Chapter 53 - The Exposition of Nirukta

[Sanskrit text for this chapter is available]

[Note: Being a technical and difficult subject, an explanation is added in the translation itself as mere literal translation would be unintelligible]

Sanandana continued:

1. I shall now delineate to you the excellent accessory (śrotrāṅga) of the Veda, called Nirukta. It is said to be of five kinds (pañcavidha) and pertains to Vedic conjugations (vaidika) or verbal formations (dhāturūpaka).

2. Therein at some places is found epenthesis (varṇāgama), in other cases is seen metathesis (varṇaviparyaya), at some places is found modification (vikāra) of the syllable (varṇa) [=varṇavikāra], at other places, (we get) haplology (varṇanāśa).

3. O Nārada, where due to the change (vikāra) or loss (nāśa) of a syllable (varṇa), light is thrown on the special connotation of the root (dhātu) it is called Saṃyoga.

4. (Now follow the illustrations of these processes:) By addition of a syllable (varṇāgama), the word Haṃsa (< √han—) ‘a swan’ is formed^[1] (by the addition of sa). The word Siṃha^[2] ‘a lion’ is formed by metathesis, gūḍhotmā (gūḍha + ātmā) is formed due to change of syllable (here the vowel ā is changed to u) while in pṛṣodara (pṛṣat + udara) there is a loss of a letter (t).

5. The fifth kind of modification should be known in words like bhramara ‘a bumble bee’ (from √bhram + ara by the uṇādi sūtra artikamibhramicami etc., cf. √kam—> kamara, √cam—> camara).

In the Vedic (language) there are changes or options in the formations of words etc. Here is the example of Punarvasū. (In the Vedas, the two stars Punarvasū may optionally be singular and connote a dual vide, Pāṇini I. 2. 61).

6. The form nabhasvat (nabhas + vat ‘like the sky’, ‘vaporous’, retains ½ S though followed by—vat (vide the Vārttika—‘nabho'ṅgiromanuṣāṃ vatyupasaṅkhyānam on Pāṇini I.4.9 or by the Vārttika Vṛṣaṇvasvaśvayoḥ/^[3] The form Vṛṣaṇśva retains ½ n ½ even though the derivation is vṛṣā aśvo yasya saḥ, ( whereas in classical Sanskrit it would be Vṛṣāśva. Sometimes the Parasmaipada is used in stead of the Ātmanepada (In Classical Sanskrit, the prepositions or proverbs like pra-, ā-precede the verb and are called gati). But in the Vedic, these pre-positions are used after the verbs (as post-positions). They are called gatis (e.g. haribhyām yāhyoka ā Ṛg Veda VII.32.4) Here ā-yāhi is the verb but ā- is placed separately after the verb in the Vedic.

7. (In the Vedic) there is a change in the cases. For example, dadhnā juhoti, ‘He offers oblations of curds (into the fire)’. Here the word dadhi is in the Instrumental case in stead of the Acc. case. But as per Pāṇini II.3.3. the object of the verb √hu—‘to sacrifice’ takes the terminations of the Instr, as well as of the Acc. case.

The Aorist form of abhi + ut + √sad— is abhyutsādayā-maka (instead of Classical Sanskrit—abhyudasīṣadat) in the Vedic (as per Pāṇini III.1.42), and dhvanayīt (<√dhvan ‘to sound’) as per Pāṇ III.1.51 (in Classical Sanskrit adidhvanat is found. But in Ṛg Veda I.162.15 we find: mā tvā'gnir dhvanayīt). These arc the prominent instances. The forms like niṣṭarkya (> nis + √kṛt— ‘to cut’) and others are similar^[4] And also formations like gṛbhāya and others. Here √gṛbh + śāyac √ gṛbhāya, the—h in grah ‘to seize’ is √ transformed to bh i.e. √grah—< √gṛbh—in Vedic (vide Pāṇini III.1.84 and S.K. on it.^[5] e.g. gṛbhāya jihvayā madhu (Ṛg Veda VIII.17.5).

8. Due to optional formations (in the Vedic) the writers on this science desire (accept) changes in declension, conjugation, Padas (Parasmaipada and Ātmanepada,) gender, person, tense, consonants, vowels, causatives and yaṅ (the intensive suffix—ya).

9. The verse lists the following peculiar Vedic forms:

rātri: rātri + ṅip in the Vedic as per Pāṇini IV.1.31 for class. Sanskrit rātri ‘a night’.

Vibhvī: fem. of vibhu as per Pāṇini IV 1.47 we get vibhu+ ṅiṣ (i.e. the affix—ī) in forming the feminine.

Kadrū: As per Pāṇini IV.1.71 the feminine affix—Ūṅ (—Ū) comes after the words Kadrū ‘tawny’ (and kamaṇḍalu ‘water bowl’).

āviṣṭya: As per Vārttika on the above sūtra the affix—lyap is added to the word āvis in the Veda e.g. āviṣṭyo vardhate /

Vājasaneyinaḥ: As per Pāṇini IV.3.10.6 the affix—ṇini comes in the sense of “enunciated by him”, after the word Śaunaka e.g. (Śaunakinaḥ “who study the chandas as per Śaunaka.” Similarly, Vājasaneya + ṇini = Vājasaneyin: who studies the Veda as per Vājasaneya.

