This chapter teaches verbs made from nouns: "he wants a son", "he behaves like Kṛṣṇa", "it becomes red", "he worships". Pāṇini forms them by adding an affix to a noun, not to a root. The result is a new root, a nāmadhātu, "a root from a noun". It then takes the lakāras like any other root. Bhaṭṭoji calls the section the nāmadhātu-prakriyā, "the working of nominal roots". After this chapter you will be able to build putrīyati "he wants a son for himself" step by step, and say why rājīyati loses its n but vācyati keeps its c. You will know the other forms of wishing (aśanāyati, aśvasyati, putrakāmyati) and of behaving (putrīyati chātram, kṛṣṇāyate, ojāyate, kumārāyate, kṛṣṇati, rājānati). You will form verbs of becoming (bhṛśāyate, sumanāyate, lohitāyati / lohitāyate, paṭapaṭāyati) and of special acts (kaṣṭāyate, romanthāyate, tapasyati, bāṣpāyate, śabdāyate, namasyati, utpucchayate, muṇḍayati). Last, you will learn how a noun before ṇic is treated as if it stood before the superlative affix iṣṭhan: prathayati, svāpayati, tvādayati, śāvayati, tirāyayati, jyāpayati. The chapter has twenty-one sūtras, SK 2657 to SK 2677.
putram ātmana icchati: putra-am + kyac → (sup luk) putra-ya → (ī for a, 2658) putrīya → putrīyati; rājānam icchati: rājan-ya → (pada, n deleted) rāja-ya → rājīyatiFour sūtras build the first nominal root. SK 2657 adds the affix kyac to a noun in the sense "wishes for". SK 2658 turns the stem's final a into ī. SK 2659 decides when the noun before kyac still counts as a word, a pada. SK 2660 lets kyac drop after a consonant before certain affixes. The running examples are putra "son" and rājan "king". "He wishes for a son for himself" becomes putrīyati; "he wishes for a king" becomes rājīyati.
Rule. The affix kyac comes optionally, in the sense of wishing. It comes after a word ending in a case-ending (sup) that names the object of the wishing. That object must be wished for the wisher himself. Bhaṭṭoji's gloss: "let kyac come optionally, in the sense of wishing, after a sup-final word that is the object of wishing and is connected with the wisher".
Bhaṭṭoji adds one sentence: "Because the case-ending is now part of a dhātu, it undergoes luk." A form ending in kyac is a dhātu by SK 2304, so the case-ending inside it is deleted.
Three words come from SK 2608 by anuvṛtti: karmaṇaḥ "after the object", icchāyām "in the sense of wishing", and vā "optionally". Bhaṭṭoji's commentators give this. They add that "object" means the object of the wishing, since wishing is the nearest action. Ātman means "self", and the self is the wisher.
Example.
putram ātmana icchati, "he wishes for a son for himself":
putra + am (the second case-ending: the son is the object)
→ this rule: kyac; k and c are markers → putra-am-ya
→ the whole is a dhātu (SK 2304), so the case-ending am undergoes luk → putra-ya
→ ī for a before kyac (SK 2658) → putrīya
→ laṭ, the ending tip, the class-sign śap → putrīya + a + ti
→ the two *a*s become one (SK 191) → putrīyati.
Kyac carries no ṅ, so the endings are parasmaipada.
Applies when / blocked by. Why "for oneself" (ātmanaḥ)? One of Bhaṭṭoji's commentators gives the counter-example parasya putram icchati, "he wishes for another's son". There no kyac comes, and only the phrase is used.
The other commentator adds a second limit. Kyac is added to a finished word, so the word must be samartha, connected in sense with nothing else. In mahāntaṃ putram icchati, "he wishes for a big son", putra depends on mahāntam. So no kyac comes after putra.
Watch out. The rule is optional. Putram ātmana icchati stays correct beside putrīyati. In putrīyati the sense of wishing is carried by kyac.
Rule. Before kyac, ī replaces the a of the stem. Bhaṭṭoji's gloss: "let ī replace a". One of Bhaṭṭoji's commentators names the sources. "Of a" (asya) comes from SK 2118, asya cvau, and "ī" from SK 2648, ī ghrādhmoḥ.
Example.
putra-ya → this rule: the a of putra becomes ī → putrīya → putrīyati.
A stem in o or au does not take ī. SK 63 gives av and āv before an affix beginning with y:
go + kyac → go-ya → av for o (SK 63) → gav-ya → gavyati, "he wishes for a cow for himself"
nau + kyac → nau-ya → āv for au (SK 63) → nāv-ya → nāvyati, "he wishes for a boat for himself".
Applies when / blocked by. Could SK 67, lopaḥ śākalyasya, now drop the v of gav-ya? No. That rule drops only a v at the end of a pada, and this v is not pada-final.
Someone may object: the stem is a pada through its deleted case-ending. Bhaṭṭoji answers with "For it is as follows", and goes straight on to the next sūtra. One of Bhaṭṭoji's commentators reads the phrase this way: SK 2659 shows how pada-status fails here.
Rule. Before kyac and kyaṅ, only a sup-final word ending in n is a pada; no other is. Bhaṭṭoji's gloss: "before kyac and kyaṅ, only an n-final is a pada, and no other".
One of Bhaṭṭoji's commentators explains the form of the rule. "Sup-final" and "pada" come by anuvṛtti from SK 29, suptiṅantaṃ padam. A sup-final word is a pada anyway, so this rule is a restriction (niyama). Why does it matter? Before kyac, the stem ends in a deleted case-ending. Without this rule it would be a pada, and the rules for the end of a pada would apply.
Here kya means kyac and kyaṅ only. Bhaṭṭoji's commentators say why kyaṣ is left out: kyaṣ does not come after a consonant-final stem.
Example.
rājānam icchati, "he wishes for a king":
rājan + am + kyac → luk of am → rājan-ya
→ the stem ends in n, so it is a pada (this rule)
→ the n at the end of the pada is deleted → rāja-ya
→ ī for a (SK 2658) → rājīya → rājīyati.
Bhaṭṭoji's commentators add a point. The loss of n is not counted as undone for the ī. SK 353 limits that to four kinds of operation, and this is not one of them.
Stems that do not end in n are not padas. So the end-of-pada rules stay away:
vāc-ya → vācyati: the c does not become k. One of Bhaṭṭoji's commentators adds that vac takes no saṃprasāraṇa by SK 2409 either. That rule works only before an affix taught after a dhātu.samidh-ya → samidhyati: the dh does not become d.adas-ya → adasyati: the s stays s. One of the commentators notes that it does not become ru, since it is not pada-final.More forms. Bhaṭṭoji works through a series of stems.
gav-ya? No. The v arose because of that y, so it cannot cause the y's loss. This is the sannipāta paribhāṣā: a rule does not destroy the condition that brought it in. So: gavyāṃcakāra, gavyitā; nāvyāṃcakāra, nāvyitā.tvadyati, madyati. This holds only for one person. For more than one: yuṣmadyati, asmadyati.gīryati, pūryati. But only after a dhātu. Div "sky" is not one, so divyati keeps its short i.kartrīyati. Gārgya: the y of the descendant affix is lost by SK 2119, then ī → gārgīyati.kavīyati.Watch out. Here kya means only kyac and kyaṅ; the commentators reject reading kyaṣ into it (see the k of kyaṣ under SK 2669).
Rule. After a consonant, kyac and kyaṅ are optionally deleted before an ārdhadhātuka affix. Bhaṭṭoji's gloss: "let kyac and kyaṅ, after a consonant, be optionally deleted before an ārdhadhātuka". One of Bhaṭṭoji's commentators gives the carried words. "After a consonant" (halaḥ) comes from SK 2631; "before an ārdhadhātuka" is the heading of SK 2307.
Example.
samidh "fuel", luṭ, 3rd sg.: samidh-ya + itā
→ this rule: deletion. By SK 44 it falls on the first sound only, the y → samidh-a-itā
→ the a is deleted before the ārdhadhātuka (SK 2308) → samidh-itā
→ that loss of a counts as if the a were still there (sthānivat). So the light penultimate i takes no guṇa → samidhitā.
Without this rule: samidhyitā.
Vārttika: māntaprakṛtikasubantād avyayāc ca kyac na. "Kyac does not come after a sup-final word whose base ends in m, nor after an indeclinable." Bhaṭṭoji's counter-examples: kim icchati, idam icchati, svar icchati. These stay phrases.
