

And if meeting with A, they won’t shrink.

Instead, A will disappear with a wink.

Yooooooo. Now bring out them basic “stop” consonants, and throw in the H…

Stops without voice are like me and you

We get what we lack when we meet a sādhu

H acts voiceless when it looks behind

To add the voice of any stop that it can find

T has got to be the hardest sound around

So ways to make soft more abundantly abound

Before stopping on the palate-roof, T transforms,

To the very same sound, it completely conforms

But, if that palate sound is S, then they both become C

And that S gets extra breath, revealing what it used to be

Yo. Alright, check it. Where them silky slinky sibilants at boy?

When at the end, S becomes the visarga

But if met without voice, it converts to that varga

When met with a voice, then the S becomes R

(But don’t let it double, that would take it too far)

When S comes with A, there’s a few things to test

Is the consonant voiced? Then make A O and drop S

With vowels, if A is the  voice than it too

disappears. If not, only S has to shoo

S can also come with a friend named “Long A”

When that happens, just  make that S go away

Oh, by the way, AS is AR’s very close friend

They act the same before the voiceless, and the same when at the end

Yo, we almost done with this joint. What we got left? Bring the basil nasals to the tables…

If you stop before a nasal, you become a nasal too
Though to your own varga you remain true

Simple M becomes a dot, way up in the air.
But with the rest there’s lots of other rules, so take care.

With a voiceless in a varga with an S defined
Use it, and make the N a dot above the line

With the voiced in such a varga, theres not much to do
Make the nasal match that varga and you’re through

A shiny Ś takes a nasal’s place
And uses CH, to hide its face

Lengthy L likes to swell and double
Forcing N above the line like a little bubble

N can double too, except the curvy sort,
When between vowels, with the first short

So now you got the keys

To the epiphonies

That make speech so easy it can flow like the breeze

Explanations

A makes everyone stronger

The rest only make themselves longer.

The others, they merely step down

Into their semi-vowel sound

But outdoors, E and AI won’t conform
They hide their vowelized form

And if meeting with A, they won’t shrink.

Instead, A will disappear with a wink.

~~~~ Explanation ~~~~~~~~~~~~~~~~~~~~~

A makes everyone stronger

(a) strengthens any vowel it meets. It increases its own length, strengthens / (i/ī) to (e), / (u/ū) to (o), / (ṛ/ṝ) to अर् (ar), (e) to (ai), and (o) to (au).

The rest only make themselves longer 

Vowels other than (a) only strengthen/lengthen themselves. += (a+a=ā), += (i+i=ī), += (u+u=ū). Further combinations cannot strengthen past this. I.E. += (a+ā=ā).

The others, they merely step down
Into their semi-vowel sound

When vowels other than (a) meet other vowels, the do not strengthen one another, but the first vowel will “step down” into its semi-vowel equivalent. य् (iy), उव् (uv), एअ[य्] (ea[y]), ओअव् (oav), ऐआ[य्] (aiā[y]), औआव् (auāv), ऋर् (r).

But outdoors, E and AI won’t conform
They hide their semi-vowel form

“Outdoors” refers to external sandhi. In that environment, / (e/ai) hide their semi-vowel component, and thus result in / (a/ā) respectively.

And if meeting with A, they won’t shrink
Instead, A will disappear with a wink

If / (e/ai) meet with (a) in external sandhi, they simply cause the (a) to disappear, leaving behind an apostrophe-like mark.

~~~~ Examples ~~~~~~~~~~~~~~~~~~~~~~~

A makes everyone stronger”

(a) strengthens any vowel it meets. It increases its own length, strengthens / (i/ī) to (e), / (u/ū) to (o), / (ṛ/ṝ) to अर् (ar), (e) to (ai), and (o) to (au).

a + i = e a + e = ai a + ai = ai

a + u = o a + o = au a + au = au

a + ṛ = ar

+ = + = + =

+ = + = + =

+ = अर्

When a vowel comes after “a/ā” they combine to produce the  guṇa vowel.

a + i = e

हा + = हे

mahā + īśa = maheśa

(Great lord)

a + u = o

म्ब + दर = म्बोदर

lamba + udara = lambodara

(Potbelly)

a + ṛ = ar

हा + षि = हर्षि

mahā + i = mahari

(Great sage)

When a guṇa vowel comes after an “a/ā” they combine to  produce the vṛddhi vowel.

a + e = ai

त्र + एव = अत्रै

atra + eva = atraiva

(Right here)

a + o = au

सा + ओदनं पचति = सौदनं पचति

sā + odana pacati = saudana pacati

(She cooks rice)

When a vṛddi vowel comes after an “a/ā” they also combine to produce the vṛddhi vowel, since it is not possible to further intensify the sound.

a + ai = ai

च्छ + ऐश्वर्यम् = गच्छैश्वर्यम्

gaccha + aiśvaryam = gacchaiśvaryam

(Attain power)

a + au = au

च्छ + दुम्बरम् = च्छौदुम्बरम्

yaccha + audumbaram = yacchaudumbaram

(Give the Audumbara fruit)

“The rest only make themselves longer.” 

