Every word derived from its Dhātupāṭha entry forward, with a rule against each change: markers deleted, augments placed, affixes elided, visarga formed. Any change to a form has a rule beside it.
The whole of the first khaṇḍa of the third muṇḍaka, in sequence, before anything is taken apart. Six of the ten are analysed word by word further down; the other four are here for context, because the argument runs continuously and the analysed verses are an excerpt from it, not a unit on their own.
Two fine-winged birds, yoked together, companions, cling to the one tree. One of them eats the sweet fruit; the other, not eating, looks on.
On that same tree the person, sunk down, grieves in his helplessness, bewildered. When he sees the other — the Lord, worshipped — and this, His greatness, his grief departs.
When the seer sees the golden-hued one — maker, Lord, Person, source of Brahmā — then, knowing, shaking off merit and demerit, unstained, he reaches the highest identity.
This is the breath that shines forth through all beings. Knowing it, the wise no longer speak in excess — at play in the Self, delighting in the Self, and active: he is the first among knowers of Brahman.
विभाति again, as in 3.1.7 — and again of the Self. The verb is becoming a refrain by this point in the khaṇḍa.
By truth is this Self won, and by austerity, by right knowledge, by continence constantly kept. Within the body he is, made of light, radiant — whom the ascetics see, their faults worn away.
Worth setting beside 3.1.8, which denies that तपसा reaches it. Here तपसा is among the means. The two are not necessarily in conflict — Śaṅkara reads 3.1.8 as denying austerity as the direct instrument of seeing — but a class should be shown the tension rather than have it smoothed over.
Truth alone prevails, not falsehood. By truth is the path of the gods laid open, by which the seers, their desires already fulfilled, ascend to where truth has its highest treasure.
सत्यमेव जयते is the national motto of India, taken from this verse. Note that जयते is again ātmanepada where the classical language would want जयति — the same ārṣa licence as पश्यते in 3.1.3 and 3.1.8, and a third instance within one khaṇḍa.
That is vast and self-luminous; its form is beyond thought. It shines subtler than the subtle. It is farther than the far, and it is here, close at hand — set down, for those with eyes to see, right here, in the cave.
It is not grasped by the eye, nor by speech, nor by the other faculties, nor by austerity or ritual action. When one's inner being has been clarified by the serenity of knowledge, then, in meditation, one sees that partless One.
This Self is subtle, and is to be known by the mind — the mind into which the breath has entered fivefold. The whole thinking of creatures is woven through by the breaths; when it is purified, this Self shines forth.
Whatever world one of purified being holds in mind, and whatever desires he desires — that world he wins, and those desires. Therefore let whoever seeks welfare honour the one who knows the Self.
विशुद्धसत्त्वः returns from 3.1.8, and संविभाति couples सम् to the same √भा. The khaṇḍa closes on the vocabulary it opened the argument with.
3.1.7 and 3.1.8 are given with their Vedic accents. The rest are unaccented — see the editor queries. Verses 3.1.4, 3.1.5, 3.1.6 and 3.1.10 were transcribed from a recitation edition that marks word junctures and nasals inline (समानं-वृँक्ष-म्परिषस्वजाते); they have been normalised to the orthography used throughout this module, and should be checked against the course's own edition before printing.
Every verbal word opens with its Dhātupāṭha entry — root as taught, gaṇa, pada, seṭ or aniṭ. The derivation begins there, not half way.
Each step highlights the part of the form that differs from the step above it. If a mark appears or vanishes, you can see exactly where — and the rule for it sits alongside.
Press / and type a word, a root, a gloss, or a sūtra number such as 6.1.87 to find every word that uses it.
Eleven rules do most of the work across these verses. Pick one below to see which words turn on it.
Some readings are disputed between editions or unsettled in the commentaries. Those carry a note on the card and are gathered under open questions.
Set unaccented — the accented text should come from the same edition used for 3.1.7–8. See editor queries.
Derivations for this verse cover the words that the later verses depend on. The remainder are staged: text, translation, anvaya and sandhi are complete, and the word cards will follow in the same format. 2 of 14 words derived so far.
Two fine-winged birds, yoked together, companions, cling to the one tree. One of them eats the sweet fruit; the other, not eating, looks on.
सयुजा सखाया द्वा सुपर्णा समानं वृक्षं परिषस्वजाते । तयोः अन्यः स्वादु पिप्पलम् अत्ति; अन्यः अनश्नन् अभिचाकशीति ॥
Ṛgveda 1.164.20 verbatim, and the frame for everything that follows. The second bird — the one that does not eat — is the subject of 3.1.7–9. Read the later verses without this one and तत् has no antecedent.
| तयोः + अन्यः | → | तयोरन्यः | ससजुषो रुःP.8.2.66visarga → र् before a vowel |
| स्वादु + अत्ति | → | स्वाद्वत्ति | इको यणचिP.6.1.77उ → व् before a dissimilar vowel |
| अनश्नन् + अन्यः | → | अनश्नन्नन्यः | ङमो ह्रस्वादचि ङमुण्नित्यम्P.8.3.32a ङम् (ङ ण न) after a short vowel takes the augment ङमुट् before a vowel — hence the doubled न् |
| अन्यः + अभिचाकशीति | → | अन्यो अभिचाकशीति | अतो रोरप्लुतादप्लुतेआद्गुणःP.6.1.113P.6.1.87Vedic: the following अ is preserved rather than elided — contrast एषोऽणुः in 3.1.9 |
In classical Sanskrit ओ + अ gives ओऽ with avagraha, as at एषोऽणुः in 3.1.9. Here the अ of अभिचाकशीति stands open. Vedic texts frequently decline the elision, and printed editions differ; this is a place to show students that sandhi is a convention of transmission as much as a rule of the language.
सु compounds with what follows; a bahuvrīhi, “having good wings”.
The compound denotes the bearer of the wings, not the wings. → सुपर्ण.
The nominative dual affix औ is introduced.
अ + औ → औ. The classical form.
In the Veda the case affixes are freely substituted, lengthened or elided. The dual in आ for औ is one of the licensed Vedic variants.
The most quoted image in the Upaniṣads, and older than this text: it is Ṛgveda 1.164.20 verbatim, and recurs at Śvetāśvatara 4.6. Whatever 3.1.7–9 say about the Self is said about the second bird.
सुपर्णा, सयुजा, सखाया — three duals in आ where classical Sanskrit wants औ. P.7.1.39 licenses the whole line at once. Students meeting Vedic for the first time should be given this rule before the verse, not after.
Taught with an initial ष् — which by convention becomes स्, and then becomes ष् again after the prefix परि. A good demonstration that the ष् of the Dhātupāṭha is notation, not sound.
The root in upadeśa.
An initial ष् of a root is replaced by स्, and the nasalised अँ is an इत् and drops.
A लिट् ending that is not पित् is treated as कित्. आते is not पित्, so it is कित् — which is what gives the next rule its trigger.
A root not marked with इ loses its penultimate nasal before a कित् or ङित् affix. → स्वज्.
लिट् for remote past.
Third person dual ātmanepada in लिट् is आते.
Reduplication before लिट्.
Only the first consonant of the reduplicate is kept: स्वज् → स.
The reduplicated perfect.
The upasarga परि is prefixed.
P.8.3.57 is the अधिकार supplying the condition — a preceding इण् or कु sound, here the इ of परि. The operative rule is P.8.3.59: this स् is itself an आदेश, put for the root's original ष् by P.6.1.64. So the ष् is restored, by a quite different route from the one that removed it.
Same operative rule as चक्षुषा — P.8.3.59 आदेशप्रत्यययोः — but reached differently: there the स् was part of an affix, here it is an आदेश put for the root's own ष् by P.6.1.64. Compare निष्कल in 3.1.8, which uses a different rule again (P.8.3.41, visarga after इ/उ). One sound change, three routes to it.
Set unaccented — the accented text should come from the same edition used for 3.1.7–8. See editor queries.
Derivations for this verse cover the words that the later verses depend on. The remainder are staged: text, translation, anvaya and sandhi are complete, and the word cards will follow in the same format. 2 of 13 words derived so far.
On that same tree the person, sunk down, grieves in his helplessness, bewildered. When he sees the other — the Lord, worshipped — and this, His greatness, his grief departs.
समाने वृक्षे निमग्नः मुह्यमानः पुरुषः अनीशया शोचति । यदा जुष्टम् अन्यम् ईशम् अस्य महिमानम् इति च पश्यति, (तदा) वीतशोकः (भवति) ॥
The turn from the first bird to the second, and the model for 3.1.8: a condition of failure, then a seeing that ends it. Note that seeing is again the operative verb, and again it is the other that is seen.
| पुरुषः + निमग्नः | → | पुरुषो निमग्नो | हशि चआद्गुणःP.6.1.114P.6.1.87visarga → ओ before a voiced consonant |
| निमग्नः + अनीशया | → | निमग्नोऽनीशया | अतो रोरप्लुतादप्लुतेआद्गुणःएङः पदान्तादतिP.6.1.113P.6.1.87P.6.1.109the full chain: रु → उ, अ+उ → ओ, then the following अ absorbed and marked by avagraha ऽ |
| पश्यति + अन्यम् | → | पश्यत्यन्यम् | इको यणचिP.6.1.77इ → य् |
| अन्यम् + ईशम् | → | अन्यमीशम् | म् retained before a vowel |
निमग्नोऽनीशया here and एषोऽणुः in 3.1.9 are the same sandhi by the same rule — and both stand against अन्यो अभिचाकशीति in 3.1.1, where the elision does not happen. Three instances, two outcomes, one khaṇḍa.
The same root gives ईशम् in the next line — the helplessness and the Lord are named from one word.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
A feminine action-noun in आ, “lordship”.
Negative tatpuruṣa.
The न् of नञ् is elided… but see the next step.
…and when the following member begins with a vowel, the augment नुट् restores an न्. This is why we have अन्- and not merely अ-. → अनीशा.
Instrumental singular of an आ-stem: टा with the आ stem gives -या.
Editions split between अनीशया and अनिशया. The long ī is the reading that makes grammatical sense — नञ् + ईशा, “lordlessness”, which is exactly what Śaṅkara glosses (अनैश्वर्येण) and what the next line answers with ईशम्. अनिशया would have to be from अनिश, “ceaseless”, which fits neither the syntax nor the commentary. Printed here as अनीशया, but the course's edition should be checked and the choice recorded.
इमनिच् forms abstract nouns from पृथु and the rest.
In इमनिँच् the nasalised इँ and the final च् are both इत् and drop, leaving इमन्.
The टि of the stem — the final vowel with what follows — is elided before the affix: महत् → मह्. → महिमन्.
The accusative singular affix is introduced.
अम् in the masculine is a सर्वनामस्थान, so the penultimate is lengthened — exactly as in आत्मानम्.
The म् of अम् survives.
Both are न्-stems, and both lengthen before a सर्वनामस्थान affix (P.6.4.8). The difference is only that आत्मा additionally loses its न् (P.8.2.7), because there the affix is gone and the stem stands as a pada. Two words, one rule, opposite surfaces.
The sūtra is P.6.4.143. It is easily confused with P.6.4.146 ओर्गुणः, which is a different rule entirely.
Set unaccented — the accented text should come from the same edition used for 3.1.7–8. See editor queries.
Derivations for this verse cover the words that the later verses depend on. The remainder are staged: text, translation, anvaya and sandhi are complete, and the word cards will follow in the same format. 2 of 14 words derived so far.