Applies when / blocked by. Bhaṭṭoji's commentators explain the wording. In the Bhāṣya the statement reads māntāvyayebhyaḥ pratiṣedhaḥ, "prohibition after m-final words and indeclinables". Read literally, it would stop putrīyati, because putram ends in m. Kyac would then come only from putrau or putrān. So Bhaṭṭoji rewords it: the base must end in m.
Watch out. Look at the base, not the inflected word. Putra does not end in m, so putrīyati stands. Kim and idam do, so they take no kyac.
aśva + asuk + kyac → aśvas-ya → aśvasyati (vaḍavā); putram ātmana icchati: putra + kāmyac → putrakāmyatiThree more sūtras stay with wishing. SK 2661 gives three ready-made verbs for hunger, thirst and greed. SK 2662 adds the augment asuk to four stems, so that "the mare desires a stallion" is aśvasyati vaḍavā. SK 2663 offers a second affix, kāmyac, beside kyac: putrakāmyati, "he wishes for a son for himself".
Rule. Aśanāya, udanya and dhanāya are given ready-made as kyac-final forms. A form given ready-made in a sūtra is a nipātana. They are used in the senses of hunger, thirst and greed, respectively. Bhaṭṭoji's gloss: "kyac-final forms given as nipātanas". "Respectively" is one commentator's word.
Example.
aśana + kyac → aśanāya → aśanāyati, "he wants to eat food". The a does not become ī (SK 2658); that is the irregular part. The ā comes by SK 2298 instead. One of Bhaṭṭoji's commentators gives this account.udaka + kyac → udan replaces udaka, and its n is not deleted → udanya → udanyati, "he wants to drink water".dhana + kyac → dhanāya → dhanāyati, "he craves wealth".Applies when / blocked by. Why "in the senses of hunger and the rest"? Outside them, the regular forms come: aśanīyati, udakīyati, dhanīyati. Bhaṭṭoji's commentators explain the contrast. In the counter-examples one wants food to keep for later, or water for some other use than drinking. One commentator says bathing, the other, watering crops. Or one wants wealth because one is poor, not out of greed.
Watch out. The sense picks the form. One who wants food to eat now says aśanāyati; one who wants food to store says aśanīyati.
Rule. Before kyac, in the sense of one's own gratification, aśva "horse", kṣīra "milk", vṛṣa "bull" and lavaṇa "salt" take the augment asuk. An augment (āgama) is a sound added to a form, not put in place of anything. Bhaṭṭoji's gloss: "these take the augment asuk before kyac".
Asuk comes by anuvṛtti from āj jaser asuk (SK 3572, taught later). Bhaṭṭoji's commentators say so. One of them adds that its k is a marker and its u is for pronunciation. Being marked with k, it is added at the end of the stem.
Example.
aśvasyati vaḍavā, "the mare desires a stallion":
aśva + kyac → this rule: asuk at the end of the stem → aśva-as-ya
→ a single short a remains where the two met → aśvas-ya
→ the stem now ends in s, not a, so SK 2658 has nothing to change → aśvasyati.
Vārttika: aśvavṛṣayor maithunecchāyām. "For aśva and vṛṣa, in the sense of desire for mating." aśvasyati vaḍavā; vṛṣasyati gauḥ, "the cow desires a bull".
Vārttika: kṣīralavaṇayor lālasāyām. "For kṣīra and lavaṇa, in the sense of craving (lālasā)." kṣīrasyati bālaḥ, "the child craves milk"; lavaṇasyati uṣṭraḥ, "the camel craves salt".
Vārttika: sarvaprātipadikānāṃ kyaci lālasāyāṃ sugasukau. "For all noun stems, before kyac in the sense of craving, the augments suk and asuk." A noun stem is a prātipadika. Bhaṭṭoji's forms:
dadhi + suk + kyac → dadhisyati; dadhi + asuk + kyac → y for i before a vowel → dadhyasyati
madhu + suk → madhusyati; madhu + asuk → v for u → madhvasyati.
One of Bhaṭṭoji's commentators reads this vārttika as giving kyac itself to every stem in the sense of craving, and then the augments. He adds that it does not repeat the kṣīra vārttika. That one is Kātyāyana's; this is "another view", which the Bhāṣya introduces with apara āha, "another says".
Watch out. Both of Bhaṭṭoji's commentators discuss vṛṣasyantī, "desiring", said of a woman. There only the desire for mating is meant. The stem's own meaning, "bull", drops away.
Rule. The affix kāmyac also comes, in the same domain as kyac (SK 2657). Bhaṭṭoji's gloss: "in the domain just stated, let kāmyac come". So it too comes optionally after a sup-final object wished for oneself.
Example.
putram ātmana icchati, "he wishes for a son for himself":
putra + am + kāmyac → luk of am → putra-kāmya
→ no ī: SK 2658 works only before kyac → putrakāmyati.
Its k is heard. Bhaṭṭoji's commentators say it is not a marker; one says this is because it is pronounced, the other because a marker k would serve no purpose.
Applies when / blocked by. SK 2631, yasya halaḥ, deletes ya after a consonant. Does it touch the ya of kāmya, which follows m? Bhaṭṭoji says no, because that y is meaningless. Yasya takes the sequence y + a as a unit, as stated under SK 2631. One of his commentators gives the contrast. In bebhidya-itā the ya is the affix yaṅ, which has a meaning, so it is deleted. In kāmyac the ya is only part of an affix.
More forms:
yaśaskāmyati, "he wishes for fame". One of Bhaṭṭoji's commentators gives s, not visarga, by so 'padādau (SK 152).sarpiṣkāmyati, "he wishes for ghee".kiṃkāmyati, svaḥkāmyati. The bar on m-final bases and indeclinables (the vārttika under SK 2660) does not reach kāmyac. Bhaṭṭoji says it outright: this affix does come after them.Watch out. Kyac and kāmyac share their domain but not their limits. kim icchati has no kyac-form, yet kiṃkāmyati is correct.
putram ivācarati chātram: putra + kyac → putrīyati chātram; kṛṣṇa ivācarati: kṛṣṇa + kyaṅ → (lengthening, SK 2298) kṛṣṇāya → kṛṣṇāyateThe sense now changes from wishing to behaving (ācāra). The thing one behaves like is the standard of comparison, the upamāna. SK 2664 gives kyac when that standard is an object: "he treats the pupil like a son", putrīyati chātram. SK 2665 gives a new affix, kyaṅ, when the standard is an agent: "he behaves like Kṛṣṇa", kṛṣṇāyate. Because kyaṅ is marked with ṅ, its verbs take the ātmanepada.
Rule. Kyac comes optionally after a sup-final word naming an object that is the standard of comparison, in the sense of behaving. Bhaṭṭoji's gloss: "let kyac come, in the sense of behaving, after a sup-final word that is an object serving as the standard".
One of Bhaṭṭoji's commentators gives the carried words. "Sup-final" and "kyac" come from SK 2657; "object" and "optionally" from SK 2608. The option shows in Bhaṭṭoji's own analytic phrase, putram iva.
Example.
putram ivācarati chātram, "he treats the pupil like a son":
putra + am (the standard, an object) + kyac → luk → putra-ya
→ ī for a (SK 2658) → putrīya → putrīyati chātram.
Chātram "the pupil" is the one actually so treated.
Likewise viṣṇum ivācarati dvijam → viṣṇūyati dvijam, "he treats the brāhmaṇa like Viṣṇu". The u is lengthened before ya (SK 2298).
Vārttika: adhikaraṇāc ceti vaktavyam. "It should be stated: also after a word for a locus (adhikaraṇa)." One of Bhaṭṭoji's commentators says this means a word in the seventh case. Bhaṭṭoji's forms:
prāsādīyati kuṭyāṃ bhikṣuḥ, "the mendicant lives in his hut as in a palace";kuṭīyati prāsāde, "he lives in a palace as in a hut".One commentator adds the feeling: happy in the hut in the first, miserable in the palace in the second.
Watch out. putrīyati is now two verbs. With SK 2657 it means "he wishes for a son for himself". With this rule, and an object beside it, it means "he treats [someone] like a son".
