तद्धिताधिकारे प्राग्दिशीयप्रकरणम्

Covers SK 1947–1973 (27 sūtras): all present. The spine is Aṣṭādhyāyī 5.3.1–5.3.25, in order, with one omission and three insertions. The omission is 5.3.20 tayor dārhilau ca chandasi (SK 3499), which is Vedic and treated in the Vaidika section; it would fall between 1968 and 1969. The next Vedic sūtra, 5.3.26 thā hetau ca chandasi (SK 3500), also goes to the Vaidika section, so the prakaraṇa ends at 5.3.25. The insertions are three rules that build these forms: 7.2.104 ku tihoḥ (SK 1954), placed after tasil; 7.2.105 kvāti (SK 1960), placed after at; and 2.4.33 etadas tratasos… (SK 1962), placed after the tral rules.

taddhitādhikāre prāgdiśīyaprakaraṇam — outline (p40, SK 1947–1973, all 27 sūtras)

p39 closed at 1946 with ahaṃśubhamor yus (5.2.140). This prakaraṇa opens the third pāda of the fifth book and runs up to, but not including, dikśabdebhyaḥ… (5.3.27, SK 1974).

• Where the prakaraṇa begins. NSP opens 1947 with "atha svārthikāḥ pratyayāḥ" (p. 318, printed 308; scan). The BM print opens the prāgdiśīya-prakaraṇa here.
• Where it ends. NSP prints the closing line "prāgdiśīyānāṃ vibhaktisaṃjñānāṃ pūrṇo 'vadhiḥ" in the mūla column after 1973 (p. 320, printed 310; scan). BM and TB close with the same words. Our kaumudī block lacks this line. The English supplies it in brackets, and the omission has been escalated to Ajai as a mūla issue. The Wikisource SK has a different colophon, "iti taddhite prāgdiśīyaprakaraṇam".
• 1947–1948: the frame. Every affix up to diś is a vibhakti. The affixes come after kim, a sarvanāman or bahu, but not after dvi and the rest.
• 1949–1952: base substitutes before these affixes: iś for idam; eta/it for idam before r/th; eta/it or an for etad; sa for sarva before d.
• 1953–1963: place and source. tasil after the fifth case, and tral, ha and at after the seventh. Then ku and kva for kim, the Vedic ha, the anudātta aś of etad, and tasil etc. after other cases.
• 1964–1970: time. dā, rhil, adhunā, dānīm, then a nipātana list of fourteen time-adverbs.
• 1971–1973: mode. thāl, and thamu after idam, etad and kim.

Three threads run through the section.

• The vibhakti designation (1947). BM gives its purposes:
• the prohibition na vibhaktau tusmāḥ (SK 190);
• tyadādy-atva, a for the final of tyad etc. (tyadādīnām aḥ, SK 265);
• the accent of ūḍidampadādi… (SK 3717).
• 1955 relies on it: tasil is substituted for tasi so that tyadādy-atva works (TB: yato 'vagacchati).
• At 1959 and 1972 BM and TB argue over how far na vibhaktau tusmāḥ reaches.
• "vā" continues (1947: "vā, however, does continue"; BM: explanation alone is the authority). So each affix stands beside the inflected form: kutaḥ / kasmāt, yataḥ / yasmāt (TB under sk.1954.t1). "samarthānām" and "prathamāt" of SK 1072 have ceased.
• The affixes are svārthika, in the base's own sense (1947). BM: no sense is stated for tasil etc., and words such as "atiśāyane" qualify the base-meaning. The derivatives are avyayas (BM on 1954: tasilādayaḥ prāk pāśapaḥ).

Only the kaumudī blocks are translated (SK-TRANSLATE §25). BM and TB readings are the ones recorded in the unit notes.

