Syntax of the Cases in the Pali Nikayas

Chapter II
The Accusative Case


§29. [General Character]

The primary function of the acc. or the second case (dutiyā) is to denote the direct or near object, in other words, that which is mainly affected by the verb-concept. Brugmann favours the term grammatical object in this connection (KVG §866). The name kamma (Skr. karman) of the local grammarians, on the other hand, strictly means the logical object, whatsoever may be its grammatical function; it implies the object of the active verb as well as the subject of the passive or the objective genitive (cp. SS p.l, f.n.). Apparently Indian grammar did not distinguish between the grammatical and the logical object, just as it did not make any difference between the subject and the agent, both being commonly denoted by the same term kattā (Skr. kartṛ) meaning literally “doer” (vide §17). The logical connection between the object and the verb, just as in the case of the subject and the latter, varies according to the nature of the action (kriyā). The acc. is more conspicuously adverbal than any other case. Even in its adnominal function the other noun to which it is connected bears a marked predicative character, being usually a verbal noun, agent-noun or some such formation.

§30. [Local Grammarians]

“That which the agent does or performs is the object” says Kaccāyana, yaṃ karoti taṃ kammaṃ. Kac. 282. while Pāṇinī karturīpsitatamaṃ. Pāṇ. I.4.49. defines it as “That which is most desired (sci. to be affected) by the agent”. Moggallāna has no special rule for this purpose but lays down in the vutti karīyati kattukiriyāyâbhisambandhīyatīti taṃ kammaṃ, on the sutta kamme dutiyā. Mog. II.2. that “what is done by, or is intimately connected with, the action of the agent is the object”. It may be remarked that none of these gives a comprehensive idea of the logical function of the object but rather tries to explain the connection between the agent and the object. This is due to the common characteristic of all Indian grammarians, [39] namely, dealing with syntax only from the point of view of the agent and not viewing the sentence as one whole psychological unit. Strictly speaking Kaccāyana’s definition holds good only for verbs like making, doing, performing etc., while Pāṇinī and Moggallāna come closer to, if not actually anticipate, the modern conception, namely, that the noun-concept that is affected by verbs of influence is the object and is put in the acc. case (vide KVG §560). But the defect of Pāṇinī’s method is evident from his next sūtra “Or that which is not desired if it is not connected with the verb”, illustrated by “viṣaṃ bhakṣyati” and “caurān paśyati”, whereby he attempts to surmount the difficulty caused by the narrow meaning of the term ‘īpsitatamaṃ’ (most desired). Patañjali however thought that this rule might be done away with. vide Philosophy of Sanskrit Grammar, p. 258. As regards Kaccāyana’s definition, the only justification is that the verb of physical action is best instanced by the root kṛ (karoti) “to do, to make”. All the examples he adduces contain either that verb or an equivalent. But Moggallāna appears to have taken a wider view, for his illustrations extend to such verbs as pacati “cooks” and passati “sees”. As usual, of course, both the Pāli grammarians for the most part merely repeat the conventional illustrations of Pāṇinī (e.g., kaṭaṃ karoti).

§31. The Acc. of External Object.

Verbs of direct agency or verbs of influence (i.e. affecting) in their primitive or non-causative state may take an external object (cp. KVG §561) that is, the noun put in the acc. case may denote an actual, physical or material, object such as a person, thing or place. This may be called the concrete external object. e.g.,

mā Tathāgataṃ vihesesi M I.332
“do not harass the Tathāgata”;

gāviṃ vadhitvā M I.58
“having killed a cow”;

gattāni anomajjāmi M I.80
“I rub down the limbs”;

te bhoge rājāno vā haranti, ... aggi vā dahati, udakaṃ vā vahati M I.86
“those riches either kings (will) seize, ... the fire burn or the water carry away”;

thusodakaṃ pivāmi M I.77
“I drink gruel”;

andhakāre telapajjotaṃ dhāreyya M I.512
“would hold a lamp in the darkness”;

kaṭṭhāni pāḷenti Sn p.104
“chop wood”.


a. The same external object may represent an abstract concept such as a mental state in which case we have an abstract external object. e.g.,

Samaṇo Gotamo dhammaṃ deseti M I.77
“The recluse Gotama preaches the doctrine”;

abrahmacariyaṃ pahāya M I.345
“having abandoned immoral conduct”.

b. Some of these verbs of influence or affecting may permit of an etymological object. e.g.,

āhāraṃ āhāreti D I.166;II.203
“takes food”;

-bhojanaṃ bhojimha M I.367
“I ate a meal”;

mantaṃ manteyya D I.104
“would utter a charm”.

§32. [Object]

There are some other verbs which can be only in an abstract sense called verbs of affecting. Such are those of seeing, knowing, perceiving, hearing and the like. With these the actual effect of the action is more or less on the agent himself and the physical object is merely the cause for that influence. e.g.,

satte passāmi M I.504
“I see beings”;

dhammapariyāyaṃ sutvā M I.83
“having heard the doctrinal system”;

jātisamudayaṃ pajānāti M I.50
“knows the origin of birth”.

Local grammarians denote this function by the term ‘viṣayatva’ or “the state of being the object of cognition”. vide Philosophy of Sanskrit Grammar, p. 199.

§33. The Adnominal Acc. of Ext. Object.

