After this chapter you will be able to derive the taddhita words of SK 2051 to SK 2138. There are 88 sūtras, the short rules of Pāṇini's grammar, from A 5.3.96 to A 5.4.67. Most affixes here are svārthika: each comes in the base's own sense and adds no new thing meant. The chapter opens with likenesses. An image like a horse is aśvakaḥ, made with the affix kan, heard as -ka. For some images that kan is deleted by lup, a named kind of deletion. So an image of Vāsudeva carried about for a living is simply vāsudevaḥ. Other affixes then say "like X": vāsteyam, śileyam, kuśāgrīyā buddhiḥ "an intellect fine as the tip of kuśa grass", and kākatālīyaḥ "come about unforeseen". Next come associations and armed bands: lauhitadhvajyaḥ, kṣaudrakyaḥ, yaudheyaḥ. Their affixes are called tadrāja, so they vanish in the plural: kṣudrakāḥ. With the fourth quarter of the fifth book come dvipadikām "two pādas at a time", sthūlakaḥ, chinnakam "cut in part", and bṛhatikā "an upper wrap". Then come affixes after particular stems: prācīnam, aṣaḍakṣīṇo mantraḥ "a consultation held by two only", brāhmaṇajātīyaḥ, pitṛsthānīyaḥ "like a father". Here Bhaṭṭoji also lists the svārthika affixes that are obligatory. For "so many times" come pañcakṛtvaḥ "five times", dviḥ "twice", sakṛt "once" and bahudhā "many times". Abundance and "for that" follow: annamayam, ātithyam. Then comes devatā, just "the deva itself". Then "X itself" and colours: avikaḥ, lohitakaḥ, and the feminine lohitikā or lohinikā. Then a message, a deed, a medicine and the words for earth: vācikam, kārmaṇam, auṣadham, mṛttikā, mṛtsnā. Two more affixes give bahuśaḥ, dviśaḥ "two at a time", and grāmataḥ "from the village". Next comes becoming what one was not: kṛṣṇīkaroti "he makes black what was not black", agnisād bhavati "it turns wholly to fire", and devatrā. Last comes an affix -ā used with verbs, from SK 2129 with kṛ "make" alone: paṭapaṭākaroti, from an imitated sound, and dvitīyākaroti "he does a second ploughing". Every rule is worked on a concrete word.
The name svārthika comes from svārthe, "in the word's own sense". One of Bhaṭṭoji's commentators opens the section at SK 2051: "now the svārthika section within the taddhitas is set out". The 1915 printed edition heads it instead "the tadrājas among the taddhitas". Bhaṭṭoji follows Pāṇini's order. He leaves out seven of Pāṇini's rules in this stretch. Three are Vedic and are taught in the Vedic section. The other four stand at SK 1100 and SK 2004, taught earlier, and at SK 3216 and SK 3219, in the section on kṛt affixes, the affixes added directly to a root. He also brings in four rules from other books (SK 2118–2120 and SK 2123). Each builds a form of the "becoming" affixes. With this chapter the taddhita section of the Kaumudī ends.
aśva iva pratikṛtiḥ → aśvakaḥ; vāsudeva-pratikṛti → vāsudevaḥThe chapter opens with the word iva "like". The first rule gives kan when the thing compared is a likeness (pratikṛti), such as a clay image. A likeness like a horse is aśvakaḥ. The next rule gives kan for a name. Three rules then delete these kan affixes by lup. An image of the god Vāsudeva, carried from house to house for a living, is just vāsudevaḥ.
Rule. In the sense of iva "like", kan comes when the thing compared is a likeness (pratikṛti). It comes after the word for the standard of comparison. One of Bhaṭṭoji's commentators says kan is carried forward from SK 2050.
Example.
aśva iva pratikṛtiḥ "a likeness like a horse" → aśva + kan
→ the n of kan is an it → aśva + ka
→ aśvakaḥ.
Applies when / blocked by. A pratikṛti is an image made of clay and the like. One of Bhaṭṭoji's commentators defines it so. The thing compared must be such an image. Bhaṭṭoji's counter-example is gaur iva gavayaḥ "the gavaya is like a cow". A gavaya is not a likeness, so no kan comes.
Watch out. Pratikṛti is feminine, yet aśvakaḥ is masculine. One of Bhaṭṭoji's commentators explains: a svārthika affix keeps the gender and number of its base, here aśva. The other commentator reports two views of what kan does here. On one, it names the likeness itself. On the other, aśva already stands for the likeness, and kan only signals that use.
Rule. Kan also comes in the sense of iva when base and affix together are a name (saṃjñā). Bhaṭṭoji says the rule is made for things that are not likenesses.
Example.
aśvasadṛśa "something resembling a horse" → aśva + kan → aśvakaḥ, as its name.
Likewise uṣṭrakaḥ, the name of something resembling a camel (uṣṭra).
Applies when / blocked by. One of Bhaṭṭoji's commentators traces the words. Kan is supplied. Ive "like" continues from SK 2051, but pratikṛtau "a likeness" has ceased. Otherwise this rule would add nothing to SK 2051.
Rule. The kan given for a name by SK 2052 undergoes lup when a man (manuṣya) is meant.
Example.
A cañcā is a man made of grass.
cañcā iva manuṣyaḥ "a man like a cañcā" → cañcā + kan (SK 2052)
→ lup of kan
→ cañcā, the man's name.
Likewise vardhrikā. One of Bhaṭṭoji's commentators says a vardhrikā is a likeness made of leather.
Applies when / blocked by. Only the kan of SK 2052 is meant, not that of SK 2051. One of Bhaṭṭoji's commentators gives the reason. A man cannot be a likeness.
Watch out. The man's name cañcā stays feminine. Both of Bhaṭṭoji's commentators say so: under lup the gender follows the base (SK 1294). One of them adds that the number follows the man. So for two men like cañcās, the singular cañcā is not used.
Rule. Kan also undergoes lup when the likeness serves a livelihood (jīvikārtha) and is not for sale (apaṇya).
Example.
vāsudeva iva pratikṛtiḥ "a likeness like Vāsudeva" → vāsudeva + kan (SK 2051)
→ lup of kan
→ vāsudevaḥ.
Likewise śivaḥ, skandaḥ.
Applies when / blocked by. Bhaṭṭoji says this applies to images of gods that are the livelihood of devalakas. Both of his commentators explain: devalakas carry such images from house to house, begging alms.
One of Bhaṭṭoji's commentators says this is the kan of SK 2051, not of SK 2052. The Bhāṣya gives the example only for a likeness.
Why "not for sale"? Bhaṭṭoji's counter-example is hastikān vikrīṇīte "he sells hastikas", likenesses of elephants. Goods for sale keep kan. Both of his commentators cite a verse on a man who sells rāma, sītā and lakṣmaṇa for his living. For such goods, they say, the correct forms keep kan: rāmakam, sītikām.
Watch out. Our text of the Kaumudī reads tasmivanvācye. The 1915 printed edition reads tasminvācye, "when that is to be expressed". We follow the print.
Rule. After devapatha and the words that follow it, the kan of SK 2051 also undergoes lup. The list is an ākṛtigaṇa.
Example.
devapatha iva pratikṛtiḥ "a likeness like the path of the gods" → devapatha + kan
→ lup of kan
→ devapathaḥ.
Likewise haṃsapathaḥ.
Applies when / blocked by. Bhaṭṭoji writes "there is lup of kan". One of his commentators calls these words a śeṣapūraṇa. The same commentator says the kan is that of SK 2051.
Watch out. Both of Bhaṭṭoji's commentators cite a verse. It extends this lup to images for worship (arcā), to paintings, and to banners. One of them gives examples: Śiva and Viṣṇu among images for worship, and the monkey (kapi) and Garuḍa on banners.
śilā iva → śileyam (ḍha) / śaileyam (ḍhañ)The sense "like" goes on, but "a likeness" does not. Each rule now names its own affix after its own words. On the running example, what is like a rock (śilā) is śileyam, with ḍha. Some also allow ḍhañ, which brings vṛddhi: śaileyam. The sense "like" runs as a heading up to SK 2066, which it does not reach.
Rule. After vasti, ḍhañ comes in the sense "like". Vasti is the region below the navel. Both of Bhaṭṭoji's commentators gloss it from Amara's lexicon.
Example.
vastir iva "like the vasti" → vasti + ḍhañ
→ the ñ is an it → vasti + ḍha
→ ḍh → ey (SK 475) → vasti + eya
→ vṛddhi of the first vowel (SK 1075) → vāsti + eya
→ the final i deleted (SK 311) → vāsteyam.
In the feminine, ṅīp (SK 470) → vāsteyī.
Applies when / blocked by. Bhaṭṭoji says iva is carried forward, and pratikṛtau has ceased (nivṛtta). One of his commentators gives the reason. Pratikṛtau was not marked svarita.
Rule. After śilā "rock", ḍha comes in the sense "like". Unlike ḍhañ, it brings no vṛddhi.
Example.
śileva "like a rock" → śilā + ḍha
→ ḍh → ey (SK 475) → śilā + eya
→ the final ā deleted (SK 311) → śileyam.
One of Bhaṭṭoji's commentators supplies what is meant: curds and the like.
Applies when / blocked by. Bhaṭṭoji reports that some split the rule (yogavibhāga) at śilāyāḥ. So ḍhañ comes too: śaileyam, with vṛddhi. One of his commentators says ḍhañ is carried forward from SK 2056 into that first part. One of his commentators gives the point of ḍhañ: vṛddhi, and ṅīp in the feminine.
Rule. After śākhā "branch" and the words that follow it, ya comes in the sense "like".
Example.
śākheva "like a branch" → śākhā + ya
→ the final ā deleted (SK 311) → śākhyaḥ.
Likewise mukhyaḥ, agryaḥ, śaraṇyaḥ, and jaghanam iva "like the hind part" → jaghanyaḥ.
Watch out. Our text of the Kaumudī and the 1915 printed edition read the affix as ya. One of Bhaṭṭoji's commentators reads yat and calls ya a faulty reading, on grounds of accent; scholars discuss this. The forms are the same either way; only the accent differs.
Rule. The word dravya is fixed by nipātana from dru "tree", in the sense "like". It is used when the thing compared is bhavya, "worthy".
Example.
dravyam ayaṃ brāhmaṇaḥ "this brāhmaṇa is dravya", a worthy man.
One of Bhaṭṭoji's commentators explains the likeness. A tree bears flowers and fruit; so he is a vessel of what is desired. Or dru is the wish-granting tree.
The same commentator names the affix ya: dru + ya → guṇa of u → dro + ya → av for o → dravyam.
Applies when / blocked by. The other commentator names the affix ṇya, heard as -ya, its ṇ being a marker; scholars discuss this.
Watch out. Our text of the Kaumudī reads dvavyam. The 1915 printed edition reads dravyam, as the sūtra itself does. We follow the print.
Rule. After kuśāgra "the tip of kuśa grass", cha comes in the sense "like".
Example.
kuśāgram iva "like the tip of kuśa grass" → kuśāgra + cha
→ ch → īy (SK 475) → kuśāgra + īya
→ the final a deleted (SK 311) → kuśāgrīyā buddhiḥ "a fine intellect".
Applies when / blocked by. One of Bhaṭṭoji's commentators says the likeness lies in fineness. The intellect is as fine as a kuśa tip.
Rule. Cha also comes after a compound (samāsa) whose domain is the sense "like". The word ca "and" draws cha in from SK 2060. One of Bhaṭṭoji's commentators says so.
Example.
kākatāla "crow-and-palm", a compound of likeness → kākatāla + cha
→ kākatāla + īya → kākatālīyaḥ.
kākatālīyo devadattasya vadhaḥ "Devadatta's killing was kākatālīya".
Likewise ajākṛpāṇīyaḥ "goat-and-knife-like".
Applies when / blocked by. Two likenesses are at work, Bhaṭṭoji explains. The compound means Devadatta's meeting with the thief, like the meeting of crow and palm fruit. The affix means his killing, brought about by that, like the crow's death. The upshot is "come about unforeseen".
The compound must be made to convey likeness. One of Bhaṭṭoji's commentators gives śastrīśyāmā as a case where cha does not come. It holds only one likeness, already in the compound. What is expressed is not expressed again.
Both of Bhaṭṭoji's commentators say the sense "like" runs as a heading up to SK 2066. It does not include that rule.
Watch out. Our text of the Kaumudī reads sadṛśas caurasamāgama, with s before c. The 1915 printed edition reads sadṛśaś caurasamāgama, as sandhi requires. We follow the print; the words are the same.
Rule. After śarkarā "grit" and the words that follow it, aṇ comes in the sense "like".
Example.
śarkareva "like grit" → śarkarā + aṇ
→ the ṇ is an it → śarkarā + a
→ vṛddhi of the first vowel (SK 1075) → śārkarā + a
→ the final ā deleted (SK 311) → śārkaram.
Watch out. Śarkarā is feminine, but śārkaram is here neuter. One of Bhaṭṭoji's commentators explains. Svārthika affixes may go beyond the gender and number of their base. The derivative follows the thing it qualifies.