10. The following Vedic forms are explained as below: Karṇebhiḥ: Instr. Pl. of Karṇa “an ear”. The termination —ebhiḥ instead of the usual—ais as per bahulaṃ chandasi (Pāṇini II.4.76.).

yaśobhagya: (yaśo-bhāgya—in the printed text of the Nārada Purāṇa is not correct) is dissolved as follows : yaśas tad eva bhagaḥ: This Karmadhāraya compound has bhaga changed to bhagya by the addition of the suffix—yat as per Pāṇini IV.4.131,32.

yaśobhagya and others include forms like veśobhagya. caturakṣaram indicates four-lettered formulas like āśrāvaya astu śrauṣaṭ.^[6]

devāsaḥ: The Nom. pl. of deva ‘a god’ (vide Pāṇini VII. 1.50) surva-devatāti: sarva-deva + tātil (pleonastic) e.g. savitā naḥ suvatu sarvatātim /—Ṛg Veda X.36.14 tvāvat ‘like you’. Here, as in asmad & yuṣmad, the affix—vatup showing similarity is added.

11. The following forms are formed thus:

ubhayāvinam: Here the possessive affix vini (i.e. vin) is added (optionally) as per Pāṇini V.2.122 and the final—a in Ubhaya is lengthened as per Vārttika: Chandovit prakaraṇe dīrghaś ceti vaktavyam /

Pralayādyāḥ is prob, a misprint.

Tṛcam: tri—Ṛc (a Sūkta with three Ṛks). This form is derived, thus: tri > tṛ by Samprasāraṇa + C < Tṛc.

apaspṛdhethām^[7]

avyāt—^[8] as in vasubhir no avyāt ayo asman mukhas tathā / If cyo be a misprint for āpo, the quotation in the Nārada Purāṇa is probably:

āpo asmān mātaraḥ śundhyantu / Ṛg Veda X.7.10 and the form is as per Pāṇ VI.1.118, the short a is retained after āpo (also juṣāṇo[?], Vṛṣṇo etc.)

12. The following forms are derived thus:

Sagarbhyaḥ: ‘a brother of the whole blood, by the same father and mother’. Here sa = samāna ‘common’ but the affix—yaṅ (—ya) is as per Pāṇini IV.4.114.

sthāpadi or asthāpadī (a misprint prob, for aṣṭāpadī?) Ṛtvyam ‘seasonal’ ṛtau bhavaṃ. This formation from ṛtu is as per Pāṇini VI.4.175 the final—u is changed to—va (similarly vāstva < vāstu)

Rajiṣṭha: ṛju > iṣṭha ‘the most straight-forward & delicate’ vide Pāṇini VI.4.162 for the change of ṛ > ra—before the affixes —īyas,—iṣṭha

tri-pañcakam:—trīṇi pañcakāni yatra bahuvrīhi comp.

Hiraṇyaya—‘golden, abounding in gold’

Hiraṇya + ya (mayaṭ)

Hiraṇyayena is a part of the mantra hiraṇyayena savitā rathena (Ṛg Veda I.35.2)

naram—(misprint? otherwise no Vedic speciality)

parame vyoman: instead of Vyomni, the Loc. sg. term is dropped as per supām suluk /

13. The explanation of the following irregular forms:

urviyā (Sanskrit uru-ṇā) But —yā is affixed instead of—nā, as per Vārttika—iyāḍiyāji etc. on Pāṇini VII.1.39.

svaprayā—(?) Is it a misprint for svapnayā? vāravadhvāt—(?) Is it a misprint for vārayadhvāt?

Aduha (?) The root √duh is obvious but the special formation—not clear.

Vadhi (?) The Vedic special form is Vadhīm &; not Vadhī. We have in Ṛg Veda 1.165.8, X.28.7 the form Vadhīm from √Vadh. The Classical Sanskrit form would be avadhiṣam.

yajadhvainam < yajadhvam + enam. These are joined after dropping—m of °dhvam.

Emasi—irregular for imaḥ by Pāṇini VII.1.46.

snātvī (as in svinnaḥ snātvī malād iva—Atharva Veda VI.115.3, VS. 20.20). Also—tvī for Classical Sanskrit—tvā vide Pāṇini VIII.1.49 (cf. pītvī < √pā—to drink)

gatvā pacāsthabhīḥ: Out of this asthabhīḥ may be a

misprint for aṣṭhabhiḥ, Instr. pl. of asthin. The rest is anybody’s guess. (possibly gatvāya cāsthabhiḥ /).

14. (The following irregular forms are explained).

Gonām—Gen. Pl. of go ‘a cow’—Pāṇini VII.1.57

Aparihvarataḥ As per Pāṇini VII 2.31 √hru is substituted for √hvar (Bhvādi 978) before a participial affix.

cāturi or āturi (?) prob. a misprint.

Grasita ppp. of √gras—(Classical Sanskrit grasta) treated as seṭ (cf. skabhita) vide Pāṇini VII.2.34.

Paśya— Nothing special.

Dadhat as in dadhad ratnāni dāśuṣe Ṛg Veda IV.15.3. It is the Vedic subjunctive of √dhā.