• TB printed under a neighbouring unit:
• 1948 at the tail of sk.1933.t1 (after "iti matvarthīyāḥ");
• 1953 and 1955 in part under sk.1954.t1;
• 1960 under sk.1959.t1;
• 1962, 1963, 1964 and 1965 under sk.1961.t1;
• 1968 under sk.1966.t1.
• TB on 1974 and after is printed under sk.1973.t1 and belongs to p41.
• No TB is printed on 1947, 1949, 1950, 1952, 1958 or 1967.
• NSP footnotes 2 and 3 on p. 318 (on 1947 and 1948) are not filed in any source.json. They were read from the scan and are recorded in the unit notes.

The p40 audit (.work/sk-run/p40-41-audit-fixes.md; reports sk1947-1961.md and sk1962-1976.md) found no MUST-FIX in this span. Its MINOR fixes are note-level, and this outline follows the corrected sense:

• Derivations the kaumudī does not give are BM's (or TB's), not Bhaṭṭoji's. The prakriyā of kutaḥ (1954), of ittham (1972) and of katham (ka by kimaḥ kaḥ, 1973) moves from the kaumudī block into the notes. They are marked below as BM/TB.
• Anuvṛtti sourcing:
• 1961: kim is from 1959 and the seventh case from 1957. Neither is a stated anuvṛtti; BM only glosses "kimaḥ saptamyantāt".
• 1969: the 1948 and 1957 identifications are the translator's; BM says only "kimādibhyaḥ saptamyantebhyaḥ".
• 1955 and 1959 are likewise marked as the translator's.
• BM and TB kept apart where they differ: 1956 (why pariṣiñcati has no tasil), 1959 (why na vibhaktau tusmāḥ does not stop the t of at being an it), and 1960 (kutra an apaśabda).
• 1962 variant: the print reads atho 'trādhīmahe; BM's pratīka and the Wikisource SK read ato 'trādhīmahe (WS also has gantāraḥ smaḥ). The English follows the print.

The digital source's corrupt forms are not repeated. The outline reads the NSP forms, which the scan confirms:

• svārthikāḥ (1947);
• prāgdiśaḥ / prāgdiśīye (1948–1950, 1952);
• pratyayaś ca, asminn ahani, ahnoḥ (1970);
• prakāreṇa (1972).

Ordering: SK order throughout. The SK reorders the AS only by the three insertions noted above.

Affix shapes and standard steps used below, cited where the units themselves do not spell them out:

• tasil → -tas (visarga: -taḥ); tral → -tra; rhil → -rhi; thāl → -thā. The l is an it-letter, which gives the accent of an l-marked affix (litsvara; BM on 1955, 1962). thamu → -tham: the u is for pronunciation, and it keeps the m from being final in upadeśa (BM on 1972).
• The case-ending of the base takes luk (luk of sup; BM, TB on 1954). A tyad-type base takes a for its final (SK 265), followed by pararūpa: yad → ya-taḥ, tad → ta-taḥ.
• ku / ka / kva for kim. ku before a vibhakti beginning with t or h (1954), ka by kimaḥ kaḥ (SK 342) elsewhere, and kva before at (1960).

A. The frame: vibhakti affixes after kim, pronominals and bahu — running example: bahubhyaḥ → bahutaḥ, but dvābhyām (no tasil)

• 1947 prāg diśo vibhaktiḥ 5.3.1 — the affixes taught before "diś" are termed vibhakti. BM: "diś" means the sūtra containing it, dikśabdebhyaḥ… (SK 1974).
• This is what licenses the substitutions: tyadādy-atva in yataḥ and tataḥ needs the tasil to be a vibhakti.
• Now come the svārthika affixes. "samarthānām" and "prathamāt" of SK 1072 have ceased, but vā continues. So bahutaḥ stands beside bahubhyaḥ.
• NSP fn 2 (scan): the sūtra is a saṃjñā-sūtra, known from the explanation; otherwise it could be taken as an adhikāra, like prāk kaḍārāt.
• 1948 kiṃsarvanāmabahubhyo 'dvyādibhyaḥ 5.3.2 — an adhikāra up to diś: the affixes come after kim, a sarvanāman and bahu, but not after dvi and the rest. "prāgdiśaḥ" is carried forward (BM).
• bahu is a base, so bahu + tasil → bahutaḥ.
• dvi is a sarvanāman but is excluded, so there is no tasil and the form stays dvābhyām (kaumudī at 1954).
• kim is named separately although it is a sarvanāman, because it is read among dvi and the rest. That reading lets kim remain under tyadādīnāṃ mithaḥ sahoktau…, as in tvaṃ ca kaś ca kau (BM, TB).
• It is an adhikāra because nothing to be enjoined is stated (BM; NSP fn 3).