When the seer sees the golden-hued one — maker, Lord, Person, source of Brahmā — then, knowing, shaking off merit and demerit, unstained, he reaches the highest identity.
यदा पश्यः रुक्मवर्णं कर्तारम् ईशं पुरुषं ब्रह्मयोनिं पश्यते, तदा विद्वान् पुण्यपापे विधूय निरञ्जनः परमं साम्यम् उपैति ॥
The verse that supplies the corroboration for पश्यते at 3.1.8. It also supplies the vocabulary 3.1.7–9 will strip away: here the Self is named with five epithets, there with none — only तत्.
| कर्तारम् + ईशम् | → | कर्तारमीशम् | म् retained before a vowel |
| विद्वान् + पुण्यपापे | → | विद्वान्पुण्यपापे | no change |
| साम्यम् + उपैति | → | साम्यमुपैति | म् retained |
| उप + एति | → | उपैति | एत्येधत्यूठ्सुP.6.1.89the rule specific to एति, एधति and ऊठ्: अ + ए → vṛddhi ऐ, not the guṇa that P.6.1.87 would give |
In उपैति the vṛddhi happens inside the word, between preverb and root, by the same rule that joins इह and एव in 3.1.7. Sandhi does not care whether a boundary is between words or within one.
Same root as पश्यत्सु and पश्यते. Three formations from one root across the khaṇḍa — participle, agent noun, finite verb.
Markers इँ and र् drop.
दृश् is replaced by पश्य् before a शित् affix.
An agent noun in अ, “one who sees”.
पश्य is a प्रातिपदिक, so the twenty-one सुप् affixes become available; the nominative singular is सुँ.
The उँ of सुँ is a nasalised vowel in upadeśa, hence an इत्, and is elided. Only स् remains.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
The figura etymologica the whole khaṇḍa turns on: the seer sees. And this is the first occurrence of the ātmanepada पश्यते, which recurs at 3.1.8 — the corroboration cited on that card. Two instances in one khaṇḍa is what makes it a habit of the text rather than a scribal slip.
The two members.
The न् of ब्रह्मन् drops in composition. → ब्रह्मयोनि.
The accusative singular affix is introduced.
The म् survives.
As a tatpuruṣa (ब्रह्मणो योनिः) it means “the source of Brahmā” — the translation most editions give. As a bahuvrīhi (ब्रह्म योनिर्यस्य) it means “he whose womb is Brahman”. The grammar permits both and the accent would decide it — which is why an accented edition matters. If the course has one for this verse, it settles a question the translations argue about.
That is vast and self-luminous; its form is beyond thought. It shines subtler than the subtle. It is farther than the far, and it is here, close at hand — set down, for those with eyes to see, right here, in the cave.
तत् बृहत् दिव्यम् अचिन्त्यरूपं च । तत् सूक्ष्मात् सूक्ष्मतरं विभाति । (तत्) दूरात् सुदूरे, तत् इह अन्तिके च । पश्यत्सु इह एव गुहायां निहितम् (अस्ति) ॥
Six predications about a subject never named — only pointed at with तत्. Each pair cancels the last: vast but subtler than the subtle, far but near. The final line resolves the paradox by locating it: not out there, but guhāyām.
| बृहत् + च | → | बृहच्च | स्तोः श्चुना श्चुःP.8.4.40dental त् → palatal च् before च् |
| तत् + दिव्यम् | → | तद्दिव्यम् | झलां जश् झशिP.8.4.53त् → द् because a झश् (द्) follows |
| सूक्ष्मात् + च | → | सूक्ष्माच्च | स्तोः श्चुना श्चुःP.8.4.40same change as line 1 |
| दूरात् + सुदूरे | → | दूरात्सुदूरे | no change: त् before स् is stable |
| तत् + इह | → | तदिह | झलां जशोऽन्तेP.8.2.39pada-final झल् → जश्. Note this is NOT 8.4.53: no झश् follows, only a vowel |
| इह + अन्तिके | → | इहान्तिके | अकः सवर्णे दीर्घःP.6.1.101अ + अ → आ |
| पश्यत्सु + इह | → | पश्यत्स्विह | इको यणचिP.6.1.77उ → व् before a dissimilar vowel |
| इह + एव | → | इहैव | वृद्धिरेचिP.6.1.88अ + ए → ऐ |
| निहितम् + गुहायाम् | → | निहितं गुहायाम् | मोऽनुस्वारःP.8.3.23pada-final म् → anusvāra before a consonant |
Lines 7–8 are the ones to dwell on. Read quickly, the stretch looks like पश्यत् / स्विहैव — but स्वि- is not a word. Undo the two sandhis in order and the padas fall out cleanly as पश्यत्सु + इह + एव: a locative plural, not a nominative singular.
The root is cited in the Dhātupāṭha with a nasalised इँ after it. That इँ is not part of the root; it is a marker, and it is the reason this word acquires — and then loses — a nasal.
The root as taught in the Dhātupāṭha (उपदेश), with its marker.
The इँ is a nasalised vowel in upadeśa, therefore an इत्, and is elided. The root proper is बृह्.
Because the root was इदित् — marked with इ — it takes the augment नुँम्.
In नुँम्, the उँ is an इत् by P.1.3.2 and the final म् is an इत् by हलन्त्यम्. Both drop. The augment is a bare न्.
An augment marked with म् (a मित्) is placed immediately after the last vowel of what it is added to. The last vowel of बृह् is ऋ, so न् goes between ऋ and ह्.
A न् or म् not standing at the end of a pada becomes anusvāra before a झल् sound. ह् is in झल्, so न् → ं. This is where the dot appears.
The Uṇādi sūtra supplies the affix अति in present sense — and, crucially, it does so by naming the finished words themselves.
This is where the dot disappears. No productive rule removes it. The Uṇādi sūtra is a निपातन: it states the forms पृषत्, बृहत्, महत्, जगत् ready-made, and whatever irregularity they contain is licensed by the mere fact of their being listed. See the note below — this step deserves scrutiny, not acceptance.
The stem is now a प्रातिपदिक and takes a case affix.
After a neuter stem, सु and अम् undergo लुक् — elision leaving no trace. The stem stands alone as the finished word.
Steps 6 and 8 are the ones to dwell on. The nasal arrives by ordinary rules — the नुम् augment, placed after the last vowel, then converted to anusvāra before ह्. It leaves by no productive rule: the Uṇādi sūtra names बृहत् outright, and a निपातन licenses its own irregularities. Traditional grammar's answer to “which sūtra?” here is निपातनात् सिद्धम्, and a student is entitled to find that unsatisfying.
There is an alternative. Several editions of the Dhātupāṭha carry a second root बृहँ alongside बृहिँ — not इदित्, so never taking नुम् at all. Derive बृहत् from that and no nasal ever appears, so none has to be explained away. Which root is intended should be settled before class.
हल्ङ्याब्भ्यो दीर्घात् सुतिस्यपृक्तं हल् (P.6.1.68) governs masculines and feminines after a consonant — मरुत्, for instance. बृहत् here agrees with neuter तत्, so the neuter-specific स्वमोर्नपुंसकात् (P.7.1.23) applies. Both routes end at बृहत्, so the choice is invisible in the result and visible only in the reasoning. Compare आत्मा in 3.1.9, where P.6.1.68 is the rule that applies.
Śaṅkara takes बृहत् as निरतिशयमहत् — great in a way admitting no comparison. The root sense “to grow” is the same one behind ब्रह्मन्, which the text is unlikely to be unaware of.
च and the words listed with it are termed निपात when they do not denote a substance.
The स्वरादि words and the निपातs are termed अव्यय, indeclinable.
Even an indeclinable is a प्रातिपदिक and formally receives a case affix.
After an अव्यय, आप् and the सुप् affixes undergo लुक्. The affix is introduced and then removed without trace — which is what “indeclinable” means procedurally.
अव्ययादाप्सुपः (P.2.4.82), which also silences the endings on न, अपि, वा, तु and एव. Note the shape of the derivation: the affix is added, then it is deleted. Pāṇini does not simply decline to inflect indeclinables.
निपात एकाजनाङ् (P.1.1.14) makes a single-vowel निपात प्रगृह्य — exempt from vowel sandhi. It does not bite here, since च is followed by consonants both times, but it is why आ stands unmerged elsewhere.
The Amarakoṣa derivation (च ← √चद् आह्लादे + ड) is a traditional nirukti. Historically च continues PIE *kʷe — Latin -que, Greek τε.
तद् is one of the सर्वादि list and is therefore termed सर्वनाम.
The nominative singular affix is introduced.
After a neuter stem, सु undergoes लुक्. Nothing of the affix remains.
Ordinarily प्रत्ययलक्षणम् (P.1.1.62) lets an elided affix still trigger rules. But when the elision is a लु-elision — लुक्, श्लु, लुप् — that is withheld for operations on the अङ्ग. So त्यदादीनामः cannot fire, and the stem is untouched.
At the end of an utterance, a झल् optionally becomes चर् — the voiced द् may surface unvoiced as त्. Both readings are correct.
In the masculine, तद् + सु does change: त्यदादीनामः (P.7.2.102) puts अ for the final and तदोः सः सावनन्त्ययोः (P.7.2.106) puts स् for त्, giving सः. Neither fires in the neuter, and step 4 is the whole reason. If P.1.1.63 did not exist, the neuter would come out as *त. Students should be able to say why it does not.
तद् stands voiced before दिव्यम् and unvoiced before सूक्ष्मतरम्. One word, two surfaces — the first thing to be able to see through.
A secondary (taddhita) derivative, not a verbal one: the base is a noun, not a root.
The base: दिव् f., heaven.
A taddhita affix य in the sense “belonging to, arising from” heaven. The licensing sūtra is flagged for the editor rather than guessed at.
The त् of यत् is an इत् by हलन्त्यम् and is elided, leaving य. → दिव्य.
The nominative singular affix is introduced.
After a neuter stem ending in अ, अम् replaces सु and अम्. सु is gone; अम् stands in its place.
The म् of अम् would be an इत् by हलन्त्यम्, but this rule exempts त थ द ध न स म at the end of a विभक्ति. The म् therefore survives into the finished word.
अ + अम्: the prior vowel alone is the substitute for both. The stem's final अ and the अ of अम् merge into one.
दिव् + य is uncontroversial as a formation, but the exact sūtra should be filled in from the भवार्थ taddhitas (P.4.3.53 ff.). Left open deliberately: an invented rule number is worse than a visible gap.
Śaṅkara glosses दिव्य as स्वयंज्योतिः. The published translations follow him. A good place to show that a translation is already a commentary.
A curādi root: it carries णिच् obligatorily, so the stem is चिन्ति-, and the णि has to be got rid of again before ण्यत्.
The root, as a curādi verb.
Curādi roots take णिच् obligatorily.
In णिच्, the initial ण् is an इत् (चुटू) and the final च् is an इत् (हलन्त्यम्). Both drop; the affix is a bare इ.
ण्यत् follows a root ending in ऋ or in a consonant, in the sense “fit to be —ed”.
Same two इत्s again: the ण् and the त् of ण्यत् drop, leaving य. The ण् is not idle, though — it marks the affix as णित्, which is why it would cause vṛddhi elsewhere.
The णि just added is elided before an ārdhadhātuka affix that takes no इट्. ण्यत् is aniṭ, so णि goes. → चिन्त्य.
