After this chapter you will be able to derive the taddhita words of SK 1974 to SK 2050. There are 77 sūtras, most of them from the third quarter of Pāṇini's fifth book. The chapter opens with words of direction. The east as the place where, or the place from which, gives purastāt and puraḥ. The north gives uttarataḥ, uttarāt, uttareṇa and uttarāhi. The eastern quarter prācī gives prāk, ūrdhva "upper" gives upari, and apara "western" gives paścāt. Next comes dhā, "in so many ways": caturdhā "in four ways", with dvidhā, dvaidham and dvedhā for "in two ways". Then come small affixes that add a shade without changing the thing meant. A contemptible physician is bhiṣakpāśaḥ, a sixth part is ṣāṣṭhaḥ, a man alone is ekākī, and one who was formerly rich is āḍhyacaraḥ. The degrees of comparison follow: laghutamaḥ and laghiṣṭhaḥ "lightest", laghutaraḥ and laghīyān "lighter", and even pacatitamām "he cooks exceedingly well". Some stems change before -iṣṭha and -īyas, as in śreṣṭhaḥ, jyāyān "elder" and bhūyān. Four affixes follow for "thoroughly", "almost" and "of the sort": paṭurūpaḥ, vidvatkalpaḥ "almost learned", bahupaṭuḥ, paṭujātīyaḥ. From SK 2025 the affix ka is the heading. Beside it, akac goes inside indeclinables, pronouns and verb-forms. An unknown horse is aśvakaḥ, and we get uccakaiḥ and sarvake. Ka also shows compassion, putrakaḥ, and the name of a pitied person may be cut down: devadatta gives devikaḥ and devilaḥ beside devadattakaḥ. Then come the small and the short: tailakam "a little oil", vṛkṣakaḥ "a short tree", vatsataraḥ "a calf past its first age". Last come kataraḥ "which of the two", katamaḥ "which of many", and a kan for the means by which someone is disparaged: vyākaraṇakena garvitaḥ, "puffed up by his grammar". Every rule is worked on a concrete word.
The name prāgivīya means "taught before iva". The affixes of this chapter all come before the sūtra ive pratikṛtau (SK 2051), which opens chapter 42. Chapter 40 gave the affixes taught before diś "direction", and diś is the first word of this chapter's first sūtra. One of Bhaṭṭoji's commentators opens the new section at SK 1974. Bhaṭṭoji mostly keeps Pāṇini's order here (A 5.3.27–95), but he moves a few rules. He also brings in nine rules from elsewhere in the grammar, each where it builds a form: a name (SK 2003), the affix āmu (SK 2004), and seven rules of stem change before -iṣṭha, -iman and -īyas (among SK 2008–2018).
pūrvasyāṃ / pūrvasyāḥ / pūrvā dik → purastāt, puraḥThe first rule gives astāti, heard as -astāt, after a word for a direction. The word may stand in the seventh case ("in the east"), the fifth ("from the east") or the first ("the east"). The next three rules are Pāṇini's A 5.3.39–41, which Bhaṭṭoji moves up to stand here. Before asi (heard as -as) and before astāti, pūrva "eastern" is replaced by pur. On the running example, the east as the place where, the place from which, or the quarter itself gives puraḥ and purastāt.
Rule. After a word for a direction, astāti comes in the word's own sense. Three conditions hold. The word must be conventionally used (rūḍha) for a direction. It must stand in the seventh, fifth or first case. And it must denote a direction, a place or a time. Astāti is heard as -astāt: one of Bhaṭṭoji's commentators says its i is only for pronunciation.
Example.
pūrvasyāṃ / pūrvasyāḥ / pūrvā dik "in the east / from the east / the east" → pūrvā + astāt
→ pur replaces pūrvā (SK 1976) → pur + astāt
→ purastāt.
One of Bhaṭṭoji's commentators gives the three uses: purastād vasati "he lives in the east", purastād āgataḥ "he has come from the east", purastād ramaṇīyam "the east is pleasant".
Applies when / blocked by. Any of the three cases goes with any of the three senses. Both of Bhaṭṭoji's commentators say there is no yathāsaṃkhya here, on different grounds. One rests on the received interpretation. The other says the rule is not marked svarita.
One of Bhaṭṭoji's commentators draws "conventionally used" from the word śabda "word" in the sūtra. He adds that this whole rule, without astāti itself, runs on as far as SK 1988. So every direction affix of SK 1975–1987 inherits its three conditions.
Watch out. Our text of Bhaṭṭoji's paraphrase has two slips: didgeśa- for digdeśa- "direction, place", and astīti for astāti. The 1915 printed edition reads digdeśa- and astāti, and this chapter follows the print. Bhaṭṭoji tests the three conditions with counter-examples under SK 1977.
Rule. After pūrva, adhara and avara, the affix asi comes in the sense of astāti. With it, pur, adh and av replace them respectively. Asi is heard as -as. One of Bhaṭṭoji's commentators explains "the sense of astāti": the three words must denote a direction, a place or a time, as in SK 1974.
Example.
pūrvā + asi → pur replaces pūrvā → pur + as → puraḥ.
Likewise adhara + asi → adh → adhaḥ; avara + asi → av → avaḥ.
One of Bhaṭṭoji's commentators gives these steps. Bhaṭṭoji himself lists the forms under SK 1976.
Watch out. This rule is A 5.3.39, taken out of Pāṇini's order. One of Bhaṭṭoji's commentators explains why Bhaṭṭoji sets it here. It shows that its words run on into the next rule, astāti ca.
Rule. Pur, adh and av also replace pūrva, adhara and avara when astāti follows. One of Bhaṭṭoji's commentators says the words of SK 1975 are carried forward.
Example.
pūrvasyāṃ / pūrvasyāḥ / pūrvā dik → pūrvā + astāt → pur for pūrvā → purastāt.
Bhaṭṭoji pairs each asi form with its astāti form: puraḥ, purastāt; adhaḥ, adhastāt; avaḥ, avastāt.
Watch out. These rules do not make purataḥ, though Amara's lexicon and the poet Bhavabhūti use it. One of Bhaṭṭoji's commentators reports two views. Many call it a slip. Some say the first a of atasuc (SK 1978) lets that affix come after other words too; scholars discuss this.
Rule. For avara, av is only optional before astāti. The sūtra's word vibhāṣā means "optionally". One of Bhaṭṭoji's commentators says astāti is carried forward from SK 1976.
Example.
avara + astāt → with av → avastāt;
without av → avarastāt.
Bhaṭṭoji adds: "Likewise for place and time."
Applies when / blocked by. Bhaṭṭoji now tests the three conditions of SK 1974, one counter-example each. In none of them does astāti come.
aindryāṃ vasati "he lives in Indra's [quarter]".pūrvaṃ grāmaṃ gataḥ "he went to the eastern village".pūrvasmin gurau vasati "he lives with his former teacher".Bhaṭṭoji then draws a jñāpaka. Asi is a special rule for pūrva, adhara and avara; astāti is the general rule. Would asi block astāti after these three words? No: astāti ca (SK 1976) gives pur before astāti, so astāti must still come after them. One of Bhaṭṭoji's commentators sets out this objection and answer.
Watch out. Our text adds a stray pair of sentences: "Why 'after seventh-case words and the rest'? aindryāṃ vasati." The 1915 printed edition does not have this pair, and this chapter leaves it out. In the print the counter-example aindryāṃ vasati answers only "why 'conventionally used for a direction'?", and pūrvaṃ grāmaṃ gataḥ answers the question about the cases.
uttarā dik → uttarataḥ, uttarāt, uttareṇa, uttarā, uttarāhiThe next ten rules give the other direction affixes. Each works under the conditions of SK 1974 and takes the place of astāti for particular words, or deletes it. On the running example, uttara "northern" ends up with five direction forms: uttarataḥ, uttarāt, uttareṇa "not far to the north", uttarā and uttarāhi. One rule of Pāṇini's in this stretch, A 5.3.33, is Vedic. Bhaṭṭoji leaves it for his section on Vedic forms.
Rule. After dakṣiṇa "southern" and uttara "northern", the affix atasuc comes, under the conditions of SK 1974. Bhaṭṭoji calls it an apavāda of astāti. Atasuc is heard as -atas.
Example.
uttara + atas → uttarataḥ.
dakṣiṇa + atas → dakṣiṇataḥ.
Watch out. One of Bhaṭṭoji's commentators discusses the first a of atasuc. A plain tasuc would give the same forms, so the a is only for clarity. That is why the Bhāṣya sets it aside. The question of purataḥ (SK 1976) turns on this a.
Rule. After para and avara, atasuc comes optionally. One of Bhaṭṭoji's commentators supplies atasuc from SK 1978. In the other option astāti comes.
Example.
para + atas → parataḥ; or para + astāt → parastāt.
avara + atas → avarataḥ; or avara + astāt → avarastāt.
Rule. After a direction-word ending in the root añc, astāti undergoes luk. Bhaṭṭoji names astāti in his own paraphrase.
Example.
prācyāṃ / prācyāḥ / prācī [dik] "in / from / the eastern [quarter]" → prācī + astāt
→ luk of astāti → the feminine affix goes too (SK 1408)
→ prāk.
Bhaṭṭoji cites SK 1408 here. One of his commentators gives the uses: prāg vasati "he lives in the east", prāg āgataḥ "he has come from the east". Likewise udak, from the northern quarter. Bhaṭṭoji adds: "Likewise for place and time."
Applies when / blocked by. The rule names añc. How does it cover the feminine prācī? One of Bhaṭṭoji's commentators answers with the liṅgaviśiṣṭa-paribhāṣā: a stem named in a rule covers its gender-marked form too.
Rule. In the domain of astāti, upa replaces ūrdhva "upper". The affix is ril or riṣṭātil, heard as -ri and -riṣṭāt.
Example.
ūrdhva + ril → upa for ūrdhva → upa + ri → upari.
ūrdhva + riṣṭātil → upa + riṣṭāt → upariṣṭāt.
Bhaṭṭoji gives upari or upariṣṭād vasati, āgataḥ, ramaṇīyaṃ vā. One of his commentators says the three show the three cases in order: "he lives above", "he has come from above", "what is above is pleasant".
Watch out. Ūrdhva is the word that gets replaced. It is not a meaning in which the affix comes. One of Bhaṭṭoji's commentators says Bhaṭṭoji's wording repeats a sentence of the Bhāṣya.
Rule. In the domain of astāti, paśca replaces apara "western", and the affix āti comes. One of Bhaṭṭoji's commentators says Bhaṭṭoji's wording repeats a sentence of the Bhāṣya, as at SK 1981. He adds that the i of āti is for pronunciation, so it is heard as -āt.
Example.
apara + āti → paśca for apara → paśca + āt → paścāt.
Rule. After uttara, adhara and dakṣiṇa, the affix āti comes, heard as -āt. One of Bhaṭṭoji's commentators calls it an apavāda of astāti.
Example.
uttara + āt → uttarāt.
Likewise adharāt, dakṣiṇāt.
Each form carries the three case-senses and the three senses of SK 1974.
