सर्वनामानि — the pronouns and their suppletive stems

A class defined by list

1.1.27

Every grammatical category so far in this series — vowel, stop, root, affix — has been defined by some property you can check. सर्वनाम is defined differently: by enumeration. That has a real, checkable consequence, and it is the first thing worth seeing.

सर्वादीनि सर्वनामानिThe definition is a list

— 1.1.27: सर्व, and the words that follow it in the same Gaṇapāṭha entry, are सर्वनाम. Not "words that mean 'all' or 'this' or 'that'" — words on the list. सर्वनाम is a संज्ञा, a technical name assigned to specific stems, exactly the way माहेश्वरसूत्र assigns अच् to nine specific vowels rather than to "vowel-like sounds in general."

And a संज्ञा is never decorative in this grammar — it is always a trigger for some other rule. Being सर्वनाम changes how a stem declines. Three sūtras spell out exactly how:

चतुर्थी-एकवचन gets -स्मै, not -आय

— 7.1.14. Compare the ordinary अकारान्त noun देव with the सर्वादि words सर्व, तद् and किम् in exactly the same cell of exactly the same paradigm:

विभक्तिः	देव (not सर्वनाम)	सर्व	तद्	rule

Read the last row carefully. The प्रथमा-बहुवचन row is not a simple ending swap — देवाः keeps the ordinary -अस् ending, but सर्वे replaces it with शी (7.1.17 ) before the usual vowel sandhi (, 6.1.87) turns अ + ई into ए. Three different mechanisms — a straight ending swap for चतुर्थी and पञ्चमी/सप्तमी, a whole-affix substitution for प्रथमा-बहुवचन — all triggered by the same one fact: the stem is on the सर्वादि list.

व्यवस्थायाम्Some words are सर्वनाम only in one sense

If सर्वनाम were tracking a meaning, this would be incoherent. Because it tracks a list instead, Pāṇini can put a word on the list conditionally — and does, for ten of the forty stems the corpus attests.

— 1.1.34: पूर्व, पर, अवर, दक्षिण, उत्तर, अपर and अधर are सर्वनाम only व्यवस्थायाम् — when used to place things in a series ("the former, the latter") — not when they simply mean "east" or "south" or "inferior." (1.1.35) does the same for स्व, restricted to "one's own" excluding kin or wealth; and (1.1.36) for अन्तर, restricted to "outside."

This is the whole argument for "class, not meaning," in one fact. पूर्व meaning "the earlier one in this list" is सर्वनाम — it takes पूर्वस्मै, not *पूर्वाय. The identical word पूर्व meaning "the east" is not — an ordinary अकारान्त noun, पूर्वाय. Same spelling, same root sense of "before/in front," different grammatical behaviour, because 1.1.34 reads the intended sense before it decides whether the sūtra 1.1.27 applies. The full list, with which of these ten are conditional, is in the reference table below.

युष्मद् and अस्मद् — one paradigm, several stems

Suppletion

"You" and "I" are also सर्वनाम, but they do not use the सर्वादि endings above at all — they are irregular in a completely different way. Their seven-case paradigm is not built from one stem with changing endings; it is assembled from several stems that happen to share a meaning. Linguists call this suppletion, and it is the single most common way the world's languages handle their pronouns — English I / me / my is the same phenomenon.

युष्मद् · अस्मद्The nominative does not predict the resta working habit, not a fear

Look at three cells of each paradigm — प्रथमा (the citation form), तृतीया-एकवचन, and षष्ठी-बहुवचन:

त्व-, त्वय्-, युष्मा(क)- are three different shapes for what is grammatically one stem; अ-/मय्-/अस्मा(क)- are three shapes for the other. A learner who expects तृतीया to look like प्रथमा with an ending stuck on will be confused exactly once by this pair, and never again once the shapes are known by sight — that is what "suppletive" is telling you to expect, not a warning that something has gone wrong.

No gender. युष्मद् and अस्मद् decline the same way whether they refer to a man, a woman, or a thing — there is no separate masculine/feminine/neuter paradigm the way there is for सर्व or तद्. The corpus stores exactly one paradigm for each.

The short forms that cannot lead

8.1.18–23

Nine cells of the युष्मद्/अस्मद् paradigm have a second, shorter form sitting beside the regular one — तुभ्यम्-ते, not just तुभ्यम्. These short forms are real Sanskrit, used constantly in both prose and verse — but they come with a placement rule that the long forms do not.