Negative tatpuruṣa compound.
The ञ् of नञ् is an इत् and drops.
The न् of नञ् is elided when it stands as the prior member of a compound. What is left of the negation is a bare अ.
Two or more words in the same case combine as a bahuvrīhi, denoting something other than either of them.
The nominative singular affix is introduced.
After a neuter stem ending in अ, अम् replaces सु and अम्. सु is gone; अम् stands in its place.
The म् of अम् would be an इत् by हलन्त्यम्, but this rule exempts त थ द ध न स म at the end of a विभक्ति. The म् therefore survives into the finished word.
अ + अम्: the prior vowel alone is the substitute for both. The stem's final अ and the अ of अम् merge into one.
The text is accented अचि॑न्त्य-, with the udātta on the prior member — exactly what बहुव्रीहौ प्रकृत्या पूर्वपदम् (P.6.2.1) predicts, since in a bahuvrīhi the first member keeps its own accent. The accent marks are evidence, not decoration: here they settle a reading the syntax alone leaves open. The same test recurs at विशुद्धसत्त्वः in 3.1.8.
रूप is neuter and so is the compound — but only by accident. A bahuvrīhi takes the gender of what it qualifies, here तत्. With a masculine referent it would be अचिन्त्यरूपः, रूप itself unchanged.
The प्रातिपदिक, “minute, fine, subtle”.
The ablative singular affix ङसिँ is introduced.
Two इत्s go: the initial ङ् and the final nasalised इँ. What is left is अस्.
After a stem in अ, the affixes टा, ङसि and ङस् are replaced by इन, आत् and स्य. Here अस् is replaced wholesale by आत्.
अ + आ, two सवर्ण vowels: the corresponding long आ replaces both.
पञ्चमी विभक्ते (P.2.3.42): the fifth case where one thing is marked off from another. The standard ablative of comparison that accompanies तरप्, and it returns at दूरात् in the next line.
तरप् and ईयसुन् follow when a comparison between two is expressed.
The final प् of तरप् is an इत् and is elided. It is not idle either: a पित् affix is anudātta, which is why the accent sits where it does. → सूक्ष्मतर.
The nominative singular affix is introduced.
After a neuter stem ending in अ, अम् replaces सु and अम्. सु is gone; अम् stands in its place.
The म् of अम् would be an इत् by हलन्त्यम्, but this rule exempts त थ द ध न स म at the end of a विभक्ति. The म् therefore survives into the finished word.
अ + अम्: the prior vowel alone is the substitute for both. The stem's final अ and the अ of अम् merge into one.
सूक्ष्मात् सूक्ष्मतरम् belongs to a family of Upaniṣadic superlatives-by-recursion: अणोरणीयान् महतो महीयान् (Kaṭha 1.2.20, Śvetāśvatara 3.20). The construction says what the vocabulary cannot — that the term is being pushed past its own scale. 3.1.9 picks the thread up again with अणुः.
A second-class root. The defining fact about अदादि is that the class sign शप् is elided, so the endings attach straight to the root.
The root as taught: भा दीप्तौ.
लट् follows a root when present action is denoted.
What is introduced is a प्रत्यय, and it stands after the root.
ल् is replaced by one of eighteen personal endings; third person singular parasmaipada is तिप्. Its प् is an इत् by हलन्त्यम् and drops, leaving ति.
शप् is introduced between root and ending when the affix denotes the agent.
In शप्, the श् is an इत् by लशक्वतद्धिते and the प् by हलन्त्यम्. Both drop; the class sign is a bare अ.
After अद् and the rest — the second class — शप् undergoes लुक्. The अ is removed entirely, and the ending now meets the root directly. This is why the form is भाति and not *भायति.
No further change: आ + ति stand as they are.
वि is termed उपसर्ग when joined to a verb, and is prefixed.
Śaṅkara reads विविधं भाति — it shines in manifold ways. That is where “shines diversely” comes from; without the gloss वि- looks like an empty preverb.
3.1.9 ends with विभवति, from √भू, not √भा. Two roots, two classes, adjacent verses, one syllable apart, same preverb. Compare the two derivations side by side — the divergence is exactly at the शप् step.
The प्रातिपदिक, “distant”.
The ablative singular affix ङसिँ is introduced.
Two इत्s go: the initial ङ् and the final nasalised इँ. What is left is अस्.
After a stem in अ, the affixes टा, ङसि and ङस् are replaced by इन, आत् and स्य. Here अस् is replaced wholesale by आत्.
अ + आ, two सवर्ण vowels: the corresponding long आ replaces both.
Words meaning “far” and “near” govern unusually freely — see दूरान्तिकार्थैः षष्ठ्यन्यतरस्याम् (P.2.3.34) and दूरान्तिकार्थेभ्यो द्वितीया च (P.2.3.35). Here the ablative is comparative (P.2.3.42), matching सूक्ष्मात्.
कु, the गति words and the प्रादि words compound with what follows. सु is one of them.
The compound stem is now a प्रातिपदिक.
The locative singular affix ङि is introduced.
The initial ङ् of ङि is an इत् — ल्, श् and the velars कु are इत् when initial in a non-taddhita affix — and is elided. The affix is इ.
अ followed by इ: guṇa ए is the single substitute for both.
Ablative of comparison against locative of position: “farther off than what is far”. The figure runs through the line — दूरात् सुदूरे against इह अन्तिके — and the second half deliberately refuses the comparative, because nearness is not a matter of degree.
Start from the pronoun इदम् in the locative.
ह replaces the locative affix after इदम्.
इदम् itself is replaced by इश् before such affixes; the श् is an इत् by लशक्वतद्धिते and drops, leaving इ. → इह.
A taddhita form that does not admit the full range of case affixes is termed अव्यय — hence no further inflection.
Śaṅkara reads इह as अस्मिन् शरीरे — in this body. Both occurrences do work: the first answers दूरात्, the second, with एव, insists the location is not figurative.
A taddhita derivative of अन्त, in the sense of proximity.
The final न् of ठन् is इत् by हलन्त्यम् and drops; the ठ् is the affix itself, not a marker. But it is not idle: a नित् affix makes the first syllable udātta, which is what the accent अ॑न्तिके records.
The ठ of a taddhita affix is replaced by इक. → अन्तिक.
The locative singular affix ङि is introduced.
The initial ङ् of ङि is an इत् — ल्, श् and the velars कु are इत् when initial in a non-taddhita affix — and is elided. The affix is इ.
अ followed by इ: guṇa ए is the single substitute for both.
The root is replaced wholesale by पश्य् before शित् affixes. This is a suppletion, not a phonetic change — दृश् and पश्य् share nothing.
The root in upadeśa, with two markers.
The nasalised इँ and the final र् are both इत् and drop. The र् is not idle: it marks the root as ऋदित्, relevant elsewhere. → दृश्.
लट् for present sense.
शतृ and शानच् replace लट् to form the present participle.
In शतृँ the initial श् is an इत् (a शित् affix); the final nasalised ऋँ is an इत् too. What remains is अत्. The श् is not idle — it is what makes the affix शित्, which triggers the next two steps.
After पा, घ्रा, ध्मा, धेट् and दृश्, श replaces शप् as the class sign.
The श् of श is an इत् and drops, leaving अ.
Before a शित् affix these roots are replaced outright: दृश् gives पश्य्. Nothing of दृश् survives.
अ followed by a guṇa vowel: the latter alone is the substitute (पररूप). The अ of श and the अ of अत् merge.
The locative plural affix is introduced.
The प् of सुप् is an इत् and drops.
No नुम् is inserted: see the note. The finished form is पश्यत्सु.
It is tempting to read पश्य॑त् and स्विहैव as two words, but स्वि- is not a word. Undoing the sandhi gives पश्यत्सु + इह + एव: the व् is the उ of the locative plural सु turned semivowel by इको यणचि (P.6.1.77). The split decides the grammar of the whole line — locative plural, not nominative singular.
The nominative plural is पश्यन्तः, with नुम् inserted by उगिदचां सर्वनामस्थानेऽधातोः (P.7.1.70). सुप् is not a सर्वनामस्थान, so no नुम् here — पश्यत्सु, not *पश्यन्त्सु. Note that this is the same नुम् that बृहत् acquired and lost: worth putting the two cards side by side.
अधिकरण, “among, in the case of”. Śaṅkara: प्राणिषु. The participle does double duty — those in whom it is set are those capable of seeing it.
निपात, not denoting a substance.
Therefore अव्यय.
A case affix is formally introduced.
…and immediately removed by लुक्.
एव restricts rather than intensifies: not “here too” but “here and nowhere else”. The whole rhetorical movement of the verse — far, near, here — lands on this particle.
Three markers on one root — डु, and ञ् — and all three drop immediately. The ञ् is what makes it उभयपद.
The root as taught, with its markers.
Initial ञि, टु, डु of a root are इत्, and a final consonant is इत्. So डु and ञ् both go. The root proper is धा.
क्त and क्तवतु follow a root in the sense of past action; they are termed निष्ठा.
The क् of क्त is an इत् and drops, leaving त. The क् is not idle: it makes the affix कित्, which blocks guṇa.
धा is replaced by हि before क्त and the like. → हित.
The upasarga नि, “down”, is prefixed.
The nominative singular affix is introduced.
After a neuter stem ending in अ, अम् replaces सु and अम्. सु is gone; अम् stands in its place.
The म् of अम् would be an इत् by हलन्त्यम्, but this rule exempts त थ द ध न स म at the end of a विभक्ति. The म् therefore survives into the finished word.
अ + अम्: the prior vowel alone is the substitute for both. The stem's final अ and the अ of अम् merge into one.
निहितं गुहायाम् is fixed Upaniṣadic diction — compare गुहायां निहितोऽस्य जन्तोः (Kaṭha 1.2.20) and Taittirīya 2.1. Students should meet it as a unit.
The name of the place is built from the act of concealing — which the commentary makes the point of the verse.
The root in upadeśa.
The nasalised ऊँ is an इत् and drops.
A कृत् affix अ forms the noun stem गुह.
A stem in अ takes टाप् in the feminine.
In टाप्, the ट् and the प् are both इत् and drop, leaving आ. The stem's final अ and this आ merge by अकः सवर्णे दीर्घः (P.6.1.101). → गुहा.
The locative singular affix is introduced.
A ङित् affix takes the augment याट् after an आप् stem.
The ट् of याट् is an इत् and drops; by आद्यन्तौ टकितौ (P.1.1.46) a ट्-marked augment stands at the beginning of what it is added to, so या precedes the affix.
ङि is replaced by आम् after a नदी, आप् or नी stem.
आ + या + म् stand as they are. The finished form.
Śaṅkara: बुद्धिगुहायाम् — the cave of the intellect, glossed elsewhere as हृदय. The root sense “that which conceals” carries the argument: the Self is hidden not by distance but by the very faculty that would look for it. 3.1.9 then makes that faculty the instrument of knowing, once purified.
It is not grasped by the eye, nor by speech, nor by the other faculties, nor by austerity or ritual action. When one's inner being has been clarified by the serenity of knowledge, then, in meditation, one sees that partless One.