Rule. After uttara, adhara and dakṣiṇa, the affix enap comes optionally, heard as -ena. Two conditions hold. The sense is "not far" (adūre). And the word is not in the fifth case. One of Bhaṭṭoji's commentators says the three words are carried forward from SK 1983.
Bhaṭṭoji explains "not far". The limit (avadhi) and what is placed with respect to it (avadhimat) are close to each other.
Example.
uttara + ena → uttareṇa "not far to the north".
Likewise adhareṇa, dakṣiṇena.
One of Bhaṭṭoji's commentators construes these with vasati "he lives" or ramaṇīyam "is pleasant".
In the fifth case there is no enap: uttarād āgataḥ "he has come from the north".
Applies when / blocked by. In the other option each word takes its own affixes. Bhaṭṭoji's commentators list them differently. One names astāti, asi and āti. The other gives dakṣiṇataḥ, uttarataḥ, adhaḥ, adhastāt, uttarāt and the like.
Bhaṭṭoji reports that some do not carry uttara and the rest forward. They give enap after any direction-word: pūrveṇa grāmam "close east of the village", apareṇa grāmam "close west of the village".
Watch out. "Not in the fifth case" reaches beyond this rule. One of Bhaṭṭoji's commentators cites the Bhāṣya: it holds up to asi, A 5.3.39 in Pāṇini's order. So it covers SK 1985–1987 as well.
Rule. After dakṣiṇa, the affix āc comes in the domain of astāti. It is heard as -ā. Bhaṭṭoji says "not after a fifth-case word" holds here too.
Example.
dakṣiṇa + ā → dakṣiṇā vasati "he lives to the south".
In the fifth case, no āc: dakṣiṇād āgataḥ "he has come from the south".
Applies when / blocked by. "Not far" does not carry over. One of Bhaṭṭoji's commentators reads this from Bhaṭṭoji's words "in the domain of astāti". He counts three forms: dakṣiṇā (āc), dakṣiṇāt (āti, SK 1983) and dakṣiṇataḥ (atasuc, SK 1978).
Rule. After dakṣiṇa, in the sense "far" (dūre), the affix āhi comes. By ca "and", āc comes as well. One of Bhaṭṭoji's commentators supplies "after dakṣiṇa".
Example.
dakṣiṇa + āhi → dakṣiṇāhi.
dakṣiṇa + ā → dakṣiṇā.
Rule. After uttara too, āc and āhi come. One of Bhaṭṭoji's commentators supplies the two affixes.
Example.
uttara + āhi → uttarāhi.
uttara + ā → uttarā.
The same commentator counts four forms with atasuc and āti: uttarāhi, uttarā, uttarataḥ and uttarāt. Add uttareṇa for "not far" (SK 1984), and uttara has the five forms of the running example.
dvi → dvidhā, dvaidham, dvedhāTwo rules give dhā, "in so many ways", after number-words. Three more give substitutes for it. A substitute (ādeśa) is an item put in the place of another. On the running example, dvi "two" gives dvidhā with dhā, dvaidham with dhamuñ, and dvedhā with edhāc. One of Bhaṭṭoji's commentators says the conditions of SK 1974 run on as far as SK 1988, the first rule of this group.
Rule. After a number-word used in the sense of vidhā "mode", the affix dhā comes in the word's own sense. Bhaṭṭoji narrows the mode to a mode of action (kriyāprakāra). Both of his commentators say this narrowing comes from the nature of expression.
Example.
catur "four" + dhā → caturdhā "in four ways".
pañcan "five" + dhā → pañcadhā "in five ways".
On the running example, dvi + dhā → dvidhā "in two ways".
One of Bhaṭṭoji's commentators says a verb such as gacchati "goes" is understood: "he goes in four ways".
Watch out. Vidhā can also mean a lump of cooked rice. One of Bhaṭṭoji's commentators says that sense is not taken, so no affix comes in ekā govidhā. That is why the sūtra says vidhārthe, "in the meaning of vidhā". It means the sense in which vidhā is well known, "mode".
Rule. Dhā also comes when a substance (adhikaraṇa) is brought to another number (vicāla). Bhaṭṭoji's paraphrase shows that "after a number-word" and dhā carry over from SK 1988.
Example.
ekaṃ rāśiṃ pañcadhā kuru "make the one heap into five": pañcan + dhā → pañcadhā.
Both of Bhaṭṭoji's commentators give the reverse case: anekam ekadhā kuru "make the many into one".
Applies when / blocked by. Why a separate rule? Both of Bhaṭṭoji's commentators answer. Here the mode belongs to the heap, not to an action. SK 1988 covers only a mode of action.
Watch out. Our text reads saṃkhyāntarāpādāne. The 1915 printed edition reads saṃkhyāntarāpādane, "in bringing about another number", and both commentators gloss it that way. This chapter follows the print.
Rule. After eka "one", dhyamuñ optionally replaces dhā. Dhyamuñ is heard as -dhyam. Its ñ brings first-vowel vṛddhi (SK 1075).
Example.
eka + dhā → dhyamuñ for dhā → eka + dhyam
→ vṛddhi, e → ai → aikadhyam.
In the other option → ekadhā.
One of Bhaṭṭoji's commentators shows both senses: aikadhyaṃ bhuṅkte for a mode of action (SK 1988), aikadhyaṃ rāśiṃ kuru "make [them] one heap" for a change of number (SK 1989).
Watch out. Dhyamuñ is a substitute, not an affix of its own. One of Bhaṭṭoji's commentators gives the reason. As an affix it would come only in the sense of SK 1989, the nearest rule. The other adds that dhā is named again in the sūtra ("dhaḥ") so that the dhā of both senses is replaced.
Rule. After dvi "two" and tri "three", dhamuñ optionally replaces dhā. It is heard as -dham, and its ñ brings vṛddhi. One of Bhaṭṭoji's commentators says "for dhā" and "optionally" are carried forward from SK 1990.
Example.
dvi + dhā → dhamuñ for dhā → dvi + dham
→ vṛddhi, i → ai → dvaidham.
In the other option → dvidhā.
Likewise traidham, tridhā.
Applies when / blocked by. A vārttika adds: after a word ending in dhamuñ, ḍa is seen in the word's own sense. Ḍa is heard as -a; its ḍ deletes the ṭi. Bhaṭṭoji's example is pathi dvaidhāni tṛṇāni "the grasses on the path are twofold".
dvaidham + ḍa → the ṭi am deleted (SK 316) → dvaidha → dvaidhāni.
One of Bhaṭṭoji's commentators gives the ṭi deletion. He says the vārttika's "is seen" (darśana) means it applies only in some places.
Watch out. Dvaidhāni is a declined adjective agreeing with tṛṇāni. It is not an indeclinable. One of Bhaṭṭoji's commentators gives two reasons. By its nature the word denotes a thing (sattvavacana). And the Bhāṣya itself uses the form dvaidhāni.
Rule. Edhāc also replaces dhā after dvi and tri. It is heard as -edhā.
Example.
dvi + dhā → edhāc for dhā → dvedhā.
Likewise tredhā.
Applies when / blocked by. Why a separate rule, and not "dhamuñ and edhāc" in SK 1991? One of Bhaṭṭoji's commentators answers: the split (yogavibhāga) rules out yathāsaṃkhya. In one rule, dhamuñ might go with dvi only and edhāc with tri only. As it stands, each comes after both.
Watch out. One of Bhaṭṭoji's commentators says the affixes from tasil (SK 1953) up to this edhāc make indeclinables. So purastāt, uttareṇa, dvidhā and dvedhā are indeclinables. The ḍa-form dvaidhāni of SK 1991 is not.
ṣaṣṭho bhāgaḥ → ṣāṣṭhaḥ / ṣaṣṭhaḥ; ṣaṣṭhakaḥ (as a measure)The next eight rules give small affixes in the word's own sense. They mark something as contemptible, as a share, as being alone, or as what it formerly was. The running example is ṣaṣṭha "sixth". A sixth part is ṣāṣṭhaḥ or ṣaṣṭhaḥ, and as a measure it is ṣaṣṭhakaḥ.
Rule. After a word used of something contemptible (yāpya), the affix pāśap comes in the word's own sense. It is heard as -pāśa. One of Bhaṭṭoji's commentators glosses yāpya as kutsita "contemptible", citing Amara's lexicon.
Example.
kutsito bhiṣak "a contemptible physician" → bhiṣaj + pāśa → bhiṣakpāśaḥ.
Applies when / blocked by. One of Bhaṭṭoji's commentators, following the Bhāṣya, limits the rule. The contempt must touch the pravṛttinimitta: here, a physician poor at being a physician. Other contempt takes the ka of SK 2029.
The other commentator explains why pāśap is not taught in SK 2029 itself. There it would come after verb-forms (tiṅanta) too.
Rule. After an ordinal in the affix tīya, the affix an comes in the word's own sense, when the word means a part (bhāga). An is heard as -a; its n is a marker.
Example.
dvitīyo bhāgaḥ "the second part" → dvitīya + an → dvitīyaḥ.
Likewise tṛtīyaḥ "the third part".
The shape does not change. Bhaṭṭoji says: "There is a difference in the accent." One of his commentators names the accent rule, A 6.1.197 (SK 3686). An affix marked with ñ or n makes the first vowel of the word udātta.
Applies when / blocked by. Two vārttikas follow. The first gives īkak, heard as -īka, after tīya in the word's own sense:
dvitīya + īka → dvaitīyīkaḥ, beside dvitīyaḥ;
tṛtīya + īka → tārtīyīkaḥ, beside tṛtīyaḥ.
The second says "not for vidyā", a branch of knowledge. So dvitīyā, tṛtīyā [vidyā] keep their plain form. One of Bhaṭṭoji's commentators says both vārttikas are found in the Bhāṣya on A 4.2.7.
One of Bhaṭṭoji's commentators explains the word pūraṇa "ordinal". Every tīya affix is an ordinal affix, so the word is there for the next rule. In words such as mukhatīya the tīya has no meaning of its own, so no an comes.
Watch out. The same commentator says bhāga is meant in the masculine. The feminine vibhakti "share" means the same, yet it gives only caturthī and pañcamī. It does not give caturthā or pañcamā with the an of SK 1995.
Rule. After the ordinals of the numbers below eleven, an comes in the sense of a part, outside the Veda. Bhaṭṭoji fills in the rest from SK 1994 (śeṣapūraṇa, "completing what remains"): "after a word ending in an ordinal affix, an in the sense of a part".
Example.
caturtho bhāgaḥ "the fourth part" → caturtha + an → caturthaḥ.
Likewise pañcamaḥ "the fifth part".
Applies when / blocked by. One of Bhaṭṭoji's commentators fixes the domain. Dvitīya and tṛtīya already take an by SK 1994. So this rule covers caturtha up to daśama "tenth". Why then SK 1994? So that it applies in the Veda too.
Watch out. Our text gives the second example as aṣṭamaḥ. The 1915 printed edition reads pañcamaḥ, and this chapter follows the print. Aṣṭama has its own rule next.
