After this chapter you will be able to derive the taddhita words for "a product of that" and "a part of that". aśmano vikāraḥ "a product of stone" is one word, āśmaḥ. mayūrasyāvayavaḥ "a part of a peacock" is māyūraḥ. A thing made of bilva wood is bailvam, and one made of mango wood is āmramayam. You will know which of three rival affixes a base takes: the default aṇ, the accent-driven añ, or mayaṭ, heard as -maya. You will see why the pitch of a base's first vowel decides between aṇ and añ. You will learn the special affixes for single words: auṣṭrakaḥ from the camel, gavyam from the cow, dravyam from the tree. Last, you will name fruits, flowers and roots, where the affix disappears: āmalakam, jambūḥ, harītakyaḥ, mallikā. There are 34 sūtras, and every one is worked on a concrete word.
gavyam).vaiyāsakiḥ).aśmano vikāraḥ → āśmaḥ; mayūrasyāvayavaḥ → māyūraḥTwo sūtras set up the senses of the whole chapter. A vikāra is a product: the original material after it has changed into another state. An avayava is a part. Every word can take the sense "product"; only words for living beings, herbs and trees also take "part". On the running example, a product of stone is āśmaḥ, and a part or a product of a peacock is māyūraḥ.
Three affixes compete through the chapter. The default is aṇ. The rival añ comes after a base ending in u (SK 1519), after a base whose first vowel is anudātta (SK 1520), and after words for living beings (SK 1532). The second rival, mayaṭ, comes optionally after any base (SK 1523) and obligatorily after some (SK 1524). Most other rules are apavādas that set one of these aside for particular words.
Rule. After a word in the sixth case (the genitive), the affixes aṇ etc. come in the sense "a vikāra of that", a product of that. "Aṇ etc." means the default aṇ and the special affixes that the following rules name.
A vārttika adds a stem change: aśmano vikāre ṭilopo vaktavyaḥ. Before an affix in this sense, aśman "stone" loses its ṭi, here an.
Example.
aśmano vikāraḥ "a product of stone" → aśman + aṇ
→ the vārttika deletes the ṭi an → aśm + a
→ aṇ is ṇit, "marked with ṇ", so first-vowel vṛddhi (SK 1075): a → ā → āśmaḥ.
bhasmano vikāraḥ "a product of ash" → bhasman + aṇ
→ no vārttika here, so the stem keeps its ṭi by an (SK 1155) → vṛddhi → bhāsmanaḥ.
mṛttikāyā vikāraḥ "a product of clay" → mṛttikā + aṇ
→ vṛddhi ṛ → ār; the final ā is deleted by yasyeti ca (SK 311) → mārtikaḥ (also written mārttikaḥ).
Applies when / blocked by. The vārttika is an apavāda of SK 1155. Without it, aśman would keep its ṭi, as bhasman does.
Why do all three bases take plain aṇ? Bhaṭṭoji's commentators explain it by accent. Aśman and bhasman are formed with the affix manin, and mṛttikā with tikan. By the accent rule for affixes marked with n, such a word is ādyudātta: its first vowel is udātta, high. The añ of SK 1520 needs a low first vowel, so it does not reach them.
Watch out. This chapter begins where the śaiṣika senses of chapter 29 end. Bhaṭṭoji's commentators note that the śaiṣika affixes no longer operate here. So hala and sīra, which take ṭhak in chapter 29 (SK 1504), take plain aṇ in the product sense: hālaḥ, sairaḥ.
Rule. After words for living beings (prāṇin), herbs (oṣadhi) and trees (vṛkṣa), the same affixes come also in the sense "an avayava of that", a part of that. The word ca "also" keeps the sense vikāra. So these words take both senses; all other words take the product sense only.
Example.
mayūrasyāvayavo vikāro vā "a part or a product of a peacock" → mayūra names a living being → mayūra + añ (the añ for living beings, SK 1532)
→ vṛddhi a → ā; the final a deleted (SK 311) → māyūraḥ.
mūrvā (a herb) → its stalk (a part) or its ash (a product) → mūrvā + aṇ
→ vṛddhi ū → au; the final ā deleted → maurvam.
pippala (a tree, the aśvattha) → pippala + aṇ → vṛddhi i → ai → paippalam.
Applies when / blocked by. Mūrvā is ādyudātta by a phiṭ rule, one of the rules that fix the accent of noun stems not formed by an affix. So the añ of SK 1520 does not reach it, and it takes the default aṇ. Pippala takes aṇ too; its accent is a point scholars discuss.
Watch out. A word like aśman "stone" names no living being, herb or tree. So āśmaḥ means only "a product of stone", never "a part of stone".
bilvasya vikāro 'vayavo vā → bailvamThese three rules give the plain aṇ to bases that would otherwise take añ or mayaṭ. The first names a list, the second a sound-shape, the third two single words. On the running example, a part or a product of the bilva tree is bailvam.
Rule. After bilva and the other words of its list, aṇ comes. It comes in the product sense, and, for words naming living beings, herbs or trees, also in the part sense. Bhaṭṭoji's commentators give the list as bilva, vrīhi, kāṇḍa, mudga, masūra, godhūma, ikṣu, veṇu, gavedhuka, kārpāsī, pāṭalī, karkandhū, kaṭī.
Example.
bilvasyāvayavo vikāro vā "a part or a product of the bilva" → bilva + aṇ
→ vṛddhi i → ai → bailva + a
→ the final a deleted (SK 311) → bailvam.