(तत्) चक्षुषा न गृह्यते, वाचा अपि न, अन्यैः देवैः तपसा कर्मणा वा न । ज्ञानप्रसादेन विशुद्धसत्त्वः ततः तु ध्यायमानः तं निष्कलं पश्यते ॥
Structurally a single sentence: four instrumentals denied, then one instrumental granted. Everything before ज्ञानप्रसादेन is what fails; everything after is what succeeds. Note that the verse does not replace the faculties — it purifies them.
| न + अपि | → | नापि | अकः सवर्णे दीर्घःP.6.1.101अ + अ → आ |
| न + अन्यैः | → | नान्यैः | अकः सवर्णे दीर्घःP.6.1.101same |
| अन्यैः + देवैः | → | अन्यैर्देवैः | ससजुषो रुःP.8.2.66final स् → रु; र् surfaces before the voiced द् |
| देवैः + तपसा | → | देवैस्तपसा | विसर्जनीयस्य सःP.8.3.34visarga → स् before a dental खर् |
| विशुद्धसत्त्वः + ततः | → | विशुद्धसत्त्वस्ततः | विसर्जनीयस्य सःP.8.3.34same |
| ततः + तु | → | ततस्तु | विसर्जनीयस्य सःP.8.3.34same |
| निष्कलम् + ध्यायमानः | → | निष्कलं ध्यायमानः | मोऽनुस्वारःP.8.3.23म् → anusvāra |
Three identical applications of विसर्जनीयस्य सः in four words make this verse the natural place to teach visarga sandhi before dentals. Set it against प्राणैश्चित्तम् in 3.1.9, where the same rule feeds स्तोः श्चुना श्चुः and produces श् instead.
निपात, not denoting a substance.
Therefore अव्यय.
A case affix is formally introduced.
…and removed by लुक्.
AK 446/11 lists four words for अभाव: नहि, नञ् (→ अ / अन् in compounds), नो, न. The koṣa derives all four from √णह् बन्धने — negation as “binding off”. That is the koṣa's own nirukti; keep it, but label it.
Historically न continues PIE *ne — Latin nē, German nein, English no. Showing the traditional and comparative accounts together is usually more interesting than either alone.
न … नापि … नान्यैः. Not redundancy: each न governs a different instrumental, building the list the second half of the verse overturns.
The ङ् marks it ātmanepada. The organ is named from the act — the eye as “that which makes distinct”.
The root in upadeśa, with two markers.
The nasalised इँ and the final ङ् are both इत् and drop. → चक्ष्.
The affix उसिँ follows चक्ष्, and is treated as though it had श् as an इत्.
The final nasalised इँ of उसिँ is an इत् and drops, leaving उस्. → चक्षुस्.
The instrumental singular affix टा is introduced.
The initial ट् of टा is an इत् (चु and टु initial in an affix), and is elided. The affix is simply आ.
The स् of a substitute or an affix becomes ष् when it stands after a sound in इण् or कु. Here the स् belongs to the affix उसिँ and follows उ, which is in इण्.
ष् + आ stand as they are. The finished form.
चक्षुषा has ष्; तपसा and चेतसा (3.1.9) have स्. The difference is the vowel in front. इण् = इ उ ऋ ऌ ए ओ ऐ औ ह य व र ल — अ is not in it. चक्षुस् has उ before the स्; तपस् and चेतस् have अ. One rule, four words.
Editions number the Uṇādi sūtras differently, and this one appears as both उ. 2.119 and उ. 3.119. The number here follows the former; it should be checked against the edition the course uses. The Uṇādi sūtras are also the one body of rules cited in this module that could not be machine-verified.
A ninth-class root. In the passive the class sign is displaced by यक् anyway, so what distinguishes this derivation is not the gaṇa but the सम्प्रसारण.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
लट् for present sense.
When the affix denotes the action itself or its object, an ātmanepada ending substitutes for ल्. Third person singular ātmanepada is त.
यक् follows the root before a sārvadhātuka affix in the bhāva or karman sense. This is the passive marker, and it displaces the class sign.
The क् of यक् is इत् as a final consonant and drops, leaving य. The क् is not idle: it makes the affix कित्, which is exactly what licenses the next step.
सम्प्रसारण: in ग्रह् and the roots listed with it, the semivowel is replaced by its corresponding vowel before a कित् or ङित् affix. The र् becomes ऋ.
When a सम्प्रसारण vowel is followed by a vowel, the prior alone remains — here the अ of ग्रह् is absorbed, giving the single syllable गृ.
The ending त is टित्, so its टि portion — the final vowel with what follows — is replaced by ए. त → ते.
The finished passive.
The class 1 root √गृहू ग्रहणे would also reach गृह्यते, since यक् displaces the class sign either way — but its stem already contains ऋ, so that route never shows the सम्प्रसारण at steps 7–8. Built here from √ग्रहँ उपादाने, the root the passive belongs to.
Watch this rule across the set: it applies here, is explicitly blocked at वाचा below, and applies again at ओतम् in 3.1.9. Three verses, three outcomes, one rule.
The passive is what licenses चक्षुषा, वाचा, देवैः, तपसा, कर्मणा: in a कर्मणि construction the agent stands in the third case. The grammar of the whole first half of the verse hangs off the single affix यक्.
अपि is one of the twenty प्रादि words.
अपि is termed कर्मप्रवचनीय in the senses: “understood”, possibility, permission, censure, aggregation. Here it is the last — aggregation, “nor even”.
अव्यय.
The case affix is introduced and elided.
The Amarakoṣa route (अपि ← नञ् + √पि गतौ + क्विप्, “that which does not go”) is a traditional gloss. Historically अपि continues PIE *h₁epi, the same word as Greek ἐπί. The remark sometimes added, that आगम तुक् is not applied “because the rule is अनित्य”, is the sort of patch a nirukti needs and an etymology does not — a good illustration of how the two differ.
न + अपि = “nor even”. The particle concedes that speech might have seemed a better candidate than the eye, and denies it anyway.
This root normally undergoes सम्प्रसारण (उक्त, उच्यते). The Uṇādi rule below exists to stop it doing so here.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
क्विप् follows वच् and the roots listed with it; the rule prescribes lengthening of the root vowel and, expressly, no सम्प्रसारण.
In क्विँप्, the क् and the nasalised इँ and the final प् are all इत्. What remains is व्, a single consonant, and that too is elided as अपृक्त. The affix vanishes entirely — but its क् has already marked the operation as कित्.
The lengthening prescribed by the same rule now applies: अ → आ. → वाच्. Had the ordinary rule operated, this would have been *उच्.
The instrumental singular affix टा is introduced.
The initial ट् of टा is an इत् (चु and टु initial in an affix), and is elided. The affix is simply आ.
च् + आ stand as they are.
√वच् normally does undergo सम्प्रसारण: उक्त, उच्यते, by वचिस्वपियजादीनां किति (P.6.1.15). The Uṇādi rule overrides it specifically for this formation, which is why we have वाच् and not *उच्. Put this card beside गृह्यते and the rule stops being a list of exceptions.
अन्य is one of the सर्वादि, hence सर्वनाम.
The instrumental plural affix भिस् is introduced.
The final स् of भिस् is protected from इत्-hood by this rule, as it ends a विभक्ति.
After a stem in अ, भिस् is replaced by ऐस्.
अ + ऐ: vṛddhi ऐ is the single substitute for both.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
The instrumental plural of a सर्वनाम is formed exactly like a noun's — compare देवैः on the next card. The pronominal declension shows itself only elsewhere (dat. अन्यस्मै, loc. अन्यस्मिन्, as at यस्मिन् in 3.1.9).
The god is named from shining — the same base as दिव् in दिव्यम् above.
The प्रातिपदिक, from the root दिव् with guṇa.
The instrumental plural affix भिस् is introduced.
The final स् of भिस् is protected from इत्-hood by this rule, as it ends a विभक्ति.
After a stem in अ, भिस् is replaced by ऐस्.
अ + ऐ: vṛddhi ऐ is the single substitute for both.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
Śaṅkara glosses देवैः as इन्द्रियैः. The sense organs are routinely called देवाः in the Upaniṣads, being seats of presiding deities — cf. Kaṭha 2.1.1 पराञ्चि खानि व्यतृणत् स्वयम्भूः. Read as “gods”, the line becomes a polemic against ritual theism; read as “senses”, it continues the list begun by चक्षुषा and वाचा. The second is almost certainly right, and the grammar cannot decide it — a good discussion prompt.
Austerity as heat — the metaphor is in the root, not the translation.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
The kṛt affix असुन् forms a neuter action-noun.
The nasalised उँ and the final न् of असुन् are इत् and drop, leaving अस्. → तपस्.
The instrumental singular affix टा is introduced.
The initial ट् of टा is an इत् (चु and टु initial in an affix), and is elided. The affix is simply आ.
No षत्व here: see the note. स् + आ stand as they are.
Why not *तपषा? आदेशप्रत्यययोः (P.8.3.59) requires the स् to stand after इण् or कु, and the vowel here is अ, which is outside इण्. Compare चक्षुषा, where the vowel is उ. The rule is not about the affix — असुन् is an affix too — it is about what precedes.
Two markers, डु and ञ्, both dropping immediately — the same pattern as डुधाञ् in निहितम्.
The root as taught, with its markers.
Initial डु and final ञ् are both इत् and drop. The root proper is कृ.
The kṛt affix मनिन् forms a neuter noun.
The nasalised इँ and the final न् of मनिन् are इत् and drop, leaving मन्. The इ is not idle: it makes the affix ङिद्वत्, blocking guṇa where relevant.
guṇa of the root's final ऋ before the ārdhadhātuka affix: ऋ → अर्. → कर्मन्.
The instrumental singular affix टा is introduced.
The initial ट् of टा is an इत् (चु and टु initial in an affix), and is elided. The affix is simply आ.
न् becomes ण् after र् or ष् within the same pada — and still does so when अट्, कु, पु, आङ् or नुम् intervene. Here the intervening म् belongs to पु (प फ ब भ म), so it does not block.
Four instances in this set, decided every time by what stands between र् and न्:
· कर्मणा — only म् (पु class) → ण्
· प्राणः (3.1.9) — only the vowel अ (अट्) → ण्
· प्रसादेन — स् and द् intervene, neither permitted → न्
· प्रजानाम् (3.1.9) — ज् intervenes, not permitted → न्
Memorise the permitted set (अट् · कु · पु · आङ् · नुम्) and all four follow.
Śaṅkara restricts कर्म here to Vedic ritual — अग्निहोत्रादि. The verse is not dismissing action as such; it places the karma-kāṇḍa outside the means of this particular knowledge, consistent with 1.2.7 प्लवा ह्येते अदृढा यज्ञरूपाः.
निपात.
अव्यय.
The case affix is introduced and elided.
Two roots in one compound. The second, षदॢँ, is taught with an initial ष् which is always replaced by स् — a convention, not a sound.
ल्युट् forms neuter action-nouns.
The ल् and the ट् of ल्युट् are both इत् and drop, leaving यु.
यु is replaced by अन. → ज्ञान.
An initial ष् of a root is replaced by स्, and the nasalised ऌँ is an इत् and drops. → सद्.
घञ् in the sense of the action itself; प्र is prefixed as upasarga.
The घ् of घञ् is इत् as an initial कु letter, the ञ् as a final consonant. Both drop, leaving अ. Neither is idle: घ् marks it घित् and ञ् marks it ञित्, and ञित् is what triggers the next step.
The penultimate अ takes vṛddhi आ before a ञित् or णित् affix. → प्रसाद.
Genitive tatpuruṣa: “the प्रसाद of knowledge”.
The instrumental singular affix is introduced.
The ट् of टा is an इत् and drops.