Babhūtha— Perfect 2nd p. sg. of √bhū in Vedic, vide Pāṇini VII.2.64.

Pramiṇānti is probably a misprint for pramiṇanti for classical Sanskrit pramīṇanti—mī—is shortened to—mi— as per Pāṇini VII.1.81

Avīvṛdhat—Aor—3rd p. sg. of √vṛdh— As per Pāṇini VII.4.8., in Vedic (Chandas), ṛ is substituted invariably for penultimate ṝ or ṛ of a causative stein in the Reduplicative Aorist.

15. (The following forms arc derived as follows:) mitrayu ‘friendly-minded’. As per Pāṇini VII.4.35, rules of lengthening or the substitution of—ī for the final vowel of the stem before the affix Kyac is not applicable in the Vedic.

Durasyu—duṣṭa ivā'carali & ‘Wishing to do harm’. As per Pāṇini VII.4.36. duṣṭa becomes duras—before the affix—kyac. [The Nārada Purāṇa gives durasvā which is prob, a mis-print).

hātvā: Abs. of √hā ‘to abandon’ √hā. > √hī (hi) optionally in Vedic (Pāṇini VII.4.44).

Sudhita: ‘well-placed’ ppp. of Su—√dhā. Pāṇ. VII.4.45 sanctions √dhā > ppp -dhita in Sudhita, Vasudhita, nemadhita.

Dadharti should have been dādharti. As per Pāṇini VI.1.4.65 in Vedic the Caus. Aor. 3 p. sg. of √dhṛ becomes dādharti cf. dardharti, dardharṣi sanctioned in the same Sūtra.

Svavadbhih: Instr. pl. of Svavas. As per Vārt. on Pāṇini VII.4.45,—t is substituted for -s in the case of Svavas. ‘He whose proteclion is good (derived from Su—√av) ‘to protect well’.

Sasūva.—Perf.—3 p. sg. of √sū ‘to give birth to’. Formation as per Pāṇini VII.4.74.

Dhiṣva—Imp. II p. sg. of √dhā— as per Pāṇini VII.4.45 (but Classical Sanskrit form is dhatsva).

16. (the following forms are explained thus:)

praprāya as in praprāyam agniḥ (Ṛg Veda VI.1.8.4) shows the reduplication of the prefix pra—as per Pāṇini VIII.1.6 Pāṇini sanctions pra, sam, upa and ut as pādapūraṇas (the completers of the foot of a verse).

Harivate—Here v is substituted for m of the affix—mat, as the stem Hari ends in—i. Pāṇini VIII.2.15 sanctions this change for stems ending in—i—ī or ṛ.

Akṣaṇvaśaḥ (a misprint for akṣaṇvantaḥ) vide Ṛg Veda X.71.7 (akṣaṇvanṭaḥ karṇavantaḥ sakhāyaḥ) /

A—kṣaṇ—vat ‘not injuring’. As per Pāṇini VIII.2.16. The augment—nuṭ is added to the possessive affix, after stems ending in—an.

Supathi—tara: Supathin + tara *> Supathi-tara (as per Pāṇini VIII.2.7) and —n ṇuṭ is augmented to tara by Pāṇini VIII.2.17.

Rathī-tarī: ‘rathin—tarī > rathi’ (Pāṇini VIII.2.7) & lengthened to -rathī + tarī as per Vārttika on VIII.2.17. nasat (? Is it nasatta < na—√sad for Pāṇini VIII.2.61 sanctions nasatta and niṣatta.

Amnar: The original word is amnas ‘unawares’. According to Pāṇini VIII.2.70, it becomes amnar before Sandhi e.g. amnareva.

Bhuvaratho: Pāṇini VIII.2.71. states that in Chandas—ru or—ra replace the final of the word Bhuvas.

17. (the special formations recorded in this verse)

Brūhyādeḥ: According to Pāṇini VIII.2.91 the first syllable of brūhi, preṣya, śrauṣaṭ, vauṣaṭ, and āvaha is pluta or prolated. Thus we say: agnaye'nubrū3hi / (By the next Sūtra) both the syllables become prolated or Pluta as in ‘O3śrā3vaya’) when the Adhvaryu (the chief sacrificial) priest gives order to the Āgnidhra priest (who is to kindle the fire.).

Dāśvān—p.pf. from √daś— ‘to honour’ √daś + Kvasu but here the reduplication, and the augment—i— are prohibited as per Pāṇini VI.1.12.

Svatavān— nom. sg. of sva-tavas ‘inherently powerful’. This formation takes place in case of yāpau but yāpau is obscure, (a misprint probably for Pāyu?)

Tribhiṣṭvam < tribhis + tvam; As per Pāṇini VII.3.103—S coming within the (Vedic) metrical foot is changed—Ṣ before the t—of tvam and of tatakṣuḥ.

nṛbhiṣṭutaḥ < nṛbhis + stutaḥ. Here ending—s of nṛbhis is changed to—ṣ— as per Pāṇini VIII.3.105 the cerebralization of st—> ṣṭ—in. ṣṭutaḥ and stoma in go-ṣṭoma (< go + stoma) is allowed.

18-19. (The following forms arc derived as explained below).

