Book 3 - Pada-kāṇḍa (8): Kriyā-samuddeśa (On Action)

3.8.1 unformalized

यावत् सिद्धमसिद्धं वा साध्यत्वेनाभिधीयते ।
आश्रितक्रमरूपत्वात् तत् क्रियेत्यभिधीयते ॥ १ ॥

yāvat siddhamasiddhaṃ vā sādhyatvenābhidhīyate |
āśritakramarūpatvāt tat kriyetyabhidhīyate || 1 ||

Whatever, accomplished or unaccomplished, is expressed as what-is-to-be-accomplished — from its having a form in which sequence is resorted to, that is expressed as 'action'.

The definition of action, and every clause is doing work. Yavat siddham asiddham va: *whatever*, *accomplished* or *unaccomplished* — the distinction is set aside at once: what a thing *is* does not decide whether it is an action. Sadhyatvena abhidhiyate: is *expressed as what-is-to-be-accomplished* — sadhyatva, the to-be-accomplished-ness that 3.7.79 made the mark of actions alone. And the reason: asrita-krama-rupatvat — from its having a form in which *sequence is resorted to*: what makes something an action is that it is presented as sequenced. Tat kriya iti abhidhiyate: *that* is *expressed* as 'action'. So action is not a kind of thing but a mode of presentation: whatever is presented as to-be-accomplished-in-sequence is an action, whether or not it is done, whether or not it takes time. This is 3.1.35 exactly — Being with sequence is action, Being with sequence withdrawn is substance — now stated as the samuddesa's opening definition, with abhidhiyate twice: it is *expressed* so.

3.8.2 unformalized

कार्यकारणभावेन ध्वनतीत्याश्रितक्रमः ।
ध्वनिः क्रमनिवृत्तौ तु ध्वनिरित्येव कथ्यते ॥ २ ॥

kāryakāraṇabhāvena dhvanatītyāśritakramaḥ |
dhvaniḥ kramanivṛttau tu dhvanirityeva kathyate || 2 ||

'It sounds' — by way of cause and effect, sequence being resorted to; but when sequence has withdrawn, it is spoken of simply as 'sound'.

The definition demonstrated on one root. Dhvanati iti asrita-kramah: '*it sounds*' — sequence *resorted to*, karya-karana-bhavena, by way of *cause and effect*: the sounding presented as a process, one phase producing the next. Dhvanih krama-nivrttau tu dhvanih iti eva kathyate: but when *sequence has withdrawn*, it is spoken of simply as '*sound*'. Same phenomenon, two words: 'dhvanati' (verb) and 'dhvanih' (noun), and the difference between them is nothing but whether krama is in or out. There is no second fact. This is 3.1.35's doctrine given the cleanest possible evidence — a root and its derived noun, differing in nothing but sequence-presentation — and it explains why the noun/verb distinction, the deepest structural fact about the language, requires no ontological difference to support it.

3.8.3 unformalized

श्वेतते श्वेत इत्येतच्छ्वेतत्वेन प्रकाशते ।
आश्रितक्रमरूपत्वादभिधानं प्रवर्तते ॥ ३ ॥

śvetate śveta ityetacchvetatvena prakāśate |
āśritakramarūpatvādabhidhānaṃ pravartate || 3 ||

'It whitens', 'white' — this shines forth as whiteness; from its having a form in which sequence is resorted to, the expression proceeds.

The same demonstration on a quality, where it is stranger. Svetate svetah iti: '*it whitens*', '*white*' — a verb and an adjective from one stem. Etat svetatvena prakasate: *this* shines forth as *whiteness* — what both present is the quality. Asrita-krama-rupatvad abhidhanam pravartate: from its having a form in which *sequence is resorted to*, the *expression* proceeds. So even a static quality can be verbed: whiteness presented as sequenced becomes a whitening, and nothing in the whiteness changed. The example is better than 3.8.2's because sounding is *naturally* processual and whitening is not — a white thing is not doing anything. Yet Sanskrit verbs it, and the verb is not false. Which shows that krama is genuinely contributed by the presentation (2.25's mind's own sequencing power) rather than found in the phenomenon.

3.8.4 unformalized

गुणभूतैरवयवैः समूहः क्रमजन्मनाम् ।
बुद्ध्या प्रकल्पिताभेदः क्रियेति व्यपदिश्यते ॥ ४ ॥

guṇabhūtairavayavaiḥ samūhaḥ kramajanmanām |
buddhyā prakalpitābhedaḥ kriyeti vyapadiśyate || 4 ||

A collection of sequentially-born parts, which are subordinate — its non-difference constructed by the intellect — is designated as 'action'.

The definition's second half: what an action is made of, and who makes it one. Guna-bhutaih avayavaih samuhah krama-janmanam: a *collection* of *sequentially-born* parts which are *subordinate* — the sub-actions: the fetching, lighting, stirring that constitute a cooking (2.448's uncounted sub-actions displayed by a single verb). Buddhya prakalpita-abhedah: its *non-difference constructed by the intellect* — the collection's *unity* is buddhi's construction; the parts are given and the whole is made. Kriya iti vyapadisyate: is *designated* as 'action'. So an action is a mind-unified sequence of subordinate parts, and both halves matter: the parts are really sequential, and their being *one action* is a construal. This is 3.7.113's finding (the causal sequence arising from the grasping of non-difference) stated as the definition of what a kriya is — and it explains why 3.7.100's delimitation of the action's limit could be a variable: where one cuts the collection decides what the action is.

3.8.5 unformalized

समूहः स तथाभूतः प्रतिभेदं समूहिषु ।
समाप्यते ततो भेदे कालभेदस्य सम्भवः ॥ ५ ॥

samūhaḥ sa tathābhūtaḥ pratibhedaṃ samūhiṣu |
samāpyate tato bhede kālabhedasya sambhavaḥ || 5 ||

That collection, being such, is completed at each division among the members; therefore, in the difference, a difference of time is possible.

The action's whole-at-each structure. Samuhah sa tathabhutah pratibhedam samuhisu samapyate: that *collection*, being such, is *completed at each division* among the *members* — pratibhedam samapyate: 2.18's own formula, where the entire sentence-meaning completes at each word. The cooking is wholly present at the stirring; each sub-action is the whole action, not a fraction of it. Tatah bhede kala-bhedasya sambhavah: *therefore*, in the *difference*, a *difference of time* is *possible*. Because the whole completes at each member, and the members are sequential, the one action can bear temporal difference — it is at the fetching and at the stirring, and those are at different times. So an action's occupying a stretch of time is explained by whole-at-each rather than by part-summing: it does not take an hour because its parts take an hour; it is wholly present at each of a sequence of moments.

3.8.6 unformalized

क्रमात् सदसतां तेषामात्मानो न समूहिनाम् ।
सद्वस्तुविषयैर्यान्ति संबन्धं चक्षुरादिभिः ॥ ६ ॥

kramāt sadasatāṃ teṣāmātmāno na samūhinām |
sadvastuviṣayairyānti saṃbandhaṃ cakṣurādibhiḥ || 6 ||

The selves of those members, which are existent and non-existent by sequence, do not come into connection with eye and the rest, which have existent things for their domain.

Why actions are not perceived. Kramat sad-asatam tesam atmanah na samuhinam: the *selves* of those *members* — which are *existent and non-existent by sequence*: at any moment, the fetching is over and the stirring is not yet; only one phase is, and the rest are not (3.3.61's being and non-being as constructions of one self). Sad-vastu-visayaih yanti sambandham caksur-adibhih: they do not come into *connection* with *eye and the rest*, which have *existent things* for their *domain*. The senses take what is; an action is never all-there to be taken. So no action is ever perceived — one sees the stirring, never the cooking. This is a sharp result and it is the samuddesa's ground: if actions are not given to sense, their unity cannot be perceived either, and 3.8.4's buddhya prakalpita-abhedah is not a philosopher's flourish but the only available account.

3.8.7-8 unformalized

यथा गौरिति सङ्घातः सर्वो नेन्द्रियगोचरः ।
भागशस्तूपलब्धस्य बुद्धौ रूपं निरूप्यते ॥ ७ ॥
इन्द्रियैरन्यथा प्राप्तौ भेदांशोपनिपातिभिः ।
अलातचक्रवद्रूपं क्रियाणां परिकल्प्यते ॥ ८ ॥

yathā gauriti saṅghātaḥ sarvo nendriyagocaraḥ |
bhāgaśastūpalabdhasya buddhau rūpaṃ nirūpyate || 7 ||
indriyairanyathā prāptau bhedāṃśopanipātibhiḥ |
alātacakravadrūpaṃ kriyāṇāṃ parikalpyate || 8 ||

(7) As the whole aggregate 'cow' is not the range of the sense-faculty — the form of what is apprehended part by part is determined in the intellect; (8) so, the senses having otherwise attained it, falling upon portions of difference, the form of actions is constructed, like a firebrand-circle.

The parallel that makes 3.8.6 tolerable. (7): yatha gauh iti sanghatah sarvo na indriya-gocarah — as the *whole aggregate* 'cow' is *not* the *range of the sense-faculty*: nobody ever sees a whole cow (2.161's durlabham sarva-avayava-darsanam; the far side, the interior, are never given). Bhagasah tu upalabdhasya buddhau rupam nirupyate: the *form* of what is *apprehended part by part* is *determined in the intellect*. So the cow, too, is a buddhi-construction on partial data. (8) applies it: tatha alata-cakra-vad rupam kriyanam parikalpyate — so the *form of actions* is *constructed*, *like a firebrand-circle*. The whirling brand: no circle is there, and every eye sees one (1.129's emblem, 2.291's touch-test that breaks it). Actions are firebrand-circles: successive phases, no whole, and an irresistible appearance of one. And the parallel with the cow disarms the objection — if constructedness disqualified actions, it would disqualify cows.

3.8.9 unformalized

यथा च भागाः पचतैरुदकासेचनादयः ।
उदकासेचनादीनां ज्ञेया भागास्तथापरे ॥ ९ ॥

yathā ca bhāgāḥ pacatairudakāsecanādayaḥ |
udakāsecanādīnāṃ jñeyā bhāgāstathāpare || 9 ||

And as the parts of 'paci' are the pouring of water and the rest — so the other parts of the pouring of water and the rest are to be known.