Applies when / blocked by. Bhaṭṭoji's commentators call this rule an apavāda of añ and of mayaṭ. One of them is more precise. Gavedhuka has k as its penultimate, so it would get aṇ anyway by SK 1517; it is listed here only to block mayaṭ. The other words are listed to block the añ of SK 1520.
The optional mayaṭ of SK 1523 still reaches this list: bilvamayam stands beside bailvam. For vrīhi, SK 1528 sets this aṇ aside when the product is a sacrificial cake.
Watch out. bailvam looks exactly as an añ-form would. The difference is the marker: aṇ is not ñit, "marked with ñ". It matters in SK 1533, where only a ñit affix lets a further añ follow. So a product of bailva is bailvamayam.
Rule. Aṇ also comes after a base whose upadhā, its penultimate sound, is k. It is an apavāda of añ.
Example.
tarku (a kind of tree) → tarku + aṇ
→ vṛddhi a → ā → tārku + a
→ the final u takes guṇa by or guṇaḥ (SK 847): u → o, and o → av before the vowel → tārkavam.
tittiḍīka → tittiḍīka + aṇ → vṛddhi i → ai; the final a deleted → taittiḍīkam.
Applies when / blocked by. Each example sets aside a different añ, as Bhaṭṭoji's commentators point out. Tarku ends in u, so orañ (SK 1519) would give añ. Tittiḍīka is madhyodātta: a middle vowel is high, so its first vowel is low, and SK 1520 would give añ.
Rule. After trapu "tin" and jatu "lac", aṇ comes, carried forward from SK 1516. The two bases also take the augment ṣuk. An augment (āgama) is a sound added to a form without replacing anything. The u and k of ṣuk are markers, so it adds only ṣ. The forms show where it stands: after trapu and jatu, at the end of the base.
Example.
trapu "tin" → a product of tin → trapu + ṣuk + aṇ
→ the augment is added → trapuṣ + a
→ vṛddhi a → ā → trāpuṣam.
Likewise jatu "lac" → jatuṣ + a → jātuṣam.
Applies when / blocked by. Both bases end in u, so this aṇ is an apavāda of orañ (SK 1519). Tin and lac are not living beings, herbs or trees. So the affix comes in the product sense only.
Watch out. The u of trapu does not become av here, as it does in tārkavam. In trāpuṣam the augment ṣ ends the base, and the affix follows the ṣ.
devadāror avayavo vikāro vā → daivadāravamNow the first rival of aṇ. Añ is heard as a and is ñit, so it too brings first-vowel vṛddhi. The written form is therefore mostly the same as with aṇ. The two affixes differ in the marker ñ, which a later rule looks for (SK 1533). Which of the two a base takes often depends on SK 1520, which looks at the pitch of the base's first vowel. That is why Bhaṭṭoji's commentators ask of each base whether its first vowel is high or low. On the running example, a part or a product of the devadāru tree is daivadāravam.
Rule. After a base ending in u, short or long, añ comes. As before, it comes in the product sense, and for words for living beings, herbs and trees also in the part sense.
Example.
devadāror avayavo vikāro vā "a part or a product of the devadāru" → devadāru + añ
→ vṛddhi e → ai → daivadāru + a
→ the final u takes guṇa (SK 847): u → o → av → daivadāravam.
bhadradāru → vṛddhi a → ā → bhādradāravam.
Applies when / blocked by. Why is this rule needed beside SK 1520? Bhaṭṭoji's commentators answer through the examples. Devadāru and bhadradāru are ādyudātta by a phiṭ rule: their first vowel is high. So SK 1520 would not give them añ; only this rule does. Bhaṭṭoji chose examples that SK 1520 does not reach.
Several later rules set this añ aside for particular u-final words: SK 1517 (tarku), SK 1518 (trapu, jatu), SK 1539 (dru) and SK 1542 (śigru, karkandhū).
Rule. Añ also comes after a base whose first vowel is anudātta, low. Such a base is called anudāttādi, "low-initial".
Example.
dadhittha (another name of the kapittha, the wood-apple tree) → the base is antodātta: its last vowel is high, so the vowels before it, the first included, are low
→ dadhittha + añ → vṛddhi a → ā; the final a deleted → dādhittham.
kapittha → in the same way → kāpittham.
The printed text reads dāghittham here, with gh. Bhaṭṭoji himself writes dādhittham at SK 1533, and that is the intended form.
Applies when / blocked by. Why these two bases are antodātta is a point scholars discuss. Several rules set this añ aside and give aṇ instead: SK 1516, 1517, 1530 and 1542.
Watch out. "Antodātta" and "anudāttādi" are not opposites. A word whose last vowel is high has a low first vowel, so it is anudāttādi and takes this añ.
Rule. After palāśa and the other words of its list, añ comes optionally. The other option is the default aṇ.
Bhaṭṭoji's commentators call this an ubhayatra-vibhāṣā, an option "in both places". Some members, palāśa, khadira, śiṃśapā and syandana, are anudāttādi, so SK 1520 had made añ obligatory for them. Others, such as karīra, śirīṣa, vikaṅkata and yavāsa, could not take añ at all. This rule makes añ optional for both groups.
