Syntax of the Cases in the Pali Nikayas

Chapter VI
The Genitive Case


§136. [General Character]

The genitive or the sixth case (Pāli chaṭṭhī = Skr. ṣaṣṭī) is on the whole a dependent case. It is used not only in connection with verbs and substantives but also with adjectives and adverbs (cp. VGS §202). But it is predominantly adnominal and in this respect appears in direct contrast to the acc. As the proper employment of the latter is to qualify the verb, so the gen. is normally used to qualify some other noun. It does so by assigning it to a particular class or description, or by distinguishing it as a part of a whole. So the fundamental notion expressed by it is to mark the belonging to or being part of. This possessive or partitive application admits of the almost universal rendering of the gen. in Pāli as in the older languages by the English of. Generally speaking, with substantives the gen. plays the part of an adj. as seen by the alternate constructions of either compounding it with the substantive qualified as kammāraputta D II.126; A V.263 “artizan-son” for kammārassa putta “son of an artizan”, or of using instead of it the derived adj. as rājā Māgadho M I.94 “the Māgadhan King” for rājā Magadhānaṃ “the King of the Magadhas”, and, porisaṃ dhuraṃ Sn 256 for purisassa dhuraṃ, or other adjectival formations such as those with the suffix -ka. As a qualifying word it expresses the most diverse logical relations between the two noun-concepts as in Skr. or even in Latin and Greek (cp. KVG §556 and SS §110).

The adverbial use on the other hand is not so diverse but presents sufficient interrelations between the noun- and the verb-concept to demand separate investigation. In all such uses the fundamental unity of conception appears to be the partitive notion. Says Brugmann: “As its fundamental character we may abstract (i.e. infer) that in the gen. the noun-concept (Nominal- [173] begriff) appears when the verb concept does not refer to its full range but when the former (noun-concept) is represented as a sphere which is only touched by the action” (KVG §529). We cannot discern the original relation between the adnominal and the adverbal uses. In spite of the unity of the principal notions expressed there are, at least in Pāli, a good many syntactical divergences.

§137. [Local Grammarians]

Owing to the predominance of the adnominal connection and its consequent remoteness from the action of the main verb, the local grammarians consider the gen. as falling outside the logical sphere denoted by the term kāraka. According to them, actual relations subsisting between the noun and the verb in a sentence are only expressed by the six genuine kārakas, viz., kartṛ, karman, karaṇa, sampradāna, apādāna and adhikaraṇa. They divide the vibhaktis into two main classes kāraka-vibhakti and upapada-vibhakti, the former having a definite relation to the action (kriyā) and the latter having none. The actions are generally performed by various agents (sādhana = efficients) which are directly or indirectly related to the verb. The term kāraka is only applied to such. The gen., according to orthodox opinion, is not a sādhana and consequently does not constitute a kāraka. Cakravarti Philosophy of Sanskrit Grammar pp. 199, 215. It is however disputable whether this observation of the ancient grammarians can be, without reservation, extended to the whole sphere of the genitive’s employment, especially to its adverbal function. On the whole their treatment of this case is not so clear-cut and exhaustive as of the others. In definition of the gen. Pāṇinī has only the loose aphorism ‘ṣaṣṭī śeṣe’ (II.3.50), which the kāśikā explains as meaning “in all other instances”, i.e. if none of the other cases enjoined (II.3.1-49) be available, one should use the sixth case (vide Speyer, SS p.82, f.n.l.). Kaccāyana attempts a clearer definition when he lays down the rule ‘yassa vā pariggaho taṃ sāmī’ (235) i.e. “that which has possession is called sāmī” and supplements it later on by saying “that the sixth case-affix is employed in denoting sāmī (possessor)”. sāmismiṃ chaṭṭhī’ Kac. 303. Moggallāna has the somewhat abstract definition ‘chaṭṭhī sambandhe’ (II.41). As for the other notions expressed by the gen. he gives only a few extra rules (305-310) which hardly compass [174] even the adnominal uses. This indifferent treatment of the gen. on the part of local grammarians is due to the fact, as we have pointed out before (§30), of their dealing with syntax solely from the point of view of the verb and not viewing the sentence as one psychological unit. The conception of kāraka is only the logical outcome of such an outlook (cp. the meaning of kāraka from kṛ “to do or make”, denoting action).

§138. [Form]

As regards form, the gen. has on the whole preserved the older case endings. It has even outgrown its legitimate sphere, and, as we have seen earlier (§91), replaced the dat. both in the singular and in the plural, the only surviving dat. form being the one in -āya of the a- declension (§4). The reason of this substitution of the gen. for the dat. in Pāli becomes apparent when we consider that even as early as in the dialect of the Brāhmaṇas the gen. (syntactically) had begun to encroach upon the proper sphere of the dat. cp. gen. with śraddhā and in the Aitareya and such uses as ‘tasya ha putro jajñe’ (VGS §202.B.2.a). In the Epics this replacement has gone even further and in the later classical language almost ousted it from its proper employ, but for a few fundamental uses (vide SS §86). In Prākṛt the dat. has become obsolete, a few traces of it being only found in the artificial dialect of the dramas (SS §100), and in the Aśokan inscriptions where a few -ehi forms are preserved (§9). A further point of contact with the dat. is found in the sympathetic use of the enclitic forms me, te and no, vo, to which phenomenon we have already alluded (§102).

§139. [Sphere of the Genitive]

Thus in Pāli the sphere of the genitive includes many uses of the dat. in the earlier language. In the great majority of cases we can assign a gen. or dat. to a particular syntactical category only on the analogy of Vedic and Classical Sanskrit. The gen. has also come into contact with the abl., in its adverbal uses such as with verbs of taking, hearing etc. and with the loc. in the partitive and absolute uses. It has, moreover, close affinities to the acc. of external object as with verbs of remembering and imitating, and, to the inst. of agent especially with participles [175] and the inst. of means with such verbs as meaning to fill etc. Most of these uses overlap one another and such divisions as the gen. of possession, of material, of distinction, and of origin or the partitive, subjective and objective gen. and others are made merely for the sake of convenience; they do not imply that absolute categories are possible. As we have remarked before, the unity of the logical functions of the various genitives remains unaffected (cp. SS §110).

§140. The Genitive with Substantives.

The Possessive Gen. represents the simplest syntactical function of that case, viz., of classifying a noun by naming its possessor. It is generally placed before the qualified noun as, for instance, in sabbe Bhagavato puttā S I.192 “all the Blessed One’s sons”. In verse or poetical prose, however, it is often found following the noun as, for instance, in Puttā Buddhassa orasā S III.83 “sons of the Buddha, self-begotten”. Though in this limited sense of denoting the possessor it is apparently a simple construction, what normally passes under the designation ‘possessive gen.’ is so varied in application that the most different logical relations may find expression by it (cp. SS §110). For instance, in such expressions as

Sundarikāya nadiyā tīre Sn p.79
“on that bank of the river S.”


brāhmaṇassa pada-saddena Sn p.80
“by the sound of the footsteps of the brahmin”,

the gen. properly speaking denotes no physical possession at all, the implication is more or less metaphorical. The latter, for instance, clearly means ‘brāhmaṇena kata-pada-saddena’ and consequently has a logical implication of agency rather than of possession. When the qualifying noun denotes a person and the qualified the result of some action on that person’s part, the notion of agency can hardly be overlooked. Thus e.g., in pitu vacanaṃ D III.181 “the father’s word”, the implied meaning is, like in the above, “pitarā bhāsitaṃ vacanaṃ”.

§141. [Various Realtions]

We may notice the following other relations expressed by this gen.: [176]

a. As in other I.E. languages the gen. in Pāli is capable of standing as the predicate of the whole sentence. e.g.,

sakaṃ te Mahārāja! D II.173
“all (is) thine, O Great King!”.

Here there is no doubt that te stands for the gen. and not the dat. (vide P.T.S. Dict. s.v. saka-).

b. With the verb ‘to be’ (bhavati) in the sense of “becoming” it is always doubtful whether the case is dat. or gen. of possession. But we may reasonably regard the following as genitives of possession since similar uses exist in the earlier languages (KVG §558.2). e.g.,

te rañño cakkavattissa anuyuttā ahesuṃ D III.62
“they became dependents of the universal monarch”.