Vowels other than (a) only strengthen/lengthen themselves. += (a+a=ā), += (i+i=ī), += (u+u=ū). Further combinations cannot strengthen past this. I.E. += (a+ā=ā).

a/i/u/ṛ + a/i/u/ṛ = ā/ī/ū/ṝ

/// + /// = ///

When two vowels of the same type come next to each other, they become one, as the elongated version of themselves.

a + a = ā

रामेण + + गच्छत् + सीता = रामेण सहागच्छत्सीता

rāmeṇa + saha + agacchat + sītā = rāmeṇa sahāgacchat sīta

[Sītā goes with Rāma.]

i + i = ī

गच्छामि + ति + वदति = गच्छामीति वदति

gacchāmi + iti + vadati = gacchāmīti vadati

[“I go,” he says.]

u + u = ū

कुरु + त्तम = कुरूत्तम

kuru + uttama = kurūttama

[The best Kuru]

“For others they merely step down
Into their semi-vowel sound”

When vowels other than (a) meet other vowels, the do not strengthen one another, but the first vowel will “step down” into its semi-vowel equivalent. य् (iy), उव् (uv), एअ[य्] (ea[y]), ओअव् (oav), ऐआ[य्] (aiā[y]), औआव् (auāv), ऋर् (r).

i y u v r

e a[y] o av

ai    ā[y] au  āv

य् उ व् ऋ र्

अ[य्] ओ अव् 

आ[य्] औ आव्

i + a = ya

आगच्छामि + हम् = आगच्छाम्यहम्

āgacchāmi + aham = āgacchāmyaham.

(I’m coming)

u + a = va

जयतु + र्यपुत्रः = जयत्वार्यपुत्रः

jayatu + āryaputraḥ = jayatryaputraḥ

(May the Āryan descendent be victorious)

ṛ + i = ri

पितृ + प्सितम् = पित्रीप्सितम्

pitṛ + īpsitam = pitpsitam

(Father’s wish)

If the first vowel is a guṇa or vṛddhi vowel, it must first be devolved into its basic components. Then the same rule can be applied.

au + e = (ā+u) + e = āve

भौ + = भावे

ubhau + eva = ubhāveva

[both]

“au” devolves into its components: “ā+u.” The “u” then mutates to its semi-vowel form, “v.”

[internal] e + a = (a+i) + a = aya

जे + = जय

je + a = jaya

(to win)

“e” devolves to its components: “a+i.” The “i” then mutates to its semi-vowel form, “y.”

o + a = (a+u) + a = ava

भो + = भव

bho + a = bhava

(to be)

“o” devolves to its components: “a+u.” The “u” then mutates to its semi-vowel equivalent, “v.”

But outdoors, E/AI won’t conform
They hide their semi-vowel form

“Outdoors” refers to external sandhi. In that environment, / (e/ai) hide their semi-vowel component, and thus result in / (a/ā) respectively.

e + i = a i

स्वर्गे + न्द्रः = स्वर् इन्द्रः

svarge + indraḥ = svarga indraḥ

(Indra is in heaven)

The “e” in svarge devolves to “a + i.” The “i” then mutates to its semivowel equivalent, giving “ay.” Finally, the “y” disappears.

ai + a = ā a

तस्मै + ददात् = तस्मा ददात्

tasmai + adadāt = tasmā adadāt

[He gave]

The “ai” in tasmai devolves to: “ā + i.” The “i” then mutates to “y” before disappearing.

And while meeting with A will not shrink.
Instead, A disappears with a wink.

If / (e/ai) meet with (a) in external sandhi, they simply cause the (a) to disappear, leaving behind an apostrophe-like mark.

e + a = e

ते + अब्रुवन् = ते ब्रुवन्

te + abruvan = te ‘bruvan

[they said]

Stops without voice are like me and you

We get what we lack when we meet a sādhu

H acts voiceless when it looks behind

To add the voice of any stop that it can find

T has got to be the hardest sound around

So ways to make soft more abundantly abound

Before stopping on the palate-roof, T transforms,

Into that very same sound, it completely conforms

But, if that palate sound is S, then they both become C

And that S gets extra breath, revealing what it used to be

~~~~ Explanation ~~~~~~~~~~~~~~~~~~~~~

Sounds without voice are like me and you
We get what we lack when we meet a sādhu

When a non-voiced sound is followed by a voiced sound, the non-voiced sound changes the first to the voiced letter in its वर्ग (varga). So it “gets what it lacks” (voice) when it meets a letter that has it (a “sādhu,” an entity with a desired characteristic)

H acts voiceless when it looks behind

To add the voice of any stop that it can find

“Looking behind” means that H blends in this manner when it is the first letter of a word, “looking back” towards the last letter of the previous word. “Adding” the voice means that it adds a voiced sound to its fundamental aspirated nature, it’s “h”-ness. The particular voice it adds is from the same वर्ग as the preceding sound.

T has got to be the hardest sound around
So ways to make it soft more abundantly abound

It is relatively difficult to produce the clear “T” sound (त्), and difficult to switch from it to a different consonant. So, there are a few rules, additional to the above, in which the T will change even when not meeting a voiced sound.

Before stopping on the palate-roof, T transforms
Into that very same sound, it completely conforms

If त् (t) comes before any palate- or roof-group consonant (/वर्ग), it becomes just like whatever follows it, minus any aspiration that might be there.

But, if that palate sound is S, then they both become C
And that S gets extra breath, revealing what it used to be

If त् (t) comes before श् (ś), the त् (t) becomes च् (c) and the श् (ś) becomes छ् (ch).

~~~~ Examples ~~~~~~~~~~~~~~~~~~~~~~~

“Sounds without voice are like me and you.
We get what we lack when we meet a sādhu.”

When a non-voiced sound is followed by a voiced sound, the non-voiced sound changes the first to the voiced letter in its वर्ग (varga).

क्रोधात् + भवति + संमोहः = क्रोधाद्भवति संमोहः

krodhāt + bhavati + saṁmohaḥ = krodhād bhavati saṁmohaḥ

(Anger gives rise to confusion)

The त् (t) of क्रोधात् (krodhāt) comes before a voiced consonant, (bh), so it changes to the voiced consonant in its own varga, द् (d)

आसीत् + असुरः = आसीदसुरः

āsīt + asura = āsīd asura

(There was a demon)

The त् (t) of आसीत् (āsīt) comes before a vowel, which is a voiced sound, so it changes to the voiced consonant in its own varga, द् (d).

