PREMABASESiddhānta-kaumudī
Learner edition · Prakaraṇa 57

57. Tiṅanta yaṅluk-prakaraṇa — Verbs: the intensive whose yaṅ is deleted

Chapter 56 built the intensive with the affix yaṅ: bobhūyate, śāśayyate. This chapter takes the same intensive and deletes its yaṅ. What is left is still doubled, and it is still a root, but it is conjugated like a root of the ad class of chapter 45: parasmaipada only, with its vikaraṇa śap deleted by luk. Bhaṭṭoji calls the section the yaṅluganta-prakriyā, "the working of what ends in a yaṅ removed by luk". After this chapter you will be able to build bobhavīti and bobhoti from bhū step by step, and to say why the first has ī and the second has none. You will know the conjugation of seventeen roots under SK 2651 (momudīti / momotti, jaṅgamīti / jaṅganti, jaṅghanīti, sāsvapīti …), and the forms of the ṛ-roots under SK 2652–2653: most come in three shapes (varvarti / varivarti / varīvarti, carkarti), and ṛ itself has two (ararti / ariyarti). You will see ūṭh at work in māmoti (SK 2654), the loss of v after r in totorti (SK 2655), and the sūtra that bars guṇa before a deleting ārdhadhātuka (SK 2656). Above all, you will learn which rules reach this doubled stem and which do not, and why. The chapter has seven sūtras, SK 2650 to SK 2656.

Before you start

A. The yaṅluganta itself: yaṅ deleted, the root still doubled, a parasmaipada adādi-type stem — running example: bhū + yaṅ → (luk of yaṅ, unconditioned and antaraṅga, so first) bhū → (doubling: still yaṅ-final by pratyayalakṣaṇa) bhū-bhū → (abhyāsa: bu, guṇa by guṇo yaṅlukoḥ) bobhū → laṭ, parasmaipada, śap deleted (adādi) → bobhoti / bobhavīti (2651)

One sūtra opens the chapter. SK 2650 deletes yaṅ by luk: always before the suffix ac, and "variously" without it. Bhaṭṭoji then sets out the order of the steps that turn bhū + yaṅ into a conjugable stem, and he says what that stem may and may not take. The running example is bhū "to be". Its yaṅ is deleted, the root is still doubled, the abhyāsa takes guṇa, and the stem bobhū is conjugated in parasmaipada, its śap deleted by luk.

SK 2650 · यङोऽचि च · yaṅo 'ci ca (A 2.4.74)

यङोऽचि च
yaṅo'ci ca · A 2.4.74
यङोऽच्प्रत्यये लुक् स्याच्चकारात्तं विनापि बहुलं लुक् स्यात् । अनैमित्तिकोऽयमन्तरङ्गत्वादादौ भवति । ततः प्रत्ययलक्षणेन यङन्तत्वाद्द्वित्वम् । अभ्यासकार्यम् । धातुत्वाल्लडादयः । शेषात्कर्तरि - SK 2149 इति परस्मैपदम् । अनुदात्तङित - SK 2157 इति तु न । ङित्त्वस्य प्रत्ययाप्रत्ययसाधारणत्वेन प्रत्ययलक्षणाप्रवृत्तेः । यत्र हि प्रत्ययस्यासाधारणं रूपमाश्रीयते तत्रैव तत् । अत एव सुदृषत्प्रासाद इत्यत्र अत्वसन्तस्य - SK 425 इति दीर्घो न । येऽपि स्पर्धशीङादयोऽनुदात्तङितस्तेभ्योपि न । अनुदात्तङित - SK 2157 । इत्यनुबन्धनिर्देशात् । तत्र च [(परिभाषा - ) श्तिपा शपा...] इति निषेधात् । अत एव श्यन्नादयो न । गणेन निर्देशात् । किंतु शबेव । चर्करीतं चेत्यादादौ पाठाच्छपो लुक् ॥

Rule. The affix yaṅ undergoes luk when the suffix ac follows. By the word ca "and", it also undergoes luk "variously" (bahulam) even without that suffix. Bahulam means that the rule applies in some places and not in others.

The sūtra is short, and two words come from elsewhere by anuvṛtti. "Luk" comes from SK 1276 (ṇyakṣatriyārṣañito yūni luk aṇiñoḥ). "Bahulam" is drawn in by the ca from bahulaṃ chandasi, a rule of the Vedic section (SK 3400). Bhaṭṭoji's gloss takes aci as the suffix ac, a kṛt affix. It does not mean the pratyāhāra ac "any vowel". His commentators give the reason, association (sāhacarya): it stands beside yaṅ, which is a suffix too.

Two kinds of luk. Bhaṭṭoji names the kind that occurs without ac anaimittika, "unconditioned". Nothing that follows causes it. Being antaraṅga, it applies first. An antaraṅga operation is one that depends on less, or on what lies nearer the root; it takes effect before its rivals.

Bhaṭṭoji's commentators supply the other kind. Both call the luk before the suffix ac naimittika, "conditioned": the ac causes it. Both tie it to SK 2656: the guṇa the suffix ac would bring is barred, because the ac causes the deletion of the yaṅ. One of them, the Bālamanoramā, adds that this luk is therefore bahiraṅga, the opposite of antaraṅga. So there doubling comes first, and the luk of yaṅ after it.

Example (without ac). Bhaṭṭoji gives the order of steps. The root is bhū "to be".

bhū + yaṅ (the intensive affix of chapter 56)

→ this rule: luk of yaṅ, unconditioned, so antaraṅga, so first → bhū

→ the yaṅ is gone, but by pratyayalakṣaṇa (SK 262) the root still counts as ending in yaṅ

→ so doubling by sanyaṅoḥ (SK 2395) → bhū bhū

→ the operations on the abhyāsa (abhyāsakārya): shortening (SK 2180) → bhu bhū; b for bh (SK 2182) → bu bhū; guṇa of the abhyāsa by guṇo yaṅlukoḥ (SK 2630) → bobhū

→ the whole is a dhātu (SK 2304), so the lakāras come: laṭ

→ parasmaipada by śeṣāt kartari parasmaipadam (SK 2159): the ending tip → bobhū + ti

→ śap comes (SK 2167) and undergoes luk, because carkarītaṃ ca is read in the ad class (SK 2423) → bobhū + ti

→ the next sūtra gives bobhavīti and bobhoti.

Example (with ac). One of Bhaṭṭoji's commentators shows the other kind on the root lū.

lū + yaṅ → the luk is conditioned, so it waits; doubling comes first → lu lū + ya

→ guṇa of the abhyāsa (SK 2630) → lolūya

→ the kṛt suffix ac follows → this rule: luk of yaṅ → lolū + a

→ guṇa of ū before ac is due, but SK 2656 bars it → uvaṅ (SK 271) → loluva.

Without the luk, the same commentator gives lolūya. Without ac, he gives the pair bobhūyate (no luk, chapter 56) and bobhavīti (luk).

Applies when / blocked by. Does na lumatāṅgasya (SK 263) not stop the doubling? An affix removed by luk conditions no operation on the base. Bhaṭṭoji's commentators answer in two ways.