This confusion is found even without the verb ‘to be’ in purely adnominal constructions. For instance, in rogānaṃ āyatanaṃ D III.182 the word rogānaṃ can mean “a province of diseases” or better perhaps “a province for diseases”.

c. Such contact between the sympathetic dat. and the possessive gen. has already been noticed (§102). It is mostly found, as pointed out there, with the enclitic forms of the personal pronouns which, though originally pure datives, are however found as genitives even in Vedic (cp. HKS §11). In the following examples it is the possessive sense that is more marked:

dibbaṃ te cakka-ratanaṃ ṭhānā cutaṃ D III.59
“your divine Gem of the Wheel has fallen from its place”;

yāva me idaṃ brahmacariyaṃ na iddhaṃ ... D II.114
“till this higher life of mine is not complete ...”;

yattha me assa chando vā ... taṃ mam’ assa musā D I.25
“where there was desire for me ... that was false of me”.

Here the parallel use of me (dat.) and mama (pure gen. form) side by side shows how far the syntactical confusion has gone.

d. With the relative pronouns the noun to be qualified may not immediately follow the gen. of possessor, which thereby assumes a role different from the above sense. e.g., ...

yassa kho pan’ assa Vāseṭṭha Tathāgate saddhā niviṭṭhā D II.84
lit. “of whom would, O Vāseṭṭha, faith be placed in the T. ...”;

santi bhante devā yesaṃ na sakkā gaṇanāya vā saṅkhāto vā āyuṃ [177] saṅkhātuṃ D III.111
“there are gods, Sir, of whom it is impossible either by reckoning or counting, to number the years (span of life)”.

§142. [The Subjective Gen.]

The subjective Gen. can also be regarded as an extension of the possessive gen. (cp. KVG §559) since the verbal noun qualified represents some action of the person denoted by the noun in the gen. (vide §140). But more particularly it denotes agency as shown by the following examples where the action is expressed by a p.p.p. in -ta:

imassa ca bhikkhuno duggahītaṃ D II.124
“also a misconception of this monk”;

paresaṃ subhāsitaṃ D I.3
“good-speech (lit. well-spoken) of others”.

We may observe from these examples that the gen. in such instances is interchangeable with the inst. of agent (§88). Orthodox grammarians are divided on the question whether these two constructions are universally interchangeable. According to Pāṇinī the inst. is of necessity if the verbal noun be attended by its subject and its object at the same time. This is understandable since otherwise the presence of two genitives would lead to confusion. Speyer considers that we may extend this observation to all such instances as where the subjective gen. would be used together with some other sixth case (SS §114). But according to other Indian authorities the gen. of the subject is nowhere forbidden (ibid). On the whole the observation of Pāṇinī seems applicable to Pāli concinnity as well. In fact here the gen. even seems to be preferred with most participles of clearly nominal standing, and in the following example, coming as it is after the verbal noun, the gen. conspicuously possesses the agent sense:

na kho Tapassi ācinnaṃ Tathāgatassa daṇḍaṃ daṇḍanti paññāpetuṃ M I.373
“it is not the practice of the Tathāgata, Tapassi, to lay down punishment as punishment”.

a. In such instances as those discussed above the gen. can be interchanged with the inst. of agent and implies kartṛ, the agent of a passive (hence originally transitive) verb. But when the verbal noun is formed from an intransitive verb the gen. seems to denote not the agent to be expressed by the inst. but the [178] subject implying an original nom. e.g.,

iti rūpassa samudayo M I.61
“so the arising of form”;

paṭigha-saññānaṃ atthaṅgamā D III.262
“by the disappearance of ideas of ill-will”;

catunnaṃ māsānaṃ accayena Sn p.102
“on the lapse of four months”.

In the first example, for instance, the idea implicit cannot be paraphrased by ‘rūpena samudayo’ as in parehi subhāsitaṃ for paresaṃ subhāsitaṃ but must be taken as “iti rūpaṃ samudeti”.

b. With other types of verbal nouns where there is no participial sense and the verbal element is less emphasized, neither the inst. of agent nor the nom. of subject can possibly be substituted. e.g.,

idaṃ pacchimakaṃ Ānanda Tathāgatassa Vesāli-dassanaṃ bhavissati D II.122
“This, Ānanda, shall be Tathāgata’s last sight of Vesāli (lit. Vesāli-seeing)”;

raññaṃ niyyānaṃ bhavissati D I.9
“there will be an exit of kings”.

Here the gen. being used along with the verb ‘to be’ has a strong possessive sense. The latter example clearly borders on the subjective and possessive gen.

§143. [The Objective Gen.]

The Objective Gen. is not so frequent as the above, since usually Pāli prefers to retain the acc. even with verbal nouns. e.g.,

Bhagavantaṃ dassanāya M II.23,46; A I.121; III.381
“for seeing the Blessed One”.

But the dat. of purpose dassanāya has greater verbal force than any other type of nomina verbalia, being more or less an infinitive (§107.a). Even in such instances, however, one not rarely comes across the gen. of object. e.g.,

ariyānaṃ dassanāya Dh 206
“for the seeing of the noble ones”

(i.e. in order to see the noble ones);

mano-bhāvanīyānaṃ pi bhikkhūnaṃ asamayo dassanāya D III.36
“it is not the time even for the seeing of self-composed monks”.

a. It is frequently found with primary nominal formations formed by adding such suffixes as -a to the root. In such cases the gen. is almost of necessity and qualifies the noun. e.g.,

catunnaṃ bhikkhave dhammānaṃ ananubodhā ... D II.122
“Brethren, owing to the non-understanding of four things”;

Tathāgatassa pūjāya D II.137
“for the honouring of the T.”;


lobho cittassa upakkileso D I.91
“greed is a defilement of the heart”.

b. With verbs of motion the normal construction is to retain the acc. of destination or the corresponding dat. But when the motion implied is towards a person the gen. is used similar to the objective gen. e.g.,

upasaṅkamanaṃ pâhaṃ bhikkhave tesaṃ bhikkhūnaṃ bahukāraṃ vadāmi S V.67
“even the approaching (going to) of those monks, brethren, I say is of advantage".

c. With the so-called nomina agentis or agent-nouns the gen. and acc. are promiscuously employed, (vide §33). It must be stated, however, that with those formed with the suffix -tṛ (Pāli -) the acc. seems to be favoured, while the gen. is the commoner idiom with those having possessive sense such as -vin and -in (f. -inī). e.g.,

Evaṃ opanāyikassa dhammassa desetāraṃ D II.222,228
“the preacher of such a redeeming doctrine”;

ariyānaṃ adassāvi M I.1
“a non-beholder of Noble Ones”;

lābhinī Bodhisatta-mātā hoti pañcannaṃ kāma-guṇānaṃ D II.13
“the mother of the ‘Being destined for Enlightenment’ is receiver of the five-fold pleasures”.

§144. [The Partitive Genitive]

The Partitive Genitive proper denotes the whole, a part of which is meant by the qualified word as in

bhāgo maraṇassa Sn 427
“a share of death”,

bhāgī āyussa A II.80, III.42
“having a share of life”


kiñcideva desaṃ vācāya A V.39
“a certain portion of the speech”.

a. But the more frequent type of this gen. in Pāli is that which carries the notion of selecting or distinguishing out of a multitude, usually of persons but sometimes even of things. In this case it is interchangeable with the loc. of the persons among whom (§167). In fact the two cases occur side by side in the same context. e.g.,

Kati jāgarataṃ suttā, kati suttesu jāgarā S I.3 (V.)
“How many are the sleeping among the wake, how many are awake among the sleeping?”.

From the following examples it may be observed that in Pāli as in Skr. (SS §116) the partitive gen. may not only attend substantives but all kinds [180] of pronouns and adjectives. Such a gen. may also appear in various logical connections.

b. With certain adjectives and nouns it denotes the person or thing that is distinguished from the rest. e.g.,

tvaṃ yeva nesaṃ eko cakkhumā D I.191
“you alone are the seeing among them”,

or that which is selected out of many. e.g.,

imesaṃ tiṇṇaṃ aṅgānaṃ jātiṃ ṭhapayāma D I.121
“of these three factors let us keep birth aside”.

c. With numerals grammatically denoted by substantives as sataṃ, sahassaṃ etc. the nouns qualifying are put in the gen. case. e.g.,

bhiyyo naṃ satasahassaṃ yakkhānaṃ payirupāsati D II.256
“a hundred-thousand (of) yakkhas worship him”.