संमोहात् + स्मृतिविभ्रमः = संमोहात्समृतिविभ्रमः

saṁmohāt + smṛtivibhramaḥ = saṁmohāt smṛtivibhramaḥ

(Confusion rattles the memory)

The (t) of संमोहात् (saṁmohāt) does not change at all, because it is followed by a sound that is not voiced.

H acts voiceless when it looks behind

To add the voice of any stop that it can find.

“Looking behind” means that H blends in this manner when it is the first letter of a word, “looking back” towards the last letter of the previous word. “Adding” the voice means that it adds a voiced sound to its fundamental aspirated nature, it’s “h”-ness. The particular voice it adds is from the same वर्ग as the preceding sound.

एतत् + हरति + चौरः = एतद्धरति चौरः

etat + harati + cauraḥ = etaddharati cauraḥ

[The thief steals that]

First, the final “t” of etat becomes “d” by following the rule, “Sounds without a voice are like me and you, they get what they lack when they meet a sādhu.” Here “h” is treated as a voiced sound. But the change also reflects on the “h” itself transforming it to the same voiced sound, but with aspiration. It the “h” of harati becomes “dh” after changed the “t” of etat into “d” – resulting in etad dharati.

वाक् + हि + देवता = वाग्घि देवता

vak + hi + devatā = vagghi devatā

[Speech is certainly a goddess]

“Before stopping on the palate-roof, T transforms,
Into that very same sound, it completely conforms.”

If त् (t) comes before any palate- or roof-group consonant (/वर्ग), it becomes just like whatever follows it, minus any aspiration that might be there.

तत् + चिकीर्षति = तच्चिकीरषति

tat + cikīrṣati = tac cikīrṣati

(He wants to do that)

The त् (t) at the end of तत् (tat) is followed by च् (c), which is a palate-group consonant, so the त् (t) becomes a च् (c). There is no aspiration in  च् (c), so that is not an issue. If the च् (c) in चिकीषति (cikīrṣati) was a छ् (ch) however, it would make no difference. The त् (t) would still become a च् (c), because the rule is to remove any aspiration.

तत् + जहाति = तज्जहाति

tat + jahāti = taj jahāti

(He gives up on that)

“Now, if that palate sound is S, then they both become C
And that S gets extra breath, revealing what it used to be.”

If त् (t) comes before श् (ś), the त् (t) becomes च् (c) and the श् (ś) becomes छ् (ch).

तत्+श्रुत्वा+कुपितस्+अभवत्+शिवः = तच्छ्रुत्वा कुपितोभवच्छिवः

tat+śrutvā+kupitas+abhavat+śivaḥ = tac chrutvā kupito ‘bhavac chivaḥ

(Hearing that enraged Śiva)

The final त् (t) of तत् (tat) becomes च् (c) because it meets the श् (ś) of श्रुत्व (śrutva), and the श् (ś) of श्रुत्व (śrutva) also changes to छ् (ch). The same thing happens at the juncture of अभवत्+शिवः (abhavat+śiva).

Also, the meeting of कुपितस् + अभवत् (kupitas + abhavat) changes per the rule “When S comes with A, there’s a few things to test: With consonant voice, make A O and drop S; With vowels, if A is the voice than it too disappears. If not, only the S has to shoo.”

When at the end, S becomes the visarga

But if met without voice, it converts to that varga

When met with a voice, then the S becomes R

(But don’t let it double, that would take it too far)

When S comes with A, there’s a few things to test

Is the consonant voiced? Then make A O and drop S

With vowels, if A is the voice than it too

disappears. If not, then only S will have to shoo

S can also come with a friend named “Long A”

When that happens, only make that S go away

Oh, by the way, AS is AR’s very close friend

They act the same before the voiceless, and the same when at the end

~~~~ Explanation ~~~~~~~~~~~~~~~~~~~~~

“When at the end, S becomes the visarga.”

When स् (s) comes at the end of a word that stands alone or is at the very end of a statement, change it to the visarga, : (ḥ).

“If met without voice, it converts to that varga.”

If स् (s) comes before an unvoiced consonant: Alter the स् (s) to match the वर्ग (varga) of the following consonant.

“When met with a voice, S becomes R.”

If स् (s) comes before a voiced sound (which includes both consonants and vowels, except /a): Change the स् (s) to र् (r).

“(But don’t let it double, that would take it too far.)”

If र्र (rr) results from the above process, drop the first र् (r) and make the previous vowel long.

“When S comes with A, there’s a few things to test”

There are special considerations for अस् (-as)…

“With consonant voice, make A O and drop S”

If अस् (-as) comes before a Voiced Consonant: Change the (a) to (o) and drop the स् (s).

“With vowels, if A is the voice than it too [disappears.]”

If अस् (-as) comes before the (a) vowel, do the same as above, and also drop the following (a).

“If not, only the S has to shoo.”

If अस् (-as) comes before a vowel other than (a), only drop the स् (s), not the vowel.

S also might come with a friend called “Long A”

When that happens, again, only S goes away.

This is a special rule for अास् (-ās). When it comes before any vowel or voiced consonant: just drop the स् (s)

Oh, by the way, AR and AS are very close friends.
They act the same before the voiceless, and the same when at the ends.

At the end of a statement, or before a non-voiced sound, अर् (-ar) behaves exactly as if it was अस् (-as). However, it does not change in other situations where -ar does.

~~~~ Examples ~~~~~~~~~~~~~~~~~~~~~~~

“When at the end, S becomes the visarga.”

When स् (s) comes at the end of a word that stands alone or is at the very end of a statement, change it to the visarga, : (ḥ).

अग्निस् = अग्निः

agnis. = agniḥ.

