26. Taddhitādhikāra: apatyādi-vikārānta-artha-sādhāraṇa-pratyayāḥ — The taddhita section opens: affixes shared by every sense from "offspring" to "product"

After this chapter you will be able to read the three headings that govern the whole taddhita section, and to derive the first taddhita forms. A taddhita is an affix added to a finished noun to make a new noun: aśvapater apatyam "the offspring of aśvapati" can also be said in one word, āśvapatam. You will know after which word such an affix goes, why the plain phrase stays correct beside it, and which affix comes by default (aṇ). You will know the rule that replaces the first vowel of the base with its vṛddhi (ā, ai, au): aśvapati → āśva-, diti → daitya. And you will know the small set of affixes that serve all the senses of this stretch — offspring, collection, "born in", "deity of", "existing in", "product of" and the rest — for special bases: daityaḥ, ādityaḥ, prājāpatyaḥ, yāmyaḥ, pārthivā / pārthivī, daivyam / daivam, bāhyaḥ / bāhīkaḥ, aśvatthāmaḥ, uḍulomāḥ, gavyam, autsaḥ, āgneyam, kāleyam. There are only seven sūtras, but they carry ten vārttikas (nine under SK 1077, one under SK 1078), and every form is worked on a concrete word.

Before you start

• This chapter opens a new part of the grammar. The compound chapters (17–25) are finished. From here on the Siddhānta-kaumudī teaches the taddhita affixes, and this is the first chapter of that part.
• What a taddhita is, in plain words: a secondary affix. A kṛt affix is added to a root; a taddhita is added to a noun that is already a finished word. The result is a new noun stem. From chapter 8 you know that a taddhita-ending form counts as a prātipadika, a noun stem (SK 179), so it takes case-endings like any other noun.
• A taddhita is always taught for a sense: "the offspring of X", "a collection of X", "born in X", "whose deity is X", "existing in X", "come from X", "belonging to X", "a product of X". The rule that states the sense usually names no affix. This chapter tells you which affix fills that gap.
• From chapter 1 you know it (a marker sound, written but not counted), upadeśa (the original statement of an affix), halantyam (SK 1: a final consonant of an upadeśa is it), ac (any vowel), vṛddhi (the name of ā, ai, au, SK 16), adhikāra (a heading whose words are read into the rules below it), anuvṛtti (carrying a word forward) and vārttika (a supplementary statement on a sūtra).
• From chapter 2 you know apavāda, an exception that sets aside a general rule. From chapters 2 and 8 you know that an affix is named by its markers: kit (marked with k), ṇit (marked with ṇ), ñit (marked with ñ).
• From chapter 3 you know how a substitute replaces a sound, and the ṭi of a word: its last vowel with any consonant after it.
• From chapter 8 you know cuṭū (SK 189: a palatal or retroflex at the start of an affix is a marker), luk (the complete disappearance of an affix, SK 260) and the bha stem (SK 231: the name a stem carries before an affix beginning with a vowel or with y). From chapter 10 you know yasyeti ca (SK 311): the final i- or a-vowel of a bha stem is deleted before a taddhita. Nearly every form below uses it.
• From chapter 15 you know the feminine affixes ṭāp (giving -ā) and ṅīp (giving -ī), and SK 470 (ṭiḍḍhāṇañ…), which gives ṅīp after a stem in aṇ or añ. From chapter 17 you know nas taddhite (SK 679): an n-final bha stem loses its ṭi before a taddhita. From chapters 19 and 23 you know the bahuvrīhi, the tadanta reading (tadanta-vidhi) and pṛṣodarādi, the list of forms accepted as they are found.

A. The three headings — after which word, optionally, and which affix by default — running example: aśvapater apatyam → āśvapatam (the phrase survives beside the derived word)

The taddhita section begins with headings. SK 1072 makes three words headings: the affix goes after the first-named of the connected words, and only optionally. SK 1073 makes aṇ a heading: the affix is aṇ unless a rule names another. Neither makes a form by itself. SK 1074 is an operative rule: it gives aṇ after aśvapati and a few similar words. On the running example, aśvapater apatyam "the offspring of aśvapati" gets the affix aṇ after aśvapati; the phrase itself stays good Sanskrit.

SK 1072 · समर्थानां प्रथमाद्वा · samarthānāṃ prathamād vā (A 4.1.82)

Rule. A taddhita affix is added, optionally, after the first among the samartha words. The three words samarthānām, prathamāt and vā are a heading (adhikāra). They are read into every taddhita rule as far as prāgdiśo vibhaktiḥ (A 5.3.1, SK 1947).