Here the nom. of apposition is the parallel construction (§24.c.).

d. With indefinite pronouns such as aññataro, aññatamo, eko etc. it denotes inclusion among a group or class. e.g.,

etesaṃ vā aññatarena D I.21
“or by one or the other of the same”;

aññataro ca kho pan’ āyasmā Kassapo arahataṃ ahosi D I.177
“the Ven. Kassapa became one among the saints”;

tesaṃ ahaṃ aññatamo M I.17
“of them I am one”.

e. With substantives and adjectives denoting mastery and power it takes a slightly different turn of meaning and can be rendered by the English over. e.g.,

evaṃ mahiddhiko kho bhikkhave sīho migarājā tiracchānagatānaṃ pāṇānaṃ, evaṃ mahesakkho ... S III.85
“so powerful indeed, monks, is the lion, the king of beasts, over beings of the animal class, so majestic ...”;

Satthā devamanussānaṃ M I.69
“The Master of gods and men”.

f. When option between two persons or things is intended the persons or things considered in making the comparison are put in the gen. case. Here it closely resembles the gen. of the persons regarding whom a statement is made (i.e. the gen. of relation, §156). e.g.,

ko nu kho āyasmantānaṃ sukhavihāritaro rājā vā Māgadho ... āyasmā va Gotamo M I.94
“of you two honourable ones, who is the more happy-living, the king of Magadha ... or the Ven. Gotama?”;

ayaṃ imesaṃ dvinnaṃ puggalānaṃ ... [181] hīnapuriso akkhāyati M I.25
“he of these two persons ... appears as the lesser one”.

g. With superlatives or adjectives having superlative sense such as those meaning first, last, foremost etc. a similar gen. of the persons (or things) of whom, from amongst or regarding whom the statement is made, is frequently found. e.g.,

khattiyo dvipadaṃ seṭṭho S I.6 (V.)
“the warrior is the best among bipeds (i.e. two-footed beings)”;

ye te ahesuṃ brāhmaṇānaṃ pubbakā D I.104
“those who were the foremost of the brahmins”;

gimhānaṃ pacchime māse M I.306
“in the last month of the summer (lit. of the warm months)”;

pañcannaṃ bhikkhusatānaṃ pacchimako bhikkhu D II.155
“the last (monk) of the five hundred monks”.

h. Even with adjectives of equality and identity or those of superiority such as sadiso and visiṭṭho this gen. is commonly employed. This sometimes gives rise to a double genitive, for the gen. of comparison (§151.b.) may also appear side by side with it, as in the following example:

nâssa hoti koci paññāya sadiso vā visiṭṭho vā sabba-sattānaṃ D III.158
“of all beings, there is no one who is equal to him or superior”.

Of course assa here may as well stand for the dat. (§110), or may be alternately expressed by the inst. (§§85,86) or the abl. (§132).

§145. [The Gen. of the Material]

The Gen. of the Material may also be regarded as expressive of the partitive notion (cp. KVG §559). This is comparable to the gen. of material or stock drawn upon, found with verbs of filling and lacking in Latin and Greek cp. Buckland Green, Notes on Greek and Latin Syntax, §45.3. and generally comes under the so-called Gentivus Materiae et Originis though in Pāli, just as in Skr. (SS §113), the pure gen. of origin is not very frequent. It is only found adverbally with the verb pahoti (Skr. prabhavati = originate) “to rise from”. e.g.,

yato câyaṃ Gaṅgā nadī pahoti S II.184
“whence arises the river Ganges”,

where yato is the abl. representing more normal construction, whereas the gen. occurs only in instances of a metaphorical nature such as

sammādiṭṭhissa, bhikkhave, sammā-saṅkappo pahoti M III.76
“right aspiration originates from (lit. of) right views”).

In Pāli the gen. usually expresses the substance or thing of which [182] something else or some object is made, consists of, full of or is laden with. The following distinctions may be observed.

a. When the qualified noun denotes an artificial product, the gen. always expresses the material of which the former is made. e.g.,

suvaṇṇassa pabhassarāni Sn 48
“ornaments of gold”.

This is, however, not found in prose; but the gen. denoting that of which something else consists occurs, though not frequently. e.g.,

aṅgulīnaṃ mālaṃ M II.98
“a necklace of fingers”.

b. When it is attended by the actual verb of making or preparing etc. the gen. seems to border on the inst. of means. It is then, so to speak, half-way between the adnominal and the adverbal constructions. e.g.,

sālīnaṃ odanaṃ (racayitvā) M I.31
“having prepared a meal of rice”;

mahantaṃ hirañña-suvaṇṇassa puñjaṃ kārāpetvā M II.63
“having caused a great heap of gold and bullion to be piled up”;

pāṇīyaṃ khādanīyaṃ bhojanīyaṃ paṭiyādāpetvā paṇḍumuṭikassa sālino ... M II.50
“having prepared excellent hard and soft food (consisting) of the choicest golden rice”.

That in such instances the verb is not of necessity and the gen. can be purely adnominal is shown by the example

udakassa dhārā D II.15
“torrents of water”,

where it borders on the descriptive gen.

c. When the qualified noun denotes a limited space to be filled, the gen. of the material is logically related to the gen. with adjectives and verbs of filling (§§151.b.ii. & 149). e.g.,

uṇhodakassa kājaṃ S I.175
“a pail of hot water”

(i.e. a pail full of ...);

pañcamattāni taṇḍulavāhasatāni pāhesi paṇḍumuṭikassa sālino M II.54
“sent about 500 cart-loads of the choicest golden rice”.

d. As partitive genitives, the above uses may be said to express logically the notion of quantity. Closely related to this is the gen. of description expressive of quality. This is generally known as the gen. of quality (KVG §559). It is however not so frequent in Pāli or Skr. as in Latin and Greek. e.g., ( kusāvatiyā)

catunnaṃ vaṇṇānaṃ dvārāni ahosi D II.170
“In Kusāvatī there were doors of four colours”.

This may also mean “there were [183] gates for the four castes”, in which case vaṇṇānaṃ would be the dat. of interest.

§146. [Abstract Usage]

It is not to be expected that the whole sphere of the adnominal genitive’s use is included in the categories established in the preceding paragraphs. No divisions can be absolute or exhaustive in treatment. There are bound to be many other uses which may or may not fall within such categories. A most abstract use of the gen. is to be found in the construction with such causal postpositions as hetu and kāraṇā, which are both ablatives (cp. §122.d. & §73). There is however the parallel construction of compounding these (especially hetu) with the preceding noun, which would otherwise be put in the gen. (cp. §136). e.g.,

atta-hetu M III.48
“due to himself’; vedanāhetu M II.216 “due to sensation”;

cīvarahetu A I.147
“for the sake of a robe”.

According to Pāṇinī the sixth case-affix is to be employed in the construction with hetu. saṣṭī hetuprayoge’ (Pāṇ. II.3.26). The vārttika illustrates this by ‘annasya hetor vasati’ “lives by reason of food”. Pāli follows the same usage with both hetu and kāraṇā. e.g.,

issariyassa kāraṇā D I.84
“by reason of glory”

(i.e. for the sake of glory);

yāsaṃ ... hetu brahmacariyaṃ carasīti M II.46
“owing to whom ... you live the Holy Life”;

accharānaṃ hetu M II.64
“owing to heavenly damsels”;

puttadārassa hetu M II.187
“for the sake of child and wife”.

On the analogy of these examples we may not hesitate to place in this category the form kissa found in the frequent interrogative phrase taṃ kissa hetu D II.14; M I.1; A II.31, as a genitive qualifying hetu. Consequently kissa here is pronominal substantive and not adj. as in kissa ... kammassa D II.185; it is not to be confused with the later adv. kissa “why?”.

§147. The Gen. with Verbs.

Several classes of verbs are construed with a gen. Most of these adverbal uses correspond to the adnominal constructions discussed above. In the majority of instances the gen. in special connection with verbs appears concurrently for an acc., inst., or abl. and sometimes even for the loc. Generally speaking it [184] is the partitive notion that underlies such functions. One hardly meets with adverbal genitives expressive of the idea of possession in Pāli (of the Nikāyas) as found in Skr. with verbs of owning and ruling such as prabhū, īś, and Vedic rāj and kṣi (VGS §202). But there is a considerable number of verbs construed with the gen. having a sense analogous to that of the acc., but, with this difference: that unlike the latter it expresses that the action affects the object not as a whole, but only in part. It is used with verbs having the following senses:

a. remembering, thinking of; e.g.,

nāgavanassa sumarati Dh 324
“thinks of the elephants’ haunts”.

b. imitating, following;

taṃ tassa anugaṃ hoti S I.72
“it follows him”;


evaṃ gihī nânukaroti bhikkhuno Sn 221
“so the householder imitates not the monk”.

In such instances, however, the prefix anu- also favours construing with the gen. e.g.,

dhammassa cânudhammaṃ vyākaronti M I.368
“preach what is consistent with the doctrine”.

A similar gen. is found with the verb anumodāmi in the sense of “appreciate”. e.g.,

N’ eva kho tyāhaṃ brāhmaṇa anumodāmi na paṭikkosāmi A II.36
“brahmin, I neither appreciate you nor blame you”.