“If met without voice, it converts to that varga.”

If स् (s) comes before an unvoiced consonant: Alter the स् (s) to match the वर्ग (varga) of the following consonant.

रामस् + चलति = रामश्चलति

rāmas + calati = rāmaś-calati

(Rāma wanders)

हरिस् + चलति = हरिश्चलति

haris + calati = hariś-calati

(Hari wanders)

विष्णोस् + छया = विष्णोश्छया

viṣṇos + chayā = viṣṇoś-chayā

(Viṣṇu’s shadow)

/ (ca/cha) is from the palate-group, so the स् (s) becomes the palate-sibilant, श् (ś).

हरिस् + टीकां करोति = हरिष्टीकां करोति

haris + ṭīkāṁ karoti = hariṣ-ṭīkāṁ karoti

(Hari writes a commentary)

(ṭa) is from the roof-group, so the स् (s) becomes the roof-sibilant, ष् (ṣ).

अग्निस् + तीक्ष्णः = अग्निस्तीक्ष्णः

agnis + tīkṣṇaḥ = agnis-tīkṣṇaḥ

(Fire is fierce)

(ta) is from the tooth-group, so the स् (s) remains as the tooth-sibilant, स् (s).

हरिस् + पश्यति = हरिः पश्यति

haris + paśyati = hariḥ paśyati

(Hari sees)

रामस् + पश्यति = रामः पश्यति

rāmas + paśyati = rāmaḥ paśyati

(Rāma sees)

(pa) is from the lip-group, so the स् (s) would become a lip-sibilant, an “f” sound, which is approximated in Sanskrit by : (ḥ).

हरिस् + खनति = हरिः खनति

haris + khanati = hariḥ khanati

(Hari digs)

(kha) is from the throat-group, so the स् (s) changes to the throat-sibilant, approximated in Sanskrit by : (ḥ).

रामस् + सीतां पश्यति = रामः सीतां पश्यति

rāmas + sītaṁ paśyati = rāmaḥ sītaṁ paśyati

(Rāma sees Sītā)

Before another sibilant, स् (s) also transforms to : (ḥ).

“When met with a voice, S becomes R.”

If स् (s) comes before a voiced sound (which includes both consonants and vowels, except /a): Change the स् (s) to र् (r).

गतिस् + नास्ति = गतिर्नास्ति

gatis + nāsti = gatir-nāsti

(Impossible)

हरेस् + गौस् = हरेर्गौः

hares + gaus = harer-gau

(Hari’s cow)

अग्निस् + इव = अग्निरिव

agnis + iva = agnir-iva

(Fire-like)

विष्णोस् + आयुधम् = विष्णोरायुधम्

viṣṇos + āyudham = viṣṇoyudham

(Viṣṇu’s weapon)

“(But don’t let it double, that would take it too far.)”

If र्र (rr) results from the above process, drop the first र् (r) and make the previous vowel long.

अग्निस् + रोचते = अग्नीरोचते

agnis + rocate = agnī-rocate

(Fire shines)

अग्निस् (agnis) first becomes अग्निर् (agnir), but because the next sound is the र् (r) of रोचते (rocate), the र् (r) at the end of अग्निर् (agnir) is dropped and the previous vowel is made long, resulting in अग्नी (agnī).

“When S comes with A, there’s a few things to test”

There are special considerations for अस् (-as)…

“With consonant voice, make A O and drop S”

If अस् (-as) comes before a Voiced Consonant: Change the (a) to (o) and drop the स् (s).

रामस् + गच्छति = रामो गच्छति

rāmas + gacchati = rāmo gacchati

(Rāma goes)

A short “a” preceded the “s”, which was followed by “g,” a voiced consonant. The “a” changed to “o.” The “s” was destroyed.

पश्यतस् + राज्ञः = पश्यतो राज्ञः

paśyatas + rājñaḥ = paśyato rājñaḥ

(While the king watches)

A short “a” preceded the “s”, which was followed by “r,” a voiced consonant. The “a” changed to “o.” The “s” was destroyed.

“With vowels, if A is the voice than it too [disappears.]”

If अस् (-as) comes before the (a) vowel, do the same as above, and also drop the following (a).

रामस् + अयम् = रामोयम्

rāmas + ayam = ramo’yam

(He is Rāma)

पश्यतस् + अर्जुनस्य = पश्यतोर्जुनस्य

paśyatas + arjunasya = paśyato’rjunasya

(While Arjuna watches)

A short “a” preceded the “s”, which was followed by “a.” The first “a” changed to “o.” The “s” was destroyed. The second “a” was also destroyed.

“If not, only the S has to shoo.”

If अस् (-as) comes before a vowel other than (a), only drop the स् (s), not the vowel.

रामस् + उवाच = राम उवाच

rāmas + uvāca = rāma uvāca

(Rāma said)

बुद्धस् + इव विद्यया = बुद्ध इव विद्यया

buddhas + iva vidyayā = buddha iva vidyayā

(Wise like Buddha)

S also might come with a friend called “Long A”

When that happens, again, only S goes away.

This is a special rule for अास् (-ās). When it comes before any vowel or voiced consonant: just drop the स् (s)

हतास् वीरास् गच्छन्ति स्वर्गलोकम् = हता वीरा गच्छन्ति स्वर्गलोकम्

hatās vīrās gacchanti svargalokam = hatā vīrā gacchanti svargalokam

(Slain heroes go to paradise)

Pronoun exception:

If the root word is सः (saḥ/“he”) or ऐषः (eṣaḥ/“that”), just drop the स् (s), don’t change the (a) to an (o).

ऐषस् + शुकस् + अस्ति = ऐष शुकोस्ति

eṣas + śukas + asti = eṣa śuko ‘sti.