The regress of sub-actions, and it has no floor. Yatha ca bhagah pacateh udaka-asecanadayah: as the *parts* of 'paci' (cooking) are the *pouring of water* and the rest — the cooking analyses into fetching, pouring, lighting, stirring. Udaka-asecanadinam jneyah bhagah tatha apare: so the *other parts* of the *pouring of water* and the rest are *to be known* — and the pouring analyses into lifting, tilting, aiming; and each of those analyses further. Tatha apare: likewise others, and the phrase does not stop. So the sub-actions of 3.8.4 are themselves collections whose unity is constructed, all the way down — and there is no level of simple actions from which the complex ones are built. This is 2.236's atom given parts (the mind that cannot stop dividing) at the level of events, and it means that 3.8.4's definition applies at every grain: an action is a construed unity of construed unities, indefinitely.

3.8.10 unformalized

यश्चापकर्षपर्यन्तमनप्राप्तः प्रतीयते ।
तत्रैकस्मिन् क्रियाशब्दः केवले न प्रयुज्यते ॥ १० ॥

yaścāpakarṣaparyantamanaprāptaḥ pratīyate |
tatraikasmin kriyāśabdaḥ kevale na prayujyate || 10 ||

And that which, not having reached the limit of subtraction, is understood — there, in that single one, the action-word is not employed when it is alone.

Where the regress of 3.8.9 stops for practical purposes. Yah ca apakarsa-paryantam anaprapta pratiyate: that which, *not having reached the limit of subtraction*, is understood — a sub-action taken short of full analysis. Tatra ekasmin kriya-sabdah kevale na prayujyate: *there*, in that *single one*, the *action-word* is *not employed* *when it is alone*. One does not call a single tilt of a jug an action; the verb needs a collection. So although the analysis descends without limit (3.8.9), the *word* 'action' attaches only where a collection has been construed — and an isolated minimum has no collection and takes no verb. This is the criterion the regress needed: actions are where verbs go, verbs go to construed collections, and the indefinitely-divisible parts are not themselves actions until a mind gathers them.

3.8.11 unformalized

पूर्वोत्तरैस्तदा भागैः समवस्थापितक्रमः ।
एकः सोऽप्यसदध्यासादाख्यातैरभिधीयते ॥ ११ ॥

pūrvottaraistadā bhāgaiḥ samavasthāpitakramaḥ |
ekaḥ so'pyasadadhyāsādākhyātairabhidhīyate || 11 ||

Then, its sequence established by the prior and subsequent parts, even that single one is expressed by finite verbs, through a superimposition of what is not there.

How the minimum of 3.8.10 nonetheless gets verbed. Purva-uttaraih tada bhagaih samavasthapita-kramah: *then*, its *sequence established* by the *prior and subsequent* parts — the single tilt, set among its neighbours, acquires a before and after from them. Ekah so 'pi asad-adhyasat akhyataih abhidhiyate: *even that single one* is *expressed by finite verbs*, through a *superimposition of what is not there* — asad-adhyasa: the superimposition of the non-existent (3.1.6's adhyaropa; 3.3.53's existence superimposed by designation). The isolated part has no internal sequence; sequence is laid onto it from its neighbours; and then it can take a verb. So even the minimal action is verbed by a fiction — and the fiction is explicitly named asat, not there. The samuddesa's method is on display: find where the definition fails, and show that the grammar proceeds there by a superimposition it can name.

3.8.12 unformalized

कालानुपाति यद्रूपं तदस्तीति प्रतीयते ।
परितस्तु परिच्छिन्नं भाव इत्येव कथ्यते ॥ १२ ॥

kālānupāti yadrūpaṃ tadastīti pratīyate |
paritastu paricchinnaṃ bhāva ityeva kathyate || 12 ||

The form that follows upon time is understood as 'it exists'; but what is delimited on all sides is spoken of simply as 'an entity'.

The two great categories distinguished by their relation to time. Kala-anupati yad rupam tad asti iti pratiyate: the form that *follows upon time* is understood as '*it exists*' — the verbal, the processual, presented as running along time's course. Paritas tu paricchinnam bhava iti eva kathyate: but what is *delimited on all sides* — bounded, closed off, complete — is spoken of *simply as 'an entity'*. So verbs present what runs with time; nouns present what is fenced. And 'asti' is on the verbal side: existence itself is a following-upon-time (3.3.47's self-bearing 'asti'), while an entity is a delimitation. This refines 3.1.35 with a spatial metaphor: substance is Being *fenced*, action is Being *running* — and paritah paricchinnam, delimited on all sides, is what a noun does to what it names.

3.8.13 unformalized

व्यवहारस्य सिद्धत्वान्न चेयं गुणकल्पना ।
उपचारो हि मुख्यस्य सम्भवादवतिष्ठते ॥ १३ ॥

vyavahārasya siddhatvānna ceyaṃ guṇakalpanā |
upacāro hi mukhyasya sambhavādavatiṣṭhate || 13 ||

And this is not a construction of the subordinate, since the practice is accomplished; for a figurative usage stands only from the possibility of a primary one.

A guard against deflating the whole account. Vyavaharasya siddhatvat na ca iyam guna-kalpana: since the *practice is accomplished* — people do use verbs and nouns, and the usage works — *this* is *not a construction of the subordinate*: the action-presentation is not a merely secondary or figurative device. And the principle: upacaro hi mukhyasya sambhavad avatisthate — for a *figurative usage* stands only *from the possibility of a primary* one. Metaphor presupposes literal; the secondary requires a primary to be secondary *to* (2.265's rival criterion, and 2.253's currency-doctrine). So if action-talk were wholly figurative, there would be nothing for it to be figurative from, and it could not work. It works; therefore it is primary. The verse protects the samuddesa's constructivism from collapsing into the claim that verbs say nothing: the construction is real, the usage is primary, and 'constructed' does not mean 'figurative'.

3.8.14 unformalized

आहितोत्तरशक्तित्वात् प्रत्येकं वा समूहिनः ।
अनेकरूपा लभ्यन्ते क्रमवन्तः इवाक्रमाः ॥ १४ ॥

āhitottaraśaktitvāt pratyekaṃ vā samūhinaḥ |
anekarūpā labhyante kramavantaḥ ivākramāḥ || 14 ||

Or, from the fact that a subsequent capacity has been deposited, the members of the collection are severally obtained as multiform — sequenceless, as though possessing sequence.

An alternative account of how the parts bear the whole. Ahita-uttara-saktitvat: from the fact that a *subsequent capacity has been deposited* — each part carries a capacity toward what follows: the fetching is already oriented to the pouring, the pouring to the boiling (1.84's ahita-bija: the seed deposited by the sounds; 3.7.33's current drawing onward). Pratyekam va samuhinah aneka-rupah labhyante: the *members* are *severally obtained as multiform* — each part, carrying its onward capacity, is many-formed. Kramavantah iva akramah: *sequenceless*, *as though possessing sequence*. And there it is: the parts are *akrama*, sequenceless in themselves, and appear sequential because each carries a deposited capacity toward the next. So sequence is not a relation between the parts; it is a capacity *in* each, and the appearance of order is the capacities showing. This is 1.48's structure (the sequenceless born as if in sequence) applied to the constituents of an action.

3.8.15 unformalized

अनन्तरं फलं यस्याः कल्पते तां क्रियां विदुः ।
प्रधानभूतां तादर्थ्याद् अन्यासां तु तदाख्यता ॥ १५ ॥

anantaraṃ phalaṃ yasyāḥ kalpate tāṃ kriyāṃ viduḥ |
pradhānabhūtāṃ tādarthyād anyāsāṃ tu tadākhyatā || 15 ||

That action whose fruit is construed as immediately following, they know as the principal; but the others, from their being for that purpose, have its designation.

Which of the collection's members is the action. Anantaram phalam yasyah kalpyate tam kriyam viduh pradhana-bhutam: that action whose *fruit* is construed as *immediately following* — they know as the *principal*: the member after which the result arrives (3.7.90's definition of the instrument, transposed: what the accomplishment immediately follows). Anyasam tu tadarthyat tad-akhyata: but the *others*, from their *being for that purpose*, have *its designation* — the fetching and pouring are called 'cooking' because they are for the cooking. So one member is the action proper and the rest borrow its name by serving it. Tadarthya — being-for-that (3.7.27's dative of purpose). And note the verb: kalpyate, is *construed* — which member's fruit counts as immediate is a construal (3.7.100's delimitation of the action's limit), and the whole naming-hierarchy follows from it.

3.8.16 unformalized

सत्सु प्रत्ययरूपोऽसौ भावो यावन्न जायते ।
तावत् परेषां रूपेण साध्यः सन्निभिधीयते ॥ १६ ॥

satsu pratyayarūpo'sau bhāvo yāvanna jāyate |
tāvat pareṣāṃ rūpeṇa sādhyaḥ sannibhidhīyate || 16 ||

That entity, whose form is a cognition among existent things — so long as it is not born, so long is it expressed as to-be-accomplished, by way of the form of the others.

Why the unfinished is sayable. Satsu pratyaya-rupo 'sau bhavah: that *entity*, whose *form is a cognition* among *existent* things — the action exists as a cognition, not as a thing among things (3.8.6's members never reaching the senses). Yavan na jayate: *so long as it is not born* — while the action is incomplete, still running. Tavat paresam rupena sadhyah sann abhidhiyate: *so long* is it expressed as *to-be-accomplished*, by way of the *form of the others* — para-rupena: through the form of what is not it, its parts and neighbours (3.5.9's arupam para-rupena dravyam akhyayate: the formless substance designated by another's form). So the unfinished action has no form of its own to be expressed by, and is expressed by its parts' forms, and is thereby sayable while unaccomplished. This is the sadhyatva of 3.8.1 given its mechanism: to be to-be-accomplished is to be presented through what is not yet you.