Rule. After aṅgulī "finger" and the words that follow it, ṭhak comes in the sense "like".
Example.
aṅgulīva "like a finger" → aṅgulī + ṭhak
→ the k is an it → aṅgulī + ṭha
→ ṭha → ika (SK 1170) → aṅgulī + ika
→ vṛddhi of the first vowel (SK 1076) → āṅgulī + ika
→ the final ī deleted (SK 311) → āṅgulikaḥ.
Likewise bharujeva "like a bharujā" → bhārujikaḥ.
Rule. After ekaśālā "a single hall", ṭhac comes optionally (anyatarasyām) in the sense "like". In the other option, ṭhak comes.
Example.
ekaśāleva "like a single hall" → ekaśālā + ṭhac
→ the c is an it → ekaśālā + ṭha
→ ṭha → ika (SK 1170) → ekaśālā + ika
→ the final ā deleted (SK 311) → ekaśālikaḥ.
With ṭhak, marked with k, vṛddhi comes too (SK 1076) → aikaśālikaḥ.
Bhaṭṭoji gives the two forms without naming their affixes. Pairing them with ṭhac and ṭhak is this book's inference, from the order in which he names the affixes.
Applies when / blocked by. One of Bhaṭṭoji's commentators explains the word anyatarasyām. It adds the ṭhak of SK 2063 as the other option. Otherwise the shorter "ṭhaj vā" would have served.
Rule. After karka and lohita "red", īkak comes in the sense "like". Bhaṭṭoji says a karka is a white horse.
Example.
karka iva "like a white horse" → karka + īkak
→ the final k is an it → karka + īka
→ vṛddhi of the first vowel (SK 1076) → kārka + īka
→ the final a deleted (SK 311) → kārkīkaḥ.
Likewise lauhitīkaḥ sphaṭikaḥ "a crystal that looks red".
Applies when / blocked by. One of Bhaṭṭoji's commentators explains the crystal. It looks red from a japā flower or the like placed beside it.
The next sūtra in Pāṇini's order, A 5.3.111, is Vedic. It is taught in the Vedic section (SK 3502).
kṣudraka → kṣaudrakyaḥ, kṣaudrakī; plural kṣudrakāḥThe sense "like" has now lapsed. The next rules give affixes in the base's own sense after the names of bands. A pūga is an association of mixed castes. An āyudhajīvisaṅgha is a band of men who live by arms. On the running example, the Kṣudraka band is kṣaudrakyaḥ, and a Kṣudraka woman is kṣaudrakī. The last rule names all these affixes tadrāja. So in the plural they vanish by luk: kṣudrakāḥ "the Kṣudrakas".
Rule. After a word denoting a pūga, ñya comes in its own sense. The word must not have a word for a grāmaṇī "leader" as its prior member. Bhaṭṭoji first says the sense of iva has lapsed (nivṛtta).
Example.
lohitadhvaja, "the pūga whose banners are red" → lohitadhvaja + ñya
→ the ñ is an it → lohitadhvaja + ya
→ vṛddhi of the first vowel (SK 1075) → lauhitadhvaja + ya
→ the final a deleted (SK 311) → lauhitadhvajyaḥ, that same pūga.
Applies when / blocked by. Bhaṭṭoji defines a pūga. It is a band (saṅgha) of men of various castes, bent chiefly on wealth and pleasure. Its conduct is not fixed (aniyatavṛtti). One of his commentators glosses this as "unruly".
The rule does not mean the word pūga itself. Both commentators say so. It means any word that names such a band.
Why "without a leader-word before it"? One of Bhaṭṭoji's commentators gives devadattakāḥ, "the band whose leader is Devadatta". There the whole compound names the pūga, so ñya does not come.
Bhaṭṭoji then cites SK 1100 (vrātacphañor astriyām), the next sūtra in Pāṇini's order. It gives the same ñya, not in the feminine. A vrāta differs from a pūga by living by bodily toil, both commentators say.
For a vrāta: kapotapāka → kāpotapākyaḥ. One of the commentators glosses kapotapāka as "one who cooks pigeons to eat".
For a cphañ-final word: kauñjāyana → kauñjāyanyaḥ; likewise brādhnāyanyaḥ.
Rule. Ñyaṭ comes in its own sense after a word for a band of men who live by arms (āyudhajīvisaṅgha). The band must be among the Vāhīkas. It must not be called brāhmaṇa or rājanya. Ñyaṭ is heard as -ya: its ñ and its ṭ are markers.
Example.
kṣudraka, an armed band of the Vāhīka country → kṣudraka + ñyaṭ
→ the ñ and ṭ are its → kṣudraka + ya
→ vṛddhi of the first vowel (SK 1075) → kṣaudraka + ya
→ the final a deleted (SK 311) → kṣaudrakyaḥ.
Likewise mālava → mālavyaḥ.
In the feminine, ṅīp comes because the affix is marked with ṭ (SK 470) → kṣaudrakī.
Applies when / blocked by. One of Bhaṭṭoji's commentators says vāhīkeṣu means among the settlements called Vāhīka. He also says "not in the feminine" of SK 1100 does not reach this rule. That is why kṣaudrakī is correct.
Bhaṭṭoji tests each word of the sūtra with a counter-example.
mallāḥ.samrāṭ.śabarāḥ.gopālakāḥ and śālaṅkāyanāḥ.Taking the first as brāhmaṇas and the second as a band called "rājanya" is this book's reading, following an older commentary. Bhaṭṭoji then remarks on the two words of the sūtra. For brāhmaṇa, its particular kinds are taken (tadviśeṣagrahaṇa). So every kind of brāhmaṇa band, such as the gopālakas, is excluded. For rājanya, only the word's own form is taken (svarūpagrahaṇa). So only a band called by the word "rājanya" itself is excluded. On this reading the śālaṅkāyanas are such a band.
Watch out. Our text of the Kaumudī has three slips here.
śabarāḥ. The 1915 printed edition has the words between: mallāḥ, "Why 'band'?", samrāṭ, "Why 'among the Vāhīkas'?". We follow the print. Without them, śabarāḥ would wrongly answer "Why 'arms'?".gopālakāḥ. We follow the print.Rule. After vṛka, as the name of a band of men who live by arms, ṭeṇyaṇ comes in its own sense. Its ṭ and ṇ are markers, so it is heard as -eṇya.
Example.
vṛka, the name of such a band → vṛka + ṭeṇyaṇ
→ the ṭ and ṇ are its → vṛka + eṇya
→ vṛddhi of the first vowel, ṛ → ār (SK 1075) → vārka + eṇya
→ the final a deleted (SK 311) → vārkeṇyaḥ.
Applies when / blocked by. Why "a band of men who live by arms"? Bhaṭṭoji says: so that the affix does not come after vṛka as a species-word. As a species-word, vṛka is the wolf. One of his commentators says so.
Rule. Cha comes in its own sense after dāmani and the words that follow it. It also comes after the trigartaṣaṣṭhas, "the groups of which Trigarta is the sixth". The words must name bands of men who live by arms.
Example.
dāmani → dāmani + cha → dāmani + īya
→ the final i deleted (SK 311) → dāmanīyaḥ; dual dāmanīyau.
Plural: the cha goes by luk, being tadrāja (SK 2072) → dāmanayaḥ.
One of Bhaṭṭoji's commentators gives this reason for the plural.
Likewise aulapi → aulapīyaḥ.
For the trigartaṣaṣṭhas: kauṇḍoparathīyaḥ, dāṇḍakīyaḥ.
Applies when / blocked by. Bhaṭṭoji quotes a verse that names the six groups: Kauṇḍoparatha, Dāṇḍaki, Krauṣṭuki, Jālamāni, Brahmagupta and Jālaki. One of his commentators reads the fifth name as Brāhmagupta; scholars discuss this.
Watch out. Our text of the Kaumudī reads trikartaḥ in the definition of the trigartaṣaṣṭhas. The 1915 printed edition reads trigartaḥ, as the sūtra and the verse do. We follow the print.
Rule. After parśu and the words that follow it, aṇ comes. After yaudheya and the words that follow it, añ comes. The pairing is "respectively" (kramāt). Both come in their own sense, for bands of men who live by arms.
Example.
parśu → a descendant of Parśu: pārśava, with the descendant aṇ (SK 1188)
→ a band of such descendants: aṇ again, by this rule → pārśavaḥ; dual pārśavau.
Plural: both aṇ affixes, the descendant one and the band one, go by luk as tadrāja → parśavaḥ.
One of Bhaṭṭoji's commentators says the present aṇ too is deleted by luk. If only the band aṇ went, the vṛddhi of pārśava would remain.
yaudheya → añ by this rule → yaudheyaḥ; dual yaudheyau.
Plural: luk of the añ → yaudheyāḥ.
Applies when / blocked by. Both of Bhaṭṭoji's commentators explain the bases. Each is already a descendant-word. Pārśava has the aṇ of SK 1188. Yaudheya is from yudhā with the ḍhak of SK 1124.
What does the ñ of añ do here? One of Bhaṭṭoji's commentators says it gives the word a high pitch (udātta) on its first vowel. The other adds a second purpose. Another rule gives an affix for "the mark of the Yaudheyas", and that affix comes only after an añ-final word.
Rule. Yañ comes in its own sense after a word ending in aṇ, added to one of seven names. The names are abhijit, vidabhṛt, śālāvat, śikhāvat, śamīvat, ūrṇāvat and śrumat. Yañ is heard as -ya.
Example.
abhijito 'patyam "the descendant of Abhijit" → ābhijita, with the descendant aṇ
→ ābhijita + yañ
→ the ñ is an it → ābhijita + ya
→ the final a deleted (SK 311) → ābhijityaḥ, that same one.
Likewise vaidabhṛtyaḥ, śālāvatyaḥ, śaikhāvatyaḥ, śāmīvatyaḥ, aurṇāvatyaḥ, śraumatyaḥ.
Both of Bhaṭṭoji's commentators give these steps.
Applies when / blocked by. The aṇ must be the descendant aṇ. So yañ does not come in ābhijito muhūrtaḥ, where the aṇ does not mean "descendant". One of Bhaṭṭoji's commentators says so, following the Bhāṣya.
The same commentator cites the Vṛtti, an older commentary on the sūtras. There the words "band of men who live by arms" cease at this rule.
Watch out. Our text of the Kaumudī reads śalāvatyaḥ. The 1915 printed edition reads śālāvatyaḥ, with the vowel of śālāvat. We follow the print.
Rule. The affixes taught from pūgāñ ñyaḥ (SK 2066) onward are named tadrāja. So, by SK 1193, they go by luk in the plural, but not in the feminine.
Example.
lauhitadhvajyaḥ → in the plural, ñya goes by luk → lohitadhvajāḥ.
Likewise kapotapākāḥ, kauñjāyanāḥ, brādhnāyanāḥ (the ñya of SK 1100).
One of Bhaṭṭoji's commentators fills in Bhaṭṭoji's "and so on":
kṣudrakāḥ (SK 2067), vṛkāḥ (SK 2068), dāmanayaḥ and kauṇḍoparathāḥ (SK 2069),
parśavaḥ and yaudheyāḥ (SK 2070), abhijitaḥ (SK 2071).
Applies when / blocked by. One of Bhaṭṭoji's commentators closes the third quarter of Pāṇini's fifth book here.
Watch out. Our text of the Kaumudī quotes SK 2066 as pūgāñyaḥ, with one ñ. The 1915 printed edition reads pūgāññyaḥ, as the sūtra itself does. We follow the print.
dvau dvau pādau → dvipadikām; sthūla → sthūlakaḥPāṇini's fourth quarter of the fifth book begins here. Its first two rules give vun, heard as -aka, after pāda or śata with a numeral before it. On the running example, "he gives two pādas at a time" is dvipadikāṃ dadāti. Then kan returns, now for a mode or sort (prakāra): sthūlakaḥ. It also comes for an action not carried right through: chinnakam "cut in part". A rule then forbids kan after sāmi "half". Last, kan after bṛhatī gives the name of a wrap.
Rule. After pāda or śata "hundred" preceded by a numeral, vun comes in its own sense. The sense must be distributive (vīpsā): "so many at a time". The final of the base is deleted (lopa) as well.
Example.
dvau dvau pādau dadāti "he gives two pādas at a time"
→ the compound dvipāda is formed first (SK 728)
→ dvipāda + vun
→ the n is an it → dvipāda + vu
→ the final a deleted by this rule → dvipād + vu
→ vu → aka (SK 1247) → dvipād + aka
→ the affix now begins with a vowel, so the stem is bha (SK 231); pad replaces pād (SK 414) → dvipad + aka
→ used only in the feminine: ṭāp, and i for the a before k (SK 463) → dvipadikā
→ dvipadikāṃ [dadāti].
Bhaṭṭoji leaves the aka and the feminine steps implicit.
Likewise dviśatikām "two hundreds at a time".