B. Base substitutes before a prāgdiśīya affix — running example: idam + tasil → itaḥ; etad + tasil → ataḥ

• 1949 idama iś 5.3.3 — iś replaces idam before a prāgdiśīya affix. "prāgdiśīye pare" is supplied from the section (BM).
• iś is ś-marked, so it replaces the whole (sarvādeśa). idam + tasil → i-taḥ → itaḥ (BM; the form is in the kaumudī at 1954).
• The same substitute gives iha (1958), adhunā (1966) and idānīm (1967).
• 1950 etetau rathoḥ 5.3.4 — for idam ("idamaḥ" carried forward, BM): eta before an affix beginning with r, and it before one beginning with th, respectively. This is an apavāda to iś.
• The kaumudī writes "eta ita". BM analyses the pair as etaś ca ic ca.
• By tadādividhi, r and th mean an affix beginning with them (BM).
• idam + rhil → etarhi (1965). idam + thamu → ittham (1972).
• 1951 etado 'n 5.3.5 — the sūtra is split (yogavibhāga).
• "etadaḥ": with 1950 supplied, etad also takes eta/it before r/th. etad + rhil → etarhi (1969); etad + thamu → ittham (1972, vārttika).
• "an": with only "etadaḥ" continuing, an replaces etad before any prāgdiśīya affix. As anekāl it replaces the whole, and its n is deleted by nalopaḥ prātipadikāntasya (SK 236). etad + tasil → an-taḥ → ataḥ; atra (BM).
• BM: the n is not an it-letter, because n-marking serves the accent only in affixes.
• TB: the Vṛttikāra reads etado 'ś. Our BM pratīka also shows 'ś, while BM's own text runs 'n.
• 1952 sarvasya so 'nyatarasyāṃ di 5.3.6 — sa optionally replaces sarva before a prāgdiśīya affix beginning with d (tadādividhi, BM).
• sarva + dā → sadā / sarvadā (1964).

C. Fifth case — tasil: "from where" — running example: kasmāt → kutaḥ

• 1953 pañcamyās tasil 5.3.7 — tasil, optionally ("vā" from SK 1072, BM), after kim etc. ending in the fifth case.
• kasmāt → kim + tasil.
• The alternates stand side by side (TB, under sk.1954.t1): yataḥ / yasmāt, tataḥ / tasmāt, itaḥ / asmāt, ataḥ / etasmāt, amutaḥ / amuṣmāt, bahutaḥ / bahubhyaḥ.
• 1954 ku tihoḥ 7.2.104 — ku replaces kim before a vibhakti beginning with t or h. "kimaḥ" is from kimaḥ kaḥ (SK 342) and "vibhaktau" from aṣṭana ā vibhaktau (SK 371) (BM). tasil counts as a vibhakti by 1947.
• kim + tas → ku-taḥ → kutaḥ.
• kasmāt is the option without tasil; ku does not apply there, since -smāt does not begin with t or h.
• The kaumudī's series: yataḥ, tataḥ, ataḥ (an, 1951), itaḥ (iś, 1949), amutaḥ, bahutaḥ. But dvābhyām, since dvi is excluded.
• The step lists are BM's: luk of sup, tyadādy-atva and pararūpa for yad/tad, and u/m for adas by adaso 'seḥ… (SK 419). TB agrees. The kaumudī gives only the forms.
• 1955 taseś ca 5.3.8 — tasil replaces tasi after kim etc. This is the tasi of pratiyoge pañcamyās tasiḥ (SK 2111) and apādāne cāhīyaruhoḥ (SK 2112) (BM, TB).
• It is stated "for the accent", i.e. the litsvara, "and for the vibhakti". Without it, tasi, applying as the later rule, would not be prāgdiśīya; so there would be no tyadādy-atva, and yato 'vagacchati would fail (BM, TB).
• TB: tasil keeps its own scope in kuto 'vahīyate, so this rule is not superfluous.
• That "tasil" runs on from 1953 is the translator's identification.
• 1956 paryabhibhyāṃ ca 5.3.9 — tasil after pari and abhi.
• Vārttika sarvobhayārthābhyām eva: only when they mean "all" and "both". paritaḥ = sarvataḥ; abhitaḥ = ubhayataḥ.
• Why is there no tasil in pariṣiñcati, abhiṣiñcati?
• BM: because of vā.
• TB: because of the nature of expression (abhidhānasvābhāvya), pari there meaning "over" and abhi "facing". Only odanaṃ pariṣiñcati, where pari does mean "all round", is left to vā.
• Recorded apart; neither example is in the kaumudī.