Abhīṣuṇaḥ— As per Pāṇini VIII. 3.107, the s > ṣ—of the particle—ṣu is permissible when it stands in the beginning of a second word and preceded by a word ending in ī etc. (in the Vedic) e.g. madhusthānam: madhuṣṭhānam.

Ṛtāvāha— obscure. Is it a misprint for Ṛtāṣāha, mentioned by Pāṇini in Saheḥ pṛtanartābhyāṃ ca? (Pāṇini VIII.3.109).

nyaṣīdat—Here—s—>—ṣ—in this form (though—s is preceded by ā) is permissible as per Pāṇini VIII.3.119.

Nṛmaṇāḥ—As per Pāṇini VIII.4.26, in the Vedic na of a second term is changed to ṇa when preceded by ṛ—even though there is a hiatus (Avagraha) between the two terms.

Optionals are of four types: sometimes it shows a tendency to add, sometimes to withhold, sometimes both are optional and sometimes a new substitute is added. Such is the word formation in the Vedic.

√bhū (I. conj.), √vā (II conj).—These are the dhātus (roots indicating different conjugations). They are remembered as being Parasmaipadī.

20.^[9] Thirty-six roots (dhātus) beginning with √edh— ‘to prosper’—are Udātta (of acute accent) and Ātmanepadī.

And thirty-eight dhātus beginning with √at:—‘to go constantly’ or ‘to obtain’ are Parasmaipadī.

21. Forty-two dhātus beginning with √śī-, √kṛ-, are spoken, of as belonging to Ātmanepada. Fifty dhātus beginning with √phakk—‘to move slowly’ are laid down as udāttet i.e. Parasmaipadī.

22. Twenty-one roots beginning with √varc—‘to shine’ are spoken as Anudāttet or Ātmanepadī, while the group of forty-two dhātus beginning with √gup—‘to protect’ are said to be udāttet (parasmaipadī).

23. Grammarians have stated that ten Dhātus beginning with √ghiṇ—‘to take, to grasp’ (cf. Pkt. √geṇha-) are anudāttet (Ātmanepada)^[10] while twenty-seven roots^[11] beginning with √aṇ—(‘to breathe’ are also Udāttet (Parasmaipada).

24. Grammarians have regarded thirty-four^[12] Dhātus beginning with √ay^[13] —‘to go’ as Anudāttet (Ātmanepada), while seventy-two (in all) Dhātus^[14] with √mavy^[15] ‘to bind’ as its first are Parasmaipadīs.

25. √dhāv—‘to run’ is declared as the only root which is svaritet while fifty-two Dhātus beginning with √kṣudh^[16] —‘to be hungry’ are spoken as Anudāttet (Ātmanepadī).

26. Eighty Dhātus^[17] beginning with √ghuṣ—‘to make noise’, ‘to proclaim’ are regarded as Udāttet (Parasmaipada) while twenty-two Dhātus^[18] from √dyut—‘to shine’ are considered as Anudāttet (Ātmanepada).

27. In the group beginning with √ghaṭ—the thirteen roots are spoken as ṣit^[19] and Anudāttet (Ātmanepada). Thereafter fifty-two^[20] dhātus from √jval—‘to burn’ are regarded as Udāttet (Parasmaipadī).

28. √rāj—to shine is both Udātta and Svarita.^[21] While the three roots √bhrāj, √bhrāś, √bhlāś, are spoken as Anudātta (Ātmanepada), while Dhātus from √syam—‘to shout’ No. 851) onwards are Udāttet (Parasmaipadī).

29. There only one Dhātu √sah—‘to bear’ (recorded as √ṣah in the Dhātupāṭha is Anudāttet and √ram—‘to play’ is Ātmenapadī. Thereafter three dhātus beginning with √ṣad (or √sad—) are udāttet, while four Dhātus e.g. √kuc—and others are Parasmaipadī.

30. Thereafter thirty-five Dhātus beginning with √hikk—‘to make an indistinct sound’ (Dhātupāṭha—ibid Nos. 886—921 up to √guh—‘to conceal) are Svaritet or both the Padas.^[22] Similarly √śrī—‘to serve’ and the four roots √bhṛ-, √hṛ-, √dhṛ- and √nī—are also Svaritet.

31. Forty-six Dhātus^[23] from √dhe—‘to drink’ are said to be Parasmaipadī, while eighteen^[24] dhātus from √smi—‘to smile’ are opined to be Ātmanepadī

32. Three roots beginning with √pū—‘to purify’ (Dhātu nos. 991-993) are described as Anudāttet (Ātmanepadī). √hṛ— ‘to take away’ is Parasmaipadī, while the three roots √gup—‘to protect’ (nos. 995-998) are Ātmanepadī.

33. √rabh—, √lahh—, √svaj—, √had—) these four roots are Anudāttet, while √kṣvid—‘to be wet’ is Udāttet and fifteen roots from √skambh—(Dhātupāṭha records √skand—) are Parasmaipadī.