(That is a parrot)

सस् + कृष्णस् = स कृष्णः

sas + kṛṣṇas. = sa kṛṣṇaḥ.

(He is Krishna)

Oh, by the way, AR and AS are very close friends.
They act the same before the voiceless, and the same when at the ends.

At the end of a statement, or before a non-voiced sound, अर् (-ar) behaves exactly as if it was अस् (-as). However, it does not change in other situations where -ar does.

पुनर् + पुनर् = पुनः पुनः

punar + punar. = puna puna.

(Again and again)

The “p” of the second “punar” is an unvoiced sound, so the “ar” at the end of the first “punar” changes to “aḥ.” The same happens to the second word, because it is at the end of a statement

पुनर् + पुनर् + दानवान्हन्ति = पुनः पुनर्दानवान्हन्ति

punar + punar + dānavān hanti = punaḥ punar dānavān hanti

(Again and again he kills demons)

The first punar is followed by an unvoiced sound (p), so it changes to punaḥ. The second punar is followed by a voiced sound, so it does not change.

प्रातर् + आगमिष्यति = प्रातरागमिष्यति

prātar + āgamiṣyati = prātar-āgamiṣyati

(He will come in the morning)

Again, the -ar followed by a voiced sound does not change.

If you stop before a nasal, you’ll become a nasal too
Though to your own varga you must remain true

M becomes a simple dot, way up in the air.
But with the rest there’s lots of other rules to tangle in your hair.

With a voiceless in a varga with an S defined
Use it, and make the N a dot above the line

With the voiced in such a varga, theres not much to do
Just make the nasal match that varga and you’re through

A shiny Ś will take a nasal’s place
And use CH, to hide its own face

Lengthy L likes to swell and double
Forcing N above the line like a little bubble

N can double too, except the curvy sort,
if it comes between two vowels, when the first is humbly short

Explanation

If you stop before a nasal, you’ll become a nasal too
Though to your own varga you must remain true

When stop-consonants meet nasal sounds, they become the nasal in their own वर्ग.

M becomes a simple dot, way up in the air.
But with the rest there’s lots of other rules to tangle in your hair.

When nasal “m” meets any consonant, it becomes anusvara ( ँ). This is a wonderfully simple rule, but the rules are a little more complicated for the other nasal sounds.

With a voiceless in a varga with an S defined
Use it, and make the N a dot above the line

If followed by a consonant in a वर्ग varga with a sibilant, and that consonant is non-voiced: convert the n to ṁ and add the वर्ग varga’s sibilant.

With the voiced in such a varga, theres not much to do
Just make the nasal match that varga and you’re through

If followed by a voiced stop, the “n” transforms to match the वर्ग (varga) of that consonant.

A shiny Ś will take a nasal’s place
And use CH, to hide its own face

If followed by the palatal sibilant (ś), the “n” becomes palatal too, (ñ) and the “ś” becomes “ch”

Lengthy L likes to swell and double
Forcing N above the line like a little bubble

If followed by “l” the “l” doubles and the n becomes anusvara.

N can double too, except the curvy sort,
if it comes between two vowels, when the first is humbly short

If “n” comes between two vowels, and the first vowel is short, the “n” doubles. This is true for any nasal except “ñ.”

Examples

If you stop before a nasal, you’ll become a nasal too
Though to your own varga you must remain true

When stop-consonants meet nasal sounds, they become the nasal in their own वर्ग.

एतत् + मा + कुरु = एतन्मा कुरु

etat + mā + kuru = etan mā kuru

[Don’t do that]

“T” meets “m,” so it converts to a nasal sound. Since “T” is produced on the teeth, it converts to the वर्ग nasal sound, produced on the teeth: “n.”

वाक् + मय = वाऩ्मय

vak + maya = vaṅ maya

[Made of words]

“K” meets “m”, so it converts to a nasal sound. Since “K” is produced in the throat it converts to the वर्ग nasal sound, produced in the throat: “ṅ.”

M becomes a simple dot, way up in the air.
But with the rest there’s lots of other rules to tangle in your hair.

When nasal “m” meets any consonant, it becomes anusvara ( ँ). This is a wonderfully simple rule, but the rules are a little more complicated for the other nasal sounds.

रामस् + वनम् + गच्छति = रामो वनं गच्छति

Rāmas + vanam + gacchati = rāmo vanaṁ gacchati

[Rāma goes to the forest]

The “m” comes before a consonant so it becomes “ṁ.” Additionally, the “-as” of “rāmas” comes before “va” in “vanam” so it changes to “rāmo,” following the rule: “When S comes with A, there’s a few things to test; With consonant voice, make A O and drop S.”

ऋषीनाम् +पुस्तकम् + पठति = ऋषीनां पुस्तकं पठति

ṛṣīnām + pustakam + paṭhati = ṛṣīnāṁ pustakaṁ paṭhati

[He reads the book of the sages]

All the “m”s at the ends of the words convert to “ṁ” because they are followed by consonants.

With a voiceless in a varga with an S defined
Use it, and make the N a dot above the line

If followed by a consonant in a वर्ग varga with a sibilant, and that consonant is non-voiced: convert the n to ṁ and add the वर्ग varga’s sibilant.

कस्मिन + चित् + नगरे + अवसत् + राजा = कस्मिंश्चिन्नगरे ‘वसद्राजा

kasmin + cit + nagare + avasat + rājā = kasmiṁś cin nagare ‘vasad rājā

[A king lives in a city]

There are many sandhis in this sentence. The one that illustrates the current rule occurs between kasmin and cit. The “c” of “cit” is वर्ग, produced on the palate – a group of sounds that include a sibilitant, “ś.” So, the “n” of “kasmin” converts to “ṁś.”