D. Seventh case — tral, ha, at: "where" — running example: kasmin → kutra / kva

• 1957 saptamyās tral 5.3.10 — tral after kim etc. ending in the seventh case, dvi etc. excluded (BM).
• kim + tral → ku-tra → kutra (ku by 1954). Also yatra, tatra, bahutra.
• BM: they are derived like kutaḥ.
• TB: tral and tasil are independent affixes, not substitutes for the case-endings. So acca gheḥ etc. do not touch kutra and bahutra.
• vā (1947) makes it optional beside yasmin etc. That is the translator's note; no commentator restates it.
• 1958 idamo haḥ 5.3.11 — ha after idam ending in the seventh case. It is an apavāda to tral.
• idam + ha → i-ha → iha, with iś by 1949 (kaumudī).
• 1959 kimo 't 5.3.12 — at after kim ending in the seventh case, optionally, with tral as the alternative.
• vā is drawn back (apakarṣa) from 1961. BM: the only authority for this is the explanation (vyākhyāna). TB: the sūtra would have been better read as kimo 'd vā.
• The t of at is an it-letter, but the commentators give different grounds for why na vibhaktau tusmāḥ (SK 190) does not block it:
• BM: the vārttika tavargasyettvapratiṣedho 'taddhite;
• TB: the u of thamu (1972) shows that the prohibition is not universal.
• "Ending in the seventh case" comes from the kaumudī's own gloss (saptamyāḥ, 1957); the sourcing is the translator's.
• 1960 kvāti 7.2.105 — kva replaces kim when at follows. "kimaḥ" is from SK 342 (BM).
• kim + at → kva + a → kva.
• kutra is the option without at: tral, with ku by 1954 (BM).
• Is kutra correct?
• BM reports some (kecit) who hold that the apakarṣa of vā has no authority, so at is obligatory and kutra is an apaśabda;
• TB defends kutra from wide usage (Śrīharṣa, Amara), against the Bhāgavṛtti.
• The English keeps kutra, as the kaumudī does.
• 1961 vā ha ca chandasi 5.3.13 — in the Veda, after kim in the seventh case, ha optionally, and by ca at and tral too (BM).
• kim + ha → ku-ha → kuha: kuha sthaḥ, kuha jagmathuḥ. ku by 1954 is implicit; the kaumudī gives no derivation.
• The rule belongs to the Vedic section. It is set out here so that the apakarṣa of vā into 1959 can be seen (BM, TB).
• kim comes from 1959 and the seventh case from 1957. That is BM's gloss "kimaḥ saptamyantāt", not a stated anuvṛtti.