After a stem in अ, टा is replaced by इन.
अ + इ: guṇa ए replaces both. → ज्ञानप्रसादेन, with न् and not ण् — see the note.
प्र- supplies an र्, yet the ending is -एन, not -एण. Between the र् and the न् stand स् and द्, neither in the permitted set of P.8.4.2. Compare रामेण, where only म् and a vowel intervene and the change goes through.
प्रसाद carries two live senses: clearness, serenity (of water, of mind) and grace, favour. Śaṅkara takes the first — the intellect becoming still and transparent once its defects are gone — and the following विशुद्धसत्त्वः supports him. Renderings such as “the favourableness of the intellect”, leans to the second. Both defensible; the class should be told it is a choice. Note too that the compound is genitive: the प्रसाद of knowledge, not knowledge as grace.
The second member सत्त्व is built not from a root but from सत्, itself the शतृ participle of √अस्.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
क्त in the past sense.
The क् of क्त is an इत् and drops, leaving त.
After a झष् sound, the त् or थ् of an affix is replaced by ध्.
A झल् before a झश् becomes जश् — the root-final ध् is devoiced-adjusted to द्. → शुद्ध.
The upasarga वि is prefixed.
त्व forms the abstract noun: “being, essence”. → सत्त्व.
Bahuvrīhi.
विशुद्धसत्त्व is a प्रातिपदिक, so the twenty-one सुप् affixes become available; the nominative singular is सुँ.
The उँ of सुँ is a nasalised vowel in upadeśa, hence an इत्, and is elided. Only स् remains.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
वि॒शुद्ध॑सत्त्व- carries its accent on the prior member, exactly as अचि॑न्त्यरूपम् does in 3.1.7 — बहुव्रीहौ प्रकृत्या पूर्वपदम् (P.6.2.1). Two independent instances in consecutive verses. If students learn one thing about why accented editions are worth the trouble, let it be this.
Not the guṇa of Sāṅkhya here. Śaṅkara reads अन्तःकरण — the inner instrument, the same faculty called चेतस् in 3.1.9 and गुहा in 3.1.7. Three names for one thing across three verses is itself the argument.
Start from तद् in the ablative.
तसिल् replaces the fifth-case affix after तद् and the rest.
The nasalised इँ and the final ल् of तसिल् are इत् and drop, leaving तस्. The ल् is not idle: a लित् affix makes the syllable before it udātta.
The final of तद् is replaced by short अ.
अ + अ: the latter alone is the substitute (पररूप).
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
A taddhita form not admitting all case affixes is termed अव्यय.
निपात.
अव्यय.
The case affix is introduced and elided.
ततस्तु is the pivot of the verse: everything before it is denial, everything after is the positive account. तु marks the reversal, not mere continuation.
तद् is a सर्वनाम.
The accusative singular affix अम् is introduced.
The म् of अम् is protected from इत्-hood, as it ends a विभक्ति.
The final द् of तद् is replaced by short अ.
अ + अम्: the prior vowel alone is the substitute for both.
A traditional alternative derives तद् from √तन् + क्विप् (“that which extends”), with गमः क्वौ, ह्रस्वस्य पिति कृति तुक् and वेरपृक्तस्य. That is a nirukti in the Yāska tradition, not the morphology of the pronoun. Worth keeping as a sidebar because it shows how the two kinds of analysis coexist in the commentarial literature.
तत् (3.1.7) → तं (3.1.8) → एष आत्मा (3.1.9). The demonstrative moves from remote, to object of a verb of seeing, to named subject. The three verses are one movement — the reason for teaching them together.
Parasmaipada by the Dhātupāṭha — which is exactly the problem this form poses.
Markers इँ and र् drop, as at पश्यत्सु.
लट् for present sense.
In Vedic, parasmaipada and ātmanepada are interchanged freely. This is what licenses an ātmanepada ending on a parasmaipada root — the whole reason this form is not पश्यति.
श replaces शप् after दृश्.
The श् of श is an इत् and drops.
दृश् is replaced by पश्य् before a शित् affix.
The टित् ending त has its टि replaced by ए: त → ते.
The finished form.
पश्यते is not a variant spelling of पश्यति. It is an ātmanepada where the classical grammar requires a parasmaipada, licensed as ārṣa by व्यत्ययो बहुलम् (P.3.1.85). Students should meet it as evidence that the text is older than the grammar describing it.
The same form appears six verses earlier, at Muṇḍaka 3.1.3: यदा पश्यः पश्यते रुक्मवर्णम्. Two occurrences in one khaṇḍa makes this a habit of the text, not a scribal accident.
The prefix निस्, “out, away”, with कला, “part, digit”.
Bahuvrīhi: “that from which parts have gone”. The feminine आ is shortened in the compound.
The स् of निस् becomes रुँ and then visarga before the following खर् sound क्.
The visarga becomes ष् when the penultimate sound is इ or उ and the word is not an affix — the rule behind निष्कल, निष्फल, दुष्कृत.
The accusative singular affix is introduced.
The म् survives as before, and अ + अम् merge into one.
Śaṅkara: निरवयवम्. Note the pairing with पश्यते — what is seen has no parts, so the seeing cannot be the assembling of an image. The word is chosen to make the verb impossible in its ordinary sense.
A root ending in ऐ. Whether that ऐ becomes आ or आय् depends entirely on whether the following affix is शित् — see steps 4–5.
The root in upadeśa.
The nasalisation is an इत् and drops.
लट् for present sense.
शानच् replaces लट् to form the ātmanepada present participle.
The श् of शानच् is इत् as an initial शित् marker, the च् as a final consonant. Both drop, leaving आन. The श् is not idle: it makes the affix शित्, which decides step 6.
शप् is introduced; its इत्s drop as usual, leaving अ.
ऐ becomes आय् before a vowel. Note what does not happen: आदेच उपदेशेऽशिति (P.6.1.45) would give ध्या-, but that rule explicitly excludes शित् affixes, and शप् is शित्. One letter in an affix name decides the shape of the word.
After a stem ending in अ, the augment मुँक् is added before आन.
In मुँक्, the उँ and the क् are इत् and drop, leaving म्. Being कित्, it stands at the end of what it is added to — i.e. after ध्याय+अ. → ध्यायमान.
ध्यायमान is a प्रातिपदिक, so the twenty-one सुप् affixes become available; the nominative singular is सुँ.
The उँ of सुँ is a nasalised vowel in upadeśa, hence an इत्, and is elided. Only स् remains.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
The participle is attributive to the same unexpressed subject as विशुद्धसत्त्वः and expresses attendant circumstance: purified, and while meditating, he sees. Not two conditions in sequence but one state described twice.
Set unaccented — the accented text should come from the same edition used for 3.1.7–8. See editor queries.
This Self is subtle, and is to be known by the mind — the mind into which the breath has entered fivefold. The whole thinking of creatures is woven through by the breaths; when it is purified, this Self shines forth.
एषः अणुः आत्मा चेतसा वेदितव्यः, यस्मिन् प्राणः पञ्चधा संविवेश । प्रजानां सर्वं चित्तं प्राणैः ओतम् (अस्ति) । यस्मिन् विशुद्धे एषः आत्मा विभवति ॥
The answer to 3.1.8. Having ruled out eye, speech, senses, austerity and ritual, the text now names the one instrument that works — चेतसा — and immediately explains why it usually does not: it is woven through with the breaths. The condition is not a different faculty but the same one, clarified.
| एषः + अणुः | → | एषोऽणुः | ससजुषो रुः → हशि च → आद्गुणः → एङः पदान्तादतिP.8.2.66P.6.1.114P.6.1.87P.6.1.109The full visarga chain: स् → रु → उ → ओ, then the following अ is absorbed and marked by अवग्रह ऽ |
| अणुः + आत्मा | → | अणुरात्मा | ससजुषो रुःP.8.2.66र् stands before a vowel |
| वेदितव्यः + यस्मिन् | → | वेदितव्यो यस्मिन् | हशि चआद्गुणःP.6.1.114P.6.1.87अः + य् → ओ य् |
| प्राणैः + चित्तम् | → | प्राणैश्चित्तम् | विसर्जनीयस्य सःस्तोः श्चुना श्चुःP.8.3.34P.8.4.40visarga → स् → श् before the palatal च् |
| ओतम् + प्रजानाम् | → | ओतं प्रजानाम् | मोऽनुस्वारःP.8.3.23म् → anusvāra |
| विभवति + एष | → | विभवत्येष | इको यणचिP.6.1.77इ → य् |
| एषः + आत्मा | → | एष आत्मा | भोभगोअघोअपूर्वस्य योऽशिलोपः शाकल्यस्यP.8.3.17P.8.3.19रु → य्, and the य् is then dropped — the hiatus is genuine, not a printing gap |
The most sandhi in one line: एषः appears twice, once merging completely (एषोऽणुः) and once leaving an open hiatus (एष आत्मा). Same word, same case, opposite outcomes — decided entirely by whether the next vowel is अ.
एतद् is a सर्वनाम.
The nominative singular affix is introduced.
The उँ of सुँ is an इत् and drops.
The final द् of एतद् is replaced by short अ.
The त् or द् of तद्/एतद् — when not final — is replaced by स् before सु.
This स् is a substitute (आदेश) and stands after ए, which is in इण्. It therefore becomes ष्.
The अ of एष and the following merge by पररूप.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
एषः completes the set. The ष् here comes not from an affix but from a substitute — P.8.3.59 covers both, which is why the rule is named आदेश-प्रत्यययोः. And once again the trigger is the preceding vowel, ए, a member of इण्.
एतद् is proximate, तद् remote. After two verses of तत् and तम्, the switch to एषः is deliberate: what was pointed at from a distance is now named at hand. The grammar of the pronoun carries the argument.
The प्रातिपदिक, “fine, minute”.
अणु is a प्रातिपदिक, so the twenty-one सुप् affixes become available; the nominative singular is सुँ.
The उँ of सुँ is a nasalised vowel in upadeśa, hence an इत्, and is elided. Only स् remains.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
सूक्ष्मात् सूक्ष्मतरम् there, अणुः here — the same claim in a smaller word, now attached to a named subject rather than a bare pronoun. Compare Kaṭha 1.2.20 अणोरणीयान् महतो महीयान्, which uses both halves of the Muṇḍaka's paradox in a single line.
The प्रातिपदिक, an न्-final stem.
The nominative singular affix is introduced.
The उँ of सुँ is an इत् and drops.
सुट् — the first five सुप् affixes — is termed सर्वनामस्थान in the non-neuter. This is the condition for the next step.
The penultimate vowel of an न्-final aṅga is lengthened before a सर्वनामस्थान affix (outside the vocative). अ → आ.
The single-consonant affix स् is elided after a stem ending in a consonant. Now compare बृहत्, where the neuter rule P.7.1.23 applied instead.
The न् at the end of a nominal stem is elided when the stem stands as a pada. → आत्मा.
Three operations, three different chapters of the Aṣṭādhyāyī, all visible in the surface form. Compare कर्मणा in 3.1.8: same stem class, neuter, and the न् survives because a case affix follows it rather than nothing. Teaching the two together shows why P.8.2.7 specifies pada-final.
The same root gives चित्तम् later in this verse — the faculty and its content, morphologically distinguished.
The root in upadeśa.
The nasalised ईँ is an इत् and drops.
The kṛt affix असुन् forms a neuter action-noun.
