This article is a brief introduction to how Pāṇini presents his Sanskrit language and grammar rules.

We learned in that, in Sanskrit, vowel sandhi rules require that (see Lesson 5 ):

Rule 1. two similar simple vowels coalesce to from the corresponding long vowel and

Rule 2. the vowels इ (i) ई (ī), उ (u) ऊ (ū) and ऋ (r̥) before a dissimilar vowel or diphthong are converted into the corresponding semivowels, य् (y), व् (v), र् (r) respectively.

Rule 1 could be framed thus:

[Note: I have used a semi-formal system for representing the rules. This may not be a standard way]

<{stem1 | word1} beginning>&{a | ā}&{a | ā}&<{stem2 |word2} end> ::= <{stem1 | word1} beginning>&ā&<{stem2 |word2} end>

For example: rāmasya ācāryaḥ  –> rāmasyācāryaḥ

<{stem1 | word1} beginning>&{i | ī}&{i | ī}&<{stem2 |word2} end> ::= <{stem1 | word1} beginning>&ī&<{stem2 |word2} end>

For example: gardabhī +iyam –> gardabhīyam

<{stem1 | word1} beginning>&{u | ū}&{u | ū}&<{stem2 |word2} end> ::= <{stem1 | word1} beginning>&ū&<{stem2 |word2} end>

For example: sādhu + uktam  –> sādhūktam

<{stem1 | word1} beginning>&{r̥ | ṝ}&{r̥ | ṝ}&<{stem2 |word2} end> ::= <{stem1 | word1} beginning>&ṝ&<{stem2 |word2} end>

Rule 2 could be framed thus:

<{stem1 | word1} beginning>&{ i | ī }&{one of (a | ā, u | ū, r̥ | ṝ)}&<{stem2 |word2} end> ::= <{stem1 | word1} beginning>&y&{one of (a | ā, u | ū, r̥ | ṝ)}&<{stem2 |word2} end>

For example: iti + āha = ityāha

<{stem1 | word1} beginning>&{ u | ū }&{one of (a | ā, i | ī, r̥ | ṝ)}&<{stem2 |word2} end> ::= <{stem1 | word1} beginning>&v&{one of (a | ā, i | ī, r̥ | ṝ)}&<{stem2 |word2} end>

For example: madhu + iva = madhviva

<{stem1 | word1} beginning>&{ r̥ | ṝ }&{one of (a | ā, u | ū, i | ī)}&<{stem2 |word2} end> ::= <{stem1 | word1} beginning>&r&{one of (a | ā, u | ū, i | ī)}&<{stem2 |word2} end>

For example: dātr̥ + asti = dātrasti

Pāṇini’s sūtras representing the above two rules:

The first rule (Rule 1 above) is stated by Pāṇini as:

अकः सवर्णे दीर्घः akaḥ savarṇe dīrghaḥ (6.1.101)

And the second rule (Rule 2 above) is stated by Pāṇini as:

इको यणचि iko yaṇaci (6.1.77)

Now, how does Pāṇini manage to state such complex rules by these terse statements?

He employs two main tricks.

He has a system by which a set of logically connected letters can be represented by just two letters.

He has a compact system for indicating the letter that needs to be replaced, the letter that replaces it and environment in which letters occur.

He has a system by which a set of logically connected letters can be represented by just two letters.

Let us look at how Pāṇini refers to letters and groups of letters of the alphabet.

He groups the letters of the alphabet in a particular way. He enumerates the letters through a set of sutras or aphorisms given in a particular order. These are:

अइउण् a i u ṇ | ऋऌक् r̥ ḷ k | एओङ् e o ṅ | ऐऔच् ai au c | हयवरट् ha ya va ra ṭ | लण् la ṇ | ञमङणनम् ña ma ṅa ṇa na m | झभञ् jha bha ñ | घढधष् gha ḍha dha ṣ | जबगडदश् ja ba ga ḍa da ś | खफछठथचटतव् kha pha cha ṭha tha ca ṭa ta v | कपय् ka pa y | शषसर् śa ṣa sa r | हल् ha l | |

These fourteen sūtras, called the pratyāhāra sūtras (or the śiva sūtras) have all the letters of the alphabet listed. The last pure consonant letter of each sūtra (ṇ , k, c etc.) above is not part of the alphabet list. It is only indicatory (इत् it) used only to show the ending of the sūtra. Also, note that the consonants are shown with the inherent “a” [ha, ya etc.] but we should only use the pure consonants when applying the rules [h, y etc.].

Note that the order in which the sūtras are listed is important.

• Now, if Pāṇini wants to refer to all the vowels he says “a k”. This means starting from “a” and going up to the letter just before “k”. If you look at the sutras above, we see that starting from “a”, we get “a”, “i”, “u”, (we don’t use the “ṇ” as it is only indicatory), “r̥”  and “ḷ”. So “a k” means all the vowels.