3.8.17 unformalized

सिद्धे तु साधनाकाङ्क्षा कृतार्थत्वान्निवर्तते ।
न क्रियावाचिनां तस्मात् प्रयोगस्तत्र विद्यते ॥ १७ ॥

siddhe tu sādhanākāṅkṣā kṛtārthatvānnivartate |
na kriyāvācināṃ tasmāt prayogastatra vidyate || 17 ||

But when it is accomplished, the expectancy of the means withdraws, from its purpose having been served; therefore, in that case, there is no employment of action-expressing words.

Why the finished cannot be verbed. Siddhe tu sadhana-akanksa krtarthatvat nivartate: when it is *accomplished*, the *expectancy of the means* *withdraws*, from its *purpose having been served* — krtartha: the word 3.7.83 used for the animal that served the rite. A completed action wants nothing more; its karakas have discharged. Na kriya-vacinam tasmat prayogah tatra vidyate: *therefore*, in that case, there is *no employment* of *action-expressing* words. The finished action takes no verb — one says 'the cooked rice', a noun, not 'cooks'. So sadhyatva (3.8.1) and akanksa are the same condition seen twice: an action is verbal exactly while it still requires, and requiring ends with accomplishment. And the pair 3.8.16-17 delimits the verb's whole domain: it operates between beginning and completion, and at neither end.

3.8.18 unformalized

स च पूर्वापरीभूत एकत्वादक्रमात्मकः ।
पूर्वापराणां धर्मेण तदर्थेनानुगम्यते ॥ १८ ॥

sa ca pūrvāparībhūta ekatvādakramātmakaḥ |
pūrvāparāṇāṃ dharmeṇa tadarthenānugamyate || 18 ||

And that, being prior-and-posterior, is by its oneness sequenceless; it is followed along by the property of the prior and posterior parts, and by their purpose.

The action's paradoxical structure, stated flatly. Sa ca purva-apari-bhutah: and *that* — the action — being *prior-and-posterior*: it has a before and an after; it is spread. Ekatvad akrama-atmakah: by its *oneness*, *sequenceless* — and being one, it has no sequence, since a unity has no internal order. Both, at once. And the resolution: purva-aparanam dharmena tad-arthena anugamyate — it is *followed along* by the *property* of the prior and posterior *parts*, and by their *purpose*. The sequence belongs to the parts and is *followed along* into the whole; the whole's own character is sequenceless. So the action is one (hence akrama) and its parts are ordered (hence purva-apara), and the order is attributed to the one by transfer. This is 1.48-49's structure exactly — the sequenceless born as if divided, the reflection following the water's movement — now at the level of what a verb means.

3.8.19 unformalized

असन्निरवर्तते तस्माद् यत् सत् तदुपलभ्यते ।
तयोः सदसतोश्चासावात्मैक इव गृह्यते ॥ १९ ॥

asanniravartate tasmād yat sat tadupalabhyate |
tayoḥ sadasatoścāsāvātmaika iva gṛhyate || 19 ||

The non-existent withdraws; therefore what is existent is apprehended; and that self of those two, the existent and the non-existent, is grasped as though one.

How the action's shifting phases present as one thing. Asan nivartate: the *non-existent* *withdraws* — the future phases are not yet, and drop away as they arrive. Tasmat yat sat tad upalabhyate: *therefore* what is *existent* is *apprehended* — only the present phase is there to be had (3.8.6's members, existent and non-existent by sequence). Tayoh sad-asatoh ca asau atma eka iva grhyate: and that *self* of those two — the *existent* and the *non-existent* — is *grasped as though one*. The action is a shifting boundary between what has been and what will be, and the mind takes the whole shifting thing as one self. Iva: as though. This is 3.3.61's doctrine (being and non-being as constructions, not other than one self) applied to the temporal structure of a single act — and it explains 3.8.4's buddhya prakalpita-abhedah: the unity is constructed *across* an existent/non-existent divide, which is why no perception could deliver it.

3.8.20 unformalized

जातिमन्ये क्रियामाहुरनेकव्यक्तिवर्तिनीम् ।
असाध्या व्यक्तिरूपेण सा साध्येवोपलभ्यते ॥ २० ॥

jātimanye kriyāmāhuranekavyaktivartinīm |
asādhyā vyaktirūpeṇa sā sādhyevopalabhyate || 20 ||

Others call action a universal, residing in many individuals; being unaccomplishable in the form of the individual, it is apprehended as though to-be-accomplished.

The first rival theory: action as universal. Jatim anye kriyam ahuh aneka-vyakti-vartinim: *others* call *action* a *universal*, residing in *many individuals* — cooking-hood, present in every cooking as cowhood in cows (3.1.25-26's jati as self-manifesting power). Asadhya vyakti-rupena: *unaccomplishable* in the *form of the individual* — a universal cannot be accomplished; it is eternal, and no cooking brings cooking-hood about. Sa sadhya iva upalabhyate: it is *apprehended as though to-be-accomplished*. So on this view the verb's sadhyatva (3.8.1) is an appearance: what the verb presents as being brought about is a universal that cannot be brought about, and the iva carries the whole theory. The position is real and it has an advantage — it explains the action's unity (3.8.4) without buddhi, by making it a jati — at the cost of making every verb systematically misleading.

3.8.21 unformalized

अन्ते या वा क्रियाभागे जातिः सैव क्रिया स्मृता ।
सा व्यक्तेरनुनिष्पादे जायमानेव गम्यते ॥ २१ ॥

ante yā vā kriyābhāge jātiḥ saiva kriyā smṛtā |
sā vyakteranuniṣpāde jāyamāneva gamyate || 21 ||

Or the universal which is in the final action-part — that very one is remembered as the action; it is understood as though being born, in the individual's arising-in-consequence.

The universal-theory refined. Ante ya va kriya-bhage jatih: or the *universal* which is in the *final action-part* — not cooking-hood in general but the universal residing in the last phase (3.8.15's principal member, after which the fruit arrives). Sa eva kriya smrta: *that very one* is *remembered* as the *action*. Sa vyakteh anunispade jayamana iva gamyate: it is understood *as though being born*, in the individual's *arising-in-consequence* — anunispada: arising-along-with (2.121's anunispadin, the entailed). So when the final phase occurs, its universal is manifested, and the manifesting presents as a birth. The refinement locates the action precisely (in the last part, per 3.8.15) and keeps the universal's eternality (it is manifested, not born) — and again the iva does the work: jayamana iva, as though being born. Two iva's in two verses is the theory's signature: it must make the verb's entire content an appearance.

3.8.22 unformalized

स्वव्यापारविशिष्टानां सत्ता वा कर्तृकर्मणाम् ।
क्रिया व्यापारभेदेषु सत्ता वा समवायिनी ॥ २२ ॥

svavyāpāraviśiṣṭānāṃ sattā vā kartṛkarmaṇām |
kriyā vyāpārabhedeṣu sattā vā samavāyinī || 22 ||

Or action is the existence of agent and object as qualified by their own operations; or it is the inherent existence in the differences of operation.

The second rival: action as existence. Sva-vyapara-visistanam satta va kartr-karmanam: or the *existence* of *agent and object* as *qualified by their own operations* — the cooking is just the cook-and-rice existing cook-wise; no further entity. Kriya vyapara-bhedesu satta va samavayini: or it is the *inherent existence* in the *differences of operation* — the satta inhering in the operations themselves. Both versions make action a mode of satta rather than a thing, and both are the Padakanda's own doctrine wearing another school's clothes: 3.1.33 made satta the sole universal, 3.1.35 made action Being-with-sequence. So this 'rival' is very close to home, and its two versions differ over where the existence sits — in the qualified relata, or in the operations. Bhartrhari reports without deciding, and the reader who has followed from 3.1.33 knows why: on his own account both are satta carved by relata.

3.8.23 unformalized

अन्त्ये वात्मनि या सत्ता सा क्रिया कैश्चिदिष्यते ।
भाव एव हि धात्वर्थ इत्यवच्छिन्न आगमः ॥ २३ ॥

antye vātmani yā sattā sā kriyā kaiścidiṣyate |
bhāva eva hi dhātvartha ityavacchinna āgamaḥ || 23 ||

Or the existence which is in the final self — that is held by some to be the action; for 'the root's meaning is bhava alone' is the received tradition, unbroken.

The third rival, and it cites scripture. Antye va atmani ya satta sa kriya kaiscid isyate: the *existence* which is in the *final self* is held by some to be the *action* — combining 3.8.21's location (the final part) with 3.8.22's category (satta). And then the authority: bhava eva hi dhatv-arthah iti avacchinna agamah — for '*the root's meaning is bhava alone*' is the *received tradition*, *unbroken*. This is the grammarians' own maxim — bhava-pradhanam akhyatam, the verb has being-as-its-principal (Yaska's Nirukta 1.1, and Panini's own bhave affixes) — and avacchinna, unbroken, is the tradition's mark of authority (1.31's avyavacchinnah panthanah; 1.41's avicchedena). So the third theory is not a speculation but the received doctrine: what a root means is bhava, existence-in-process. And the Padakanda has been building toward it since 3.1.34 (satta as the meaning of both stem and root).

3.8.24 unformalized

बुद्धिं तज्जातिमन्ये तु बुद्धिसत्तामथापरे ।
प्रत्यस्तरूपां भावेषु क्रियेति प्रतिजानते ॥ २४ ॥

buddhiṃ tajjātimanye tu buddhisattāmathāpare |
pratyastarūpāṃ bhāveṣu kriyeti pratijānate || 24 ||

But others hold the intellect, or its universal; and yet others hold the intellect's existence, whose form is dissolved away in things, to be action.

The remaining theories, and they are idealist. Buddhim taj-jatim anye: *others* hold the *intellect* — action is a cognition, nothing more (3.8.16's pratyaya-rupo bhavah) — or *its universal*: cognition-hood (3.1.102's universals among cognitions). Buddhi-sattam atha apare: and *yet others* the *intellect's existence*. Praty-asta-rupam bhavesu kriya iti pratijanate: *whose form is dissolved away* in things — pratyasta-rupa: 3.6.7's phrase for direction, whose own form is effaced in what it orders — they *maintain* to be *action*. So the last theories make action a mode of mind, effaced into the things it construes. And with 3.8.20-24 the catalogue is complete: action as universal, as existence of the relata, as inherent existence, as final existence, as intellect, as intellect's universal, as intellect's existence. Seven positions, canvassed without decision — and every one of them is a version of something the Padakanda has already argued in its own voice.