There is a second objection. Doubling by SK 2175 names "one vowel" (ekāc). The verse under SK 2246 says such rules do not hold in yaṅ-luk. Yet the root is doubled. Bhaṭṭoji's commentators answer that guṇo yaṅlukoḥ (SK 2630) prescribes guṇa for the abhyāsa in yaṅ-luk. That rule would be pointless if there were no doubling. So it is a jñāpaka: a sign in the wording of a sūtra that shows a further principle. Here the principle is that the prohibition in the verse is anitya, not invariable. Section B returns to this with han.

Parasmaipada only. Bhaṭṭoji says the stem takes parasmaipada by SK 2159, and not ātmanepada by anudāttaṅita ātmanepadam (SK 2157). Why not? Yaṅ is ṅit, marked with ṅ, and SK 2157 gives ātmanepada after a ṅit. But being ṅit is common to suffixes and non-suffixes. Bhaṭṭoji's commentators give citraṅ as a non-suffix with the marker ṅ. Pratyayalakṣaṇa holds only where a property exclusive (asādhāraṇa) to suffixes is relied upon. So it does not carry the ṅ of the deleted yaṅ.

Bhaṭṭoji gives a parallel. In sudṛṣatprāsādaḥ "[a palace] with fine stones", the lengthening of SK 425 does not take place. Bhaṭṭoji's commentators explain: the ending as that was deleted occurs both as a suffix and as a non-suffix. So the stem does not count as ending in as by pratyayalakṣaṇa.

What of roots that are themselves anudātta-marked or ṅit, such as spardh and śīṅ? Anudātta is the low pitch-accent. A root whose marker vowel is anudātta takes ātmanepada by SK 2157. Bhaṭṭoji says they too take no ātmanepada in yaṅ-luk. SK 2157 cites them by their marker (anubandha). The paribhāṣā śtipā śapā… excludes such a citation in yaṅ-luk. A paribhāṣā is a rule about how other rules are read. This one is the verse of SK 2246. One of Bhaṭṭoji's commentators says why Bhaṭṭoji does not use this second answer for bhū as well. Bhū is not cited with a marker. The ṅ of yaṅ belongs to the suffix, not to the whole form.

The vikaraṇa. By the same paribhāṣā, Bhaṭṭoji says, śyan and the other class-signs (vikaraṇa) do not come. They are taught by citing a class (gaṇa). Only śap comes. One of his commentators adds that śap comes because its rule, SK 2167, lies outside the prohibition. Then śap undergoes luk. The reason is that the line carkarītaṃ ca is read in the ad class. One of Bhaṭṭoji's commentators names the rule: adiprabhṛtibhyaḥ śapaḥ (SK 2423).

Watch out. The luk this chapter conjugates is the unconditioned one. It comes before doubling, not after. Do not confuse it with the conditioned luk before ac, which comes after doubling and which SK 2656 concerns. Some carried the word chandasi "in the Veda" into this sūtra as well, as if yaṅ-luk were Vedic only. One of Bhaṭṭoji's commentators refutes them; the other says that even on their view yaṅ-luk is sometimes used in ordinary speech. Bhaṭṭoji's own proof of that comes under SK 2651.

B. īṭ optional before a hal-initial pit sārvadhātuka: the paradigms of seventeen roots — running example: bobhū + tip → (īṭ) bobhū-ī-ti → (guṇa, av) bobhavīti; (no īṭ) bobhū-ti → (guṇa) bobhoti

One sūtra, SK 2651, governs the endings of the whole conjugation. After a stem that ends in yaṅ, even a yaṅ removed by luk, the augment īṭ may be added to a consonant-initial pit ending. So almost every form with tip, sip or mip comes in two shapes: bobhavīti with ī, bobhoti without. Bhaṭṭoji first proves that the guṇa of bobhoti is right. Then he runs through seventeen roots: bhū, spardh, gādh, nāth, dadh, skund, mud, kurd, vañc, gam, han, car, khan, yu, nu, hā, svap. Each root shows a different rule reaching, or failing to reach, the doubled stem. The running example is bhū.

SK 2651 · यङो वा · yaṅo vā (A 7.3.94)