Example.
palāśa → antodātta by the phiṭ rule for the ghṛta list, so anudāttādi → palāśa + añ (or aṇ)
→ vṛddhi a → ā; the final a deleted → pālāśam.
karīra (a shrub) → formed from the root kṝ with the affix īran, so ādyudātta by the n-marker accent → karīra + añ (or aṇ) → kārīram.
Watch out. The optional añ and the alternative aṇ give the same written form, pālāśam either way. The two options differ only in the marker ñ, which SK 1533 looks for.
Rule. After śamī, the affix ṣlañ comes, in both senses. Its ṣ and ñ are markers, so it is heard as la. Being ñit, it brings first-vowel vṛddhi.
Example.
śamī + ṣlañ → śamī + la
→ vṛddhi a → ā → śāmīla → śāmīlaṃ bhasma "ash of śamī wood".
In the feminine: the affix has ṣ as its marker, so ṅīṣ follows (SK 498)
→ śāmīla + ī → the final a deleted (SK 311) → śāmīlī sruk "a ladle of śamī wood".
Applies when / blocked by. One authority, Mādhava, reads the affix as ṭit, marked with ṭ, so that the feminine takes ṅīp instead (SK 470). The written form śāmīlī is the same either way. Which marker is right is a point scholars discuss (see SK 1533).
aśman + mayaṭ → aśmamayam (optionally āśmanam)The second rival of aṇ is mayaṭ. Its ṭ is a marker, so it is heard as maya. It has no ñ, ṇ or k, so it brings no vṛddhi. It comes optionally after any base (SK 1523) and obligatorily after some (SK 1524). Five single words then get their own treatment. On the running example, a thing made of stone is aśmamayam, or, in the other option, āśmanam.
Rule. In the ordinary language (bhāṣā), as opposed to the Veda, mayaṭ comes optionally in these two senses, product and part. It comes after any base whatever (prakṛtimātrāt). It does not come when the thing meant is food or clothing.
The words etayoḥ "in these two" seem unnecessary: the heading already supplies the two senses. Bhaṭṭoji's commentators say they are stated so that mayaṭ reaches even the bases for which special affixes are taught. That is why Bhaṭṭoji writes "after any base whatever". So bilvamayam stands beside bailvam (SK 1516), and lohamayam beside the añ of SK 1532.
Example.
aśman + mayaṭ → aśman + maya
→ before mayaṭ the stem counts as a pada, a finished word, according to Bhaṭṭoji's commentators; so its final n is dropped
→ aśmamayam.
In the other option Bhaṭṭoji gives āśmanam. His commentators derive it from Aśman as a personal name: a son of Vasiṣṭha, also called Aśmaka.
Aśman + aṇ → the ṭi is kept by an (SK 1155) → vṛddhi → āśmanam.
The ṭi-deletion of SK 1514 is not expected here. It is for aśman in its usual sense, "stone".
Applies when / blocked by. Why "not food or clothing"? With food or clothing there is no mayaṭ, and the other affixes stand alone: maudgaḥ sūpaḥ "a soup of mudga beans", kārpāsam ācchādanam "a cotton garment". (Mudga and kārpāsī are both in the bilva list of SK 1516.)
Watch out. "Optional after any base" is the general mayaṭ. The next rule makes it obligatory for some bases, and later rules take it away from others.
Rule. After a vṛddha word, and after śara and the other words of its list, mayaṭ comes obligatorily. A vṛddha word is one whose first vowel is a vṛddhi vowel (chapter 29). This rule is an apavāda of the option in SK 1523. Bhaṭṭoji's commentators give the śara list as śara, darbha, mṛd, kuṭī, tṛṇa, soma, balvaja.
Example.
āmra "mango" → the first vowel is ā, so the word is vṛddha → āmra + mayaṭ → āmramayam "made of mango, or part of the mango".
śara "reed" → śara + mayaṭ → śaramayam "made of reed".
A vārttika adds ekāco nityam: after a monosyllabic word, a word of one vowel, mayaṭ is also obligatory.
tvac "skin" → tvac + maya → tvaṅmayam "made of skin".
vāc "speech" → vāṅmayam "consisting of speech".
ap "water" → ammayam "consisting of water".
The final consonant of each base appears as a nasal, ṅ or m, before the m of maya.
Applies when / blocked by. Bhaṭṭoji's commentators say the vārttika is obtained by yoga-vibhāga, splitting the sūtra: the word nityam is read a second time on its own. They add that mṛd is in the śara list only for clarity. It has one vowel, so it would take obligatory mayaṭ anyway.
Bhaṭṭoji raises an objection. Ap has one vowel, so it must take mayaṭ. How then is the usage āpyam correct beside ammayam? His answer: āpyam is formed another way.
apām idam "this belongs to water" → aṇ by tasyedam (SK 1500) → āpam
→ then ṣyañ (heard as ya) in the sense of the base itself (svārthe), adding no new meaning → āpyam.
Several later rules set this obligatory mayaṭ aside: SK 1530 (tāla), SK 1533 (śāmīla) and SK 1539 (dru).
Rule. After go "cow" too, mayaṭ comes when its dung (purīṣa) is meant. Bhaṭṭoji names no affix; the form gomayam shows that it is mayaṭ, as in the two rules before.
Example.
goḥ purīṣam "the dung of a cow" → go + mayaṭ → go + maya → gomayam.
Applies when / blocked by. Whether dung counts as a product of the cow at all is a point scholars discuss. Why "dung"? For other products of the cow, SK 1538 gives yat: gavyaṃ payaḥ "cow's milk".