3.8.25 unformalized

आविर्भावतिरोभावौ जन्मनाशौ तथापरैः ।
षट्सु भावविकारेषु कल्पितौ व्यावहारिकौ ॥ २५ ॥

āvirbhāvatirobhāvau janmanāśau tathāparaiḥ |
ṣaṭsu bhāvavikāreṣu kalpitau vyāvahārikau || 25 ||

By others, manifestation and disappearance, and birth and destruction, are constructed as practical, among the six modifications of being.

The six vikaras revisited (1.3, 3.1.36's sad-avasthah), and a reduction begun. Avirbhava-tirobhavau janma-nasau tatha aparaih: *manifestation* and *disappearance*, and *birth* and *destruction* — two pairs, and the first pair is the satkarya-vadin's (things manifest and hide; nothing is made) while the second is the asatkarya-vadin's (things are born and perish). Sat su bhava-vikaresu kalpitau vyavaharikau: are *constructed* as *practical*, among the *six modifications of being*. Kalpitau: constructed; vyavaharikau: practical, belonging to transaction (2.297's sabdah loka-nibandhanah). So both pairs are practical constructions on the six — which means the great dispute between satkarya and asatkarya (3.7.49's two objects) is again a difference of construal, and both are vyavaharika. The verse sets up the reduction of 3.8.26-27.

3.8.26 unformalized

ताभ्यां सर्वप्रवृत्तीनामभेदेनोपसंग्रहः ।
जन्मैवाश्रितसारूप्यं स्थितिरित्यभिधीयते ॥ २६ ॥

tābhyāṃ sarvapravṛttīnāmabhedenopasaṃgrahaḥ |
janmaivāśritasārūpyaṃ sthitirityabhidhīyate || 26 ||

By those two, all operations are comprehended without difference; birth alone, having resorted to a likeness of form, is spoken of as 'standing'.

The reduction's first step. Tabhyam sarva-pravrttinam abhedena upasangrahah: by *those two* — birth and destruction — *all operations* are *comprehended without difference*: every one of the six vikaras (born, exists, changes, grows, declines, perishes) reduces to a birth or a destruction. Six to two. And then: janma eva asrita-sarupyam sthitih iti abhidhiyate — *birth alone*, having *resorted to a likeness of form*, is spoken of as '*standing*'. Existence is birth continuing to look like itself: what we call a thing's *standing* is its being born again in the same shape (the sarupya of successive moments). So even sthiti, the second of the six and the most static, is a birth. Two to one, nearly — and 3.8.27 completes it.

3.8.27 unformalized

जायमानान्न जन्मान्याद् विनाशेऽप्यपदार्थता ।
अतो भावविकारेषु सत्तैका व्यवतिष्ठते ॥ २७ ॥

jāyamānānna janmānyād vināśe'pyapadārthatā |
ato bhāvavikāreṣu sattaikā vyavatiṣṭhate || 27 ||

Birth is not other than what is being born; and in destruction too there is non-thinghood. Therefore, among the modifications of being, one Existence stands established.

The reduction completed, and it lands on the Padakanda's central doctrine. Jayamanan na janma anyat: *birth* is *not other than* *what is being born* — there is no birth over and above the thing arising; the two are one (3.7.107's coiling snake: the coil is not a further item). Vinase 'pi apadarthata: and in *destruction* too there is *non-thinghood* — destruction is not a thing either; a nothing does not become an entity by being an ending (3.3.76's regress: reify absence and the examination has no floor). So birth is nothing over its thing, and destruction is nothing at all. Atah bhava-vikaresu satta eka vyavatisthate: *therefore*, among the *modifications of being*, *one Existence* stands *established*. The six vikaras reduce to two, the two dissolve, and what remains is satta — 3.1.33's sole universal, in which all words stand established; 3.1.36's Being attaining the six states by its own powers. The Kriya-samuddesa has reached the Jati-samuddesa's conclusion by the road of the verb.

3.8.28 unformalized

पूर्वभागस्तु यज्जातात् तज्जन्म व्यपदिश्यते ।
आश्रितक्रमरूपेण निमित्तत्वे विवक्षिते ॥ २८ ॥

pūrvabhāgastu yajjātāt tajjanma vyapadiśyate |
āśritakramarūpeṇa nimittatve vivakṣite || 28 ||

But that prior part of what has been born — that is designated as 'birth', when its being an occasioning ground is intended, by way of a form in which sequence is resorted to.

Birth relocated, having been dissolved. Purva-bhagas tu yaj jatat tad janma vyapadisyate: that *prior part* of *what has been born* — *that* is *designated* as '*birth*': birth is not an event added to the thing; it is the thing's own earlier phase, named. Nimittatve vivaksite: when its *being an occasioning ground* is *intended* — the phase is called a birth when we mean it as what the later phase came from. Asrita-krama-rupena: by way of a *form in which sequence is resorted to* — 3.8.1's definition verbatim. So birth is a *part* of the thing, construed as ground, presented sequence-wise. Everything the samuddesa has built converges: 3.8.27 dissolved birth into satta, and 3.8.28 explains what we are doing when we nonetheless say 'it is born' — naming a phase as an origin, under a sequence-presentation. The dissolution and the usage coexist (2.238's untrue path).

3.8.29 unformalized

आख्यातशब्दैरर्थोऽसावेवंभूतोऽभिधीयते ।
नामशब्दाः प्रवर्तन्ते संहरन्त इति क्रमम् ॥ २९ ॥

ākhyātaśabdairartho'sāvevaṃbhūto'bhidhīyate |
nāmaśabdāḥ pravartante saṃharanta iti kramam || 29 ||

That meaning, being of such a kind, is expressed by finite-verb words; noun-words proceed as withdrawing the sequence.

The samuddesa's thesis, restated as its two-word conclusion. Akhyata-sabdaih arthah asau evam-bhutah abhidhiyate: *that meaning*, being *of such a kind* — the phase-construed-as-ground, sequence-presented (3.8.28) — is *expressed by finite-verb words*. Nama-sabdah pravartante samharantah iti kramam: *noun-words* proceed as *withdrawing* the *sequence*. Samharantah: gathering in, drawing back, withdrawing — the verb 3.1.35 used (krama-rupasya samhare tat sattvam: in the withdrawal of the form of sequence, that is substance). So the entire noun/verb distinction is one operation and its reverse: verbs deploy sequence, nouns retract it, and the same satta is meant by both. This is the Padakanda's most economical statement of what its two great categories are, and it has taken the Kriya-samuddesa twenty-nine verses to earn what 3.1.35 asserted.

3.8.30 unformalized

फलं फलापदेशो वा वस्तु वा तद्विरोधि यत् ।
तदन्यदेव पूर्वेषां नाश इत्यपदिश्यते ॥ ३० ॥

phalaṃ phalāpadeśo vā vastu vā tadvirodhi yat |
tadanyadeva pūrveṣāṃ nāśa ityapadiśyate || 30 ||

The fruit, or the designation of the fruit, or whatever thing is opposed to that — that quite other thing is designated as the 'destruction' of the prior ones.

Destruction relocated, as birth was at 3.8.28. Phalam phala-apadeso va: the *fruit*, or the *designation of the fruit* — the outcome, or its name. Vastu va tad-virodhi yat: or whatever *thing* is *opposed to* that. Tad anyad eva purvesam nasah iti apadisyate: *that quite other thing* is *designated* as the *'destruction'* of the *prior* ones. So destruction is not an event that happens to what perishes; it is the *arrival of something else* — the ash, the fruit, the successor — named as the predecessor's end. The wood's destruction is the ash's presence (3.7.50's prakrty-uccheda-sambhutam). Anyad eva: *quite other*: nothing happened to the wood at all; something else is here now, and we call that the wood's perishing. This completes 3.8.27's apadarthata (destruction's non-thinghood) with an account of what we are naming when we nonetheless speak of it.

3.8.31 unformalized

नैवास्ति नैव नास्तीति वस्तुनो ग्रहणाद् विना ।
कल्पते पररूपेण वस्त्वन्यदनुगम्यते ॥ ३१ ॥

naivāsti naiva nāstīti vastuno grahaṇād vinā |
kalpate pararūpeṇa vastvanyadanugamyate || 31 ||

'It neither is nor is not' — apart from the apprehension of a thing, it is constructed by another's form; another thing is followed along.

The samuddesa's tetralemmic moment, and it is brief. Na eva asti na eva na asti iti: '*it neither is nor is not*' — the third and fourth corners, applied to the perishing thing of 3.8.30: the wood, at the moment of its 'destruction', neither is (it is gone) nor is not (something is there). Vastuno grahanad vina: *apart from* the *apprehension of a thing* — no thing is grasped there; nothing is available to bear either predicate (3.3.79's non-thing having neither priority nor state). Kalpyate para-rupena: it is *constructed by another's form* — the perishing is construed through the successor's form (3.8.30's anyad eva; 3.5.9's arupam para-rupena). Vastv anyad anugamyate: *another thing* is *followed along*. So the perished thing is neither existent nor non-existent, and what we track when we speak of it is its successor. This is 3.2.12's tetralemma in the ordinary case of something burning down.

3.8.32 unformalized

आवाभावौ घटादीनामस्पृशन्नपि पाणिना ।
कश्चिद्वेदाप्रकाशेऽपि प्रकाशे तत एव वा ॥ ३२ ॥

āvābhāvau ghaṭādīnāmaspṛśannapi pāṇinā |
kaścidvedāprakāśe'pi prakāśe tata eva vā || 32 ||

Someone knows the presence and absence of pots and the like without even touching them with the hand — whether in the absence of light, or from that very light.