यङो वा
yaṅo vā · A 7.3.94
यङन्तात्परस्य हलादेः पितः सार्वधातुकस्य ईड्वा स्यात् । भुसुवोः - SK 2224 इति गुणनिषेधो यङ्लुकि भाषायां न । बोभूतु तेतिक्ते SK 3596 इति छन्दसि निपातनात् । अत एव यङ्लुग्भाषायामपि सिद्धः । न च यङ्लुकि अप्राप्त एव गुणाभावो निपात्यतामिति वाच्यम् । [(परिभाषा - ) प्रकृतिग्रहणेन यङ्लुगन्तस्यापि ग्रहणात्] । द्विः प्रयोगो द्विर्वचनं षाष्ठम् इति सिद्धान्तात् । बोभवीति । बोभोति । बोभूतः । बोभुवति । बोभवांचकार । बोभविता । अबोभवीत् । अबोभोत् । अबोभूताम् । अबोभवुः । बोभूयात् । बोभूयाताम् । बोभूयास्ताम् । गातिस्था - SK 2223 इति सिचो लुक् । यङो वा - SK 2651 इतीट्पक्षे गुणं बाधित्वा नित्यत्वाद्वुक् । अबोभूवीत् । अबोभोत् । अबोभूताम् । अभ्यस्ताश्रयो जुस् । नित्यत्वाद्रुक् । अबोभूवुः । अबोभविष्यदित्यादि । पास्पर्धीति । पास्पर्धि । पास्पर्धः । पास्पर्धति । पास्पर्त्सि । हुझल्भ्यो हेर्धिः SK 2425 । पास्पर्धि । लङ् । अपास्पतर् । अपास्पदर् । सिपि दश्च SK 2468 इति रुत्वपक्षे रो रि SK 173 । अपास्पाः । जागाद्धि । जाधात्सि । अजाघात् । सिपि रुत्वपक्षे । अजाघाः । नाथृ । नानात्ति । नानात्तः । दध । दादद्धिः । दाधत्सि । अदाधत् । अदादद्धाम् । अदादधुः । अदाधः । अदाधत् । लुङि । अदादाधीत् । अदादधीत् । चोस्कुन्दीति । चोस्कुन्ति । अचोस्कुन् । अचोस्कुन्ताम् । अचोस्कुन्दुः । मोमुदीति । मोमोत्ति । मोमोदांचकार । मोमोदिता । अमोमुदीत् । अमोमोत् । अमोमुत्ताम् । अमोमुदुः । अमोमुदीः । अमोमोः । अमोमोत् । लुङि गुणः । अमोमोदीत् । चोकूर्ति । चोकूर्दीति । लङ् तिप् । अचोकूतर् । अचोकूर्दीत् । सिप्पक्षे । अचोकूः । अचोखूः । अजोगूः । वनीवञ्चीति । वनीवङ्क्ति । वनीवक्तः । वनीवचति । अवनीवञ्चीत् । अवनीवन् । जङ्गमीति । जङ्गन्ति । अनुदात्तोपदेश - SK 2428 इत्यनुनासिकलोपः । जङ्गन्तः । जङ्ग्मति । म्वोश्च SK 2309 । जङ्गन्मि । जङ्गन्वः । एकाज्ग्रहणेनोक्तत्वान्नेण्निषेधः । जङ्गमिता । अनुनासिकलोपस्याभीयत्वेनासिद्धत्वान्न हेर्लुक् । जङ्गहि ॥मो नो धातोः SK 341 अजङ्गन् । अनुबन्धनिर्देशान्न च्लेरङ् । ह्म्यन्त - SK 2299 इति न वृद्धिः । अजङ्गमीत् । अजङ्गमिष्टाम् । हन्तेर्युङ्लुक् । अभ्यासाच्च SK 2430 इति कुत्वं यद्यपि हो हन्तेः SK 358 इत्यतो हन्तेरित्यनुवर्त्य विहितं, तथापि यङ्लुकि भवत्येवेति न्यासकारः । [(परिभाषा - ) श्तिपा शपा...] इति निषेधस्तवनित्यः गुणो यङ्लुकोः SK 2630 इति सामान्यापेक्षज्ञापकादिति भावः । जङ्घनीति । जङ्घन्ति । जङ्घतः । जङ्घ्नति । जङ्घनिता । श्तिपा निर्देशाज्जादेशो न । जङ्घहि । अजङ्घनीत् । अजङ्घन् । जङ्घन्यात् । आशिषि तु । वध्यात् । अवधीत् । अवधिष्टामित्यादि । वधादेशस्य द्वित्वं तु न भवति । स्थानिवत्त्वेनानभ्यासस्येति निषेधात् । तद्धि समानाधिकरणं धातोर्विशेषणम् । बहुव्रीहिबलात् । आङ्पूर्वात्तु आङो यमहनः SK 2695 इत्यात्मनेपदम् । आजङ्घते इत्यादि । उत्परस्य - SK 2637 इति तपरत्वान्न गुणः । हलि च SK 354 इति दीर्घस्तु स्यादेव । तस्यासिद्धत्वेन तपरत्वनिवर्त्यत्वायोगात् । चञ्चुरीति । चञ्चूर्ति । चञ्चूर्तः । चञ्चुरति । अचञ्चूरीत् । अचञ्चूः । चङ्खनीति । चङ्खन्ति । जनसन - SK 2504 इत्यात्वम् । चङ्खातः । गमहन - SK 2363 इत्युपधालोपः । चङ्ख्नति । चङ्खाहि । चङ्खनानि । अचङ्खनीत् । अचङ्खन् । अचङ्खाताम् । अचङ्ख्नुः । ये विभाषा SK 2319 । चङ्खायात् । चङ्खन्यात् । अचङ्खनीत् । अचङ्खानीत् । उतो वृद्धिः SK 2443 इत्यत्र नाभ्यस्तस्येत्यनुवृत्तेरुतोवृद्धिर्न । योयोति । योयवीति । अयोयवीत् । अयोयोत् । योयुयात् । आशिषि दीर्घः । योयूयात् । अयोयावीत् । नोनवीति । नोनोति । जाहेति । जाहाति । ई हल्यघोः SK 2497 । जाहीतः । इह जहातेश्च SK 2498 आच हौ SK 2499 लोपो यि SK 2500 घुमास्था - SK 2462 एर्लिङि - SK 2374 इत्येते पञ्चापि न भवन्ति । श्तिपा निर्देशात् । जाहति । जाहासि । जाहेषि । जाहीथः । जाहीथ । जाहीति । अजाहेत् । अजाहात् । अजाहीताम् । अजाहासिष्टाम् । अजाहिष्यत् । लुका लुप्ते प्रत्ययलक्षणाभावात् स्वपिस्यमि - SK 2645 इत्युत्वं न । रुदादिभ्यः - SK 2464 इति गणनिर्दिष्टत्वादिण्न । सास्वपीति । सास्वप्ति । सास्वप्तः । सास्वपति । असास्वपीत् । असास्वप् । सास्वप्यात् । आशिषि तु वचिस्वपि - SK 2409 इत्युत्वम् । सासुप्यात् । असास्वापीत् । असास्वपीत् ॥

Rule. After a stem ending in yaṅ, the augment īṭ is optionally added to a following sārvadhātuka that begins with a consonant (hal) and is pit. A sārvadhātuka is a verb-ending or ś-marked affix taught after a root (SK 2166). Pit means marked with p: the endings tip, sip, mip and a few others. Īṭ is the augment ī.

Three words come by anuvṛtti. Bhaṭṭoji's commentators name their sources: "pit sārvadhātuka" from nābhyastasyāci piti sārvadhātuke (SK 2503), "consonant-initial" (hali) from uto vṛddhir luki hali (SK 2443), and īṭ from bruva īṭ (SK 2452). The yaṅ is deleted in this chapter, but by pratyayalakṣaṇa the stem still ends in yaṅ. One commentator rejects a reading on which yaṅ here stands for the yaṅ-luk stem itself. The other notes that īṭ, being marked with ṭ, goes at the front of the ending.

Example.

bobhū + tip (Section A) → p is a marker → bobhū + ti

→ this rule: īṭ at the front of ti → bobhū + īti

→ guṇa of ū before the sārvadhātuka (SK 2168) → bobho + īti

→ av for o before a vowel (SK 61) → bobhavīti.

Without īṭ:

bobhū + ti → guṇa (SK 2168) → bobhoti.

Why guṇa at all? Bhaṭṭoji's argument. Bhūsuvos tiṅi (SK 2224) bars guṇa of bhū before a verb-ending. Should it not bar the guṇa in bobhoti? Bhaṭṭoji says the bar does not hold in yaṅ-luk in ordinary speech (bhāṣā). His proof is a Vedic form.

The objection and the paribhāṣā. Could one say instead that SK 2224 never reaches the doubled stem, so that bobhūtu teaches a new absence of guṇa? Bhaṭṭoji rejects this. By a paribhāṣā, the mention of a base (prakṛti) includes its yaṅ-luk form as well: prakṛtigrahaṇe yaṅlugantasyāpi grahaṇam. So "bhū" in SK 2224 includes bobhū.

This paribhāṣā rests on a settled conclusion (siddhānta): the doubling taught in the sixth book of the Aṣṭādhyāyī (ṣāṣṭha) is a pronouncing twice (dviḥ prayoga). The doubled stem is the same root said twice, not a new item. One of Bhaṭṭoji's commentators concludes that the paribhāṣā is established by reasoning. It is not a fresh statement.

bhū → bobhū

The stem was built in Section A. The forms:

abobhū + īt (with this rule's īṭ)

→ guṇa is due before the sārvadhātuka, but vuk (SK 2174) adds v after bhū

→ vuk is nitya: it applies whether or not guṇa has applied, so it wins → abobhūv + īt → abobhūvīt.

Without īṭ: abobhot. 3rd du. abobhūtām. 3rd pl.: the jus that replaces jhi here is the one that depends on the doubled stem (abhyasta, SK 2226). Vuk, again nitya, sets guṇa aside → abobhūvuḥ.

One of Bhaṭṭoji's commentators rejects those who would make vuk not nitya here and so give abobhavīt. That contradicts the Bhāṣya's own words: "because vuk of bhū is nitya". Both explain the jus. The restriction of SK 2227, which allows jus after a deleted sic only for roots ending in ā, governs only the jus due to a following sic, not the jus due to the abhyasta.

spardh → pāspardh

spardh spardh → śarpūrvāḥ khayaḥ (SK 2259): in a cluster headed by a sibilant, the stop is kept → pa spardh

→ dīrgho 'kitaḥ (SK 2632) lengthens the abhyāsa → pāspardh.