Applies when / blocked by. Why is the deletion stated in the rule? Yasyeti ca (SK 311) would delete the a anyway. Bhaṭṭoji answers: the deletion here is not conditioned by what follows (anaimittika). So it is not sthānivat, and pādaḥ pat (SK 414) can apply.
One of Bhaṭṭoji's commentators spells this out. The deletion by SK 311 is caused by the affix after it. So by SK 50 it would count as the vowel it removed, and pad would not replace pād.
Bhaṭṭoji also says that a word ending in vun is used only in the feminine. One of his commentators says this is by nature.
A vārttika says the mention of pāda and śata is pointless, because the affix is seen elsewhere too. Bhaṭṭoji's example: dvimodakikām. One of his commentators analyses it as "he gives two modakas (sweets) at a time".
Watch out. The same things without distribution take no affix. One of Bhaṭṭoji's commentators gives counter-examples. pādaṃ pādaṃ dadāti has no numeral. dvau pādau dadāti has no distribution. He also says forms like "two māṣas at a time" do not occur, by anabhidhāna.
Rule. The same vun, with deletion of the final, also comes when a fine (daṇḍa) or a giving away (vyavasarga) is meant. Bhaṭṭoji says this rule is for cases without distribution.
Example.
dvau pādau daṇḍitaḥ "he has been fined two pādas" → dvipāda + vun (as in SK 2073)
→ dvipadikāṃ [daṇḍitaḥ].
Likewise dviśatikām.
Applies when / blocked by. Bhaṭṭoji glosses vyavasṛjati as "he gives". One of his commentators explains daṇḍa as the taking of goods by force. He supplies vun and the deletion of the final, as in SK 2073.
Watch out. Bhaṭṭoji gives no verb for dviśatikām. His gloss of "gives", which follows it, suggests a gift. The text does not say so.
Rule. After sthūla "stout" and the words that follow it, kan comes when something having a mode (prakāra) is meant. Bhaṭṭoji calls it an apavāda of jātīyar (SK 2024).
Example.
sthūla + kan → sthūlakaḥ.
Likewise aṇu + kan → aṇukaḥ.
Applies when / blocked by. Both of Bhaṭṭoji's commentators say prakāra means both a variety and a likeness. Kan, like jātīyar, names the thing that has the mode, not the mode alone.
A vārttika adds cañcat and bṛhat to the list: cañcatkaḥ, bṛhatkaḥ. One of Bhaṭṭoji's commentators glosses them. cañcatkaḥ is what seems to move though it does not, like a gem in rippling water. bṛhatkaḥ is what seems big though it is not.
A gaṇasūtra in the list says: after surā, only when a snake is meant. It is a niyama, one of Bhaṭṭoji's commentators says. A snake the colour of surā is surakaḥ. Its ā is shortened by ke 'ṇaḥ (SK 834), the same commentator says.
Rule. After a word ending in kta, kan comes when the action is not carried right through (anatyantagati). One of Bhaṭṭoji's commentators supplies kan; tracing it to SK 2075 is this book's inference.
Example.
chinna "cut" → chinna + kan → chinnakam "cut in part".
Likewise bhinnakam, abhinnakam.
Applies when / blocked by. One of Bhaṭṭoji's commentators defines the complete action. It reaches every part of the thing, with nothing left over. Where that is lacking, kan comes. The other glosses chinnakam as "slightly cut".
Rule. Kan does not come after a kta-final word when sāmi "half", or a word of the same meaning, is the upapada.
Example.
sāmikṛtam "half-done" — no kan.
ardhakṛtam "half-done" — no kan.
Applies when / blocked by. Both of Bhaṭṭoji's commentators say the word vacana in the sūtra takes in synonyms of sāmi. So ardha "half" counts.
Bhaṭṭoji raises an objection. Here sāmi itself expresses the incompleteness. So SK 2076 would not give kan anyway. His answer: even so, this prohibition is a jñāpaka. It shows that a wholly svārthika kan also exists (atyantasvārthika). Such a kan leaves the base's sense just as it is: bahutarakam is just bahutaram, one of his commentators says. Both of his commentators say it is that kan which this rule forbids, not the kan of SK 2076.
Watch out. One of Bhaṭṭoji's commentators adds that the Bhāṣya rejects this sūtra. It derives bahutarakam another way. The two commentators also differ on whether sāmikṛta counts as ending in kta; scholars discuss this.
Rule. After bṛhatī, kan comes when a covering (ācchādana) is meant. One of Bhaṭṭoji's commentators says kan is supplied.
Example.
bṛhatī, a covering → bṛhatī + kan → bṛhatikā, an upper wrap.
Bhaṭṭoji quotes a line: prāvāra and uttarāsaṅga are synonyms, and so is bṛhatikā. One of his commentators says the line is from Amara's lexicon.
Applies when / blocked by. Why "a covering"? Bhaṭṭoji's counter-example is bṛhatī chandaḥ, the metre named bṛhatī. It takes no kan.
Watch out. This kan is obligatory (nitya). Both of Bhaṭṭoji's commentators say it is the "bṛhatī" of the Bhāṣya's list of obligatory svārthikas (SK 2079).
prāk → prācīnam; aṣaḍakṣa → aṣaḍakṣīṇaḥSix rules give affixes in the base's own sense after named stems. Kha, heard as -īna, comes after a few compounds. It is obligatory there, and Bhaṭṭoji uses the point to list the svārthika affixes that are obligatory. On the running example, a consultation "at which there are not six akṣi" is aṣaḍakṣīṇo mantraḥ. Kha is optional after a word ending in añc: prāk gives prācīnam. Then cha after words ending in jāti and sthāna, ṭhak after anugādin, and aṇ after visārin.
Rule. Kha comes in its own sense after aṣaḍakṣa, āśitaṅgu, alaṃkarma and alampuruṣa. It also comes after a word whose final member (uttarapada) is adhi.
Example.
aṣaḍakṣa "at which there are not six akṣi" → aṣaḍakṣa + kha
→ kh → īn (SK 475) → aṣaḍakṣa + īna
→ the final a deleted (SK 311) → aṣaḍakṣīna
→ the n after ṣ becomes ṇ (SK 235, SK 197) → aṣaḍakṣīṇo mantraḥ.
By those two rules of chapter 8, ṇ replaces an n that follows r or ṣ in the same word, even with vowels and some consonants between.
Bhaṭṭoji explains: a consultation held by two persons only.
āśitā gāvo 'smin "in which cows have been fed" → āśitaṅgavīnam araṇyam, a forest.
Bhaṭṭoji says the prior member takes the augment mum by nipātana. An augment (āgama) is a sound added to an item without replacing anything; mum adds an m.
alaṃ karmaṇe "equal to work" → alaṃkarmīṇaḥ, with ṇ after the r. Likewise alampuruṣīṇaḥ, with ṇ after the ṣ, and īśvarādhīnaḥ.
Applies when / blocked by. Both of Bhaṭṭoji's commentators take the six akṣi as ears, not eyes: the ears of three persons. The word aṣaḍakṣa is a bahuvrīhi, one of them says, with ṣac by SK 852.
Bhaṭṭoji says this kha is obligatory (nitya). The next sūtra says vibhāṣā "optionally". One of his commentators explains the reasoning. The great option would already make SK 2080 optional. So its word vibhāṣā shows that the great option does not reach this rule.
Bhaṭṭoji then quotes the Bhāṣya. Some other svārthika affixes too are held to be obligatory:
Both commentators give these ranges.
Watch out. The Bhāṣya writes "bṛhatī and jātyanta" in the plural. One of Bhaṭṭoji's commentators takes the plural to include pāśap (SK 1993) and the affixes after it. The other says only the listed affixes are meant; scholars discuss this.
Our text of the Kaumudī reads iṣyate, singular. The 1915 printed edition reads iṣyante, "are held", as the plural subject requires. We follow the print.
Rule. Kha comes optionally, in its own sense, after a stem ending in añc. It does not come when the stem refers to a feminine that is a direction.
Example.
prāk "in the east": its stem is pra + ac, from añc with its n lost → pra + ac + kha
→ kh → īn (SK 475) → pra + ac + īna
→ the a of ac deleted before the bha ending (SK 416) → pra + c + īna
→ the vowel before c lengthened (SK 417) → prācīnam.
One of Bhaṭṭoji's commentators names these two steps.
Likewise pratyak → pratīcīnam; avāk → avācīnam.
Bhaṭṭoji quotes a line of synonyms, among them arva. One of his commentators says it is Amara's list of words for "low".
arvantam añcati "it goes toward the low" → arvāk → arvācīnam.
Applies when / blocked by. Both of Bhaṭṭoji's commentators read adikstriyām as "not a feminine that is a direction". Bhaṭṭoji tests each part with a counter-example.
prācī dik "the eastern quarter", udīcī dik "the northern quarter". These are feminine directions, so kha does not come.prācīnā brāhmaṇī. The brāhmaṇī is feminine but not a direction, so kha comes.prācīnaṃ grāmād āmrāḥ "the mango trees are east of the village". The base names a direction but is not feminine, so kha comes.Watch out. Do not take prāk as the form without kha. Both of Bhaṭṭoji's commentators say prāk has its own affix, astāti (SK 1974), an affix for a direction, place or time. That affix is deleted by luk (SK 1980). It means "in the eastern place". prācīnam, one of them says, refers by nature to the thing located there.
In the last example the base is that indeclinable prāk. Both commentators say so. An indeclinable is not feminine.
Rule. After a stem ending in jāti "genus", cha comes in its own sense when the stem denotes a bandhu. Bhaṭṭoji defines a bandhu: a substance (dravya) that makes a genus manifest.
Example.
brāhmaṇajāti → brāhmaṇajāti + cha → brāhmaṇajāti + īya
→ the final i deleted (SK 311) → brāhmaṇajātīyaḥ, a person of the brāhmaṇa genus.
Applies when / blocked by. Why "a bandhu"? Bhaṭṭoji's counter-example is brāhmaṇajātiḥ śobhanā "the brāhmaṇa genus is excellent". There the word names the genus itself, not a substance.
Watch out. Bandhu here does not mean "kinsman". Both of Bhaṭṭoji's commentators say so. The sūtra's bandhuni is neuter, and the word for "kinsman" is masculine.
Rule. After a word ending in sthāna, cha comes optionally. It comes if the word is meaningful as sasthāna, which Bhaṭṭoji glosses "equal" (tulya).
Example.
pitṛsthāna "whose standing is like a father's" → pitṛsthāna + cha → pitṛsthāna + īya
→ the final a deleted (SK 311) → pitṛsthānīyaḥ.
In the other option: pitṛsthānaḥ.
Both of Bhaṭṭoji's commentators gloss this "like a father".
Applies when / blocked by. Why "equal"? Bhaṭṭoji's counter-example is goḥ sthānam "the place of a cow", a cow-stall. No equality is meant, so cha does not come.
One of Bhaṭṭoji's commentators supplies cha; tracing it to SK 2081 is this book's inference.
The next two sūtras of Pāṇini's order are not taught here. A 5.4.11 is SK 2004, in chapter 41. A 5.4.12 is Vedic (SK 3503).
Rule. After anugādin, ṭhak comes in its own sense.
Example.
anugadati "he speaks after [another]" → anugādin, the base
→ anugādin + ṭhak
→ the k is an it → anugādin + ṭha
→ ṭha → ika (SK 1170) → anugādin + ika
→ the ṭi in deleted (SK 679) → anugād + ika
→ vṛddhi of the first vowel (SK 1076) → ānugādikaḥ, that very person.
One of Bhaṭṭoji's commentators names SK 679.
Applies when / blocked by. Both of Bhaṭṭoji's commentators say this ṭhak is obligatory. One of them places it in the Bhāṣya's list, among "ām and the rest, before mayaṭ" (SK 2079). So the word anugādin only displays the base. The other adds that the bare word is not fit for use.
The next two sūtras of Pāṇini's order, A 5.4.14–15, are taught in the section on kṛt affixes (SK 3216, 3219).
Rule. After visārin, aṇ comes in its own sense when a fish (matsya) is meant. Both of Bhaṭṭoji's commentators say aṇ is carried forward from aṇinuṇaḥ (SK 3219), the sūtra before it in Pāṇini's order.
Example.
visārin → visārin + aṇ
→ the ṇ is an it → visārin + a
→ vṛddhi of the first vowel (SK 1075) → vaisārin + a
→ the n after r becomes ṇ (SK 235, SK 197) → vaisāriṇaḥ, a fish.
Applies when / blocked by. Why "a fish"? Bhaṭṭoji's counter-example is visārī devadattaḥ. Devadatta is not a fish, so no aṇ comes.
Watch out. The ṭi in is not deleted here, though SK 679 would delete it. Both of Bhaṭṭoji's commentators give the reason: by SK 1245 the stem keeps its form before an aṇ that does not mean "descendant".
pañca → pañcakṛtvaḥ; dvi → dviḥ; eka → sakṛtFour rules count how many times an action happens. After a number-word the affix is kṛtvasuc, heard as -kṛtvas: "he eats five times" is pañcakṛtvo bhuṅkte. After dvi, tri and catur it is suc, heard as -s: dviḥ "twice". For eka "one", sakṛt replaces the word: sakṛt "once". After bahu "many", dhā may come for times close together: bahudhā.