Rule. The affix kyaṅ comes optionally after a sup-final word naming an agent (kartṛ) that is the standard of comparison, in the sense of behaving. If that agent-word ends in s, its s is deleted. Bhaṭṭoji's gloss: "let kyaṅ come optionally… and of an s-final agent-word let there optionally be deletion".
"The standard" and "in the sense of behaving" come from SK 2664; "optionally" from SK 2608. Bhaṭṭoji's commentators say so.
How to read the sūtra. One of Bhaṭṭoji's commentators reads kartuḥ twice. First as a fifth case: "after the agent, kyaṅ". Then as a sixth case, qualified by sa: "of the agent ending in s, deletion". The two commentators take the word ca "and" in slightly different ways; the sense is the same.
The two options. Bhaṭṭoji explains them. Kyaṅ is optional, so the analytic phrase also stands. The loss of s is taught together with kyaṅ (kyaṅsanniyogaśiṣṭa). It never happens without kyaṅ. And the loss of s has a fixed distribution (vyavasthita), set out by the vārttika below.
Example.
kṛṣṇa ivācarati, "he behaves like Kṛṣṇa":
kṛṣṇa + su (the agent, first case) + kyaṅ → luk of su → kṛṣṇa-ya
→ the a is lengthened before ya (SK 2298) → kṛṣṇāya
→ kyaṅ is marked with ṅ, so the ātmanepada comes → kṛṣṇāya + a + te → kṛṣṇāyate.
One commentator gives these steps (lengthening, ātmanepada from the ṅ) for apsarāyate, after its s is lost; kṛṣṇāyate is built the same way.
Vārttika: ojaso 'psaraso nityam itareṣāṃ vibhāṣayā. "For ojas and apsaras, [the loss of s] is obligatory; for the rest, optional."
ojas + kyaṅ → s deleted → oja-ya → ojāya → ojāyate, "he behaves like a vigorous man".apsarāyate, "behaves like an apsaras".yaśāyate or yaśasyate, "he behaves like a famous man".vidvāyate or vidvasyate, "he behaves like a learned man". One of Bhaṭṭoji's commentators notes that in vidvasyate the s does not become ru. The stem is not a pada (SK 2659).Yuṣmad, asmad (SK 1373 again): tvadyate, madyate, "he behaves like you / like me". For more than one: yuṣmadyate, asmadyate.
Feminine stems as masculine. By kyaṅmāninoś ca (SK 837), a feminine stem is treated as masculine before kyaṅ (puṃvadbhāva). Bhaṭṭoji's forms:
kumāra-ya → kumārāyate.haritāyate. Here hariṇī is the feminine of the colour-word harita "tawny". Treated as masculine, it loses both the n of SK 496 and the feminine ī. One of Bhaṭṭoji's commentators explains.gurūyate.yuvāyate. One commentator says age is not a class (jāti); the other says age-words take class-treatment only optionally. Either way the masculine treatment is not barred.paṭvīmṛdūyate. Only the last member stands before kyaṅ, so only it is treated as masculine. Bhaṭṭoji's commentators point this out.Sapatnīva, "like a sapatnī", gives three forms: sapatnāyate, sapatīyate, sapatnīyate. Bhaṭṭoji's commentators explain that sapatnī here stands for three different words:
sapatnāyate.sapatīyate.sapatnīyate.Applies when / blocked by. Na kopadhāyāḥ (SK 838) bars the masculine treatment when the penultimate is k. So pācikā gives pācikāyate, not pācakāyate.
Watch out. With kyac (SK 2664), the standard is an object and the verb is parasmaipada: putrīyati. With kyaṅ (this rule), the standard is an agent and the verb is ātmanepada: kṛṣṇāyate. The ṅ is what makes the difference. This unit also carries the kvip vārttikas; they follow in group D.
rājeva ācarati: rājan + kvip → (2666) rājān → (no pada, n kept) rājānati; kṛṣṇa ivācarati: kṛṣṇa + kvip → kṛṣṇa-a-ti → (ato guṇe) kṛṣṇatiUnder SK 2665 Bhaṭṭoji gives two vārttikas with a third affix for behaving: kvip. Kvip is a kṛt affix of which no sound remains. So the noun stem itself becomes a root, and takes śap and the endings directly: kṛṣṇati, "he behaves like Kṛṣṇa". SK 2666 then lengthens the penultimate vowel of a nasal-final stem before kvi: rājānati, "he behaves like a king". Bhaṭṭoji uses these rules to run a whole series of such roots, from a (Viṣṇu) to dru (a tree).
Vārttika: ācāre 'vagalbhaklībahoḍebhyaḥ kvib vā vaktavyaḥ. "In the sense of behaving, kvip should be stated optionally after avagalbha, klība and hoḍa." Bhaṭṭoji's comment: because of the word vā "optionally", kyaṅ comes as well. One of his commentators gives that form: avagalbhāyate.
The three words end in the kṛt affix ac of the pac-list (pacādi). Bhaṭṭoji says that, together with kvip, this ac is stipulated to be anudātta and nasal. So its a is a marker (it), and a root marked with an anudātta vowel takes the ātmanepada (taṅ):
avagalbha + kvip → the final a is a nasal anudātta marker and is lost → avagalbh
→ the marker brings the ātmanepada; laṭ with śap → avagalbh + a + te → avagalbhate, "he behaves like a bold man".
Likewise klībate, hoḍate.
ām after a stem once polysyllabic. Bhaṭṭoji gives the liṭ forms avagalbhāṃcakre, klībāṃcakre, hoḍāṃcakre. Once the marker a is gone, klīb and hoḍ have one vowel. So does galbh, since ava is split off (the principle of SK 2667, says one commentator). Ām needs more than one vowel (anekāc). Why does it still come? Bhaṭṭoji says: ām comes even after a root that was polysyllabic. He proves it from the vārttika itself (ārambhasāmarthya, "the force of its being undertaken"). The vārttika must have a purpose. That purpose is not merely the ātmanepada form avagalbhate. The plain roots galbh and the rest could themselves carry the sense "behaves like", since a root has many meanings. So its purpose is this ām.
Applies when / blocked by. The vārttika names galbha with ava. So Mādhava and others hold that plain galbha, or galbha with another upasarga, takes only kyaṅ. Bhaṭṭoji finds it more proper to say that such stems simply get no ātmanepada. Bhaṭṭoji's commentators explain: the next vārttika gives kvip to every stem anyway. Only after ava is the marker attached.
Vārttika: sarvaprātipadikebhyaḥ kvib vā vaktavyaḥ. "Kvip should be stated optionally after all noun stems." Bhaṭṭoji adds two remarks. The previous vārttika only serves to attach the marker; kvip is merely repeated in it. And because this one names the prātipadika, the word "sup-final" of SK 2657 is not connected here. So no operation belonging to a pada applies.
Bhaṭṭoji's commentators illustrate. rājānati keeps its n (one commentator). Tad iva ācarati gives tatati, with no d for the t; tvag iva ācarati gives tvacati, with no k for the c (the other).
The forms. Bhaṭṭoji runs through many stems with kvip.
Kṛṣṇa. kṛṣṇa ivācarati, "he behaves like Kṛṣṇa":
kṛṣṇa + kvip → nothing of kvip remains → root kṛṣṇa
→ laṭ, śap, tip → kṛṣṇa + a + ti
→ the two *a*s become one by pararūpa (SK 191) → kṛṣṇati.
A (a name of Viṣṇu). a ivācarati, "he behaves like Viṣṇu": ati, ataḥ, anti.
For liṭ, SK 2306 would give ām after a root ending in an affix. But the vārttika on SK 2306 drops the word "affix" and says "polysyllabic" (anekācaḥ). The root a has one vowel, so no ām: au, atuḥ, uḥ. In liṭ the root is doubled. The first copy is the abhyāsa, such as bu in bubhāva. Bhaṭṭoji lists the steps for au:
a + ṇal (the ending a) → doubling → a a a
→ the root's a and the ending's a become one (SK 191) → a a
→ the a of the abhyāsa is lengthened (ata ādeḥ, SK 2248) → ā a
→ au replaces ṇal → ā au
→ vṛddhi joins ā and au → au.
In atuḥ and the rest, the ā is deleted by āto lopa iṭi ca (SK 2372), leaving the bare endings. One of Bhaṭṭoji's commentators names the rule for au: āta au ṇalaḥ (SK 2371).
