भावकर्मतिङ्प्रकरणम्

Covers SK 2756–2765 (10 sūtras): all present. Every unit has its own kaumudī. 2758 (the survey of yak forms) and 2765 (num for labh, then the dvikarmaka verse) are long. Each is outlined under its own number; 2765's dvikarmaka matter continues in its own group (F → G). The spine is a selection of Aṣṭādhyāyī rules gathered for the bhāva/karma lakāra:

• the affix rules 3.1.67 (2756), 3.1.66 (2758) and 3.1.65 (2760);
• the aṅga rules 6.4.62 (2757), 6.4.44 (2759), 7.3.33 (2761), 6.4.93 (2762), 7.3.34 (2763), 6.4.33 (2764) and 7.1.69 (2765).

Every AS-ref below is checked against references/ashtadhyayi/sutraani-data.json, and all 10 match our source.json.

• Anuvṛtti from later in the SK.
• bhāvakarmaṇoḥ comes into 2756 from ciṇ bhāvakarmaṇoḥ (2758, the next sūtra but one) (BM, TB).
• karmakartari is drawn into 2760 by "ca" from acaḥ karmakartari (3.1.62, SK 2768, p64) (TB; BM in Wikisource).
• na in 2760 comes from na rudhaḥ (3.1.64, SK 2770) (BM, Wikisource).
• āt comes into 2759 from viḍvanor anunāsikasyāt (SK 2982).
• Anuvṛtti from earlier prakaraṇas:
• dhātoḥ ← SK 2629;
• cleḥ ← cleḥ sic (3.1.44, SK 2222);
• ciṇ / te ← ciṇ te padaḥ (3.1.60, SK 2513);
• vibhāṣā ← ye vibhāṣā (6.4.43, SK 2319);
• ñṇiti ← aco ñṇiti (7.2.115, SK 254);
• ṇau ← SK 2604; upadhāyāḥ ← SK 2364 / SK 2282; mitām ← mitāṃ hrasvaḥ (6.4.92, SK 2568);
• vṛddhiḥ ← SK 2473;
• na (n-elision) ← śnān nalopaḥ (6.4.23, SK 2544);
• vibhāṣā ← jāntanaśāṃ vibhāṣā (6.4.32, SK 3330);
• labheḥ ← labheś ca (7.1.64, SK 2582); num ← idito num dhātoḥ (7.1.58, SK 2262).

bhāvakarmatiṅprakaraṇam — outline (p63, SK 2756–2765, all 10 sūtras)

p62 (tiṅantaparasmaipadaprakaraṇam) closed at 2755 with "samāptā padavyavasthā" and the colophon "iti parasmaipadaprakriyā", which is supplied in brackets.

• Where the prakaraṇa begins. The print heads it "tiṅante bhāvakarmaprakriyā 60" (NSP p. 463, scan 0473). In kaumudī type it opens with "atha bhāvakarmaṇor laḍādayaḥ | bhāvakarmaṇoḥ 2679 iti taṅ |". Ours lacks that line, and the English supplies it in brackets at the head of sk.2756.k1. So taṅ for bhāva and karma comes from bhāvakarmaṇoḥ (1.3.13, SK 2679). The name bhāvakarmatiṅprakaraṇam and the number 63 are ours.
• Where it ends. 2765 closes with the dvikarmaka verse and "māsam āsyate māṇavakaḥ". The print then has "iti bhāvakarmaprakriyā" (NSP p. 467, scan 0477). Ours lacks it, so the English supplies it in brackets. p64 (karmakartṛtiṅprakaraṇam) opens with a prose paragraph and then karmavat karmaṇā tulyakriyaḥ (3.1.87, SK 2766).
• Contents in outline.
• 2756: yak before a sārvadhātuka in bhāva or karma. bhūyate, third person singular only, with the agent in the third case.
• 2757–2758: ciṇ and the ciṇvat-treatment.
• ciṇ in luṅ: abhāvi.
• In sya, sic, sīyuṭ and tās, the vowel-final roots and han, grah and dṛś are optionally treated as before ciṇ, with iṭ: bhāvitā / bhavitā.
• Then a survey of yak forms: karma-lakāras, causatives, desideratives, intensives, and the stem rules before yak.
• 2759–2760: root-specific adjustments. ā for tan before yak (tāyate / tanyate); no ciṇ for tap in remorse or as karmakartṛ (anvatapta).
• 2761: yuk for ā-finals before ciṇ and ñit/ṇit kṛt. dāyitā, adāyi. Then the ciṇvat forms of han, grah and dṛś, and the 104 forms of gṝ.
• 2762–2763: upadhā length before ciṇ. Optional lengthening for mit roots in ṇi (śāmitā / śamitā / śamayitā). No vṛddhi for m-final udātta roots (aśami), with the ācam/kam/vam exceptions.
• 2764–2765: the root's nasal before ciṇ. Optional n-loss for bhañj (abhāji / abhañji); optional num for labh (alambhi / alābhi, obligatory prālambhi).
• 2765 (continued): which karma the lakāra goes to with dvikarmaka roots, and with time-objects of intransitive roots.