Rule. After a number-word (saṃkhyā), kṛtvasuc comes in its own sense. The number-word must count the occurrences (abhyāvṛtti) of an action (kriyā). Bhaṭṭoji glosses abhyāvṛtti as "coming into being" (janma).
Example.
pañca "five", counting eatings → pañca + kṛtvasuc
→ the c is an it, and the u is only for pronunciation → pañca + kṛtvas
→ pañcakṛtvo bhuṅkte "he eats five times".
Applies when / blocked by. Every occurrence counts, the first included. Both of Bhaṭṭoji's commentators explain the gloss "coming into being". Suppose only the repeats were counted. Then, one of them says, cooking done four times would be triḥ pacati "he cooks three times".
Why "after a number-word"? Bhaṭṭoji's counter-example is bhūrivārān bhuṅkte "he eats many times". Bhūri "many" is not a saṃkhyā. Both of his commentators give the reason. The word bahu in SK 258 restricts the name saṃkhyā. So synonyms of bahu, such as bhūri, are not number-words.
Watch out. A word in -kṛtvas is an avyaya. One of Bhaṭṭoji's commentators says so, by SK 448.
Rule. After dvi "two", tri "three" and catur "four", suc comes in counting the occurrences of an action. Bhaṭṭoji calls it an apavāda of kṛtvasuc.
Example.
dvi + suc
→ the c is an it, and the u is only for pronunciation → dvi + s
→ dvir bhuṅkte "he eats twice".
Likewise triḥ "three times".
catur + s → the s after r is dropped by rāt sasya (SK 280) → catur → catuḥ "four times".
Applies when / blocked by. One of Bhaṭṭoji's commentators says "counting the occurrences of an action" continues from SK 2085. He also says these words are avyayas, like the kṛtvas words.
Rule. For eka "one", sakṛt is the substitute (ādeśa). By ca "and", suc comes as well.
Example.
eka + suc → sakṛt replaces eka → sakṛt + s
→ the s dropped by saṃyogāntasya lopaḥ (SK 54) → sakṛt
→ sakṛd bhuṅkte "he eats once".
Applies when / blocked by. Only "action" (kriyā) continues from SK 2085, not "occurrences". One action done once does not recur. So eka must qualify the action. One of Bhaṭṭoji's commentators cites Kaiyaṭa, an earlier commentator on the Bhāṣya: the rule does not apply in eko bhuṅkte "one person eats".
Watch out. Bhaṭṭoji says the s of suc goes by SK 54, not by SK 252. SK 252 drops the endings su, ti and si. Its si does not take in sic, the affix of the verb form abhaitsīt. In the same way, its su does not take in suc. Both of Bhaṭṭoji's commentators explain it so. One says the three words, taken together, name only endings.
Rule. After bahu "many", dhā comes optionally in counting the occurrences of an action. The occurrences must fall at a time not far apart (aviprakṛṣṭa). Bhaṭṭoji glosses aviprakṛṣṭa as "near".
Example.
bahu + dhā → bahudhā divasasya bhuṅkte "he eats many times in the day".
In the other option, kṛtvasuc: bahukṛtvo divasasya bhuṅkte.
One of Bhaṭṭoji's commentators gives this other option.
Applies when / blocked by. Why "at a near time"? Bhaṭṭoji's counter-example is bahukṛtvo māsasya bhuṅkte "he eats many times in the month". A month is not a near time, so only kṛtvasuc comes.
Watch out. Which rule gives the sixth case in divasasya is disputed; scholars discuss this.
anna → annamayam; atithi → ātithyam; deva → devatāWith SK 2088 the rules for "so many times" ended. This group turns to particular senses. First comes abundance. Food presented in abundance is annamayam, with mayaṭ. When the things presented are many, the affixes of a collection may also come. Then ñya comes after four words in their own sense, as in ānantyam. Next comes the sense "for that" (tādarthya). What is for a guest is ātithyam, and water for the feet is pādyam. The group ends with tal after deva. The deva itself is devatā.
Rule. Mayaṭ comes after a first-case word. Its sense is tatprakṛtavacana, "the stating of that as prakṛta". Bhaṭṭoji explains prakṛta as "presented in abundance".
The word vacana "stating" can be read in two ways, Bhaṭṭoji says. It is made with the affix lyuṭ. That lyuṭ may denote the action itself (bhāva) or its locus (adhikaraṇa). So the rule has two senses.
Example. First sense, the action:
prakṛtam annam "food presented in abundance" → anna + mayaṭ
→ annamayam.
Likewise apūpamayam "cakes in abundance" and yavāgūmayī "gruel in abundance".
yavāgū + maya → mayaṭ is marked with ṭ, so the feminine takes ṅīp (SK 470) → yavāgūmayī.
One of Bhaṭṭoji's commentators gives this reason.
Second sense, the locus:
annamayo yajñaḥ "a sacrifice in which food is stated to be abundant".
apūpamayaṃ parva "a festival in which cakes are stated [to be made] in abundance".
Applies when / blocked by. Both of Bhaṭṭoji's commentators take tat "that" as marking the first case. In the first sense the affix only makes the abundance explicit. Being svārthika, it keeps the base's gender and number. In the second sense the affix names the place where the thing is stated. So the word takes the gender of the thing it describes: annamayo yajñaḥ is masculine, like yajña. Both commentators say this.
Watch out. The editor of the printed edition adds a note. The word tat, read as a separate statement, gives a wholly svārthika mayaṭ too. That mayaṭ adds no meaning at all and needs no abundance. So a form like cinmayaṃ brahma is correct.
Rule. When the thing presented in abundance is many things, the collective affixes come, "as for a collection" (samūhavat). By ca "and", mayaṭ comes too. The collective (sāmūhika) affixes are those taught under tasya samūhaḥ "a collection of that" (SK 1243).
Bhaṭṭoji calls this an atideśa, an extension. Affixes taught for one sense are made to hold in another.
Example.
modakāḥ prakṛtāḥ "sweets (modaka) presented in abundance" → modaka + ṭhak (the collective affix of SK 1256)
→ ṭha → ika (SK 1170); first-vowel vṛddhi o → au; the final a deleted (SK 311)
→ maudakikam.
With mayaṭ: modakamayam.
Likewise from śaṣkulī: śāṣkulikam and śaṣkulīmayam.
In the second sense, the locus:
maudakiko yajñaḥ or modakamayaḥ "a sacrifice at which sweets are presented in abundance".
Applies when / blocked by. One of Bhaṭṭoji's commentators says tatprakṛtavacane is carried forward from SK 2089. So both of that rule's senses hold here. The base must stand for many things (bahuṣu). Both commentators name the ṭhak of maudakikam as the collective affix of SK 1256.
Watch out. maudakikam is neuter, though modaka is masculine. One of Bhaṭṭoji's commentators notes that this svārthika affix departs from its base's gender.
Rule. After ananta, āvasatha, itiha and bheṣaja, ñya comes in the base's own sense.
Example.
ananta eva "ananta itself" → ananta + ñya
→ vṛddhi a → ā; the final a deleted (SK 311) → ānantyam.
āvasatha eva "the dwelling itself" → āvasathyam.
itiha + ñya → vṛddhi i → ai → aitihyam.
bheṣajam eva "medicine itself" → vṛddhi e → ai → bhaiṣajyam.
Applies when / blocked by. Itiha is not one word. Bhaṭṭoji calls it a nipātasamudāya, a group of particles. A nipāta is a particle, one of the indeclinable words (SK 447). One of Bhaṭṭoji's commentators says the group is used for a tradition of instruction. The other adds that the affix comes after it only because the rule names it: the group is not a nominal stem.
One commentator explains ananta: anta is destruction, and ananta is its absence.
Watch out. One of Bhaṭṭoji's commentators notes that ānantyam, though svārthika, oversteps its base's gender.
Rule. After a word ending in devatā "deity", yat comes. The sense is tādarthya, "being for that".
Bhaṭṭoji explains tādarthya. It is just tadartha "what is for that", with the affix ṣyañ in its own sense. One of Bhaṭṭoji's commentators says ṣyañ comes because the word belongs to the list headed by caturvarṇa.
Example.
agnidevatāyai idam "this is for the deity Agni" → agnidevatā + yat
→ the final ā deleted (SK 311) → agnidevatyam.
Likewise pitṛdevatyam "for the ancestors as deities".
Applies when / blocked by. What is a devatā here? One of Bhaṭṭoji's commentators answers. It is the one to whom an offered substance is addressed, and who is praised by the mantra. So pitṛdevatyam is covered. If devatā meant one class of beings known from scripture, the gods as distinct from ancestors and others, it would not be. The editor of the printed edition gives the same definition.
Rule. After pāda "foot" and argha "respectful offering", yat comes in the sense "for that". One of Bhaṭṭoji's commentators supplies "tādarthye yat" from SK 2092.
Example.
pādārtham udakam "water for the feet" → pāda + yat
→ the final a deleted (SK 311) → pādyam.
Likewise arghyam "water for the offering".
Applies when / blocked by. Bhaṭṭoji adds four vārttikas, each with new affixes.
First: nū replaces nava "new", and the affixes tnap, tanap and kha come. So nūtnam, nūtanam and navīnam, all "new". One of Bhaṭṭoji's commentators says these affixes add no meaning at all. He gives the steps of the third:
nava + kha → kha → īna (SK 475); nū for nava → nū + īna
→ guṇa of the ū (SK 847), then av → navīnam.
Second: naś ca purāṇe prāt, "and na after pra in the sense 'old'". By ca, the affixes just named come too. So praṇam, pratnam, pratanam and prīṇam, all "old". One commentator says prīṇam is the form with kha.
Third: after bhāga "share", rūpa "form" and nāman "name", dheya comes. So bhāgadheyam, rūpadheyam, nāmadheyam.
Fourth: after āgnīdhra and sādhāraṇa, añ comes. The words look the same: āgnīdhram, sādhāraṇam. Bhaṭṭoji gives the point: in the feminine, ṅīp (SK 470). So āgnīdhrī and sādhāraṇī. The other commentator adds a second purpose, the accent. Without añ, he says, the feminine would be āgnīdhrā śālā.
Watch out. Our text of the Kaumudī prints a whole sentence as the second vārttika: naś ca purāṇe prāt purāṇārthe vartamānāt praśabdān no vaktavyaḥ. The 1915 printed edition sets only naś ca purāṇe prāt as the vārttika. The rest is Bhaṭṭoji's gloss: "na is to be stated after the word pra used in the sense 'old'". We follow the print.
Rule. After atithi "guest", ñya comes in the sense "for that". Bhaṭṭoji says "tādarthye still holds".
Example.
atithaye idam "this is for the guest" → atithi + ñya
→ vṛddhi a → ā; the final i deleted (SK 311) → ātithyam.
Applies when / blocked by. One of Bhaṭṭoji's commentators says the base is the fourth-case word, atithaye "for the guest". This is the last rule on which the sense "for that" runs.
Rule. After deva, tal comes. One of Bhaṭṭoji's commentators calls it wholly svārthika. He says the sense "for that" has ceased here.
Example.
deva eva "the deva itself" → deva + tal → devatā.
Watch out. devatā is feminine, though deva is masculine. Both of Bhaṭṭoji's commentators note it. One says the svārthika affix oversteps its base's gender. This is the same word devatā that ended the base of SK 2092.
lohita → lohitakaḥ; feminine lohitikā / lohinikāSeven rules add an affix that leaves the sense as it was. Ka comes after avi, and kan after yāva and its list. Then kan comes after lohita "red" in three senses: a jewel, a passing colour, and a dyed thing. Kāla "black" follows in the last two. On the running example, lohitakaḥ kopena is "red with anger". A vārttika gives the feminine two forms, lohitikā and lohinikā. The group ends with ṭhak after vinaya and its list.
Rule. After avi, ka comes. One of Bhaṭṭoji's commentators says it is purely svārthika. Ka is heard as -ka and brings no vṛddhi.
Example.
avir eva "just avi" → avi + ka → avikaḥ.
Watch out. Bhaṭṭoji does not say which avi is meant. One of his commentators quotes a lexicon line that gives three senses: mountain, ram and sun.
Rule. After yāva and the words of its list, kan comes in their own sense.
Example.
yāva eva "just yāva" → yāva + kan → yāvakaḥ.
Likewise maṇikaḥ, from maṇi.
Applies when / blocked by. One of Bhaṭṭoji's commentators gives two senses of yāva. It is either a dish made of barley, such as boiled grain, or lac. The next four rules also give kan, with lohita and kāla.
Rule. After lohita "red", when it refers to a jewel, kan comes in its own sense. One of Bhaṭṭoji's commentators names the affix kan.
Example.
lohita eva maṇiḥ "the red one itself, [being] a jewel" → lohita + kan → lohitakaḥ.