Mālā "garland". mālevācarati: mālāti.
mālāṃcakāra. Laṅ amālāt. SK 252 does not delete the t. By association with ṅī, that rule deletes only su after āp.amālāsīt.Kavi "poet". kavir ivācarati: kavayati; āśīrliṅ kavīyāt.
akavayīt, without vṛddhi. In sici vṛddhiḥ (SK 2297) the word "dhātu" is carried forward and read as "what is purely a dhātu". A nominal root is not that. This is the view of Kaiyaṭa and others.akavāyīt.Vi "bird": vayati; liṭ vivāya, vivyatuḥ; luṅ avayīt or avāyīt.
Śrī: śrayati; liṭ śiśrāya, śiśriyatuḥ.
Pitṛ "father": pitarati; in the benedictive, with riṅ, pitriyāt.
Bhū "earth". bhūr ivācarati: bhavati; liṭ bubhāva; luṅ abhāvīt.
Three rules do not reach this root: gātisthā… (SK 2223), bhuvo vuk (SK 2174) and bhavater aḥ (SK 2181). By the maxim of "what is manifest" (abhivyakta), they take only the root bhū of the Dhātupāṭha. So no loss of sic, no vuk, and no a in the abhyāsa.
Dru "tree". drur ivācarati: dravati. ṇiśri… (SK 2312) does not give caṅ, for the same reason. So luṅ adrāvīt.
Watch out. These kvip roots look exactly like ordinary roots: bhavati, śrayati, dravati. But they are not the roots of the Dhātupāṭha. Rules that name a Dhātupāṭha root do not apply to them.
Rule. A long vowel replaces the penultimate vowel (upadhā) of a stem ending in a nasal. This happens before kvi, and before an affix that begins with a jhal and is marked with k or ṅ. Jhal is the pratyāhāra of all consonants except nasals and semivowels. Bhaṭṭoji's gloss: "a long vowel replaces the penultimate of a nasal-final stem, before kvi and before a jhal-initial kit or ṅit affix".
"The penultimate" comes from nopadhāyāḥ (SK 370) and "long" from ḍhralope (SK 174). Bhaṭṭoji's commentators give both. One of them adds examples before a jhal: śāntaḥ, dāntaḥ.
Example.
rājeva ācarati, "he behaves like a king":
rājan + kvip → this rule: the penultimate a is lengthened → rājān
→ the stem is not a pada (the vārttika above), so the n stays
→ laṭ with śap → rājān + a + ti → rājānati.
idam ivācarati, "he behaves like this": idam → idām → idāmati.
Bhaṭṭoji adds pathīnati (panthā iva, "like a road"), mathīnati and ṛbhukṣīṇati. For pathīnati, scholars discuss this: one of Bhaṭṭoji's commentators prefers pathenati, by the restriction inhan… (SK 356).
More kvip forms. Bhaṭṭoji continues the series here.
devati. The commentators explain: the stem is not a pada, so diva ut (SK 337) gives no u. Bhaṭṭoji holds that dyavati is proper, with ūṭh (SK 2561). Bhaṭṭoji's commentators give the steps: ūṭh replaces the v → di-ū → y for i → dyū → with śap, guṇa and av → dyavati.kati. Its liṭ is cakau, according to Haradatta. Mādhava gives caka instead. He relies on the statement ṇyallopau…: the loss of a sets aside vṛddhi before ṇal.svati; liṭ sasvau, sasva. Svāmī gives svāṃcakāra, with ām. Bhaṭṭoji calls that without basis in the sources (anākara). Bhaṭṭoji's commentators note that the Bhāṣya never drops the word "affix" from SK 2306. On that reading ām would be available after an affix-final stem; scholars discuss this.Watch out. The lengthening needs a nasal at the end of the stem. Rājan and idam end in a nasal. Kṛṣṇa does not, so kṛṣṇati keeps its short vowels.
abhṛśo bhṛśo bhavati: bhṛśa + kyaṅ → bhṛśāya → bhṛśāyate; lohita + kyaṣ → lohitāya → lohitāyati / lohitāyateThe sense changes again, from behaving to becoming (bhū). SK 2667 gives kyaṅ after a list of words beginning with bhṛśa "intense": "what was not intense becomes intense", bhṛśāyate. Under it Bhaṭṭoji draws an important principle from the word saṃgrāma. A first member that looks like an upasarga is split off when a root is being made. SK 2668 gives a fourth affix, kyaṣ, after lohita "red" and after stems in ḍāc. SK 2669 makes the parasmaipada optional after it: lohitāyati or lohitāyate.
Rule. Kyaṅ comes after bhṛśa and the rest, in the sense of becoming, when they do not end in cvi. Those of them that end in a consonant lose that consonant. Bhaṭṭoji's gloss: "let kyaṅ come, in the sense of bhavati, after bhṛśa and the rest, when they concern becoming what one was not; and let consonant-final ones among them lose [the consonant]".
Kyaṅ comes by anuvṛtti from SK 2665. One of Bhaṭṭoji's commentators gives this, and the other lists the group: bhṛśa, śīghra, manda, paṇḍita, durmanas, sumanas, unmanas and so on.
"Not ending in cvi." The words acveḥ "not cvi-final" exclude the affix cvi of chapter 42. That exclusion is a paryudāsa: it means "something like a cvi-final, but other than it". By its force the rule is limited to abhūtatadbhāva, "becoming what one was not". Bhaṭṭoji draws this out himself.
Example.
abhṛśo bhṛśo bhavati, "what was not intense becomes intense":
bhṛśa + su + kyaṅ → luk → bhṛśa-ya
→ lengthening before ya (SK 2298) → bhṛśāya
→ kyaṅ is marked with ṅ → ātmanepada → bhṛśāyate.
sumanas, "good-minded" (a bahuvrīhi, says one commentator):
sumanas + kyaṅ → this rule: the final s is lost → sumana-ya → sumanāya → sumanāyate.
Applies when / blocked by. No kyaṅ where nothing new comes to be. The counter-example is kva divā bhṛśā bhavanti (from the Bhāṣya, says one commentator). Bhaṭṭoji explains: "where by day are those that are intense at night, the stars and so on?" One commentator glosses "intense" as "bright". The stars are already bright; they do not become so.
The saṃgrāma principle. Bhaṭṭoji now proves a rule that this chapter will use often.
So the augment aṭ comes after such a member, before the rest:
sumanāya + laṅ → su is separated → aṭ before manāya → su-a-manāyata → svamanāyata
unmanāyate; laṅ ud-a-manāyata → udamanāyata.
Ava is separated in the same way under the kvip of SK 2665: avāgalbhata, avāgalbhiṣṭa.
Only cases of the same kind. Bhaṭṭoji limits the principle. The sign covers only cases like its own (sajātīyaviṣaya). The member is separated only where the whole upasarga is heard. It is not separated where a substitute has swallowed it.
oḍhāyitvā, "having behaved like him", with -tvā.unmanāyya and avagalbhya. There ud and ava are separated and compound with the gerund. In a compound the gerund takes lyap (-ya). One of Bhaṭṭoji's commentators names the rules: SK 761 for the compound, SK 3332 for lyap.Vārttika: usyomāṅkṣv āṭaḥ pratiṣedhaḥ. "Prohibition for āṭ before us, om and āṅ." Bhaṭṭoji: before these, the augment āṭ does not merge by pararūpa. It takes vṛddhi instead. Bhaṭṭoji cites this to prove the limit just stated.
usrīya + laṅ → āṭ + u → not pararūpa (as SK 2214 would give) → vṛddhi → ausrīyat.auṅkārīyat (from oṅkāra), and auḍhīyat (from oḍha): no pararūpa by SK 80.Bhaṭṭoji adds that, as laid down in the sixth book, this already follows from the word ca in āṭaś ca (SK 269), which prescribes vṛddhi again.
How does this prove the limit? Bhaṭṭoji's commentators explain. Suppose ā were separated even inside oḍha. Then aṭ would come after the ā, and no pararūpa could ever arise there. The prohibition for āṅ would be pointless. Since it is stated, ā inside oḍha is not separated.
Watch out. The principle needs the upasarga's full form. In svamanāyata the augment sits inside the word. In auḍhīyat it stands at the front, because the ā is no longer heard.
Rule. The affix kyaṣ comes after lohita "red" and the rest, and after a stem ending in ḍāc, in the sense of becoming. Bhaṭṭoji's gloss: "let kyaṣ come, in the sense of bhavati, after lohita and the rest and after a ḍāc-final". "In the sense of becoming" (bhuvi) comes from SK 2667, one of Bhaṭṭoji's commentators says.