The उँ and the न् of असुन् are इत् and drop.
guṇa of the light penultimate इ before the ārdhadhātuka affix: इ → ए. Note it is this rule and not सार्वधातुकार्धधातुकयोः (P.7.3.84), which requires the aṅga to end in an इक् vowel — चित् ends in त्. → चेतस्.
The instrumental singular affix टा is introduced.
The initial ट् of टा is an इत् (चु and टु initial in an affix), and is elided. The affix is simply आ.
No षत्व — the स् follows अ, which is outside इण्.
चक्षुषा (ष) · तपसा (स) · चेतसा (स) · एषः (ष). Two vowels decide all four: उ and ए are in इण्, अ is not.
The word the argument rests on. 3.1.8 denied चक्षुषा, वाचा, देवैः, तपसा, कर्मणा — all instrumentals. 3.1.9 grants one: चेतसा. The argument is conducted entirely in the third case.
seṭ — it admits the इट् augment, which is why the form is वेदितव्य and not *वेत्तव्य.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
The कृत्य affixes तव्यत्, तव्य and अनीयर् express necessity or fitness.
The final त् of तव्यत् is an इत् and drops, leaving तव्य. It is not idle: a तित् affix is svarita-accented.
An ārdhadhātuka affix beginning with a वल् sound takes the augment इट्. This is the seṭ property of the root in action.
guṇa of the light penultimate इ before the affix: इ → ए. → वेदितव्य.
वेदितव्य is a प्रातिपदिक, so the twenty-one सुप् affixes become available; the nominative singular is सुँ.
The उँ of सुँ is a nasalised vowel in upadeśa, hence an इत्, and is elided. Only स् remains.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
A कृत्य affix is passive and carries obligation: is to be known. It takes its agent in the instrumental — hence चेतसा. Set this against गृह्यते in 3.1.8, also passive, also governing instrumentals. The two verses use the same sentence-shape to opposite ends: what fails to grasp, and what must know.
यद् is a सर्वनाम.
The locative singular affix is introduced.
The ङ् of ङि is an इत् and drops.
After a सर्वनाम stem in अ, ङसि and ङि are replaced by स्मात् and स्मिन्.
The final द् of यद् is replaced by short अ.
अ + स्मिन्: the stem's अ and the following merge by पररूप. → यस्मिन्.
First (यस्मिन् प्राणः पञ्चधा संविवेश): what is the antecedent — चेतसि, the mind just named, or शरीरे, the body, understood? Śaṅkara supplies the body; the syntax allows either, and published translations split on it. Second (यस्मिन् विशुद्धे): a locative absolute — यस्य च भावेन भावलक्षणम् (P.2.3.37) — “when that is purified”, where the antecedent is almost certainly चित्तम्. Open, and worth discussing.
A two-letter root that all but disappears into its own derivative.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
घञ् in the sense of the action; प्र is prefixed as upasarga.
The घ् of घञ् is इत् as a कु letter initial in a non-taddhita affix; the final ञ् is इत् by हलन्त्यम्. Both drop, leaving अ.
The अ of प्र and the अ of अन् are सवर्ण: long आ replaces both. → प्रान.
The र् of प्र causes न् → ण्. Only the vowel आ intervenes, and vowels belong to अट्, which does not block.
प्राण is a प्रातिपदिक, so the twenty-one सुप् affixes become available; the nominative singular is सुँ.
The उँ of सुँ is a nasalised vowel in upadeśa, hence an इत्, and is elided. Only स् remains.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
The cleanest of the four: nothing but a vowel between र् and न्. Teach प्राणः first, then कर्मणा (म् permitted), then प्रसादेन and प्रजानाम् (स्/द् and ज् not permitted). In that order the rule assembles itself.
धा follows a numeral in the sense “in so many ways”.
The final च् of धाच् is इत् by हलन्त्यम् and drops, leaving धा. It is not idle: a चित् affix makes the last syllable udātta.
The टि portion — the final न् with its vowel — of an अन्-final stem is elided before a taddhita affix. → पञ्चधा.
Therefore अव्यय.
प्राण, अपान, व्यान, उदान, समान. The word पञ्चधा does not list them; it presumes the student knows the scheme from 2.1.8 and from the Praśna.
The perfect is the one tense where the root is doubled — and most of the derivation is about cutting the copy back down.
The root in upadeśa.
The nasalised अँ is an इत् and drops.
लिट् for remote past.
ल् is replaced by the perfect endings; third singular parasmaipada is णल्.
The ण् and the ल् of णल् are इत् and drop, leaving अ. The ण् is not idle — a णित् affix causes vṛddhi or guṇa, which is step 8.
Before लिट्, the first syllable of the root is doubled. The copy is called अभ्यास.
Of the reduplicate, only the first consonant is kept; the rest is dropped. विश् → वि.
guṇa of the light penultimate इ of the root before the णित् affix: इ → ए. Note it affects the root, not the अभ्यास. → विवेश.
The upasarga सम् is prefixed.
Pada-final म् becomes anusvāra before a consonant.
The perfect here describes a settled condition rather than a remote event — the breath's fivefold entry is the standing state of every embodied creature, not something that happened once. Common in the Upaniṣads; flag it before students translate it as a simple past.
The stem, derived as at प्राणः above.
The instrumental plural affix भिस् is introduced.
The final स् of भिस् is protected from इत्-hood by this rule, as it ends a विभक्ति.
After a stem in अ, भिस् is replaced by ऐस्.
अ + ऐ: vṛddhi ऐ is the single substitute for both.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
The instrument of the weaving. Note the shift from singular प्राणः (what entered) to plural प्राणैः (what does the weaving) within two lines — the same breath counted as one and as five.
The same root as चेतसा above — but with क्त instead of असुन्, and with no guṇa, because क्त is कित्.
Markers dropped, as at चेतसा.
क्त in the past sense.
The क् of क्त is an इत् and drops. Being कित्, the affix blocks guṇa — which is exactly why this word has चि- where चेतसा has चे-.
त् + त stand as they are. → चित्त.
The nominative singular affix is introduced.
After a neuter stem ending in अ, अम् replaces सु and अम्. सु is gone; अम् stands in its place.
The म् of अम् would be an इत् by हलन्त्यम्, but this rule exempts त थ द ध न स म at the end of a विभक्ति. The म् therefore survives into the finished word.
अ + अम्: the prior vowel alone is the substitute for both. The stem's final अ and the अ of अम् merge into one.
One root, two formations, one verse. चेतस् takes असुन्, an ārdhadhātuka that permits guṇa (चे-); चित्त takes क्त, which is कित् and blocks it (चि-). The single letter क् in the affix name accounts for the whole difference in shape.
सर्व heads the सर्वादि list and is a सर्वनाम.
The nominative singular affix is introduced.
After a neuter stem ending in अ, अम् replaces सु and अम्. सु is gone; अम् stands in its place.
The म् of अम् would be an इत् by हलन्त्यम्, but this rule exempts त थ द ध न स म at the end of a विभक्ति. The म् therefore survives into the finished word.
अ + अम्: the prior vowel alone is the substitute for both. The stem's final अ and the अ of अम् merge into one.
Almost nothing of the root survives into the finished word: व → उ by सम्प्रसारण, then उ absorbed into the prefix.
The root in upadeśa.
The final ञ् is an इत् and drops. It marks the root उभयपद.
क्त in the past sense.
The क् of क्त is an इत् and drops. Being कित्, it licenses the next step.
सम्प्रसारण: the व् of वे is replaced by its corresponding vowel उ before the कित् affix. → उत.
The upasarga आ is prefixed.
आ + उ: guṇa ओ is the single substitute for both. → ओत.
The nominative singular affix is introduced.
After a neuter stem ending in अ, अम् replaces सु and अम्. सु is gone; अम् stands in its place.
The म् of अम् would be an इत् by हलन्त्यम्, but this rule exempts त थ द ध न स म at the end of a विभक्ति. The म् therefore survives into the finished word.
अ + अम्: the prior vowel alone is the substitute for both. The stem's final अ and the अ of अम् merge into one.
गृह्यते: applies. वाचा: blocked by an Uṇādi rule. ओतम्: applies. Three verses, three outcomes; the deciding factor each time is the affix — कित्/ङित् or not, and whether a special rule overrides. The set is worth teaching together.
Warp and woof — a fixed pair. Compare Bṛhadāraṇyaka 3.8.7, where everything is ओतं च प्रोतं च in the imperishable. The Muṇḍaka uses only the first half, but the image is the same loom.
“Creatures” as “those born forth” — the etymology is transparent and the commentary uses it.
The root in upadeśa.
The nasalised ईँ is an इत् and drops.
जन्, सन् and खन् substitute आ for their final न् before a झल्-initial कित्/ङित् affix. This step has to come first, because the next rule requires an आ-final root.
The कृत् affix क follows a root ending in आ when an upasarga precedes.
The क् of क is इत् as an initial कु letter and drops, leaving अ. Being कित्, it is also what licensed P.6.4.42 above.
The आ of an आ-final aṅga is elided before an ārdhadhātuka affix beginning with a vowel. → प्रज.
A stem in अ takes टाप् in the feminine; its ट् and प् are इत् and drop, leaving आ, which merges with the stem's अ. → प्रजा.
The genitive plural affix आम् is introduced.
आम् takes the augment नुट् after a short-vowel stem, a नदी stem or an आप् stem. By आद्यन्तौ टकितौ (P.1.1.46) a ट्-marked augment stands first, so न् precedes आम्.
No णत्व, though an र् is present — see the note. The finished form.
The र् of प्र- is present and the न् of नुट् is present, yet no ण् — because ज् stands between them, and palatals are not in the permitted set of P.8.4.2. Pair with प्रसादेन in 3.1.8: two blockers, different consonants, same principle.
आतश्चोपसर्गे prescribes क, and requires a root ending in आ — which जन् becomes only by जनसनखनां सञ्झलोः (P.6.4.42). So 6.4.42 must come first, and it is itself licensed by the affix being कित्. Some commentaries give a ड route instead; if the course prefers it, the substitution should be shown rather than assumed.
Same participle as in विशुद्धसत्त्वः, 3.1.8 — but here standing on its own.
As derived in full at विशुद्धसत्त्वः, 3.1.8.
The upasarga वि is prefixed.
The locative singular affix ङि is introduced.
The initial ङ् of ङि is an इत् — ल्, श् and the velars कु are इत् when initial in a non-taddhita affix — and is elided. The affix is इ.
अ followed by इ: guṇa ए is the single substitute for both.
The seventh case marks the circumstance whose occurrence characterises another — the locative absolute.
विशुद्ध appeared in 3.1.8 inside the bahuvrīhi विशुद्धसत्त्वः, describing the person. Here it stands in a locative absolute, describing the condition. One participle, two syntactic roles, adjacent verses.
The first root in the Dhātupāṭha, and the one every student derives first. Compare विभाति in 3.1.7 — a different root, and the divergence is at शप्.
The root as taught: भू सत्तायाम्.
लट् follows a root when present action is denoted.
What is introduced is a प्रत्यय, and it stands after the root.
ल् is replaced by one of eighteen personal endings; third person singular parasmaipada is तिप्. Its प् is an इत् by हलन्त्यम् and drops, leaving ति.
शप् between root and ending.
The श् and प् of शप् are इत् and drop, leaving अ. Here — unlike विभाति — the class sign stays, because भू is भ्वादि, not अदादि.
guṇa of the root's final vowel before the sārvadhātuka affix: ऊ → ओ.