What the three words mean:

• samartha, "capable": a word that is fully settled, that is, one on which the sandhi operations have already been performed. Bhaṭṭoji's commentators explain why this gloss is needed. The other sense of "capable", being joined in one meaning, is already supplied for every word-rule by the paribhāṣā samarthaḥ padavidhiḥ (A 2.1.1). A paribhāṣā is a rule about how other rules are read. So the word would be idle in that sense. A word is capable of conveying its meaning only once its sandhi is done.
• prathamāt, "after the first": after the word named first in each sense-rule. In the rule tasyāpatyam, "his offspring", that word is tasya, "his" — the genitive.
• vā, "optionally": the affix need not be used, so the phrase with separate words, aśvapater apatyam, stays correct.

Bhaṭṭoji's commentators add two points. Samarthānām is a partitive (nirdhāraṇa) genitive: "among the samartha words". And the sūtra is a heading, not an operative rule (vidhi), because it names nothing to be added.

Example.

aśvapater apatyam "the offspring of aśvapati" → the sense-rule tasyāpatyam names the genitive word first → aśvapateḥ, a finished word, is the samartha → the affix goes after it: aśvapati + aṇ (SK 1073–1076) → āśvapatam.

By vā, the phrase aśvapater apatyam is also correct.

Watch out. The affix goes after the first-named word of the sense-rule, not after the word for the new sense. In aśvapater apatyam the affix follows aśvapati; apatya "offspring" is the meaning the affix expresses.

SK 1073 · प्राग्दीव्यतोऽण् · prāgdīvyato 'ṇ (A 4.1.83)

Rule. The affix aṇ is a heading. It governs every rule up to, but not including, tena dīvyati (A 4.4.2, SK 1550). The word dīvyat in the sūtra quotes part of that later rule's word dīvyati, "he plays"; scholars discuss exactly how the quotation marks the boundary.

Example.

aśvapater apatyam → the sense-rule tasyāpatyam names no affix → it asks "which affix?"; this heading answers aṇ → it asks "after which word?"; SK 1072 answers aśvapati → aśvapati + aṇ.

The ṇ of aṇ is final in the upadeśa, so it is a marker by SK 1: the affix attaches as a and is ṇit.

Applies when / blocked by. Aṇ is the general rule (utsarga). Where a later rule names a specific affix, such as iñ, that specific affix wins, as an apavāda of aṇ. The senses in the stretch before tena dīvyati are called prāgdīvyatīya, "those before dīvyat". Bhaṭṭoji's commentators explain the name itself as a taddhita form: "belonging to [the stretch] before dīvyat".

SK 1074 · अश्वपत्यादिभ्यश्च · aśvapatyādibhyaś ca (A 4.1.84)

Rule. After aśvapati and the rest of its list, aṇ occurs in the prāgdīvyatīya senses. The ca "and" draws aṇ in from SK 1073. Bhaṭṭoji's commentators give the list: aśvapati, gaṇapati, rāṣṭrapati, kulapati, gṛhapati, paśupati, kṣetrapati.

Example.

aśvapater apatyam → aśvapati is in the list → aśvapati + aṇ → the form is finished in Section B → āśvapatam.

gaṇapati → also in the list → gaṇapati + aṇ → gāṇapatam.

Applies when / blocked by. Why is this needed, if aṇ is the default anyway? Because SK 1077 will teach ṇya after every word whose last member is pati. All the words of this list end in pati. So this rule is an apavāda of that ṇya: without it, aśvapati would take ṇya.

B. The vṛddhi every such form needs — first-vowel vṛddhi before a ñit, ṇit or kit taddhita — running example: aśvapati + aṇ → āśvapatam (kit case: bahis + īkak → bāhīkaḥ, agni + ḍhak → āgneyam)

Most forms in this chapter take vṛddhi of the first vowel: its vṛddhi (ā, ai or au) replaces it. So a becomes ā, i or e becomes ai, u becomes au; the vṛddhi of ṛ comes with r attached, ār. It comes only when the affix is marked with ñ or ṇ (SK 1075) or with k (SK 1076). Two rules from the seventh book of the Aṣṭādhyāyī do this, and the Siddhānta-kaumudī reads them here. Where an affix has none of these markers (yat, rūpya, mayaṭ, the a after sthāman), there is no vṛddhi. On the running example, aṇ is ṇit, so the first a of aśvapati becomes ā: āśvapatam.