Rule. After ṣaṣṭha "sixth" and aṣṭama "eighth", the affix ña comes, and by ca "and", an too. Both come in the domain of SK 1995, the sense "part". Ña is heard as -a, and its ñ brings first-vowel vṛddhi (SK 1075).
Example.
ṣaṣṭho bhāgaḥ "the sixth part" → ṣaṣṭha + ña → vṛddhi, a → ā → ṣāṣṭhaḥ;
ṣaṣṭha + an → ṣaṣṭhaḥ.
Likewise āṣṭamaḥ (ña) and aṣṭamaḥ (an).
Applies when / blocked by. Both affixes come after both words. One of Bhaṭṭoji's commentators says there is no yathāsaṃkhya here, on the received interpretation.
Rule. After ṣaṣṭha and aṣṭama, kan comes when a measure (māna) is meant, and luk when a part of an animal (paśvaṅga) is meant. By ca, ña and an come as well. Kan is heard as -ka.
Example.
ṣaṣṭho bhāgaḥ, if it is a measure → ṣaṣṭha + ka → ṣaṣṭhakaḥ.
aṣṭamo bhāgaḥ "an eighth part", if it is part of an animal → aṣṭama + ña or an → luk → aṣṭamaḥ.
By ca, also ṣaṣṭhaḥ, ṣāṣṭhaḥ; aṣṭamaḥ, āṣṭamaḥ.
Applies when / blocked by. Bhaṭṭoji draws a jñāpaka. The great option would already allow the plain word with no affix. So the statement of luk shows that ña and an in SK 1996 are obligatory. One of his commentators explains: only then does a luk that applies in some cases have a purpose.
Rule. After eka meaning "without a companion" (asahāya), the affix ākinic comes in the word's own sense. It is heard as -ākin. By ca, kan and luk come too. They are drawn in from SK 1997 by the "and" (anukarṣaṇa), not carried forward by anuvṛtti.
Example.
eka + ākin → ekākī "alone".
eka + ka → ekakaḥ.
With luk → ekaḥ.
One of Bhaṭṭoji's commentators says the luk is of ākinic or of kan.
Watch out. "Without a companion" rules out the numeral "one". The same commentator draws the result. Two or many people can each be without a companion, so the dual ekākinau and the plural ekākinaḥ are possible. The numeral "one" could not stand in them.
Rule. After a word meaning "formerly [such]" (bhūtapūrva), the affix caraṭ comes in the word's own sense. It is heard as -cara. One of Bhaṭṭoji's commentators says the base is a nominal stem.
Example.
āḍhyo bhūtapūrvaḥ "one who was formerly rich" → āḍhya + cara → āḍhyacaraḥ.
Rule. After a sixth-case word, in the sense "formerly", the affix rūpya comes, and caraṭ too. One of Bhaṭṭoji's commentators says bhūtapūrve is carried forward from SK 1999.
Example.
kṛṣṇasya bhūtapūrvo gauḥ "the cow that formerly was Kṛṣṇa's" → kṛṣṇa + rūpya → kṛṣṇarūpyaḥ;
kṛṣṇa + cara → kṛṣṇacaraḥ.
Applies when / blocked by. What is "former" here? Not Kṛṣṇa. Both of Bhaṭṭoji's commentators attach it elsewhere. One says the relation of belonging is past. The other says the cow, the principal thing, is qualified.
Watch out. A feminine base keeps its feminine form before rūpya: śubhrāyā bhūtapūrvaḥ [gauḥ] → śubhrārūpyaḥ. Bhaṭṭoji gives the reason. Rūpya is not counted among tasil and the affixes listed with it, so there is no puṃvadbhāva. One of his commentators names the list: the vārttika tasilādayaḥ prāk pāśapaḥ, "tasil and the rest, before pāśap", which serves SK 836.
laghu → laghutamaḥ / laghiṣṭhaḥ; laghutaraḥ / laghīyānSix rules give the degrees of comparison. For one of many, the affixes are tamap (heard as -tama) and iṣṭhan (heard as -iṣṭha). For one of two, they are tarap (-tara) and īyasun (-īyas, first case -īyān). On the running example, laghu "light" gives laghutamaḥ and laghiṣṭhaḥ "the lightest", laghutaraḥ and laghīyān "the lighter". Two rules brought in from elsewhere in the grammar name tarap and tamap "gha" and add āmu after them: pacatitamām "he cooks exceedingly well". The stem changes before iṣṭha and īyas follow this group.
Rule. Tamap and iṣṭhan come in the word's own sense after a word whose meaning is qualified by excellence (atiśaya). Both of Bhaṭṭoji's commentators say the excellence qualifies the base's meaning. The affix only makes it manifest (dyotaka).
Example.
ayam eṣām atiśayena āḍhyaḥ "this one is rich in excess among these" → āḍhya + tama → āḍhyatamaḥ "the richest".
laghu + tama → laghutamaḥ "the lightest".
laghu + iṣṭha → the ṭi u deleted (SK 1786) → laghiṣṭhaḥ.
One of Bhaṭṭoji's commentators gives this last step. The u is deleted, not replaced by its guṇa o.
Applies when / blocked by. This rule is for one of many. Both of Bhaṭṭoji's commentators reason by elimination. One of two is given to tarap and īyasun by SK 2005, an apavāda. So tamap and iṣṭhan are left with one singled out from many.
Both commentators take the base to be a subanta, a word with its case-ending. One points to the kept seventh-case ending in pūrvāhṇetamām "very early in the forenoon". Without a subanta base, that form could not be explained.
Watch out. Excellence is not the affix's own meaning. One of Bhaṭṭoji's commentators says that reading would make laghutama mean "the excelling of the light". Another point from the other commentator: no second affix of excellence comes after a word that already has one. Usage does not express the sense that way (anabhidhāna). The form śreṣṭhatamāya karmaṇe is Vedic.
Rule. After a tiṅanta, tamap comes when excess is to be shown. Only tamap comes. Both of Bhaṭṭoji's commentators say iṣṭhan is kept out by SK 2006.
Example.
pacati "he cooks" + tamap → pacatitama
→ āmu added (SK 2004) → pacatitamām "he cooks exceedingly well".
Bhaṭṭoji gives the form under SK 2004.
Applies when / blocked by. One of Bhaṭṭoji's commentators says why the rule is needed. A verb-form is not a nominal stem, so no taddhita could reach it otherwise.
Rule. Tarap and tamap are called gha. This saṃjñā rule belongs to Pāṇini's first book (A 1.1.22). Bhaṭṭoji places it here, just before SK 2004, the rule that uses the name.
Example.
In pacatitama the tama is a gha. So SK 2004 can add āmu: pacatitamām.
Applies when / blocked by. Both of Bhaṭṭoji's commentators draw a jñāpaka from the rule's place. Pāṇini could have named the affixes with a shorter rule at the end of this section. Instead he made a separate, heavier sūtra in another section. That reveals a tarap used purely in the word's own sense, with no excellence. An example is alpāctaram "having few vowels". One commentator adds that tarap there does not mean one of two.
Rule. Āmu comes after a word ending in a gha, when the gha follows kim "what", a word ending in e, a tiṅanta, or an indeclinable. It does not come when the excellence (prakarṣa) is of a substance (dravya). Āmu is heard as -ām.
Example.
kim + tama + ām → kiṃtamām.
prāhṇe "in the forenoon" + tama + ām → prāhṇetamām "very early in the forenoon".
pacati + tama + ām → pacatitamām "he cooks exceedingly well".
uccaiḥ "loudly" + tama + ām → uccaistamām "very loudly".
Excellence of a substance takes no āmu: uccaistamas taruḥ "a very tall tree".
Applies when / blocked by. One of Bhaṭṭoji's commentators explains the examples. In kiṃtamām the tamap is purely in the word's own sense: one cannot say "the most 'who' among these". In prāhṇetamām the day excels through its early part, and a day is not a substance. In pacatitamām it is the action of cooking that excels. In uccaistamas taruḥ the height is seen in the tree, a substance.
Watch out. Why the u of āmu? Both of Bhaṭṭoji's commentators answer. Without the u, the rule that adds the augment nuṭ before ām after a short final would apply. An augment (āgama) is a sound added to an item without replacing anything. It would give the wrong pacatitarāṇām. A paribhāṣā keeps the marked āmu out of that rule: an item named without markers does not include one with markers.
Rule. Tarap and īyasun come when a word stated nearby (the upapada) expresses two, or names the group from which something is divided off (vibhajya). Bhaṭṭoji calls this an apavāda of the two rules before, SK 2001–2002. The affixes come after a subanta or a tiṅanta.
Example.
ayam anayor atiśayena laghuḥ "this one is light in excess of these two" → laghu + tara → laghutaraḥ;
laghu + īyas → the ṭi u deleted (SK 1786) → laghīyas → laghīyān "the lighter".
udīcyāḥ prācyebhyaḥ paṭutarāḥ, paṭīyāṃsaḥ "the northerners are cleverer than the easterners".
Both of Bhaṭṭoji's commentators say prācyebhyaḥ is in the fifth case by SK 639. They also give the steps of the first case laghīyān. The augment num comes, because īyasun counts as marked with u. Then come lengthening (SK 317) and two deletions (SK 252, SK 54).
Applies when / blocked by. "Expressing two" does not mean the dual ending. One of Bhaṭṭoji's commentators says a word that conveys two things is enough. Both commentators say there is no yathāsaṃkhya: each affix comes with either kind of upapada. One also reports that some allow these affixes even with no word for two, citing Bhāṣya usages; scholars discuss this.
Watch out. Vibhajya means "what is to be divided off": the group from which one part is set apart. In the second example that group is the easterners, prācyebhyaḥ. The upapada here is simply a word stated nearby in the analysing sentence. One of Bhaṭṭoji's commentators says it is not the technical upapada.
Rule. The two vowel-initial affixes, iṣṭhan and īyasun, come only after a quality-word (guṇavacana).
Example.
Laghu "light" is a quality-word, so it takes all four affixes (SK 2001, 2005).
Pācaka "cook" is not: pācaka + tara → pācakataraḥ; pācaka + tama → pācakatamaḥ. No iṣṭhan or īyasun.
One of Bhaṭṭoji's commentators calls pācaka an action-word (kriyāśabda).
Applies when / blocked by. The eva "only" restricts the affix, not the base (pratyayaniyama). Both of Bhaṭṭoji's commentators say so. Without eva, the rule could be read the other way: "after a quality-word, only the vowel-initial affixes". Then paṭutaraḥ and paṭutamaḥ would be lost. This rule is also why a tiṅanta takes no iṣṭhan (SK 2002).