In the above example the acc. is used adverbally, that is, the noun-concept is the object of a proper verb which is the predicate of the sentence or the clause. Corresponding to each of these we may have an adnominal construction, where the acc. represents the object not of an actual verb but only one implied by a verbal noun, agent-noun or a similar formation of a verbal character. In such cases there is the alternate construction with the gen. of the object (§143). Certain dependent (tatpuruṣa) determinative compounds which preserve the acc. of the first member (aluk-samāsa), such as piyaṃ-vada D II.163 “pleasant-speaking”, also belong to this class. In the adnominal construction too we may have all the logical differences as were found in the adverbal such as the concrete, the abstract and so on. Here the second member [41] which is really the verb-concept or the predicative element though employed nominally may be either an agent-noun, verbal substantive, verbal adjective, or any other secondary formation such as those with the suffixes -ka and -ika. e.g.; Agent-noun:

samaṇaṃ vā brāhmaṇaṃ vā apasādetā D III.44
“reproacher (of) recluse or brahmin”;

samagga-karaṇiṃ vācaṃ bhāsitā D I.64; 154
“speaker (of) uniting speech”;

aññe samaṇabrāhmaṇe pucchitā D I.51
“questioner (of) other recluses and brahmins”;

attānaṃ āvikattā M II.125
“revealer (of) himself’;

pāṇaṃ jīvitā voropetā M II.103
“depriver of life (of) being(s)”;

Verbal noun:

Gotamaṃ dassanāya D I.113
“for the purpose of seeing Gotama”;

dassana-kāmā hi mayaṃ taṃ Gotamaṃ D I.150
“we are desirous of seeing that Gotama”.

Here the expressions dassanāya and dassana-kāmā are almost infinitives of purpose (cp. PLS §77). Verbal adjective:

bhattaṃ bhuttāvissa D II.127
“of one-who-has-taken (his) meal”;

Secondary formations:

ārādhako hoti ñāyaṃ dhammaṃ kusalaṃ M II.197
“is one undertaking (accomplishing) the method, the law and what is good”.

§34. The Acc. of Internal Object.

When the noun-concept is represented as the result of an action and that result outlasting the process of the action exists separately, it is called the acc. of internal object (cp. KVG §561.2.). This is known as nirvavttya karman according to orthodox grammarians. In an example like ghaṭam karoti “makes a pot” it is held that the pot is not logically supposed to have existence prior to the action, and, therefore, it is not strictly correct to take ghaṭa as an example of karmakāraka. vide Philosophy of Sanskrit Grammar, p. 218. The contention appears to be reasonable when we consider the fact that it is actually not the pot which is made but the clay which is made into a pot. But this however does not affect the actual grammatical relation existing between karoti and the object. Naturally enough, therefore, the internal object is only found with such verbs as those of making, building, constructing, creating and the like, e.g.,

nivāsanāni māpetuṃ D II.87
“to build dwellings”;

Samīti yānakāraputto rathassa nemiṃ tacchati M I.31
“Samīti the [42] cartwright(-son) fashions a rim for the chariot”;

app’ekacce uddhanāni khaṇanti Sn p.104
“some dig out fireplaces”.

a. Just as in the case of the external object this can also appear in an abstract sense. e.g.,

kopaṃ na pātukaroti M I.125
“does not manifest anger”.

Here the noun and the verb are frequently etymologically related. e.g.,

udānaṃ udānesi D II.186
“uttered a solemn utterance”;

anuttaraṃ sammāsambodhiṃ abhisambuddho D II.140
“realized unsurpassed and perfect Enlightenment (lit. awoke to the highest Awakenment)”.

b. These constructions may occur adnominally as well. e.g.,

na dhūmaṃ kattā hoti M I.220
“is not a producer of smoke”;

pāpakaṃ cittaṃ uppādetā M I.79
“producer (of) evil thought”.

§35. The Object of Contents.

The same internal object may represent a nominal concept that exists only during the verbal action. The noun put in the acc. thus denotes the process by which the action is manifested. Since it is contained within the action it has been called the object of contents (KVG §561.2.b.). By their very nature of being simultaneous with the verbal process they consist generally of cognate uses. e.g.,

vitakkaṃ vitakkessati M I.122 lit.
“will think a thought”

i.e., “reflect”;

loka-cintaṃ cinteti S V.447
“thinks a worldly thought”;

mā pāpakaṃ akusalaṃ cittaṃ cinteyyātha S V.418
“do not think any evil, unmeritorious thought”;

dussutaṃ assumha M II.185
“we have heard a bad thing (lit. a bad hearing)”;

vedanaṃ vedeti M I.90
“he feels a sensation”;

sīhanādaṃ nadeyyaṃ D III.23
“roars a lion’s roar”;

karuṇaṃ jhānaṃ jhāyati D II.239
“he meditates (a meditation) on compassion”;

kumāra-kīḷikaṃ kīḷi D II.96
“he played children’s games”;

brahmacariyaṃ caritvā D II.208
“having lived the Higher Life”;

caratha bhikkhave cārikaṃ D II.45,250
“wander forth, O monks, (lit. the wandering)”;

nāgâpalokitaṃ Vesāliṃ apaloketvā D II.122
“having taken (lit. looked) a majestic (lit. elephant’s) look at Vesāli”.

a. Sometimes the verb used may not be the same as the radical element of the noun but may be cognate only in sense. e.g.,

cetiya-cārikaṃ [43] āhiṇḍantā D II.141
“travelling on pilgrimage”;

pahāraṃ dadeyya M I.124
“would give a blow”;

seyyaṃ kappemi M I.78
“I make my bed”.

b. To this class also belong certain compound verbs where the first member is the acc. singular of a substantive standing as internal object, in most cases cognate, to the verb karoti, which here does not contain the full meaning of “to do” or “to make” but merely expresses the cognate idea contained in the noun. e.g.,

padakkhiṇaṃ katvā D II.40,163
“having gone round to the right”

(i.e. having paid his respects);

dukkhass’ antaṃ karonti D II.252
“they make an end of Sorrow”;

ātappaṃ akaruṃ D II.256
“they made an effort”;

rajjaṃ kāreyya D II.140
“would administrate the government”;

kālaṃ karissati D II.93,140
“he will die”;

in verse the acc. may appear after the verb, thus:

taṃ jano kurute piyaṃ Dh 217
“him the people love”,

which should normally stand as taṃ jano piyaṃ kurute; hence the Comy. glosses it by piyaṃ karoti (vide Dh Comy. III.286).