34. √kit—‘to desire’ is Udāttet (Parasmaipadī) while √dān—(‘to cut” and √śaṇ—to shine are Ubhayapadī (i.e. are found both Parasamaipadī and Ātmanepadī). Nine roots beginning with √pac—‘to cook’ (Dhātu nos. 1021-1029 from √pac— to √vah— ‘in Pāṇini Dhātupāṭha) are Svaritet. They are regarded as Parasmaipadī, (and Ātmanepadī) as well.

35. Three more dhātus are Svaritet—.√vad— and √vac—both meaning ‘to speak’ are Parasmaipadī. In this way, one thousand and six are said to be the roots of the group of Dhātus beginning with √bhū (i.e. of the 1st conjugation).

Dhātus of the 2nd conjugation

36. √ad^[25] —‘to eat’ and √han—‘to kill or to go’ are spoken of as belonging to Parasmaipada). And the four Dhātus beginning with √dviṣ—‘to hate’ (i.e. Dhātu Nos. 3-6 of Pāṇini Dhātupāṭha) are regarded as Svaritet (both Par. pada and Ātmanepada).

37. Here the Dhātu √cakṣ ‘to speak’ is regarded as Ātmanepadī,^[26] while thirteen roots from √īr—‘to go’ (Nos. 8-20 in Pāṇini Dhātupāṭha) are Anudāttet (Ātmanepada).

38. O sage! grammarians have proclaimed √sū— to give birth to and √śī—‘to sleep’ as Ātmanepadīs, while seven Dhātus beginning with √ṣū (it should be √yu as per Dhātupāṭha).

39. O lord of sages! Only one root viz. √ūrṇu ‘to cover’ is said to be svarita (i.e. both Parasmaipada & Ātmanepada), while three dhātus beginning with √ghu (a misprint for √/dyu as evidenced in the Dhātupāṭha—Dhatu No. 30) are spoken as Parasmaipadīs.^[27]

40. O Nārada! √stu—‘to praise’ is the only root pronounced and remembered as Parasmaipada. by the grammarians.

41. Eighteen Dhātus beginning with √rā—‘to give’ (prob, from √yā ‘to go’ to √vac ‘to speak’, in the Dhātupāṭha) are Parasmaipada O Nārada! Only √i—‘to study’ (adhyavane) is declared as Ātmanepada).

42. Four Dhātus from √vid—‘to know’ to √rud—‘to weep’, nos. 54-57 in the Dhātupāṭha are considered as Parasmaipadī.

√svap—‘to sleep’ is pointed as Parasmaipada.

43. √śvas—‘to breathe’ etc. (svamādayaḥ in the Nārada Purāṇa is probably a misprint as per series of roots in the Dhātupāṭha) have been told by me as Parasmaipadī.

√dīdhī—‘to shine’ and √vevī—‘to go’ (rarely used in Classical Sanskrit) are Ātmanepadī, O sage.

44. Three roots beginning from √pṛth—(prob. a misprint for √prath—1st conj. are Ātmanepada and the Pāṇini Dhātupāṭha reads: ‘68 √ṣasa, 69 √sasti svapne, 70 √vaśa kāntau’ as the three Dhātus in the serial list of roots of 2nd conj. are proclaimed as Udātta (Parasmaipada)^[28]. Carkarīta^[29] is a term for any intensive formed without the syllable ya like carkarīti < √kṛ—.

√hnu—‘to cover’ is declared as Anudātta, O excellent sage.

45. Thus seventy-three roots have been said to be included in the 2nd conjugation (adādi gaṇa).

(Now begins the list of the 3rd conjugation—juhotyādayaḥ)

Dhātus of the III conjugation

The four roots from √dā (√hu?) are regarded as Parasmaipadī.^[30]

47. √bhṛ is svaritet (both Parasmaipada & Ātmanepada) & ‘ohāk’ (√hā ‘to abandon’) is udāttet.

√mā—‘to measure’ and √hā— ‘to go’ are Anudāttet.

√dā—‘to give’; √dhā—‘to hold’, ‘to contain’ are symbolised as ‘it’ indicating their being svarita (both Parasmaipada and Ātmanepada).

47. Three roots beginning with vānitir (probably a misprint for ṇijir in the Dhātupāṭha) are spoken of as Svarita (both Parasmaipada and Ātmanepada), while twelve dhātus from √ghṛ—‘to besprinkle’ are regarded as Parasmaipadīs.

48a. Thus twenty-two roots have been mentioned in. the 3rd conjugation (hvādika gaṇa).

Dhātus of the IVconjugation

48b. Twenty-five Dhātus beginning with √div—‘to shine’ have been spoken of as being Parasmaipadī.

49. O Nārada! There are two Ātmanepadi Dhātus viz. √sū—‘to give birth to’ and √dū—‘to trouble’. The seven Dhātus beginning with √pū—are regarded as Ātmanepadī.

50. O Brāhmaṇa! Roots beginning with √dī—‘to perish’ ‘to die’ are known here as Ātmanepadī, while the four Dhātus beginning with √so—‘to kill’ are Parasmaipadī.