Spardh is cited with an anudātta marker, yet it takes no ātmanepada here (SK 2650).

apāspardh + t → the one-sound ending is lost after a consonant (SK 252) → apāspardh

→ jaś at the end of the word (SK 84) → apāspard → optional car at a pause (SK 206) → apāspart / apāspard.

apāspardh + s → jaś → apāspard + s

→ daś ca (SK 2468) optionally gives ru (r) for the d → apāspar + r + s

→ s is lost (SK 252) → ro ri (SK 173) deletes the first r → apāspa + r

→ ḍhralope (SK 174) lengthens the vowel before the lost r → apāspā + r → visarga → apāspāḥ.

This order of steps is one commentator's. The other notes that the option of ru implies the plain forms apāspart / apāspard as well.

gādh → jāgādh

gādh gādh → halādiḥ śeṣaḥ (SK 2179): only the first consonant of the abhyāsa stays → gā gādh

→ shortening (SK 2180) → ga → kuhoś cuḥ (SK 2245): a velar of the abhyāsa becomes a palatal → ja

→ dīrgho 'kitaḥ → jāgādh.

nāth → nānāth

nāth nāth → halādiḥ śeṣaḥ, shortening, dīrgho 'kitaḥ → nānāth. laṭ nānātti, nānāttaḥ: th becomes t before t (SK 121).

dadh → dādadh

dadh dadh → da dadh → dīrgho 'kitaḥ → dādadh.

skund → coskund

skund skund → śarpūrvāḥ khayaḥ keeps the stop → ku skund → kuhoś cuḥ → cu

→ guṇa of the abhyāsa (SK 2630) → coskund.

mud → momud

mud mud → mu mud → guṇa of the abhyāsa (SK 2630) → momud. This root shows how īṭ and the guṇa of a light penultimate exclude each other.

momud + īti → the ending now begins with a vowel and is pit

→ nābhyastasyāci piti sārvadhātuke (SK 2503) bars guṇa of the light u → momudīti.

momud + ti → no vowel follows, so SK 2503 does not apply (one of Bhaṭṭoji's commentators)

→ guṇa of the light penultimate (SK 2189) → momod + ti → car (SK 121) → momotti.

kurd → cokūrd

The abhyāsa takes c by kuhoś cuḥ and guṇa by SK 2630; Bhaṭṭoji's forms show the stem cokūrd.

vañc → vanīvañc

vañc vañc → va vañc → the augment nīk (SK 2642), marked with k, so added at the end of the abhyāsa (SK 36) → vanīvañc.

gam → jaṅgam

gam gam → halādiḥ śeṣaḥ → ga gam → kuhoś cuḥ → ja gam

→ the augment nuk (SK 2643): the abhyāsa ends in a and the root ends in a nasal; nuk is marked with k, so it goes at the end of the abhyāsa → jan gam

→ n becomes anusvāra (SK 123) and then ṅ before g (SK 124) → jaṅgam.

jaṅgam + hi → the nasal is deleted (SK 2428) → jaṅga + hi

→ ato heḥ (SK 2202) would now delete hi after a stem in a

→ but the nasal deletion belongs to the ābhīya zone (chapter 44), so for SK 2202 it counts as not having taken effect (asiddha) → hi stays → jaṅgahi.

han → jaṅghan

han han → ha han → kuhoś cuḥ and abhyāse carca (SK 2182) → ja han

→ nuk (SK 2643) → jaṅ han

→ abhyāsāc ca (SK 2430): after the abhyāsa, the h of han becomes the velar gh → jaṅghan.

Does SK 2430 reach the yaṅ-luk stem? Bhaṭṭoji raises the question himself. SK 2430 gets its word "of han" (hanteḥ) by anuvṛtti from ho hanter ñṇinneṣu (SK 358). Hanteḥ is a citation with śtip, the ending -ti. The paribhāṣā śtipā śapā… excludes such citations in yaṅ-luk. Even so, says the author of the Nyāsa (the Nyāsakāra), the velar does come in yaṅ-luk. The reason: the prohibition of that paribhāṣā is anitya. Guṇo yaṅlukoḥ (SK 2630) is the jñāpaka, and it shows this with regard to the general case (sāmānyāpekṣa). Bhaṭṭoji closes: "that is the point". One of his commentators makes clear that the point is the Nyāsakāra's.

How does SK 2630 show this? Bhaṭṭoji's commentators explain. The mention of "one vowel" in SK 2175 would deny doubling in yaṅ-luk. Yet SK 2630 gives guṇa to the abhyāsa in yaṅ-luk, which presupposes doubling. So the "one vowel" exclusion sometimes fails. By kinship, all the exclusions of the same verse sometimes fail too. One of the commentators cites a usage that depends on this: rājā vṛtraṃ jaṅghanat.

The forms:

car → cañcur

car car → ca car → nuk by caraphaloś ca (SK 2636) → can car → anusvāra and ñ before c (SK 123, 124) → cañcar

→ ut parasyātaḥ (SK 2637): u for the a after the abhyāsa → cañcur.

Bhaṭṭoji makes two points about that u.

The forms:

khan → caṅkhan

khan khan → kha khan → abhyāse carca (SK 2182): an aspirate of the abhyāsa loses its aspiration → ka khan

→ kuhoś cuḥ → ca khan → nuk (SK 2643), anusvāra, ṅ → caṅkhan.

yu → yoyu, nu → nonu

yu yu → guṇa of the abhyāsa (SK 2630) → yoyu. Likewise nonu.

In the ad class, yu takes vṛddhi before a consonant-initial pit ending: yauti (SK 2443). Not here. Bhaṭṭoji explains: "nābhyastasya" "not of a doubled stem" is carried into SK 2443 from SK 2503. So the doubled stem takes only guṇa (SK 2168).

hā → jāhā

hā hā → shortening → ha hā → kuhoś cuḥ and abhyāse carca → ja hā → dīrgho 'kitaḥ → jāhā.

Two roots are heard as hā: hāṅ, with the marker ṅ, and hāk, with the marker k. Bhaṭṭoji's commentators say that in yaṅ-luk both give the same forms, since ātmanepada due to ṅ does not come here. Is the abhyāsa of hāk lengthened by dīrgho 'kitaḥ, "of an abhyāsa without k"? Bhaṭṭoji's commentators say yes, reading akitaḥ as a bahuvrīhi. One of them adds that, doubling being a pronouncing twice, the k belongs to the doubled whole, not to each copy.

Bhaṭṭoji names five rules of hā that do not apply here, because each is taught with śtip:

The forms:

Some copies skip from ajāhītām straight to ajāhāsiṣṭām. The printed text has the four forms between.

svap → sāsvap

svap svap → halādiḥ śeṣaḥ → sa svap → dīrgho 'kitaḥ → sāsvap.

Bhaṭṭoji makes two points.

The forms:

Applies when / blocked by. Īṭ may come before any consonant-initial pit sārvadhātuka: tip, sip, mip in laṭ, t, s in laṅ and luṅ (abobhūvīt), tu in loṭ. Before tas, jhi and the other ṅit endings it does not come: bobhūtaḥ, amomuttām.