§36. Secondary Uses of the Acc. of Object.

Many uses of this acc. seem to border on the logical spheres of other cases. The following division may be admitted:

a. With verbs of asking, begging, imploring, learning etc. the person asked and so on is denoted by the acc.. Here the abl. as found used in Sanskrit (SS §95.5.) appears to be the more logical construction. e.g.,

bhikkhate pare Dh 266
“he begs (from) others”;

Bhagavantaṃ yāci D II.104
“he implored the Blessed One”;

sakaṃ ācariyakaṃ uggahetvā D II.104,105
“having learnt from his teacher”.

With verbs of fearing the source of fear, if it is a thing, is optionally construed with an abl. (§122.e.) or a gen. (§150.c.), but if it is a person the acc. seems to be the more proper case. e.g.,

bhāyasi maṃ samaṇâti S I.207; Sn p.48
“do you fear me, recluse?”

b. The inst. is the usual case to signify the person with whom one converses or talks (§63.a.iv), but the acc. is used to denote the person talked to, implied by the prefix ā- (cp. Kac. 309). e.g., [44]

sace maṃ samaṇo Gotamo ālapissati ahaṃ pi taṃ ālapissāmi S I.77
“if the recluse Gotama talks to me I shall also talk to him”.

Other examples of such verbs taking the acc. will be given later (§58.c.). Logically related to such uses is the acc. of the person addressed (cp. SS §42). e.g.,

bhikkhū āmantesi D II.90
“addressed the monks”.

c. Sometimes the acc. is used where the dat. is also admissible. With the verb saddahati “to have faith in or to trust” the normal idiom is the dat. (§94.a.) or the loc.(§166.c.) of the person trusted, but the acc. of the thing. e.g.,

saddahati Tathāgatassa bodhiṃ A II.66
“has faith in the Enlightenment of the Tathāgata”.

With verbs of teaching the older idiom is the dat. of the person taught but the acc. is also found. e.g.,

amhe vācessasi S I.120
“you will teach us”.

It has to be remarked here, however, that the verb has the causative form and the acc. may denote the original subject of the primitive verb, in which case the meaning will be “you will cause us to recite”.

d. In Pāli just as in Sanskrit (vide SS §42) certain verbs normally regarded as ‘intransitives’ occur with a different sense construed with the acc. denoting an object. The acc. with most of these seems originally to have implied relation. Such verbs are: roceti in the secondary sense of “approve of”. e.g.,

kassa vā tvaṃ dhammaṃ rocesi M I.170; S I.133
“whose doctrine do you approve of”;

cinteti in the sense of “think about, reflect over, consider”. e.g.,

diṭṭhigatāni cintayanto Sn 834
“thinking about the heresies”;

rodati in the sense of “cry about or at, lament over”. e.g.,

mataṃ vā amma rodanti ... kasmā maṃ amma rodasîti Th 1. 44
“they weep for one dead ... why mother do you weep for me”;

arahati in the sense of “be as worthy as”. e.g.,

arahasi vā maṃ tvaṃ na vā maṃ tvaṃ arahasîti D I.99
“are you held as worthy as I or are you not ... ?”;

gacchati in the ‘transitive’ sense of “walk, traverse”. e.g.,

maggaṃ kho pana me gacchantassa kāyo kilamissati D III.255
“of me going the way the body will be tired”,

which may logically be regarded as a cognate use. [45]

§37. The Acc. of Direction.

Verbs which imply motion towards any object which may be a person, place or condition take an acc. of such a noun-concept. This has been also called the acc. of the aim (SS §40). Verbs of this type are usually those of pure motion such as the following frequently employed in the Nikāyas: gacchati with or without the prefixes upa-, anu-, adhi-, etc.; kamati usually with abhi-, ava- upa+saṃ-; yāti with abhi-, ni(s)-, ud-, abhi+ud-; eti with upa-; sarati with anu-; dhāvati with anu-; ruyhati with ā; gāhati with ava-; pajjati (pad) with ā-, ud-, paṭi-, upa-; visati with or without prefixes; vattati with pa-; kirati with ava-; vassati usually with abhi+ni-. In most of these, especially in the last three examples, the acc. is usually regarded as being due to the prepositional force of the prefixes. But in the earlier language a verb of the type of varṣati was capable of taking the acc. even without any prefix (cp. SS 42). These prefixes are generally supposed to change the so-called intransitive verbs into transitives. According to this view the acc. with the above verbs when they occur with any of the said prefixes is due to its being the object. But in Old Indo-Aryan such indeclinables as ā, adhi, anu, pari, pra etc. are found both as prepositions and postpositions just as in Old Greek. In Indo-European the so-called prepositions were merely directional adverbs referring to an action but later came to attach themselves and ‘govern’ particular case forms of nouns like the acc., inst., abl., gen., and loc., thus losing their separate and independent existence as help-words in the sentence. vide S. Chatterjee, Origin and Development of the Bengali Language, Vol. II §509. Their pre-verbal use is due to the peculiar syntactical relation they bore to the verb as adverbs. Classical Sanskrit shows fewer particles with a prepositional or postpositional employ than Vedic (vide VSS §§85 - 90). In Pāli they are hardly used as adnominal prepositions.