SK 1075 · तद्धितेष्वचामादेः · taddhiteṣv acām ādeḥ (A 7.2.117)

Rule. When a taddhita affix marked with ñ or ṇ follows, vṛddhi replaces the first of the vowels of the base. Bhaṭṭoji's commentators show where the missing words come from: "ñ or ṇ" and "vowel" are carried by anuvṛtti from aco ñṇiti (A 7.2.115), and "vṛddhi" from mṛjer vṛddhiḥ (A 7.2.114). Acām, "of the vowels", is a partitive genitive: the first vowel among them.

Example.

aśvapati + aṇ → aṇ is ṇit (its ṇ is a marker) → the first vowel of aśvapati is the a of aśva → it becomes ā → āśvapati + a

→ the final i is deleted before the taddhita by yasyeti ca (SK 311) → āśvapata

→ neuter, nominative singular → āśvapatam "the offspring (and the rest) of aśvapati".

Watch out. The vṛddhi falls on the first vowel of the whole base, whatever vowel that is. Normally a rule that says "vṛddhi" without naming the sound replaced acts only on i, u, ṛ, ḷ, by iko guṇavṛddhī (A 1.1.3). Here the rule names the sound it replaces, "the first vowel", so that restriction does not come in. That is why the a of aśva takes vṛddhi.

SK 1076 · किति च · kiti ca (A 7.2.118)

Rule. And when a taddhita marked with k follows, the same happens: vṛddhi replaces the first vowel of the base. Bhaṭṭoji's commentators say the rule is set down here, off its proper topic, for ease of explanation, and for forms such as bāhīkaḥ below.

Example.

Bhaṭṭoji's own examples here close the aṇ derivation of Section A:

aśvapater apatyādi "the offspring and the rest of aśvapati" → āśvapatam. The "and the rest" takes in the other senses, such as "collection".

gaṇapati + aṇ → gāṇapatam.

The case with k proper:

bahis + īkak → the ṭi is is dropped (a vārttika under SK 1077) → bah + īka → īkak is kit → the first vowel a of bahis becomes ā → bāhīkaḥ (Section D).

agni + ḍhak → ḍhak is kit → a becomes ā → āgneyam (Section E).

Watch out. Bhaṭṭoji calls the form gāṇapatyo mantraḥ, "a mantra whose deity is gaṇapati", with the affix ṇya, simply an error. Gaṇapati is in the aśvapati list (SK 1074), so it takes aṇ, not ṇya: gāṇapatam. Scholars discuss whether the form can be defended in another way.

C. ṇya after diti, aditi, āditya and pati-final words (and yama) — running example: diter apatyam → daityaḥ

One sūtra teaches the affix ṇya for a short list of bases and for every word ending in pati. Its ṇ is a marker, so the affix attaches as ya and brings first-vowel vṛddhi. On the running example, diti + ṇya gives daityaḥ "a son of diti". A vārttika adds yama.

SK 1077 · दित्यदित्यादित्यपत्युत्तरपदाण्ण्यः · dityadityādityapatyuttarapadāṇ ṇyaḥ (A 4.1.85)

Rule. After diti, aditi and āditya, and after a word whose last member (uttarapada) is pati, the affix ṇya occurs in the prāgdīvyatīya senses. It is an apavāda of aṇ (SK 1073). Bhaṭṭoji's commentators note that "in the prāgdīvyatīya senses" is supplied by anuvṛtti from SK 1073.

Example.

diter apatyam "the offspring of diti" → diti + ṇya → the ṇ is initial and retroflex, so it is a marker by cuṭū (SK 189); the affix is ṇit → first-vowel vṛddhi by SK 1075: i → ai → daiti + ya → the final i deleted by yasyeti ca (SK 311) → daitya → daityaḥ.

aditi + ṇya → vṛddhi a → ā → yasyeti ca → ādityaḥ "of aditi".

āditya + ṇya → the first vowel is already ā, so vṛddhi changes nothing → the final a deleted by yasyeti ca → ādity + ya = ādityya

→ optionally one y dropped by halo yamāṃ yami lopaḥ (A 8.4.64) → ādityaḥ "of āditya" (or, with two y, ādityyaḥ).

prajāpati + ṇya (a pati-final word) → vṛddhi a → ā → yasyeti ca → prājāpatyaḥ.