Rule. After piṣṭa "flour", mayaṭ comes, in the product sense.
Example.
piṣṭa + mayaṭ → piṣṭamayaṃ bhasma "ash from flour".
Applies when / blocked by. Bhaṭṭoji raises an objection. The liquor made from flour is called paiṣṭī surā, with aṇ, not mayaṭ. How is that correct? His answer: there the relation is meant in general, "this is its", not as "product". In that general sense aṇ comes by tasyedam (SK 1500).
piṣṭa + aṇ → vṛddhi i → ai; the final a deleted (SK 311) → paiṣṭa
→ in the feminine, the ordinary ṅīp after a stem in aṇ (SK 470) → paiṣṭa + ī
→ the final a deleted again (SK 311) → paiṣṭī.
Rule. "After piṣṭa" is carried forward (anuvṛtti). After piṣṭa, the affix kan comes when a name (saṃjñā) is meant: the name of one particular product of flour. The n of kan is a marker; it is heard as ka and brings no vṛddhi.
Example.
piṣṭasya vikāraviśeṣaḥ "a particular product of flour" → piṣṭa + kan → piṣṭakaḥ.
Bhaṭṭoji's commentators explain that the particular product is the apūpa, a cake. Bhaṭṭoji quotes a lexicon line, pūpo 'pūpaḥ piṣṭakaḥ syāt: pūpa, apūpa and piṣṭaka are names of one thing. His commentators identify the line as the Amarakośa's.
Watch out. A puroḍāśa, a sacrificial cake, is not a piṣṭaka, say Bhaṭṭoji's commentators, because it is not an apūpa. A rice puroḍāśa is instead called brīhimayaḥ, with mayaṭ, by the next rule.
Rule. After vrīhi "rice", mayaṭ comes necessarily when the product is a puroḍāśa. It is an apavāda of the aṇ that vrīhi takes as a member of the bilva list (SK 1516).
Example.
vrīhi + mayaṭ → vrīhi + maya → brīhimayaḥ puroḍāśaḥ "a puroḍāśa made of rice".
The printed text writes b here and v elsewhere; b and v alternate in the transmission.
Any other product of rice keeps the aṇ of SK 1516:
vrīhi + aṇ → vṛddhi ī → ai; the final i deleted (SK 311) → vraiham.
Applies when / blocked by. Bhaṭṭoji's commentators ask why vrīhi is in the bilva list at all, if this rule gives it mayaṭ. Bhaṭṭoji's vraiham is the answer: for products other than a puroḍāśa, the aṇ holds.
Rule. After tila "sesame" and yava "barley", mayaṭ comes necessarily when no name is meant. "Necessarily mayaṭ" is carried forward from SK 1524.
Example.
tila + mayaṭ → tilamayam.
yava + mayaṭ → yavamayam.
Applies when / blocked by. When a name is meant, there is no mayaṭ: Bhaṭṭoji gives tailam and yāvakaḥ. His commentators derive yāvakaḥ in two steps:
yava + aṇ (product) → vṛddhi a → ā → yāva
→ then kan in the sense of the base itself, by yāvādibhyaḥ, the rule adding kan after yāva and the other words of its list → yāvakaḥ.
They give no steps for tailam.
rajatasya vikāraḥ → rājatam (not rajatamayam)These four rules give aṇ or añ where mayaṭ, or a different añ, would otherwise come. On the running example, a product of silver is rājatam, with añ, and not rajatamayam.
Rule. After tāla and the other words of its list, aṇ comes. It is an apavāda of añ and of mayaṭ. A gaṇasūtra, a rule set inside the word-list itself, limits one member: tālād dhanuṣi, after tāla, only when the product is a bow.
Example.
tāla "palmyra" → vṛddha, so mayaṭ would come by SK 1524 → for a bow, tāla + aṇ instead
→ vṛddhi (the ā stays ā); the final a deleted → tālaṃ dhanuḥ "a bow of palmyra wood".
Any other product of tāla keeps the mayaṭ: tālamayam.
indrāyudha → antodātta by the accent of compounds, so anudāttādi → the añ of SK 1520 would come → aṇ instead
→ vṛddhi i → ai → aindrāyudham.
Applies when / blocked by. Bhaṭṭoji's commentators show how "an apavāda of añ and of mayaṭ" divides over the list. Tāla and śyāmāka are vṛddha, so this aṇ sets aside their mayaṭ. Most other members are anudāttādi, so it sets aside their añ. Bārhiṇa, a product of the peacock (barhin) formed with the añ of SK 1532, would take a further añ by SK 1533; that too is set aside.
Rule. After words for gold (jātarūpa), aṇ comes when the product is a measure (parimāṇa), such as a niṣka. The sūtra says "words" in the plural, so the synonyms of jātarūpa are meant, not the word jātarūpa alone.
Example.
hāṭaka (a word for gold) → a product of hāṭaka having the measure of a niṣka → hāṭaka + aṇ
→ vṛddhi (the ā stays); the final a deleted → hāṭakaḥ "a niṣka of hāṭaka gold".
Likewise tapanīya → tāpanīyaḥ, and suvarṇa → vṛddhi u → au → sauvarṇo niṣkaḥ.
Applies when / blocked by. Why "a measure"? When no measure is meant, mayaṭ comes: hāṭakamayī yaṣṭiḥ "a staff made of gold". Which rival affix this aṇ sets aside for each base, mayaṭ or añ, is a point scholars discuss.