A datum about knowing that resists the ordinary account. Ava-bhavau ghatadinam: the *presence and absence* of pots and the like. Asprsann api panina: without even *touching* them *with the hand* — no contact, and yet the presence and absence are known. Kascid veda aprakase 'pi: *someone knows*, *even in the absence of light* — the dark room's furniture, known by no sense one can name; prakase tatah eva va: *or from that very light* — the ordinary case. So knowing presence and absence does not require contact and does not always require illumination, and the samuddesa records this without explaining it. The verse opens a short movement on capacities that operate without their expected conditions — preparing 3.8.33's subtlety and 3.8.34's alternative account of what an action is.

3.8.33 unformalized

व्यापि सौक्ष्म्यं क्वचिद् याति क्वचित् संहन्यते पुनः ।
अकुर्वाणोऽथवा किञ्चित् स्वशक्त्यैवं प्रकाशते ॥ ३३ ॥

vyāpi saukṣmyaṃ kvacid yāti kvacit saṃhanyate punaḥ |
akurvāṇo'thavā kiñcit svaśaktyaivaṃ prakāśate || 33 ||

Pervading, it goes somewhere into subtlety; somewhere it is condensed again; or, doing nothing at all, it shines forth thus by its own capacity.

The three modes of a capacity's operation. Vyapi saukymam kvacid yati: *pervading*, it goes *somewhere into subtlety* — the power thins, spreads, becomes imperceptible. Kvacit samhanyate punah: *somewhere* it is *condensed* again — and thickens, gathers (samhan: the verb of gathering, cognate with 3.8.29's samharantah). Akurvano 'thava kimcit sva-saktya evam prakasate: *or*, *doing nothing at all*, it *shines forth* thus *by its own capacity*. Three modes: thinning, thickening, and simply showing without acting. The third is the striking one — a capacity that manifests without operating (3.7.28's sahaja sakti standing as long as the substance, its assisting only occasional). So the capacities that 3.8.32's knower detects are these: pervading subtly, condensed, or merely shining. And none of the three is an action on the pot.

3.8.34 unformalized

सर्वरूपस्य तत्त्वस्य यत् क्रमेणेव दर्शनम् ।
भागैरिव प्रक्लृप्तिश्च तां क्रियामपरे विदुः ॥ ३४ ॥

sarvarūpasya tattvasya yat krameṇeva darśanam |
bhāgairiva praklṛptiśca tāṃ kriyāmapare viduḥ || 34 ||

That seeing of the all-formed reality which is, as it were, in sequence — and the construction, as it were, by parts — others know as the action.

Another account of action, and it is thoroughly idealist. Sarva-rupasya tattvasya yat kramena iva darsanam: that *seeing* of the *all-formed reality* which is, *as it were*, *in sequence* — the one reality, omniform (3.3.56's sarvartha-rupata: the purity of cognition being its having all objects for its form), seen as-though-sequentially. Bhagaih iva praklrptih ca: and the *construction*, *as it were*, *by parts*. Tam kriyam apare viduh: *that* others know as the *action*. So action is the sequential-seeming seeing of what is neither sequential nor parted — two iva's in a single verse, and both mark the appearance. This is 3.8.24's idealist entries developed: action is not in the world and not even in the mind as a thing; it is the *mode* in which the mind sees an omniform reality. And 3.8.35 immediately gives that reality its name.

3.8.35 unformalized

सत्ता स्वशक्तियोगेन सर्वरूपा व्यवस्थिता ।
साध्या च साधनं चैव फलं भोक्ता फलस्य च ॥ ३५ ॥

sattā svaśaktiyogena sarvarūpā vyavasthitā |
sādhyā ca sādhanaṃ caiva phalaṃ bhoktā phalasya ca || 35 ||

Existence, by connection with its own capacities, stands established as all-formed: as what is to be accomplished, as the means, as the fruit, and as the enjoyer of the fruit.

The Padakanda's master-doctrine, stated in the Kriya-samuddesa's own terms. Satta sva-sakti-yogena sarva-rupa vyavasthita: *Existence*, by connection with its *own capacities*, stands established as *all-formed* (3.1.33's satta carved by relata; 3.1.22's sarva-sakty-atma-bhutatvam). And the four: sadhya ca — as *what is to be accomplished* (the action); sadhanam ca eva — as the *means* (the karakas); phalam — the *fruit*; bhokta phalasya ca — and the *enjoyer of the fruit*. So the whole structure of purposive action — the doing, the instruments, the outcome, the one who gets it — is satta wearing four faces. Nothing in the list is other than the others. This is 3.2.14's fourfold (that is the seen, the seeing, the seer, and the purpose in the seeing) transposed from cognition to action, and it is the same one differentiating itself into the whole apparatus by which anything gets done.

3.8.36 unformalized

क्रियामन्ये तु मन्यन्ते क्वचिदप्यनपाश्रिताम् ।
साधनैकार्थकारित्वे प्रवृत्तिमनपायिनीम् ॥ ३६ ॥

kriyāmanye tu manyante kvacidapyanapāśritām |
sādhanaikārthakāritve pravṛttimanapāyinīm || 36 ||

But others consider action to be an operation not resorting to anything anywhere, and unfailing, in the means' effecting of a single purpose.

The process-account, offered as the alternative to 3.8.35. Kriyam anye tu manyante kvacid api anapasritam: *others* consider *action* to be an operation *not resorting to anything anywhere* — anapasrita: leaning on nothing, resting on no substrate; the operation is primary and free-standing (3.7.33's pravrttir eva prathamam kvacid api anapasrita — the operation itself first, resorting to nothing, drawing like a current). Sadhana-eka-artha-karitve pravrttim anapayinim: an operation *unfailing* — anapayin, not lapsing — in the *means'* *effecting of a single purpose*. So action is the pure running-toward-one-end, needing no bearer and never failing while it runs. Against 3.8.35's satta-account (action as one of Being's four faces), this makes action fundamental and Being derivative. It is the boldest of the seven positions and it is stated in three words: kvacid api anapasritam.

3.8.37 unformalized

सामान्यभूता सा पूर्वं भागशः प्रविभज्यते ।
ततो व्यापाररूपेण साध्येव व्यवतिष्ठते ॥ ३७ ॥

sāmānyabhūtā sā pūrvaṃ bhāgaśaḥ pravibhajyate |
tato vyāpārarūpeṇa sādhyeva vyavatiṣṭhate || 37 ||

That, being general at first, is divided out part by part; thereafter, in the form of an operation, it stands as though to-be-accomplished.

The process-account's genesis. Samanya-bhuta sa purvam: *that* — the operation of 3.8.36 — being *general at first*: undifferentiated running, no parts, no direction. Bhagasah pravibhajyate: is *divided out part by part* — the analysis carves it. Tatah vyapara-rupena sadhya iva vyavatisthate: *thereafter*, in the *form of an operation*, it stands *as though to-be-accomplished*. So the sequence is: undivided operation, then carving, then the appearance of sadhyatva. And the iva again: sadhya *iva* — the to-be-accomplished-ness that 3.8.1 made the very definition of action is here an appearance, arriving only after the carving. Every account the samuddesa canvasses converges on this: whatever action is, its being *to-be-done* is not in it (3.8.20's asadhya, 3.8.21's jayamana iva) but in how it is taken.

3.8.38 unformalized

प्रकृति साधनानां सा प्रथमं तच्च कारकम् ।
व्यापाराणां ततोऽन्यत्लमपरैरुपवर्ण्यते ॥ ३८ ॥

prakṛti sādhanānāṃ sā prathamaṃ tacca kārakam |
vyāpārāṇāṃ tato'nyatlamaparairupavarṇyate || 38 ||

That is the material cause of the means; it is first, and that is the karaka; thereafter, the otherness of the operations is described by others.

The process-account's ordering. Prakrtih sadhananam sa: *that* — the general operation — is the *material cause of the means*: the karakas are made out of it, not it out of them. Prathamam tac ca karakam: it is *first*, and *that* is the *karaka* — the primal operation is the original karaka, from which the six are differentiated (3.7.37's sodha kartrtvam eva: agency alone, sixfold). Tatah anyatvam vyaparanam aparaih upavarnyate: *thereafter*, the *otherness of the operations* is *described by others*. So the order is: one operation, then karakas as its parts, then the operations' mutual otherness as a further description. This inverts 3.7.1's picture (the means as capacity in the completing of actions) — there the karakas served the action; here they are made of it. And prakrti — material cause — is the word 3.2.15 gave the supreme reality (prakrtim param).

3.8.39 unformalized

बहूनां सम्भवेऽर्थनां केचिदेवोपकारिणः ।
संसर्गे कश्चिदेषां तु प्राधान्येन प्रतीयते ॥ ३९ ॥

bahūnāṃ sambhave'rthanāṃ kecidevopakāriṇaḥ |
saṃsarge kaścideṣāṃ tu prādhānyena pratīyate || 39 ||

Among many possible meanings, only certain ones are assisting; and in the commingling, some one of these is understood as principal.

How one operation among many becomes the verb. Bahunam sambhave arthanam kecid eva upakarinah: among *many possible meanings*, only *certain ones* are *assisting* — a situation contains countless operations, and few contribute (3.7.2's sarvada sarvatha bhavat: the universe existing always and in every way, and something intended somewhere). Samsarge kascid esam tu pradhanyena pratiyate: and in the *commingling*, *some one* of these is understood as *principal*. So the verb picks one operation from a crowd of assisting ones and makes it the principal, and the rest become its karakas (3.8.15's pradhana-bhutam kriyam; the others taking its designation). Kascid — some one: the indefinite is the samuddesa's constant, and it means the selection is not dictated by the crowd.

3.8.40 unformalized

साध्यत्वात्तत्र चाख्यातैर्व्यापाराः सिद्धसाधनाः ।
प्राधान्येनाभिधीयन्ते फलेनापि प्रवर्तिताः ॥ ४० ॥

sādhyatvāttatra cākhyātairvyāpārāḥ siddhasādhanāḥ |
prādhānyenābhidhīyante phalenāpi pravartitāḥ || 40 ||

And there, from being to-be-accomplished, the operations whose means are accomplished are expressed as principal by the finite verbs, impelled also by the fruit.