51. O sage! Fifteen roots beginning with √jan— ‘to be born’ (Dhātus from 43-49) are Ātmanepadī while the five dhātus beginning with √mṛṣ—‘to bear’ ‘to endure’ arc svaritet (both Parasmaipada & Ātmanepada)

52. Eleven dhātus from √pad—‘to go’ (Root Nos. 63-74) are considered as Atmanepadīs^[31] while eight Dhātus from √radh—‘to hurt’, ‘to injure’, are known as Parasmaipadī.^[32]

53. Forty-six Dhātus from √śam—‘to pacify’ are spoken of as Udāttet (Parasmaipada). Thus one hundred and forty Dhātus are regarded as belonging to the divādi gaṇa (IV conjugation).

Dhātus of the V conjugation (svādi-gaṇa)

55. Nine roots beginning with √su—‘to extract soma juice’ are described as being svaritet (both Parasmaipada & Ātmanepada), while seven roots such as dunoti (√du—‘to trouble’) are Parasmaipadīs.

56-57a. The two roots √aś—‘to pervade’ and √stigh—‘to mount over’ are known as Anudāttet (Ātmanepada), while fourteen. Dhātus beginning with √tik—‘to go’, ‘to assail’ are Parasmaipadī here. O excellent Brāhmaṇa! Thirty-two dhātus are spoken of as constituting the svādika gaṇa (V conjugation).^[33]

Dhātus of the VI conjugation (tudādi-gaṇa)

57b. O excellent sage! Six dhātus^[34] beginning with √tud—‘to trouble’ are said to be svaritet (both Parasmaipada & Ātmanepada)

58. √ṛṣ—‘to go’ is udāttet (Parasmaipada), while nine roots from √juṣ—‘to be pleased’ are Ātmanepada, and one hundred and five Dhātus^[35] beginning from √vraśc—‘to cut’ are declared as Udāttet.

59. O chief of sages! Only one Dhātu √gur—‘to make exertion’ is declared as Udāttet. Four roots beginning with √nu—‘to praise arc considered as Parasmaipadī.

60. The Dhātu √ku—‘to sound’ is spoken of as Anudāttet (Ātmanepada). Here ends the Kuṭādi gaṇa.

√pṛ—‘to be active’ and √mṛ—‘to die’ are Ātmanepadīs, while six Dhātus from √pi-, √ri—‘to move’, ‘to go’ are Parasmaipadīs.

61. Two roots viz. √dṛ—‘to honour’, and √dhṛ—‘to exist, to live’ are spoken, of as Ātmanepadīs, while sixteen. Dhātus from √pracch—‘to ask’ are Parasmaipadīs.

62. Thereafter, O sage, six roots beginning with √mil— ‘to join’ are spoken, of as svaritet (both Parasmaipada and Ātmanepada) And Three Dhātus beginning with √kṛt—‘to cut’ are Parasmaipadīs.

63a. Thus there are one hundred and fifty seven dhātus in the VI conjugation (tudādi gaṇa).

Roots of the VII conjugation

63b. Nine roots beginning with √rudh—‘to obstruct are svaritet (both Parasmaipada and Ātmanepada), Kṛtī (Veṣṭane) is parpadī.

64. Three Dhātus from √indh—‘to kindle’ (& √khid—& √vid) are said to be Anudāttet (Ātmanepada) (Twelve-Dhātus from) √śiṣ—‘to leave as remainder’, √piṣ—‘to reduce to powder’ are udāttet. In this way, there are twenty-five loots in the VII conjugation (rudhādi gaṇa).

Dhātus of the VIII conjugation (tanādi-gaṇa)

65. Seven Dhātus^[36] beginning with √tan—‘to spread’ are proclaimed as being Svarita (both Parasmaipadī & Ātmanepadīs)

√van—‘to beg’, ‘to request’ and √man—‘to think’ are Ātmanepadī and √kṛ—(technically called ḍukṛñ) ‘to do’ is Svarita (both Parasmaipada and Ātmanepada).

66a. In this way, ten dhātus have been proclaimed as belonging to the eighth conjugation.

Dhātus of the IX conjugation (kryādi-gaṇa)

66b-67. Seven Dhātus^[37] beginning with √krī—‘to buy’ belong to both the Padas. Four dhātus beginning with √stambh and which are used (by Pāṇini) in the Sūtras are Parasmaipadī. O chief of sages, twenty two dhātus from √krudh—‘to get angry’ (This is obviously a misprint for √knūñ as stated in the Dhātupāṭha) arc Udāttet (Parasmaipada).

68. √vṛ—‘to select’ is Ātmanepadī. Twenty-one Dhātus from √śranth—‘to loosen’ are Parasmaipadī while √grah—‘to take’ is svaritet (both Parasmaipadī & Ātmanepadī).

69a. In this way fifty-two roots^[38] have been enumerated in the IX conjugation by the learned.

Dhātus of the X conjugation: (gurādi-gaṇa)

69B. One hundred and thirty-six Dhātus beginning from √cur—‘to steal’ belong to (both) Parasmaipada and Ātmanepada.

70. O sage! Eighteen (but actually thirty-nine roots from 140 √cit—to 178 √kusm—) Dhātus are Ātmanepadī. O sage! Roots from √carc—‘to study’ upto √dhṛṣ (Dhātu nos. 318) are declared as belonging to both Parasmaipada and Ātmanepada.

71. Fortyeight Dhātus from √kath—(Dhātus 319-365)—known as adantas, and ten Dhātus beginning from √pad— (i.e. Dhātus 366-375) are Ātmanepadī.