Threads that run through the prakaraṇa:

• The bhāva lakāra (2756). bhāva = bhāvanā, the act of bringing about (utpādanā), i.e. kriyā. Every dhātu already denotes it as a dhātu, so the bhāva-lakāra only restates (anūdyate) it (BM, TB).
• There is no co-reference with yuṣmad or asmad, so the verb is third person.
• The action expressed by a tiṅ is not a substance (asattvarūpa), so it is singular only. The singular is the default (autsargika).
• TB calls this a general statement only (prāyovāda), citing the Bhāṣya's uṣṭrāsikā āsyante. TB also reports that the Manoramā reads āsikāḥ as a second-case kriyāviśeṣaṇa, which saves "ekavacanam eva" (audit MINOR). BM makes no qualification.
• The agent is unexpressed, so it is in the third case. In a karma-lakāra the tiṅ expresses the karma, so the karma is in the first case (2758).
• What the ciṇvat-treatment (2757) does. It is a kāryātideśa of aṅga operations only (the aṅgasya adhikāra). The kārikā "ciṇvad vṛddhir yuk ca hanteś ca ghatvaṃ…" (BM, TB) lists its uses: vṛddhi (bhāvitā), yuk (dāyitā, 2761), gh for han (ghānitā), and the iṭ, which counts as asiddha.
• No vadha, ga or gāṅ. Those are not aṅga operations (TB on 2757; kaumudī 2761).
• "upadeśe yo 'c". The gloss is "an ac that stands in upadeśa", not "ajanta in upadeśa". So a ṇyanta (bhāvitā) and a root that has taken guṇa (āritā, smāritā) still qualify (BM, TB). TB notes smāritā is still a difficulty under this gloss.
• The iṭ comes jointly (sanniyogaśiṣṭa) with the treatment. It is nitya, and blocks the valādi iṭ.
• "ārdhadhātuke" qualifies sīyuṭ only. That keeps the rule out of vidhiliṅ bhūyeta.
• "ñṇiti" qualifies kṛt, not ciṇ (2761, 2763). ciṇ is always ṇit (BM, TB).
• Where the commentary lies. Only the kaumudī blocks are translated.
• TB on 2756–2758 is at the tail of sk.2753.t1, after p62's colophon. Our OCR drops TB's opening on 2756 (scan 0473: the anuvṛtti and "bhāvo bhāvanā").
• TB on 2761 is at the tail of sk.2760.t1.
• TB on 2761's count and on 2762 is in sk.2760.n1. That is TB body text, mislabelled as an editorial note.
• TB on the tail of 2762 is in sk.2761.t1.
• sk.2765.t1 runs past the colophon into TB on the p64 opener.
• No TB on 2764 (print).
• BM is missing from our corpus for 2760, and truncated for 2764. Both are in Wikisource (uttara-3/0005–0006, 0011).