Watch out. The word ajādī, "the two vowel-initial ones", runs on through the next group. Here it is a first-case word. From SK 2009 on, both of Bhaṭṭoji's commentators say it is carried forward and recast in the seventh case: "before iṣṭhan and īyasun". There the two affixes mark where a stem is replaced, as praśasya by śra. Wherever group F says "before iṣṭhan and īyasun", that is this word at work.
praśasya → śreṣṭhaḥ, śreyān; jyeṣṭhaḥ, jyāyānSK 2006 kept iṣṭhan and īyasun for quality-words alone. This group opens with one rule that lets them back in, and then changes the stem before them. Many bases do not simply add -iṣṭha and -īyas. Some lose a part, some are replaced by a short substitute, and some keep a ṭi that would otherwise go. The rules brought in from Pāṇini's sixth book name the affixes without their markers: iṣṭha, iman and īyas. On the running example, praśasya "praiseworthy" is replaced by śra or by jya. So "most praiseworthy" is śreṣṭhaḥ or jyeṣṭhaḥ, and "more praiseworthy" is śreyān or jyāyān.
Rule. In the Veda (chandas), iṣṭhan and īyasun come after a word ending in tṛ. Bhaṭṭoji glosses tṛ as the affixes tṛn and tṛc, which make words like kartṛ "doer". Such a word is no quality-word, yet it takes the pair.
Example. Bhaṭṭoji gives the form under the next rule.
atiśayena kartā "a doer in a high degree" → kartṛ + iṣṭhan
→ the tṛ deleted (SK 2008) → kariṣṭhaḥ.
Applies when / blocked by. SK 2006 shut iṣṭhan and īyasun out after any word that is not a quality-word. This rule lets them back in after a tṛ-final word. Both of Bhaṭṭoji's commentators call it a pratiprasava: a rule that re-admits what a restriction had shut out. It is not a new rule of its own. The word ajādī is carried forward from SK 2006.
Watch out. Why does a Vedic rule stand among the rules of ordinary speech? Bhaṭṭoji's commentators give different reasons. One says the place shows that it is a pratiprasava of SK 2006. The other says a later extension, ṇāv iṣṭhavat, uses it in ordinary speech too. That extension treats a base before the verb-forming affix ṇi as if it stood before iṣṭha. His example is kartāram ācaṣṭe "he calls him a doer" → kārayati.
Rule. Before iṣṭha, iman and īyas, the element tṛ is deleted. The whole of tṛ goes, not just its last sound.
Example.
atiśayena kartā "a doer in a high degree" → kartṛ + iṣṭha
→ the whole tṛ deleted → kar + iṣṭha
→ kariṣṭhaḥ.
iyam anayor atiśayena dogdhrī "this one of the two is the better milker" → dohīyasī dhenuḥ "the better-milking cow". Bhaṭṭoji's commentators give the steps.
dogdhrī → first treated as masculine (puṃvadbhāva), so the feminine ī goes → dogdhṛ + īyas
→ the whole tṛ deleted → the gh and g made for h go back, their cause being gone → doh + īyas
→ the feminine ṅīp, since īyasun is marked with u → dohīyasī.
The o of doh stays. One commentator says the deleted affix still counts for that guṇa.
Applies when / blocked by. One of Bhaṭṭoji's commentators argues for "the whole". The deletion of the ṭi ṛ before iṣṭha is already taught by ṭeḥ (SK 1786). A rule that did only that would be pointless. So this rule must delete the whole tṛ.
Watch out. This is a sixth-book rule, set here beside SK 2007 because it builds that rule's forms. Its word iṣṭhemeyassu "before iṣṭha, iman and īyas" is carried forward into SK 2010 and SK 2015–2018. One of Bhaṭṭoji's commentators says so.
Rule. Before iṣṭhan and īyasun, śra replaces praśasya "praiseworthy". Bhaṭṭoji writes "before the two vowel-initial ones".
Example.
praśasya + iṣṭhan → śra for praśasya → śra + iṣṭha.
praśasya + īyasun → śra + īyas.
Bhaṭṭoji gives the finished forms under SK 2010: śreṣṭhaḥ, śreyān.
Applies when / blocked by. One of Bhaṭṭoji's commentators asks why the rule says "before the vowel-initial ones". So that śra does not come before tarap and tamap: praśasyataraḥ, praśasyatamaḥ.
Watch out. Praśasya is not a quality-word, so SK 2006 would keep the pair away from it. One of Bhaṭṭoji's commentators calls it an action-word. This rule is a jñāpaka: by giving a substitute before iṣṭhan and īyasun, it shows that praśasya does take them.
Rule. Before iṣṭha, iman and īyas, a base with a single vowel (ekāc) stays as it is (prakṛtyā). So its ṭi is not deleted.
Example.
śra + iṣṭha → the ṭi a would be deleted by ṭeḥ (SK 1786)
→ śra has one vowel, so it stays as it is → śreṣṭhaḥ "most praiseworthy".
śra + īyas → śreyān "more praiseworthy".
Applies when / blocked by. "Before iṣṭha and the rest" is carried forward from SK 2008. One of Bhaṭṭoji's commentators says this "staying as it is" (prakṛtibhāva) bars every deletion of the section. He names four deletions of a bha stem:
The other commentator says that in śreṣṭhaḥ neither ṭeḥ nor yasyeti ca applies.
Watch out. Our text of the Kaumudī reads iṣṭhātiṣu in Bhaṭṭoji's gloss. The 1915 printed edition reads iṣṭhādiṣu "before iṣṭha and the rest", and so do Bhaṭṭoji's commentators. We follow the print.
Rule. Jya also replaces praśasya before iṣṭhan and īyasun. The word praśasya is carried forward from SK 2009.
Example.
praśasya + iṣṭhan → jya for praśasya → jya + iṣṭha
→ jya has one vowel, so its ṭi stays (SK 2010) → jyeṣṭhaḥ "most praiseworthy".
Watch out. Before īyasun the same steps would give jyeyān. That is wrong. One of Bhaṭṭoji's commentators points this out, and the next rule mends it.
Rule. After jya, ā replaces the first sound of īyas.
Example.
praśasya + īyasun → jya (SK 2011) → jya + īyas
→ ā for the ī of the affix → jya + āyas
→ jyāyān "more praiseworthy".
Applies when / blocked by. Bhaṭṭoji cites ādeḥ parasya (SK 44): a substitute taught for what follows replaces its first sound. One of his commentators explains why the citation is needed. Without it, the ā would replace the last sound of īyas.
Rule. Jya also replaces vṛddha "old" before iṣṭhan and īyasun. One of Bhaṭṭoji's commentators says the plain word vṛddha is meant. It is not vṛddha, the technical name for a word whose first vowel is a vṛddhi vowel (SK 1335).
Example.
ayam anayor atiśayena vṛddhaḥ "this one of the two is exceedingly old" → vṛddha + īyasun
→ jya → jya + āyas (SK 2012) → jyāyān "elder".
vṛddha + iṣṭhan → jyeṣṭhaḥ "eldest".
Applies when / blocked by. The sūtra itself says only "and of vṛddha". Bhaṭṭoji supplies the rest: "the substitute jya before the two vowel-initial ones". One of his commentators calls this śeṣapūraṇa.
Watch out. Vṛddha has a second substitute, varṣi, by SK 2016. One of Bhaṭṭoji's commentators says it is one option beside jya. So "eldest" is also varṣiṣṭhaḥ.
Rule. Before iṣṭhan and īyasun, neda replaces antika "near", and sādha replaces bāḍha "intense". Bhaṭṭoji supplies "before the two vowel-initial ones" (śeṣapūraṇa).
Example.
antika + iṣṭhan → neda → nediṣṭhaḥ "nearest".
antika + īyasun → nedīyān "nearer".
bāḍha + iṣṭhan → sādha → sādhiṣṭhaḥ "most intense".
bāḍha + īyasun → sādhīyān "more intense".
One of Bhaṭṭoji's commentators glosses bāḍha as "intense", citing the lexicon of Amara.
Rule. The rule names six words: sthūla "stout", dūra "far", yuvan "young", hrasva "short", kṣipra "quick" and kṣudra "small". Before iṣṭha, iman and īyas, their later part that begins with a yaṇ is deleted. Yaṇ is the short name for the semivowels y, v, r, l. The vowel before that part takes guṇa.
Example.
sthūla + iṣṭha → the later part la deleted → sthū + iṣṭha
→ guṇa of ū → stho + iṣṭha → sthav + iṣṭha → sthaviṣṭhaḥ "stoutest".
yuvan + iṣṭha → van deleted → yu → guṇa → yaviṣṭhaḥ "youngest".
Likewise daviṣṭhaḥ (from dūra), hrasiṣṭhaḥ (from hrasva), kṣepiṣṭhaḥ (from kṣipra), kṣodiṣṭhaḥ (from kṣudra).
Bhaṭṭoji says "likewise before īyas". One of his commentators lists sthavīyān, davīyān, yavīyān, hrasīyān, kṣepīyān, kṣodīyān.
Hrasva, kṣipra and kṣudra are in the list of pṛthu, so they also take iman (SK 1784): hrasimā, kṣepimā, kṣodimā.
Applies when / blocked by. Why "the later" (param) part? Both of Bhaṭṭoji's commentators answer. In yuvan the whole word also begins with a yaṇ, y. In hrasva there is an earlier part that begins with r. Without param, those earlier parts would be deleted.
Why does the rule name guṇa at all? There is a separate guṇa rule for a bha stem in u, or guṇaḥ (SK 847). Both commentators say it cannot do the work here. The deletion counts as not yet made (asiddha) for that rule. One adds that for kṣipra and kṣudra that rule would give no guṇa at all.
Watch out. Our text of the Kaumudī reads evam īyam "likewise īyam". The 1915 printed edition reads evam īyas "likewise īyas", and one of Bhaṭṭoji's commentators quotes the words so. We follow the print.
Rule. Before iṣṭha, iman and īyas, ten words are replaced by ten substitutes, in order:
priya → pra, sthira → stha, sphira → spha, uru → var, bahula → baṃhi,
guru → gar, vṛddha → varṣi, tṛpra → trap, dīrgha → drāghi, vṛndāraka → vṛnda.
Example.
priya + iṣṭha → pra for priya → preṣṭhaḥ "dearest".
Bhaṭṭoji lists the rest: stheṣṭhaḥ, spheṣṭhaḥ, variṣṭhaḥ, baṃhiṣṭhaḥ, gariṣṭhaḥ, varṣiṣṭhaḥ, trapiṣṭhaḥ, drāghiṣṭhaḥ, vṛndiṣṭhaḥ.
Likewise before īyas: preyān. One of his commentators continues: stheyān, spheyān, varīyān, and so on.
Priya, uru, bahula, guru and dīrgha are in the list of pṛthu, so they also take iman: premā. One commentator adds varimā, baṃhimā, garimā, drāghimā.
Applies when / blocked by. "Before iṣṭha and the rest" is carried forward from SK 2008. One of Bhaṭṭoji's commentators says why stheṣṭhaḥ keeps its a. Stha has one vowel, so it stays as it is (SK 2010).
Watch out. Three small points of text. First, our text of the Kaumudī repeats a word, preṣṭhādiṣu, before preṣṭhaḥ. The 1915 printed edition does not have it, and we leave it out. Second, our text reads pṛtvādi. The print reads pṛthvādi "the list of pṛthu", as in SK 2015, and we follow the print. Third, the sūtra in the heading writes trab, the p voiced before d. The substitute is trap either way.