Rule. After words for living beings, and after rajata "silver" and the other words of its list, añ comes. Bhaṭṭoji's commentators name rajata, sīsa, loha and udumbara among the list.
Example.
śuka "parrot" → śuka + añ → vṛddhi u → au; the final a deleted → śaukam "a part or a product of a parrot".
baka "crane" → bākam.
rajatasya vikāraḥ "a product of silver" → rajata + añ → vṛddhi → rājatam.
Applies when / blocked by. Bhaṭṭoji's commentators explain the choice of examples. Śuka and baka are ādyudātta by the phiṭ rule prāṇināṃ kupūrvam. So SK 1520 would not give them añ; only this rule does.
Rajata is different. It is anudāttādi, so SK 1520 already gives it añ. It is listed again to block mayaṭ. Without this rule, mayaṭ, taught later than SK 1520, would win as the later rule (para). One commentator adds that this rule also blocks mayaṭ after vṛddha words for living beings.
The camel and the doe get their own affixes instead of this añ (SK 1535, 1537).
Rule. Some words already end in a product or part affix that is ñit. After such a word, a further añ comes, in the same two senses. It is an apavāda of mayaṭ.
Example.
śāmīla (from śamī + ṣlañ, SK 1522) → vṛddha, so mayaṭ would come by SK 1524 → śāmīla + añ instead
→ vṛddhi (the ā stays); the final a deleted → śāmīlam "a product or a part of a śāmīla thing".
dādhittha (from dadhittha + añ, SK 1520) → dādhittha + añ → dādhittham. Likewise kāpittham.
Applies when / blocked by. Why "ñit"? Bailva has the aṇ of SK 1516, which is not ñit. So this rule does not apply, and mayaṭ comes: bailvamayam.
The Bhāṣya, the great commentary on Pāṇini, rejects this sūtra. It holds that a product of a product takes no second affix of this kind. What that means for the marker of ṣlañ (SK 1522) is a point scholars discuss.
Watch out. The new form can look exactly like its base: śāmīlam from śāmīla. The vṛddhi is already there, the final a is deleted (SK 311), and the affix a follows. Only the sense has changed.
uṣṭrasya vikāro 'vayavo vā → auṣṭrakaḥThis group gives affixes other than aṇ, añ and mayaṭ. After a word for a measure, the affixes of the sense "bought with" come. After the camel, flax, wool and the doe, single affixes come. On the running example, a part or a product of a camel is auṣṭrakaḥ.
Rule. After a word for a measure, the affixes taught in the sense "bought with that" (krīta) come also in the product sense. They come with the same conditions (upādhi) as there, such as a condition on the base. This is an atideśa, a rule that makes one thing behave like another. It is an apavāda of aṇ and the rest.
Example.
niṣkeṇa krītam "bought with a niṣka" → niṣka + ṭhak (asamāse niṣkādibhyaḥ, A 5.1.20, SK 1682)
→ ṭh → ik; the k brings vṛddhi i → ai; the final a deleted → naiṣkikam.
So also niṣkasya vikāraḥ "a product of a niṣka" → naiṣkikaḥ.
śatasya vikāraḥ "a product of a hundred" → yat or ṭhan (śatāc ca ṭhanyatāv aśate, A 5.1.21, SK 1686)
→ śata + ya → the final a deleted → śatyaḥ
→ or śata + ika → śatikaḥ. Neither affix brings vṛddhi.
Applies when / blocked by. Bhaṭṭoji's commentators name these two "bought" rules for the examples. Bhaṭṭoji says the "bought" affixes are those taught from prāg vahateṣ ṭhak (SK 1548) onward. The "bought" rules actually stand under a later heading, prāg vateṣ ṭhañ (SK 1680). So the printed reference is probably a slip.
Rule. After uṣṭra "camel", vuñ comes, in both senses. It is an apavāda of the añ for living beings (SK 1532).
Example.
uṣṭrasya vikāro 'vayavo vā "a part or a product of a camel" → uṣṭra + vuñ
→ vu → aka (SK 1247) → uṣṭra + aka
→ vṛddhi u → au; the final a deleted → auṣṭrakaḥ.
Rule. After umā "flax" and ūrṇā "wool", vuñ comes optionally.
Example.
umā + vuñ → umā + aka → vṛddhi u → au; the final ā deleted → aumakam.
ūrṇā + vuñ → vṛddhi ū → au → aurṇakam.
When vuñ does not come, Bhaṭṭoji says, aṇ and añ come yathākramam, "respectively": the first with the first, the second with the second.
umā + aṇ → aumam.
ūrṇā + añ → the printed text has orṇam. The expected form, with vṛddhi as in aurṇakam, is aurṇam.
Applies when / blocked by. Bhaṭṭoji's commentators explain "respectively". Umā is ādyudātta by a phiṭ rule, so SK 1520 does not reach it, and it takes the default aṇ. Ūrṇā is antodātta, hence anudāttādi, so it takes the añ of SK 1520.
Rule. After the feminine eṇī "doe", ḍhañ comes, in both senses. It is an apavāda of the añ for living beings (SK 1532).
Example.
eṇī + ḍhañ → ḍh → ey → eṇī + eya
→ the ī is deleted by yasyeti ca (SK 311); vṛddhi e → ai → aiṇeyam "a part or a product of a doe".