72. Here eight roots beginning from √sūtr—‘to cover’ are spoken as belonging to both Parasmaipada and Ātmanepada by the wise. In the sense of a root (verb) the prātipadika gets the affix ṇic (—i) and is formed in most cases like the application of the—iṣṭha.

73-74. This affix—i is in the sense ‘does for the sake of’ or ‘tells the other’. In the sense of causative, this ṇic is added. Eight roots beginning with √cira (root no. 385 onwards) upto √saṅgrāma (root no. 390) are spoken as Anudāttet by the learned grammarians. Sixteen Dhātus from √stoma—‘to praise’, are the illustrations of adanta dhātus.

75. The meaning of “optionally” (bahulam etan nidarśanam—as stated in the Gaṇapāṭha) is that it includes all the remaining dhātus irrespective of their use in (Pāṇini’s) Sūtras, laukika Sanskrit and the Chāndasa (Vedic) language. All dhātus belong to all conjugations and have different meanings.

76. In addition to these roots, dhātus ending in the affixes beginning with —san (i.e. the twelve affixes viz.—san, —kyac, —kāmyac,—kyaṅ,—kyaṣ,—ācāra-kvip,—ṇic,—yaṅ,—yak,—āya,—iyaṅ and—niṅ.) are also called roots, and also Denominatives (e.g. Garuḍāyate—acts like a Garuḍa). In this way innumerable number of Dhātus are formed here. O Nārada. Here it is explained in a nutshell. The details are given in the relevant treatises.

77. In Dhātus ending in the vowels, all Dhātus with the exception of those ending in—ū and -ṛ and the roots √yu,√ru, √kṣṇu, √śī, √snu, √nu, √kṣu, √śvi, √ḍī, √śri, √vṛ (of both Padas)—are regarded as Anudātta or Aniṭ.

78. The following is the list of one hundred and two Dhātus ending in consonants which are Anudātta or Aniṭ.

√śakḷ, √pac, √muc, √ric, √vac, √vic, √sic, √pracch, √tyaj, √nij, √bhaj, √bhañj, √bhuj, √bhrasj, √masj, √yaj, √yuj, √ruj, √rañj, vi √ji, √svañj, √sañj, √sṛj;

79. √ad, √kṣud, √khid, √chid, √tud, √nud, √pad, (4th conj.), √bhid, √vid, vi- √nad, √śad, √sad, √svid, √skand, √had, √krudh, √budh;

80. √bandh, √yudh, √rudh, √rādh, √vyadh, √śudh, √sādh, √sidh, √man (IV conj.), √han, √āp, √kṣip, √chup, √tap, √tip, √tṛp, √dṛp;

81. √lip, √lup, √vap, √śap, √svap, √sṛp, √yabh, √rabh, √labh, √gam, √nam, √yam, √ram, (not rabhiḥ i.e. √rabh— as in the text of the Nārada Purāṇa), √kruś, √daṃs, √diś, √dṛś, √mṛś, √riś, √ruś, √liś, √viś, √spṛṣ, √kṛṣ.

82. √tviṣ, √tuṣ, √duṣ, √puṣ, (IV conj.) √piṣ, √viṣ, √śiṣ, √śuṣ, √śliṣ, √ghas, (= ad), √vas, √dah, √dih, √duh, √nah, √mih, √ruh, (√lih, √vah.)

83b-85. Ca and others are called Nipātas.^[39] The ‘pre’-positions like pra—are known as gatis (gavayaḥ in the Nārada Purāṇa should have been gatayaḥ); words which have been created in different quarters, regions and periods have a number of connotations. And, O Brāhmaṇa, they are characterised by all genders. Here Gaṇapāṭha, Sūtrapāṭha Dhātupāṭha Anunāsika Pāṭha are called Pārāyaṇa.

O Nārada! All words whether laukika (current) or Vedic—are permanent and imperishable.

86a. The vocabulary collected by grammarians is hence intended for recitation.

86b.-88. The formation of the (already formed, permanent) words is briefly explained by (the study of) its original form (the noun base or root), terminations, substitutes, the augment, etc.

In this manner a bit of Nirukta has been briefly related to you. Who is capable of recounting it in its entirety, as it is infinite (limitless), O Nārada?

Footnotes and references:

[back to top]

[1]:

Hantīti haṃsaḥ.

[2]:

√hīṃs ‘to kill’: hiṃs—a—hiṃs a & metathesis by Pāṇini VI.3.109. Bhaṭṭojī Dīkṣita quotes this very verse in the Nārada Purāṇa to elucidate the above sūtra or Pāṇini. If Bhattoji Dīkṣita is not presumed to have consulted the Nārada Purāṇa both S.K. (Sīddhānta Kaumudī) and the Nārada Purāṇa used a common source.

[3]:

vṛṣaṇvasvaśvayoḥ

[4]:

As a matter of fact though the form niṣṭarkya is sanctioned by Pāṇini III.1.123 the formation is extremely irregular.

nis + √kṛt + ya (ṇyat instead of the regular kṣap) and the root √kṛt becomes √tṛk and we get niṣṭarkya.