§38. The Acc. of the Aim.

Local grammarians consider the varied logical functions of the acc. with such verbs under the common category of gati i.e. motion (vide Kac. 302). Nevertheless according to the various [46] meanings implied by the verb, with or without prefixes, these uses differ from one another in the syntactical relations expressed.

a. The simplest use of this acc. is to denote the person, place or thing to or towards which motion is directed. This may be called the terminal acc. e.g.,

pokkharaṇiṃ āgamissati M I.76
“will go to the pond”;

so āḷāhanaṃ gantvā gantvā M II.60
“he having gone continually to the cemetery”;

dakkhiṇaṃ janapadaṃ gantvā D I.96
“having gone to the southern country”;

khattiya-parisaṃ upasaṃkamitvā D II.109
“having approached (lit. walked to) the assembly of princes”;

uyyāna-bhūmiṃ niyyāsi D II.179
“went (lit. down) to the pleasure-ground”;

yaṃ paṭidhāveyyātha M III.9
“whom you would run back to”;

maṃ Mithilaṃ paṭinetu M II.80
“lead me back to Mithilā”;

na heṭṭhāpāsādaṃ orohati D II.21
“he does not descend to the lower storey”;

pāsādaṃ āruyha D II.39
“having ascended to the terrace”;

paṭhavikāyaṃ anupeti D I.55,180
“goes to an earthly body”.

Sometimes two accusatives of aim may appear with the same verb in the same sentence. e.g.,

abhikkāmuṃ bhikkhūṇaṃ samitiṃ vanaṃ D II.256
“they went to the meeting of the monks to the forest (i.e. in the forest)”.

A similar double acc. construction is found with the stock phrase saraṇaṃ gacchati. e.g.,

Bhagavantaṃ saraṇaṃ gacchāmi A III.242; M I.290; D I.116
“I go to the Blessed One as refuge”;

saraṇaṃ taṃ upema Sn 31
“we go to you as refuge”.

Commenting on the frequent phrase Buddhaṃ saraṇaṃ gacchāmi Buddhaghosa makes the following observation: Gamanīyadīpanāyaṃ codako āha: Buddhaṃ saraṇaṃ gacchāmīti ettha, yo Buddhaṃ saraṇaṃ gacchati, esa Buddhaṃ vā gaccheyya saraṇaṃ vā; ubhayathā pi ca ekassa vacanaṃ niratthakaṃ, kasmā gamanakiriyāya kammadvayâbhāvato, na h’ ettha ‘ajaṃ gāmaṃ netîti’ ādisu viya dikammakattaṃ akkharacintakā icchanti, gacchat’ eva pubbaṃ disaṃ, gacchati pacchimaṃ disanti ādisu sâtthakaṃ evâti ce” Paramatthajotikā I.pp. 17-18. “In elucidation of the gamana-formula a critic has urged: In the phrase Buddhaṃ saraṇaṃ gacchāmi, he who goes to the Buddha for (lit. to) refuge, may go either to the Buddha or for (lit. to) refuge. In either case the (use) of one word is meaningless, because there is no double object for the action of going; nor do the grammarians find here a double object as in ‘ajaṃ gāmaṃ neti’ etc.. It is however (a) proper (construction) as in the case of ‘he goes to the east and he goes to the west’”. Thus he regards both as accusatives of direction.

b. Sometimes however the termination denoted by the acc. may be unspecified. In other words the mere direction may be [47] only implied, which is generally due to the indefinite nature of the noun-concept. e.g.,

puratthimaṃ disaṃ pavattati D II.172
“rolled on towards the eastern direction”;

nabhaṃ abbhussukkamāno D II.183
“ascending to(-wards) the sky”.

c. With certain verbs of going having the prefix abhi- it denotes the person or party against whom one marches etc. e.g.,

asurā deve abhiyaṃsu S I.216
“the demons marched against the gods”;

rājānaṃ Pasenadiṃ Kosalaṃ abbhuyyāsi S I.93
“he marched against King Pasenadī of Kosala”.

d. When they have the prefix anu- the acc. denotes the object or person that is pursued. This application may cover such English usages as “in the wake of” or “after”. e.g.,

mā sandiṭṭhikaṃ hitvā kālikaṃ anudhāvi S I.9
“do not leave the present and run after the future”;

phalaṃ anusarati S IV.303
“follows up the result”.

It is also found with the verb anubandhati where the original root (bandh “tie, connect”) is not one of pure motion but gains such an implication by the addition of the prefix anu- through the figurative sense of “start after”, hence “follow”. e.g.,

Bhagavantaṃ anubaddhā honti D II.102 (cp. D I.1)
“they were following (lit. started after) the Blessed One”.

e. When the prefix adhi- is added to the verb of motion the acc. logically borders on that of the external object. Such verbs are generally treated as ‘transitives’. e.g.,

madhupiṇḍikaṃ adhigaccheyya M I.114
“he would obtain a ball of honey (lit. come to)”;

kusalaṃ dhammaṃ adhigaccheyya D I.224
“he would attain to something good”, lit. “he would come to or up to something good”.