Applies when / blocked by. Bhaṭṭoji's commentators add three points. The one-y step above is theirs: in ādityya, a later optional doubling of the y and the optional deletion of a y by A 8.4.64 give a spelling with one y or with two. A different deletion of the y, taught in A 6.4.151, does not apply, because it needs an inner affix of offspring, and the ṇya of āditya is not one. And the plural daiteyāḥ for "sons of diti" is simply incorrect; at most it could be defended as pṛṣodarādi.

Watch out. Aśvapati and gaṇapati also end in pati, but they do not take ṇya. SK 1074 has already given them aṇ.

Vārttika yamāc ca — "and after yama, ṇya"

Bhaṭṭoji adds that this vārttika is read so in the Kāśikā, the old running commentary on the Aṣṭādhyāyī. His commentators take this to mean it is not found in the Bhāṣya, the great commentary.

yama + ṇya → vṛddhi a → ā → yasyeti ca → yāmyaḥ.

Bhaṭṭoji's commentators add that yāmya could be formed in another way too, even without this vārttika.

D. Special affixes for special bases — the vārttikas under SK 1077 for pṛthivī, deva, bahis, sthāman, loman and go — running example: bahir bhavaḥ → bāhyaḥ / bāhīkaḥ

SK 1077 carries eight more vārttikas. Most assign a special affix to one base, in place of aṇ or of another affix it would take. One removes an affix again, and one concerns offspring only. On the running example, bahis "outside" takes yañ or īkak and loses its ṭi: bāhyaḥ or bāhīkaḥ "existing outside".

Vārttika pṛthivyā ñāñau — after pṛthivī, ña and añ

Both affixes attach as a and both bring vṛddhi. The ñ of ña is initial, a marker by cuṭū; the ñ of añ is final, a marker by SK 1.

pṛthivī + ña → vṛddhi of the first vowel ṛ → ār → the final ī deleted by yasyeti ca → pārthiva → feminine with ṭāp, because it ends in a → pārthivā.

pṛthivī + añ → the same steps → pārthiva → feminine with ṅīp by SK 470, because the stem ends in añ → pārthivī.

Why teach two affixes that look the same? Bhaṭṭoji's commentators answer: to get both feminines. With ña alone there would be no ṅīp; with añ alone, no ṭāp. That is why Bhaṭṭoji gives the feminine forms as examples.

Vārttika devād yañañau — after deva, yañ and añ

deva + yañ → vṛddhi e → ai → the final a deleted by yasyeti ca → daivyam.

deva + añ → the same → daivam.

Vārttika bahiṣaṣ ṭilopo yañ ca — after bahis, yañ, and the ṭi of the base is dropped

Bahis "outside" is an indeclinable. Its ṭi is is.

bahir bhavaḥ "existing outside" → bahis + yañ → the ṭi is is dropped → bah + ya → vṛddhi a → ā (SK 1075, yañ is ñit) → bāhyaḥ.

Bhaṭṭoji's commentators draw a lesson from the express mention of the ṭi-loss. A general statement, avyayānāṃ bhamātre ṭilopaḥ, already drops the ṭi of an indeclinable bha stem. If that statement always applied, this vārttika would not need to say it. So the wording is a jñāpaka: it shows that the general rule does not always apply. Hence ārād bhavaḥ → ārātīyaḥ, with the ṭi kept.

Vārttika īkak ca — and īkak after bahis, with the same ṭi-loss

bahis + īkak → the ṭi is is dropped → bah + īka → īkak is kit, so vṛddhi a → ā by SK 1076 → bāhīkaḥ.

This is the form for which SK 1076 was read early.

Vārttika sthāmno 'kāraḥ — after sthāman, the affix a

Sthāman means "stance". The vārttika applies to a word ending in sthāman, by the tadanta reading. Bhaṭṭoji's commentators rest this on the example of the Bhāṣya; scholars discuss the reason. The example word is a bahuvrīhi: aśvasyeva sthāmā yasya "whose stance is like a horse's" → aśvasthāman.

aśvasthāman + a → the ṭi an is dropped by nas taddhite (SK 679) → aśvasthām + a → s becomes t, because the word belongs to the pṛṣodarādi (A 6.3.109) → aśvatthāmaḥ "the offspring of aśvatthāman", "born in him", and so on.

Watch out. There is no vṛddhi: aśvatthāmaḥ, not āśvatthāmaḥ. The affix a has no ñ, ṇ or k. With aṇ, Bhaṭṭoji's commentators point out, there would have been vṛddhi.