The sandhi between cit and nagare illustrates the rule that a stop-consonant before a nasal will convert to the nasal of its own वर्ग. T is वर्ग, produced on the teeth, so it converts to the dental nasal, “n.”

The sandhi between nagare and avasat illustrates the rule that, “[E] meeting A will not shrink. Instead, A disappears with a wink.”

The sandhi between avasat and rājā illustrates the rule, “Sounds without voice are like me and you. We get what we lack when we meet a sādhu.” The “t” ending “avasat” has no voice, but the “r” begining “rājā” does, so, when it meets the “t” it gives it a voice, transforming it into “d.”

मूर्खान् + त्यजति + पण्डितः = मूर्खांस्त्यजति पण्डितः

mūrkhān + tyajati + paṇḍitaḥ = mūrkhāṁs tyajati paṇḍitaḥ

[The wise leave fools.]

The “n” of “mūrkhān” meets “t,” a dental. So it transforms into “ṁs” (ṁ plus the dental sibilant).

With the voiced in such a varga, theres not much to do
Just make the nasal match that varga and you’re through

If followed by a voiced stop, the “n” transforms to match the वर्ग (varga) of that consonant.

दानवान् + जयति + इन्द्रः = दानवाञ्जयतीन्द्रः

dānavān + jayati + indraḥ = dānavāñ jayatīndraḥ

[Indra conquers demons.]

The “n” of dānavān meets “j”, a voiced stop, so it converts to the “j”’s वर्ग (varga), वर्ग (palatal), so it converts to the palatal nasal, “ñ.” The “i”s of jayati and indraḥ join, following the rule, “The rest only make themselves stronger.”

Note that “n” is already a dental sound, so if it meets “d,” the voiced dental stop, there will be no change.

A shiny Ś will take a nasal’s place
And use CH, to hide its own face

If followed by the palatal sibilant (ś), the “n” becomes palatal too, (ñ) and the “ś” becomes “ch”

मधुरान् + शब्दान् + शृणोति = मधुराञ्छब्दाञ्छृणोति

madhurān + śabdān + śṛnoti = madhurāñ chabdañ chṛnoti

[He hears sweet sounds]

Lengthy L likes to swell and double
Forcing N above the line like a little bubble

If followed by “l” the “l” doubles and the n becomes anusvara.

उत्तमान् + लोकान् + लभते + धर्मज्ञः = उत्तमांल्लोकांल्लभते धर्मज्ञः

uttamān + lokān + labhate + dharmajñaḥ = uttamāṁ llokāṁ llabhate dharmajñaḥ

[The knower of morality attains the highest realms]

N can double too, except the curvy sort,
if it comes between two vowels, when the first is humbly short

If “n” comes between two vowels, and the first vowel is short, the “n” doubles. This is true for any nasal except “ñ.”

स्तुवन् + आगच्छति = स्तुवन्नागच्छति

stuvan + āgacchati = stuvann āgacchati

[He comes, praising]

हसन् + इव = हसन्निव

hasan + iva = hasann iva

[As if laughing]

Complete Sandhi Mnemonic, Explanation, and Examples for स् (-s)

~~~~ Rhymes ~~~~~~~~~~~~~~~~~~~~~~~~

When at the end, S becomes the visarga.
If met without voice, it converts to that varga.

When met with a voice, S becomes R.
(But don’t let it double, that would take it too far.)

When S comes with A, there’s a few things to test
With consonant voice, make A O and drop S

With vowels, if A is the voice than it too
disappears. If not, only the S has to shoo.

S also might come with a friend called “Long A”
When that happens, again, only S goes away.

~~~~ Explanation ~~~~~~~~~~~~~~~~~~~~~

“When at the end, S becomes the visarga.”

When स् (s) comes at the end of a word that stands alone or is at the very end of a statement, change it to the visarga, : (ḥ).

“If met without voice, it converts to that varga.”

If स् (s) comes before an unvoiced consonant: Alter the स् (s) to match the वर्ग (varga) of the following consonant.

“When met with a voice, S becomes R.”

If स् (s) comes before a voiced sound (which includes both consonants and vowels, except /a): Change the स् (s) to र् (r).

“(But don’t let it double, that would take it too far.)”

If र्र (rr) results from the above process, drop the first र् (r) and make the previous vowel long.

“When S comes with A, there’s a few things to test”

There are special considerations for अस् (-as)…

“With consonant voice, make A O and drop S”

If अस् (-as) comes before a Voiced Consonant: Change the (a) to (o) and drop the स् (s).

“With vowels, if A is the voice than it too [disappears.]”

If अस् (-as) comes before the (a) vowel, do the same as above, and also drop the following (a).

“If not, only the S has to shoo.”

If अस् (-as) comes before a vowel other than (a), only drop the स् (s), not the vowel.

S also might come with a friend called “Long A”
When that happens, again, only S goes away.