अनुदात्तम्Toneless, and therefore never first

— 8.1.18: [a finite verb, and certain other words,] all become अनुदात्त — toneless, unaccented — except at the start of a पाद. 8.1.20–8.1.23 name exactly which cells of युष्मद्/अस्मद् fall under this: (द्वितीया/चतुर्थी/षष्ठी-द्विवचन → वाम्/नौ), (…-बहुवचन → वः/नः), (चतुर्थी/षष्ठी-एकवचन → ते/मे), and (द्वितीया-एकवचन → त्वा/मा).

Being toneless is exactly what a word at the head of a clause cannot be in Vedic recitation — so in practice, the working rule a reader can hold onto is simple: these words are never the first word. They lean on whatever comes before them.

रूपम्	पूर्णरूपम्	किम्	विभक्ति-वचनम्	rule

All three example sentences below put the clitic in second position, which is where it belongs:

What breaks. is not Sanskrit anyone writes — opening on ते is exactly what 8.1.18 rules out. If a clitic needs to lead — because it is the only word before a pause, say — the long form is used instead: तुभ्यम्, not ते.

A related, non-enclitic restriction. इदम् and एतद् also carry a second oblique base, एन- (visible in the tables below as the second or third alternant: इमम्-एनम्, एतम्-एनम्, अनयोः-एनयोः…). (2.4.32) licenses it for इदम् specifically, in अन्वादेश — a renewed reference to someone already named in the sentence. एन- forms are never nominative and, like the enclitics above, never open a sentence: they exist only to point back.

इदम् · एतद् · तद् · अदस् — one system, four distances

Deixis

These four are usually taught as four unrelated irregular paradigms. They are not. Each has an oblique stem that takes the सर्वादि endings from §1 exactly like सर्व does — and each has a nominative that does not match its own oblique stem, exactly like the suppletion in §2. Once you see that shared skeleton, what is left to learn is only the four bases and the four distances they mark.

इदम्-एतद्-तद्-अदस्-किम्The shared skeletonsame endings, four bases

शब्दः	अर्थः	प्रथमा-एक	प्रथमा-बहु	चतुर्थी-एक	पञ्चमी-एक	सप्तमी-एक

Read down the last three columns: every base ends चतुर्थी-एकवचन in -स्मै (or, after उ, its retroflex twin -ष्मै), and सप्तमी-एकवचन in -स्मिन्/-ष्मिन् — the same 7.1.14/7.1.15 that gave सर्वस्मै. Now read the first two columns: not one of the five nominatives is built on the oblique base in the same row. That break is not an accident particular to one word — it is the shape every सर्वादि deictic takes.

किम् belongs to the same family. It is not a distance word — it is the interrogative member — but it takes the identical skeleton: कस्मै, कस्मात्, कस्मिन्, and a nominative (कः) that no more resembles क- than सः resembles त- or अयम् resembles अ-.

दूरत्वम्Four distances, not four synonymsnear → visible → known → yonder

इदम् — this, right here, at or near the speaker.
एतद् — this, pointed at or just now in view — closer to the act of pointing than to the speaker's body.
तद् — that: the ordinary anaphoric pronoun, picking up something already named in the discourse, without necessarily being far away in space at all.
अदस् — that, yonder: physically remote, often out of sight.

Try the dial below: one sentence, the pronoun swapped through all four, the English shifting with it.

The deictic dial

Try it

Same sentence, same case, same gender — only the pointing word changes. Watch the English change with it.

Full paradigms

Every stem this page discusses, all seven cases, all three genders where the corpus attests one — pulled from the same 9,007-paradigm corpus as every form above.

Every सर्वादि stem the corpus attests

Coverage

शब्दः	Gloss	Status

Where this comes from. Every शब्द entry the corpus holds carries an opts field naming the sūtras used to build it. This table is every entry whose opts include सर्वादीनि सर्वनामानि — the corpus reporting its own working, not a Gaṇapāṭha retyped from memory. Feminine-only headwords (सर्वा, अन्या, …) and two compound teaching-examples (परमसर्व, परमतद्) are folded out so the count below is the forty base stems themselves.

Check yourself

अभ्यासः

Six questions.