Watch out. The sūtra names the feminine eṇī on purpose, as Bhaṭṭoji shows. The masculine eṇa "antelope" does not take ḍhañ. It takes the añ of SK 1532: eṇa + añ → aiṇam.
dror vikāro 'vayavo vā → dravyamTwo short rules give yat, and a third gives vaya. Yat has t as its marker and is heard as ya; it brings no vṛddhi. On the running example, a part or a product of a tree (dru) is dravyam.
Rule. After go "cow" and payas "milk", yat comes.
Example.
gor vikāro 'vayavo vā "a product or a part of a cow" → go + yat → go + ya
→ o → av before the y of an affix, by vānto yi pratyaye (SK 63) → gavyam.
payaso vikāraḥ "a product of milk" → payas + ya → payasyam.
Applies when / blocked by. For go, yat was already available: the vārttika gor ajādiprasaṅge yat (under SK 1077) gives it wherever an affix beginning with a vowel would come. Bhaṭṭoji's commentators say the rule is taught again to block the optional mayaṭ of SK 1523. For the cow's dung, SK 1525 gives mayaṭ: gomayam.
Rule. After dru "tree", yat comes too. Bhaṭṭoji's commentators call it an apavāda of two rules. Dru has one vowel, so the obligatory mayaṭ of the vārttika under SK 1524 would come. And dru ends in u, so orañ (SK 1519) would come.
Example.
dror vikāro 'vayavo vā "a product or a part of a tree" → dru + yat → dru + ya
→ the u takes guṇa by or guṇaḥ (SK 847): u → o → dro + ya
→ o → av by vānto yi pratyaye (SK 63) → dravyam.
Watch out. The word dravya that means "substance" is a different word. Bhaṭṭoji's commentators derive it from the root dru: "that which is resorted to by qualities". The affix is yat, added by aco yat, "yat after a vowel", the rule that adds yat to a root ending in a vowel such as dru.
Rule. "After dru" is carried forward. After dru, the affix vaya comes when the product is a measure (māna). It is an apavāda of the yat of SK 1539.
Example.
dru + vaya → druvayam, a measure made of wood, such as a prastha.
Applies when / blocked by. Bhaṭṭoji quotes a line: yautavaṃ druvayaṃ pāyyam iti mānārthakaṃ trayam, "yautava, druvaya, pāyya: these three denote a measure". No source is named for it.
āmalakyāḥ phalam → āmalakamA fruit is both a part and a product of its tree. So the product or part affix would name it. But when the fruit is meant, that affix usually disappears. The kind of disappearance decides the gender. After luk, the word agrees with the thing it qualifies: "fruit" (phalam) is neuter, so the word is neuter. After lup, the word keeps the gender and number of its base (SK 1294). On the running example, the fruit of the āmalakī is āmalakam, a neuter.
Rule. When the fruit (phala) is meant, the affix of the product or part sense disappears by luk.
Example.
āmalakyāḥ phalam "the fruit of the āmalakī" → Bhaṭṭoji's commentators start from āmalakī + mayaṭ. (Āmalakī is vṛddha, SK 1524; its ī is the ṅīṣ of the gaura list, SK 498.)
→ luk of mayaṭ by this rule
→ luk of the feminine ṅīṣ by luk taddhitaluki (SK 1408) → āmalaka
→ neuter, agreeing with phalam → āmalakam.
Applies when / blocked by. One commentator adds a point. Yasyeti ca (SK 311) deletes a final a before ī as well as before a taddhita. If the removed feminine ī still counted, the final a of āmalaka would be deleted. But what luk removes is not sthānivat, "like the original". So the a stays.
This luk has several exceptions: SK 1542 keeps the affix, and SK 1545–1546 give lup instead.
Rule. After plakṣa and the other words of its list, aṇ comes, in both senses. When the fruit is meant, this aṇ does not disappear by luk. Bhaṭṭoji gives the reason as vidhāna-sāmarthya, "the force of the express teaching".
Example.
plakṣa + aṇ → vṛddhi a → ā; the final a deleted → plākṣam "the fruit of the plakṣa".
Applies when / blocked by. Bhaṭṭoji's commentators give the list as plakṣa, nyagrodha, aśvattha, iṅgudī, śigru, karkandhū, bṛhatī. The rule is an apavāda of añ. Śigru and karkandhū end in u and would take orañ (SK 1519). The rest would take the añ of SK 1520. The commentators quote a lexicon line that lists plākṣa, naiyagrodha, āśvattha and aiṅguda among the names of fruits.
Rule. This sūtra comes from the seventh book of the Aṣṭādhyāyī. Bhaṭṭoji reads it here because nyagrodha is in the plakṣa list. There it stands right after SK 1098 (na yvābhyām…). For nyagrodha standing alone (kevala), there is no vṛddhi, and the augment aic (ai or au) comes before the y.
Example.
nyagrodha + aṇ (SK 1542)
→ the vowel after the y takes no vṛddhi
→ ai is placed before the y → naiyagrodha + a
→ the final a deleted → naiyagrodham.
Applies when / blocked by. Bhaṭṭoji's commentators discuss the rule's status. If nyagrodha is derived, from nyak rohati, its y is pada-final. Then SK 1098 already applies, and this sūtra is a niyama, a restriction: the treatment is for nyagrodha alone and not in a compound. If the word is underived, its y is not pada-final. Then this sūtra is a vidhi, an injunction that teaches something new.
