Hal-sandhi-prakaraṇam — outline (p05, SK 111–149, all 39 sūtras)
Covers SK 111–149 (39 sūtras): all present. Grouped by consonant-change family; groups run roughly in SK order (noted departures below). The voicing rule jhalāṃ jaśo'nte (A 8.4.53 = SK 84) lies outside this span but is cross-referenced repeatedly where our sūtras feed on it. Reorder notes: 120 (śaścho'ṭi) is pulled up next to 121 (khari ca) into group C so tac chivaḥ is finished in one place; the augment-tuk of che ca (146–149) is kept in its own SK-order group I rather than merged with the earlier augments of group D; the suṭ augment (135) is placed with the ru→visarga engine (group H) it opens, not with the augments.
A. ścutva — dental → palatal before a palatal — running examples: sat + cit → sac cit, tat + śivaḥ → tac chivaḥ
- 111 stoḥ ścunā ścuḥ 8.4.40 — s and the dental ta-varga (stu), in conjunction with ś or the palatal ca-varga (ścu), become ś / ca-varga. In
rāmas-śete the pada-final s before ś → ś: rāmaś śete; in śārṅgin-jaya the n before j → ñ: śārṅgiñ-jaya. In sat-cit: t →(jaśtva, SK 84)→ d →(this rule, before c)→ j →(cartva, 121)→ c: sac-cit. - 112 śāt 8.4.44 — limit: a ta-varga standing after śa does not undergo 111. In
viśnaḥ, praśnaḥ the n after ś stays n (does not become ñ).
B. ṣṭutva — dental → retroflex before a retroflex — running example: rāmas + ṭīkate → rāmaṣ ṭīkate; limit ṣaṭ santaḥ
- 113 ṣṭunā ṣṭuḥ 8.4.41 — s and the dental ta-varga (stu), in conjunction with ṣ or the retroflex ṭa-varga (ṣṭu), become ṣ / ṭa-varga. Pada-final s before ṭ → ṣ:
rāmaṣ ṭīkate; s before ṣ → ṣ: rāmaṣ ṣaṣṭhaḥ; internally peṣṭā, taṭ ṭīkā, and ṇ for the retroflex-nasal in cakriṇ ḍhaukase. - 114 na padāntāṭ ṭoranām 8.4.42 — limit: a stu standing after a pada-final ṭu does not become ṣṭu — except before nām (anām). So s after word-final ṭ stays:
ṣaṭ santaḥ, ṣaṭ te; but before nām ṣṭutva does apply — ṣaṇṇām. Vārtika anāmnavatinagarīṇām: likewise ṣaṇṇavatiḥ, ṣaṇṇagaryaḥ. - 115 toḥ ṣi 8.4.43 — limit: a ta-varga before ṣ takes no ṣṭutva;
san ṣaṣṭhaḥ (the n is not retroflexed). The block then pivots to voicing (jhalāṃ jaśo'nte, SK 84): vāgīśaḥ (k → g), cidrūpam (t → d) — jaśtva, exploited throughout this prakaraṇa though its own sūtra is out of span.
C. Pada-end voicing residue: cartva (voiceless) & chatva (ś → ch) — running examples: sac cit (cartva), tac chivaḥ (chatva)
- 120 śaścho'ṭi 8.4.63 — a ś after a pada-final jhay (stop), when an aṭ (vowel/semivowel) follows, optionally → ch. After ścutva has made a pada-final dental into c (e.g.
tac), the following ś of śivaḥ → ch: tac chivaḥ (alt. tac śivaḥ). [Block cites d → j by ścutva as the feeding step.] - 121 khari ca 8.4.55 — cartva: a jhal becomes the corresponding car (voiceless unaspirate) before a khar (voiceless). The j of
saj-cit → c: sac cit; likewise closes tac chivaḥ. Vārtika chatvam ami (ś → ch before an am): tac chlokena (alt. tac ślokena); but with no following am, ś is untouched — vāk ścyotati.
D. Pada-final augments before a following consonant (kuk/ṭuk, dhuṭ, tuk, ṅamuṭ) — running example: saN + śambhuḥ → sañ chambhuḥ and kin
- 130 ṅṇoḥ kukṭuk śari 8.3.28 — after ṅ / ṇ, the augments kuk / ṭuk (respectively) optionally arise before a śar (ś ṣ s); being asiddha, they block jaśtva.