Rule. After bahu "much", the first sound of iman and of īyas is deleted. And bhū replaces bahu.
Example.
bahu + iman → bhū for bahu, the i of the affix deleted → bhū + man → bhūmā "muchness".
ayam anayor atiśayena bahuḥ → bahu + īyas → bhū + yas → bhūyān "more".
Applies when / blocked by. The word iṣṭhemeyassu is carried forward from SK 2008, but iṣṭha is left out here. One of Bhaṭṭoji's commentators says why. The next rule gives iṣṭha an operation of its own. So Bhaṭṭoji writes "of iman and īyas following bahu". The deletion falls on the first sound by ādeḥ parasya (SK 44).
Rule. After bahu, the first sound of iṣṭha is deleted too, and the augment yiṭ is added. Bhū replaces bahu, as in SK 2017.
Example.
bahu + iṣṭha → bhū for bahu (SK 2017) → the i of iṣṭha deleted → bhū + ṣṭha
→ the augment yiṭ (heard as yi) at the start of the affix → bhū + yi + ṣṭha
→ bhūyiṣṭhaḥ "most".
Applies when / blocked by. One of Bhaṭṭoji's commentators says the whole preceding sūtra is carried forward. That is why bhū comes here too. The ṭ of yiṭ is an it. An augment marked with ṭ stands at the start of what it is taught for.
Watch out. The Bhāṣya allows a second account. There the augment is y, taking the place of the deletion; the i of yiṭ is only for pronunciation. One of Bhaṭṭoji's commentators says both views rest on the Bhāṣya. Bhaṭṭoji's own words follow the first. The form is the same.
Rule. Before iṣṭhan and īyasun, kan optionally replaces yuvan "young" and alpa "small".
Example.
yuvan + iṣṭhan → kan for yuvan → kaniṣṭhaḥ "youngest".
alpa + īyasun → kan for alpa → kanīyān "smaller".
In the other option: yaviṣṭhaḥ (by SK 2015) and alpiṣṭhaḥ. One of Bhaṭṭoji's commentators adds yavīyān, alpīyān.
Applies when / blocked by. One of Bhaṭṭoji's commentators says yuvan here is the plain word "young". It is not the technical yuvan, a kind of descendant. The pairing with alpa shows this. He also says the substitute comes only before iṣṭhan and īyasun, never before tarap and tamap.
Watch out. This kan is a substitute for the base, not the affix kan. It takes the place of yuvan or alpa, and iṣṭha or īyas is added to it. Also, our text of the Kaumudī prints the form as kaṃnīyān, with a stray nasal dot. The 1915 printed edition reads kanīyān, and we follow the print.
Rule. Before iṣṭhan and īyasun, there is luk of the possessive affixes vin and matup.
Example.
atiśayena sragvī "garlanded in a high degree" → sragvin + iṣṭhan
→ luk of vin → sraj + iṣṭha → srajiṣṭhaḥ; likewise srajīyān.
atiśayena tvagvān "possessed of skin in a high degree" → tvagvat + iṣṭhan
→ luk of matup → tvac + iṣṭha → tvaciṣṭhaḥ; likewise tvacīyān.
Applies when / blocked by. Why sraj and not srag? In sragvin the j of sraj became the velar g because it stood at the end of a pada, a finished word. Replacing a palatal by a velar is called kutva. One of Bhaṭṭoji's commentators says that with vin gone, that pada status is gone too, and so is the g. The other says the stem is now bha before the vowel-initial affix, so no pada rule applies.
Watch out. Words in vin and matup are not quality-words. Both of Bhaṭṭoji's commentators read this rule as a jñāpaka. By giving their luk before iṣṭhan and īyasun, it shows that they do take the pair.
paṭu → paṭurūpaḥ, paṭukalpaḥ, bahupaṭuḥ, paṭujātīyaḥFour rules give six more svārthikas. Each adds a shade to the base's own meaning. Rūpap marks the meaning as complete and excellent. Kalpap, deśya and deśīyar mark it as slightly short of complete. Bahuc does the same, but it stands before the word. Jātīyar means "of that sort". On the running example, paṭu "clever" gives paṭurūpaḥ "an excellently clever one", paṭukalpaḥ or bahupaṭuḥ "nearly clever", and paṭujātīyaḥ "of the clever sort".
The tiṅaś ca of SK 2002 still runs here. So rūpap and kalpap can follow a tiṅanta as well as a subanta. SK 2023 then stops it.
Rule. Rūpap comes after a subanta or a tiṅanta whose own meaning is excellent (praśaṃsā). Its p is an it, so it is heard as -rūpa.
Example.
praśastaḥ paṭuḥ "an excellent clever one" → paṭu + rūpap → paṭurūpaḥ.
praśastaṃ pacati "he cooks excellently" → pacati + rūpap → pacatirūpam.
Applies when / blocked by. Bhaṭṭoji says "after a subanta and after a tiṅanta". One of his commentators says tiṅaś ca (SK 2002) and "after a nominal stem" are both carried forward. With a tiṅanta the excellence belongs to the action. He cites the Bhāṣya: a verb has action as its chief meaning.
The form after a verb is neuter and singular. The same commentator says the singular is the default and the neuter comes from usage. So there is no dual or plural: pacatorūpam "the two cook excellently", pacantirūpam "they cook excellently".
Watch out. Praśaṃsā here is not "praise". One of Bhaṭṭoji's commentators takes it as the completeness of the base's meaning. So the affix comes even with a bad thing done thoroughly: caurarūpaḥ "a thorough thief", one who steals even the collyrium from the eyes.
Rule. Kalpap, deśya and deśīyar come when the meaning is slightly short of complete (īṣadasamāpti). They are heard as -kalpa, -deśya and -deśīya.
Example.
īṣadūno vidvān "a learned man slightly short [of full learning]" → vidvatkalpaḥ, vidvaddeśyaḥ, vidvaddeśīyaḥ.
yaśas + kalpap → yaśaskalpam "fame not quite complete".
yajus + kalpap → yajuḥkalpam "a yajus not quite complete".
pacati + kalpap → pacatikalpam "he cooks, not quite completely".
Applies when / blocked by. Pacatikalpam shows that tiṅaś ca still runs. One of Bhaṭṭoji's commentators says it is neuter and singular for the same reason as pacatirūpam (SK 2021).
A svārthika keeps the gender of its base. One of Bhaṭṭoji's commentators gives vṛṣabhakalpa iyaṃ gauḥ "this cow is nearly a bull": the derivative stays masculine, like vṛṣabha, though gauḥ is feminine. But usage also has guḍakalpā drākṣā "grapes that are nearly jaggery", feminine, overstepping the gender of its base guḍa. Bhaṭṭoji's commentators account for this in different ways; scholars discuss this.
Watch out. Our text writes yajuḥ kalpam, with a space that splits off the affix. The 1915 printed edition gives yajuḥkalpam, one word with visarga. One of Bhaṭṭoji's commentators reads yajuṣkalpam, with ṣ by SK 153, as SK 152 gives the s of yaśaskalpam. Scholars discuss this.
Rule. In the same sense, slightly short of complete, bahuc comes optionally. It stands before the subanta, never after it. Bahuc is heard as bahu-; its c is an it.
Example.
īṣadūnaḥ paṭuḥ "slightly short of clever" → bahuc placed before the subanta paṭu
→ bahupaṭuḥ.
In the other option: paṭukalpaḥ (SK 2022).
One of Bhaṭṭoji's commentators gives the steps. Bahuc is set before paṭu with its case ending. The whole is now a nominal stem, so that inner ending is deleted by luk. A new ending comes after the whole.
Applies when / blocked by. Why say "after a sup" (supaḥ)? Bhaṭṭoji's counter-example is yajatikalpam. Yajati is a tiṅanta, so bahuc does not come before it. One of his commentators says supaḥ is there to stop the running tiṅaś ca of SK 2002.
Why say vibhāṣā "optionally", when the great option (mahāvibhāṣā, the vā of SK 1072) already runs? The same commentator answers. Without it, only the phrase īṣadūnaḥ paṭuḥ would remain in the other option. With it, kalpap and its companions come there: paṭukalpaḥ.
Watch out. Do not read "after the subanta". In an affix rule a fifth-case word normally names what the affix follows. Here Bhaṭṭoji writes subantāt, a fifth case, but with prāg eva "only before". One of his commentators says subantāt goes with prāk "before". So it means "before the subanta", as one says prāg grāmāt "before the village". The sūtra's tu restricts: only before, not after. The Bhāṣya reads tu in another way; scholars discuss this.
Rule. Jātīyar comes when a mode (prakāra), a sort, is to be expressed. It is heard as -jātīya. Bhaṭṭoji says it comes only in the sense of "that which has the mode" (prakāravat).
Example.
paṭuprakāraḥ "of the clever sort" → paṭu + jātīyar → paṭujātīyaḥ.
Applies when / blocked by. Bhaṭṭoji contrasts thāl of SK 1971, which also has prakāra in its sūtra. Thāl expresses the mode alone. Jātīyar expresses the thing that has the mode. One of his commentators draws the result: jātīyar can follow a thāl-word, as in tathājātīyaḥ, yathājātīyaḥ.
Watch out. The other commentator holds that both affixes mean "that which has the mode", since the wording does not differ. Scholars discuss this.
aśva → aśvakaḥ; uccaiḥ → uccakaiḥ; sarve → sarvakeNow begins the ka section. Three headings open it. Ka is added after a subanta. Akac is placed inside an avyaya, a pronoun or a tiṅanta, before its ṭi. In a k-final avyaya, the k also becomes d. Then the senses begin: unknown, despised, and a name born of contempt. On the running example, a horse whose owner is not known is aśvakaḥ. The avyaya uccaiḥ "loudly" becomes uccakaiḥ, and the pronoun sarva "all" gives sarvake.
Rule. This is an adhikāra. From here up to, but not including, ive pratikṛtau (SK 2051), the affix is ka. The word iva in the sūtra names that later sūtra by one of its words.
Example. The heading has no example of its own. Its first sense comes at SK 2028: aśva + ka → aśvakaḥ.
Applies when / blocked by. One of Bhaṭṭoji's commentators says supaḥ "after a subanta" is carried forward from SK 2023. So ka never follows a tiṅanta.
Watch out. The range of this heading gives the section its name. Prāgivīya means "taught before iva", that is, before SK 2051. The section closes at SK 2050.
Rule. In an avyaya, a pronoun (sarvanāman) or a tiṅanta, akac is placed before the ṭi. In akac the c is an it and the a after the k is only for pronunciation. So -ak- is what goes in.
Example.
uccaiḥ "loudly" → ṭi aiḥ → ucc + ak + aiḥ → uccakaiḥ.
sarva "all" → ṭi a → sarv + ak + a → sarvaka
→ plural ending -e → sarvake.
The akac goes inside the pronoun stem sarva, not the finished word sarve; see SK 2028.
pacati "he cooks" → ṭi i → pacat + ak + i → pacataki.