Skr. grammarians are at a loss to explain these uses. They regard them as anomalies. vide, Philosophy of Sanskrit Grammar p. 220.

c. It is also found with verbs having the sense of desiring of, expecting of etc. e.g.,

yassa dāni devassa icchissati Brahmā Sanaṅkumāro D II.210
“of whichever god the Brahmā S. now desires (anything)”.

A similar construction is found with the gerundive pāṭikaṅkhaṃ. e.g.,

Saddhassa hi Sāriputta ariyasāvakassa etaṃ pāṭikaṅkhaṃ S V.226
“Of the faithful Aryan disciple, Sāriputta, this should be expected”.

d. In the gāthā literature are found a few verbs of harming, robbing, plundering and conquering, construed with a similar gen. e.g.,

ahāsi me Dh 3
“he robbed me”;

ālopati sahasā yo paresaṃ Th 1.743
“who forcibly plunders others”;

uccāvaceh’ upāyehi paresaṃ abhijigīsati Th 1.743
“by various means he (cheats, harms or) overcomes others”.

As rendering of this rare verb Mrs. Rhys Davids has ‘cheat’ and Neumann ‘vernichten’. But [185] P.T.S. Dict, prefers ‘overcome’ or ‘covet’, after Comy. ‘jinituṃ icchati’ (on J. VI.193). It should, however, literally mean “desire to kill”, since it is the desiderative of han, normally reading ‘jigiṃsati’.

§148. [Gen. with Jānāti]

A peculiar idiom is found with the verb jānāti, which with a sense different from the usual “to know” (like vid-), seems to be construed with a gen.. According to Pāṇinī (II.3.51) jñā is to be employed in agreement with the gen. of the instrument (karaṇa); then jñā must not be equivalent to vid but, says the kāśikā, must mean “to ween, fancy or have an opinion of”. In certain contexts in the Nikāyas jānāti construed with the gen. seems to exhibit this last sense, and, hence imply a partitive notion. e.g.,

na tvaṃ tāta Raṭṭhapāla kassaci dukkhassa jānāsi M II.56 (cp. 60)
“dear R., you know not of any sorrow”.

Here, the Comy. points out, the exact meaning is “you have no experience even of a little of sorrow” (‘appamattakam pi kālabhāgaṃ dukkhassa’). In this example the gen. certainly expresses the thing of or regarding which there is knowledge, and is comparable with the English expression “to know of any sorrow”. But there are other uses of this verb, which, though similar in meaning to the above, are still logically different. e.g.,

ko nu kho pana bho jānāti jīvitānaṃ D II.233 (246) lit.
“who indeed knows of the living”.

Here the context shows that the meaning is “entitled to have an opinion of”, as the P.T.S. translation goes: “who indeed can answer for the (survival of) the living?” With ājānāti however the case is the dat. since the same is found in Skr. e.g.,

Āmagandhe ... bhāsamānassa na ājānāmi D II.242
“I do not understand you speaking about ... ‘the smell of raw flesh’”.

§149. [Inst.-like Gen.]

An inst.-like gen. is found with verbs having the sense of fullness and satisfaction, (cp. SS §123.6). Though these verbs pūreti, tappati etc. are often construed with a gen. the more logical construction seems to be with the inst. of means. e.g.,

dvinnaṃ ... dhammānaṃ atitto ... mātugāmo kālaṃ karoti [186] A I.78
“unsatiated with two ... things, the female dies”;


puriso pāyāsassa tappati J I.185
“the man is satisfied with the milk-rice”;

dhīro pūrati puññassa Dh 122
“the wise man is filled with merit”;

nivesanāni ... nānādhaññassa pūretvā Sn 305
“having filled the dwellings ... with various kinds of grain”.

With verbal nouns, especially nomina agentis, formed from these roots, the gen. of filling falls into the category of the objective gen. e.g.,

labhati lūkhassa vā paṇītassa vā bhojanassa yāvadatthaṃ pāripūriṃ D III.257
“he obtains satiety (lit. fulfilment) in (lit. of) gross or excellent food”.

In such cases even the loc. is admissible.

§150. [Abl.-like Gen.]

The abl.-like gen. (SS §125) occurs with certain verbs where some notion of separation or distance or some such ablatival function seems to be implied. We have already referred to the gen. used with local adverbs such as dakkhiṇato, uttarena and sammukhā denoting the point from which distance or direction is reckoned (§130) and the gen. of origin employed instead of the corresponding abl. with the verb pahoti (§145). As Speyer points out (SS §125) the gen. is admissible in such instances when there is at the same time room for the conception of belonging to (being part of) and that of proceeding from.

a. With such verbs as paṭigaṅhāti the person from whom something is received is denoted by the gen. In the following example the presence of the participle gives a semi-absolute appearance to the construction:

paṭigaṅhāti ... na dvinnaṃ bhuñjamānānaṃ M I.307
“he does not receive ... from two people at meals”.

The same verb in its metaphorical sense of “accepting a confession” follows a similar construction (cp. SS §83.6 & Pāṇ. I.IV.41). e.g.,

yo ca accayaṃ desentassa yathā-dhammaṃ na paṭigaṅhāti S I.239
“who does not pardon (lit. accept of) one confessing a transgression, according to the Doctrine”;

or with semi-absolute sense:

parassa kho pana accayaṃ desentassa yathā-dhammaṃ nappaṭigaṅhāti A I.103
“he does not pardon one confessing ...” etc. or “he does not accept when one makes a confession.”

The verb ādiyanti in the sense of “taking the [187] word” (Comys. ‘vacanaṃ na karonti’ i.e. “obey”) may admit of a similar gen. e.g.,

mahārājānaṃ na ādiyanti D III.204
“they do not heed or obey the great kings”;

later on the same page: n’eva rañño Māgadhassa ādiyanti. It is however not quite certain whether this is the gen. or the dat. with verbs of listening to like sussūsati etc. (§94.b.).

b. With verbs of hearing, especially suṇāti, the gen. seems to be the general construction, the abl. being hardly ever found in the Nikāyas (§125.b.). This may be regarded as a gen. expressive of the origin of sound or the source from which the perception comes. e.g.,

āyasmato Sāriputtassa sutvā bhikkhu dhāressanti M I.14,46
“having heard from (lit. heard of, i.e. learnt of) the ven. S. the monks will take (it) to heart”

(cp. D II.2,148);

na aññassa samaṇassa vā brāhmaṇassa vā sutvā A I.142; M III.186
“not having heard from another recluse or brahmin ...”;

Tesaṃ sutvā D III.61
“hearing from them”.

Even with passive forms of this verb the gen. is preserved. e.g.,

tesañca sotabbaṃ maññissanti A IV.16 lit.
“they think it should be heard of them”

i.e. “they think they should be heard”. In the above examples the gen. denotes the person from whom one learns or hears and is, therefore, logically parallel to the abl. But when the gen. is attended by participles of verbs denoting talking, sounding, howling, conversing etc. the construction not only appears to be semi-absolute but seems to be used in place of an acc. rather than an abl. though at the same time it expresses the source of sound. e.g.,

assuttha no tumhe ... siṅgālassa vassamānassa S II.271
“did you not hear the jackal howling ... ?”;

nāmāni me kittayato suṇātha M III.69
“hear me repeating the names”

i.e. “listen while I am repeating ...” (absolute);

assosi kho āyasmā Anuruddho dāyapālassa Bhagavatā saddhiṃ mantayamānassa M I.205
“the ven. Anuruddha heard the Blessed One talking with the park-keeper”;

assosi kho āyasmā Sāriputto Sunakkhattassa Licchaviputtassa parisatiṃ vācaṃ bhāsamānassa M I.68
“the ven. Sāriputta heard Sunakkhatta the Licchavi speaking these words among the rabble”.

In such cases the Comys. supply the implied object, usually ‘saddaṃ[188] in their paraphrase. e.g., for sutvā devassa vassato Sn 30, the Comy. has ‘vassato saddaṃ sutvā’ (Pj. II. p.42). But the acc. of the object may not be always implied, since the verb suṇāti can occur with an intransitive sense, as for instance in:

kinti te sutaṃ brāhmaṇānaṃ ... bhāsamānānaṃ D I.104 (cp. D II.237)
“What! have you heard from the brahmins ... speaking”.

c. Similar to the above ablatival gen. is that which is expressive of the source of fear with verbs of fearing such as bhāyati “fear”, tasati “tremble at” and āsaṅkati “suspect”. The abl. too is found in the Nikāyas, but only with noun bhayaṃ and never adverbally, (§122.e.). Here the local grammarians regard the abl. as the proper case. Kaccāyana’s rule dutiyā pañcamīnaṃ (311) is interpreted by the vutti as meaning that the sixth case-affix is used sometimes (kvaci) instead of the second and the fifth, examples for the latter being those such as ‘sukhassa bhāyāmi’ etc. The acc. however is rarely found with the verb bhāyati as in

bhāyasi maṃi samaṇā ti Sn p.48
“do you fear me, recluse?”;

it appears to be almost of necessity when the object to be feared is a person (§36.a.). Otherwise the gen. is the commoner adverbal construction. e.g.,

kiṃ nu kho ahaṃ tassa sukhassa bhāyāmi M I.247
“what! indeed, do I fear that happiness?”;

na bhāyitabbaṃ etassa sukhassāti vadāmi M I.454; III.233
“I declare that one should not fear this happiness”.