Why "standing alone"? In a compound the ordinary vṛddhi comes: nyāgrodhamūlāḥ śālayaḥ "rice grown at the root of a nyagrodha".
Rule. After jambū, aṇ comes optionally when the fruit is meant.
Example.
jambū + aṇ → vṛddhi a → ā → the final ū, like the u of tārkavam, appears as av → jāmbavam.
In the other option, jambū takes añ by orañ (SK 1519), and phale luk (SK 1541) removes it
→ jambū
→ neuter, agreeing with phalam, so the final vowel is shortened → jambu.
Applies when / blocked by. One commentator quotes the Amarakośa for the result: for the fruit, jambū (feminine), jambu and jāmbava. The feminine form comes by the next rule.
Rule. After jambū, the fruit-affix may also disappear by lup, optionally.
Example.
jambvāḥ phalam "the fruit of the jambū" → jambū + the fruit-affix → lup of that affix
→ by lupi yuktavat (SK 1294), gender and number of the base: feminine singular → jambūḥ.
Applies when / blocked by. Why teach lup, when luk already removes the affix? Under lup, the word keeps the base's gender and number, not those of the qualified word. So even "the fruits of the jambū" are jambūḥ, feminine singular.
Two vārttikas extend the lup.
Phalapākaśuṣām upasaṃkhyānam: lup is added (upasaṃkhyāna, an addition to a rule) after words for plants that wither when their fruit ripens. It is an apavāda of phale luk (SK 1541).
"The fruits of the plants called vrīhi", many plants → vrīhi + aṇ (SK 1516) → lup of aṇ
→ by yuktavat, the gender and number of the base: masculine plural → vrīhayaḥ "grains of rice".
Likewise mudgāḥ "mudga beans".
Puṣpamūleṣu bahulam: the lup comes bahulam, "variously", when flowers or roots are meant. Bahula means a rule applies in some cases, not in others, and sometimes gives something else. Three results follow.
mallikāyāḥ puṣpam "the flower of the mallikā" → mallikā; jātyāḥ puṣpam → jātī; vidāryā mūlam "the root of the vidārī" → vidārī.pāṭalāni puṣpāṇi "flowers of the pāṭalī"; sālvāni mūlāni "sālva roots". (The printed text reads pāṭālāni; the form from pāṭalī is pāṭalāni.)aśokam "the aśoka flower", karavīram "the karavīra flower". Under luk the word follows the neuter puṣpam.Watch out. Had lup come for aśoka and karavīra, yuktavat would make them masculine: aśokaḥ puṣpam. Bhaṭṭoji's commentators give this as the point of the luk.
Rule. After harītakī and the other words of its list, the fruit-affix disappears by lup. A vārttika, as Bhaṭṭoji's commentators identify it, limits the effect: harītakyādīnāṃ liṅgam eva prakṛtivat. For these words, only the gender follows the base; the number does not.
Example.
harītakyāḥ phalāni "the fruits of the harītakī"
→ harītakī is antodātta, so anudāttādi → harītakī + añ (SK 1520)
→ lup of añ by this rule, in place of the luk of SK 1541
→ gender from the base: feminine; number from phalāni: plural → harītakyaḥ.
Applies when / blocked by. One commentator gives the rule's full range. The lup removes the mayaṭ of SK 1524 after vṛddha words such as drākṣā, the añ after anudāttādi words, and the aṇ after the rest.
Watch out. Compare SK 1545. There the fruits of the jambū are jambūḥ, singular even for many fruits. Here the fruits of the harītakī are harītakyaḥ, plural, because only the gender follows the base.
kaṃsīyasya vikāraḥ → kāṃsyamThe chapter ends with one rule for two words that are themselves taddhita derivatives. Each means "good for" something. Their product-words drop the earlier affix. On the running example, a product of kaṃsīya is kāṃsyam.
Rule. After kaṃsīya and paraśavya, yañ and añ come respectively, in the product sense. At the same time, the affixes cha and yat inside those bases disappear by luk. Yañ is heard as ya and is ñit.
Example.
kaṃsāya hitam "good for bell-metal" → kaṃsa + cha (tasmai hitam, under prāk krītāc chaḥ, A 5.1.1) → ch → īy → kaṃsīyam.
kaṃsīyasya vikāraḥ "a product of that" → kaṃsīya + yañ
→ luk of cha → kaṃsa + ya
→ vṛddhi a → ā; the final a deleted by yasyeti ca → kāṃsyam.
paraśave hitam "good for an axe" → paraśu + yat (ugavādibhyo yat, A 5.1.2) → guṇa u → o → av → paraśavyam.
paraśavyasya vikāraḥ → paraśavya + añ
→ luk of yat → paraśu + a
→ the u again → o → av (SK 847); vṛddhi a → ā → pāraśavaḥ.
Applies when / blocked by. Whose luk is meant? Bhaṭṭoji's commentators argue by pariśeṣa, elimination. It cannot be luk of yañ and añ: then teaching them would be pointless. It cannot be luk of the bases, since luk removes only affixes. So it is luk of cha and yat, which sit inside the bases.
Paraśavya is anudāttādi, so it would take añ anyway (SK 1520). One commentator says the rule names it so that the luk of yat comes along with the añ.