Applies when / blocked by. Bhaṭṭoji says tiṅaś ca (SK 2002) is carried forward. Both of his commentators add that it comes by the frog's leap (maṇḍūkapluti). It skips SK 2023–2025, where supaḥ had stopped it. So the tiṅanta takes akac, though it cannot take ka.
Watch out. This rule is also a heading, up to iva (SK 2051). One of Bhaṭṭoji's commentators says so. So every sense from SK 2028 on gives ka after a subanta and akac inside an avyaya, a pronoun or a tiṅanta.
Rule. In an avyaya ending in k, d replaces the final k. By ca "and", akac comes as well.
Example.
dhik "fie" → ṭi ik → dh + ak + ik → final k → d → dhakit.
hiruk → hirakut.
Applies when / blocked by. Only avyaya is carried forward from SK 2026, not "pronoun" or "tiṅanta". The two commentators give different grounds and agree on the result. One says the grammar marks just that part of the compound as continuing. The other says a pronoun or a tiṅanta cannot end in k anyway.
Watch out. The d never comes alone. It is taught only together with akac (saṃniyogaśiṣṭa, "taught in the same breath"). One of Bhaṭṭoji's commentators says so, and says this rule too is a heading up to SK 2051.
Rule. The first sense of the headings: the thing is unknown (ajñāta). Ka comes after a subanta; akac goes inside an avyaya, a pronoun or a tiṅanta.
Example.
kasyāyam aśvaḥ "whose is this horse?" → aśva + ka → aśvakaḥ.
One of Bhaṭṭoji's commentators says the question acts out the horse's being unknown.
uccaiḥ → uccakaiḥ; nīcaiḥ "low" → nīcakaiḥ.
sarve → sarvake; viśve → viśvake.
pacati → pacataki; jalpati → jalpataki.
dhik → dhakit; hiruk → hirakut (SK 2027).
Applies when / blocked by. A vārttika fixes where akac goes in a pronoun. Before a case ending beginning with o, s or bh, it goes before the ṭi of the pronoun stem. So with yuṣmad "you" and asmad "I":
yuvakayoḥ, āvakayoḥ; yuṣmakāsu, asmakāsu; yuṣmakābhiḥ, asmakābhiḥ.
Elsewhere, akac goes before the ṭi of the whole word with its ending: tvayakā, mayakā "by you", "by me".
One of Bhaṭṭoji's commentators explains the need. Since supaḥ also runs into SK 2026, akac could fall before the ṭi of the ending, giving yuvayakoḥ. Or it could fall in the stem everywhere, giving tvakayā. The vārttika fixes which.
Both commentators say the restriction concerns yuṣmad and asmad only. Other pronouns take akac in the stem: sarvakeṇa.
Two more vārttikas concern tūṣṇīm "silently". In the akac section, kām comes for tūṣṇīm. It is marked with m, so it goes after the last vowel (SK 37): tūṣṇīkām āste "he sits silent". One of Bhaṭṭoji's commentators takes kām as an augment, not an affix. The second vārttika: for a disposition (śīla), ka comes and the m is deleted. tūṣṇīṃśīlaḥ "silent by nature" → tūṣṇīkaḥ.
Watch out. The words "elsewhere, before the ṭi of the whole word" continue the same vārttika. They are not a second vārttika. The 1915 printed edition does not have them. One of Bhaṭṭoji's commentators gives the same sense as his gloss of the first half.
Rule. The affixes of the headings come when the thing is despised (kutsita).
Example.
kutsito 'śvaḥ "a despised horse" → aśva + ka → aśvakaḥ.
One of Bhaṭṭoji's commentators says the contempt comes from its poor running. Pronouns, avyayas and tiṅantas take akac as before.
Applies when / blocked by. Compare pāśap (SK 1993). Pāśap needs contempt that touches the word's ground of application (pravṛttinimitta): a poor physician as physician. This rule needs no such thing. One of Bhaṭṭoji's commentators draws the contrast.
Watch out. Why was pāśap not taught here, under "despised"? One of Bhaṭṭoji's commentators answers: here it would reach tiṅantas too.
Rule. In the sense "despised", kan comes when the word so formed is a name (saṃjñā). One of Bhaṭṭoji's commentators says this is a name that arises from contempt.
Example.
śūdra + kan → śūdrakaḥ.
rādha + kan → rādhakaḥ.
Applies when / blocked by. Kutsite "despised" is carried forward from SK 2029.
Watch out. Bhaṭṭoji says only that the rule is stated "for the accent" (svara). As at SK 1994, an n-marked affix makes the first vowel udātta (A 6.1.197, SK 3686).
devadatta (anukampitaḥ) → devikaḥ, deviyaḥ, devilaḥ, devadattakaḥTwo senses come next: compassion (anukampā) and policy (nīti). Ka serves both. For a person's name there are more affixes: ṭhac, ghan and ilac, heard as -ika, -iya and -ila. Before these the name is cut down. On the running example, Devadatta regarded with compassion is devikaḥ, deviyaḥ, devilaḥ or devadattakaḥ. A string of vārttikas cuts other names in other places.
Bhaṭṭoji changes Pāṇini's order here. He gives 5.3.83 (SK 2035) before 5.3.80–81, and 5.3.84 (SK 2038) before 5.3.82 (SK 2039).
Rule. Ka comes, in the word's own sense, after a word for something regarded with compassion (anukampā). One of Bhaṭṭoji's commentators glosses anukampā as dayā, pity.
Example.
anukampitaḥ putraḥ "a son regarded with compassion" → putra + ka → putrakaḥ.
Rule. When policy (nīti) is in play, ka comes after a word for what is connected with compassion. Policy is a means such as conciliation (sāman) or a gift (dāna). The connection may be indirect.
Example.
hanta te dhānakāḥ "Ah, [my son], grains will be given to you" → dhānā + ka → dhānakāḥ.
One of Bhaṭṭoji's commentators says hanta shows compassion, and "[my son]" is understood. The son is pitied directly. Through him the grains too are connected with compassion.
Likewise guḍakāḥ, from guḍa "jaggery".
With akac in a tiṅanta: ehi "come" → ehaki; addhi "eat" → addhaki.
Applies when / blocked by. Bhaṭṭoji states the difference from SK 2031. By SK 2031 the affix follows a word for what is itself pitied. By this rule it follows a word for what is connected with the pitied one, even indirectly. One of his commentators adds that SK 2026 and SK 2027 are carried forward here, so akac comes too.
Watch out. Is dhānakāḥ feminine, like dhānā, or masculine? Bhaṭṭoji's commentators disagree; scholars discuss this.
Rule. After a person's name (manuṣyanāman) with many vowels (bahvac), ṭhac comes optionally. Bhaṭṭoji says it applies in the domain of the two preceding rules: compassion and policy. In the other option, ka comes. Ṭhac is heard as -ika, since ṭha → ika (SK 1170).
Example.
anukampito devadattaḥ "Devadatta, regarded with compassion" → devadatta + ṭhac
→ the name cut down (SK 2035) → devikaḥ.
In the other option: devadatta + ka → devadattakaḥ.
Applies when / blocked by. The sūtra writes ṭhaj: the c of ṭhac is voiced before vā. One of Bhaṭṭoji's commentators says the rule gives ṭhac, and ka too by its vā.
Rule. In the same domain, after the same kind of name, ghan and ilac come too. Ghan is heard as -iya, since gha is replaced by iy (SK 475). Ilac is heard as -ila.
Example.
devadatta + ghan → cut down (SK 2035) → deviyaḥ.
devadatta + ilac → devilaḥ.
Applies when / blocked by. Bhaṭṭoji's gloss is tatraiva, "in that same domain". One of his commentators spells it out: the domain of SK 2033, that is, of compassion and policy.
Watch out. What does ca "too" add here? Ṭhac and ka are already given by SK 2033. One of Bhaṭṭoji's commentators, following an earlier scholar, finds the point of ca doubtful. Scholars discuss this.
Rule. In this section, before ṭha or a vowel-initial affix, everything after the base's second vowel is deleted. "This section" means the rules of compassion and policy.
Example.
devadatta + ṭhac → the second vowel is the a of va → everything after it, datta, deleted
→ deva + ṭha → ṭha → ika → devikaḥ.
devadatta + ghan → deviyaḥ; devadatta + ilac → devilaḥ.
Before ka nothing is deleted: devadattakaḥ; vāyudatta + ka → vāyudattakaḥ.
Applies when / blocked by. Ūrdhvam "above, after" means all of it. Both of Bhaṭṭoji's commentators say so. Without it, ādeḥ parasya (SK 44) would delete only the first sound after the second vowel. The word "deletion" is not in this sūtra. Both commentators say it comes from ajināntasyottarapadalopaś ca (SK 2039). That is Pāṇini's 5.3.82, before this rule in his order.
Why name ṭha at all, since ika is vowel-initial anyway? Bhaṭṭoji answers: for a name whose second vowel is an uk vowel (u, ṛ or ḷ). One of his commentators gives the steps. Because ṭha is named, datta goes while the affix is still ṭha. Then ṭha after the uk vowel becomes ka (SK 1221).
vāyudatta + ṭhac → datta deleted → vāyu + ṭha → ka → vāyukaḥ.
pitṛdatta + ṭhac → pitṛkaḥ.
A ślokavārttika, a vārttika in verse, adds more cuts. Bhaṭṭoji gives it piece by piece, then whole.
bṛhaspatidatta → bṛhaspatikaḥ.devadattakaḥ or devakaḥ.dattikaḥ, dattiyaḥ, dattilaḥ, dattakaḥ.devadatta → dattaḥ, devaḥ; satyabhāmā → bhāmā, satyā.bhānudatta → bhānulaḥ. La is the earlier teachers' name for deletion. One of Bhaṭṭoji's commentators says only the i of ila goes.savitriyaḥ, savitṛlaḥ. One of Bhaṭṭoji's commentators rejects this piece, since the Bhāṣya does not have it.Watch out. Three slips in our text of the Kaumudī. It reads pratye in the gloss; the 1915 printed edition reads pratyaye "affix". It reads urdhvasya in the first vārttika; the print reads ūrdhvasya. And it lacks the base word before bhānulaḥ; the print gives bhānudattaḥ. We follow the print in all three.
Rule. After a many-voweled name beginning with upa, aḍac and vuc come too. This is in the domain of SK 2033. By ca, the affixes already available come as well. Aḍac is heard as -aḍa; its c is an it. Vuc is heard as -aka (SK 1247).
Example. From upendradatta, regarded with compassion, Bhaṭṭoji gives six forms:
upaḍaḥ (with aḍac), upakaḥ (with vuc),
upikaḥ (with ṭhac), upiyaḥ (with ghan), upilaḥ (with ilac),
upendradattakaḥ (with ka).
Applies when / blocked by. One of Bhaṭṭoji's commentators says ca brings in ṭhac, ghan and ilac. Bhaṭṭoji says the word prācām "in the view of the eastern grammarians" is for honour (pūjārtha). One of his commentators explains: the option already runs from the vā of SK 2033, so prācām adds none.