In verse however the gen. is found even with personal nouns. e.g.,

sabbe bhāyanti Maccuno Dh 129
“all fear Death”,

or adnominally. e.g.,

kālassa meghassa bhayena tajjitā Th 1.308
“oppressed with fear of the black cloud”.

The verb tasati occurs only in verse in this construction. e.g.,

sabbe tasanti daṇḍassa Dh 129
“all tremble at punishment”.

But āsaṅkati is found even in prose with the gen. e.g.,

Tena khopana samayena rājā Māgadho Ajātasattu Vedehiputto Rājagahaṃ paṭisaṅkhārāpeti rañño Pajjotassa āsaṅkamāno M III.7
“at that time Ajātasattu King of Magadha, son of Vedehi, was fortifying Rājagaha out of suspicions for King Pajjota”.

In Skr. however śaṅk “to suspect” takes the gen. with the prefix abhi- and the acc. when it is used with ā- (cp. Cappeller’s Dict.) [189]

§151. The Genitive with Adjectives.

We have already referred to the gen. used with verbal adjectives implying ‘possession’ such as those ending in -vin (f. -vinī) and -in (f. -inī). When formed from transitive roots these agree with an objective gen. (§143.c.). A similar gen. is found with upādāya which, though a gerund in form, is still capable of qualifying a noun as much as an adj. e.g.,

catunnaṃ mahābhūtānaṃ upādāya rūpaṃ M I.185
“whatever form depends on the four great elements”.

But apart from being used parallel to the acc., the gen. with some other adjectives appears in place of a loc., inst. or dat., at least from a logical point of view.

a. With adjectives of knowledge, skill and experience the gen. of the thing in which one is skilled etc. is logically akin to the acc. with transitive verbs (SS §124.III.). But the concurrent idiom is the loc. of relation, owing to the fact that the noun in the gen. with such adjectives expresses that regarding which one is skilled etc. (§175.a.). Kaccāyana enjoins the use of the gen. in place of the loc. with such words by the rule ‘chaṭṭhī ca’ (310), which according to the vutti is to be taken as implying that the sixth case-affix is employed instead of the third and seventh (laid down in the previous rule ‘tatiyā sattamīnañca’ - 309) optionally (kvaci). It is found in the Nikāyas with such adjectives as: kevalī “perfected, expert”; e.g.,

brahmacariyassa kevalī A II.23
“perfected in the Higher Life” or “expert of the Higher Life”;

kusala “clever, skilled”; e.g.,

puriso kusalo bherisaddassa A II.185
“a man clever in the tone of the drum”;

kusalo ahaṃ rathassa aṅgapaccaṅgānaṃ M I.395
“I am skilled in the parts and accessories of the chariot”;

imesaṃ dhammānaṃ sukusalo D I.180
“extremely gifted in these things”;

akusalā imassa lokassa M I.225
“having no knowledge of (unversed in matters of) this world”;

kovido “adept, proficient”; e.g.,

yogakkhemassa pathassa kovido Th 1.69
“proficient in the path of deliverance”

ariyadhammassa akovido M I.1
“with no knowledge of the Noble Doctrine”.

With kovido the gen. is logically quite similar to the objective gen.. But the concurrent construction [190] is with the loc. as seen by the corresponding gloss ‘ariyadhamme akusalo’ (Ps. I. p.22). The addition of prefixes (negative etc.) to these various adjectives does not necessitate the alteration of the construction, as may be seen from the above uses with akovido, akusalo, sukusalo etc.

b. With adjectives of likeness, equality and similarity and of fullness or completeness, the gen. is parallel to the inst., of comparison in the former case, and of means (in the narrower sense) in the latter case.

i. The notion of comparison can be diversely signified in Pāli as in Skr. In certain nuances the inst. (§§85,86) or the abl. (§132) of comparison is even preferred. With those denoting superiority or inferiority and even equality etc. the dat. is also perhaps used in the Nikāyas (vide §110). With such adjectives as sadisa, sama, samasama and sādhāraṇa, however, the case is more likely to be the gen., considering the popularity of this construction in Skr. (SS §§124.4. & 61). Here the inst., just as in the earlier language, is particularly concurrent, but the abl. cannot be applied as all these adjectives denote “similarity”, thus implying no “distinction” (vibhatta) for the latter to be permissible. e.g.,

so mātu pi sadiso pitu pi sadiso M II.153
“he is like his mother and also his father”;

assa ... sadiso D III.158
“equal to him”;

attano samasamaṃ D I.174
“equal to himself”.

With nīcataraṃ, acc. adverb from the comparative stem, nīcatara-, it is either the dat. or the gen. that is used. e.g.,

nīcataraṃ Bhagavato S I.144
“lower than the Blessed One”.

Here the abl. would be the more logical case, but the gen. is frequent with other local adverbs in the Nikāyas (§153).

ii. The gen. found with adjectives of fullness such as puṇṇa and pūra can be regarded as the adnominal corresponding to the adverbal use discussed above (§149). The parallel idiom with the inst., however, cannot be used with the adjectives, since these (not being verbs) do not imply means. e.g.,

nagaraṃ ... puṇṇaṃ sāsapānaṃ S II.182
“the city ... full of mustard”;

pūraṃ hiraññasuvaṇṇassa kumbhiṃ D II.176
“a pot full of gold and bullion”;


ayaṃ Aciravatī nadī pūrā udakassa D I.224
“this river A. is full of water”;

pūrā aṅgārānaṃ M I.365
“full of ambers";

pūran nānappakārassa asucino M I.51
“full of all kinds of filth”.

This seems to be one of the oldest functions of the gen. in I.E. (cp. Latin gen. c. plenus).

c. We have already discussed the gen. employed with certain verbs derived from the root jñā prefixed by anu- or ā- (§148). A similar gen. is found with verbal adjectives belonging to the same root, especially with samanuñña “approving of” or “favourable to”. e.g.,

na pāṇaṃ atipātayato samanuñño hoti D III.48
“he is not approving of one taking life”;

samanuñño me Satthā S I.1
“the Teacher is favourable to me”.

The gen. in these cases, however, is not a certainty since even in Skr. the dat. is applicable with some verbs from jñā. For instance, with ājānāti “to learn or understand” the acc. of the thing, or the dat. or the loc. of the person, is quite frequent. (vide Cappeller’s Dict, s.v.). But the thing approved of is found in the Nikāyas with the loc. e.g.,

samphappalāpe ca samanuñño hoti A V.305
“he is approving of frivolous talk”;

adinnâdāna veramaṇiyā ca samanuñño hoti A II.253
“he is also approving of abstinence from taking what is not given”.

§152. Dative-like Genitive with Adjectives.

We have seen earlier (§138) how the gen. in the older language had encroached upon the syntactical sphere proper to the dat. and ousted it from many of its legitimate uses. The fact was noticed even by early Skr. grammarians (cp. Pāṇinī ‘caturthyarthe bahulaṃ’ II.3.62). This replacement has proceeded in the adjectives on a wider scale. Speyer calls such uses in Skr., where the forms show the actual case unlike in Pāli in which the dat. and gen. are for the most part identical, the dat.-like gen. So with adjectives of friendship and enmity, good and evil, fitness and unfitness etc. we find a gen. (in Skr.) apparently for the dat.. Since there is no reason to suppose that Pāli retained the dat. with such adjectives, it seems not unjustifiable to regard the following -ssa forms as genitives. The dat. if used in such cases would express the point of view (cp. KVG §554.4). [192]

a. It is especially the case with adjectives having the sense of pleasing to. e.g.,

pitā puttānaṃ piyo hoti D II.178
“the father is dear to (or beloved of) his sons”;

bahuno janassa piyo ahosi D II.19
“he was beloved of many people”;

so even with compounds where the first member is such an adj.:

piyadassino honti bahuno janassa D III.167
“has a pleasing appearance for many people”

(lit. has an appearance beloved of many people). With the enclitic pronouns it is very doubtful whether the case is gen. at all. e.g.,

na kho me taṃ paṭirūpaṃ D II.30
“it is indeed not fit for me”

(cp. dat. §108).

b. With passive participles used as adjectives the gen. not only denotes the people concerned but borders on the function of agency (§154). e.g.,

Samaṇo ... Gotamo rañño Pasenadī Kosalassa sakkato garukato mānito pūjito D I.116
“the recluse ... Gotama is respected, honoured, revered and worshipped of King Pasenadī of Kosala”;

āvāha-vivāhakānaṃ apatthito hoti mittāmaccānaṃ paribhūto hoti D III.183
“unwanted of those giving or taking in marriage and despised of friends and colleagues”;

sādhusammato bahujanassa D I.47; II.150; Sn p.92
“well-revered of many people”;

ekesaṃ samaṇa-brāhmaṇānaṃ sāmañña-saṅkhāta ... D I.166
“regarded as compatible with recluseship of (i.e. by) some recluses and Brahmins”.