Watch out. The exclusion of the verse under SK 2246 is not invariable, but it is not empty either. On one and the same root, han, the velar of SK 2430 comes (jaṅghanīti) while the ja of SK 2431 does not (jaṅghahi). Each rule must be tested as Bhaṭṭoji tests it. Note also the pair under mud: with īṭ there is no guṇa (momudīti), without it there is (momotti). Final vowels take guṇa either way: bobhavīti / bobhoti.

C. ruk, rik and rīk for the abhyāsa of ṛ-penultimate and ṛ-final roots — running example: vṛt → (doubling; abhyāsa ṛ → a by urat, r, halādiḥ śeṣaḥ) va-vṛt → + ruk / rik / rīk → var-vṛt / vari-vṛt / varī-vṛt → (no īṭ, guṇa) varvarti / varivarti / varīvarti

Two sūtras give the abhyāsa of a ṛ-root three augments in yaṅ-luk: ruk (r), rik (ri) and rīk (rī). An augment (āgama) is a sound added to a form, not put in place of anything. SK 2652 covers roots whose penultimate is ṛ, such as vṛt "to turn". SK 2653 covers roots that end in short ṛ, such as kṛ "to do" and ṛ "to go". So most forms come in three shapes, and Bhaṭṭoji marks these with the figure "(3)". Ṛ itself has only two. Under SK 2653 he also gives the roots that look similar but take no augment (kṝ, tṝ, grah), the root pracch, and the roots ending in y or v. The running example is vṛt.

SK 2652 · रुग्रिकौ च लुकि · rugrikau ca luki (A 7.4.91)

रुग्रिकौ च लुकि
rugrikau ca luki · A 7.4.91
ऋदुपधस्य धातोरभ्यासस्य रुक् रिक् रीक् एते आगमाः स्युर्यङ्लुकि ॥

Rule. In yaṅ-luk, the abhyāsa of a root whose penultimate is ṛ takes the augments ruk and rik, and by the ca, rīk. Bhaṭṭoji gives no form here. The forms come under the next sūtra.

The sūtra leans on its neighbours. Both of Bhaṭṭoji's commentators say the ca draws in rīk from rīg ṛdupadhasya ca (SK 2644). One of them, the Bālamanoramā, names the sources of the rest. "Of a root whose penultimate is ṛ" (ṛdupadhasya) comes from the same sūtra. "Of the abhyāsa" comes from atra lopo 'bhyāsasya (SK 2620). "In yaṅ-luk" comes from guṇo yaṅlukoḥ (SK 2630). Hence the word luki "in luk" means luk of yaṅ.

Which sounds are real? Bhaṭṭoji's commentators say the u of ruk is only there to make it pronounceable (uccāraṇārtha). One of them adds that the i of rik and the ī of rīk are actually heard. Each augment is marked with k, so it is added at the end of the abhyāsa (SK 36): var-, vari-, varī-.

Example. vṛt "to turn":

vṛt + yaṅ → luk of yaṅ (SK 2650) → doubling (SK 2395) → vṛt vṛt

→ urat (SK 2244): the ṛ of the abhyāsa becomes a, followed by r → vart vṛt

→ halādiḥ śeṣaḥ (SK 2179): only the first consonant of the abhyāsa stays → va vṛt

→ this rule: ruk → var vṛt; or rik → vari vṛt; or rīk → varī vṛt

→ laṭ, tip, no īṭ: guṇa of the light penultimate ṛ (SK 2189), ar → varvarti / varivarti / varīvarti.

Watch out. Do not lengthen the abhyāsa. An abhyāsa that carries one of these k-marked augments is "kit", so dīrgho 'kitaḥ (SK 2632) does not apply: var-, never vār-. One of Bhaṭṭoji's commentators makes this point under the root ṛ.

SK 2653 · ऋतश्च · ṛtaś ca (A 7.4.92)

ऋतश्च
ṛtaśca · A 7.4.92
ऋदन्ताद्धातोरपि तथा । वर्वृतीति । वरिवृतीति । वरीवृतीति । वर्वर्ति । वरिवर्ति । वरीवर्ति । वर्वृतः 3 । वर्वृतति 3 । वर्वर्तामास 3 । वर्वर्तिता । गणनिर्दिष्टत्वात् न वृद्भ्यश्चतुर्भ्यः SK 2348 इति न । वर्वर्तिष्यति 3 । अवर्वृर्तीत् 3 । अवर्वतर् 3 । सिपि दश्च SK 2468 इति रुत्वपक्षे रो रि SK 173 । अवर्वाः 3 । गणनिर्दिष्टत्वादङ् न । अवर्वर्तीत् । चर्करीति 3 । चर्कर्ति । चरिकर्ति । चरीकर्ति । चर्कृतः 3 । चर्क्रति 3 । चर्करांचकार 3 । चर्करिता 3 । अचर्करीत् । अचर्कः 3 । चर्कृयात् 3 । आशिषि रिङ् । चर्क्रियात् 3 । अचर्कारीत् 3 । ऋतश्च SK 2653 इति तपरत्वान्नेह । कॄ विक्षेपे । चाकर्ति । तातर्ति । तातीर्तः । तातिरति । तातीर्हि । तातराणि । अतातरीत् । अतातः । अतातीर्ताम् । अतातरुः । अतातारीत् । अतातारिष्टामित्यादि । अर्तेर्यङलुकि द्वित्वेऽभ्यासस्योरदत्वं रपरत्वम् । हलादिः शेषः SK 2179 । रुक् । रिग्रीकोस्तु अभ्यासस्यासवर्णे SK 2290 इति इयङ् । अरर्ति । अरियर्ति । अरियर्ति । अररीति । अरियरीति । अर्ऋतः । अरियृतः । झि अत् । यण् । रुको रोरि SK 1731 इति लोपः । न च तस्मिन् कर्तव्ये यणः स्थानिवत्त्वम् । पूर्वत्रासिद्धीये तन्निषेधात् । आरति । अरिय्रति । लिङि श्तिपा निर्देशात् गुणोऽर्ति - SK 2380 इति गुणो न । रिङ् । रलोपः । दीर्घः । आरियात् । अरिय्रियात् । अरारीत् । आरियारीत् । गृहू ग्रहणे । जर्गृहीति 3 । जर्गर्ढि 3 । जर्गृढः 3 । जर्गृहति 3 । अजर्घटर् 3 । गृह्णातेस्तु । जाग्रहीति । जाग्राढि ॥ तसादौ ङिन्निमित्तं संप्रसारणम् । तस्य बहिरङ्गत्वेनासिद्धत्वान्नरुगादयः । जागृढः । जागृहति । जाग्रहीषि । जाघ्रक्षि । लुटि । जाग्रहिता । ग्रहोऽलिटि SK 2562 इति दीर्घस्तु न । तत्रैकाच इत्यनुवृत्तेः । माधवस्तु दीर्घमाह तद्भाष्यविरुद्धम् । गृधु अभिकाङ्क्षायाम् । जर्गृधीति 3 । जर्गर्द्धि 3 । जर्गृद्धः 3 । जर्गृधति 3 । जर्गृधीषि 3 । जर्घर्त्सि 3 । अजर्गृधीत् 3 । ईडभावे गुणः । हल्ङ्यादिलोपः । भष्भावः । जश्त्वचर्त्वे । अजर्घतर् 3 । अजर्गृद्धाम् 3 । सिपि दश्च SK 2668 इति पक्षे रुत्वम् । अजर्घाः 3 । अजर्गर्धीत् 3 । अजर्गर्धिष्टाम् 3 । पाप्रच्छीति । पाप्रष्टि । तसादौ ग्रहिज्या - SK 2412 इति संप्रसारणं न भवति । श्तिपा निर्देशात्. च्छ्वोः शूड् - SK 2561 इति शः । व्रश्च - SK 294 इति षः । पाप्रष्टः । पाप्रच्छति । पाप्रश्मि । पाप्रच्छ्वः । पाप्रश्मः । यकारवकारान्तानां तूठ्भाविनां यङ्लुङ् नास्तीति च्छ्वोः - SK 2561 इति सूत्रे भाष्ये ध्वनितम् । कैयटेन च स्पष्टीकृतम् । इदं च्छ्वोः - SK 2561 इति यत्रोठ् तद्विषयकम् । ज्वरत्वर - SK 2564 इत्यूठ्भाविनोः स्रिविमव्योस्तु यङ्लुगस्त्येवेति न्याय्यं माधवादिसंमतं च । मव्य बन्धने । अयं यान्त ऊठ्भावी । तेवृ देवृ देवने इत्यादयो वान्ताः । हय गतौ । जाहयीति । जाहति । जाहतः । जाहयति । जाहयीषि । जाहसि । वलि लोपे यञादौ दीर्घः । जाहामि । जाहावः । जाहामः । हर्य गतिकान्त्योः । जाहर्यीति । जाहर्ति । जाहर्तः । जाहर्यति । लोटि । जाहर्हि । अजाहः । अजाहर्ताम् । अजाहर्युः । मव बन्धने ॥