Watch out. Bhaṭṭoji gives only the forms. How the cut reaches back to upa in upendradatta is a point scholars discuss. The last form, with ka, is not cut, since ka is neither ṭha nor vowel-initial.
Rule. After a class-word (jāti) used as a man's name, kan comes, in compassion or policy. Bhaṭṭoji says "after a man's name" still holds.
Example.
siṃha, an animal-class word used as a man's name → siṃha + kan → siṃhakaḥ.
Likewise śarabhakaḥ (from śarabha) and rāsabhakaḥ (from rāsabha).
Applies when / blocked by. One of Bhaṭṭoji's commentators calls this an apavāda of the affixes of SK 2031–2032. Both commentators say these words are well known for another class and are used among men as names.
Three vārttikas follow, on cutting names.
kahoḍa, a sage's name → kah + ika → kahikaḥ. One of Bhaṭṭoji's commentators says the diphthong goes too, not only what follows it.vāgāśīrdatta → vāc + ika → vācikaḥ.ṣaḍaṅgulidatta → ṣaḍikaḥ? The vārttika: for ṣaṣ, the sūtra's own cut after the second vowel (SK 2035) applies.Watch out. The last question is about the ḍ. One of Bhaṭṭoji's commentators explains it. Before ika the stem is bha, so ṣ would not be voiced to ḍ. With the cut of SK 2035 the steps run differently, and the ḍ stands.
Rule. For men's names beginning with śevala, supari, viśāla, varuṇa, aryaman and the like, the cut comes after the third vowel. It applies before ṭha or a vowel-initial affix. Bhaṭṭoji calls it an apavāda of SK 2035.
Example.
anukampitaḥ śevaladattaḥ "Śevaladatta, regarded with compassion" → śevaladatta + ṭhac
→ the third vowel is the a of la → datta deleted → śevalikaḥ.
With ghan and ilac: śevaliyaḥ, śevalilaḥ.
Likewise suparikaḥ (from suparidatta), viśālikaḥ, varuṇikaḥ, aryamikaḥ.
Applies when / blocked by. One of Bhaṭṭoji's commentators says "before ṭha or a vowel-initial affix" and "after" come from SK 2035. "Deletion" comes from SK 2039, through SK 2035. Bhaṭṭoji's commentators also cite a vārttika from the Bhāṣya: the vowels are counted before sandhi, the joining of sounds at the seam, is made. So suparyāśīrdatta → suparikaḥ.
Watch out. Our text of the Kaumudī reads sevalilaḥ, with s. The 1915 printed edition reads śevalilaḥ, and so do Bhaṭṭoji's commentators. We follow the print.
Rule. After a man's name ending in ajina "hide", kan comes in compassion. And the final member (uttarapada), ajina, is deleted.
Example.
anukampito vyāghrājinaḥ "Vyāghrājina, regarded with compassion" → vyāghrājina + kan
→ ajina deleted → vyāghrakaḥ.
siṃhājina + kan → siṃhakaḥ.
One of Bhaṭṭoji's commentators takes this siṃhakaḥ as the form of siṃhājina, not of siṃha as in SK 2037.
Applies when / blocked by. Bhaṭṭoji supplies "a man's name", "in compassion" and "kan" from the rules before.
Watch out. This is Pāṇini's 5.3.82, given after 5.3.83–84. Its word lopa "deletion" is the one that SK 2035 and SK 2038 borrow.
vṛkṣa → vṛkṣakaḥ; vatsa → vatsataraḥThe next senses are "small" and "short". The affixes are those already prescribed: ka after a subanta, akac inside an avyaya, a pronoun or a tiṅanta. A few words take special affixes: ra, ḍupac and ṣṭarac. The group ends with ṣṭarac for a thing that has passed beyond its kind. On the running example, a short tree is vṛkṣakaḥ. A calf that has reached its second age is vatsataraḥ.
Rule. The affixes as prescribed (yathāvihitam) come when the sense is "small" (alpa).
Example.
alpaṃ tailam "a small quantity of oil" → taila + ka → tailakam.
One of Bhaṭṭoji's commentators adds sarvakam, uccakaiḥ, pacataki. So akac comes here too.
Rule. The same affixes come when the sense is "short" (hrasva).
Example.
hrasvo vṛkṣaḥ "a short tree" → vṛkṣa + ka → vṛkṣakaḥ.
Applies when / blocked by. How does "short" differ from "small"? One of Bhaṭṭoji's commentators answers. Small is the opposite of large; short is the opposite of long. The other says we do not call oil short or long, which is why the example is a tree.
Watch out. vṛkṣakaḥ could also come from SK 2040. One of Bhaṭṭoji's commentators reports that the verbal cognition (śābdabodha) differs. That is the meaning a hearer takes from the word.
Rule. Kan comes when a name arises from shortness. Bhaṭṭoji calls it an apavāda of the preceding rule, so it sets aside ka and the rest.
Example.
vaṃśa "bamboo" → vaṃśa + kan → vaṃśakaḥ, a particular short kind of bamboo.
Likewise veṇukaḥ.
One of Bhaṭṭoji's commentators glosses vaṃśaka.
Watch out. The sūtra repeats SK 2030 word for word. There the name arose from contempt. Here Bhaṭṭoji says the name arises from shortness.
Rule. After kuṭī "hut", śamī and śuṇḍā, the affix ra comes when the sense is "short". Hrasve still holds; one of Bhaṭṭoji's commentators says so.
Example.
hrasvā kuṭī "a short hut" → kuṭī + ra → kuṭīraḥ.
Likewise śamīraḥ, śuṇḍāraḥ.
Applies when / blocked by. Kuṭīraḥ is masculine, though kuṭī is feminine. Both of Bhaṭṭoji's commentators cite a statement: a svārthika sometimes oversteps the gender of its base.
Watch out. Our text of the Kaumudī reads hrasvaḥ kuṭī, with a masculine "short". The 1915 printed edition reads hrasvā kuṭī, feminine, agreeing with kuṭī. We follow the print.
Rule. After kutū, a leather flask for oil, the affix ḍupac comes when the sense is "short". Its ḍ and c are markers (it), so it is heard as -upa.
Example.
hrasvā kutūḥ "a short oil-flask" → kutū + ḍupac
→ the ṭi ū deleted before the ḍ-marked affix (SK 316) → kut + upa
→ kutupaḥ.
One of Bhaṭṭoji's commentators gives these steps.
Applies when / blocked by. Bhaṭṭoji quotes a verse, which one of his commentators identifies as Amara's. Kutū is a vessel of hide for oil; a short one is kutupa, masculine. So here too the feminine is given up.
Rule. After kāsū and goṇī, the affix ṣṭarac comes when the sense is "short". Its ṣ and c are markers (it), so it is heard as -tara.
Example.
hrasvā kāsūḥ "a short kāsū" → kāsū + ṣṭarac → kāsūtara
→ the feminine ṅīṣ (heard as -ī) → kāsūtarī.
Likewise goṇītarī.
Bhaṭṭoji glosses kāsū as a particular weapon.
Applies when / blocked by. One of Bhaṭṭoji's commentators says ṅīṣ comes because the affix is marked with ṣ.
Watch out. Our text of the Kaumudī reads āyudhaviśaiṣaḥ. The 1915 printed edition reads āyudhaviśeṣaḥ "a particular weapon". We follow the print.
Rule. After vatsa "calf", ukṣan "young bull", aśva "horse" and ṛṣabha "bull", ṣṭarac comes for diminution (tanutva). Bhaṭṭoji says the diminution must be of the word's ground of application (pravṛttinimitta).
Example.
vatsa + ṣṭarac → vatsataraḥ: a calf that has reached the second age.
ukṣan + ṣṭarac → ukṣataraḥ: a bull past its youth.
aśva + ṣṭarac → aśvataraḥ: a mule, born of a donkey on a mare.
ṛṣabha + ṣṭarac → ṛṣabhataraḥ: a bull weak at bearing loads.
Bhaṭṭoji's commentators give these senses.
Applies when / blocked by. A vatsa is a calf in its first age. Its "diminution" is having passed into the next age. So a thin calf is not a vatsatara. One of Bhaṭṭoji's commentators says this is why Bhaṭṭoji adds "only when the ground of application is diminished".
Watch out. Here "short" (hrasve) stops running. One of Bhaṭṭoji's commentators says so. Ṣṭarac alone is carried on from SK 2045.
kim → kataraḥ (of two), katamaḥ (of many)The last group asks "which one?". When one is singled out (nirdhāraṇa) from two, kim, yad and tad take ḍatarac. From many, they take ḍatamac. Eka takes both. On the running example, "which of the two?" is kataraḥ, and "which of the many?" is katamaḥ. The section ends with one more kan, for the means by which a person is disparaged.
Rule. When one of two is singled out (nirdhāraṇa), ḍatarac comes after kim "which", yad "which (relative)" and tad "that". Ḍatarac is heard as -atara; its ḍ deletes the ṭi.
Example.
anayoḥ kataro vaiṣṇavaḥ "which of these two is a Vaiṣṇava?" → kim + ḍatarac
→ the ṭi im deleted before the ḍ-marked affix (SK 316) → k + atara
→ kataraḥ.
Likewise yataraḥ, tataraḥ.
By the great option (mahāvibhāṣā) the plain words also stand: kaḥ, yaḥ, saḥ.
Applies when / blocked by. Both of Bhaṭṭoji's commentators add the condition "denoting the one singled out". In kayor anyataro devadattaḥ "Devadatta is the other of which two?" kim denotes the two, not the one singled out. So no ḍatarac comes after it. One commentator adds that the singling out may be by a quality, an action or a name.
Watch out. Our text of the Kaumudī reads mahāvibhāṣāyā. The 1915 printed edition reads mahāvibhāṣayā "by the great option", as our text itself does at SK 2048. We follow the print. The Devanāgarī heading keeps our source's spelling, nirddhāraṇe with a doubled dh; the IAST follows the print, nirdhāraṇe. The word is the same.
Rule. When one of many is singled out, ḍatamac comes optionally after kim, yad and tad. The sūtra adds "for a class (jāti) or a question (paripraśna)". Ḍatamac is heard as -atama.
Example.
katamo bhavatāṃ kaṭhaḥ "which of you is a Kaṭha?" → kim + ḍatamac → katamaḥ.
Likewise yatamaḥ, tatamaḥ.
Applies when / blocked by. Both of Bhaṭṭoji's commentators say jāti goes with all three words, paripraśna with kim alone. The question excludes kim in the sense of contempt.
Bhaṭṭoji then says the words jātiparipraśne are rejected in the source-work (ākara). One of his commentators names it: a vārttika with the gloss of Kaiyaṭa, a commentator on the Bhāṣya. So ḍatamac also comes when one is singled out by an action, a quality or a name: katama eṣāṃ pācakaḥ "which of these is the cook?".
Why does the sūtra say vā "optionally"? Bhaṭṭoji: for akac. So the pronouns also give yakaḥ, sakaḥ. One of his commentators explains. Under the great option alone, akac would not come back once ḍatamac had set it aside. And by the great option the plain yaḥ, saḥ stand too.
A vārttika: in this domain kim also takes ḍatarac: kataraḥ, of one among many.