§153. The Genitive with Adverbs.

In connection with the abl.-like gen. (§150) we had occasion to mention its use with adverbs derived from nouns such as dakkhiṇato etc. where the gen. marks that of which a region or direction is considered, thus psychologically coinciding with the ablatival notion of that from which direction or distance is reckoned. The use of the gen. in such instances is due to the slight partitive sense implied, whereas the abl. should be the more logical construction. However the local grammarians regard the gen. as of necessity in such cases, (vide Pāṇ. II.3.30), an observation supported by the fact that in the actual literature the abl. though logically proper is never found with these regional adverbs (whether in -to or -ena), the gen. being the only [193] construction. It is so even in the Vedas (VGS §202.D.) and the language of the Nikāyas is no exception.

a. Adverbs in -to are always construed with the gen. e.g.,

dakkhiṇato nagarassa D II.321
“on the south of the city”;

pācīnato Rājagahassa D II.263
“on the east of Rājagaha”;

With those ending in -ena, the kāśikā on Pāṇ. II.3.31 allows optional construction with gen. or acc. (§53). But the former seems to be the more usual even here with those signifying regions or quarters. e.g.,

puratthimena nagarassa M I.343
“on (by) the east of the city”;

uttarena Manasākaṭassa D I.235
“to (lit. by) the north of the Manasākaṭa”.

But even with other adverbs in -ena the gen. is not unusual though the acc. is by far the more frequent. e.g., antarena yamakasālānaṃ D II.169 (134,137).

b. Pāṇinī allows optional construction with abl. or gen. of all words meaning far and near (II.3.34). Pāli grammarians permit the use of the abl. acc. or inst. (vide vutti on Kac. 277 ‘dūratthe’ and ‘antikatthe’). But the only cases attested in the Nikāya prose are the abl. and the gen., the former being restricted to pronominal forms (ito and tato, vide §131.b.) and the latter being used with all types of nouns. e.g.,

Bhagavato santike D II.152
“near the Blessed One (lit. in the proximity of the Blessed One)”;

Vedehi-puttasssa avidūre D I.94
“at no distance from the son of Vedehi”;

Anuruddhassa sammukhe S V.294
“in the presence of A.”;

so nâtidūre nâccāsanne āsanassa parivattati M II.138
“he takes his position neither at too great a distance nor in too close proximity of the seat”;

avidūre ambavanassa M II.141
“in the vicinity of the mango-grove”.

c. Finally there is the gen. employed with adverbial prepositions such as heṭṭhā “below” and upari “above” (cp. VGS §202.D.). Here too the abl. is the concurrent idiom. e.g.,

tassa eva pāsādassa heṭṭhā D I.198
“below that storey”.

In the post-canonical works this gen. is widely used to denote many turns of expression. With reference to space it is used in the sense of “on top of, on, upon” as in kassa upari sāpo patissati Dh A.41, or with reference to time as in catunnaṃ māsānaṃ upari. [194]

§154. The Instrumental-like Genitive.

We have seen how the subjective gen. can sometimes be used for the agent (cp. SS §§66;129.R.2) when the noun qualified is a p.p.p. (§142). When the participle has a predicative force, as is frequently the case in Pāli just as in Skr., the gen. is clearly used instead of the inst. of agent. This is found with all types of verbs. e.g.,

Mayhaṃ kho bhikkhave ... anuttarā vimutti anuppattā S I.105
“the highest emancipation, brethren ... has been attained by me”;

suto nu bhavataṃ Asito Devalo isī ti? M II.156
“have you heard of the sage A.D.?”;

amataṃ tesaṃ bhikkhave aparibhuttaṃ yesaṃ kāyagatāsati aparibhuttā A I.45
“immortality has not been realized by them, by whom full awareness of body has not been experienced”;

kodhā ca pana assa esā vācā bhāsitā M I.68
“through anger were these words spoken by him”;

Tathāgatassa ... cattāro iddhipādā bhāvitā D II.103
“the four bases of supernormal powers have been cultivated by the T.”;

Itthaṃ Bhagavā Sakkassa devānaṃ indassa pañhaṃ puṭṭho vyākāsi D II.279
“in this wise did the Blessed One explain being questioned by Sakka, the lord of gods”;

viditaṃ hi bhante tassa purisassa ... M I.365
“it is known, Sir, by that man ...”;

dīghaṃ addhānaṃ sandhāvitaṃ saṃsaritaṃ mamañc’ eva tumhākañcā ti D II.90
“a long course (of lives) has been traversed and gone through both by me and you”

(Comy. ‘mayā ca tumehi ca’ Sum. II). We may observe the following other idioms expressed by the agent-like gen.:

a. Its employment with participles used as adjectives denoting respect or the opposite has been referred to earlier (§152.b.). A similar gen. is frequently found with verbs (p.p.p.) denoting praise, homage and salutation. e.g.,

Api ca āyasmā Maha-Kāccāno Satthu c’ eva saṃvaṇṇito sambhāvito ca viññūnaṃ sabrahmacārīnaṃ M I.111
“Even so the ven. Mahā-Kaccāna is praised by the Master and esteemed by his co-celibates”;


ye puggalā aṭṭha sataṃ pasatthā Sn 227; Kh 6
“those eight people who are praised by the good”,

where the Comy. has the inst. ‘sappurisehi ... pasatthā’ (Pj. I. p.182). [195]

b. The same gen. is employed with certain other participles which are not strictly passives. e.g.,

ādiso va tesaṃ aparaddhaṃ D I.180
“they are at fault from the very start”;

yāva aparaddhañca te idaṃ ācariyassa ... D I.103
“how deeply have you been wronged by your teacher ...”;

tuyh’ eva etaṃ dukkaṭaṃ D II.115
“this is misdone by you”;

tumh’ evetaṃ aparaddhaṃ D I.222
“by you yourself (it) has been wronged”.

In the last three examples it is very much like the subjective genitive.

c. When the noun in the gen. is attended by another participle, the construction approximates to a semi-absolute gen. e.g.,

aparāmasato c’ assa paccattaṃ yeva nibbuti viditā D I.22 (III.28)
“by him (when he is) free from clinging tranquillity is realized”;

tadapi tesaṃ bhavataṃ ... ajānataṃ apassataṃ vedayitaṃ D I.40
“that too has been experienced by you as such even without knowing or seeing on your part”.

Here the sense “even without your knowing ...”, of the participial phrase shows how closely related it is to the gen. absolute implying disregard (anādara, vide §158.a.), for “even without your knowing” can also be expressed by “in spite of your not knowing ...”. But the fact that the participles agreeing with tesaṃ bhavataṃ merely qualify the ‘agent’ of the verb viditaṃ and are therefore strictly speaking not detached from the rest of the construction shows its difference from the actual gen. absolute.

§155. [Inst.-like Gen.]

This contact of the gen. and the inst. is not restricted to the agent-like gen. described in the preceding paragraph. The gen. is also used in certain other places where in the normal course of concinnity we would have expected an inst. denoting means or the sociative notion.

a. Thus the gen. is found in place of the inst. with certain nouns and verbs which logically must admit of an inst. of means. In such cases it seems to be used on the analogy of the subjective gen. e.g.,

mantassa ājīvino D III.65
“living (by means) of the mantras”;

sabba-cetaso samannāharitvā D II.204 et seq.
“having considered well with his mind”.

Here the v.l. cetasā appears like an attempt at ‘correction’. [196]

b. It may also appear, as remarked above, in place of the sociative inst.. It is even found with or ‘governed’ by the sociative prep. saddhiṃ. e.g.,

mama saddhiṃ sammodiṃsu D I.157
“they conversed with me”.

The idiom ‘sammukhī-bhāvaṃ gacchati’ “comes face to face” or “meets with” would normally admit of an inst. of the person met with coming under the sociative class (cp. missi-bhāvaṃ gato tayā etc. §63.a.). In the following example the gen. mama can also imply possession in a metaphorical sense:

mama sammukhī-bhāvaṃ āgantuṃ D III.13,19
“to come face to face with me”

or “to come to my presence”.