The verb's office. Sadhyatvat tatra akhyataih vyaparah siddha-sadhanah: *there*, from being *to-be-accomplished*, the *operations* whose *means are accomplished* — the karakas standing ready, siddha (3.7.79's siddha-rupa) — pradhanyena abhidhiyante: are expressed as *principal* by the *finite verbs*. So the finite verb's whole job is to present one operation as principal and to-be-accomplished, with its means already standing. Phalena api pravartitah: *impelled also by the fruit* — and what drives the presentation is the outcome wanted (3.7.128's karma as the action's impeller; 2.113's prayojana). The verb is thus a selection (3.8.39) presented under sadhyatva (3.8.1) and motivated by a fruit — three conditions, and none of them in the world.

3.8.41 unformalized

एकत्वावृत्तिभावाभ्यां भेदाभेदसमन्वये ।
संख्यास्तत्रोपलभ्यन्ते संख्येयावयवक्रियाः ॥ ४१ ॥

ekatvāvṛttibhāvābhyāṃ bhedābhedasamanvaye |
saṃkhyāstatropalabhyante saṃkhyeyāvayavakriyāḥ || 41 ||

In the concomitance of difference and non-difference, through oneness and repetition — there numbers are apprehended, whose countables are the parts of the action.

How actions get counted. Ekatva-avrtti-bhavabhyam bheda-abheda-samanvaye: through *oneness* and *repetition*, in the *concomitance of difference and non-difference* — the action is one (hence countable as one) and its parts recur (hence countable as many). Samkhyah tatra upalabhyante samkhyeya-avayava-kriyah: *numbers* are *apprehended* there, whose *countables* are the *parts of the action*. So 'he cooked twice' counts the whole; 'he stirred ten times' counts a part; and both are available because the action is one-and-many (3.8.4's construed unity of sequential parts). Number attaches wherever the carving stops — which is 3.6.28's finding (oneness would not stand if one did not construct manifoldness) applied to events, and 3.1.51-52's number-doctrine given its temporal case.

3.8.42 unformalized

सिद्धस्यार्थस्य पाकादेः कथं साधनयोगिता ।
साध्यत्वे वा तिङन्तेन कृतां भेदो न कश्चन ॥ ४२ ॥

siddhasyārthasya pākādeḥ kathaṃ sādhanayogitā |
sādhyatve vā tiṅantena kṛtāṃ bhedo na kaścana || 42 ||

How can an accomplished meaning like 'paka' have connection with means? Or, if it is to-be-accomplished, there is no difference whatever from what is done by a finite verb.

A dilemma for the action-noun. Take 'paka' — cooking, the noun formed by the ghan-affix. Siddhasya arthasya pakadeh katham sadhana-yogita: how can an *accomplished* meaning like 'paka' have *connection with means*? A noun presents its object as siddha, finished (3.8.17: the accomplished takes no verb, its expectancy withdrawn) — yet 'paka' takes karakas: 'the cooking of rice by Devadatta'. An accomplished thing should want nothing. Sadhyatve va tinantena krtam bhedo na kascana: *or*, if it is *to-be-accomplished*, there is *no difference whatever* from what is done by a *finite verb* — grant it sadhyatva and the noun is a verb, and the ghan-affix is idle. So: accomplished and it cannot take karakas; to-be-accomplished and it is not a noun. The dilemma targets the samuddesa's own central distinction (3.8.29's verbs deploying sequence, nouns withdrawing it).

3.8.43-44 unformalized

तत्र कारकयोगाया यद्याख्यातं निबन्धनम् ।
षष्ठ्याः सा लेन संबन्धे व्युदस्ता कर्तृकर्मणोः ॥ ४३ ॥
एकाभिधान एकोऽर्थो युगपच्च द्विधर्मभाक् ।
न सम्भवति सिद्धत्वे स साध्यः स्यात् कथं पुनः ॥ ४४ ॥

tatra kārakayogāyā yadyākhyātaṃ nibandhanam |
ṣaṣṭhyāḥ sā lena saṃbandhe vyudastā kartṛkarmaṇoḥ || 43 ||
ekābhidhāna eko'rtho yugapacca dvidharmabhāk |
na sambhavati siddhatve sa sādhyaḥ syāt kathaṃ punaḥ || 44 ||

(43) If the finite verb is there the ground for connection with karakas — the sixth case is excluded, in connection with the 'la'-affix, for agent and object. (44) In a single expression, one meaning does not simultaneously partake of two properties; being accomplished, how then could it be to-be-accomplished?

The dilemma pressed to a contradiction. (43): if the *finite verb* is the ground for karaka-connection, then sasthyah sa lena sambandhe vyudasta kartr-karmanoh — the *sixth case* is *excluded*, in connection with the la-affix, for *agent and object*: where a finite verb governs, one does not say 'Devadatta's cooking of rice's' — the karakas take their own cases and the genitive is barred (3.7.159's genitive prescribed for the exclusion of the compound). But 'paka' *does* take genitives. So 'paka' is not governed by a finite verb, and its karaka-connections need another ground. (44) then closes: eka-abhidhane eko 'rthah yugapat ca dvi-dharma-bhak na sambhavati — in a *single expression*, *one meaning* does not *simultaneously* partake of *two properties*: it cannot be both siddha and sadhya. Siddhatve sa sadhyah syat katham punah: *being accomplished*, how then could it be *to-be-accomplished*? The contradiction is flat, and 3.8.45 answers it by denying its presupposition.

3.8.45 unformalized

एतावत् साधनं साध्यमेतावदिति कल्पना ।
शास्त्र एव न वाक्येऽस्ति विभागः परमार्थतः ॥ ४५ ॥

etāvat sādhanaṃ sādhyametāvaditi kalpanā |
śāstra eva na vākye'sti vibhāgaḥ paramārthataḥ || 45 ||

'This much is the means, this much the to-be-accomplished' — this construction exists in the science alone, not in the sentence; there is no division in the highest sense.

The samuddesa's methodological summit, and it dissolves 3.8.42-44 at a stroke. Etavat sadhanam sadhyam etavad iti kalpana: '*this much* is the means, *this much* the to-be-accomplished' — the whole sadhya/sadhana architecture, on which the Sadhana- and Kriya-samuddesas have both been built. Sastre eva na vakye 'sti: this *construction* exists *in the science alone*, *not in the sentence*. Vibhagah paramarthatah: there is *no division* in the *highest sense*. So the dilemma of 3.8.42-44 assumed that siddha and sadhya are properties a meaning must have or lack — and they are not properties at all; they are the sastra's postings on an undivided sentence-meaning. 'Paka' is neither accomplished nor to-be-accomplished; it is a word in a sentence whose meaning is one (2.143's flash), and the science divides it for its operations. This is 2.88's settlement (abstraction according to capacities, the practice preserved, the constitution untouched) applied to the categories that generate every verb and noun in the language.

3.8.46 unformalized

आख्यातशब्दे भागाभ्यां साध्यसाधनवर्तिता ।
प्रकल्पिता यथा शास्त्रे स घञादिष्वपि क्रमः ॥ ४६ ॥

ākhyātaśabde bhāgābhyāṃ sādhyasādhanavartitā |
prakalpitā yathā śāstre sa ghañādiṣvapi kramaḥ || 46 ||

As, in the finite-verb word, the standing-in-to-be-accomplished-and-means is constructed by two portions in the science — that same procedure is in 'ghan' and the rest.

The consequence for both morphologies. Akhyata-sabde bhagabhyam sadhya-sadhana-vartita prakalpita yatha sastre: as, in the *finite-verb word*, the standing-in-*to-be-accomplished-and-means* is *constructed* by *two portions* *in the science* — the grammar analyses a verb into a root (giving the action) and an ending (giving the karaka-relation), and the analysis is the science's (3.7.13's assistance understood by the case-endings alone). Sa ghan-adisu api kramah: *that same procedure* is in *'ghan'* and the rest — the action-noun is analysed identically. So verb and action-noun are the same construction differently distributed, and neither's siddha/sadhya character is in it. This completes 3.8.45: since the division is the sastra's, it can be run on both morphologies, and 3.8.42's dilemma (why does 'paka' behave like a verb?) has the obvious answer — because the same postings are made on it.

3.8.47 unformalized

साध्यत्वेन क्रिया यत्र धातुरूपनिबन्धना ।
सत्वभावस्तु यस्तस्याः स घञादिनिबन्धनः ॥ ४७ ॥

sādhyatvena kriyā yatra dhāturūpanibandhanā |
satvabhāvastu yastasyāḥ sa ghañādinibandhanaḥ || 47 ||

Where action, as to-be-accomplished, is grounded in the root's form — but that state of thinghood of it is grounded in 'ghan' and the rest.

The division of labour, stated exactly. Sadhyatvena kriya yatra dhatu-rupa-nibandhana: where *action*, *as to-be-accomplished*, is *grounded in the root's form* — the dhatu carries the sadhyatva; that is what a root is for. Sattva-bhavah tu yah tasyah sa ghan-adi-nibandhanah: but that *state of thinghood* of it is *grounded in 'ghan'* and the rest — the affix confers sattva, thing-hood, on what the root gave as action. So 'paka' = root (action, sadhya) + ghan (thinghood, siddha), and the noun's paradox (3.8.42) is that it wears both because it is made of both. And by 3.8.45, both are postings — so the noun is a construction of two constructions. This is 3.1.35 arrived at from the morphology's side: Being with sequence is action, Being with sequence withdrawn is substance, and the ghan-affix is where the withdrawing happens.

3.8.48 unformalized

बन्धुता भेदरूपेण बन्धुशब्दे व्यवस्थिता ।
समूहो बन्ध्ववस्था तु प्रत्ययेनाभिधीयते ॥ ४८ ॥

bandhutā bhedarūpeṇa bandhuśabde vyavasthitā |
samūho bandhvavasthā tu pratyayenābhidhīyate || 48 ||

Kinship, in the form of difference, stands settled in the word 'bandhu'; but the collection, the state of being kin, is expressed by the affix.