prāṅ + ṣaṣṭhaḥ → prāṅ kṣaṣṭhaḥ / (vārtika cayo dvitīyāḥ) prāṅ kṣaṣṭhaḥ / bare prāṅ ṣaṣṭhaḥ; and sugaṇ + ṣaṣṭhaḥ → sugaṇ ṭṣaṣṭhaḥ / sugaṇ ṭaṣaṣṭhaḥ / sugaṇ ṣaṣṭhaḥ. - 131 ḍaḥ si dhuṭ 8.3.29 — an s after ḍ optionally takes the augment dhuṭ (dh→t):
ṣaṭ tsantaḥ (alt. ṣaṭ santaḥ). - 132 naśca 8.3.30 — likewise an s after n optionally takes dhuṭ:
san tsaḥ (alt. san saḥ). - 133 śi tuk 8.3.31 — a pada-final n before ś optionally takes the augment tuk (t); the new t then feeds śaścho'ṭi (120, ch/no-ch) and jharo jhari (t-loss), yielding four forms
sañ chambhuḥ / sañ chambhuḥ / sañ chambhuḥ / sañ śambhuḥ (ñ-ch, ñ-c-ch, ñ-c-ś, ñ-ś). - 134 ṅamo hrasvādaci ṅamuṇnityam 8.3.32 — a pada-final ṅam (ṅ ṇ n) after a short vowel, before a vowel, obligatorily doubles (augment ṅamuṭ):
pratyaṅ + ātmā → pratyaṅṅātmā; sugaṇ + īśaḥ → sugaṇṇīśaḥ; san + acyutaḥ → sannacyutaḥ.
E. m / n before a consonant → anusvāra → parasavarṇa nasal (+ m-before-h options) — running examples: harim + vande → hariṁ vande; anusvāra → śāntaḥ
- 122 mo'nusvāraḥ 8.3.23 — a pada-final m before a hal (consonant) → anusvāra:
harim-vande → hariṁ vande. Pada-final only: mid-word gam-yate → gamyate untouched. - 123 naścāpadāntasya jhali 8.3.24 — a non-pada-final n (and m) before a jhal → anusvāra:
yaśas-i internal n → yaśāṃsi; ā-kran-syate → ākraṃsyate. Only before a jhal: man-yate → manyate (before y, no change). - 124 anusvārasya yayi parasavarṇaḥ 8.4.58 — an anusvāra before a yay → the nasal homorganic with the following stop:
aṅkitaḥ, añcitaḥ, kuṇṭhitaḥ, śāntaḥ, gumphitaḥ. (Being asiddha, it also shields kurvanti from ṇatva.) - 125 vā padāntasya 8.4.59 — for a pada-final anusvāra the parasavarṇa is only optional:
tvaṅ karoṣi / tvaṃ karoṣi; saṃyantā/nasalized-y, saṃvatsaraḥ, yaṃ lokam — anusvāra alternating with the nasal semivowel. - 126 mo rāji samaḥ kvau 8.3.25 — the m of sam stays m (no anusvāra) before rāj ending in kvip:
sam + rāṭ → samrāṭ. - 127 he mapare vā 8.3.26 — before an h that is itself followed by m, pada-final m optionally stays m:
kim hmalayati / kiṃ hvalayati. Vārtika: before an h followed by y/v/l, the y/v/l optionally so remain. - 128 yathāsaṃkhyamanudeśaḥ samānām 1.3.10 — paribhāṣā: equal-numbered later items map to earlier ones in order. It steers the 127-vārtika's y/v/l pairing:
kiy̐ hyaḥ/kiṃ hyaḥ, kib̐ hvalayati/kiṃ hvalayati, kil̐ hlādayati/kiṃ hlādayati. - 129 napare naḥ 8.3.27 — before an h followed by n, pada-final m optionally → n:
kin hnute / kiṃ hnute.
F. Pūrva-savarṇa of h (voiced aspirate for h after a stop) — running example: vāk + hariḥ → vāg ghariḥ
- 119 jhayo ho'nyatarasyām 8.4.62 — after a jhay (stop), h optionally becomes the sound homorganic with the preceding — namely that stop's voiced-aspirate (4th) member.
vāk-hariḥ → jaśtva vāg-hariḥ, then h → gh: vāg ghariḥ (alt. vāg hariḥ).
G. Further junction assimilations: nasalization, savarṇa-of-l, ud-pūrvasavarṇa — running examples: etat + murāriḥ → etan murāriḥ; tat + layaḥ → tallayaḥ; ut-sthānam → utthānam
- 116 yaro'nunāsike'nunāsiko vā 8.4.45 — a pada-final yar before a nasal optionally becomes nasal (its own homorganic nasal):
etad-murāriḥ → etan murāriḥ (alt. etad murāriḥ); r, already homorganic-realised, is exempt → caturmukhaḥ. Before a suffix in the spoken language it is obligatory: tan-mātram, cin-mayam. (madodagrāḥ, kakudmantaḥ keep d by yavādi nipātana.) - 117 tor li 8.4.60 — a ta-varga before l → parasavarṇa, i.e. l:
tad-layaḥ → tallayaḥ; a preceding n yields the nasalized l: vidvān-likhati → vidvām̐l likhati. - 118 udaḥ sthāstambhoḥ pūrvasya 8.4.61 — for sthā / stambh after ud, the initial s → pūrvasavarṇa (homorganic with the preceding d = the voiceless dental t/th):
ut-sthānam → utthānam; ut-stambhanam → uttambhanam (with optional loss of the resulting th under jharo jhari).