The p62–66 audit (.work/sk-run/p62-66-audit/b-sk2756-2765.md) found no MUST-FIX here. A fixer is applying the MINOR items in parallel. This outline already follows the audit:

• 2760: the anuvṛtti is cleḥ (SK 2222), ciṇ (SK 2513) and na (from na rudhaḥ, SK 2770), with karmakartari drawn in by ca (SK 2768). This is BM in Wikisource, which also supports the construal of jagle as "no prohibition, so ā".
• 2764: BM (Wikisource) itself states the anuvṛtti. "na" comes from SK 2544 as a word whose sixth-case ending has been dropped (luptaṣaṣṭhīka), together with lopa. vibhāṣā comes from jāntanaśāṃ vibhāṣā (SK 3330). It is not the translator's supply.
• 2761: the operation is kutva (ku-substitution), not "velarization".
• 2757: the labels [Before a sārvadhātuka:] and [āśīrliṅ:] are the translator's.
• 2756: TB's report of the Manoramā on "ekavacanam eva" (above).
• 2765: the Wikisource kaumudī reads māso māse vā, agreeing with BM's "māsam iti tv apapāṭhaḥ" against the print.

Print readings (NSP, scan-confirmed) that the outline follows:

• 2756: the opening line; sāmānādhikaraṇya.
• 2758: syāt (ours "svāt"); bubhūṣyate (ours "bubhūpyate"; the print's ṣ/p is unclear).
• 2759: ye vibhāṣā = SK 2319 (ours cross-refers SK 2391).
• 2760: tapaḥ cleḥ; eś (ours "eṣa").
• 2761: hanyate, ghāniṣyate, agrāhiṣātām, trayāṇām, iṭ, aṣṭānām, śamyate.
• 2762: ṇyantatvābhāve śamyate.
• 2763: upadhāyā. The vārttika sign stands before anācamikamivamīnām (k2); the k3 "(vārtikam)" tag is editorial.
• 2765: māṇavako dharmam; the colophon.

Still open for Ajai (mūla):

• 2761: "grahiṣīṣṭa" (print = ours, short). graho 'liṭi dīrghaḥ (SK 2562) requires grahīṣīṣṭa, as Wikisource reads.
• 2761: "aśīrḍhvam" (print = ours, with ś). The root gṝ requires agīrḍhvam, which is BM's gloss and Wikisource's reading. The step "[i for ṝ]" is supplied in brackets from BM. Neither ours nor the print has it; Wikisource has "itvaṃ".
• 2760: "…ābhāvān na tasmin ātvam" (print = ours). It is construed as "na tasmin [niṣedhaḥ]; ātvam" because jagle needs ā. Wikisource lacks the na.
• 2765: "māso māsaṃ vā" (print = ours) against "māse" (Wikisource, BM). The English keeps māsam.
• Every mūla correction listed under Print readings stands in ours unedited (source.json has not been changed).

Ordering: strict SK order. 2765 is split across groups: num for labh is in F, and the dvikarmaka verse and the time-object rule are in G.

A. yak in bhāva and karma: the impersonal third singular — running example: bhū + laṭ (bhāva) → ta (taṅ, SK 2679) → bhū + yak + te → bhūyate; tvayā mayā 'nyaiś ca bhūyate

• 2756 sārvadhātuke yak 3.1.67 — yak comes after a dhātu when a sārvadhātuka denoting bhāva or karma follows. dhātoḥ is carried forward from SK 2629 and bhāvakarmaṇoḥ from 2758 (BM, TB).
• In bhūyate, bhū is the dhātu, te (the taṅ ending in bhāva) is the sārvadhātuka trigger, and ya(k) goes between them.
• The bhāva-lakāra is third person and singular only (see the threads above).
• The agent, unexpressed, is in the third case: tvayā mayā 'nyaiś ca bhūyate ("there is becoming by you, by me and by others").
• liṭ is not sārvadhātuka, so it has no yak: babhūve.