Watch out. Do not look for a shorter route. One might delete the final a of paraśavya by yasyeti ca (SK 311). Then one might drop its y, which follows the consonant v, by halas taddhitasya (A 6.4.150, SK 472). That rule drops the y of a taddhita after a consonant, but only before ī. Bhaṭṭoji's commentators point out that the word īti, "before ī", is carried into that rule. Here no ī follows, so it cannot remove the y.
gavyam).trāpuṣam, jātuṣam (SK 1518).āpyam (SK 1524).| SK | Sūtra | What it does | Example |
|---|---|---|---|
| 1514 | tasya vikāraḥ | Aṇ etc. after a genitive: "a product of that"; vārttika drops the ṭi of aśman. | āśmaḥ, bhāsmanaḥ, mārtikaḥ |
| 1515 | avayave ca prāṇyoṣadhivṛkṣebhyaḥ | Also "a part of that", for words for living beings, herbs, trees. | māyūraḥ, maurvam, paippalam |
| 1516 | bilvādibhyo 'ṇ | Aṇ after the bilva list, setting aside añ and mayaṭ. | bailvam |
| 1517 | kopadhāc ca | Aṇ after a base with penultimate k. | tārkavam, taittiḍīkam |
| 1518 | trapujatunoḥ ṣuk | Aṇ after trapu, jatu, with the augment ṣ. | trāpuṣam, jātuṣam |
| 1519 | orañ | Añ after a base in u/ū. | daivadāravam |
| 1520 | anudāttādeś ca | Añ after a base whose first vowel is low. | dādhittham, kāpittham |
| 1521 | palāśādibhyo vā | Añ optionally after the palāśa list. | pālāśam, kārīram |
| 1522 | śamyāḥ ṣlañ | Ṣlañ (-la) after śamī; feminine in ṅīṣ. | śāmīlaṃ bhasma, śāmīlī sruk |
| 1523 | mayaḍ vaitayor bhāṣāyām abhakṣyācchādanayoḥ | Mayaṭ optionally after any base, not for food or clothing. | aśmamayam / āśmanam; maudgaḥ sūpaḥ |
| 1524 | nityaṃ vṛddhaśarādibhyaḥ | Mayaṭ obligatorily after vṛddha words, the śara list, and (vārttika) one-vowel words. | āmramayam, śaramayam, vāṅmayam |
| 1525 | goś ca purīṣe | Mayaṭ after go for dung. | gomayam |
| 1526 | piṣṭāc ca | Mayaṭ after piṣṭa. | piṣṭamayaṃ bhasma |
| 1527 | saṃjñāyāṃ kan | Kan after piṣṭa as a name. | piṣṭakaḥ |
| 1528 | vrīheḥ puroḍāśe | Mayaṭ after vrīhi for a puroḍāśa. | brīhimayaḥ puroḍāśaḥ; else vraiham |
| 1529 | asaṃjñāyāṃ tilayavābhyām | Mayaṭ after tila, yava when no name is meant. | tilamayam; names tailam, yāvakaḥ |
| 1530 | tālādibhyo 'ṇ | Aṇ after the tāla list; tāla only for a bow. | tālaṃ dhanuḥ, aindrāyudham |
| 1531 | jātarūpebhyaḥ parimāṇe | Aṇ after words for gold, for a measure. | hāṭakaḥ, sauvarṇo niṣkaḥ; else hāṭakamayī |
| 1532 | prāṇirajatādibhyo 'ñ | Añ after words for living beings and the rajata list. | śaukam, bākam, rājatam |
| 1533 | ñitaś ca tatpratyayāt | Añ after a word already in a ñit product/part affix. | śāmīlam, dādhittham; but bailvamayam |
| 1534 | krītavat parimāṇāt | After a measure, the "bought" affixes for "product". | naiṣkikaḥ, śatyaḥ, śatikaḥ |
| 1535 | uṣṭrād vuñ | Vuñ (-aka) after uṣṭra. | auṣṭrakaḥ |
| 1536 | umorṇayor vā | Vuñ optionally after umā, ūrṇā; else aṇ, añ. | aumakam / aumam; aurṇakam / aurṇam (printed orṇam) |
| 1537 | eṇyā ḍhañ | Ḍhañ (-eya) after eṇī. | aiṇeyam; but aiṇam |
| 1538 | gopayasor yat | Yat after go, payas. | gavyam, payasyam |
| 1539 | droś ca | Yat after dru. | dravyam |
| 1540 | māne vayaḥ | Vaya after dru for a measure. | druvayam |
| 1541 | phale luk | Luk of the affix when the fruit is meant. | āmalakam |
| 1542 | plakṣādibhyo 'ṇ | Aṇ after the plakṣa list, kept for the fruit. | plākṣam |
| 1543 | nyagrodhasya ca kevalasya | For nyagrodha alone: no vṛddhi, ai before the y. | naiyagrodham; but nyāgrodhamūlāḥ |
| 1544 | jambvā vā | Aṇ optionally after jambū for the fruit; else añ with luk. | jāmbavam / jambu |
| 1545 | lup ca | Lup optionally after jambū; vārttikas for withering plants, flowers, roots. | jambūḥ, vrīhayaḥ, mallikā, aśokam |
| 1546 | harītakyādibhyaś ca | Lup after the harītakī list; only gender follows the base. | harītakyaḥ |
| 1547 | kaṃsīyaparaśavyayor yañañau luk ca | Yañ, añ after kaṃsīya, paraśavya, with luk of cha, yat. | kāṃsyam, pāraśavaḥ |