Applies when / blocked by. One of Bhaṭṭoji's commentators says only a ruby is meant. A crystal made red by a japā flower beside it is another matter. That falls under SK 2100.
Rule. After lohita, kan comes when the word names an impermanent colour. Both of Bhaṭṭoji's commentators explain "impermanent": the colour ends while its bearer lasts.
Example.
lohitakaḥ kopena "[Devadatta is] red with anger".
Applies when / blocked by. A vārttika follows: lohitāl liṅgabādhanaṃ vā. After lohita, kan's blocking of the feminine affix is optional. So the feminine has two forms: lohitikā kopena and lohinikā kopena, "[she is] red with anger". One of Bhaṭṭoji's commentators gives the steps.
Where kan blocks the feminine affix: lohita + kan → lohitaka
→ the feminine ṭāp, and i for the a before k (SK 463) → lohitikā.
Where it does not: the feminine of lohita is lohinī, with n for t (A 4.1.39) → lohinī + kan
→ the ī shortened before ka (SK 834) → lohinika → ṭāp → lohinikā.
Without the vārttika, he says, only lohitikā would result.
Watch out. Is a ruby's redness not also impermanent? It perishes with the ruby. One of Bhaṭṭoji's commentators answers. A colour that lasts in its bearer from its origin to its end counts as permanent.
The same commentator defends the vārttika as a pointer to a wider principle. The other calls it unnecessary. Scholars discuss this.
Rule. After lohita, kan comes when the word is used of something dyed (rakta) with lac or the like.
Example.
lohitikā śāṭī or lohinikā śāṭī "a red-dyed sari".
Bhaṭṭoji says the vārttika liṅgabādhanaṃ vā of SK 2099 holds here too. So the feminine again has two forms.
Applies when / blocked by. Why a separate rule? Both of Bhaṭṭoji's commentators answer. A dyed cloth keeps its redness as long as the cloth lasts. So that colour is permanent, and SK 2099 would not reach it.
Rule. After kāla "black", kan comes. Bhaṭṭoji says the two rules before are both carried forward: varṇe cānitye (SK 2099) and rakte (SK 2100). So the black must be a passing colour or a dye.
Example.
A passing colour: kālakaṃ mukhaṃ vailakṣyeṇa "a face darkened by embarrassment".
A dye: kālakaḥ paṭaḥ "a cloth dyed black"; kālikā śāṭī "a sari dyed black".
kāla + kan → kālaka → in the feminine, i for the a before k (SK 463) → kālikā.
Applies when / blocked by. One of Bhaṭṭoji's commentators pairs the examples with the two rules. He adds the causes: shame, jealousy and the like for the face, and indigo for the cloth.
Rule. After vinaya and the words of its list, ṭhak comes in their own sense.
Example.
vinaya eva "vinaya itself" → vinaya + ṭhak
→ ṭha → ika; vṛddhi i → ai; the final a deleted (SK 311) → vainayikaḥ.
Likewise sāmayikaḥ, from samaya.
Applies when / blocked by. A gaṇasūtra in the list adds one word: upāyo hrasvatvaṃ ca. Upāya takes ṭhak, and its long vowel is shortened.
upāya → upaya → upaya + ṭhak → vṛddhi u → au → aupayikaḥ.
One of Bhaṭṭoji's commentators says the long vowel must be meant. Shortening a short vowel would be pointless.
Watch out. The other commentator adds a word from the list: ākasmikam, from akasmāt. Its ṭi goes by avyayānāṃ bhamātre ṭilopaḥ: an indeclinable loses its ṭi before a taddhita that makes it a bha stem.
mṛd → mṛttikā; praised: mṛtsā, mṛtsnāSix rules, each for one base or one short list. Ṭhak comes after vāc "speech" when it is a message, giving vācikam. Then aṇ comes after karman "action", after oṣadhi "plant", and after prajña "knowing" and its list. The last two rules are about mṛd "earth". Earth itself is mṛttikā, with tikan. Earth that is praised is mṛtsā or mṛtsnā, with sa or sna.
Rule. After vāc "speech", ṭhak comes in its own sense. The speech must be vyāhṛtārthā: one whose content has been given as a message. Bhaṭṭoji's gloss names the affix ṭhak.
Example.
vāc + ṭhak → ṭha → ika; the first vowel is already ā → vācikam.
Bhaṭṭoji quotes a line: "A message-speech is [called] vācikam."
Applies when / blocked by. One of Bhaṭṭoji's commentators explains vyāhṛta. It is what is told to a messenger in the form "this is to be said to him". The other gives a counter-example: madhurā vāg devadattasya "Devadatta's speech is sweet". That speech carries no message, so no ṭhak comes.
Watch out. vācikam is neuter, though vāc is feminine. One of Bhaṭṭoji's commentators cites the principle that svārthika affixes sometimes go beyond their base's gender.
Rule. After karman "action", aṇ comes in its own sense. The action must be connected with that message (vācika).
Example.
karmaiva "just the action" → karman + aṇ
→ vṛddhi a → ā; the ṭi an kept by an (SK 1155) → kārmaṇam.
Bhaṭṭoji gives the meaning: an action performed after hearing a message.
Applies when / blocked by. One of Bhaṭṭoji's commentators gives the reason the ṭi stays. The base "stays as it is" (prakṛtibhāva) by an (SK 1155), so its ṭi is not deleted. He adds that the action is done exactly so, after hearing the envoy's words.
Rule. After oṣadhi "plant", aṇ comes in its own sense. The word must not name a species (jāti).
Example.
oṣadhi + aṇ → vṛddhi o → au; the final i deleted (SK 311) → auṣadham.
auṣadhaṃ pibati "he drinks medicine".
Applies when / blocked by. Why "not naming a species"? Bhaṭṭoji's counter-example is oṣadhayaḥ kṣetre rūḍhāḥ "the plants have sprung up in the field". There oṣadhi names a kind of plant, so no aṇ comes.
One of Bhaṭṭoji's commentators explains both. The medicine drunk is a powder of dry ginger, pepper and the like, mixed with water. That is not a species. The plants in the field are grain crops such as rice.
Rule. After prajña "knowing" and the words of its list, aṇ comes in its own sense. One of Bhaṭṭoji's commentators supplies "svārthe 'ṇ".
Example.
prajña eva "the knowing one himself" → prajña + aṇ
→ vṛddhi a → ā; the final a deleted (SK 311) → prājñaḥ.
In the feminine, ṅīp, since the stem ends in aṇ (SK 470): prājñī strī "a knowing woman".
Bhaṭṭoji also gives daivataḥ and bāndhavaḥ, without naming their bases.
Watch out. Do not confuse prājñī with prājñā. Both of Bhaṭṭoji's commentators explain the second. "She who has wisdom (prajñā)" takes a different affix, ṇa. It is heard as -a, and its ṇ is a marker that brings vṛddhi (SK 1075). Then ṭāp follows. So she is prājñā.
Rule. After mṛd "earth", tikan comes in its own sense. Tikan is heard as -tika.
Example.
mṛd eva "just earth" → mṛd + tikan
→ the n is an it → mṛd + tika
→ d before t becomes t (SK 121) → mṛttika
→ the feminine ṭāp → mṛttikā.
Bhaṭṭoji gives only the result; the d → t change and the ṭāp are left implicit.
Watch out. The i is written into the affix itself. One of Bhaṭṭoji's commentators asks why, since SK 463 would supply it before ṭāp anyway. He gives two reasons. It is simpler, and the i is then heard even when ṭāp is deleted by luk. His example is pañcamṛttikaḥ paṭaḥ "a cloth bought for five mṛttikās".
Rule. When praise (praśaṃsā) is meant, sa and sna come in their own sense. One of Bhaṭṭoji's commentators supplies the base mṛd "earth". Bhaṭṭoji calls them an apavāda of rūpap (SK 2021), the general affix of praise.
Example.
praśastā mṛd "praised earth" → mṛd + sa
→ d before s becomes t (SK 121) → mṛtsa
→ the feminine ṭāp → mṛtsā.
mṛd + sna → likewise → mṛtsnā.
Bhaṭṭoji gives only the results; these steps are left implicit.
Applies when / blocked by. Bhaṭṭoji says this injunction is nitya, obligatory. The reason is the next sūtra, which says anyatarasyām "optionally". One of Bhaṭṭoji's commentators agrees. He adds that mere mṛd does not convey praise.
Watch out. The other commentator reports a contrary view. Sa and sna are not in the Bhāṣya's list of obligatory svārthika affixes. So, "they say", they should be optional. Scholars discuss this. The next sūtra in Pāṇini's order is a Vedic rule, taught in the Vedic section (SK 3504).
bahūni dadāti → bahuśaḥ; dvau dvau → dviśaḥ; grāmāt → grāmataḥTwo affixes fill this group. Śas is heard as -śas. It comes after words for "much" or "few", and after numbers in a distributive sense. bahūni dadāti "he gives many things" → bahuśaḥ [dadāti]. Tasi is heard as -tas. It comes after a word in the fifth, third or sixth case, in set senses. grāmād āgacchati "he comes from the village" → grāmataḥ [āgacchati].
Both affixes make indeclinables. One of Bhaṭṭoji's commentators says so under SK 2110. Śas and every affix after it, up to ḍāc (SK 2128–2138), are listed as avyayas: śasprabhṛtayaḥ prāk samāsāntebhyaḥ. So bahuśaḥ never changes its form.
Rule. After a word meaning "much, many" (bahu) or "little, few" (alpa), śas optionally comes. The word must denote a kāraka. One of Bhaṭṭoji's commentators adds "in its own sense".
Example.
bahūni dadāti "he gives many [things]" → bahu + śas → bahuśaḥ [dadāti].
alpāni "few [things]" → alpaśaḥ.
Applies when / blocked by. The sūtra says "meaning" (artha). So words of the same meaning count too: bhūriśaḥ. Both of Bhaṭṭoji's commentators say this. One of them adds that any kāraka will do: bahubhyo dadāti "he gives to many" → bahuśaḥ too. He also gives two counter-examples. gāṃ dadāti "he gives a cow" has no word for much or few. bahūnāṃ svāmī "owner of many" is a sixth case, not a kāraka.
A vārttika narrows the rule. Śas comes only where the auspicious or the inauspicious is expressed. So it does not come in bahūni dadāty aniṣṭeṣu "he gives much on unwelcome occasions". Nor in alpaṃ dadāty ābhyudayikeṣu "he gives little on occasions for prosperity".
Watch out. How these counter-examples fit the vārttika's wording is a point scholars discuss. One of Bhaṭṭoji's commentators would read the vārttika as "auspicious" alone.
Our text of the Kaumudī reads bahvalpārthan at the start of the vārttika. The 1915 printed edition reads bahvalpārthān, "after [a word] meaning much or few". We follow the print.
Rule. Śas optionally comes in a distributive sense (vīpsā). It comes after a number-word, or after an ekavacana, a word expressing a single thing. Bhaṭṭoji says "kārakāt still holds": the word must denote a kāraka.
Example.
dvau dvau dadāti "he gives two at a time" → dvi + śas → dviśaḥ.
māṣaṃ māṣaṃ [dadāti] "[he gives] a māṣa at a time" → māṣaśaḥ.
Likewise prasthaśaḥ, a prastha at a time.
Applies when / blocked by. Bhaṭṭoji explains ekavacana. Words of quantity (parimāṇa), such as māṣa and prastha, are always single in sense inside the derived word. Those are the words meant.
He then gives three counter-examples.
ghaṭaṃ ghaṭaṃ dadāti "he gives pot after pot": ghaṭa is neither a number nor a quantity-word.dvau dadāti "he gives two": no distribution.dvayor dvayoḥ svāmī "owner of every two": no kāraka.One of Bhaṭṭoji's commentators explains the doubling. In dvau dvau dadāti the word is doubled for distribution. In dviśaḥ there is no doubling, since śas itself expresses the distribution.
Watch out. The restriction to quantity-words is the view of Jayāditya, both of Bhaṭṭoji's commentators report. Vāmana found no authority for it. He lets śas come after a genus-word too, when it is single in sense: kuṇḍaśo dadāti. For him, ekavacana only shuts out ghaṭau ghaṭau dadāti "he gives two pots at a time".
Rule. Tasi comes after a fifth-case word, when that fifth case is prescribed in connection with prati as a karmapravacanīya. Prati takes that name in the sense "representative" or "exchange" (SK 599). The fifth case is prescribed with it by SK 600. One of Bhaṭṭoji's commentators names both rules.
Example.
kṛṣṇāt (fifth case) → kṛṣṇa + tasi → kṛṣṇataḥ.
pradyumnaḥ kṛṣṇataḥ prati "Pradyumna is Kṛṣṇa's counterpart".
Applies when / blocked by. A vārttika adds more bases: tasi comes after ādi "beginning" and the words like it. One of Bhaṭṭoji's commentators says this tasi comes after any case.
ādau "at the beginning" → āditaḥ.
Likewise madhyataḥ, pṛṣṭhataḥ, pārśvataḥ.
svareṇa "by the accent" → svarataḥ; likewise varṇataḥ.