B. ciṇ in the luṅ and the ciṇvat-treatment, with a survey of yak forms — running example: bhū + luṅ (bhāva), ta → cli → ciṇ → (vṛddhi, ta luk) → abhāvi; bhū + tās (ciṇvat: vṛddhi + iṭ) → bhāvitā / bhavitā

• 2757 syasicsīyuṭtāsiṣu bhāvakarmaṇor upadeśe 'jjhanagrahadṛśāṃ vā ciṇvad iṭ ca 6.4.62 — when sya, sic, sīyuṭ or tās follows and bhāva or karma is understood, the aṅga of a root with an ac in upadeśa, and of han, grah and dṛś, is optionally treated as before ciṇ (ciṇīva, vati on a seventh-case word), and sya etc. take iṭ.
• Construal (BM): "bhāvakarmaṇoḥ" is a seventh case of domain (viṣayasaptamī). The iṭ belongs to sya etc., not to the aṅga. There is no yathāsaṃkhya.
• In bhāvitā:
• the ū of bhū is the upadeśa ac;
• tās is the trigger;
• treated as before ciṇ, the ū takes vṛddhi (bhāv-) and tās takes the joint iṭ: bhāvitā.
• Without the treatment, the valādi iṭ and guṇa give bhavitā. Likewise bhāviṣyate / bhaviṣyate and āśīrliṅ bhāviṣīṣṭa / bhaviṣīṣṭa.
• There is no ciṇvat-treatment before a sārvadhātuka: bhūyatām, abhūyata, and vidhiliṅ bhūyeta ("ārdhadhātuke" qualifies sīyuṭ).
• 2758 ciṇ bhāvakarmaṇoḥ 3.1.66 — ciṇ replaces cli before ta denoting bhāva or karma. cleḥ is carried forward from SK 2222 and te from SK 2513 (BM, TB).
• In abhāvi, cli of the luṅ is the target, and ta (3rd singular, in bhāva) is the trigger. ciṇ gives vṛddhi of ū, and ta is deleted (ciṇo luk, 6.4.104, SK 2329): abhāvi.
• Before ātām there is no ciṇ, so sic with the optional ciṇvat-treatment: anvabhāviṣātām / anvabhaviṣātām (BM: why "te"). lṛṅ: abhāviṣyata / abhaviṣyata.
• Karma-lakāra. The karma is expressed by the tiṅ, so it is not in the second case:
• anubhūyate ānandaś caitreṇa tvayā mayā ca;
• anubhūyete [sukhaduḥkhe], anubhūyante [sukhāni];
• tvam anubhūyase, aham anubhūye;
• anvabhāvi.
• Causative. ṇi is elided before yak: bhāvyate; bhāvayāṃcakre / -babhūve / -āmāse.
• In -āmāse there is no h for the s of ās by ha eti (SK 2250). Like tās, as takes h only before a sārvadhātuka e, and this e (eś for ta, e for the first-person iṭ) is ārdhadhātuka.
• bhāvitā: the ciṇvad-iṭ is ābhīya and so asiddha, so "aniṭi" does not bar the ṇi-elision (ṇer aniṭi, 6.4.51, SK 2313). With the valādi iṭ: bhāvayitā.
• Likewise bhāviṣyate / bhāvayiṣyate, bhāviṣīṣṭa / bhāvayiṣīṣṭa and abhāviṣātām / abhāvayiṣātām.
• Desiderative: bubhūṣyate, bubhūṣāṃcakre, bubhūṣitā.
• Intensive: bobhūyyate. With yaṅ-luk: bobhūyate, bobhāvitā / bobhavitā.
• Stem rules before yak:
• lengthening by akṛtsārvadhātukayor dīrghaḥ (7.4.25, SK 2298): stūyate viṣṇuḥ; stāvitā / stotā, astāvi;
• guṇa by guṇo 'rtisaṃyogādyoḥ (7.4.29, SK 2380): aryate, smaryate. Guṇa is para and nitya, so it comes first; the ciṇvad-iṭ still applies "because upadeśa is mentioned": āritā / artā, smāritā / smartā;
• no guṇa in saṃskriyate, since "nitya" is carried into SK 2380 and kṛ is saṃyogādi only with the suṭ;
• n-elision by aniditām… (6.4.24, SK 415): srasyate, but idit nandyate;
• saṃprasāraṇa: ijyate;
• ayaṅ yi kṅiti (SK 2649): śayyate.

