The Anubandha


Recall that an anubandha is a "technical" letter that has no meaning in real Sanskrit. This is the definition that we used when talking about the Shiva Sutras. But this definition obscures the real power of the anubandha. Let me redefine the term: an anubandha is an "indicatory" letter that is given a special meaning within the Ashtadhyayi itself, and words behave differently based on the anubandha attached to them. By using a considerable number of anubandha letters, Panini concisely describes the grammatical behavior of all of Sanskrit grammar.

The anubandha might seem like a vague concept, but it's not too difficult to grasp.

An Illustration

Consider three sets of verbs: "sing" and "swim," "teach" and "catch," and "walk" and "wonder." The verbs in the first set will change their vowel when used in the past tense: "sing" becomes "sang," and "swim" becomes "swam." The verbs in the second set will use "aught" for their ending: "teach" becomes "taught," and "catch" becomes "caught." Finally, the verbs in the third set just have "ed" added to the end of them: "walk" becomes "walked," and "wonder" becomes "wondered."

Based on these divisions, we can assign different names to the different sets. We could call the first set the "vowel change verbs," the second set the "-ght verbs," and the third set the "regular change verbs." But imagine that we wanted to talk about these verbs with the same brevity and concision that Panini uses in his Ashtadhyayi. These names are all much too long. What do we do? The answer is simple: we make these names shorter! We could name the first set "vowel," the second set "ght," and the third set "ed." But suppose we wanted to make this even briefer. We could reduce these names to one letter: "V" for the first set, "G" for the second set, and "E" for the third set. With these terms, we could talk about "singV," "catchG," or "walkE" and know exactly how each verb behaves. Each single-letter terms is essentially an anubandha.


Panini's special term for the anubandha is it, a contraction of the word iti. We can talk about a particular anubandha by attaching it to the end of it. For example, the anubandha letter k is called kit.

The examples in the next lesson will explore the details of kit, which is used to describe both the tvā suffix of the gerund and the ta suffix of the PPP. Since Panini uses hardly any verbs in the Ashtadhyayi, there's no need to wait until you finish the chapter on verbs. You can proceed to the next lesson whenever you're ready.

1.3.2 - 1.3.9 — Which letters are anubandhas?

In seven rules, Panini states which letters are actually used as anubandhas. Of these seven rules, the first relies on a term defined in 1.1.8. ?et's take a look at that rule.

1.1.8 — anunāsika

This is a saṃjñā rule defining anunāsika, meaning "nasalized." Something is anunāsika if it is a vacana, or utterance, made with mukha-nāsika. This term is a dvandva compound meaning "the mouth and the nose." So, a sound that uses both the mouth and the nose is called anunāsika. The anusvāra, which is a "pure nasal" sound, is not anunāsika. However, the nasal consonants are anunāsika, as are nasalized vowels, which are not used in Classical Sanskrit.

Now we can read the rules below. These rules define the letters that are used as anubandhas. I've provided all seven of these rules so that you can see how parts of one rule carry over to the next. Together, these rules show that the order of rules is not arbitrary.

1.3.2 — Nasalized vowels

This rule may not make much sense. I provide it here because parts of it carry over to the seven rules that follow.

upadeśa is a technical term for an "original" expression, one of the supposedly fundamental parts of Sanskrit. Every upadeśa has an anubandha, or perhaps two. The word is used in the locative case, but there is no substitution going on here. Rather, the locative shows that the rules that follow occur "in," or within the scope of, an upadeśa.

'janunāsika is originally ac-anunāsika, meaning a nasalized vowel. The last word of the rule is it, meaning an anubandha.

So, this is the rule: In an upadeśa, a nasalized vowel is an anubandha.

1.3.3 — Final consonants

With the sandhi split, this rule becomes: hal antyam, meaning "a final consonant." The rule doesn't make any sense — that is, unless we consider the terms upadeśe and it. Then, the rule is plain: In an upadeśa, a final consonant is an anubandha.

1.3.4 — Exceptions for noun endings

vibhakti refers to the "case endings" added to the end of a noun. tu-s-māḥ has three parts:

So, what does the rule mean? It's tempting to translate the rule like this: tavarga, s, and m are not indicatory in a case affix. The problem with this rendition is that it talks about all indicatory letters even though it follows halantyam, which is a rule that only deals with final consonants. If na vibhaktau tusmāḥ were really so general, then it would come before halantyam, not after it.

Therefore, this rule must be a limiting case of halantyam. We should translate the rule like this: tavarga, s, and m are not indicatory when at the end of a case affix.

1.3.5 — Initial syllables

We have two words here. The first word is ādiḥ, meaning "the beginning" or "initial." This word shows that we're not dealing with a final consonant any more. The second word is ñi-ṭu-ḍavaḥ, which is formed of three anubandha terms: ñi, ṭu, and ḍu. So, we should translate the rule like this: ñi, ṭu, and ḍu are all indicatory when at the beginning of a word. Consider this rule and note this important point: an anubandha does not have to be just a single letter.

Perhaps you're wondering: If tu refers to all of tavarga, then shouldn't ṭu refer to ṭavarga? Yes, but not in this context. I can only explain why this is the case when we study rule 1.3.7 below.

1.3.6 — ṣa in a suffix

A pratyaya is a suffix. ādi carries over. So, we should translate the rule like this: The initial of a suffix is indicatory.

1.3.7 — cavarga and ṭavarga

ādi and pratyaya both carry over. The one word in this rule is cuṭu, meaning cavarga and ṭavarga. Therefore, we should translate the rule like this: In a suffix, initial cavarga and ṭavarga letters are indicatory.

Perhaps you're wondering why 1.3.7 isn't just part 1.3.6. According to one translator, the reason is that this rule isn't always true: although cavarga and ṭavarga can be indicatory at the beginning of a word, they aren't always. I don't like that this ambiguity exists, but we won't have to deal with it for now.

When we studied rule 1.3.5, I mentioned that ṭu referred to the syllable ṭu and not to all of ṭavarga. You should be able to see why that's the case by studying this rule. If the first ṭu did refer to all of ṭavarga, then the ṭu in this rule would be redundant. So, it must be the case that the first ṭu is a syllable. We can also infer this fact from the term's context; the first ṭu exists with two other anubandha terms, neither of which is a single letter.

1.3.8 — More initial letters

ādi and pratyaya both carry over. This rule is composed of two words. The first is laśaku, which refers to l, ś, and the sounds in kavarga. The second word is ataddhite. Recall that taddhita is the term for a secondary suffix, like tva or maya. So, ataddhite means "not in a secondary suffix."

So, we can translate the rule like this: l, ś,, and the letters in kavarga are indicatory at the beginning of a sufffix, as long as the suffix is not a taddhita suffix.

Now, let's look at one last rule.

1.3.9 — Removing the anubandha

This highly significant rule describes how it letters behave. It is a substitution rule: An anubandha is replaced with lopa. What is lopa? It means "dropping" or "removal," and it shows that the term is lost. So, we can translate the rule as simply as this: The anubandha is always deleted. I've added the "always" here to clarify that this rule describes the anubandha's behavior.

This rule clarifies what a anubandha really is: a technical marker that has absolutely no meaning in real Sanskrit. These markers are just used in Panini's system of analysis.