• If he wants to say all vowels and diphthongs he will say “a c”. Starting from “a”, we get “a”, “i”, “u”, (we don’t use the ṇ as it is only indicatory), “r̥”, “ḷ” (we don’t use the k as it is only indicatory), “e”, “o”, (we don’t use ṅ) and “ai” and  “au”. We go up to “c”. This therefore indicates all vowels and diphthongs.
• If he wants to list the semivowels, he says “ya ṇ”

We can define groups of letters in this way. (Pāṇini says this: आदिरन्त्येन सहेता ādirantyena sahetā 1.1.71)

• “ha l” means all consonants ha ya va ra ṭ | la ṇ | ña ma ṅa ṇa na m | jha bha ñ | gha ḍha dha ṣ | ja ba ga ḍa da ś | kha pha cha ṭha tha ca ṭa ta v | ka pa y | śa ṣa sa r | ha l | [Note that in Pāṇini’s scheme “ha” appears twice.]
• “ña m” means the nasals ña ma ṅa ṇa na m

A letter doublet standing for a list (eg. “a c”, “ha l” etc.) is called a प्रत्याहार pratyāhāra. Though numerous pratyāhāras can be formed, only 42 are in actual use.

He has a compact system for indicating the letter that needs to be replaced, the letter that replaces it and environment in which letters occur.

• He uses the genitive case (ṣaṣṭhī) to indicate “that which is to be replaced” (षष्ठी स्थानेयोगा ṣaṣṭhī sthāneyogā 1.1.49)
• He uses the locative case (saptamī) to indicate “before this” (ie. the environment after the item to be replaced) or “that which follows (the item to be replaced) ” (तस्मिन्निति निर्दिष्टे पूर्वस्य tasminniti nirdiṣṭe pūrvasya 1.1.66)
• He uses the ablative case (pañcamī) to indicate “after this” (the environment before the item to be replaced) or “that which precedes (the item to be replaced)” (तस्मादित्युत्तरस्य tasmādityuttarasya 1.1.67)
• The actual replacement is rendered in the nominative case (prathamā).

So, if Pāṇini wanted to say “if X occurs before Y, then replace X by Z”, he would put X in the genitive case, Y in the locative case and Z in the nominative. That is, his rule would be (in any order), “X(genitive)Z(nominative)Y(locative)”.

Let us take Rule 2. The vowels इ (i) ई (ī), उ (u) ऊ (ū) and ऋ (r̥) before a dissimilar vowel or diphthong are converted into the corresponding semivowels, य् (y), व् (v), र् (r) respectively.

Pāṇini expresses this rule by इको यणचि iko yaṇaci (6.1.77)

In its uncombined form the rule is इकः यण् अचि ikaḥ yaṇ aci

Now, इकः ikaḥ is in the genitive case (of इक् ik), which indicates “that which is to be replaced”

अचि aci is in the locative case (of अच् ac), which indicates “before” or the environment after the item to be replaced

यण् yaṇ is in the nominative case (यण् yaṇ), which indicates the replacement.

So what Pāṇini is saying is “Before ac replace ik by (corresponding) yaṇ”

Or, before a (dissimilar) vowel replace “i”, “u” etc. by “y” “v” etc. (Note: Rule 1 takes care of similar vowels)

Let us take Rule 1. Two similar simple vowels coalesce to from the corresponding long vowel

Pāṇini expresses rule 1 by अकः सवर्णे दीर्घः akaḥ savarṇe dīrghaḥ (6.1.101)

Now अकः akaḥ is in the ablative case (of अक् ak) which indicates “that which precedes”. ak means all simple vowels.

सवर्णे savarṇe is in the locative case (of सवर्ण savarṇa) which indicates “that which follows”  “savarṇa” means similar

दीर्घः  dīrghaḥ is in the nominative case (दीर्घ  dīrgha) which indicates “the replacement”. “dīrgha” means long.

So what Pāṇini is saying is: “When a simple vowel is followed by a homogeneous vowel, the corresponding long vowel is the single substitute for both the precedent and the subsequent vowels”

[Note 1: this sūtra (6.1.101) is qualified by an earlier sūtra (6.1.84) which says एकः पूर्वपरयोः ekaḥ pūrvaparayoḥ. Here ekaḥ is in the nominative case (of एक eka) , and pūrvaparayoḥ is in the genitive case (dual) (of पूर्वपर pūrvapara). And it means “One is substituted in place of the former and the latter”]

[Note 2: Both above sūtras are qualified by 1.1.50 which says स्थानेऽन्तरतमः  sthāne’ntaratamaḥ (स्थाने अन्तरतमः sthāne antaratamaḥ) which means “When making a substitution (if there are several possibilities by the rule), the nearest (most likely, corresponding, similar) one is the right substitute]

[Note 3: Interestingly, for consonant ending stems like ac, ak etc (ie. for all the pratyāhāras), the genitive and the ablative cases are the same. But it is normally not difficult to figure out whether it is ablative or genitive from the context]

The Magic

Now you can see how Pāṇini has managed to squeeze very complicated rules into simple aphorisms. This was important in the days when writing down things was frowned upon and so you had to get all rules by heart.

Pāṇini is considered to be the father of linguistics and grammar. His Aṣṭādhyāyī, written around 6th century BCE, is the first currently-available formal grammar described of any language. It consists of 3959 Sūtras, or aphorisms set out in eight chapters (hence the name Aṣṭādhyāyī), that sets out in terse style the rules of the Sanskrit language as was used then.

It is not easy to understand or use Pāṇini’s sutras for simple grammar study. It is quite complicated as you could see from the examples given above.

If you want to understand a bit more of how the Aṣṭādhyāyī is set out, please refer “The Aṣṭādhyāyī of Pāṇini”, Volumes I and II, by Srīśa Candra Vasu, published by Motilal Banarasidas.