§39. Acc. of Direction in Abstract Sense.

In certain metaphorical constructions we find the acc. used with similar verbs of motion, implying the aim in an abstract sense. e.g.,

āgato imaṃ saddhammaṃ M I.47
“come to this good doctrine”;

so yasaṃ paramaṃ patto Sn 138
“he attaining to the highest glory”;

paṭhamajjhānaṃ samāpajji D II.156
“attained to the first ecstatic state”;

uccâvacaṃ āpajjati D II.283
“comes now [48] to the high now to the low”;

pallomaṃ āpadiṃ M I.20
“came to (i.e. got) confidence”;

bhavadiṭṭhiṃ upagatā M I.65
“come to the false doctrine of existence”;

gārayhaṃ thānaṃ āgacchati D I.161
“comes to a blameworthy position”;

vuddhiṃ virūḷhaṃ ... āpajjissathâti D II.63
“would attain to growth and magnitude”;

vosānaṃ āpādi M I.196
“came to the end”;

visādaṃ vā pāpuṇāti D I.248
“he comes to grief (lit. dejection)”.

Various other nuances of the same use may be observed:

a. Certain other abstract uses of the acc. of direction consist of stereotyped idioms where it merely forms part of the verb. Such compound verbs generally have the implied sense of considering or reckoning. Logically they can be compared to the compound verbs with karoti (§35.b.). e.g.,

saṅkhaṃ upeti S III.93
“is reckoned as (lit. goes to the category)”;

saṅkhaṃ gacchati D I.200
“is considered as”;

saṅgahaṃ gacchati M I.184
“is considered as (within)”;

samodhānaṃ gacchati M I.184; A I.234
“is included”.

b. The acc. construed co-ordinately with the inst. in idioms of the type of kālena kālaṃ also belongs to this group. e.g.,

te kālena kālaṃ upasaṅkamitvā D III.60
“they having come from time to time”,

where the Comy. has ‘kālena kālaṃ ti kāle kāle’ (Sum.III.851). Even in the following the acc. really belongs to the idiom aṅkena aṅkaṃ and is not necessarily the acc. of direction with the verb pariharīyati “is carried”:

aṅkena aṅkaṃ pariharīyati D II.20
“is taken care of (or carried) from hip to hip”.

But the normal sense of pariharati in the Nikāyas is generally the former, that is, “attends to, takes care of, etc.” (vide P.T.S. Dict. s.v.).

§40. The Acc. of the Goal.

The acc. may also denote the place or state entered, that is, the actual goal into which motion is implied. This is mostly found with verbs of descending, entering, falling, throwing, being born into and the like. It is only to be expected here that the loc. is, if not the more frequent, at least the concurrent construction (§165). e.g.,

mātu-kucchiṃ okkamati D II.108; III.103
“he enters into the mother’s womb”;

pokkharaṇiṃ otarituṃ S I.123
“to descend into [49] the pond”;

ogahe te thanūdaraṃ D II.266(V.)
“I would enter into thine bosom”;

samuddaṃ ajjhogahetvā D II.173
“having descended into the ocean”;

gehaṃ pavisante M II.178; D I.83
“entering the house”;

vihāraṃ pavisitvā D II.143
“having entered the monastery”;

upari-vehāsaṃ khitto S II.184
“thrown into the sky above”;

aṅgārakāsuṃ papatissāmi M I.65
“I shall fall into the pit of charcoal”.

When the noun-concept in the acc. denotes a person the use is more or less abstract. e.g.,

Tatra vata maṃ bhayaṃ vā sārajjaṃ okkamissati M I.72
“In that matter, fear or flurry will come upon me” (lit. enter into me).

a. With the verb uppajjati “be born”, the acc. and the loc. are promiscuously employed in the Nikāyas. Logically however the former’s application seems to be due to the original meaning of “attain to” (uppajj- < Skr. ud+pad = step up) which develops into the sense of “be born into” (cp. Skr. utpāda = birth), and, the latter (loc.) is construed with it as denoting the place where one is born (§163.a.iii) according to this derived meaning of the verb. In any case the acc. here signifies direction in the wider sense. e.g.,

Tusitaṃ kāyaṃ uppajjati D II.14
“is born into the Happy Abode”;

Vessavaṇassa Mahārājassa sahavyataṃ uppajjāmi D II.206
“I am born into the company of Vessavaṇa, the Great King”;

saggaṃ lokaṃ uppajjanti D II.142
“they are born into the heavenly world”;

nirayaṃ uppannaṃ D I.162; III.264 (cp. D II.208; A IV.75)
“(him) born into hell”.

b. The acc. denoting the surface on to which motion is implied also comes under this category. It is found with verbs of the type of abhivassati “rain on to, upon” and okirati “scatter upon”. Here too the loc. can be optionally employed, e.g.,

mahiṃ iva sūro abhivassaṃ D III.160
“like the rain showering on the earth”;

sarīraṃ okiranti D II.137
“fall on (to) the body” (lit. scatter).

§41. Secondary Uses of the Acc. of Direction.