The division of labour of 3.8.47 shown on a nominal case. Bandhuta bheda-rupena bandhu-sabde vyavasthita: *kinship*, in the *form of difference*, stands settled in the *word 'bandhu'* — the stem carries the relation, and relations are differences (a kinsman is kin *to* someone). Samuho bandhv-avastha tu pratyayena abhidhiyate: but the *collection* — the *state of being kin* — is *expressed by the affix*: the -ta of bandhuta gathers the difference into a state. So the stem gives the differentiated relation and the affix gives the gathered condition — which is 3.8.47's structure (root gives action, ghan gives thinghood) in nominal morphology, and 3.8.29's (verbs deploy sequence, nouns withdraw it) at the level of a single derivation. Samuha — collection — is 3.8.4's word for what an action is: a collection whose non-difference the intellect constructs. The abstract suffix does to a relation what the mind does to sub-actions.

3.8.49 unformalized

तत्र यं प्रति साध्यत्वमसिद्धा तं प्रति क्रिया ।
सिद्धा तु यस्मिन् साध्यत्वं न तमेव पुनः प्रति ॥ ४९ ॥

tatra yaṃ prati sādhyatvamasiddhā taṃ prati kriyā |
siddhā tu yasmin sādhyatvaṃ na tameva punaḥ prati || 49 ||

There, with respect to what it is to-be-accomplished, the action is unaccomplished; but that in which it is accomplished — with respect to that same one it is not again to-be-accomplished.

The relativity of sadhyatva, stated as a rule. Tatra yam prati sadhyatvam asiddha tam prati kriya: with respect to *what* it is *to-be-accomplished*, the action is *unaccomplished* — sadhyatva is a relation, not a property: an action is to-be-done *for* someone, *toward* something. Siddha tu yasmin sadhyatvam na tam eva punah prati: but that in which it is *accomplished* — with respect to *that same one* it is *not again* to-be-accomplished. So one action is siddha here and sadhya there, relative to different terms — which resolves 3.8.44's contradiction (one meaning cannot bear two properties simultaneously) by showing that the two are not properties but relations, and relations to different terms do not compete. This is 3.8.45's finding (the division is the science's, not the sentence's) given its logical form: siddha and sadhya are two-place, and the dilemma treated them as one-place.

3.8.50 unformalized

राज्ञः पुत्रस्य नप्तेति न राज्ञि व्यतिरिच्यते ।
पुत्रस्यार्थः प्रधानत्वं न चास्य विनिवर्तते ॥ ५० ॥

rājñaḥ putrasya napteti na rājñi vyatiricyate |
putrasyārthaḥ pradhānatvaṃ na cāsya vinivartate || 50 ||

'The grandson of the king's son' — it is not separated in respect of the king; the son's meaning is principality, and this of his does not withdraw.

The relativity of 3.8.49 exhibited in a chain of genitives. Rajnah putrasya napta iti: '*the grandson of the king's son*'. Na rajni vyatiricyate: it is *not separated* in respect of the *king* — the king does not drop out of the chain; the grandson is still his. Putrasya arthah pradhanatvam na ca asya vinivartate: the *son's* meaning is *principality*, and *this of his* does *not withdraw* — the son is principal relative to the grandson and subordinate to the king, and his principality is not cancelled by his subordination. This is 3.7.162 exactly (having obtained principality in its own subordinate, it goes to remainder-hood in the principal; therefore principality is not lost) — and its recurrence here confirms that the samuddesa's structural insight is general: relational statuses are two-place, and a term holds opposite ones toward different terms without contradiction.

3.8.51 unformalized

मृगो धावति पश्येति साध्यसाधनरूपता ।
तथा विषयभेदेन सरणस्योपपद्यते ॥ ५१ ॥

mṛgo dhāvati paśyeti sādhyasādhanarūpatā |
tathā viṣayabhedena saraṇasyopapadyate || 51 ||

'See — the deer runs!' — the character of to-be-accomplished and means; so, by difference of domain, it holds for the running.

One action bearing both statuses in a single sentence. Mrgo dhavati pasya iti: '*see — the deer runs*'. Sadhya-sadhana-rupata: the *character of to-be-accomplished and means* — the running is *sadhya* (it is what the deer is doing, presented as in-process) and *sadhana* (it is the means by which the seeing gets its object: you see the deer *by* its running). Tatha visaya-bhedena saranasya upapadyate: *so*, by *difference of domain*, it holds for the *running*. One running, two statuses, two domains — relative to the deer it is to-be-accomplished, relative to the seeing it is a means. This is 3.8.49's rule in a sentence anyone might say, and it is the samuddesa's cleanest evidence that the sadhya/sadhana division is a matter of what one relates the action to (3.8.45's sastre eva na vakye).

3.8.52 unformalized

लकृत्यक्तखलर्थानां तथाव्ययकृतामपि ।
रूढिनिष्ठाघञादीनां धातुः साध्यस्य वाचकः ॥ ५२ ॥

lakṛtyaktakhalarthānāṃ tathāvyayakṛtāmapi |
rūḍhiniṣṭhāghañādīnāṃ dhātuḥ sādhyasya vācakaḥ || 52 ||

For the 'la'-affixes, the krtya, the kta, the khal-meanings, and likewise for the indeclinable-krt forms, the conventional, the nistha, the 'ghan' and the rest — the root is the expressor of the to-be-accomplished.

The root's invariance across every morphology. La-krtya-kta-khal-arthanam: for the *la*-affixes (finite verbs), the *krtya* (gerundives), the *kta* (past participles), the *khal*-meanings; tatha avyaya-krtam api: and likewise for the *indeclinable-krt* forms (gerunds, infinitives); rudhi-nistha-ghan-adinam: the *conventional*, the *nistha*, the *'ghan'* and the rest (action-nouns). Dhatuh sadhyasya vacakah: the *root* is the *expressor* of the *to-be-accomplished*. Every derivative of a root, however nominal, however frozen into convention — all of them get their sadhyatva from the root, and the affixes do everything else. This is 3.8.47's division stated exhaustively: whatever a form does, its action-content is the dhatu's, and its thing-content is the affix's. The list's length is the point — no derivative escapes.

3.8.53 unformalized

साध्यस्यापरिनिष्पत्तेः सोऽयमित्यनुपग्रहः ।
तिङन्तैरन्तरेणेवमुपमानं ततो न तैः ॥ ५३ ॥

sādhyasyāpariniṣpatteḥ so'yamityanupagrahaḥ |
tiṅantairantareṇevamupamānaṃ tato na taiḥ || 53 ||

From the non-completion of the to-be-accomplished, there is no taking-up as 'that is this'; therefore there is no comparison by finite verbs, without something else.

Why verbs cannot be compared. Sadhyasya aparinispatteh: from the *non-completion* of the *to-be-accomplished* — an action, while presented as sadhya, is not finished (3.8.16-17: sadhyatva holds exactly between beginning and completion). So 'yam iti anupagrahah: there is no *taking-up* as '*that is this*' — the so 'yam identification (2.40, 2.128's abhijalpa) requires something settled to identify. Tinantaih antarena evam upamanam tatah na taih: *therefore* there is *no comparison* *by finite verbs*, *without something else*. One cannot say 'he runs like he runs' — comparison needs two completed terms, and verbs present nothing completed. This is a genuine linguistic observation: Sanskrit's comparison-apparatus attaches to nouns, and the samuddesa explains why from its own doctrine of sadhyatva.

3.8.54 unformalized

साधनत्वं प्रसिद्धं च तिङूक्षु संबन्धिनां यतः ।
तेनाध्यारोप एव स्यादुपमा तु न विद्यते ॥ ५४ ॥

sādhanatvaṃ prasiddhaṃ ca tiṅūkṣu saṃbandhināṃ yataḥ |
tenādhyāropa eva syādupamā tu na vidyate || 54 ||

And since means-hood is well established for the relata in finite-verb utterances — therefore there would be superimposition alone; but comparison does not exist.

What happens instead of comparison. Sadhanatvam prasiddham ca tinuktesu sambandhinam yatah: since *means-hood* is *well established* for the *relata* in *finite-verb* utterances — in a verbal sentence the participants are already fixed as karakas (3.7.13's assistance understood by the case-endings). Tena adhyaropa eva syat upama tu na vidyate: *therefore* there would be *superimposition* alone; but *comparison* does *not exist*. Adhyaropa — superimposition (3.1.6's central term): where one might expect a comparison between two verbal situations, what actually happens is that one is laid over the other. So verbal 'comparison' is really projection: 'he flies like a bird' does not compare two flyings; it superimposes the bird's on his. The distinction is fine and it follows from 3.8.53: no completed terms, no comparison — only overlay.

3.8.55 unformalized

न्यूनेषु च समाप्तार्थमुपमानं विधीयते ।
क्रिया चैवाश्रये सर्वा तत्र तत्र समाप्यते ॥ ५५ ॥

nyūneṣu ca samāptārthamupamānaṃ vidhīyate |
kriyā caivāśraye sarvā tatra tatra samāpyate || 55 ||

And comparison is prescribed for things of completed meaning among the deficient; and every action is completed in each substratum, there and there.

The two conditions comparison needs, and actions fail one. Nyunesu ca samapta-artham upamanam vidhiyate: *comparison* is prescribed for things of *completed meaning* among the *deficient* — one compares a lesser thing to a complete standard (2.63's upamana-upameya pair, where the standard always retains a property the compared lacks). Kriya ca eva asraye sarva tatra tatra samapyate: and *every* action is *completed* in *each substratum*, *there and there*. Every action completes wholly in each of its bearers (3.8.5's pratibhedam samapyate: the whole at each division). So no action is *deficient* relative to another: each is complete where it is, and comparison needs a deficient term. Hence the impossibility of 3.8.53, now given its second ground: not only are verbs unfinished (3.8.53), their actions are also never lacking (here) — and comparison requires both a complete standard and a lacking subject.

3.8.56 unformalized

येनैव हेतुना हंसः पततीत्यभिधीयते ।
आतौ तस्य समाप्तत्वादुपमार्थो न विद्यते ॥ ५६ ॥

yenaiva hetunā haṃsaḥ patatītyabhidhīyate |
ātau tasya samāptatvādupamārtho na vidyate || 56 ||

By that very cause by which it is said 'the goose flies' — from its being completed in the coming, the purpose of comparison does not exist.