[5]:

As a matter of fact the change—bh>;—h is accepted by modern linguists and Vedic √gṛbh—>Classical Sanskrit √gṛh. grah—is more natural linguistically.

[6]:

In the Vedic, the following seventeen letters are implied in akṣara-samūha (collection of letters viz. āśrāvaya (4 letters), astu śrauṣaṭ (4 letters), ye yajāmahe (5 letters), yaja (2 letters) and vaṣaṭ (2 letters) =17 letters in all.

[7]:

apaspṛdhethām—Imp. 2nd p. dual of apa + √spardh ‘to challenge’). As per Pāṇini VI. 1.36, the reduplication of the root, vocalisation of ra and the elision of a irregularly is accepted in the Vedic.

[8]:

ayyāt—As per Pāṇini VI.1.116, the e or o retain their original form in the middle of Vedic verse in the case of avyāt (avadyāt, avakramuḥ, etc) though ‘a’ in these has -a ‘v’ & ‘y’ following it.

[9]:

verses from 20 onwards are based on Pāṇinīya dhātupāṭha and the verses are practically quotations from it. For example, after listing 36 dhātus from edh—‘to prosper’ to √katth—‘to boast’. 

Pāṇini Dhātupāṭha remarks:

edhādaya udāttā anudāttetaḥ ātmanebhāṣāḥ. 

And after listing 37 (No. 38 to 74) dhātus from √at ‘to move constantly’ to √śundh—‘to purify or to get purified’, the Dhātupāṭha concludes: 

atādaya udātta udāttetaḥ parasmai-bhāṣāḥ.

Nārada Purāṇa repeats or paraphrases in verse these remarks of Pāṇini’s Dhātupāṭha.

[10]:

vide Dhātu Nos. 435-444(both inclusive) in the Dhātupāṭha (conj. I).

[11]:

vide Dhātu Nos. 445-474. (ibid).

[12]:

vide Dhātu Nos. 475-508 (ibid).

[13]:

amādayaḥ is a misprint in the Nārada Purāṇa

[14]:

vide Dhātu Nos. 509-601 (ibid).

[15]:

Dhātu Nos. 602.

[16]:

The Dhātupāṭha records √/dhukṣ—to be kindled and not √kṣudh— vide the group from dhātu Nos. 603-653.

[17]:

vide, ibid Dhātu nos. 654-741.

[18]:

vide, ibid Dhātu nos. 742-763.

[19]:

ṣit—an affix attached to roots to show the verb-activity. The Gaṇa sūtra in the Dhātupāṭha includes the roots from √ghaṭ—to √tvar (nos. 764-776) as ṣit for the purpose of adding the Kṛt. affix—a e.g. ghaṭā, vyathā, tvarā.

[20]:

vide, ibid Dhātu Nos. (806-846?)

[21]:

vide, ibid Dhātu No. 847.

[22]:

A Dictionary of Sanskrit Grammar (GOS 134, 1961) p. 409.

[23]:

vide—Dhātupāṭha (ibid) Dhātu Nos. 927-972.

[24]:

vide ibid Dhātu nos. 973-990.

[25]:

√vad is a misprint as this conjugation is called adādi and as such √ad— & not √vad (of the 1st conj.) can be the original.

[26]:

the Dhātupāṭha states udātto anudātted atmanepadi /

[27]:

Pāṇini Dhātupāṭha. 2nd conj. enumerates four dhātus viz. √dyu, √su-, √ku and √stu. Nārada Purāṇa records the first three in this verse and relegates √stu to the next (no. 40) verse.

[28]:

Ṣasādaya udāttā udāttetaḥ parasmai bhāṣāḥ—Pāṇini Dhātupāṭha.

[29]:

This is listed as Dhātu No. 71 in the Dhātupāṭha.

[30]:

The text of this line is as follows:

dādayo dhātavo vedāḥ parasmai-padino matāḥ /

The 3rd conjugation is called juhotyādi i.e. √hu dānādanayoḥ is implied in dādayo. Verse 46a shows that here in verse 45b dhātus √hu √bhi, √hrī and √pṛ are implied, the number ‘four’ seems to be implied in Vedāḥ.

[31]:

√rādh—‘to prosper’ is also found in the 5th & 10th conjugation (Svādi-curādike).

[32]:

Thirteen roots from √rādh to √sidh in Pāṇini Dhātupāṭha) are udāttet (tudādya) in the Nārada Purāṇa is a mistake & Pāṇini Dhātupāṭha docs not support it).

[33]:

Actually the Dhātupāṭha gives 35 dhātus.

[34]:

Viz. from √tud—‘to trouble’ to √kṛṣ—‘to plough’, ‘to draw’.

[35]:

Viz. from No. 12 √vraśc to No. √bhṛd—‘to plunge’, in Pāṇini Dhātupāṭha.

[36]:

i.e. Dhātus from (1) √tan—to (7) √ghṛṇ—

[37]:

viz. Dhātus from (1) √krī—to (7) √yu—‘to join’ or ‘to bind’.

[38]:

Pāṇiniya Dhātupāṭha lists sixty one roots in this conjugation.

[39]:

Pāṇini gives no definition of a Nīpāta, but states a number of panicles with no gender, number or sense beginning with ca: cādayo'sattve /