With verbs having the sense of bending towards, leaning, depending, hanging on, clinging on to and tending to the acc. logically implies direction just as in the examples adduced in the preceding [50] paragraphs. Speyer puts such uses on the boundary between the acc. of near object and the acc. of the aim (SS §40). Though these verbs do not imply any motion in the fundamental sense, still, being of a dynamic character, they are capable of taking an acc. of the thing on to or towards which ‘influence’ (in the sense that Brugmann (KVG §560) uses the term) is meant by the action. Speyer (loc. cit.) says that in Skr. the acc. can be used after the primary verb śrayati without any prefix in the sense of bending towards, e.g., tvāṃ śrayati “he bends to you”. But in Pāli it always occurs with prefixes, e.g.,

Brahmaṃ ajjhosissāmi M I.328
“I am attached to (lit. bent on) Brahma”;

dvayaṃ nissita S II.17; III.134
“depending on both”;

vibhavadiṭṭhiṃ ajjhositā M I.65
“holding on to the view of non-becoming”;

bhavadiṭṭhiṃ allīnā M I.65
“clinging on to the view of becoming”;

kapisīsaṃ ālambitvā D II.143
“hanging on to or leaning against the door post”;

sattā dhātuṃ abhinivisanti D II.282
“beings hold on to an element”.

The primary verb namati however can take the acc. even without any prefix in the sense of “bowing”. e.g.,

Tathāgataṃ namassantā D II.20
“bowing (down) to the Tathāgata”

(cp. Skr. śrayati, above). It is to be remarked that in the adnominal construction with namo it is not the acc. but the dat. that is commonly employed (§112). Even adverbally the latter is concurrently used (§96.b.). The verb atthu (imperative) in conjunction with nouns of blessing and greeting as well as their opposites may take an acc. of the person unto or on whom such a wish is meant to be conferred. e.g.,

bhavaṃ atthu bhavantaṃ Jotipālaṃ mānavaṃ D II.231
“May good fortune attend the honourable Jotipāla”.

It is clear here that the acc. is due to the ‘direction’ implied in atthu, a construction parallel to the dat. of advantage which is elsewhere frequently applied in such connections (cp. Comy. ‘bhoto Jotipālassa ... hotu ... ‘ Sum.II. p.660). We may compare with this the usage dhī-r-atthu mama jīvitaṃ Sn 440 “fie! on my life” where too the acc. is due to the ‘direction’ implied in atthu rather than to the force of the exclamatory particle dhi (cp. Skr. dhik c. acc., SS §417.2.). [51]

§42. The Acc. of Relation.

We have already referred to certain uses of the acc. with verbs of speaking etc., where it seems to have originally implied the idea of relation as denoting the thing or person referred to by the action rather than the object proper, (§36.d). Though this construction is undoubtedly pre-Indo-Aryan (cp. KVG §561.5., Latin Acc. of Respect), still in Skr. it had for the most part gone into disuse, the loc. being there the normal case for that function. Moreover, such verbs as those capable of being thus construed became ‘transitives’ owing to the addition of prefixes (§37), and the acc. came to be regarded as being due to their prepositional force rather than to the special character of the verb. In most of these instances Indian grammarians explain the acc. as being due to the accented prefixes (karmapravacanīya cp. Pāṇ.II.38). A striking example of this is found in the frequent passage

taṃ kho pana bhavantaṃ Gotamaṃ evaṃ kalyāṇo kittisaddo abbhuggato D I.87; M II.83
“to this effect has the good report arisen concerning that venerable Gotama”,

which the vutti on Kaccāyana 301 takes as an instance for the construction ‘kammappavacanīyayutte. But the same occurring at Sn p.103 the commentator (Buddhaghosa) styles as ‘itthambhūtayoge’ (vide Paramatthajotikā II.2. p.441).

a. The acc. of the person referred to with the verb vadati can therefore be regarded as originally denoting the person concerning whom a statement is made. The noun clause ending with iti consisting of that statement is here the actual object. e.g.,

kulaputto ti bhikkhave Nandaṃ sammā vadamāno vadeyya A IV.166
“one could, brethren, rightly say speaking of Nanda that he is the scion of a noble family”;

that the iti clause is the actual object and not the noun denoting the person, is shown by the use of taṃ in the following:

Sāriputtaṃ eva taṃ sammā vadamāno vadeyya M III.29
“one could, rightly speaking, say this of Sāriputta”;

asammohasatto loke uppanno ... sukhāya devamanussānanti maṃ eva taṃ sammā vadamāno vadeyya M I.21
“one could say this, rightly speaking, of me: that an undeluded being is born in the world ... for the happiness of gods and men”.


b. With the verb vyākaroti the acc., though logically denoting the person about whom something is declared, can also be regarded as being the external object of that verb in the sense of “reveal”. e.g.,

paricārake abbhatīte kālakate uppattīsu vyākaroti D II.200
“makes declarations as to the rebirths of (lit. concerning) such followers as had passed away”;

iti maṃ jano jānātûti sāvake abbhatīte kālakate uppattīsu byākaroti M I.465
“makes declarations as to the births of such disciples as had passed away (saying): thus may the people know me”.

c. A more logical instance of an acc. of relation occurs in the following:

te aññamaññaṃ migasaññaṃ paṭilabhissanti D III.73
“they receive the impression of a deer with respect to each other”;

or adnominally:

aññamaññaṃ agāravā Th 1.976
“disrespectful towards each other”.

From these translations it becomes clear that the acc. here stands parallel to the loc. known as nimitta-sattamī (§177.a). This is supported by the v.l.

aññamaññamhi at D III.73
“with regard to each other”,

which also shows that it is unnecessary to consider the acc. aññamaññaṃ as an adverb.

§43. Acc. of Relation with Reflexive Participles.