The doctrine of 3.8.55 in an example. Yena eva hetuna hamsah patati iti abhidhiyate: by that *very cause* by which it is said '*the goose flies*' — whatever makes the flying the goose's flying. Atau tasya samaptatvat: from its being *completed* in the *coming* — the flying is complete in the goose's own arriving; nothing is lacking in it. Upama-artho na vidyate: the *purpose of comparison* does *not exist*. So one cannot say 'he flies as the goose flies', because the goose's flying is not a standard against which anything is deficient — it is simply complete where it is, as every action is complete in its bearer (3.8.55). The example is well chosen: the goose is the tradition's stock simile-vehicle, and the verse denies that its flying can serve as one.

3.8.57 unformalized

क्रियाणां जातिभिन्नां सादृश्यां नावधार्यते ।
सिद्धेश्च प्रक्रमे साध्यमुपमातुं न शक्यते ॥ ५७ ॥

kriyāṇāṃ jātibhinnāṃ sādṛśyāṃ nāvadhāryate |
siddheśca prakrame sādhyamupamātuṃ na śakyate || 57 ||

A similarity of actions, differing in universal, is not ascertained; and in the commencing of the accomplished, the to-be-accomplished cannot be compared.

The two impossibilities gathered. Kriyanam jati-bhinnanam sadrsyam na avadharyate: a *similarity* of actions *differing in universal* is *not ascertained* — two actions of different kinds have no common measure; flying and running share no jati under which they could be alike (3.1.14's anuvrtti: recurrence being the universal's mark, and there is none across kinds). Siddheh ca prakrame sadhyam upamatum na sakyate: and in the *commencing of the accomplished*, the *to-be-accomplished* cannot be *compared* — where the comparison-apparatus starts from settled terms, the unsettled cannot enter it. So actions cannot be compared across kinds (no shared universal) or within the verbal presentation (nothing settled). Between them the verse closes the case that 3.8.53-56 built: comparison is a nominal operation, and the reason is what verbs are.

3.8.58 unformalized

वनं वृक्षा इति यथा भेदाभेदव्यपाश्रयात् ।
अर्थात्मा भिद्यते भावे स बाह्याभ्यन्तरे क्रमः ॥ ५८ ॥

vanaṃ vṛkṣā iti yathā bhedābhedavyapāśrayāt |
arthātmā bhidyate bhāve sa bāhyābhyantare kramaḥ || 58 ||

As 'forest' and 'trees' — from a resort to difference and non-difference, the meaning's self is differentiated; in the entity, that sequence is outer and inner.

The samuddesa's most compact image for its whole doctrine. Vanam vrksah iti yatha: as *'forest'* and *'trees'* — one thing, two words: the forest is the trees, and 'forest' takes them as one while 'trees' takes them as many. Bheda-abheda-vyapasrayat artha-atma bhidyate: from a *resort to difference and non-difference*, the *meaning's self* is *differentiated* — the same reality carved one way or the other by which word one uses. Bhave sa bahya-abhyantare kramah: in the *entity*, that *sequence* is *outer and inner*. So the forest/trees pair is the noun/verb pair: gather and the sequence goes inside (the forest's trees are its internal manifold); disperse and it comes outside (the trees stand in a row). Krama is the same krama, relocated by the word. This is 3.8.29 (verbs deploy sequence, nouns withdraw it) shown in a case where both words are nouns — which proves the operation is not about word-class but about gathering.

3.8.59 unformalized

सामान्ये भाव इत्यत्र यल्लिङ्गमुपलभ्यते ।
भेदानामनुमेयत्वान्न तत् तेषु विवक्ष्यते ॥ ५९ ॥

sāmānye bhāva ityatra yalliṅgamupalabhyate |
bhedānāmanumeyatvānna tat teṣu vivakṣyate || 59 ||

The gender-mark which is apprehended in 'bhava' as the general — since the differences are to be inferred, that is not intended among them.

A grammatical detail with doctrinal weight. Samanye bhava iti atra yal lingam upalabhyate: the *gender-mark* apprehended in *'bhava'* (being, existence) *as the general* — 'bhava' is masculine, and the abstract noun for existence-in-general carries that gender. Bhedanam anumeyatvat na tat tesu vivaksyate: since the *differences* are *to be inferred*, *that* is *not intended* among them. The particular actions subsumed under bhava (cooking, going, whitening) do not inherit its masculinity; the gender belongs to the general word and does not distribute. So 'bhava' is masculine and cooking is not masculine — gender is a property of the word, not of what it means (3.6.10's direction-words, feminine, with nothing in direction to be engendered). And this prepares 3.8.60-62's argument that the root's meaning is one across all its gendered nominalizations.

3.8.60 unformalized

निर्देशे चरितार्थत्वाल्लिङ्गं भावेऽविवक्षितम् ।
उपमासविधत्वाच्च भावादन्यत् पचादिषु ॥ ६० ॥

nirdeśe caritārthatvālliṅgaṃ bhāve'vivakṣitam |
upamāsavidhatvācca bhāvādanyat pacādiṣu || 60 ||

From its purpose having been served in the designation, the gender-mark is unintended in 'bhava'; and from its proximity to comparison, something other than 'bhava' is in 'paci' and the rest.

The gender-mark's status settled. Nirdese caritarthatvat lingam bhave avivaksitam: from its *purpose having been served* in the *designation*, the *gender-mark* is *unintended* in 'bhava' — caritartha: spent, its errand discharged (2.54's words that vanish into their meaning; 3.7.83's animal that served the rite). The gender got the word said and means nothing. Upama-savidhatvat ca bhavad anyat pacadisu: and from its *proximity to comparison*, something *other than* 'bhava' is in *'paci'* and the rest — because 'paci' can enter comparison-constructions in a way 'bhava' cannot (3.8.53-57's analysis), there is something in the particular roots beyond bare existence. So the verse concedes a residue: the roots are not simply bhava, or they would all behave alike. And 3.8.61-62 will locate the residue exactly.

3.8.61 unformalized

भवतौ यत् पचादीनां तावदत्रोपदिश्यते ।
न च लिङ्गं पचादीनां भवतौ समवस्थितम् ॥ ६१ ॥

bhavatau yat pacādīnāṃ tāvadatropadiśyate |
na ca liṅgaṃ pacādīnāṃ bhavatau samavasthitam || 61 ||

Only so much of 'paci' and the rest as is in 'bhavati' is taught here; and the gender-mark of 'paci' and the rest does not stand settled in 'bhavati'.

The residue located. Bhavatau yat pacadinam tavad atra upadisyate: *only so much* of 'paci' and the rest as is in *'bhavati'* (it exists, it becomes) is *taught here* — what the particular roots share with the existence-verb is what the doctrine addresses: their bhava-content. Na ca lingam pacadinam bhavatau samavasthitam: and the *gender-mark* of 'paci' and the rest does *not stand settled* in 'bhavati' — the particulars' grammatical properties do not transfer to the general verb. So the roots have bhava in common and their own marks besides, and the marks do not travel in either direction (3.8.59's non-distribution, now stated symmetrically). The verse is careful bookkeeping in service of the identification 3.8.62 is about to make.

3.8.62 unformalized

एकश्च सोऽर्थः सत्ताख्यः कथञ्चित् कैश्चिदुच्यते ।
लिङ्गानि चास्य भिद्यन्ते पचिरूपादिभेदवत् ॥ ६२ ॥

ekaśca so'rthaḥ sattākhyaḥ kathañcit kaiściducyate |
liṅgāni cāsya bhidyante pacirūpādibhedavat || 62 ||

And that meaning is one, called Existence; it is somehow spoken of by some. And its marks are differentiated, like the differences of 'paci's form and the rest.

The identification, and it is the Padakanda's constant. Ekah ca so 'rthah satta-akhyah: and that *meaning* is *one*, called *Existence* — the root's meaning, all roots' meaning, is satta (3.1.34's tam pratipadika-artham ca dhatv-artham ca pracaksate; 3.8.23's bhava eva hi dhatv-arthah, the unbroken tradition). Kathamcit kaiscid ucyate: it is *somehow* spoken of by *some* — the shrug (2.171, 3.7.34's kathamcana): the schools name it variously and none says it well. Lingani ca asya bhidyante paci-rupadi-bhedavat: and its *marks* are *differentiated*, like the *differences* of 'paci's form and the rest. So: one meaning, Existence; many marks, differentiated as the roots' forms are. The residue 3.8.60 conceded is thus placed — it is in the *marks*, not in the meaning. Every root means satta, and they differ in how they mark it, and that is the whole story of the Sanskrit verb.

3.8.63 unformalized

आचार्यो मातुलश्चेति यथैको व्यपदिश्यते ।
संबन्धिभेदादर्थात्मा स विधिः पक्तिभावयोः ॥ ६३ ॥

ācāryo mātulaśceti yathaiko vyapadiśyate |
saṃbandhibhedādarthātmā sa vidhiḥ paktibhāvayoḥ || 63 ||

As one and the same is designated 'teacher' and 'maternal uncle' — the meaning's self, by difference of relata: that is the rule for 'pakti' and 'bhava'.

The samuddesa's last verse, and it closes with a homely simile. Acaryo matulah ca iti yatha ekah vyapadisyate: as *one and the same* man is designated '*teacher*' and '*maternal uncle*' — one person, two names, and the names are not synonyms: they say quite different things. Sambandhi-bhedad artha-atma: the *meaning's self*, by *difference of relata* — he is teacher relative to his pupil and uncle relative to his sister's son, and the two designations differ because the relata differ, not because the man does. Sa vidhih pakti-bhavayoh: *that* is the *rule* for '*pakti*' (cooking) and '*bhava*' (being). So 'cooking' and 'being' name one thing under different relations — as 'teacher' and 'uncle' do. The Kriya-samuddesa ends by making its central identification (every root means satta, 3.8.62) intuitive: nobody is puzzled that one man is teacher and uncle, and nobody should be puzzled that one Existence is cooking and being. The relata differ; the self does not.