Rule. The abhyāsa of a root ending in short ṛ also takes ruk, rik and rīk in yaṅ-luk. Bhaṭṭoji's gloss: "likewise also [the abhyāsa] of a root ending in ṛ". One of his commentators says that, strictly, ṛtaḥ qualifies the word aṅga "base", carried from the heading. "Of a root ending in ṛ" only states the resulting sense.

Example. kṛ "to do":

kṛ kṛ → urat → kar kṛ → halādiḥ śeṣaḥ → ka kṛ → kuhoś cuḥ → ca kṛ

→ this rule: ruk → carkṛ (or carikṛ, carīkṛ)

→ laṭ, tip, no īṭ: guṇa of the final ṛ (SK 2168) → carkarti / carikarti / carīkarti.

The rest of Bhaṭṭoji's passage runs root by root. "(3)" after a form means it has the three shapes car-, cari-, carī- (or var-, vari-, varī-).

vṛt → varvṛt (3)

Vṛt has penultimate ṛ, so its augments come from SK 2652. Its stem was built above.

avarvṛt + t → guṇa → avarvart + t → the ending is lost (SK 252) → avarvart

→ the final t of a cluster would be lost (SK 54), but rāt sasya (SK 280) allows only s to be lost after r → avarvart (3). One of Bhaṭṭoji's commentators gives this step.

kṛ → carkṛ (3)

Long ṝ takes no augment: kṝ, tṝ

The sūtra says ṛt, with t after the vowel (tapara). A tapara vowel names only its own length. So a root ending in long ṝ takes no ruk, rik or rīk.

ṛ (arti) → ar-ṛ / ariy-ṛ

One of Bhaṭṭoji's commentators says both roots ṛ are meant, that of the bhū class and that of the hu class. Bhaṭṭoji gives the steps:

ṛ ṛ → urat: the abhyāsa's ṛ becomes a, followed by r → ar ṛ

→ halādiḥ śeṣaḥ → a ṛ

→ with ruk → ar ṛ → guṇa before tip → ararti.

With rik or rīk:

a ṛ → ari ṛ / arī ṛ

→ iyaṅ: the abhyāsa's i or ī becomes iy before the unlike vowel ṛ (abhyāsasyāsavarṇe, SK 2290) → ariy ṛ → guṇa → ariyarti.

So rik and rīk give the same forms here: two shapes, not three. One of Bhaṭṭoji's commentators adds that some give the abhyāsa guṇa by SK 2630 instead of urat; the result is the same.

ar ṛ + jhi → jhi becomes at → ar ṛ + ati

→ yaṇ: r for ṛ before the vowel → ar r ati

→ ro ri (SK 173) deletes the r of ruk → a r ati

→ ḍhralope (SK 174) lengthens the a → ārati.

With rik / rīk: ariy ṛ + ati → yaṇ → ariyrati.

ar ṛ + yāt → riṅ (SK 2367) → ar ri yāt → ro ri deletes the first r → a ri yāt → lengthening (SK 174) → āriyāt.

With rik / rīk: ariy ṛ + yāt → riṅ → ariyriyāt.

gṛh → jargṛh (3), and grah → jāgrah

gṛh "to take" has penultimate ṛ:

gṛh gṛh → urat → gar gṛh → halādiḥ śeṣaḥ → ga → kuhoś cuḥ → ja → ruk → jargṛh.

Contrast grah, the root of gṛhṇāti:

grah grah → ga grah → kuhoś cuḥ → ja → dīrgho 'kitaḥ → jāgrah.

gṛdh → jargṛdh (3)

gṛdh "to desire", built like gṛh:

ajargṛdh + t → guṇa (SK 2189) → ajargardh + t

→ the ending is lost (SK 252) → ajargardh

→ bhaṣ: g becomes gh (SK 326) → ajarghardh

→ jaś (SK 84) → ajarghard → car (SK 206) → ajarghart (3).

pracch → pāpracch

Pracch is not a ṛ-root, but Bhaṭṭoji places it here.

pracch pracch → halādiḥ śeṣaḥ → pa → dīrgho 'kitaḥ → pāpracch.

pāpracch + tas → ś for cch (cchvoḥ śūḍ anunāsike ca, SK 2561) → ṣ for ś (SK 294) → ṭ for t → pāpraṣṭaḥ.

Roots ending in y or v

Bhaṭṭoji now states a limit. Roots ending in y or v that undergo ūṭh have no yaṅ-luk. Ūṭh is the substitute ū, and an ūṭhbhāvin is a root that is to take it. Bhaṭṭoji gives his authorities: this is hinted (dhvanita) in the Bhāṣya on cchvoḥ… (SK 2561), and made explicit by Kaiyaṭa.

How is it hinted? Bhaṭṭoji's commentators explain. The Bhāṣya there weighs whether "before a kit or ṅit" continues into SK 2561. It concludes "only so much is the difference" (etāvān eva viśeṣaḥ): only the ṣ for ch is affected. Were there yaṅ-luk for such roots, div would give either dedeti or dedyoti according to the choice. That would be a further difference, and "only" would be false.

Bhaṭṭoji limits the statement in two ways.

He gives examples of the first kind:

One of his commentators says outright that these have no yaṅ-luk. Which root the mav of SK 2654 is, mav or mavya, scholars discuss.

haya → jāhay. Haya "to go" ends in y but takes no ūṭh, so it has yaṅ-luk.

hay hay → ha hay → kuhoś cuḥ, abhyāse carca → ja → dīrgho 'kitaḥ → jāhay.

jāhay + mi → y lost before m (SK 873) → jāha + mi

→ a lengthened before m (ato dīrgho yañi, SK 2170) → jāhāmi.