E. Beyond the fifth and seventh: the anudātta aś, and tasil/tral before bhavat — running examples: etasmād grāmāt → ataḥ (anudātta); sa bhavān → tatobhavān, tatrabhavān

• 1962 etadas tratasos tratasau cānudāttau 2.4.33 — in re-mention (anvādeśa), aś replaces etad before tra or tas, and the aś is anudātta. The tra and tas are anudātta too. "anvādeśe", "aś" and "anudāttaḥ" are carried forward from idamo 'nvādeśe… (SK 350) (BM).
• etasmin grāme sukhaṃ vasāmaḥ; atho 'trādhīmahe; ato na gantāsmaḥ ("we shall not go from here").
• TB: ataḥ = etasmād grāmāt, not "therefore".
• etad + tasil → a-taḥ → ataḥ. The form already comes from an (1951), so the rule is for the accent (BM).
• Once the anudātta aś is made, the litsvara no longer applies, and tra/tas would become udātta by the affix accent. So their anudātta has to be stated (BM).
• Variant: BM's pratīka and the Wikisource SK read ato 'trādhīmahe.
• 1963 itarābhyo 'pi dṛśyante 5.3.14 — tasil etc. are seen after kim etc. ending in case-endings other than the fifth and seventh. "kimādeḥ" is supplied (BM).
• Vārttika dṛśigrahaṇād bhavadādiyoga eva: only in construction with bhavat etc.
• First case: sabhavān, tatobhavān, tatrabhavān. Second case: taṃbhavantam, tatobhavantam, tatrabhavantam.
• So too with dīrghāyuḥ, devānāṃpriyaḥ and āyuṣmān (BM: tato dīrghāyuḥ, tatra dīrghāyuḥ; TB adds tatobhavatā etc.).

F. Time — dā, rhil, adhunā, dānīm: "when" — running example: tasmin kāle → tadā, tadānīm, tarhi

• 1964 sarvaikānyakiṃyattadaḥ kāle dā 5.3.15 — dā after sarva, eka, anya, kim, yad and tad, in the seventh case, denoting time. Only "saptamyāḥ" continues, on the authority of the explanation (BM).
• tad + dā → tadā. sarvasmin kāle → sadā (sa by 1952) / sarvadā. ekadā, anyadā, yadā.
• kim + dā → kadā, with ka by kimaḥ kaḥ (SK 342) (BM, TB).
• Why "time"? sarvatra deśe.
• 1965 idamo rhil 5.3.16 — rhil after idam, in the seventh case and denoting time ("ity eva" in the kaumudī). It is an apavāda to ha (1958).
• asmin kāle → idam + rhil → eta-rhi → etarhi, with eta before r by 1950 (BM, TB).
• Why "time"? iha deśe.
• 1966 adhunā 5.3.17 — the affix adhunā after idam, in the seventh case and denoting time. BM names the carried words idamaḥ, saptamyāḥ and kāle; the sūtra numbers (1965, 1957, 1964) are the translator's.
• idam → iś (1949) → i + adhunā. The i is deleted by yasyeti ca (SK 311), leaving only the affix: adhunā.
• TB records the Vṛttikāra's rival analysis, aś + dhunā, with the accent by SK 3717. The kaumudī follows the Bhāṣya's analysis, which takes the affix as adhunā, middle-acute by nipātana.
• 1967 dānīṃ ca 5.3.18 — dānīm also, after idam in the seventh case denoting time, in its own sense (BM).
• idam → iś + dānīm → idānīm (BM).
• 1968 tado dā ca 5.3.19 — after tad in the seventh case denoting time, dā and dānīm (BM).
• tadā, tadānīm.
• Vārttika tado dāvacanam anarthakaṃ vihitatvāt: dā after tad is already given by 1964. TB: the sūtra should simply read tadaś ca.
• [5.3.20, the Vedic tayor dārhilau ca chandasi, SK 3499, falls here.]
• 1969 anadyatane rhil anyatarasyām 5.3.21 — rhil, optionally, after kim etc. in the seventh case, for a time not of today. In the other option dā comes (BM).
• tarhi / tadā, karhi / kadā, yarhi / yadā.
• etasmin kāle → etad + rhil → etarhi: eta for etad before r, by the split rule "etadaḥ" (1951) (BM, TB). Compare 1965, where etarhi comes from idam.
• The 1948 and 1957 identifications are the translator's.