An acc. is also found in the Nikāyas with certain reflexive participles, denoting originally an external object with the indicative forms of the verb but with the participles appearing as accusatives of relation. This is the exact counterpart of the Latin construction already referred to (cp. manūs victus = bound with respect to the hands). e.g.,

cakkhundriyaṃ asaṃvutaṃ viharantaṃ D I.170; A II.16
“dwelling unrestrained with respect to the sense of sight”,

where the loc. is concurrently used (cp. cakkhundriye saṃvaraṃ āpajjati);

yaṃ pan’ assa khamati taṃ gathito mucchito ... D III.42
“being fettered and infatuated as to that which pleases him”,

where too the loc. is parallel (vide P.T.S. Dict. s.v.);

sīsaṃ nahātassa D II.160,172
“of him (who was) bathed with respect to his head”;

sīsaṃ pārutaṃ S I.167
“covered as to the head”;

here the v.l. sasīsaṃ looks like an attempt to make the concinnity more normal regarding the acc. ending -aṃ as adverbial. A similar acc. occurs with puṭṭho “asked”: e.g.,

labujaṃ vā puṭṭho ambaṃ [53] vyākareyya D I.55,56
“inquired as to a gourd, replies about a mango”;


yo atthaṃ pucchito santo anatthaṃ anusāsati Sn 126
“who being questioned (regarding) what is useful admonishes (regarding) what is useless”.

§44. Acc. of Extension in Space or Time.

This use of the acc. shows most clearly the original idea of the case, viz., to restrict the application of a verb or a noun to a certain length of space or time. With regard to the latter conception it is usually designated the acc. of duration of time, which Macdonell considered to be in origin only a special form of the cognate acc. (VGS §197.2.). He adduces such examples as

śataṃ jīva śarádo várdhamānaḥ
“live prospering a hundred autumns”

(R.V. X.16104), where certainly the cognate sense is clear enough. But in the example

tásmāt sárvān ṛtūn varṣati
“therefore it rains during all the seasons”

no cognate idea is manifest, because the verb-concept varṣati has no semantic connection whatever with the noun-concept ṛtūn. This shows that the cognate acc. is not the only source from which the duration idea may have developed. Pāli grammarians call this kāladdhānamaccanta-saṃyoge dutiyā vibhatti, “the second case applied to (denote) the extension in time and space”, (Kac. 300; Mog. II.3; cp. Pāṇ. II.3.5.).

a. Extent of Space

This may denote the space traversed or the range over which an action is executed. Here the acc. is clearly seen to limit the application of the verb. e.g.,

yugamattañ ca pekkhati M III.137
“he fixes his gaze within the range of a yoke”;

samantā yojanaṃ passati D II.20
“he sees for a yojana on every side”;


sattaporisam pi mahāsamudde udakaṃ saṇthāti A IV.102
“there is water in the ocean to a depth of seven times the height of a man”;


dvādasa yojanāni āyāmena D II.146
“twelve leagues in breadth”.


b. Extent in Time

This usually signifies the time during which an action continues or is carried on. e.g.,

satipaṭṭhāne evaṃ bhāveyya satta māsāni M I.63
“and so for seven months one should develop the bases of mindfulness”;

so cattāre māse parivasati D I.176; II.152
“he dwells for four months”;

tiṭṭhatu Bhagavā kappaṃ D II.103
“may the Blessed One live for an aeon”;

imaṃ rattiṃ dukkhaṃ sessati S I.83
“he will lie in grief during this night”;

api pana tumhe āyasmanto ekaṃ vā rattiṃ ekaṃ vā divasaṃ ekanta-sukhiṃ attānaṃ sañjānāthâti D I.194
“would ye, brethren, realize yourselves complete bliss for one night or one day?”;

bahu-d-eva rattiṃ dhammiyā kathāya sandassetvā M I.354
“For the greater part of the night, having exhorted (them) with discourse on the Doctrine”.

This use of the acc. is not far removed from its adverbial function as seen from the following examples:

ayaṃ vammīko rattiṃ dhūmāyati M I.142
“this ant-hill smokes during the night or by night”;

tena hi bho muhuttaṃ āgametha D III.20
“therefore, friend, wait for a moment”.

[It is also found in certain cognate uses. e.g., divā-vihāraṃ nisīdissāmi D II.182 “I shall sit through the day’s sitting”.]

In the following the conceptions of time and space are linked together:

dīghaṃ addhānaṃ sandhāvitaṃ D II.90
“run through a long period”.

It is used adnominally in

āyasmā Upavāno dīgharattaṃ Bhagavato upaṭṭhako D II.139 (cp. D II.3)
“the venerable Upavāna (was) the attendant on the Blessed One for a long period”;

kīvaciraṃ pabbajito si, āvuso Samiddhi? na ciraṃ, āvuso, tīṇi vassānîti M II.207; D I.152
“How long have you been ordained, venerable Samiddhi? Not long, friend, three years”.

§45. Acc. of Place Where.

The acc. may also denote the place where something happens. The few examples found, however, seem to border on the sphere of the adverbial acc. (of place). e.g.,

upasaṅkamitvā vehāsaṃ aṭṭhaṃsu S I.23
“having approached they stood in the air (or adv. airily)”;

ekamantaṃ nisīdi D II.91,102
“sat on a side (cp. adv. aside)”;

ekamantaṃ aṭṭhāsi D II.112
“stood on a side or [55] aside”.