So too jāhāvaḥ, jāhāmaḥ.

harya → jāhary. Harya "to go, to desire":

Bhaṭṭoji ends the passage with "mava — to bind". That is the root of the next sūtra's example.

D. ūṭh for the penultimate and the v of jvara, tvara, srivi, avi, mava — running example: mav: māmav-ti → (ūṭh for a + v together) māmū-ti → (tip is pit, so guṇa) māmoti

SK 2654 replaces the penultimate vowel and the final v of five roots by ū, before certain affixes. One of those roots is mav, which, Bhaṭṭoji has just said, keeps its yaṅ-luk (Section C). His example is the yaṅ-luk of mav: māmoti.

SK 2654 · ज्वरत्वरस्रिव्यविमवामुपधायाश्च · jvaratvarasrivyavimavām upadhāyāś ca (A 6.4.20)

ज्वरत्वरस्रिव्यविमवामुपधायाश्च
jvaratvarasrivyavimavāmupadhāyāśca · A 6.4.20
ज्वरादीनामुपधावकारयोरूठ् स्यात् क्वौ झलादावनुनासिकादौ च प्रत्यये । अत्र क्ङितीति नानुवर्तते । अवतेस्तुनि कृते ओतुरिति दर्शनात् । अनुनासिकग्रहणं चानुवर्तेते । अवतेर्मन्प्रत्यये तस्य टिलोपे ओमिति दर्शनात् । ईडभावे ऊठि पिति गुणः । मामोति । मामवीति । मामूतः । मामवति । मामोषि । मामोमि । मामावः । मामूमः । मामूहि । मामवानि । अमामोत् । अमामोः । अमामवम् । अमामाव । अमामूम । तुर्वी हिंसायाम् । तोतूर्वीति ॥

Rule. Of the roots jvar, tvar, srivi, av and mav, the penultimate and the v are replaced by ūṭh. This happens before kvi, before a suffix that begins with a jhal, and before a suffix that begins with a nasal. Ūṭh is the substitute ū; its ṭh is a marker. Kvi is a kṛt affix of which no sound remains. The penultimate (upadhā) is the sound before the last.

The sūtra borrows much. One of Bhaṭṭoji's commentators lists it. From cchvoḥ śūḍ anunāsike ca (SK 2561) come "v", "ūṭh" and "before a nasal". From anunāsikasya kvijhaloḥ kṅiti (SK 2666, taught later) comes "before kvi or a jhal". The ca of this sūtra adds the v to the penultimate.

Two points on the anuvṛtti. Bhaṭṭoji tests each by a word.

av + tu → ūṭh for a and v → ū + tu

→ guṇa, possible only because tun is not kit → otu → otuḥ.

One of Bhaṭṭoji's commentators says the earlier commentary (the vṛtti) has kṅiti here, and that it is a mistake.

av + man → ṭi (an) deleted → av + m

→ ūṭh before the nasal → ū + m → guṇa → om.

Example. mav "to bind", in yaṅ-luk:

mav mav → halādiḥ śeṣaḥ → ma mav → dīrgho 'kitaḥ → māmav

→ laṭ, tip, no īṭ → māmav + ti

→ this rule: ūṭh before the jhal t. On one view, one ū replaces a and v together → māmū + ti

→ tip is pit, so not ṅit, and the ū takes guṇa (SK 2168) → māmoti.

Both views are from one of Bhaṭṭoji's commentators, the Bālamanoramā. The first is the one just shown. On the second, each of a and v takes ūṭh, and the two ū merge. It places both views in the Bhāṣya on ekaḥ pūrvaparayoḥ.

The forms:

The same commentator adds the other roots: jājvarīti, tātūrti, sesroti. Av has no yaṅ, so he illustrates it with kvip: ūḥ.

Bhaṭṭoji's passage ends with the root turvī "to injure" and the form totūrvīti. That leads into the next sūtra.

E. v / ch deleted after r, and guṇa not barred before tip: turvī, murchā; the close — running example: toturv-ti → (v deleted after r, 2655) totur-ti → (laghūpadha guṇa: tip is not ārdhadhātuka, so 2656 does not bar it) totorti

Two sūtras close the chapter. SK 2655 deletes ch and v after r before the same affixes as SK 2654. SK 2656 bars guṇa and vṛddhi before an ārdhadhātuka that causes part of a root to be deleted. Bhaṭṭoji uses it to show that it does not bar the guṇa of totorti. He ends with a note on aj, which has no yaṅ-luk. The running example is turvī "to injure": totorti.

SK 2655 · राल्लोपः · rāl lopaḥ (A 6.4.21)

राल्लोपः
rāllopaḥ · A 6.4.21
रेफात्परयोश्छ्वोर्लोपः स्यात् क्वौ झलादावनुनासिकादौ च प्रत्यये । इति वलोपः । लघूपधगुणः ॥

Rule. After r, ch and v are deleted before kvi, a jhal-initial suffix, or a nasal-initial suffix. "ch and v" and "before a nasal" come from SK 2561, and "before kvi or a jhal" from SK 2666. One of Bhaṭṭoji's commentators adds that ūṭh is not carried, nor "before a kit or ṅit".

Example. turv (turvī) "to injure":

turv turv → halādiḥ śeṣaḥ → tu turv → guṇa of the abhyāsa (SK 2630) → toturv.

With īṭ: toturv + īti → the u is lengthened (upadhāyāṃ ca, SK 2265, one of Bhaṭṭoji's commentators) → totūrvīti.

Without īṭ:

toturv + ti → this rule: v after r is deleted before the jhal t → totur + ti

→ tip is pit, so not ṅit: guṇa of the light penultimate (SK 2189) → totorti.

Bhaṭṭoji: "thus the deletion of v; then guṇa of the light penultimate."

Applies when / blocked by. Is this rule needed in totorti? One of Bhaṭṭoji's commentators, the Tattvabodhinī, notes that lopo vyor vali (SK 873) would already delete the v before t here. The other cites this rule at totorti as the proper one. Both say the rule is needed in totūrtaḥ:

toturv + tas → ūṭh by cchvoḥ… (SK 2561) would replace the v, and would set lopo vyor vali aside

→ this rule deletes the v instead → totur + taḥ

→ hali ca (SK 354) lengthens the u → totūrtaḥ.

One of Bhaṭṭoji's commentators gives forms with kvi: tūḥ, dhūḥ, mūḥ.

SK 2656 · न धातुलोप आर्धधातुके · na dhātulopa ārdhadhātuke (A 1.1.4)

न धातुलोप आर्धधातुके
na dhātulopa ārdhadhātuke · A 1.1.4
धात्वंशलोपनिमित्ते आर्धधातुके परे इको गुणवृद्धी न स्तः इति नेह निषेधः । तिबादीनामनार्धधातुकत्वात् । तोतोर्ति । हलि च SK 354 इति दीर्घः । तोतूर्तः । तोतूर्वति । तोथोर्ति । दोदोर्ति । दोधोर्ति । मुर्छा । मोमूर्च्छीति । मोमोर्ति । मोमूर्तः । मोमूर्छतीत्यादि । आर्धधातुक इति विषयसप्तमी । तेन यङि विवक्षिते अजेर्वी । वेवीयते । अस्य यङ्लुङ्नास्ति । लुकापहारे विषयत्वाभावेन वीभावस्याप्रवृत्तेः ॥

Rule. Guṇa and vṛddhi do not replace an ik vowel before an ārdhadhātuka that causes the deletion of part of a dhātu. An ārdhadhātuka is any affix after a root other than the sārvadhātukas (SK 2187). "Guṇa and vṛddhi in place of ik" comes by anuvṛtti from iko guṇavṛddhī (SK 34). Bhaṭṭoji's gloss: "when an ārdhadhātuka follows that is the cause of the deletion of part of a dhātu (dhātvaṃśalopanimitte), guṇa and vṛddhi do not come in place of ik".