"Not cvi-final" also comes from SK 2667. But it qualifies lohita alone. Bhaṭṭoji says this under the next sūtra, where he gives the forms.
Example.
lohita + kyaṣ → lohita-ya → lengthening before ya (SK 2298) → lohitāya
→ SK 2669 then gives lohitāyati or lohitāyate.
Watch out. Kyaṣ has no ṅ. So, unlike kyaṅ, it does not by itself give the ātmanepada. The next sūtra settles its endings.
Rule. After a stem ending in kyaṣ, the parasmaipada is optional. "Parasmaipada" comes from śeṣāt kartari parasmaipadam (SK 2159), one of Bhaṭṭoji's commentators says. In the other option the ātmanepada comes.
Example.
lohitāya → parasmaipada → lohitāyati; ātmanepada → lohitāyate. One of Bhaṭṭoji's commentators glosses: "what was not red becomes red".
What "not cvi-final" covers. Bhaṭṭoji explains the anuvṛtti of acveḥ into SK 2668. It gives the restriction to becoming what was not. That restriction qualifies the word lohita alone.
Vārttika: lohitaḍājbhyaḥ kyaṣvacanaṃ bhṛśādiṣv itarāṇi. "Kyaṣ is to be prescribed for lohita and for ḍāc-stems; the others [belong] among bhṛśa and the rest."
The k of kyaṣ. Bhaṭṭoji adds: one might think the k of kyaṣ is heard, like the k of kāmyac, since it is pronounced. No: the Bhāṣya rejects that k too. One of his commentators draws the result: SK 2659 and SK 2660 do not take in kyaṣ.
ḍāc-stems. paṭapaṭāyati, paṭapaṭāyate. Here ḍāc comes without kṛ, bhū or as, though SK 2128 asks for them. Bhaṭṭoji's reason: kyaṣ is prescribed after ḍāc-stems, and that prescription would be idle otherwise. One of his commentators gives the steps:
paṭat doubled for ḍāc → paṭat paṭat
→ the t and the following p give a single p (pararūpa) → paṭapaṭat
→ ḍāc is marked with ḍ, so the ṭi at is lost → paṭapaṭ-ā → paṭapaṭā
→ kyaṣ → paṭapaṭāya → paṭapaṭāyati / paṭapaṭāyate.
Applies when / blocked by. Some read a list here: lohita, śyāma, duḥkha, harṣa, garva, sukha and others. On that basis they give kyaṣ forms in both padas from śyāma and the rest. Bhaṭṭoji rejects this: it conflicts with the Bhāṣya and the vārttika. Those words take only the kyaṅ of SK 2667: śyāmāyate, "becomes dark". Inside the derived word, duḥkha and the like mean "one possessing that quality".
By the liṅgaviśiṣṭa paribhāṣā, kyaṣ also comes after the feminine lohinī: lohinīyati, lohinīyate. One of Bhaṭṭoji's commentators contrasts kyaṅ: it would give only the ātmanepada, with the stem treated as masculine.
Watch out. Only lohita and ḍāc-stems have both padas. lohitāyati is right; śyāmāyati is not.
śabdaṃ karoti: śabda + kyaṅ → śabdāya → śabdāyateFive sūtras give kyaṅ after named words, each in a sense of its own. SK 2670 takes kaṣṭa in the fourth case: "he is eager for what is hard", kaṣṭāyate. SK 2671 to SK 2674 take their words as objects (karman): "it chews the cud" (romanthāyate), "it gives off vapour" (bāṣpāyate), "he makes a sound" (śabdāyate), "he is aware of [his] happiness". Kyaṅ is carried forward from SK 2665 through all five. The running example is śabda "sound": śabdaṃ karoti, "he makes a sound", becomes śabdāyate.
Rule. Kyaṅ comes after the word kaṣṭa "hard" in the fourth case, in the sense of zeal (utsāha). Bhaṭṭoji's gloss: "let kyaṅ come after the word kaṣṭa, ending in the fourth case, in the sense of zeal".
Why kyaṅ and not kyaṣ, the nearer affix? Bhaṭṭoji's commentators answer: kyaṅ in SK 2665 is marked svarita, the sign that a word is to be carried forward; kyaṣ is not.
Example.
kaṣṭāya kramate, "he sets about what is hard":
kaṣṭa + ṅe (the fourth case-ending) + kyaṅ → luk → kaṣṭa-ya
→ lengthening (SK 2298) → kaṣṭāya → ātmanepada → kaṣṭāyate.
Bhaṭṭoji gives the meaning: "he is eager to commit sin".
One of Bhaṭṭoji's commentators explains the gloss. No one is eager for pain itself, so sin, the means to it, is meant. The other gives the rules behind the phrase. The fourth case comes by SK 581, and kramate takes the ātmanepada by vṛttisargatāyaneṣu kramaḥ (SK 2711, taught later).
Vārttika: satrakakṣakaṣṭakṛcchragahanebhyaḥ kaṇvacikīrṣāyām iti vaktavyam. "It should be stated: after satra, kakṣa, kaṣṭa, kṛcchra and gahana, in the sense of wishing to do kaṇva." Bhaṭṭoji explains:
Forms: satrāyate, kakṣāyate, and so on.
Watch out. The sūtra needs the fourth case (kaṣṭāya). The vārttika takes the second case (kaṣṭam and the rest). One of Bhaṭṭoji's commentators adds that only kaṣṭa has an analysis in its own words, kaṣṭāya kramate.
Rule. Kyaṅ comes after romantha "cud" and tapas "austerity", as objects, in the senses of turning over and practising, respectively. Bhaṭṭoji's gloss: "let kyaṅ come after romantha and tapas as objects, in the senses of turning over and practising, respectively".
Example.
romanthaṃ vartayati, "it turns over the cud":
romantha + am + kyaṅ → luk → romantha-ya → romanthāya → romanthāyate.
Vārttika: hanucalana iti vaktavyam. "It should be stated: when the jaw moves." Bhaṭṭoji: "that is, in chewing again what has been chewed, after drawing it up". So not here: kīṭo romanthaṃ vartayati, "the insect rolls its romantha". There romantha is matter that has come out of the anal region. Kaiyaṭa takes the meaning as "it eats that". The author of the Nyāsa and Haradatta take it as "it makes it round".
Vārttika: tapasaḥ parasmaipadaṃ ca. "And after tapas, the parasmaipada." tapaś carati, "he practises austerity" → tapasyati.
One of Bhaṭṭoji's commentators explains. Tapas takes kyaṅ from the sūtra. The vārttika then gives the parasmaipada, setting aside the ātmanepada that the ṅ would bring. He adds that the s does not become ru: the stem does not end in n, so it is not a pada (SK 2659).
tapas + kyaṅ → tapas-ya (no pada, s stays) → parasmaipada by the vārttika → tapasyati.
Watch out. tapasyati has a kyaṅ in it, yet parasmaipada endings. The vārttika overrides the ṅ.
Rule. Kyaṅ comes after bāṣpa "vapour" and ūṣman, as objects, in the sense of emitting (udvamana). Bhaṭṭoji's gloss: "let kyaṅ come after these two as objects". One of his commentators explains "as objects": karmaṇaḥ comes forward from SK 2671.
Example.
bāṣpam udvamati, "it gives off vapour":
bāṣpa + am + kyaṅ → luk → bāṣpa-ya → bāṣpāya → bāṣpāyate.
ūṣman ends in n, so before kyaṅ it is a pada (SK 2659) and loses the n:
ūṣman-ya → ūṣma-ya → ūṣmāya → ūṣmāyate.
Vārttika: phenāc ceti vācyam. "It should be stated: also after phena." phenāyate, "it gives off foam", in the gloss of one of Bhaṭṭoji's commentators.
Rule. Kyaṅ comes after śabda "sound", vaira, kalaha, abhra, kaṇva "sin" and megha, as objects, in the sense of karoti "makes". Bhaṭṭoji's gloss: "let kyaṅ come after these as objects, in the sense of karoti". One of Bhaṭṭoji's commentators gives the anuvṛtti of "as objects" from SK 2671.
Example.
śabdaṃ karoti, "he makes a sound":
śabda + am + kyaṅ → luk → śabda-ya → śabdāya → śabdāyate.