Vārttika bhavārthe tu lug vaktavyaḥ — but in the sense "existing in", luk of that affix

In the one sense bhava, "existing in", the affix a of the previous vārttika disappears by luk.

aśvatthāman + a (sense: "existing in") → luk of a → aśvatthāman → aśvatthāmā.

Vārttika lomno 'patyeṣu bahuṣv akāraḥ — after loman, when the offspring are many, the affix a

Loman means "hair". Again the vārttika applies to a word ending in loman. Bhaṭṭoji's commentators analyse the example word: uḍūni nakṣatrāṇīva lomāni yasya "whose hairs are like stars" → uḍuloman.

uḍuloman + a (many offspring) → the ṭi an dropped by nas taddhite (SK 679) → uḍulom + a → nominative plural uḍulomāḥ, accusative plural uḍulomān.

Applies when / blocked by. This a is an apavāda of the iñ that uḍuloman would otherwise take as a member of the bāhvādi list (A 4.1.96). Why "when many"? With one offspring, iñ does come: uḍulomno 'patyam → uḍuloman + iñ → ṭi dropped, vṛddhi u → au → auḍulomiḥ. An editor's note in the printed text adds that this vārttika concerns offspring alone; it stands among the common affixes because it continues the affix a of the sthāman vārttika.

Vārttika gor ajādiprasaṅge yat — after go, where an affix beginning with a vowel would come, yat

Go means "cow". Where some rule would add an affix beginning with a vowel (an ajādi affix) after go, yat comes instead. It is an apavāda of aṇ. Its t is a marker; it attaches as ya.

go + yat (the o of go becomes av before the y of an affix, by vānto yi pratyaye, SK 63 in chapter 3) → gavyam "existing in cows", "come from a cow", "whose deity is the cow", "belonging to a cow", and so on.

Bhaṭṭoji's commentators read the vārttika with one more word, sarvatra, "everywhere": in all the prāgdīvyatīya senses, not only "offspring". Scholars discuss how far this reaches.

Applies when / blocked by. Why "an affix beginning with a vowel"? An affix beginning with a consonant is untouched.

gobhyo hetubhya āgatam "come from cows as the cause" → go + rūpya (A 4.3.81) → gorūpyam.

go + mayaṭ → gomayam. Bhaṭṭoji's commentators cite different rules for this mayaṭ.

Watch out. There is no vṛddhi in gavyam: yat has no ñ, ṇ or k, so SK 1075–1076 do not apply.

E. añ after the utsa list; ḍhak after agni and kali — running example: utsa + añ → autsaḥ (and agner apatyādi → āgneyam)

The chapter closes with one more list-rule and one more vārttika. The utsa list takes añ, so utsa "spring" gives autsaḥ. Agni "fire" and kali take ḍhak in every sense of the stretch: āgneyam, kāleyam.

SK 1078 · उत्सादिभ्योऽञ् · utsādibhyo 'ñ (A 4.1.86)

Rule. After utsa and the rest of its list, the affix añ occurs in the prāgdīvyatīya senses. Bhaṭṭoji's commentators supply that sense phrase. They add that añ is an apavāda of aṇ, and also of aṇ's own apavādas such as iñ.

Example.

utsa + añ → the ñ is a marker, so the affix is ñit → vṛddhi of the first vowel by SK 1075: u → au → autsa + a → the final a deleted by yasyeti ca → autsaḥ.

Applies when / blocked by. Bhaṭṭoji's commentators quote a class-rule (gaṇa-sūtra) inside the list, baṣkayāse: baṣkaya takes añ only when it is not in a compound. So bāṣkayaḥ, but from a compound gaurbaṣkayiḥ.

Vārttika agnikalibhyāṃ ḍhak vaktavyaḥ — after agni and kali, ḍhak is to be stated

Ḍhak is kit: its k is a marker. Its ḍh is replaced by eya (A 7.1.2).

agner apatyādi "the offspring and the rest of agni" → agni + ḍhak → ḍh → eya → agni + eya → vṛddhi a → ā by SK 1076, ḍhak being kit → the final i deleted by yasyeti ca → āgneyam.

kali + ḍhak → the same steps → kāleyam.

Applies when / blocked by. The "and the rest" covers every prāgdīvyatīya sense. Bhaṭṭoji's commentators list them for āgneyam: agner apatyam "offspring of agni", agnir devatā asya "whose deity is agni", agninā dṛṣṭaṃ sāma "a chant seen by agni", agner idam "belonging to agni", agnau bhavam "existing in agni", agner āgatam "come from agni". They explain the placing. In the Bhāṣya this vārttika stands under a later rule and carries the words "there, everywhere". Bhaṭṭoji drops those words and reads it here, among the common affixes, so that one statement gives ḍhak in all these senses. It thereby replaces two narrower statements for agni and kali given later.