C. Two root-specific adjustments, to yak and to ciṇ — running example: tan + yak → (ā for n, optional) → tāyate / tanyate; anu + tap, luṅ → (no ciṇ) sic → anvatapta pāpena

• 2759 tanoter yaki 6.4.44 — before yak, ā optionally replaces the final of tan. āt is carried forward from SK 2982 and vibhāṣā from ye vibhāṣā (SK 2319) (BM, TB).
• In tāyate, the n of tan is the final, and yak is the trigger: tan → tā-ya-te. In the other option: tanyate.
• By ye vibhāṣā (SK 2319) itself: jāyate / janyate.
• Why "yaki"? The y of yaṅ does not trigger it: taṃtanyate (TB).
• 2760 tapo 'nutāpe ca 3.1.65 — for tap, ciṇ does not replace cli when the karma is agent (karmakartṛ, drawn in by ca from SK 2768) and in the sense of remorse (anutāpa).
• anvatapta pāpena. There are two readings:
• luṅ in the karma. Sin is the agent, and the meaning is "he was struck by it" (anu + tap made transitive by the upasarga; BM, Wikisource).
• luṅ in bhāva. pāpena = "by a sinful man", and the meaning is "there was grieving" (TB; BM).
• The karmakartṛ example is given in p64 at 2771 (atapta).
• Two more forms before yak:
• ī by ghumāsthā… (6.4.66, SK 2462): dīyate, dhīyate.
• In ādeca… (SK 2370), "aśiti" is a karmadhāraya: "before an affix whose initial ś is an it". eś has no initial it-ś, so there is no prohibition and ā is substituted: jagle (glai, liṭ).

D. yuk for ā-finals, and the ciṇvat forms of han, grah, dṛś and gṝ — running example: dā + tās (ciṇvat) → dā-y-i-tā (yuk) → dāyitā / dātā; dā + luṅ ciṇ → adāyi

• 2761 āto yuk ciṇkṛtoḥ 7.3.33 — an aṅga ending in ā takes the augment yuk before ciṇ, or before a kṛt marked with ñ or ṇ. ñṇiti is carried forward from SK 254 and qualifies kṛt only (BM, TB).
• In dāyitā, the ā of dā is the aṅga-final. tās is the trigger, treated as ciṇ by 2757; yuk is one of the operations that treatment brings (the kārikā's "yuk"). So dā-y-itā.
• Also:
• dāyiṣīṣṭa / dāsīṣṭa, adāyi, adāyiṣātām;
• without the ciṇvad-iṭ, sthāghvor icca (1.2.17, SK 2389) gives adiṣātām;
• adhāyiṣātām / adhiṣātām, aglāyiṣātām / aglāsātām.
• Counter-examples (TB): cauḍiḥ, bālākiḥ and dadau (not ciṇ/kṛt); pānīyam (not ñit/ṇit).
• han (hanyate):
• with the ciṇvat-treatment there is no t for n (hanas to 'ciṇṇaloḥ, 7.3.32, SK 2574), and kutva by ho hanteḥ… (7.3.54, SK 358): ghānitā / hantā, ghāniṣyate / haniṣyate;
• in the āśīrliṅ, the clause "ārdhadhātuke sīyuṭi" makes the ciṇvat-treatment the apavāda of vadha (hano vadha liṅi, 2.4.42, SK 2433): ghāniṣīṣṭa; otherwise vadhiṣīṣṭa. BM explains this by the rule's four clauses (niravakāśa); TB says sīyuṭ is ātmanepada-only while liṅ belongs to both padas. This is recorded neutrally;
• luṅ: aghāni; aghāniṣātām / ahasātām, or with vadha avadhi, avadhiṣātām;
• vadha itself is never extended: it is not an aṅga operation.
• grah (gṛhyate): the ciṇvad-iṭ is not lengthened. graho 'liṭi dīrghaḥ (SK 2562) lengthens only the valādi iṭ in context. So grāhitā / grahītā, grāhiṣyate / grahīṣyate, grāhiṣīṣṭa / grahiṣīṣṭa (read grahī-; see Still open), agrāhi, agrāhiṣātām / agrahīṣātām.
• dṛś (dṛśyate): adarśi, adarśiṣātām. sic is kit (liṅsicāv ātmanepadeṣu, 1.2.11, SK 2300), so there is no am: adṛkṣātām.
• gṝ, luṅ dhvam: 104 forms.
• Three bases: agāridhvam (ciṇvad-iṭ, unlengthened), and agarīdhvam / agaridhvam (valādi iṭ, optionally long by vṝto vā, SK 2391).
• Five optional operations multiply them to 96: latva, ḍhatva and three doublings.
• With no iṭ (liṅsicoḥ, SK 2528): kit by uś ca (SK 2368), [i for ṝ], r, lengthening by hali ca, and ḍh obligatorily: agīrḍhvam. Doubling ḍh, v and m gives 8 more, so 96 + 8 = 104. The verse "iḍ dīrghaś ciṇvadiḍ…" sums it up.
• Causative in the karma: śamyate moho mukundena (ṇi elided before yak). A hetumat-ṇyanta is always transitive (BM).