This is a special rule for अास् (-ās). When it comes before any vowel or voiced consonant: just drop the स् (s)

~~~~ Examples ~~~~~~~~~~~~~~~~~~~~~~~

“When at the end, S becomes the visarga.”

When स् (s) comes at the end of a word that stands alone or is at the very end of a statement, change it to the visarga, : (ḥ).

अग्निस् = अग्निः

agnis. = agniḥ.

“If met without voice, it converts to that varga.”

If स् (s) comes before an unvoiced consonant: Alter the स् (s) to match the वर्ग (varga) of the following consonant.

रामस् + चलति = रामश्चलति

rāmas + calati = rāmaś-calati

(Rāma wanders)

हरिस् + चलति = हरिश्चलति

haris + calati = hariś-calati

(Hari wanders)

विष्णोस् + छया = विष्णोश्छया

viṣṇos + chayā = viṣṇoś-chayā

(Viṣṇu’s shadow)

/ (ca/cha) is from the palate-group, so the स् (s) becomes the palate-sibilant, श् (ś).

हरिस् + टीकां करोति = हरिष्टीकां करोति

haris + ṭīkāṁ karoti = hariṣ-ṭīkāṁ karoti

(Hari writes a commentary)

(ṭa) is from the roof-group, so the स् (s) becomes the roof-sibilant, ष् (ṣ).

अग्निस् + तीक्ष्णः = अग्निस्तीक्ष्णः

agnis + tīkṣṇaḥ = agnis-tīkṣṇaḥ

(Fire is fierce)

(ta) is from the tooth-group, so the स् (s) remains as the tooth-sibilant, स् (s).

हरिस् + पश्यति = हरिः पश्यति

haris + paśyati = hariḥ paśyati

(Hari sees)

रामस् + पश्यति = रामः पश्यति

rāmas + paśyati = rāmaḥ paśyati

(Rāma sees)

(pa) is from the lip-group, so the स् (s) would become a lip-sibilant, an “f” sound, which is approximated in Sanskrit by : (ḥ).

हरिस् + खनति = हरिः खनति

haris + khanati = hariḥ khanati

(Hari digs)

(kha) is from the throat-group, so the स् (s) changes to the throat-sibilant, approximated in Sanskrit by : (ḥ).

रामस् + सीतां पश्यति = रामः सीतां पश्यति

rāmas + sītaṁ paśyati = rāmaḥ sītaṁ paśyati

(Rāma sees Sītā)

Before another sibilant, स् (s) also transforms to : (ḥ).

“When met with a voice, S becomes R.”

If स् (s) comes before a voiced sound (which includes both consonants and vowels, except /a): Change the स् (s) to र् (r).

गतिस् + नास्ति = गतिर्नास्ति

gatis + nāsti = gatir-nāsti

(Impossible)

हरेस् + गौस् = हरेर्गौः

hares + gaus = harer-gau

(Hari’s cow)

अग्निस् + इव = अग्निरिव

agnis + iva = agnir-iva

(Fire-like)

विष्णोस् + आयुधम् = विष्णोरायुधम्

viṣṇos + āyudham = viṣṇoyudham

(Viṣṇu’s weapon)

“(But don’t let it double, that would take it too far.)”

If र्र (rr) results from the above process, drop the first र् (r) and make the previous vowel long.

अग्निस् + रोचते = अग्नीरोचते

agnis + rocate = agnī-rocate

(Fire shines)

अग्निस् (agnis) first becomes अग्निर् (agnir), but because the next sound is the र् (r) of रोचते (rocate), the र् (r) at the end of अग्निर् (agnir) is dropped and the previous vowel is made long, resulting in अग्नी (agnī).

“When S comes with A, there’s a few things to test”

There are special considerations for अस् (-as)…

“With consonant voice, make A O and drop S”

If अस् (-as) comes before a Voiced Consonant: Change the (a) to (o) and drop the स् (s).

रामस् + गच्छति = रामो गच्छति

rāmas + gacchati = rāmo gacchati

(Rāma goes)

A short “a” preceded the “s”, which was followed by “g,” a voiced consonant. The “a” changed to “o.” The “s” was destroyed.

पश्यतस् + राज्ञः = पश्यतो राज्ञः

paśyatas + rājñaḥ = paśyato rājñaḥ

(While the king watches)

A short “a” preceded the “s”, which was followed by “r,” a voiced consonant. The “a” changed to “o.” The “s” was destroyed.

“With vowels, if A is the voice than it too [disappears.]”

If अस् (-as) comes before the (a) vowel, do the same as above, and also drop the following (a).

रामस् + अयम् = रामोयम्

rāmas + ayam = ramo’yam

(He is Rāma)

पश्यतस् + अर्जुनस्य = पश्यतोर्जुनस्य

paśyatas + arjunasya = paśyato’rjunasya

(While Arjuna watches)

A short “a” preceded the “s”, which was followed by “a.” The first “a” changed to “o.” The “s” was destroyed. The second “a” was also destroyed.

“If not, only the S has to shoo.”

If अस् (-as) comes before a vowel other than (a), only drop the स् (s), not the vowel.

रामस् + उवाच = राम उवाच

rāmas + uvāca = rāma uvāca

(Rāma said)

बुद्धस् + इव विद्यया = बुद्ध इव विद्यया

buddhas + iva vidyayā = buddha iva vidyayā

(Wise like Buddha)

S also might come with a friend called “Long A”
When that happens, again, only S goes away.

This is a special rule for अास् (-ās). When it comes before any vowel or voiced consonant: just drop the स् (s)

हतास् वीरास् गच्छन्ति स्वर्गलोकम् = हता वीरा गच्छन्ति स्वर्गलोकम्

hatās vīrās gacchanti svargalokam = hatā vīrā gacchanti svargalokam

(Slain heroes go to paradise)

Pronoun exception:

If the root word is सः (saḥ/“he”) or ऐषः (eṣaḥ/“that”), just drop the स् (s), don’t change the (a) to an (o).

ऐषस् + शुकस् + अस्ति = ऐष शुकोस्ति

eṣas + śukas + asti = eṣa śuko ‘sti.

(That is a parrot)

सस् + कृष्णस् = स कृष्णः

sas + kṛṣṇas. = sa kṛṣṇaḥ.

(He is Krishna)

Easy to Memorize Vowel Sandhi

Vowels strengthen themselves.
(a) even strengthens others.

Other vowels are at least polite,
they become semi-vowels, to make themselves lite.

But (e)’s not polite to the “a” (अ).
 (e) scares (a) away


Vowels strengthen themselves.

a/i/u/ṛ + a/i/u/ṛ = ā/ī/ū/ṝ

/// + /// = ///

(a) even strengthens others.