Why "part of" a dhātu? One of Bhaṭṭoji's commentators tests the literal readings.

So the ārdhadhātuka must itself cause the deletion. How the compound dhātulopa yields this sense, scholars discuss.

Example: where the sūtra applies. Bhaṭṭoji's commentators give its own domain: the conditioned luk of yaṅ before the kṛt suffix ac (SK 2650). One of them names loluvaḥ, popuvaḥ and marīmṛjaḥ.

lolūya + ac → ac causes the luk of yaṅ, part of the dhātu lolūya → lolū + a

→ guṇa of ū would come before ac (SK 2168)

→ this rule: ac is an ārdhadhātuka that caused part of the dhātu to be deleted, so no guṇa

→ uvaṅ (SK 271) → loluva → loluvaḥ.

Example: where it does not. Bhaṭṭoji: "this prohibition does not apply here, because tip and the rest are not ārdhadhātuka". So in totur + ti the guṇa stands → totorti. The luk of yaṅ in this chapter is the unconditioned one in any case (SK 2650).

Bhaṭṭoji's remaining forms:

The status of the sūtra. Bhaṭṭoji's commentators report that the Bhāṣya rejects this sūtra (pratyākhyāta): it is not needed. One of them gives the reason: the deleted a is treated as still present (sthānivadbhāva), and that already prevents the guṇa and vṛddhi. The other adds that Haradatta reinstates the sūtra and that the Kaustubha rejects it again.

A note on aj. Bhaṭṭoji closes with the root aj. The rule ajer vy aghañapoḥ (SK 2292) replaces aj by vī before an ārdhadhātuka. In it, the carried word ārdhadhātuke is a locative of domain (viṣayasaptamī), not of something that actually follows. So vī replaces aj as soon as yaṅ is intended, before yaṅ is added: vevīyate. One of Bhaṭṭoji's commentators adds why this matters. Vī begins with a consonant, so yaṅ can now come at all.

But aj has no yaṅ-luk, says Bhaṭṭoji. When the yaṅ is removed by luk, there is no domain of an ārdhadhātuka, so vī does not come. One of his commentators puts it thus: a yaṅ removed by luk cannot be the thing intended in advance.

Here ends the yaṅluganta-prakriyā.

Watch out. SK 2656 is a sūtra of the first book (A 1.1.4), though Bhaṭṭoji teaches it here. Its prohibition needs two things: an ārdhadhātuka, and a deletion that this ārdhadhātuka causes. The endings tip and the rest are sārvadhātuka, so it cannot touch totorti, momotti or bobhoti. And where an ārdhadhātuka does follow, as tās in momoditā, the luk of this chapter was not caused by it, so the guṇa stands (SK 2651).

Terms used in this chapter

Summary

SKWhat it doesExample
2650luk of yaṅ before the suffix ac, and variously without it. The unconditioned luk comes first; the stem still doubles; parasmaipada only; śap deleted as in the ad classbhū + yaṅ → bobhū; loluva
2651optional īṭ before a consonant-initial pit sārvadhātuka after a yaṅ-final stembobhavīti / bobhoti
2652ruk, rik, rīk to the abhyāsa of a root with penultimate ṛ, in yaṅ-lukvarvarti, varivarti, varīvarti
2653the same for a root ending in short ṛcarkarti (3), ararti / ariyarti
2654ūṭh for the penultimate and v of jvar, tvar, srivi, av, mav before kvi, a jhal or a nasalmāmoti, otuḥ, om
2655ch and v deleted after r before kvi, a jhal or a nasaltotorti, totūrtaḥ, momorti
2656no guṇa or vṛddhi of ik before an ārdhadhātuka that causes part of the dhātu to be deletedloluvaḥ; not in totorti

The roots of SK 2651 and SK 2653, with the step that marks each:

RootStemFormsWhat to notice
bhūbobhūbobhavīti / bobhoti, abobhūvītguṇa allowed in speech (bobhūtu is a Vedic niyama); vuk is nitya
spardhpāspardhpāspardhīti / pāspardhi, apāspāḥno ātmanepada; ru, ro ri, lengthening
gādhjāgādhjāgāddhi, jāghātsi, ajāghāḥkuhoś cuḥ, dīrgho 'kitaḥ
nāth, dadhnānāth, dādadhnānātti; dādaddhi, adādhaḥ / adādhatlengthened abhyāsa
skundcoskundcoskundīti / coskuntiśarpūrvāḥ khayaḥ
mudmomudmomudīti / momotti, momoditāSK 2503 with īṭ; guṇa without it
kurdcokūrdcokūrti, acokūḥru option
vañcvanīvañcvanīvaṅkti, vanīvaktaḥnīk; nasal kept before pit
gamjaṅgamjaṅganti, jaṅgataḥ, jaṅgamitā, jaṅgahiekāc rule fails; nasal loss asiddha for ato heḥ; no aṅ
hanjaṅghanjaṅghanīti, jaṅghahi, vadhyātgh by SK 2430 (Nyāsa); no ja; vadha not doubled
carcañcurcañcūrti, acañcūḥtapara u: no guṇa, but hali ca
khancaṅkhancaṅkhātaḥ, caṅkhāyāt / caṅkhanyātā by SK 2504
yu, nuyoyu, nonuyoyoti / yoyavīti, nonotino vṛddhi by SK 2443
hājāhājāheti / jāhāti, jāhītaḥfive śtip-cited rules fail
svapsāsvapsāsvapīti / sāsvapti, sāsupyātno u from SK 2645; no iṭ from SK 2474
vṛtvarvṛt (3)varvarti, varvṛtīti, avarvart, varvartiṣyatiSK 2348 and aṅ fail (class citations)
kṛcarkṛ (3)carkarti, carkriyāt, acarkārītSK 2653
kṝ, tṝcākṝ, tātṝcākarti, tātartitapara ṛt: no augment
ṛarṛ, ariyṛararti / ariyarti, ārati, āriyātiyaṅ; ro ri; SK 2380 fails
gṛh / grahjargṛh (3) / jāgrahjargarḍhi; jāgrāḍhi, jāgṛḍhaḥ, jāgrahitābahiraṅga saṃprasāraṇa takes no ruk; no long ī
gṛdhjargṛdh (3)ajarghart, ajarghāḥbhaṣ, jaś, car
pracchpāpracchpāpraṣṭi, pāpraṣṭaḥ, pāpraśmino saṃprasāraṇa (śtip)
haya, haryajāhay, jāharyjāhati, jāhāmi; jāhartiy lost before val; ūṭh roots of SK 2561 have no yaṅ-luk
mavmāmavmāmoti, māmāvaḥ, māmūmaḥSK 2654
turv, murchtoturv, momurchtotorti, totūrtaḥ, momortiSK 2655; SK 2656 does not bar guṇa
← textbook contentsreading page (Technical) →