अव्ययप्रकरणम्

Covers SK 447–452 (6 sūtras): all present.

avyaya-prakaraṇa — outline (p14, SK 447–452, all 6 sūtras)

Assigning the avyaya (indeclinable) designation and then cashing it out. Five saṃjñā-sūtras name the indeclinables — first the ready-made lexicon (svarādi + cādi), then three derivational routes (a taddhita form, a kṛt form, an avyayībhāva compound) — and one operation-sūtra draws the payoff: the case-affixes (āp, sup) enjoined after an avyaya undergo luk. Four outcome-groups, kept in strict SK order (447→452). The whole point of the saṃjñā is group D: an avyaya is "one-form-only," so its would-be endings are dropped.

A. The off-the-shelf indeclinables — particles and adverbs made avyaya by list — running example: uccaiḥ (listed → avyaya, no derivation)

• 447 svarādinipātam avyayam 1.1.37 — the words of the svarādi list and the nipātas (the cādi particles, cādayo 'sattve 1.4.57) receive the saṃjñā avyaya. svarādi (an ākṛtigaṇa — open/illustrative): svar, antar, prātar, punar, uccais, nīcais, śanais, ṛte, pṛthak, divā, sāyam, joṣam, svayam, vṛthā, svasti, alam, vaṣaṭ, namas, dhik…; cādi (also ākṛtigaṇa): ca, vā, ha, aha, eva, evam, nūnam, khalu, kila, sma…, plus by gaṇasūtra the upasarga-, vibhakti-, and svara-lookalikes (upasarga-vibhakti-svara-pratirūpaka). So uccaiḥ, svasti, ca are avyaya straight off the list — no derivation. (These feed group D: uccaiḥ supplies the gauṇa test at 452; svasti, prātar are the svarādi words there shown to carry a kāraka.)

B. A taddhita-derivate declared avyaya — the parigaṇana gate — running example: pacatirūpam (an asarva-vibhakti taddhita, yet kept OUT)

• 448 taddhitaś cāsarvavibhaktiḥ 1.1.38 — a taddhita-ending word that is asarva-vibhakti is likewise avyaya. The bahuvrīhi is read not as "that from which all case-endings fail" (avyayas do take all seven vibhaktis — only ever in the singular), but, taking sarva as avayava-kārtsnya (as in sarvaḥ paṭo dagdhaḥ), as "that from which the whole — three-number — vibhakti does not arise; only the singulars do." Because the class is really pinned by a parigaṇana (vārttika), the sūtra is strictly dispensable. Enumerated: tasil-group up to pāśap (pañcamyās tasil 5.3.7 … dvitryoś ca dhamuñ 5.3.45); śas-group up to the samāsāntas (bahv-alpārthāt 5.4.42 … avyaktānukaraṇāt 5.4.57); am; ām; the "times" suffixes (saṃkhyāyāḥ… kṛtvasuc 5.4.17, dvi-tri-caturbhyaḥ suc 5.4.18, vibhāṣā bahoḥ dhā 5.4.20); tasi & vati (tena tulyaṃ… vatiḥ 5.1.115); nānā & ñau (vinañbhyāṃ nānāñau 5.2.27). The gate keeps pacatirūpam and pacatikalpam OUT (rūpap/kalpap taddhitas, asarva-vibhakti, but un-enumerated) and blocks ati-vyāpti to ubhaya — none gets the saṃjñā. (The purposeful pratyāhāra sup in 452 is what secures "seven vibhaktis, all singular" here.)

C. A kṛt- or compound-derivate declared avyaya — running example: smāraṃ smāram (ṇamul, an m-final kṛt → avyaya)

• 449 kṛnmejantaḥ 1.1.39 — a kṛt-affix ending in m or in an ec (e o ai au) — and, since a bare kṛt cannot stand alone, by tadanta-vidhi the word ending in such a kṛt — is avyaya. smāraṃ smāram: ṇamul on √smṛ (ābhīkṣṇye ṇamul ca 3.4.22), vṛddhi ṛ → ār with raparatva (aco ñṇiti 7.2.115), then doubling (nityavīpsayoḥ 8.1.4) → an m-final kṛt → avyaya. jīvase: the affix ase (tumarthe se-sen-ase… 3.4.9), an e-final kṛt. pibadhyai: the affix śadhyai (same rule; śit → sārvadhātuka → piba-substitution by pāghrā-dhmā… 7.3.78), an ai-final kṛt (no śap, as the avyaya-kṛt is bhāve, not agentive).
• 450 ktvā-tosun-kasunaḥ 1.1.40 — a word ending in ktvā, tosun, or kasun is avyaya. One per affix: kṛtvā (ktvā), udetoḥ (tosun), visṛpaḥ (kasun).
• 451 avyayībhāvaś ca 1.1.41 — and an avyayībhāva compound is avyaya (the saṃjñā carried by anuvṛtti; ca adds the compound to it). adhihari — an avyayībhāva in a vibhakti sense = harau ("in/on Hari"). [Segmentation note: this unit's Bālamanoramā gloss (supplying the elided avyaya-saṃjñaḥ syāt and glossing adhihari) is printed under sk.450's b1; k1 here carries the example.]

D. The payoff — luk of the āp and sup enjoined after an avyaya — running example: tatra (āp-luk; śālāyām only names the feminine)

• 452 avyayād āp-supaḥ 2.4.82 — after an avyaya, the āp and the sup enjoined after it undergo luk (luk by anuvṛtti from ṇyakṣatriyārṣa-ñito yūni luk 2.4.58; āp-sup a samāhāra-dvandva). tatra (avyaya) → āp dropped (the underlying śālā's āp is gone; śālāyām is added only to convey feminine gender); atha → sup dropped. The word vihita ("enjoined-after," not merely "following") blocks atyuccaisau: there the sup is enjoined after the compound aty-uccaiḥ (atyādayaḥ, under kugatiprādayaḥ 2.2.18), not directly after the avyaya — and aty-uccaiḥ as a whole is not avyaya, since only the bare uccaiḥ is listed in svarādi; the saṃjñā's tadanta-vidhi does not reach a subordinate member (na gauṇe — the sarvanāman-prohibition on upasarjana carried into the avyaya-saṃjñā).
• 452 (vārttika, k2) āb-grahaṇaṃ vyartham aliṅgatvāt — the mention of āp in 2.4.82 is idle, since avyayas have no gender (aliṅgatva); i.e. avyayād āb-luk-vacanānarthakyam.
• 452 (śruti + Bhāguri appendix, k3) — the Atharvaṇa-śruti sadṛśaṃ triṣu liṅgeṣu sarvāsu ca vibhaktiṣu / vacaneṣu ca sarveṣu yan na vyeti tad avyayam (Mahābhāṣya-cited) is reconciled with aliṅgatva by reading its locatives via ṣaṣṭhī cānādare 2.3.38: "disregarding gender, kāraka, number, [that which] does not vary." The liṅga-kāraka-saṅkhyā-abhāva holds strictly of nipātas; some svarādi do take a kāraka (svasti vācayati, prātar yajate). Then, prasaṅgāt, the Bhāguri verse vaṣṭi bhāguriḥ…: al-lopa of ava/api → vagāhaḥ/avagāhaḥ, pidhānam/apidhānam, and āp-luk of consonant-final stems (vācā niśā diśā) — though the SK itself doubts the verse stands in the Bhāṣya (cf. Bhāṣya on ṅyāp-prātipadikāt 4.1.1) and holds vācā to be asādhu. Thus close the avyayas; the Avyayī-bhāva-prakaraṇa follows.