a + i = e. :|: . a + e = ai . :|: . a + ai = ai

a + u = o. :|: . a + o = au. :|: . a + au = au

a + ṛ = ar

+ = ए. :|: . + = ऐ. :|: . + =

+ = ओ. :|: . + = औ. :|: . + =

+ = अर्

Other vowels [not /a] are at least polite,
they become semi-vowels, to make themselves lite.

i y. :|: . u v. :|: . r

e a[y]. :|: . o av

ai    ā[y]. :|: . au  āv

य्. :|: . उ व्. :|: . ऋ र्

अ[य्]. :|: . ओ अव् 

आ[य्]. :|: . औ आव्

But (e) is not polite to (a).
 (e) scares (a) away

a + e = a

+ =

Gita 2.12 – Simple Sanskrit Sandhi Evaluation

When speaking, we start with words, and then blend them using sandhi to arrive at finished sentences. Listening or reading, however, goes in reverse. We start with finished sentences, and need to undo the sandhi to arrive at the words, and thus understand the meaning.

It is only in the early stages of learning and using a language that this seems tedious. Even in English we do this all the time (though not while reading, since English sandhi doesn’t change how words are spelled).

Let’s see which words we can pick out from famous fully formed sentences quoted from Bhagavad-Gītā. Let’s look specifically at text 12 of chapter 2. Since we haven’t learned consonant sandhi yet, we’ll leave these out.

It is impossible to do reverse sandhi without knowing some vocabulary words, but this verse mostly uses very simple words.

न त्वेवाहं जातु नासं न त्वं नेमे जनाधिपाः
न चैव न भविष्यामः सर्वे वयमतः परम् ।।

na tvevāhaṁ jātu nāsaṁ
na tvaṁ neme janādhipāḥ
na caiva na bhaviṣyāmaḥ
sarve vayam ataḥ param

The First Pada त्वेवाहं जातु नासं

There are tons of sandhis in the first line alone. Take the second word, त्वेवाहं (tvevāhaṁ). the first “e” with a “v” in front of it alerts us that त्व (tv) was originally तु (tu). The rule here is: “If the adjoining vowels are dissimilar and the first is not “a/ā” the combination changes the first vowel to its semi-vowel equivalent.”

“U” and “e” are dissimilar (“e” is a complex vowel, guṇa, in the “i” group), so the first vowel in the combination (“u” in this case) will morph into its semi-vowel equivalent (“v” in this case). Hence त्वेव (tveva) is really तु एव (tu eva).

In एवाहं (evāhaṁ) We find a long ā, and recognize some simple words on either side of it, so we can see that एवाहं (evāhaṁ) is a combination of two words, एव (eva) and अहं (ahaṁ), joined by the simplest sandhi rule that “a/ā + a/ā = ā.”

जातु (jātu) is a word, not a combination of smaller words. It means, “ever.” If we didn’t know that we might try to split it into (ja) and अतु (atu) or आतु (ātu). But if we search a dictionary we will find that atu / ātu are not words, and jātu is.

नासं (nāsaṁ) is a combination of two words, (na – “not”) and आसं (āsaṁ – “I was”). These are joined with the same simplest sandhi rule, “a/ā + a/ā = ā.” How do we know the original word is आसं (āsaṁ) and not असं (asaṁ)? We don’t, not by sandhi anyway. By sandhi alone either word could be intended. However, there is no word असं (asaṁ) in use in this grammatical context in Sanskrit, so we know it must be आसं (āsaṁ).

The first pada is therefore na (not) तु tu (but) एव eva (definitely) अहं ahaṁ (I) जातु jātu (ever) na (not) आसं (āsaṁ) I was. Rendered in natural English, this means “But, definitely, I was never non-existent.”

The Second Pada त्वं नेमे जनाधिपाः

The next quarter of the śloka begins, त्वं (na tvaṁ). We might want to change this “v” into a “u” but (a) tvam is a word, and (b) aṁ is not. त्वं (na tvaṁ) means, “Nor you.”

The next word, नेमे (neme) is a combination of two words, (na – not) and इमे (ime – these). The rule is “a + i = e” (or, “a + another vowel intensifies that vowel to its guṇa or vṛddhi version”).

Next, जनाधिपाः janādhipāḥ (protectors of people, “kings”) is a compound word formed by joining जन (jana – “people”) with अधिपाः (adhipāḥ – “protectors”) using again the simple and frequently used rule, “a/ā + a/ā = ā”.

So we now can render the English as: “But I was definitely never non-existent; nor you; nor these kings.”

The Third Pada  चैव भविष्यामः

The third pada begins चैव (na caiva). चैव (caiva) is a combination of two simple words, (ca) and एव (eva), joining with the simple sandhi rule of “a + e = ai” (Essentially the same rule as for नेमे neme.)

The other two words in the pada are not combined – (na, not) and भविष्यामः (bhaviṣyāmaḥ, we will be). It is tempting to think that the long ā is a result of sandhi but familiarity with the grammatical declinations will help us avoid wasting time pursuing that hunch.

We can now render this line as, “Nor, definitely, will we ever not exist.”

The Fourth Pada सर्वे वयमतः परम्

The fourth pada has no vowel sandhi.

सर्वे (sarve, “all”) वयम् (vayam, “us”) अतः (ataḥ, “hence”) परम् (param, “ultimate/supreme”). Perhaps this can be rendered, “Hence we are all supreme.” However, it is more commonly rendered with ataḥ param meaning “from hence, evermore”, giving us, “We exist, and so shall we evermore.”

So, we can understand the entire śloka to render in English as:

But definitely, I was never non-existent;
Nor were you; Nor any of these kings.

Nor definitely will we ever not exist;
All of us will exist forevermore.

Sanksrit for the rest of us