Applies when / blocked by. One of Bhaṭṭoji's commentators calls this kyaṅ an exception (apavāda) to the ṇic of tat karoti (the list-rule under SK 2573). Bhaṭṭoji reports that the Nyāsa also allows that ṇic sometimes: śabdayati. The same commentator reads this report as a reservation, since the Bhāṣya does not give it; scholars discuss this.
Vārttika: sudinadurdinanīhārebhyaś ca. "And after sudina, durdina and nīhāra." One of Bhaṭṭoji's commentators supplies "kyaṅ, in the sense of karoti". Bhaṭṭoji's form: sudināyate.
Watch out. With these six words, kyaṅ is the regular affix for "makes". The ṇic form śabdayati rests only on the Nyāsa.
Rule. Kyaṅ comes after sukha "happiness" and the rest, as objects, in the sense of awareness (vedanā). The condition: the happiness and the rest belong to the agent of the awareness himself. Bhaṭṭoji's gloss: "let kyaṅ come… provided that sukha and the rest belong to the agent of the awareness himself".
Both commentators say vedanā means knowing. One adds that kartṛ "agent" is a separate word whose sixth-case ending has been dropped. One commentator lists the group: sukha, duḥkha, tṛpra, kṛcchra and others.
Example. Bhaṭṭoji gives the phrase that the derived verb replaces: sukhaṃ vedayate, "he is aware of [his] happiness". Vedayate is the curādi vid "to know", one of his commentators says. Bhaṭṭoji gives no derived form here.
Applies when / blocked by. Why the word "agent"? Bhaṭṭoji's counter-example: parasya sukhaṃ vedayate, "he is aware of another's happiness". One of his commentators explains. The knower is not the one in whom the happiness lies, so only the phrase is used.
Watch out. The happiness must be the knower's own. Awareness of someone else's happiness gives no verb.
namas + kyac → namasya → namasyati devān (= pūjayati); muṇḍaṃ karoti: muṇḍa + ṇic → muṇḍayatiThe word karaṇe, "in [some] action", runs on from SK 2673. Three last sūtras use it with three different affixes. SK 2675 gives kyac to namas, varivas and citra: "he worships the gods", namasyati devān. SK 2676 gives ṇiṅ to puccha "tail" and two other words: "it raises its tail", utpucchayate. SK 2677 gives ṇic to muṇḍa "shaven" and nine others: "he makes [someone] shaven", muṇḍayati. Under SK 2677 Bhaṭṭoji also asks why these words are named at all, since chapter 53 already gave ṇic after any noun.
Rule. Kyac comes after namas, varivas and citra, in particular actions: worship, service and wonder. Bhaṭṭoji's gloss: "because karaṇe is carried forward, [it comes] in particular actions: in worship, service and wonder".
Citra is cited with the marker ṅ (citraṅ). Bhaṭṭoji's commentators say the ṅ is for the ātmanepada. One of them asks why this is kyac and not kyaṅ. His answer: so that ī comes by SK 2658.
Example.
namasyati devān, "he worships the gods":
namas + kyac → namas-ya (the stem does not end in n, so it is no pada, and the s stays) → namasyati.
varivasyati gurūn, "he serves the teachers".
citra + kyac → ī for a (SK 2658) → citrīya → the ṅ of citraṅ gives the ātmanepada → citrīyate, "he is amazed". Others take it as "he causes amazement".
Applies when / blocked by. Why the second case in devān? One of Bhaṭṭoji's commentators answers: a case-ending that expresses a kāraka role is stronger than the fourth case that the word namas would otherwise call for. A kāraka is a participant in the action, such as the object.
Watch out. Citra has kyac, not kyaṅ, and yet its verb is ātmanepada. The ṅ sits on the stem citraṅ, not on the affix.
Rule. The affix ṇiṅ comes after puccha "tail", bhāṇḍa "vessel" and cīvara "robe", for particular actions. Bhaṭṭoji gives no gloss of the sūtra. He goes straight to three vārttikas, which name the actions. One of his commentators says that "particular actions" comes from karaṇe (SK 2673).
Why the marker ṅ? The same commentator gives two reasons. It lets ṇer aniṭi (SK 2313), which names ṇi, cover ṇiṅ along with ṇic. And it makes the ātmanepada invariable.
Vārttika: pucchād udasane vyasane paryasane ca. "After puccha, in the senses of throwing up, throwing variously and throwing around." Bhaṭṭoji defines vyasana: throwing up in various or contrary ways. Forms: utpucchayate, "it throws up its tail"; vipucchayate; paripucchayate.
Vārttika: bhāṇḍāt samācayane. "After bhāṇḍa, in the sense of collecting together." saṃbhāṇḍayate: "he collects vessels together", that is, he heaps them up.
The aorist is samababhāṇḍata. Note where the augment a stands: after sam, not before it. One of Bhaṭṭoji's commentators explains. Sam has the form of an upasarga, so it is separated (the principle of SK 2667). The augment comes before bhāṇḍa.
Vārttika: cīvarād arjane paridhāne ca. "After cīvara, in the senses of acquiring and putting on." saṃcīvarayate bhikṣuḥ: "the mendicant acquires robes, or puts them on".
Watch out. The ṅ lets rules that name ṇi, such as SK 2313, take in ṇiṅ too. The difference is the pada: a ṇiṅ verb is always ātmanepada.
Rule. Ṇic comes after muṇḍa, miśra, ślakṣṇa, lavaṇa, vrata, vastra, hala, kala, kṛta and tūsta, in the sense of kṛ "to make". Bhaṭṭoji's gloss: "in the sense of kṛñ". One of his commentators says karaṇe comes forward from SK 2673, and Bhaṭṭoji's words fill it out.
Example.
muṇḍaṃ karoti, "he makes [someone] shaven":
muṇḍa + ṇic → the final a goes before ṇi → muṇḍi → laṭ, śap, guṇa and ay → muṇḍayati.
Vārttika: vratād bhojanatannivṛttyoḥ. "After vrata, in the senses of eating and of abstaining from it." payo vratayati, "he eats milk". Both of Bhaṭṭoji's commentators gloss it so. And śūdrānnaṃ vratayati, "he abstains from food given by a śūdra".
Vārttika: vastrāt samācchādane. "After vastra, in the sense of covering completely." saṃvastrayati, "he covers [himself] completely with a garment". One of Bhaṭṭoji's commentators notes that this vārttika is not found in the Bhāṣya.
Vārttika: halyādibhyo grahaṇe. "After hali and the rest, in the sense of taking." With it Bhaṭṭoji gives a statement: "for hali and kali, ending in a is given by nipātana". So the i of hali and kali becomes a before ṇic.
haliṃ kaliṃ vā gṛhṇāti, "he takes hold of a hali or a kali" → halayati, kalayati. Bhaṭṭoji: a large plough (hala) is a hali.
Now the aorist. Bhaṭṭoji says: vṛddhi comes first, being the later rule (paratva). Even so, the treatment "like iṣṭhan" (see group H) deletes an ak: the ā produced by vṛddhi. So the stem has lost an ak:
hala + ṇic → vṛddhi of the final a → halā-i → the ṭi ā is lost, as before iṣṭhan → hal-i → halayati
luṅ: hal-i + caṅ → doubling → the stem has lost an ak, so no sanvadbhāva (SK 2316) and no lengthening (SK 2318) → ajahalat; likewise acakalat.
Why make hali and kali end in a? One of Bhaṭṭoji's commentators explains. With i kept, vṛddhi would give ai, and losing ai is not losing an ak. The other commentator sees in the nipātana a sign that the loss of the ṭi follows vṛddhi.
More forms.
kṛtayati.vitūstayati. Tūsta means hair, say some; matted hair, say others; sin, say yet others.Why are these words named? Bhaṭṭoji raises the question himself.
muṇḍayati māṇavakam, "he shaves the boy"; miśrayaty annam, "he mixes the food"; ślakṣṇayati vastram, "he makes the cloth smooth"; lavaṇayati vyañjanam, "he salts the relish".One of Bhaṭṭoji's commentators gives the contrast for an ordinary stem. With the qualifier ramaṇīyam beside paṭum, one says only ramaṇīyaṃ paṭum ācaṣṭe, never paṭayati. The other commentator reports a view that the mention is elaboration only; scholars discuss this.