§156. The Genitive of Relation.

In Pāli, as we have seen earlier, the notion of relation (that concerning whom) can be expressed by cases like the acc. or the inst.. The proper case however would be the loc.. There are however certain instances where the gen. too seems to be employed to denote the person concerning whom a statement is made. It is found both adverbally and adnominally.

a. Adverbally it signifies the person regarding whom something is said or meant. e.g.,

Tañ ca kho sīlavato vadāmi no dussīlassa D III.259
“I say it of the virtuous man, not of the evil”;

The frequent phrase ‘ko pana vādo’ also ‘governs’ a similar gen. e.g.,

Idha bhikkhave asappuriso, yo hoti parassa avaṇṇo taṃ apuṭṭho pi pātukaroti: ko pana vādo puṭṭhassa A II.77
“Here, monks, an evil person even unquestioned discloses bad reports of others; what talk of (him when) questioned?”

(i.e. how much more when questioned or what would you say about him if he is questioned); similarly, ko pana vādo manussa bhūtassa M I.227 “what (use is there to) talk of the human being?” (i.e. not to mention the human being)

b. Adnominally it denotes the person to whom something is attributed or of whom some qualification is said to exist. Thus it resembles the possessive gen. e.g.,

acchariyaṃ idaṃ āyasmato Kosiyassa D II.270
“this is wonderful of (or regarding) the ven. K.”;

abbhutaṃ idaṃ āyasmato Janavasabhassa yakkhassa [197] D II.206
“this is marvellous of the spirit, ven. J.”.

A similar adnominal gen. is found with compound ‘vutta-vādino’, where the gen. can also be regarded as used instead of the agent with vutta (p.p.p. of vac). But the more likely explanation is that it denotes the person regarding whom something is said by the speaker (vādin). e.g.,

kacci te bhoto Gotamassa vutta-vādino D I.161
“what! are you true reporters concerning the ven. Gotama?”

(lit. tellers of what is said); vutta-vādī c’ eva Bhagavato homi D III.115 “I am a true reporter (lit. a teller of what has been said) of the Blessed One”. There is one example of a similar gen. which (if it is not a nom. sg. formed from the dat.-gen. stem *bhikkhu-) is to be regarded as a gen. of relation, viz.

Idha bhikkhave bhikkhuno cakkhunā rūpaṃ disvā na nimittaggāhī hoti nânuvyañjanaggāhī hoti yatvādhikaraṇaṃ enaṃ cakkhundriyaṃ asaṃvutaṃ viharantaṃ abhijjhādomanassā pāpakā akusalā dhammā anvāssaveyyuṃ A II.16
“Here, brethren, (speaking) of a monk, he does not take an object (as a theme for reflection) having seen it, ...”.

c. A gen. is found with the idiom ‘kittisaddo abbhuggato’ “a report has been broadcast” denoting the person concerning whom something is reported. Comparing with the gen. found in the phrase ‘ko pana vādo’ discussed above, it may be regarded as denoting the person concerning whom the report is broadcast. But according to ancient grammarians the gen. is due to the accented prefix (karmapravacanīya), viz. abhi- in abbhuggato, as much as the acc. which is concurrently used in such contexts (vide acc. of relation §42.). e.g.,

Mayhaṃ kho ayyāya evaṃ kalyāṇo kittisaddo abbhuggato M I.125
“Concerning (cp. English ‘of’) my lady such a good report has arisen”;

evaṃ te kalyāṇo kittisaddo abbhuggato M I.394
“thus of you a good report has arisen”;

dussīlassa ... pāpako kittisaddo abbhuggato Ud 86
“an evil report arose ... about the unvirtuous one”;

idaṃ me dānaṃ dadato kalyāṇo kittisaddo abbhuggacchati D III.258
“about me ...”.


§157. The Genitive of Time.

The gen. sg. of time-denoting words such as cira, kāla, and divā (§2) is used adverbially to denote after what time or within or during what time an action takes place. In the former, i.e. when it expresses after what time, the gen. is parallel to the abl. as found in cirā (Skr. cirāt) etc., but in the latter sense it is clearly concurrent with the inst. as in kālena, cirena and divā. In the Nikāyas however the abl. of these words is hardly found even in the sense of time after which (even cirā is not attested by cirena), for the inst. has replaced it in such functions. Though in Skr., therefore, it may be said that the gen. here “always stands on the ground of the abl.” (SS §128), the same cannot be stated for Pāli, where the gen. of time may be taken as standing for the inst. with the corresponding temporal sense (§76.a.).

a. The gen. sg. cirassa is only found in the negative phrase ‘na cirass’ eva’ and can be substituted for ‘aciren’ eva’. Similarly it has the sense of “not long after”. e.g.,

na cirass’ eva kālaṃ akāsi D II.195
“not long after he passed away”;

cp. D I.177,202; II.l1,35,153 (cp. Skr. cirasya Macdonell Skr. Gr. §202.5). The compound form sucirass’ eva is similarly found meaning “after a very long time”. e.g.,

atha kho āyasmā Aññāsi Koṇḍañño sucirass’ eva yena Bhagavā ten’ upasaṅkami S I.193
“then the ven. A.K. after a very long time came whither the Blessed One was”.

The extended form cirassaṃ is frequently found in place of cirassa with the same sense. Here we have an interesting instance of the interplay of morphological and syntactical development, for it is certainly the result of a want felt in the mind of the Pāli speakers as regards the form cirassa (with an ending unusual for adverbs) to which they added the nasal (-aṃ) in order to bring it into uniformity with the adv. acc. on the analogy of forms like ciraṃ and muhuttaṃ, though the acc. is logically out of place here. e.g.,

cirassaṃ kho bhante Bhagavā imaṃ pariyāyaṃ akāsi yadidaṃ idhâgamanāya D I.179; S I.142
“It is a long time since (i.e. after a long time) the Blessed One has thought of coming this way”.


b. The form kālassa is found in the sense of “early” very much like the inst. kālena (§76.e.2.). e.g.,

kālass’ eva vuṭṭhahanto A V.263
“rising quite early (in the day)”;

Siṅgālako ... kālass’ eva vuṭṭhāya D III.180
“Siṅgālaka ... having arisen early”.

c. As regards the term divassa, it is always found in the compound expression ‘divā-divassa’ which means “in the day”. The form divā is also an adverb from Vedic divā (§2). e.g.,

Atha kho Sandhāno gahapati divādivass’ eva Rājagahā nikkhami ... D II.36
“then the householder S. set out from R. very early in the day”.

On this passage the Comy, has ‘divassa divā nāma majjhaṇhâtikkamo’ according to which it should mean “after the mid-day”. But our rendering fits in better with the context. Similarly:

Sāvatthiyā niyyāsi divādivassa M I.174 (S I.89)
“he set out from S. early in the day”.

The expression seems to be parallel to a reduplicated form divā-divā with an intensive sense [cp. udagga-udagga etc. W. Stede ‘Reduplikationskomposita im Pāli’ (ZfB Vol.6 (1925) p.89)].

§158. The Genitive Absolute.

The absolute use of the gen. in Pāli is restricted, just as in Skr. (SS §369), to a few standing phrases. Though it is sometimes concurrent with the loc. absolute, it is still far from possessing the general character of the latter. It has been observed with regard to Skr., by writers on syntax, vide, Grammaire Sanscrite §226, by Louis Renou (Paris-1930) and Speyer §369 Sanskrit Syntax, both referring to the exhaustive treatise by F de Saussure ‘de I’emploi de genitif absolu en Sanscrit’. that there are at least two principal conditions governing the use of the gen. absolute, especially with regard to the character of the subject- and predicate-factors that constitute the absolute clause. Firstly, the substantive is almost always the name of a person (or a personal pronoun), very rarely to be supplied. Secondly, the predicate must have a durative sense, that is, it may be either an ordinary present participle or an adj. or a verbal formation having the value of an adj.. Though there are not sufficient examples in the Nikāyas to make a detailed investigation, we may fairly observe that the above conditions are for the most part fulfilled even in Pāli. [200]

a. Strictly speaking, the construction seems to be limited to the expression of action going on but not cared for while performing the main action. Hence local grammarians denote this nuance by the term anādara, i.e. disregard. This nuance is said to express ‘garva rasa’, Renou ibid. In describing the employment of the gen. Kaccāyana lays down the rule anādare ca (307) which the vutti explains as meaning that the sixth case-affix is employed to express the action not fully attended to while performing the main action or the seventh case (anādare ca chaṭṭhi vibhatti hoti sattamī ca). He is here relying on Pāṇinī sūtra (II.38) which says ṣaṣṭau cânādare, the conjunction ca being expressive of ‘option’ as this is primarily the province of the loc. absolute, which is enjoined by the preceding sūtra (37) viz. yasya ca bhāvena bhāvalakṣaṇaṃ (cp. SS p.287.f.n.2). As pointed out above it occurs but rarely in the Nikāyas, and, may be rendered by such expressions as “though, not withstanding, in spite of” and the like. e.g.,

So kho ahaṃ ... akāmakānaṃ mātāpitunnaṃ assumukhānaṃ rudantānaṃ kesamassuṃ ohāretvā ... agārasmā anagāriyaṃ pabbajiṃ M I.163
“despite the parents unwilling and crying I shaved off my hair and beard ... and set forth from home to homelessness”;

seyyathāpi brāhmaṇa puriso daliddo assako anāḷhiyo, tassa akāmassa bilaṃ olaggeyyuṃ M II.178,181,
“it is as (if there were) a poor, needy and destitute wretch and (they) in spite of his not wanting would drop him into a hole”.