E. The upadhā vowel before ciṇ: optional length for mit roots, no vṛddhi for m-final udātta roots — running example: śam + ṇic + tās (ciṇvat) → śāmitā / śamitā / śamayitā; śam + luṅ ciṇ → (no vṛddhi) aśami

• 2762 ciṇṇamulor dīrgho 'nyatarasyām 6.4.93 — when a ṇi followed by ciṇ or ṇamul follows, the penultimate of a mit root is optionally long. ṇau comes from SK 2604, upadhāyāḥ from SK 2364 and mitām from SK 2568 (BM).
• In śāmitā, the a of śam is the upadhā of a mit root, and the ṇi before the ciṇvad-iṭ is the trigger: optionally long, śāmitā. In the other option it stays short, śamitā. With no ciṇvat-treatment, śamayitā.
• Also śāmiṣyate / śamiṣyate / śamayiṣyate.
• Why "dīrgha" and not just "the short vowel optionally"? With ṇi after a ṇyanta, the elided first ṇi would be sthānivat and intervene in a rule of shortening. na padānta… (SK 51) bars sthānivattva in a rule of lengthening.
• Bhāṣya: the pūrvatrāsiddhīya principle makes the "dīrgha" superfluous, and "ṇau" is jātipara. Which "dīrghagrahaṇa" is dropped is disputed: BM says the one in na padānta-, TB the one in this rule. TB then dissents (heḍ: ahiḍi / ahīḍi). This is recorded neutrally.
• Yaṅanta + ṇic: śaṃśamyate; śaṃśāmitā / śaṃśamitā / śaṃśamayitā. On the Bhāṣya's view there is no lengthening with the ciṇvad-iṭ, since the elided a of yaṅ is sthānivat. TB: since the results differ, the rejection is not right.
• Without ṇic the root is intransitive, so the lakāra is in bhāva: śamyate muninā.
• 2763 nodāttopadeśasya māntasyānācameḥ 7.3.34 — for an m-final root that is udātta in upadeśa, except ācam, there is no vṛddhi of the upadhā before ciṇ or a ñit/ṇit kṛt. vṛddhiḥ comes from SK 2473, upadhāyāḥ from SK 2282, ciṇkṛtoḥ from 2761 and ñṇiti from SK 254 (BM).
• In aśami, the a of śam is the upadhā, and ciṇ is the trigger. ata upadhāyāḥ (7.2.116, SK 2282) would give vṛddhi, and this rule bars it: aśami, adami.
• Counter-examples:
• agāmi: gam is anudātta in upadeśa;
• avādi: vad is not m-final;
• ācāmi: excepted.
• vārttika (k2) anācamikamivamīnām: kam and vam are also excepted.
• akāmi: kam without ṇiṅ by āyādayaḥ… (3.1.31, SK 2305). With ṇiṅ or ṇic the form is the same, since ṇi is elided.
• avāmi.
• avadhi: consonant-final vadh, no vṛddhi by janivadhyoś ca (7.3.35, SK 2512).
• ajāgāri: jāgro 'viciṇṇalṅitsu (7.3.85, SK 2480) excludes ciṇ, so there is no guṇa and vṛddhi stands.