Bhaṭṭoji calls this list an ākṛtigaṇa.
Watch out. The other commentator points to evidence for the vārttika inside Pāṇini. SK 8, tasyādita udāttam ardhahrasvam, itself uses the form āditaḥ.
Rule. Tasi comes after a fifth-case word that stands for an apādāna. The apādāna must not be connected with hīya or ruh.
Example.
grāmād āgacchati "he comes from the village" → grāma + tasi → grāmataḥ [āgacchati].
Applies when / blocked by. Why "not with hīya or ruh"? Bhaṭṭoji gives two counter-examples. svargād dhīyate "he falls from heaven". parvatād avarohati "he climbs down from the mountain". Neither takes tasi.
One of Bhaṭṭoji's commentators divides the word as a-hīya-ruhoḥ. Hīya imitates the passive stem, as in hīyate. The other adds a further example: adhyayanāt parājayate "he recoils from study" → adhyayanataḥ.
Rule. After a third-case word that does not denote the agent, tasi optionally comes. The sense must be one of three: atigraha "being taken as surpassing", avyathana "not wavering", or kṣepa "censure". Bhaṭṭoji's gloss says "optionally" (vā).
Example. Each sense, with both forms:
cāritreṇa atigṛhyate or cāritrato 'tigṛhyate "by his conduct he is regarded as surpassing [others]".
vṛttena na vyathate or vṛttato na vyathate "by [reason of] his conduct he does not waver".
vṛttena kṣiptaḥ or vṛttataḥ kṣiptaḥ "he is reproached for his conduct".
Applies when / blocked by. Why "not the agent"? devadattena kṣiptaḥ "reproached by Devadatta". Devadatta is the agent, so no tasi.
One of Bhaṭṭoji's commentators says the third case here expresses a cause. He derives vyathana from vyath in the sense "to move", so avyathana is "not moving".
Watch out. Our text of the Kaumudī reads cāritreṇānyanatikramya in Bhaṭṭoji's gloss of the first example. The 1915 printed edition reads cāritreṇānyān atikramya, "by his conduct surpassing others". Both of Bhaṭṭoji's commentators quote anyān atikramya. We follow the print.
Rule. Tasi optionally comes after a third-case word, not denoting the agent, that is connected with hīyamāna "falling short" or pāpa "bad".
Example.
vṛttena hīyate "he falls short in conduct" → vṛttato hīyate.
vṛttena pāpaḥ "he is bad in conduct" → vṛttataḥ pāpaḥ.
Bhaṭṭoji gives the single form vṛttataḥ for both.
Applies when / blocked by. Bhaṭṭoji says this rule is for when censure is not intended. Where there is censure, SK 2113 already gives the form. One of Bhaṭṭoji's commentators adds: here the fact is merely stated.
Why "not the agent"? devadattena hīyate "he is abandoned by Devadatta". Devadatta is the agent of the passive verb, so no tasi.
Rule. Tasi optionally comes after a sixth-case word. The sense is vyāśraya: one side is taken among several. Bhaṭṭoji glosses it as "resorting to one of several sides".
Example.
arjunasya (sixth case) → arjuna + tasi → arjunataḥ.
devā arjunato 'bhavan "the gods were on Arjuna's side".
ādityāḥ karṇato 'bhavan "the Ādityas were on Karṇa's side".
Applies when / blocked by. Why "when a side is taken"? vṛkṣasya śākhā "the branch of the tree" takes no tasi. One of Bhaṭṭoji's commentators explains taking a side. One gives up what is common to all and adopts one side as one's own.
Watch out. Our text of the Kaumudī reads ṣaṣṭhyāntād vā in Bhaṭṭoji's gloss. The 1915 printed edition reads ṣaṣṭhyantād vā, "optionally after a sixth-case word". We follow the print.
Rule. Tasi optionally comes after a sixth-case word naming a disease (roga). The sense is its removal (apanayana). Bhaṭṭoji glosses the sense as cikitsā "treatment".
Example.
pravāhikāyāḥ (sixth case) → pravāhikā + tasi → pravāhikātaḥ.
pravāhikātaḥ kuru "make a remedy for the pravāhikā".
Applies when / blocked by. Why "removal"? pravāhikāyāḥ prakopanaṃ karoti "he aggravates the pravāhikā" takes no tasi. Both of Bhaṭṭoji's commentators identify pravāhikā with viṣūcikā. One adds pracchardikātaḥ kuru, for the disease of vomiting.
Watch out. Our text of the Kaumudī reads pravāhikātaḥ kuruḥ. The 1915 printed edition reads pravāhikātaḥ kuru, with the imperative "make". Bhaṭṭoji's own gloss, kuru, agrees. We follow the print.
akṛṣṇaḥ kṛṣṇaḥ saṃpadyate, taṃ karoti → kṛṣṇīkaroti; kṛtsnaṃ śastram agniḥ saṃpadyate → agnisād bhavatiThis group is about change. Something comes to be what it was not, and the verb is kṛ "make", bhū "become" or as "be". The affix is cvi. Nothing of it is heard. Its c is an it, its i is only for pronunciation, and its v is deleted (SK 375). But the stem before it changes. A final a becomes ī, and other vowels are lengthened. On the running example, "he makes one who is not black become black" is kṛṣṇīkaroti.
Three of these stem rules come from Pāṇini's sixth and seventh books (SK 2118–2120). Bhaṭṭoji sets them here because they build the forms of cvi. A fourth, from the eighth book, explains the s of sāti (SK 2123).
Sāti, heard as -sāt, is used when the change is total or pervasive. "The whole weapon becomes fire" is agnisād bhavati. Trā, heard as -trā, is used for what is handed over to someone. The great option still runs, so an ordinary sentence is always possible too.
Rule. In connection with kṛ, bhū or as, cvi comes. It comes after a word applied to the saṃpadyakartṛ, "the agent of becoming": the thing that is becoming something.
A vārttika adds a condition: abhūtatadbhāve, "when [something] comes to be what it was not". Bhaṭṭoji then glosses the whole. There is a source (prakṛti) that is taking on the nature of a product (vikāra). The word for the product is applied to the source. After that word, cvi optionally comes in its own sense, with karoti "makes" and the others.
Example. Bhaṭṭoji gives the forms under the next rule. The frame is this:
akṛṣṇaḥ kṛṣṇaḥ saṃpadyate "one who is not black becomes black".
The one becoming is the source. "Black" is the product. The word kṛṣṇa is applied to him, so cvi comes after kṛṣṇa.
Applies when / blocked by. The option is the great option, Bhaṭṭoji says at SK 2120. One of his commentators says that option is still running here.
The verb must be kṛ, bhū or as. One of Bhaṭṭoji's commentators gives a counter-example with another verb: aśuklaḥ śuklo jāyate "one not white is born white". No cvi comes.
Saṃpadya is a noun, "becoming", made from sam + pad by this very rule. One commentator reads the compound as "the agent of becoming"; the other as "the agent that is the one becoming". Scholars discuss this.
Watch out. The source itself must be meant as having become the product. One of Bhaṭṭoji's commentators says cvi comes only where the source is taken as not different from the product. The editor of the printed edition says the same of abhūtatadbhāve.
Rule. Before cvi, ī replaces a or ā. One of Bhaṭṭoji's commentators says ī is carried forward from ī ghrādhmoḥ (SK 2648). Bhaṭṭoji adds two things. The v of cvi is deleted. And a word ending in cvi is an indeclinable (avyaya).
This rule is from Pāṇini's seventh book. Bhaṭṭoji inserts it here to build the forms of SK 2117.
Example.
akṛṣṇaḥ kṛṣṇaḥ saṃpadyate, taṃ karoti "one not black becomes black; [someone] makes him [so]" → kṛṣṇa + cvi
→ ī for the final a → kṛṣṇī + v
→ the v deleted (SK 375) → kṛṣṇī
→ kṛṣṇīkaroti.
One of Bhaṭṭoji's commentators pictures an actor who is not black but is made up as black.
brahmībhavati "[he] becomes brahman". The same commentator gives the steps. Brahman counts as a pada, so its n is deleted; then ī replaces the a.
gaṅgī syāt "[it] would become the Gaṅgā": gaṅgā → gaṅgī.
Applies when / blocked by. Why is a cvi-word indeclinable? One of Bhaṭṭoji's commentators answers. It is a nipāta by SK 762, and nipātas are indeclinable by SK 447. The other commentator agrees. So kṛṣṇī never takes case-endings.
A vārttika says there is no ī for an indeclinable before cvi:
doṣābhūtam ahaḥ "a day become night".
divābhūtā rātriḥ "a night become day".
Doṣā "at night" and divā "by day" are indeclinables, so their ā stays. Bhaṭṭoji says the Bhāṣya states this vārttika under avyayībhāvaś ca (SK 451). One commentator explains the scenes. The day turns dark under heavy cloud; the night turns bright with moonlight.
Watch out. Our text of the Kaumudī reads cvyantātvād avyayatvam, with a stray long ā. The 1915 printed edition reads cvyantatvād avyayatvam, "because it ends in cvi, it is an indeclinable". We follow the print.
Rule. When kya or cvi follows, the y of a descendant (āpatya) affix is deleted. The y must come after a consonant. Bhaṭṭoji analyses the sūtra word as "before kya and before cvi" (kye cvau ca). Kya covers three affixes: kyac, kyaṅ and kyaṣ.
This rule is from Pāṇini's sixth book, inserted here for the forms of cvi.
Example.
agārgyo gārgyaḥ saṃpadyate "one not a Gārgya becomes a Gārgya" → gārgya + cvi
→ ī for the final a (SK 2118) → gārgyī + v
→ the y of the descendant affix yañ deleted → gārgī + v
→ the v deleted (SK 375) → gārgībhavati.
One of Bhaṭṭoji's commentators gives these steps.
Applies when / blocked by. One of Bhaṭṭoji's commentators names the words carried forward. "Deletion" comes from SK 234, "after a consonant" from SK 472, "y" from SK 499, and "descendant" from SK 1082.
Why is the rule needed? Both of Bhaṭṭoji's commentators answer. SK 1082 deletes the y of a descendant affix before a taddhita. Here ī stands in between, so SK 1082 cannot act. One adds that SK 472 needs an ī that carries meaning (arthavat). The ī here carries none.
Watch out. Do not read the sūtra as "before kyac and cvi". Its first member is kya, and it takes in kyac, kyaṅ and kyaṣ alike.
Rule. Before cvi, the preceding vowel is lengthened (dīrgha). One of Bhaṭṭoji's commentators says "lengthened" is carried forward from SK 2298.
This rule is from Pāṇini's seventh book, inserted here for the forms of cvi.
Example.
śuci + cvi → i lengthened → śucībhavati.
paṭu + cvi → u lengthened → paṭūsyāt.
mātṛ + cvi → rīṅ ṛtaḥ (SK 1234): rī for the ṛ → mātrīkaroti.
Applies when / blocked by. Some say an indeclinable is not lengthened. Bhaṭṭoji calls this baseless. One of his commentators says it is not found in the Bhāṣya.
What then of svasti syāt, with short i? Bhaṭṭoji answers in steps.
svastīsyāt would come too. Then let it come.Rule. The final sound of arus, manas, cakṣus, cetas, rahas and rajas is deleted, and cvi comes. Bhaṭṭoji's gloss: "these are deleted, and cvi [comes]". One of his commentators says "the final" is meant.
Example.
anaruḥ aruḥ saṃpadyate, tat karoti → arus + cvi
→ the final s deleted → aru
→ u lengthened (SK 2120) → arūkaroti.
unmanas + cvi → the final s deleted → unmana
→ ī for the a (SK 2118) → unmanīsyāt.
Likewise uccakṣūkaroti, vicetīkaroti, virahīkaroti and virajīkaroti.
One of Bhaṭṭoji's commentators gives these steps.
Applies when / blocked by. Cvi is already given by SK 2117. Both of Bhaṭṭoji's commentators say the word ca only restates it. This restatement is an anuvāda: repeating what another rule has already taught. It is done so that the deletion is taught together with cvi (saṃniyogaśiṣṭa).
Watch out. Our text of the Kaumudī reads virājīkaroti. The 1915 printed edition reads virajīkaroti, from rajas. One of Bhaṭṭoji's commentators reads it so too. We follow the print.
Rule. In the domain of cvi, sāti optionally comes when totality (kārtsnya) is meant. Bhaṭṭoji glosses totality as sākalya "wholeness". Sāti is heard as -sāt.
One of Bhaṭṭoji's commentators explains "the domain of cvi". It means the conditions of SK 2117. Something comes to be what it was not. The word applies to the one becoming. The verb is kṛ, bhū or as.
Example. Bhaṭṭoji gives this rule's examples after the next sūtra, SK 2123:
kṛtsnaṃ śastram agniḥ saṃpadyate "the whole weapon becomes fire" → agni + sāti → agnisād bhavati.