(Here the phrase ‘bilaṃ olaggeyyuṃ’ is of very doubtful sense. The P.T.S. Translation has ‘reserve a joint he does not want’, agreeing with Neumann’s ‘man nötigte ihm gegen seinen Willen einen Bissen auf: da hast du, lieber Mann, ein Stück Fleisch zu essen ...’. The reading too being doubtful (vide M II.178) either of the translations seems admissible). In the former example, we may notice, the predicate-factor is a present participle (rudantānaṃ etc.) but in the latter only an adj. (akāmassa).

b. But, as shown by the parallelism with the loc. absolute, in its simple temporal employment the gen. absolute denotes an action going on or a situation existing at the time when the action of the main clause intervenes. Then it may be rendered [201] by “while” or “as” or the participial construction in English (“this being so ...”). Instances of this are however rare in the Nikāyas, being superseded by the loc. absolute. e.g.,

Tatra me brahmaṇa viharato mago vā āgacchati moro vā kaṭṭhaṃ pāteti M I.20
“As I dwell there, brahmin, an animal may approach or a peacock cause a twig to fall”;

tesaṃ vo bhikkhave evaṃ samaññānaṃ sataṃ evaṃ paṭiññānaṃ sataṃ: yā samaṇa-sāmīcīpaṭipadā taṃ M II.281
“When, brethren, you are so famous and acknowledged (for your recluse-ship), whatever right course of life there is for the recluses that ...”.

Sometimes the notion of anādara is faintly implied (cp. SS §369R). e.g.,

āturakāyassa me sato cittaṃ anāturaṃ bhavissati S III.1
“even while I am sickly (in spite of my being sick) of body, my mind is not diseased”.

Here the gen. can also be regarded as possessive (§159.c.).

c. In the above examples the predicative factor has the durative notion and is almost always a present participle. There is however one past participle (pakkanta-) involved in a similar gen. absolute construction. Logically it seems to be a development of the gen. of time already described (§157). Its purely temporal character and origin are to be seen from the fact that the same phrase occurs as a loc. absolute (§183.a.iv.). The only difference (from the simple gen. of time) is that the time-denoting term is not one word but an expression consisting of a personal noun and a participle. The implied sense, however, is fundamentally the same, viz. the time after which or since which (cp. SS §128.R.s). It occurs in the stock-phrase ‘acira-pakkantassa ...’ and the presence of the time-denoting word acira (cp. acirassa()) makes quite clear the parallelism with the simple ‘noun + participle’ construction as found in Skr. ‘ciraḥ kālo maitrasya Vasantasenāyāḥ sakāśaṃ gatasya’ (SS ibid). It occurs in the following passages:

atha kho te paribbājakā acirapakkantassa Bhagavato Poṭṭhapādaṃ sañjambhariyaṃ akaṃsu D I.189
“Then those wandering ascetics, not long after the Blessed One had departed, began to tease P.”;

cp. D II.8; M I.110;

atha kho āyasmā Ānando acirapakkantassa rañño Pasenadissa Kosalassa yena [202] Bhagavā ten’ upasaṅkami M II.117 (cp. D I.86; S III.95)
“then the ven. Ānanda, not long after King P. of K. has departed came where the Blessed One was”.

The use of this past participle in the gen. absolute construction is a later development in Pāli and is contrary to the general laws prevailing in Skr. mentioned above.

§159. [Examples of Genitive Absolute]

The above usage shows, in fact, the transitional stage in the formation of the gen. absolute. Speyer points out that apart from the genuine gen. absolute Skr. upon the whole shows a preference for employing the gen. of participle either as dat.-like gen. or when depending on some substantive. This may also be said of Pāli where there are many instances of the ‘gen. + participle’ which seem to be only semi-absolute constructions (cp. SS §§370; Renou Grammaire Sanscrite §226.Note 1.). The logical relation between such genitives and the main sentence (or a word in the main sentence) “though not wholly wanting is very loose indeed”.

a. The following examples seem to border on the dat. of concern:

tesaṃ bhikkhave sattānaṃ evaṃ kāmānaṃ ... aniṭṭhā ... dhammā abhivaḍḍhanti M I.309
“For such beings desiring thus ... unpleasant mental states increase”

or “notwithstanding their desiring thus ...”;

Imaṃ kho me somanassaṃ sevato akusalā dhammā parihāyanti D II.278
“For me indulging in a state of mental ease evil things decrease”;

gocare bhikkhave carataṃ sake pettike visaye na lacchati Māro otāraṃ D III.58
“To you wandering (or as you are wandering) in such pastures, your own paternal range, the Evil One will not find occasion to enter”.

b. It may even border on the adverbal dat., as for instance, the dat. with verbs of trusting and relying on (having faith in). e.g.,

ye kho pana bhikkhave Sunettassa Satthuno Brahmaloka-sahavyatāya dhammaṃ desentassa cittāni na pasādesuṃ A IV.135
“those who, monks, did not engender faith in their minds when the Master was (or in the Master) preaching doctrine (conducive) to the companionship in the Brahma-world”.


c. Or it may have the possessive sense at the same time. e.g.,

kittāvatā nu kho āvuso Satthu pavivittassa viharato sāvakā vivekaṃ nânusikkhanti M I.14
“how far, friends, do the disciples of the Master living in seclusion not train themselves for seclusion (or whereas the Master lives in seclusion)”.

A curious construction with the singular of the noun and the plural of the participle is found in the following where, contrary to the law prevailing in Skr., the gen. is of a common noun and not of a personal noun, though it is personified by the presence of -rāja-:

Sinerussa ca pabbatarājassa jhāyamānānaṃ dayhamānānaṃ acci vātena khittā yāva Brahma-lokā pi gacchati A IV.103
“Of the King of Mountains, Sineru, (sci. from amongst those) burning and ablaze, the sparks shot forth by the wind go up as far as the Brahma-world”.

Similar is the following:

jhāyamānassa pana sarīrassa, yaṃ ahosi chavīti vā ... tassa n’ eva chārikā paññāyittha na masi D II.164
“of the burning body (or while the body was burning) whatever was the skin etc ... of that there was seen neither soot nor ashes”

(cp. last ex.§158.b.).

d. There are some other instances where the participle alone does duty for the whole construction. This happens when the personal pronoun is of the third person, which is usually dropped. e.g.,

sabhāgatassa vacanaṃ na rūhati D III.185
“Of him (or when he is) gone in the midst of an assembly the evidence is not valid”.

The Comy. supports absolute idea. (‘sakkhipuṭṭhassa sato’).

e. Sometimes it borders on some other use of the gen. itself, as for instance the gen. of origin. e.g.,

Seyyathâpi āvuso sakalikaggikassa jhāyamānassa aññā vā acci uppajjati A V.9
“just as, friends, from (a fire of) logs burning (or while logs are burning) another spark arises.”

f. Elsewhere (§144) we have referred to a loc.-like gen. denoting the persons among whom something happens, or takes place (cp. KVG §537.3). A similar gen. is found involved in a semi-absolute construction. e.g.,

Purimāni bhante divasāni purimatarāni nānātitthiyānaṃ samaṇabrāhmaṇānaṃ kutūhalasālāyaṃ [204] sannisinnānaṃ sannipatitānaṃ, ayaṃ antarā kathā udapādi M II.2 (D I.180)
“In former days quite of yore, among the recluses and brahmins of various sects gathered together (or while they were gathered) in the hall of inquiry, this side-talk arose”;

atha kho Vāseṭṭha-Bhāradvajānaṃ jaṅghā-vihāraṃ anucaṅkamantānaṃ maggāmagge kathā udapādi D I.235
“then while Vāseṭṭha and Bhāradvāja were engaged in a walk, (or between them engaged in a walk) this talk arose on the very way”.