ओ becomes अव् before a vowel. → भवति.
The upasarga वि is prefixed.
3.1.7 ends its second line with विभाति, from √भा दीप्तौ, class 2 — शप् is elided by P.2.4.72, so the root stands bare and nothing further happens. 3.1.9 ends with विभवति, from √भू सत्तायाम्, class 1 — शप् survives, and it drives first guṇa and then the semivowel change. Two roots, two classes, adjacent verses, one syllable apart, same preverb. The two ladders side by side show the divergence, which is at one step.
Two of this module's grammatical arguments rest on the accent marks in the printed text — the compound type of अचि॑न्त्यरूपम् and वि॒शुद्ध॑सत्त्वः is settled by where the udātta falls. That is only an argument if you can read the marks. Here is the system.
The raised pitch. In the standard printed convention it carries no mark of its own — you locate it by finding the syllable before a marked svarita, or by the anudātta marks around it. This is the commonest source of confusion.
The lowered pitch. Everything not udātta or svarita is anudātta by अनुदात्तं पदमेकवर्जम् (P.6.1.158) — in a pada, all syllables but one are low.
A contour tone, high falling to low. It arises automatically on the syllable following an udātta — उदात्तादनुदात्तस्य स्वरितः (P.8.4.66) — which is why a marked svarita is a pointer backwards to the accented syllable.
Practical rule for reading this text: find the ॑ mark, then look one syllable to its left. That syllable carries the udātta. In बृह॑त् the ॑ sits on -ह-, so the accent is on बृ-. In अचि॑न्त्यरूपम् it sits on -चि-, so the accent is on अ- — the first syllable of the prior member, exactly as a bahuvrīhi requires.
| P.6.1.158 | अनुदात्तं पदमेकवर्जम् | One syllable of a pada is accented; all the rest are anudātta. Accent is a property of the word, not of each syllable. |
| P.6.1.159 | कर्षात्वतो घञोऽन्त उदात्तः | A घञ् derivative is end-accented — relevant to प्रसाद and प्राण in this module. |
| P.6.2.1 | बहुव्रीहौ प्रकृत्या पूर्वपदम् | In a bahuvrīhi the prior member keeps the accent it had on its own. The diagnostic used twice in these verses. |
| P.6.1.223 | समासस्य | The default for a compound is end-accent — which is what P.6.2.1 overrides, and why the bahuvrīhi test works. |
| P.8.4.66 | उदात्तादनुदात्तस्य स्वरितः | An anudātta following an udātta becomes svarita. The rule that makes the ॑ mark a backward pointer. |
| P.3.1.3 | आद्युदात्तश्च | An affix is by default first-syllable accented — which is why anubandhas like the प् of तरप् and the ल् of तसिल् matter even after they are deleted. |
Anubandhas and accent are the same subject. A पित् affix is anudātta (P.3.1.4), a ञित् or नित् affix is first-accented (P.6.1.197), a लित् affix accents the syllable before it (P.6.1.193). Every one of those markers is deleted by P.1.3.9 long before the word is finished — and every one of them leaves its trace in the pitch. The इत् thread and the accent thread are two views of one mechanism.
Hand-checked against the standard paradigms rather than generated, and worth a second pair of eyes before class. The point of printing them is not reference — students have Kale for that — but placement: each table shows the form that occurs in these verses inside the system it belongs to, so that चक्षुषा stops being a fact to memorise and becomes a consequence.
self — masculine न्-stem
| singular | dual | plural | |
|---|---|---|---|
| नom. | आत्मा | आत्मानौ | आत्मानः |
| acc. | आत्मानम् | आत्मानौ | आत्मनः |
| inst. | आत्मना | आत्मभ्याम् | आत्मभिः |
| dat. | आत्मने | आत्मभ्याम् | आत्मभ्यः |
| abl. | आत्मनः | आत्मभ्याम् | आत्मभ्यः |
| gen. | आत्मनः | आत्मनोः | आत्मनाम् |
| loc. | आत्मनि | आत्मनोः | आत्मसु |
The strong cases lengthen the penultimate (P.6.4.8); the nominative singular additionally loses the final न् (P.8.2.7). आत्मा in 3.1.9 is the model.
that — pronoun, neuter
| singular | dual | plural | |
|---|---|---|---|
| नom. | तत् | ते | तानि |
| acc. | तत् | ते | तानि |
| inst. | तेन | ताभ्याम् | तैः |
| dat. | तस्मै | ताभ्याम् | तेभ्यः |
| abl. | तस्मात् | ताभ्याम् | तेभ्यः |
| gen. | तस्य | तयोः | तेषाम् |
| loc. | तस्मिन् | तयोः | तेषु |
तत् in 3.1.7. Compare the masculine below: the neuter nominative escapes both त्यदादीनामः and तदोः सः because the सु is elided by लुक् and P.1.1.63 withholds प्रत्ययलक्षणम्.
that — pronoun, masculine
| singular | dual | plural | |
|---|---|---|---|
| नom. | सः | तौ | ते |
| acc. | तम् | तौ | तान् |
| inst. | तेन | ताभ्याम् | तैः |
| dat. | तस्मै | ताभ्याम् | तेभ्यः |
| abl. | तस्मात् | ताभ्याम् | तेभ्यः |
| gen. | तस्य | तयोः | तेषाम् |
| loc. | तस्मिन् | तयोः | तेषु |
तम् in 3.1.8. Note सः against neuter तत् — the whole difference is P.1.1.63, and यस्मिन् in 3.1.9 follows this pattern exactly with य- for त-.
eye — neuter अस्/उस्-stem
| singular | dual | plural | |
|---|---|---|---|
| नom. | चक्षुः | चक्षुषी | चक्षूंषि |
| acc. | चक्षुः | चक्षुषी | चक्षूंषि |
| inst. | चक्षुषा | चक्षुर्भ्याम् | चक्षुर्भिः |
| dat. | चक्षुषे | चक्षुर्भ्याम् | चक्षुर्भ्यः |
| abl. | चक्षुषः | चक्षुर्भ्याम् | चक्षुर्भ्यः |
| gen. | चक्षुषः | चक्षुषोः | चक्षुषाम् |
| loc. | चक्षुषि | चक्षुषोः | चक्षुःषु |
चक्षुषा in 3.1.8. The ष् appears wherever the स् stands after उ and before a vowel or voiced sound (P.8.3.59). Set this beside तपस् below — same shape, different vowel, different outcome throughout.
austerity — neuter अस्-stem
| singular | dual | plural | |
|---|---|---|---|
| नom. | तपः | तपसी | तपांसि |
| acc. | तपः | तपसी | तपांसि |
| inst. | तपसा | तपोभ्याम् | तपोभिः |
| dat. | तपसे | तपोभ्याम् | तपोभ्यः |
| abl. | तपसः | तपोभ्याम् | तपोभ्यः |
| gen. | तपसः | तपसोः | तपसाम् |
| loc. | तपसि | तपसोः | तपःसु |
तपसा in 3.1.8; चेतस् in 3.1.9 declines identically. Every form keeps स्, because the preceding vowel is अ, which is outside इण्. Set against चक्षुस् above, the pair isolates the one conditioning factor.
seeing — शतृ participle, masculine
| singular | dual | plural | |
|---|---|---|---|
| नom. | पश्यन् | पश्यन्तौ | पश्यन्तः |
| acc. | पश्यन्तम् | पश्यन्तौ | पश्यतः |
| inst. | पश्यता | पश्यद्भ्याम् | पश्यद्भिः |
| dat. | पश्यते | पश्यद्भ्याम् | पश्यद्भ्यः |
| abl. | पश्यतः | पश्यद्भ्याम् | पश्यद्भ्यः |
| gen. | पश्यतः | पश्यतोः | पश्यताम् |
| loc. | पश्यति | पश्यतोः | पश्यत्सु |
पश्यत्सु in 3.1.7. नुम् is inserted only before a सर्वनामस्थान affix (P.7.1.70) — which is why the nominative plural is पश्यन्तः but the locative plural is पश्यत्सु, with no nasal.
Every rule invoked in a derivation or sandhi table, in Aṣṭādhyāyī order, with the number of times it is used. Search any number above to jump to the words that use it.