F. The root's nasal before ciṇ: n-loss for bhañj, num for labh — running example: bhañj + luṅ ciṇ → (n elided, optional) abhāji / abhañji; labh + ciṇ → (num, optional) alambhi / alābhi

• 2764 bhañjeś ca ciṇi 6.4.33 — for bhañj before ciṇ, the n is optionally elided. na comes from śnān nalopaḥ (SK 2544) and vibhāṣā from jāntanaśāṃ vibhāṣā (SK 3330); see the audit list above.
• In abhāji, the ñ (n) of bhañj is the target, and ciṇ is the trigger. With n gone, the upadhā a takes vṛddhi: abhāji. Otherwise abhañji.
• No TB (print).
• 2765 vibhāṣā ciṇṇamuloḥ 7.1.69 — labh optionally takes the augment num before ciṇ or ṇamul. labheḥ comes from SK 2582 and num from SK 2262 (BM, TB).
• In alambhi, labh takes num, and the trigger is ciṇ: alambhi. Without num, the upadhā takes vṛddhi: alābhi.
• The option is vyavasthita. After an upasarga num is obligatory: prālambhi (BM; the Bhāṣya carries "upasargāt" and "na").
• ṇamul: labhaṃlabham / lābhaṃlābham, but only pralambhaṃpralambham (TB). The audit notes that TB's "labhaṃlabham" (= print) should be lambhaṃlambham.

G. 2765 (continued): which karma the lakāra goes to with dvikarmaka roots, and with time-objects — running example: duh (gauṇa karma gauḥ expressed) → gaur duhyate payaḥ

• 2765, the verse. "gauṇe karmaṇi duhyādeḥ pradhāne nīhṛkṛṣvahām | buddhibhakṣārthayoḥ śabdakarmakāṇāṃ nijecchayā | prayojyakarmaṇy anyeṣāṃ ṇyantānāṃ lādayo matāḥ". The lakāras (and the kṛtya, kta and khalartha affixes, per BM and TB) go to:
• the secondary karma for duh etc. BM and TB list twelve: duh, yāc, pac, daṇḍ, rudh, pracch, ci, brū, śās, ji, math, muṣ. The gauṇa karma is the one named by akathitaṃ ca (1.4.51, SK 539). Example: gaur duhyate payaḥ. gauḥ, the gauṇa karma, is expressed and in the first case; payaḥ, unexpressed, is in the second.
• the principal karma for nī, hṛ, kṛṣ and vah: ajā grāmaṃ nīyate, hriyate, kṛṣyate, uhyate.
• either karma, at choice, for roots of cognition, eating and roots whose karma is speech: bodhyate māṇavakaṃ dharmaḥ or māṇavako dharmam.
• the prayojya karma for the other roots in ṇi (roots of motion, intransitives, and hṛ and kṛ of SK 541): devadatto grāmaṃ gamyate ("is made to go").
• 2765, time-objects. Intransitive roots with time etc. as karma take the lakāra in the karma or in bhāva: māso māsaṃ vā āsyate devadattena. BM calls māsam a wrong reading (apapāṭha) for māse; Wikisource reads māse; the print reads māsaṃ.
• From the ṇijanta the lakāra goes to the prayojya: māsam āsyate māṇavakaḥ.
• Then [iti bhāvakarmaprakriyā].