Satya with āpuk:
satyaṃ karoty ācaṣṭe vā, "he makes or declares [something] true":
satya + āpuk + ṇic → satyāp-i → satyāpayati.
One of Bhaṭṭoji's commentators notes that there is no loss of the ṭi here; the prescription of āpuk forbids it.
Vārttika: arthavedayor apy āpug vaktavyaḥ. "The augment āpuk should be stated also for artha and veda." arthāpayati, vedāpayati.
Watch out. Ṇic after a noun normally needs the noun to stand alone. muṇḍayati māṇavakam is correct only because muṇḍa is named in this sūtra.
pṛthum ācaṣṭe: pṛthu + ṇic → (iṣṭhavat: ra for ṛ, ṭi-lopa) prath-i → prathayati; luṅ apiprathat (ṭi-lopa after vṛddhi) / apaprathat (before)The rest of SK 2677 is a long run of forms. They come from the list-rule of chapter 53, prātipadikād dhātvarthe bahulam iṣṭhavac ca. It gives ṇic after any noun stem in a verbal sense, and the stem is treated "like before iṣṭhan" (iṣṭhavat). Iṣṭhan is the superlative affix of chapter 41. So whatever a stem undergoes before iṣṭhan, it undergoes before ṇi too. This extension of operations is an atideśa. The running example is pṛthu "broad". Before iṣṭhan its ṛ becomes ra and its ṭi is lost. So "he calls it broad" is prathayati.
The stems of the satyāpa list. Bhaṭṭoji first gives ṇic forms of the words in SK 2563. One of his commentators says the ṇic here comes by prātipadikād dhātvarthe.
vipāśayati.rūpayati.upavīṇayati.anutūlayati. Bhaṭṭoji: "he strokes the tip of a blade of grass with cotton".upaślokayati.abhiṣeṇayati, with ṣ by SK 2270. Laṅ abhyaṣeṇayat, with ṣ by SK 2276. Desiderative abhiṣiṣeṇayiṣati, with ṣ by SK 2277.anulomayati.tvacayati. Bhaṭṭoji derives tvaca from the root tvaca "to cover" with the affix gha. One of his commentators says why: from tvac, the loss of the ṭi would give tvayati.saṃvarmayati. varṇaṃ gṛhṇāti, "he takes colour" → varṇayati. cūrṇair avadhvaṃsate, "he sprinkles [it] with powders" → avacūrṇayati.Treatment as masculine. Bhaṭṭoji: "because of the extension 'like iṣṭhan', treatment as masculine (puṃvadbhāva) and the rest take effect".
etayati. One of Bhaṭṭoji's commentators gives the steps: the feminine affix and the n that came with it both go, and then the ṭi is lost.dāradayati. Both commentators explain why Bhaṭṭoji adds this one. Etayati would follow from the loss of the ṭi alone. But darad becomes dārada again only by the masculine treatment. Without it, one adds, the form would be darayati.ra for ṛ. pṛthum ācaṣṭe, "he calls [it] broad":
pṛthu + ṇic → as before iṣṭhan: ra for the ṛ → prathu
→ the ṭi u is lost → prath-i → prathayati.
The aorist depends on the order of two steps. Bhaṭṭoji allows both orders:
apiprathat.apaprathat.Which order is primary, scholars discuss this.
The same with other ṛ-stems:
mradayati; luṅ amimradat, amamradat.bhraśayati, kraśayati, draḍhayati; luṅ ababhraśat, acakraśat, adadraḍhat. One of Bhaṭṭoji's commentators explains why there is only one aorist each. The lost vowel is a or its vṛddhi ā, an ak either way.parivraḍhayati; luṅ paryavavraḍhat.ūḍhi and ūḍha. ūḍhim ākhyat, "he spoke of an ūḍhi" → aujiḍhat. The changes that produced ḍh from h + ti are asiddha, counted as not done. So the syllable doubled is hti, not ḍhi, and the abhyāsa's h becomes j. The principle that asiddha does not hold for doubling has been said to be not invariable. Others double ḍhi: auḍiḍhat.
ūḍham ākhyat → aujaḍhat, auḍaḍhat.
Why a in the abhyāsa, and not i? Bhaṭṭoji points to the sign in oḥ puyaṇjy apare (SK 2577). Its wording shows that, while doubling is due, no substitute replaces a vowel before ṇi. He had stated this under ūnayati (chapter 53). Bhaṭṭoji's commentators explain: were the ṭi lost first, the doubled syllable would not keep its a.
One-vowel stems keep their form. By prakṛtyaikāc (SK 2010), a one-vowel stem does not lose its ṭi:
sva + ṇic → no loss of the ṭi → vṛddhi → svā-i → the augment puk → svāp-i → svāpayati.
tvam and aham. tvāṃ māṃ vā ācaṣṭe, "he speaks of you" or "of me". Tva and ma replace the part of yuṣmad and asmad up to m (SK 1373). Two derivations follow.
tvāpayati, māpayati.tvādayati, mādayati.tva + ad + ṇic → pararūpa first, being antaraṅga → tvad → one vowel, so SK 2010 keeps the ṭi → the penultimate a takes vṛddhi → tvād-i → tvādayati.
Bhaṭṭoji adds a warning. Do not think the Bhāṣya rejects SK 2010. It only shows that certain examples, such as preṣṭha, can be explained another way.
For two: yuvām āvāṃ vā ācaṣṭe → yuṣmayati, asmayati.
śvan "dog". śvānam ācaṣṭe → śāvayati.
śāvayati.Others say nas taddhite is not extended here, since with iṣṭhan only ṭeḥ is seen. They give śunayati. Scholars discuss this.
vidvas "learned". vidvāṃsam ācaṣṭe, three views:
vidvayati: no saṃprasāraṇa, by the aṅgavṛtta paribhāṣā (some). That principle says: once an operation on a stem has been done, a further operation on it does not follow;vidāvayati: saṃprasāraṇa, vṛddhi and āv (others);vidayati: saṃprasāraṇa first, being nitya; then pūrvarūpa, being antaraṅga; then the loss of the ṭi (yet others).One of Bhaṭṭoji's commentators keeps the first; scholars discuss this.
The añc stems. Here the first member is separated, by the saṃgrāma principle (SK 2667). Ṇic is added to the one-vowel ac alone, so SK 2010 keeps its ṭi. Bhaṭṭoji's commentators explain this.
udīcayati; luṅ udaicicat.pratīcayati; luṅ pratyacicat. With prakṛtibhāva by SK 91, also pratiacicat.samīcayati; luṅ samyacicat, samiacicat.In udaicicat the augment comes after ud, and ci is doubled. Bhaṭṭoji's commentators add that the ṇi is not lost there before doubling.
tiryañc. tiryañcam ācaṣṭe → tirāyayati.
tiri → vṛddhi and āy → tirāy-i → tirāyayati.atitirāyat: the stem has lost an ak, so its penultimate is not shortened before caṅ.Bhaṭṭoji lists the other -añc stems without derivation:
sadhrāyayati; luṅ asasadhrāyat;aviviṣvadrāyat;devadrāyayati; luṅ adidevadrāyat;adadadrāyat;adamuāyayati; luṅ ādadamuāyat;amumuāyayati; with caṅ, āmumuāyayat.More aorists. Bhaṭṭoji gives a set of luṅ forms:
bhāvayati; abībhavat.abubhravat. Bhaṭṭoji's commentators say why the u stays. SK 2577 needs a labial or semivowel followed by a right after the abhyāsa, and that is not so here.aśiśrayat; gām → ajūgavat; rāyam → arīrayat; nāvam → anūnavat.svāśaśvat. Bhaṭṭoji's commentators explain: su is separated, the second syllable of aśva is doubled, and the augment āṭ comes.svar. Bhaṭṭoji cites the statement avyayānāṃ bhamātre ṭilopaḥ: an indeclinable loses its ṭi wherever it is merely bha. So svar → svayati. Its aorists are asasvat and asisvat. Bhaṭṭoji's commentators explain the two. In asasvat, doubling is due, so no vowel-substitute comes first, and svar is doubled. In asisvat, the ṭi is lost first, and svi is doubled with the ṇi.
bahu. bahūn ācaṣṭe, "he speaks of many" → bhāvayati, with bhū as before iṣṭhan. Others give bahayati; one of Bhaṭṭoji's commentators rejects it, and scholars discuss this.