Applies when / blocked by. The great option already makes the affix optional. So why does the sūtra say vibhāṣā "optionally"? One of Bhaṭṭoji's commentators answers. It lets the general cvi still come where the special sāti could come. Both affixes then stand side by side (samāveśa). So agnībhavati stands beside agnisād bhavati.
Rule. The s of sāt is not replaced by ṣ. Nor is an s at the beginning of a pada. The replacement of s by ṣ is called ṣatva. One of Bhaṭṭoji's commentators says "not" and "by ṣ" are carried forward from earlier rules.
This rule is from Pāṇini's eighth book. Bhaṭṭoji sets it between SK 2122 and that rule's examples. It explains why the s of agnisāt stays s.
Example.
dadhi siñcati "he sprinkles curd": the s begins the pada siñcati, so it stays.
agni + sāti → agnisāt, not agniṣāt.
One of Bhaṭṭoji's commentators explains both. The s of siñcati is itself a substitute for the root's ṣ. The s of sāt belongs to an affix. So each would otherwise be replaced by ṣ. This rule forbids it.
Now SK 2122's examples:
kṛtsnaṃ śastram agniḥ saṃpadyate "the whole weapon becomes fire"
→ with sāti: agnisād bhavati
→ with cvi: agnībhavati, the i lengthened (SK 2120)
→ by the great option, the sentence itself too.
One of Bhaṭṭoji's commentators glosses kṛtsna "whole" as "with all its parts".
Applies when / blocked by. Why "totality"? ekadeśena śuklībhavati paṭaḥ "the cloth becomes white in one part". The change is partial, so only cvi comes.
Watch out. Our text of the Kaumudī reads kārtsye kim, "why totality?". The 1915 printed edition reads kārtsnye kim, matching the word of SK 2122. We follow the print.
Rule. When pervasion (abhividhi) is meant, sāti optionally comes. It comes in connection with sampad "turn into", and also with kṛ, bhū and as. Bhaṭṭoji glosses abhividhi as vyāpti "pervading". One of his commentators says the ca "and" gathers in kṛ, bhū and as. This gathering-in is called samuccaya.
In the other option, cvi comes with kṛ, bhū and as. With sampad, the other option is only the plain sentence.
Example.
agnisāt saṃpadyate or agnisād bhavati śastram "the weapons turn into fire".
With cvi: agnībhavati.
jalasāt saṃpadyate "the salt turns into water"; with cvi: jalībhavati lavaṇam.
Applies when / blocked by. Bhaṭṭoji contrasts this rule with SK 2122. Totality (kārtsnya) is one individual changed in all its parts. Pervasion (abhividhi) is many individuals changed in some part. One of Bhaṭṭoji's commentators says śastram "weapon" here names the class: all the weapons.
Why no cvi with sampad? One of Bhaṭṭoji's commentators answers. SK 2117 gives cvi only with kṛ, bhū and as.
Watch out. Our text of the Kaumudī reads jalābhavati. The 1915 printed edition reads jalībhavati, with ī for a by SK 2118. We follow the print.
Our text also skips some words in the contrast. It reads "the change of one individual in all its parts is abhividhi". The 1915 printed edition has two sentences. The change of one individual in all its parts is totality; the change of many individuals in some part is pervasion. We follow the print. One of Bhaṭṭoji's commentators draws the same contrast.
Rule. Sāti comes, with kṛ, bhū, as and sampad, when "dependent on that" (tadadhīna) is expressed.
Example.
rājasāt karoti "he makes [it] the king's".
rājasāt saṃpadyate "[it] becomes the king's".
Bhaṭṭoji gives the meaning: "dependent on the king" (rājādhīnam).
Applies when / blocked by. One of Bhaṭṭoji's commentators says abhūtatadbhāve has ceased here. The thing need not come to be what it was not. Bhaṭṭoji's gloss does not repeat "optionally". The next rule shows sāti alternating with trā.
Rule. When what is dependent on someone is a thing to be given (deya), trā comes, and sāti too. Both come with kṛ, bhū and the others. One of Bhaṭṭoji's commentators says only tadadhīnavacane is carried forward. He takes "the others" as as and sampad.
Example.
viprādhīnaṃ deyaṃ karoti "he makes [a thing] to be given dependent on the brāhmaṇa" → vipra + trā → vipratrā karoti.
Likewise vipratrā saṃpadyate.
In the other option: viprasāt karoti.
Applies when / blocked by. Why "to be given"? rājasād bhavati rāṣṭram "the kingdom becomes the king's". A kingdom is not given here, so only sāti comes.
Rule. After deva, manuṣya, puruṣa, puru and martya, trā comes variously (bahulam). The words must be in the second or the seventh case. One of Bhaṭṭoji's commentators calls this trā wholly svārthika. He says sāti and "with kṛ, bhū, as" have ceased.
Example.
devān vande "I pay homage to the gods" → devatrā vande.
deveṣu rame "I delight among the gods" → devatrā rame.
Both of Bhaṭṭoji's commentators add manuṣyatrā, puruṣatrā, purutrā and martyatrā. They say puru means "many".
Applies when / blocked by. Because the sūtra says bahulam, trā comes after other words too. Bhaṭṭoji's example is bahutrā jīvato manaḥ. One of his commentators explains it: the mind of a living being goes out to many objects.
Watch out. The trā is carried forward from SK 2126, but here it is wholly svārthika. It needs no verb kṛ, bhū or as, and there is nothing to be given.
paṭat → paṭapaṭākaroti; dvitīyaṃ karṣaṇaṃ karoti → dvitīyākarotiThe last eleven rules give ḍāc, heard as -ā. Its ḍ deletes the base's ṭi (SK 316). The word it makes is joined to the verb that follows: dvitīyākaroti "he does a second ploughing". The first rule is for imitations of sound, such as paṭat. The sound-word is doubled first, so paṭat gives paṭapaṭākaroti.
The verb matters. In SK 2128, "with kṛ, bhū or as" comes back by a frog's leap (maṇḍūkapluti) from SK 2117. From SK 2129 on, the verb is kṛ alone. One of Bhaṭṭoji's commentators says kṛñ, the root kṛ "make", is carried forward from SK 2129 through SK 2138. So every ḍāc from SK 2129 to the end of the chapter comes with karoti.
Rule. Ḍāc comes after an avyaktānukaraṇa: an imitation of an inarticulate sound. The imitation's half must have at least two vowels. The verb is kṛ, bhū or as. And iti must not follow (anitau).
Bhaṭṭoji explains dvyajavara: two vowels as the lower limit, and not fewer. In short, more than one vowel (anekāc). One of his commentators defines an inarticulate sound: one in which sounds like a are not told apart.
Example. Two vārttikas are needed first.
paṭat → doubled, since ḍāc is intended → paṭat paṭat
→ t + p → p, by pararūpa → paṭapaṭat
→ + ḍāc → the ṭi at deleted (SK 316) → paṭapaṭā
→ paṭapaṭākaroti.
One of Bhaṭṭoji's commentators gives the whole sequence.
Applies when / blocked by. Bhaṭṭoji tests each word of the sūtra.
dṛṣat karoti: dṛṣat is not an imitation, so no ḍāc.śrat karoti: śrat has one vowel.kharaṭakharaṭākaroti, trapaṭatrapaṭākaroti. Here each half has three vowels, and ḍāc still comes.paṭiti karoti. There pararūpa comes by SK 81 instead.Why "when ḍāc is intended", and not "when ḍāc follows"? Both of Bhaṭṭoji's commentators answer. Doubling would wait for ḍāc, and ḍāc would wait for the doubled half. Each would wait on the other. This fault is anyonyāśraya, mutual dependence.
Watch out. Bhaṭṭoji criticises the wording. The sūtra would be better as simply anekācaḥ "after [an imitation] of more than one vowel". Then the vārttika could read "when ḍāc follows", a parasaptamī. A parasaptamī is a seventh case that means "when X follows". One of his commentators gives the reason. Paṭat already has more than one vowel before doubling. So ḍāc could come first, and the doubling after it.
Our text of the Kaumudī reads pararupam, with a short u. The 1915 printed edition reads pararūpam. We follow the print. One of Bhaṭṭoji's commentators reads the counter-example as gharaṭagharaṭākaroti; the print has kharaṭa-, as we do.
Rule. After dvitīya "second", tṛtīya "third", śamba and bīja "seed", ḍāc comes with kṛ only. The sense is ploughing (kṛṣi). Bhaṭṭoji says "only in connection with kṛñ".
Example.
dvitīyaṃ karṣaṇaṃ karoti "he does a second ploughing" → dvitīya + ḍāc
→ the ṭi a deleted (SK 316) → dvitīyā → dvitīyākaroti.
Likewise tṛtīyākaroti "he does a third ploughing".
Śamba means "against the grain". anulomaṃ kṛṣṭaṃ kṣetraṃ punaḥ pratilomaṃ karṣati "he ploughs again, against the grain, a field ploughed with the grain" → śambākaroti.
bījena saha karṣati "he ploughs together with the seed" → bījākaroti.
Applies when / blocked by. No doubling here. Bhaṭṭoji says the bahulam "variously" of SK 2128's vārttika limits doubling before ḍāc to imitations of sound. So dvitīyākaroti, not a doubled form.
One of Bhaṭṭoji's commentators says why bhū and as drop out. The sūtra names kṛñ, so only kṛ remains. Kṛñ then runs to madrāt parivāpaṇe (SK 2138).
The other commentator asks how one can plough with the seed without ploughing the seed up. He answers that here bīja means a scratching of the soil together with sowing.
Watch out. Our text of the Kaumudī reads dvitāyādibhyaḥ. The 1915 printed edition reads dvitīyādibhyaḥ, "after dvitīya etc.". We follow the print.
Rule. After a numeral ending in guṇa "-fold", ḍāc comes, with kṛ, for ploughing. Bhaṭṭoji's gloss keeps "for ploughing" in force.
Example.
dviguṇa + ḍāc → the ṭi a deleted → dviguṇā.
dviguṇākaroti kṣetram. Bhaṭṭoji gives the meaning: "he does a double ploughing, with the field as its object".
Rule. After samaya "the appointed time", ḍāc comes with kṛ, for letting the time pass (yāpanā). Bhaṭṭoji says "for ploughing" has ceased here.
Example.
samaya + ḍāc → samayā → samayākaroti.
Bhaṭṭoji gives the meaning: "he lets the time pass" (kālaṃ yāpayati).
Applies when / blocked by. One of Bhaṭṭoji's commentators pictures the scene. Told "this must be done today", a man raises some obstacle and causes delay. The other reports Haradatta, a commentator on the Vṛtti. Haradatta says the word is used of one who says, day after day, "not today; tomorrow or the day after".
Rule. After sapattra "with feathers" and niṣpattra "feathers out", ḍāc comes with kṛ. The sense is ativyathana "severe wounding". One of Bhaṭṭoji's commentators reads the word as ativyadhana "piercing through".
Example.
sapattrākaroti mṛgam "he makes the deer sapattra". Bhaṭṭoji explains: he drives in the arrow together with its feathered end.
niṣpattrākaroti. The feathered end of the arrow passes out at the other side.
Applies when / blocked by. Why "severe wounding"? sapatraṃ niṣpatraṃ vā karoti bhūtalam "he makes the ground sapatra or niṣpatra". No ḍāc comes.
Both of Bhaṭṭoji's commentators say patra is the feathering on the arrow, by which it flies to its target. For the counter-example, one keeps feathers and the other takes patra as leaves. Scholars discuss this.
Watch out. Our text of the Kaumudī has three slips here, and we follow the 1915 printed edition over each.
The sūtra in the heading writes sapattra with tt but niṣpatra with one t. The print writes tt in both. The words are the same.
Rule. After niṣkula, ḍāc comes with kṛ, for niṣkoṣaṇa: taking out the inner parts. One of Bhaṭṭoji's commentators supplies ḍāc.
Example.
niṣkula + ḍāc → niṣkulā → niṣkulākaroti dāḍimam "he clears out the pomegranate".
Applies when / blocked by. Bhaṭṭoji says niṣkula is a bahuvrīhi. It means "that from which the kula has gone out". Here kula is the mass of inner parts.
One of Bhaṭṭoji's commentators adds a counter-example: niṣkulaṃ karoti śatrum "he makes the enemy kinless". The sense is not taking out inner parts, so no ḍāc.
Rule. After sukha "happy" and priya "dear", ḍāc comes with kṛ, for compliance (ānulomya). One of Bhaṭṭoji's commentators explains compliance: following the wishes of one to be pleased, such as a teacher.
Example.
sukha + ḍāc → sukhākaroti.
priya + ḍāc → priyākaroti gurum.
Bhaṭṭoji gives the meaning: "by agreeable conduct he gladdens [his teacher]".
Applies when / blocked by. The verb is kṛ here too. One of Bhaṭṭoji's commentators says kṛñ runs from SK 2129 through SK 2138, and this rule lies inside that range.
Rule. After duḥkha "pain", ḍāc comes with kṛ, for opposition (prātilomya). One of Bhaṭṭoji's commentators supplies ḍāc. He says opposition is acting against one who is to be served.