Terms used in this chapter

• taddhita — a secondary affix: one added to a finished noun to make a new noun stem.
• kṛt — an affix added to a root.
• prātipadika — a noun stem; a taddhita-ending form is one (SK 179).
• apatya — offspring; the first of the senses of the taddhita section.
• vṛddhi — the name of ā, ai, au (SK 16); here, the replacement of a base's first vowel by its vṛddhi (the vṛddhi of ṛ comes with r attached, ār).
• adhikāra — a heading whose words are read into the rules below it, without making a form itself.
• samartha — "capable": in this section, a word whose sandhi operations are already done (SK 1072).
• paribhāṣā — a rule about how other rules are to be read.
• prathama — "first": the word named first in a sense-rule, after which the affix goes (SK 1072).
• tasyāpatyam — the sense-rule "his offspring", whose first-named word is the genitive tasya.
• vā — "optionally"; here it keeps the phrase with separate words correct beside the derived word.
• nirdhāraṇa (partitive) genitive — a genitive meaning "among": "the first among the vowels".
• vidhi — an operative rule, one that actually adds, replaces or names something.
• aṇ — the default taddhita affix of this stretch; attaches as a, marked with ṇ (SK 1073).
• it — a marker sound, written in the upadeśa but not part of the affix as attached.
• upadeśa — the original statement of an affix as the grammar teaches it, markers included.
• halantyam — SK 1: a final consonant of an upadeśa is it; "a marker by SK 1" means this.
• ṇit / ñit / kit — an affix marked with ṇ / ñ / k; each brings first-vowel vṛddhi here (SK 1075–1076).
• anuvṛtti — carrying a word of an earlier rule forward into a later one, where it is read though not repeated.
• utsarga — a general rule, such as the default aṇ.
• apavāda — an exception: a specific rule that sets aside a general one.
• iñ — a specific offspring affix that sets aside aṇ, for example after the bāhvādi words.
• prāgdīvyatīya — "before dīvyat": the senses taught before the rule tena dīvyati (SK 1550), where aṇ is the default.
• gaṇa — a list of words named by its first member plus "and the rest" (aśvapatyādi, utsādi, bāhvādi).
• ṇya — the affix of SK 1077; attaches as ya, ṇit.
• uttarapada — the last member of a compound; "pati-final" means having pati as the uttarapada.
• cuṭū — SK 189: an initial palatal or retroflex of an affix is a marker.
• bha — the name a stem carries before an affix beginning with a vowel or with y (SK 231).
• yasyeti ca — SK 311: the final i- or a-vowel of a bha stem is deleted before a taddhita.
• vārttika — a supplementary statement on a sūtra, completing or correcting it.
• Kāśikā / Bhāṣya — the old running commentary on the Aṣṭādhyāyī / the great commentary on it.
• pṛṣodarādi — the list of forms accepted as they are found, without a full derivation (A 6.3.109).
• ña / añ / yañ — affixes that attach as a, a, ya, all ñit.
• ṭāp / ṅīp — the feminine affixes giving -ā / -ī; ṅīp comes after a stem in añ (SK 470).
• ṭi — the last vowel of a word with any consonant after it (is of bahis, an of sthāman).
• ṭi-lopa — deletion of the ṭi.
• avyaya — an indeclinable, such as bahis.
• jñāpaka — a wording that reveals a principle stated nowhere else.
• īkak — the affix of the bahis vārttika; attaches as īka, kit.
• tadanta-vidhi — the reading by which a rule naming a word also covers a word ending in it (aśvasthāman, uḍuloman).
• bahuvrīhi — a compound naming something outside itself ("whose stance is like a horse's").
• nas taddhite — SK 679: an n-final bha stem loses its ṭi before a taddhita.
• luk — the complete disappearance of an affix (SK 260).
• bhava — the sense "existing in".
• ajādi — beginning with a vowel.
• yat — the affix after go; attaches as ya, with no vṛddhi.
• rūpya / mayaṭ — consonant-initial affixes, not blocked by yat (gorūpyam, gomayam).
• gaṇa-sūtra — a small rule placed inside a word-list.
• ḍhak — the affix after agni, kali; kit, with ḍh replaced by eya (A 7.1.2).

Summary