H. The ru → anusvāra / visarga → sibilant & jihvāmūlīya/upadhmānīya engine — running example: sam + suṭ + kartā → saṃskartā; puṃs → puṃskokilaḥ
- 135 samaḥ suṭi 8.3.5 — for sam, ru is substituted when the augment suṭ (s) follows:
sam-s-kartā → (ru →) saṃskartā; opens the ru-section. - 136 atrānunāsikaḥ pūrvasya tu vā 8.3.2 — adhikāra: here in the ru-section the sound preceding ru optionally takes a nasal (anunāsika) realisation.
- 137 anunāsikāt paro'nusvāraḥ 8.3.4 — when the nasal option is set aside, an anusvāra augment stands after that sound and before ru (feeding the visarga rules via kharavasānayor visarjanīyaḥ, SK 76).
- 138 visarjanīyasya saḥ 8.3.34 — a visarjanīya (visarga) before a khar → s: this closes
saṃskartā. Vārtika sampuṃkānāṃ saḥ gives the s directly for sam/puṃs/kān words: saṃskartā/sam̐skartā, etc. (the block enumerates the many optional s/anusvāra/doubling forms). - 139 pumaḥ khayyampare 8.3.6 — the final of puṃs → ru before a khay that is itself followed by an am:
puṃskokilaḥ/puṃskokilaḥ, puṃsputraḥ (via 138's s). Not before a plain khay: puṃkṣīram (khay not am-followed), puṃdāsaḥ (not a khay). - 140 naśchavyapraśān 8.3.7 — a pada-final n → ru before a chav followed by an am — but not the n of praśān:
śārṅgiñ chindhi, cakriṃs trāyasva (via visarga→s→ścutva). Contrast hanti (not pada-final n), sant saruḥ (khay not am-followed), praśāntanoti (excepted word). - 141 nṝn pe 8.3.10 — for the word nṝn, ru is optional before p: feeds the p-varga visarga rule below.
- 142 kupvoḥ ≍ka≍pau ca 8.3.37 — the visarjanīya before a ku (guttural) / pu (labial) → jihvāmūlīya (≍ka) / upadhmānīya (≍pa) respectively, or (by ca) stays visarga:
nṝm̐ ≍ pāhi / nṝm̐ḥ pāhi / nṝn pāhi. Apavāda only of 138, not of the śarpare rule — so vāsaḥ kṣaumam keeps its visarga. - 143 kānāmreḍite 8.3.12 — the n of kān → ru before an āmreḍita (reduplicated portion), yielding s by the sampuṃkānām vārtika.
- 144 kaskādiṣu ca 8.3.48 — in the kaskādi list the visarga → s (apavāda of the jihvāmūlīya/upadhmānīya of 142):
kaskaḥ, kautaskutaḥ, sarpiṣkuṇḍikā, dhanuṣkapālam, kāṃskān (an ākṛti-gaṇa).
I. The augment tuk before ch, in saṃhitā — running example: sva + chāyā → svacchāyā
- 145 saṃhitāyām 6.1.72 — adhikāra: the following operations take effect only in saṃhitā (close juncture); the first saṃhitā-adhikāra, governing 146 ff.
- 146 che ca 6.1.73 — a short vowel before ch takes the augment tuk (t) in saṃhitā:
sva-chāyā → svat-chāyā → (jaśtva d, ścutva j, cartva c) → svacchāyā; likewise śivacchāyā. (Kutva is blocked because ścutva is asiddha.) - 147 āṅmāṅośca 6.1.74 — after āṅ / māṅ before ch, tuk is added (apavāda of the optional 149):
ā-chādayati → ācchādayati; mā-chidat → mācchidat. - 148 dīrghāt 6.1.75 — after a long vowel before ch, tuk is added — the augment attaching to the vowel, not to ch (per the senāsurā... jñāpaka):
cecchidyate. - 149 padāntādvā 6.1.76 — after a pada-final long vowel before ch, tuk is optional:
lakṣmī-chāyā → lakṣmīcchāyā / lakṣmīchāyā.