Example.
duḥkha + ḍāc → duḥkhā → duḥkhākaroti svāminam.
Bhaṭṭoji gives the meaning: "he torments [his] master".
Rule. After śūla "spit", ḍāc comes with kṛ, for cooking (pāka). One of Bhaṭṭoji's commentators supplies ḍāc.
Example.
śūla + ḍāc → śūlā → śūlākaroti māṃsam.
Bhaṭṭoji gives the meaning: "he cooks the meat on a spit". One of his commentators says karoti here means "cooks".
Applies when / blocked by. The other commentator adds a counter-example. In śūlaṃ karoti the śūla is a disease of the belly, colic. There is no cooking, so no ḍāc.
Rule. After satya "true", ḍāc comes with kṛ, when no oath (śapatha) is meant. One of Bhaṭṭoji's commentators supplies ḍāc.
Example.
satya + ḍāc → satyā → satyākaroti bhāṇḍaṃ vaṇik "the merchant satyākaroti the goods".
Bhaṭṭoji gives the meaning: he makes it true that [the goods] are to be bought.
Applies when / blocked by. In the sense of an oath there is no ḍāc: satyaṃ karoti vipraḥ "the brāhmaṇa swears".
Watch out. What exactly does the merchant do? One of Bhaṭṭoji's commentators says he states the true price, and no higher. The other says he binds the purchase with earnest-money. Scholars discuss this.
Rule. After madra, ḍāc comes with kṛ, for parivāpaṇa. Bhaṭṭoji glosses both words. Madra means "auspicious", and parivāpaṇa is shaving. One of Bhaṭṭoji's commentators supplies ḍāc.
Example.
madra + ḍāc → madrā → madrākaroti.
Bhaṭṭoji gives the meaning: "he consecrates [someone] by an auspicious shaving". Both of his commentators name the tonsure rite (caula).
Applies when / blocked by. A vārttika adds bhadra: bhadrākaroti. Bhaṭṭoji says the meaning is as before.
Why "shaving"? madraṃ karoti, bhadraṃ karoti "he does what is good". One of Bhaṭṭoji's commentators says no shaving is meant there, so no ḍāc.
This is the last sūtra with kṛ. One of Bhaṭṭoji's commentators, at SK 2129, named this rule as the end of the run of kṛñ.
With this sūtra the taddhita section ends. Both of Bhaṭṭoji's commentators end their accounts of the taddhitas here. Pāṇini's next sūtra, samāsāntāḥ, heads the affixes added at the end of compounds. Bhaṭṭoji teaches it with the compounds (SK 676). The next chapter, 43, is on doubling (dvirukta). It opens with sarvasya dve (SK 2139).
Watch out. Our text of the Kaumudī lacks a closing line here. The 1915 printed edition reads iti taddhiteṣu tadrājaḥ, "thus, among the taddhitas, the tadrāja section". The print headed the section with nearly the same words at SK 2051. We follow the print.
dravyam (SK 2059).nūtnam, nūtanam, praṇam, nāmadheyam.devatā (SK 2095).devatā, deva + tal (SK 2095).lohitikā / lohinikā.mṛttikā (SK 2107).mṛtsā, mṛtsnā (SK 2108).prājñā "she who has wisdom" (SK 2106).| SK | What it does | Example |
|---|---|---|
| 2051 | kan "like", when the thing compared is a likeness | aśvakaḥ |
| 2052 | kan "like", when base and affix are a name | aśvakaḥ, uṣṭrakaḥ |
| 2053 | lup of the kan of SK 2052 when a man is meant | cañcā, vardhrikā |
| 2054 | lup of the kan of SK 2051 for a likeness serving a livelihood, not for sale | vāsudevaḥ; but hastikān vikrīṇīte |
| 2055 | lup of that kan after devapatha etc. | devapathaḥ, haṃsapathaḥ |
| 2056 | ḍhañ "like" after vasti | vāsteyam, vāsteyī |
| 2057 | ḍha "like" after śilā; by yogavibhāga ḍhañ too | śileyam, śaileyam |
| 2058 | ya "like" after śākhā etc. | śākhyaḥ, jaghanyaḥ |
| 2059 | dravya by nipātana, for "worthy" | dravyam ayaṃ brāhmaṇaḥ |
| 2060 | cha "like" after kuśāgra | kuśāgrīyā buddhiḥ |
| 2061 | cha after a compound of likeness | kākatālīyaḥ, ajākṛpāṇīyaḥ |
| 2062 | aṇ "like" after śarkarā etc. | śārkaram |
| 2063 | ṭhak "like" after aṅgulī etc. | āṅgulikaḥ, bhārujikaḥ |
| 2064 | ṭhac optionally after ekaśālā; else ṭhak | ekaśālikaḥ / aikaśālikaḥ |
| 2065 | īkak "like" after karka, lohita | kārkīkaḥ, lauhitīkaḥ |
| 2066 | ñya after a word naming a pūga; "like" lapses; SK 1100 cited | lauhitadhvajyaḥ; kāpotapākyaḥ, kauñjāyanyaḥ |
| 2067 | ñyaṭ after an armed band of the Vāhīkas, not brāhmaṇa or rājanya | kṣaudrakyaḥ, kṣaudrakī; not gopālakāḥ |
| 2068 | ṭeṇyaṇ after vṛka as a band's name | vārkeṇyaḥ |
| 2069 | cha after dāmani etc. and the trigartaṣaṣṭhas | dāmanīyaḥ, plural dāmanayaḥ |
| 2070 | aṇ after parśu etc., añ after yaudheya etc. | pārśavaḥ, parśavaḥ; yaudheyaḥ, yaudheyāḥ |
| 2071 | yañ after the aṇ-final descendant-words of seven names | ābhijityaḥ, śālāvatyaḥ |
| 2072 | the affixes from SK 2066 are named tadrāja: luk in the plural, not in the feminine | lohitadhvajāḥ, kṣudrakāḥ |
| 2073 | vun and deletion of the final after numeral + pāda/śata, distributive | dvipadikām, dviśatikām |
| 2074 | the same for a fine or a gift | dvipadikāṃ daṇḍitaḥ |
| 2075 | kan after sthūla etc. for a mode; vārttika, gaṇasūtra | sthūlakaḥ, bṛhatkaḥ, surakaḥ |
| 2076 | kan after a kta-word, action not complete | chinnakam |
| 2077 | no kan with sāmi or a synonym; reveals a wholly svārthika kan | sāmikṛtam; bahutarakam |
| 2078 | kan after bṛhatī, a covering; obligatory | bṛhatikā |
| 2079 | kha, obligatory, after aṣaḍakṣa etc.; the Bhāṣya's list of obligatory svārthikas | aṣaḍakṣīṇaḥ, alaṃkarmīṇaḥ, īśvarādhīnaḥ |
| 2080 | kha optionally after an añc-word, not a feminine direction | prācīnam; not prācī dik |
| 2081 | cha after a jāti-final word denoting a bandhu | brāhmaṇajātīyaḥ |
| 2082 | cha optionally after a sthāna-final word meaning "equal" | pitṛsthānīyaḥ / pitṛsthānaḥ |
| 2083 | ṭhak after anugādin | ānugādikaḥ |
| 2084 | aṇ after visārin for a fish; ṭi kept | vaisāriṇaḥ |
| 2085 | kṛtvasuc after a number-word counting an action's occurrences | pañcakṛtvaḥ |
| 2086 | suc after dvi, tri, catur | dviḥ, triḥ, catuḥ |
| 2087 | sakṛt for eka, with suc | sakṛt |
| 2088 | dhā optionally after bahu, for times close together | bahudhā / bahukṛtvaḥ |
| 2089 | mayaṭ after a first-case word, "stated as presented in abundance" (action or locus) | annamayam; annamayo yajñaḥ |
| 2090 | for many things, the collective affixes, and mayaṭ | maudakikam, modakamayam |
| 2091 | ñya after ananta, āvasatha, itiha, bheṣaja, in their own sense | ānantyam, aitihyam, bhaiṣajyam |
| 2092 | yat after a devatā-final word, "for that" | agnidevatyam, pitṛdevatyam |
| 2093 | yat after pāda, argha, "for that"; vārttikas nū + tnap/tanap/kha, pra + na, dheya, añ | pādyam, arghyam; nūtanam, pratnam, nāmadheyam, āgnīdhrī |
| 2094 | ñya after atithi, "for that" | ātithyam |
| 2095 | tal after deva, wholly svārthika | devatā |
| 2096 | ka after avi, purely svārthika | avikaḥ |
| 2097 | kan after yāva etc. | yāvakaḥ, maṇikaḥ |
| 2098 | kan after lohita used of a jewel | lohitakaḥ |
| 2099 | kan after lohita for an impermanent colour; vārttika: blocking of the feminine optional | lohitakaḥ kopena; lohitikā / lohinikā |
| 2100 | kan after lohita for a dyed thing | lohitikā / lohinikā śāṭī |
| 2101 | kan after kāla, passing colour or dye | kālakaṃ mukham, kālikā śāṭī |
| 2102 | ṭhak after vinaya etc.; gaṇasūtra upāya with shortening | vainayikaḥ, sāmayikaḥ, aupayikaḥ |
| 2103 | ṭhak after vāc for a message | vācikam |
| 2104 | aṇ after karman connected with a message | kārmaṇam |
| 2105 | aṇ after oṣadhi not naming a species | auṣadham |
| 2106 | aṇ after prajña etc. | prājñaḥ, prājñī, daivataḥ, bāndhavaḥ |
| 2107 | tikan after mṛd | mṛttikā |
| 2108 | sa, sna after mṛd for praise; nitya; apavāda of rūpap | mṛtsā, mṛtsnā |
| 2109 | śas optionally after words for much/few denoting a kāraka; vārttika: only for the (in)auspicious | bahuśaḥ, alpaśaḥ |
| 2110 | śas optionally after a number or quantity-word, for distribution | dviśaḥ, māṣaśaḥ, prasthaśaḥ |
| 2111 | tasi after a fifth case with prati; vārttika: after ādi etc. | kṛṣṇataḥ prati; āditaḥ, svarataḥ |
| 2112 | tasi after an apādāna fifth, not with hīya, ruh | grāmataḥ |
| 2113 | tasi optionally after a non-agent third, for surpassing, not wavering, censure | cāritrataḥ, vṛttataḥ |
| 2114 | the same with hīyamāna, pāpa, without censure | vṛttataḥ |
| 2115 | tasi optionally after a sixth, taking a side | arjunataḥ |
| 2116 | tasi optionally after a sixth-case disease-word, for its removal | pravāhikātaḥ kuru |
| 2117 | cvi with kṛ, bhū, as after a word for the one becoming; vārttika: what it was not | kṛṣṇīkaroti |
| 2118 | ī for a/ā before cvi; cvi-words indeclinable; not for an indeclinable | kṛṣṇīkaroti, brahmībhavati; divābhūtā |
| 2119 | y of a descendant affix deleted before kya or cvi | gārgībhavati |
| 2120 | lengthening before cvi | śucībhavati, paṭūsyāt, mātrīkaroti |
| 2121 | final of arus, manas etc. deleted, with cvi | arūkaroti, unmanīsyāt |
| 2122 | sāti optionally for totality, in cvi's domain | agnisād bhavati |
| 2123 | no ṣ for the s of sāt or a pada-initial s | agnisāt, dadhi siñcati |
| 2124 | sāti optionally for pervasion, with sampad and kṛ, bhū, as | jalasāt saṃpadyate, jalībhavati |
| 2125 | sāti for "dependent on that" | rājasāt karoti |
| 2126 | trā and sāti for a thing to be given | vipratrā karoti, viprasāt karoti |
| 2127 | trā variously after deva etc. in the 2nd or 7th case | devatrā vande, bahutrā |
| 2128 | ḍāc after a sound-imitation with a two-vowel half; doubling, pararūpa | paṭapaṭākaroti |
| 2129 | ḍāc with kṛ after dvitīya, tṛtīya, śamba, bīja, for ploughing | dvitīyākaroti, bījākaroti |
| 2130 | ḍāc with kṛ after a guṇa-final numeral, for ploughing | dviguṇākaroti kṣetram |
| 2131 | ḍāc with kṛ after samaya, letting time pass | samayākaroti |
| 2132 | ḍāc with kṛ after sapattra, niṣpattra, for severe wounding | sapattrākaroti mṛgam |
| 2133 | ḍāc with kṛ after niṣkula, taking out the insides | niṣkulākaroti dāḍimam |
| 2134 | ḍāc with kṛ after sukha, priya, for compliance | priyākaroti gurum |
| 2135 | ḍāc with kṛ after duḥkha, for opposition | duḥkhākaroti svāminam |
| 2136 | ḍāc with kṛ after śūla, for cooking | śūlākaroti māṃsam |
| 2137 | ḍāc with kṛ after satya, not an oath | satyākaroti bhāṇḍam |
| 2138 | ḍāc with kṛ after madra (vārttika: bhadra), for shaving | madrākaroti, bhadrākaroti |