| P.1.1.27 | सर्वादीनि सर्वनामानि | ×6 show words → |
| P.1.1.37 | स्वरादिनिपातमव्ययम् | ×6 show words → |
| P.1.1.38 | तद्धितश्चासर्वविभक्तिः | ×3 show words → |
| P.1.1.43 | सुडनपुंसकस्य | ×1 show words → |
| P.1.1.46 | आद्यन्तौ टकितौ | ×1 show words → |
| P.1.1.47 | मिदचोऽन्त्यात्परः | ×1 show words → |
| P.1.1.63 | न लुमताङ्गस्य | ×1 show words → |
| P.1.2.5 | असंयोगाल्लिट् कित् | ×1 show words → |
| P.1.3.2 | उपदेशेऽजनुनासिक इत् | ×37 show words → |
| P.1.3.3 | हलन्त्यम् | ×32 show words → |
| P.1.3.4 | न विभक्तौ तुस्माः | ×14 show words → |
| P.1.3.5 | आदिर्ञिटुडवः | ×2 show words → |
| P.1.3.7 | चुटू | ×10 show words → |
| P.1.3.8 | लशक्वतद्धिते | ×20 show words → |
| P.1.3.9 | तस्य लोपः | ×79 show words → |
| P.1.3.13 | भावकर्मणोः | ×1 show words → |
| P.1.4.57 | चादयोऽसत्त्वे | ×5 show words → |
| P.1.4.58 | प्रादयः | ×1 show words → |
| P.1.4.59 | उपसर्गाः क्रियायोगे | ×8 show words → |
| P.1.4.96 | अपिः पदार्थसम्भावनान्ववसर्गगर्हासमुच्चयेषु | ×1 show words → |
| P.2.2.6 | नञ् | ×2 show words → |
| P.2.2.8 | षष्ठी | ×1 show words → |
| P.2.2.18 | कुगतिप्रादयः | ×2 show words → |
| P.2.2.24 | अनेकमन्यपदार्थे | ×4 show words → |
| P.2.3.37 | यस्य च भावेन भावलक्षणम् | ×1 show words → |
| P.2.4.72 | अदिप्रभृतिभ्यः शपः | ×1 show words → |
| P.2.4.82 | अव्ययादाप्सुपः | ×6 show words → |
| P.3.1.1 | प्रत्ययः | ×2 show words → |
| P.3.1.2 | परश्च | ×2 show words → |
| P.3.1.25 | सत्यापपाशरूपवीणा… चुरादिभ्यो णिच् | ×1 show words → |
| P.3.1.67 | सार्वधातुके यक् | ×1 show words → |
| P.3.1.68 | कर्तरि शप् | ×3 show words → |
| P.3.1.85 | व्यत्ययो बहुलम् | ×1 show words → |
| P.3.1.96 | तव्यत्तव्यानीयरः | ×1 show words → |
| P.3.1.124 | ऋहलोर्ण्यत् | ×1 show words → |
| P.3.1.136 | आतश्चोपसर्गे | ×1 show words → |
| P.3.1.137 | पाघ्राध्माधेट्दृशः शः | ×2 show words → |
| P.3.2.102 | निष्ठा | ×5 show words → |
| P.3.2.115 | परोक्षे लिट् | ×2 show words → |
| P.3.2.123 | वर्तमाने लट् | ×6 show words → |
| P.3.2.124 | लटः शतृशानचावप्रथमासमानाधिकरणे | ×2 show words → |
| P.3.3.18 | भावे | ×2 show words → |
| P.3.3.115 | ल्युट् च | ×1 show words → |
| P.3.4.77 | लस्य | ×5 show words → |
| P.3.4.78 | तिप्तस्झिसिप्थस्थमिब्वस्मस्… · तिप्तस्झि… | ×4 show words → |
| P.3.4.79 | टित आत्मनेपदानां टेरे | ×2 show words → |
| P.3.4.82 | परस्मैपदानां णलतुसुस्थलथुसणल्वमाः | ×1 show words → |
| P.4.1.1 | ङ्याप्प्रातिपदिकात् | ×44 show words → |
| P.4.1.2 | स्वौजसमौट्… | ×44 show words → |
| P.4.1.4 | अजाद्यतष्टाप् | ×2 show words → |
| P.5.1.119 | तस्य भावस्त्वतलौ | ×1 show words → |
| P.5.1.122 | पृथ्वादिभ्य इमनिज्वा | ×1 show words → |
| P.5.3.3 | इदम इश् | ×1 show words → |
| P.5.3.7 | पञ्चम्यास्तसिल् | ×1 show words → |
| P.5.3.11 | इदमो हः | ×1 show words → |
| P.5.3.42 | सङ्ख्याया विधार्थे धा | ×1 show words → |
| P.5.3.57 | द्विवचनविभज्योपपदे तरबीयसुनौ | ×1 show words → |
| P.6.1.8 | लिटि धातोरनभ्यासस्य | ×2 show words → |
| P.6.1.16 | ग्रहिज्यावयिव्यधिवष्टिविचतिवृश्चतिपृच्छतिभृज्जतीनां ङिति च | ×2 show words → |
| P.6.1.64 | धात्वादेः षः सः | ×2 show words → |
| P.6.1.67 | वेरपृक्तस्य | ×1 show words → |
| P.6.1.68 | हल्ङ्याब्भ्यो दीर्घात् सुतिस्यपृक्तं हल् | ×1 show words → |
| P.6.1.77 | इको यणचि | ×4 show words → |
| P.6.1.78 | एचोऽयवायावः | ×2 show words → |
| P.6.1.87 | आद्गुणः | ×9 show words → |
| P.6.1.88 | वृद्धिरेचि | ×5 show words → |
| P.6.1.89 | एत्येधत्यूठ्सु | ×1 show words → |
| P.6.1.97 | अतो गुणे | ×4 show words → |
| P.6.1.101 | अकः सवर्णे दीर्घः | ×6 show words → |
| P.6.1.107 | अमि पूर्वः | ×9 show words → |
| P.6.1.108 | सम्प्रसारणाच्च | ×1 show words → |
| P.6.1.109 | एङः पदान्तादति | ×1 show words → |
| P.6.1.113 | अतो रोरप्लुतादप्लुते | ×2 show words → |
| P.6.1.114 | हशि च | ×2 show words → |
| P.6.3.73 | नलोपो नञः | ×2 show words → |
| P.6.3.74 | तस्मान्नुडचि | ×1 show words → |
| P.6.4.8 | सर्वनामस्थाने चासम्बुद्धौ | ×2 show words → |
| P.6.4.24 | अनिदितां हल उपधायाः क्ङिति | ×1 show words → |
| P.6.4.42 | जनसनखनां सञ्झलोः | ×1 show words → |
| P.6.4.51 | णेरनिटि | ×1 show words → |
| P.6.4.64 | आतो लोप इटि च | ×1 show words → |
| P.6.4.143 | टेः | ×1 show words → |
| P.6.4.144 | नस्तद्धिते | ×1 show words → |
| P.7.1.1 | युवोरनाकौ | ×1 show words → |
| P.7.1.9 | अतो भिस ऐस् | ×3 show words → |
| P.7.1.12 | टाङसिङसामिनात्स्याः | ×3 show words → |
| P.7.1.15 | ङसिङ्योः स्मात्स्मिनौ | ×1 show words → |
| P.7.1.23 | स्वमोर्नपुंसकात् | ×2 show words → |
| P.7.1.24 | अतोऽम् | ×7 show words → |
| P.7.1.39 | सुपां सुलुक्पूर्वसवर्णाच्छेयाडाड्यायाजालः | ×1 show words → |
| P.7.1.54 | ह्रस्वनद्यापो नुट् | ×1 show words → |
| P.7.1.58 | इदितो नुम् धातोः | ×1 show words → |
| P.7.2.35 | आर्धधातुकस्येड्वलादेः | ×1 show words → |
| P.7.2.82 | आने मुक् | ×1 show words → |
| P.7.2.102 | त्यदादीनामः | ×4 show words → |
| P.7.2.106 | तदोः सः सावनन्त्ययोः | ×1 show words → |
| P.7.2.116 | अत उपधायाः | ×1 show words → |
| P.7.3.50 | ठस्येकः | ×1 show words → |
| P.7.3.78 | पाघ्राध्मास्थाम्नादाण्दृश्यर्तिसर्तिशदसदां पिबजिघ्रधमतिष्ठमनयच्छपश्यर्च्छधौशीयसीदाः · पाघ्राध्मास्थाम्नादाण्दृश्यर्तिसर्ति… पश्य… | ×3 show words → |
| P.7.3.84 | सार्वधातुकार्धधातुकयोः | ×2 show words → |
| P.7.3.86 | पुगन्तलघूपधस्य च | ×3 show words → |
| P.7.3.113 | याडापः | ×1 show words → |
| P.7.3.116 | ङेराम् नद्याम्नीभ्यः | ×1 show words → |
| P.7.4.42 | दधातेर्हिः | ×1 show words → |
| P.7.4.60 | हलादिः शेषः | ×2 show words → |
| P.8.2.7 | न लोपः प्रातिपदिकान्तस्य · नलोपः प्रातिपदिकान्तस्य | ×2 show words → |
| P.8.2.39 | झलां जशोऽन्ते | ×1 show words → |
| P.8.2.40 | झषस्तथोर्धोऽधः | ×2 show words → |
| P.8.2.66 | ससजुषो रुः · ससजुषो रुः → हशि च → आद्गुणः → एङः पदान्तादति | ×16 show words → |
| P.8.3.15 | खरवसानयोर्विसर्जनीयः | ×12 show words → |
| P.8.3.17 | भोभगोअघोअपूर्वस्य योऽशि | ×1 show words → |
| P.8.3.19 | लोपः शाकल्यस्य | ×1 show words → |
| P.8.3.23 | मोऽनुस्वारः | ×4 show words → |
| P.8.3.24 | नश्चापदान्तस्य झलि | ×1 show words → |
| P.8.3.32 | ङमो ह्रस्वादचि ङमुण्नित्यम् | ×1 show words → |
| P.8.3.34 | विसर्जनीयस्य सः | ×4 show words → |
| P.8.3.41 | इदुदुपधस्य चाप्रत्ययस्य | ×1 show words → |
| P.8.3.57 | इण्कोः | ×1 show words → |
| P.8.3.59 | आदेशप्रत्यययोः | ×3 show words → |
| P.8.4.1 | रषाभ्यां नो णः समानपदे | ×2 show words → |
| P.8.4.2 | अट्कुप्वाङ्नुम्व्यवायेऽपि | ×2 show words → |
| P.8.4.40 | स्तोः श्चुना श्चुः | ×3 show words → |
| P.8.4.53 | झलां जश् झशि | ×3 show words → |
| P.8.4.56 | वाऽवसाने | ×1 show words → |
| Uṇ. 2.57 | क्विब्वचिप्रच्छ्यायतस्तुकटप्रुजुश्रीणां दीर्घोऽसम्प्रसारणं च · …दीर्घोऽसम्प्रसारणं च | ×2 show words → |
| Uṇ. 2.84 | निपातनात् सिद्धम् · वर्तमाने पृषद्बृहन्महज्जगत् | ×2 show words → |
| Uṇ. 2.119 | चक्षेः शिच्च | ×1 show words → |
Points that are unsettled, disputed between editions, or that need checking against the text the course uses.
3.1.7 and 3.1.8 are set with their Vedic accents; the rest are not. The udātta and anudātta placement should be supplied from a single accented edition rather than reconstructed. Two arguments in this module — the compound type of अचि॑न्त्यरूपम् and वि॒शुद्ध॑सत्त्वः — turn on the accent, and would gain further instances if the other verses were marked.
Editions number the Uṇādi sūtras differently. The three cited here (2.57, 2.84, 2.119) follow one common numbering and should be checked against the edition in use. They are also the only rules in this module that could not be checked against a machine-readable text.
दिव् + य is uncontroversial as a formation, but the licensing sūtra should be filled in from the भवार्थ taddhitas (P.4.3.53 ff.). Left open rather than guessed at: an invented rule number would be worse than a visible gap.
DP 1.771 gives बृहिँ वृद्धौ, which is इदित् and therefore takes नुम् — so the derivation has to account for a nasal that the finished word does not show. Several editions also carry a non-इदित् बृहँ, from which बृहत् derives with no nasal at any point. The card presents both; the course should settle which root it teaches from.
Where a rule has a familiar number in Ballantyne's Laghu Siddhānta Kaumudī, it is given alongside the Pāṇini reference. The coverage is partial; completing it would help students working from that text.
The English renderings are original to this module, so it can be distributed without a permissions question. Where a published translation is preferred, it will need attribution and, for a public course, a licence.
Notes state findings flat and neutral: no praise of the reader, no superlatives about how useful a point is, no manufactured drama, and none of the intensifiers that inflate without adding. If a contrast is worth teaching, showing it is the argument. Please keep additions to the same standard.
Eighteen transitions drawn automatically from the derivations above, one per distinct rule, in fixed random order. Each shows a form before and after; name the sūtra that did it. Because these are generated from the same data as the ladders, they cannot drift out of step with the content.
The क् of क्त is an इत् and drops. Being कित्, it licenses the next step.
The instrumental plural affix भिस् is introduced.
The त् or द् of तद्/एतद् — when not final — is replaced by स् before सु.
An initial ष् of a root is replaced by स्, and the nasalised ऌँ is an इत् and drops. → सद्.
Pada-final र् becomes visarga before a खर् sound or at a pause.
Pada-final म् becomes anusvāra before a consonant.
The उँ of सुँ is a nasalised vowel in upadeśa, hence an इत्, and is elided. Only स् remains.
Pada-final स् is replaced by रुँ; the उँ is an इत् and drops at once, leaving र्.
क्त and क्तवतु follow a root in the sense of past action; they are termed निष्ठा.
Initial ञि, टु, डु of a root are इत्, and a final consonant is इत्. So डु and ञ् both go. The root proper is धा.
The affix उसिँ follows चक्ष्, and is treated as though it had श् as an इत्.
The final प् of तरप् is an इत् and is elided. It is not idle either: a पित् affix is anudātta, which is why the accent sits where it does. → सूक्ष्मतर.
अ + इ: guṇa ए replaces both. → ज्ञानप्रसादेन, with न् and not ण् — see the note.
guṇa of the light penultimate इ before the ārdhadhātuka affix: इ → ए. Note it is this rule and not सार्वधातुकार्धधातुकयोः (P.7.3.84), which requires the aṅga to end in an इक् vowel — चित् ends in त्. → चेतस्.