Watch out. There is no kakaḥ beside yakaḥ and sakaḥ. One of Bhaṭṭoji's commentators says why: kim with akac is still replaced by ka.
Rule. After eka "one", ḍatarac and ḍatamac come, in the domain of the two preceding rules.
Example.
anayor ekataro maitraḥ "of these two, the one is Maitra" → eka + ḍatarac → ekataraḥ.
eṣām ekatamaḥ "of these [many], the one" → eka + ḍatamac → ekatamaḥ.
Applies when / blocked by. Bhaṭṭoji supplies "ḍatarac and ḍatamac come" (śeṣapūraṇa). One of his commentators says prācām is for honour; the option runs on anyway.
Watch out. This prācām does not bring akac, unlike the vā of SK 2048. One of Bhaṭṭoji's commentators says so, citing the Bhāṣya: the option already comes from the great option, so prācām need not have been said.
Rule. Kan comes after a word for the means by which another person is disparaged (avakṣepaṇa).
Example.
vyākaraṇakena garvitaḥ "puffed up by his grammar" → vyākaraṇa + kan → vyākaraṇakaḥ.
Bhaṭṭoji: "That by which another is disparaged is the example here."
Applies when / blocked by. Grammar is not contemptible in itself. One of Bhaṭṭoji's commentators explains the example. Once studied, it breeds pride in the student, and so brings contempt on him. It is the means of his disparagement. Both commentators read avakṣepaṇa as "that by which one is disparaged". It is formed with the affix lyuṭ, an affix forming nouns from verb roots, heard as -ana. Here lyuṭ is in the sense of the instrument (karaṇa).
Watch out. Do not read vyākaraṇakaḥ as "paltry grammar". Where the thing is contemptible in itself, Bhaṭṭoji says, the affix is the ka of kutsite (SK 2029). One of his commentators adds that the two differ in accent.
With this sūtra Bhaṭṭoji closes the section: prāgivīyānāṃ pūrṇo 'vadhiḥ, "the range of the affixes taught before iva is complete". Our text of the Kaumudī lacks the line; the 1915 printed edition has it, and both commentators close with the same words. The next chapter, with more svārthika affixes, opens with ive pratikṛtau (SK 2051). That is the sūtra that has marked the limit of the ka heading since SK 2025.
laghiṣṭhaḥ (SK 2001); the ū of sthūla takes guṇa in sthaviṣṭhaḥ (SK 2015).hrasimā, premā, bhūmā (SK 2015–2017).kārayati "he calls him a doer" (SK 2007).| SK | What it does | Example |
|---|---|---|
| 1974 | astāti after a direction-word in the 7th, 5th or 1st case, for direction, place, time | purastāt |
| 1975 | asi after pūrva, adhara, avara, with pur, adh, av for them | puraḥ, adhaḥ, avaḥ |
| 1976 | pur, adh, av before astāti too | purastāt, adhastāt, avastāt |
| 1977 | av optional for avara before astāti; counter-examples to SK 1974 | avastāt / avarastāt |
| 1978 | atasuc after dakṣiṇa, uttara | dakṣiṇataḥ, uttarataḥ |
| 1979 | atasuc optionally after para, avara | parataḥ / parastāt |
| 1980 | luk of astāti after an añc-final direction-word | prāk, udak |
| 1981 | upa for ūrdhva, with ril or riṣṭātil | upari, upariṣṭāt |
| 1982 | paśca for apara, with āti | paścāt |
| 1983 | āti after uttara, adhara, dakṣiṇa | uttarāt, dakṣiṇāt |
| 1984 | enap optionally for "not far", not in the 5th case | uttareṇa, dakṣiṇena |
| 1985 | āc after dakṣiṇa, not in the 5th case | dakṣiṇā vasati |
| 1986 | āhi and āc after dakṣiṇa, for "far" | dakṣiṇāhi, dakṣiṇā |
| 1987 | āhi and āc after uttara | uttarāhi, uttarā |
| 1988 | dhā after a number-word, for a mode of action | caturdhā, dvidhā |
| 1989 | dhā when a substance is brought to another number | ekaṃ rāśiṃ pañcadhā kuru |
| 1990 | dhyamuñ optionally for dhā after eka | aikadhyam / ekadhā |
| 1991 | dhamuñ optionally for dhā after dvi, tri; vārttika ḍa | dvaidham / dvidhā; dvaidhāni |
| 1992 | edhāc for dhā after dvi, tri | dvedhā, tredhā |
| 1993 | pāśap for something contemptible | bhiṣakpāśaḥ |
| 1994 | an after a tīya ordinal meaning a part; vārttikas īkak, not for vidyā | dvitīyaḥ; dvaitīyīkaḥ |
| 1995 | an after ordinals below eleven, for a part, outside the Veda | caturthaḥ, pañcamaḥ |
| 1996 | ña and an after ṣaṣṭha, aṣṭama | ṣāṣṭhaḥ / ṣaṣṭhaḥ |
| 1997 | kan for a measure, luk for an animal's part, after ṣaṣṭha, aṣṭama | ṣaṣṭhakaḥ; aṣṭamaḥ |
| 1998 | ākinic, kan, luk after eka "alone" | ekākī, ekakaḥ, ekaḥ |
| 1999 | caraṭ, "formerly" | āḍhyacaraḥ |
| 2000 | rūpya and caraṭ after a 6th-case word, "formerly"; no puṃvadbhāva | kṛṣṇarūpyaḥ, kṛṣṇacaraḥ; śubhrārūpyaḥ |
| 2001 | tamap, iṣṭhan for excellence, one of many | āḍhyatamaḥ; laghutamaḥ, laghiṣṭhaḥ |
| 2002 | tamap alone after a tiṅanta | pacatitamām |
| 2003 | tarap, tamap are called gha | pacatitama |
| 2004 | āmu after a gha following kim, an e-final word, a tiṅanta, an indeclinable; not for a substance | kiṃtamām, uccaistamām; uccaistamas taruḥ |
| 2005 | tarap, īyasun for one of two, or with a group divided off | laghutaraḥ, laghīyān; paṭutarāḥ |
| 2006 | iṣṭhan, īyasun only after a quality-word | pācakataraḥ, pācakatamaḥ |
| 2007 | iṣṭhan, īyasun after a tṛ-final word, in the Veda; a pratiprasava of SK 2006 | kariṣṭhaḥ |
| 2008 | the whole tṛ deleted before iṣṭha, iman, īyas | kariṣṭhaḥ, dohīyasī |
| 2009 | śra for praśasya before iṣṭhan, īyasun | śra + iṣṭha |
| 2010 | a one-vowel base stays as it is before iṣṭha etc. | śreṣṭhaḥ, śreyān |
| 2011 | jya also for praśasya | jyeṣṭhaḥ |
| 2012 | after jya, ā for the first sound of īyas | jyāyān |
| 2013 | jya for vṛddha "old" | jyeṣṭhaḥ, jyāyān |
| 2014 | neda for antika, sādha for bāḍha | nediṣṭhaḥ, sādhīyān |
| 2015 | later yaṇ-part of sthūla etc. deleted, guṇa before it | sthaviṣṭhaḥ, yaviṣṭhaḥ, hrasimā |
| 2016 | ten substitutes: pra for priya, etc. | preṣṭhaḥ, gariṣṭhaḥ, premā |
| 2017 | after bahu, first sound of iman, īyas deleted; bhū for bahu | bhūmā, bhūyān |
| 2018 | after bahu, i of iṣṭha deleted, augment yiṭ | bhūyiṣṭhaḥ |
| 2019 | kan optionally for yuvan, alpa | kaniṣṭhaḥ, kanīyān / yaviṣṭhaḥ, alpiṣṭhaḥ |
| 2020 | luk of vin, matup before iṣṭhan, īyasun | srajiṣṭhaḥ, tvacīyān |
| 2021 | rūpap for excellence, after a subanta or tiṅanta | paṭurūpaḥ, pacatirūpam |
| 2022 | kalpap, deśya, deśīyar: slightly short of complete | vidvatkalpaḥ, vidvaddeśīyaḥ, pacatikalpam |
| 2023 | bahuc optionally, placed before the subanta | bahupaṭuḥ; not with yajatikalpam |
| 2024 | jātīyar, "of that sort", in what has the mode | paṭujātīyaḥ |
| 2025 | heading: ka after a subanta, up to SK 2051 | aśvakaḥ |
| 2026 | heading: akac before the ṭi of an avyaya, pronoun, tiṅanta | uccakaiḥ, sarvake, pacataki |
| 2027 | heading: k → d in a k-final avyaya, with akac | dhakit, hirakut |
| 2028 | sense "unknown"; vārttikas on yuṣmad/asmad and tūṣṇīm | aśvakaḥ, yuvakayoḥ, tvayakā, tūṣṇīkām |
| 2029 | sense "despised" | aśvakaḥ |
| 2030 | kan for a name born of contempt; for the accent | śūdrakaḥ |
| 2031 | ka for what is pitied | putrakaḥ |
| 2032 | ka in policy, after what is connected with the pitied one | hanta te dhānakāḥ, ehaki |
| 2033 | ṭhac optionally after a many-voweled personal name; else ka | devikaḥ / devadattakaḥ |
| 2034 | ghan, ilac too | deviyaḥ, devilaḥ |
| 2035 | before ṭha or a vowel-initial affix, all after the 2nd vowel deleted; vārttikas | devikaḥ, vāyukaḥ, bṛhaspatikaḥ, bhānulaḥ |
| 2036 | aḍac, vuc too after an upa-initial name | upaḍaḥ, upakaḥ … six forms |
| 2037 | kan after a class-word used as a man's name; vārttikas | siṃhakaḥ, kahikaḥ, vācikaḥ, ṣaḍikaḥ |
| 2038 | for śevala etc., the cut after the 3rd vowel | śevalikaḥ, suparikaḥ |
| 2039 | kan after an ajina-final name, ajina deleted | vyāghrakaḥ |
| 2040 | the affixes as prescribed, for "small" | tailakam |
| 2041 | the same, for "short" | vṛkṣakaḥ |
| 2042 | kan for a name born of shortness | vaṃśakaḥ, veṇukaḥ |
| 2043 | ra after kuṭī, śamī, śuṇḍā for "short" | kuṭīraḥ |
| 2044 | ḍupac after kutū for "short" | kutupaḥ |
| 2045 | ṣṭarac after kāsū, goṇī for "short" | kāsūtarī, goṇītarī |
| 2046 | ṣṭarac after vatsa, ukṣan, aśva, ṛṣabha, for diminution | vatsataraḥ, aśvataraḥ |
| 2047 | ḍatarac after kim, yad, tad, one of two singled out | kataraḥ; kaḥ |
| 2048 | ḍatamac optionally, one of many; vā for akac; vārttika ḍatarac | katamaḥ, yakaḥ, kataraḥ |
| 2049 | ḍatarac, ḍatamac after eka | ekataraḥ, ekatamaḥ |
| 2050 | kan for the means of another's disparagement | vyākaraṇakaḥ |