G. Fourteen time-adverbs by nipātana — running example: samāne 'hani → sadyaḥ

• 1970 sadyaḥ-parut-parāry-aiṣamaḥ-paredyavy-adya-pūrvedyur-anyedyur-anyataredyur-itaredyur-aparedyur-adharedyur-ubhayedyur-uttaredyuḥ 5.3.22 — fourteen ready-made forms (nipātana). The kaumudī unpacks them with seven vārttikas; BM calls each of them a Bhāṣya statement (bhāṣyavākya).
• sa for samāna, affix dyas, "day": samāne 'hani → sa + dyas → sadyaḥ.
• para for pūrva / pūrvatara, affixes ut / āri, "year": parut (last year) and parāri (the year before). The pairing "respectively" is the translator's, drawn from these examples.
• iś for idam, affix samasaṇ, "year": aiṣamaḥ, this year. The ṇ is an it-letter, so the first vowel takes vṛddhi (BM).
• edyavi after para, "day": paredyavi, the next day.
• aś for idam, affix dya: a + dya → adya, today.
• edyus after the eight pūrva, anya, anyatara, itara, apara, adhara, ubhaya and uttara (BM's wording), "day": pūrvedyuḥ, anyedyuḥ, ubhayedyuḥ.
• dyus after ubhaya as well: ubhayadyuḥ.
• TB: the nipātana fixes the domain, hence "ahani". Where the Bhāṣya wording quoted by BM differs (udārī, idamo 'śbhāvo dyaś ca), the print agrees with our source.

H. Mode — thāl and thamu: "thus, how" — running example: tena prakāreṇa → tathā; kena prakāreṇa → katham

• 1971 prakāravacane thāl 5.3.23 — thāl after kim etc. when they denote a mode (prakāra), in the base's own sense. The seventh case in the sūtra stands for the fifth, i.e. prakāravṛttibhyaḥ (BM).
• tena prakāreṇa → tad + thāl → ta-thā → tathā (tyadādy-atva). Also yathā.
• prakāra is a particular (viśeṣa) that differentiates a general, not similarity (BM, TB). TB: the "similarity" of yathā under SK 652 belongs to the compound, not to bare thāl.
• TB:
• thāl does not come after a first-case word;
• jātīyar (SK 2024) is added to the thāl-form: tathājātīyaḥ.
• 1972 idamas thamuḥ 5.3.24 — thamu after idam denoting a mode. It is an apavāda to thāl.
• Vārttika etado 'pi vācyaḥ: after etad too.
• anena / etena prakāreṇa → ittham. For idam, it comes by 1950; for etad, by the split rule "etadaḥ" (1951). This derivation is BM's and TB's; the kaumudī gives only the form.
• The u of thamu keeps the m from being an it-letter. The prohibition 'na vibhaktau' would have protected it anyway, so the u indicates that the prohibition is not universal (BM; cf. TB at 1959).
• TB: thamu also comes after a second-case word, as in itthambhūtaḥ, which the sūtra SK 552 itself uses.
• 1973 kimaś ca 5.3.25 — thamu after kim denoting a mode. BM supplies "prakāravṛtteḥ thamuḥ".
• kena prakāreṇa → katham. ka comes by kimaḥ kaḥ (SK 342); that step is BM's.
• TB: the rule is split off from 1972 so that only kim, not idam, continues into thā hetau ca chandasi (SK 3500).
• Close: "prāgdiśīyānāṃ vibhaktisaṃjñānāṃ pūrṇo 'vadhiḥ" (NSP; missing from our kaumudī, see above). The vibhakti designation ends here. p41 opens with dikśabdebhyaḥ saptamīpañcamīprathamābhyo digdeśakāleṣv astātiḥ (5.3.27, SK 1974).