Loss of vin and matup. By vinmatoḥ luk, as before iṣṭhan:
srajayati. There is no vṛddhi, Bhaṭṭoji says, because the operation is enjoined through a technical name (saṃjñāpūrvaka).śrayayati; luṅ aśiśrayat.payasayati. The ṭi is not lost, because the luk, its exception, has already applied.Loss from the semivowel on. sthūla "stout" → sthavayati; dūra "far" → davayati.
How then is the usage dūrayaty avanate vivasvati explained? Bhaṭṭoji: from a stem dūrāt, "one that goes far". The phrase means "when the sun is making itself a dūrāt".
Substitutes. Each stem takes the substitute it takes before iṣṭhan:
yavayati, or with kan by SK 2019, kanayati;nedayati; bāḍha "firm" → sādhayati;praśasyayati. Here śra and jya do not come, because the upasarga pra is separated;jyāpayati, varṣayati;prāpayati; sthira "steady" → sthāpayati; sphira "abundant" → sphāpayati;varayati, vārayati;baṃhayati; guru "heavy" → garayati;trāpayati. One of Bhaṭṭoji's commentators reads trapayati; scholars discuss this;drāghayati; vṛndāraka "eminent" → vṛndayati.Here ends the procedure for nominal roots: "iti nāmadhātuprakriyā".
Watch out. Before ṇic by this list-rule, do not just drop the final vowel. First ask what the stem does before iṣṭhan. Sva keeps its vowel, pṛthu takes ra, yuvan loses van, priya becomes pra. The same happens before ṇi.
putrīya, kṛṣṇāya, muṇḍi. Bhaṭṭoji's section on them is the nāmadhātu-prakriyā.gavyitā).samidhitā still blocks guṇa (SK 2660).aśanāyati, udanyati, dhanāyati).aśvasyati, dadhisyati.putrakāmyati.kṛṣṇāyate, bhṛśāyate, śabdāyate.kumārāyate, etayati.kṛṣṇati (vārttikas under SK 2665).avagalbhāṃcakre).rājānati, tatati).bubhāva.mālāti, lohinīyati.bhavati from bhū "earth").dyavati.unmanāyya.paṭapaṭā.ajahalat, apaprathat.muṇḍayati māṇavakam.satyāpayati, arthāpayati.tvādayati).śāvayati).vidvayati, tirāyayati).srajayati).svāpayati.| SK | What it does | Example |
|---|---|---|
| 2657 | kyac after a sup-final object wished for oneself; the inner case-ending undergoes luk | putrīyati |
| 2658 | ī for a before kyac | putrīyati; gavyati, nāvyati (SK 63) |
| 2659 | before kyac and kyaṅ, only an n-final stem is a pada | rājīyati; vācyati, samidhyati, tvadyati, gīryati, divyati |
| 2660 | kyac and kyaṅ optionally lost after a consonant before an ārdhadhātuka; vārttika: no kyac from m-final bases or indeclinables | samidhitā / samidhyitā; kim icchati |
| 2661 | aśanāya, udanya, dhanāya ready-made for hunger, thirst, greed | aśanāyati against aśanīyati |
| 2662 | augment asuk for aśva, kṣīra, vṛṣa, lavaṇa before kyac; vārttikas extend suk / asuk | aśvasyati vaḍavā, kṣīrasyati, dadhisyati / dadhyasyati |
| 2663 | kāmyac in the domain of kyac; no m / indeclinable bar | putrakāmyati, kiṃkāmyati |
| 2664 | kyac after an object as standard, in the sense of behaving; vārttika: also a locus | putrīyati chātram, prāsādīyati kuṭyām |
| 2665 | kyaṅ after an agent as standard; s lost with it, fixed by vārttika; puṃvat by SK 837 | kṛṣṇāyate, ojāyate, yaśāyate / yaśasyate, kumārāyate, pācikāyate |
| 2665 (vārttikas) | kvip after avagalbha, klība, hoḍa (ātmanepada), and after any noun stem; no pada operations | avagalbhate, avagalbhāṃcakre; kṛṣṇati, ati / au, mālāti, kavayati, bhavati |
| 2666 | long penultimate in a nasal-final stem before kvi and a jhal-initial kit / ṅit affix | rājānati, idāmati; śāntaḥ |
| 2667 | kyaṅ after bhṛśa etc. for becoming what was not; final consonant lost; the saṃgrāma sign | bhṛśāyate, sumanāyate, udamanāyata; oḍhāyitvā; ausrīyat |
| 2668 | kyaṣ after lohita and ḍāc-stems for becoming | lohitāya- |
| 2669 | parasmaipada optional after kyaṣ | lohitāyati / lohitāyate, paṭapaṭāyati; but śyāmāyate only |
| 2670 | kyaṅ after fourth-case kaṣṭa for zeal; vārttika: satra etc. for wishing to sin | kaṣṭāyate, satrāyate |
| 2671 | kyaṅ after objects romantha, tapas for chewing, practising; vārttikas: jaw movement; parasmaipada for tapas | romanthāyate, tapasyati |
| 2672 | kyaṅ after objects bāṣpa, ūṣman for emitting; vārttika: phena | bāṣpāyate, ūṣmāyate, phenāyate |
| 2673 | kyaṅ after objects śabda, vaira … in the sense of karoti; vārttika: sudina etc. | śabdāyate (Nyāsa also śabdayati), sudināyate |
| 2674 | kyaṅ after objects sukha etc. for the agent's own awareness | sukhaṃ vedayate; not parasya sukham |
| 2675 | kyac after namas, varivas, citraṅ for worship, service, wonder | namasyati devān, varivasyati, citrīyate |
| 2676 | ṇiṅ after puccha, bhāṇḍa, cīvara; always ātmanepada | utpucchayate, saṃbhāṇḍayate / samababhāṇḍata, saṃcīvarayate |
| 2677 | ṇic after muṇḍa … tūsta in the sense of kṛ, even when the word depends on another; vārttikas for vrata, vastra, hali; āpuk | muṇḍayati māṇavakam, vratayati, halayati / ajahalat, satyāpayati, arthāpayati |
| 2677 (iṣṭhavat forms) | ṇic after any noun, the stem treated as before iṣṭhan | prathayati / apiprathat / apaprathat, svāpayati, tvādayati, śāvayati, tirāyayati, jyāpayati |
The affixes that make a nominal root:
| Affix | Pada | Sense | Example |
|---|---|---|---|
| kyac | parasmaipada (but citrīyate, from citraṅ) | wishing for oneself; behaving towards as; worship and the like | putrīyati, namasyati |
| kāmyac | parasmaipada | wishing for oneself | putrakāmyati |
| kyaṅ | ātmanepada (ṅ) (but tapasyati, by vārttika) | behaving like; becoming (bhṛśādi); special senses | kṛṣṇāyate, bhṛśāyate, śabdāyate |
| kyaṣ | either (SK 2669) | becoming (lohita, ḍāc-stems) | lohitāyati / lohitāyate |
| kvip | parasmaipada; ātmanepada with the marker of avagalbha etc. | behaving like | kṛṣṇati, rājānati, avagalbhate |
| ṇiṅ | ātmanepada (ṅ) | particular acts with puccha etc. | utpucchayate |
| ṇic | as for any ṇic-stem (chapter 53) | making; any verbal sense, stem as before iṣṭhan | muṇḍayati, prathayati |
The stem changes before ṇic by iṣṭhavat, with the form each gives:
| Change | Stem | Form |
|---|---|---|
| treated as masculine | enī, darad | etayati, dāradayati |
| ra for ṛ, ṭi lost | pṛthu, mṛdu, dṛḍha | prathayati, mradayati, draḍhayati |
| ṭi kept (SK 2010) | sva, tvad, ac in udañc | svāpayati, tvādayati, udīcayati |
| ṭi lost by SK 679 | śvan | śāvayati (others śunayati) |
| vin / matup lost | sragvin, śrīmat, payasvinī | srajayati, śrayayati, payasayati |
| part from the semivowel lost | sthūla, dūra, yuvan | sthavayati, davayati, yavayati |
| substitutes | antika, bāḍha, vṛddha, priya, sthira, guru, dīrgha | nedayati, sādhayati, jyāpayati, prāpayati, sthāpayati, garayati, drāghayati |
| bhū for bahu | bahu | bhāvayati |