These two examples are clearly adverbial, the stem anta- being hardly used in the Nikāyas except in adverbs of a historical nature. It is not common either in Vedic or in Classical Sanskrit and Buddhaghosa commenting on the phrase ekamantaṃ nisīdi D II.55 says that it is either an adverbial use (bhāva-napuṃsaka-niddeso) as visamaṃ in ‘visamaṃ candima-suriyā paharanti’, or, is an acc. used in the sense of the loc. (bhummatthe – Sum. II. p.483). Similarly on ekamantaṃ aṭṭhāsi Sn p.13 he has the remark: ekamantanti bhāvanapuṃsaka-niddeso, ekokāsaṃ ekapassanti vuttaṃ hoti, bhummatthe vā upayoga-vacanaṃ (Pj. II. p.140). The loc. itself occurs in the Nikāyas though not very frequently. e.g.,

ekamante tiṭṭheyya D I.104
“would stand on a side”.

The fact that ekamantaṃ is an adverbial use, however, does not preclude the possibility of it being at the same time an acc. of place where. It may even stand as an acc. of direction. e.g.,

ekamantaṃ apanetvā D I.221
“having led ... aside (lit. to a side)”,

or any other logical type. Adverbial uses always arise from such original functions as these. It may be noted however that the fundamental notion underlying these various applications is the same, viz., the spatio-temporal conception.

a. There is also a class of accusatives denoting place where which had already assumed the role of adverbs in the earlier language. Local grammarians treat them as indeclinables or nipātas. Such are the accusatives in the constantly occurring phrase uddhaṃ adho tiriyaṃ D I.153,II.250 “above, below and across” (vide §2).

b. A similar acc. is found with certain verbs of motion like carati and anucaṅkamati where it denotes the space over which one wanders etc. Here it borders on the acc. of extent in space (§44.a). e.g.,

gāmaṃ vā nigamaṃ vā piṇḍāya caranto D III.255
“going out for alms through (over) village and hamlet”;

Vesāliṃ piṇḍāya caritvā
“having wandered through Vesāli for alms”;

ekamantaṃ anucaṅkamamāno M II.158
“walking up and down on a side”.

The loc. is here the concurrent idiom (§164.b). [56]

§46. Acc. of Time When.

Just as the acc. of place where is connected to the idea of extension in space, so is the acc. of time when related to the notion of extension in time. Therefore the renderings “in” and “through” are both permissible in such instances as the following: e.g.,

ye pi te bhikkhave ahesuṃ atītaṃ addhānaṃ arahanto sammāsambuddhā D II.144
“whichever holy ones, perfectly enlightened Buddhas, there were during (or in) the long past”;

ye hi keci atītaṃ addhānaṃ samaṇā ... paccanubhosuṃ D II.213
“whichever ... recluses experienced.. during the long past”;

yaṃ pi bhikkhave Tathāgato purimaṃ jātiṃ purimaṃ bhavaṃ purimaṃ niketaṃ pubbe manussabhūto samāno ... D III.145
“that the Tathāgata in (or during) a previous birth, existence or life, being then a human being”;

yaṃ nūnâhaṃ bhante aññaṃ jātiṃ kodhanā ahosiṃ A II.204
“were I, Sir, in another birth (or during another life) irritable”;

purimāni bhante divasāni purimatarāni ... devā ... sannisinnā honti D II.207
“some few days ago ... the gods ... were assembled” (lit. on or during some previous days ... etc.);

yañca Ānanda rattiṃ Tathāgato anuttaraṃ sammāsambodhiṃ anusambujjhati D II.134
“in or during which night the Tathāgata realized unique and perfect enlightenment”.

a. In some other examples however the notion of extension in time is hardly implicit for they merely signify the time when. The most prominent employ of such an acc. is found in the stock phrase ekaṃ samayaṃ occurring at the beginning of all suttas. Buddhaghosa commenting on this says it is only an acc. used to suit the sense determined by the context (vide Papañcasūdanī I. p.9-10) and parallel to the more usual loc. (bhummaṃ eva attho, Sum. I. p.33). In the latter Comy. he discusses the use as follows: ‘Kasmā pana ettha yathā Abhidhamme yasmiṃ samaye kāmâvacaranti ca, ito aññesu Sutta-padesu yasmiṃ samaye bhikkhave bhikkhu vivicceva kāmehîti ca bhumma-vacanena niddeso kato, Vinaye ca: Tena samayena Buddho Bhagavā ti, karaṇavacanena, tathā akatvā ekaṃ samayanti upayoga-vacanena niddeso kato ti? Tattha tathā idha ca aññathā attha-sambhavato’. Then he goes on to say that in the first instance the loc. is used [57] to denote the state (bhāva); in the second, the inst. is employed because the time was itself the occasion (tena samayena hetubhūtena karaṇabhūtena) for laying down the precept; in the third, it is a peculiar use of the acc. of extension of time (yaṃ hi samayaṃ Bhagavā imaṃ aññaṃ vā suttantaṃ desesi accantaṃ eva taṃ samayaṃ karuṇā-vihārena vihāsi tasmā tadattha-jotanatthaṃ idha upayogavacananiddeso kato ti). He is therefore quite aware of the logical connection between the acc. of time when and that of extension. But the following examples, as pointed out above, have purely the notion of time when;

atha kho Bhagavā pubbaṇhasamayaṃ nivāsetvā D I.178; II.102,122
“thereupon the Blessed One having robed himself in the forenoon”;

pubbaṇhasamayaṃ abhiruhitvā D II.174
“having ascended ... in the forenoon”.

It is significant, however, that even in such applications Buddhaghosa sees the implication of extension. On pubbaṇhasamayaṃ Sn p.13, he makes the following observation: ‘pubbaṇhasamaye ti attho ... pubbaṇhe ekaṃ khaṇanti vuttaṃ hoti, evaṃ accanta-saṃyoge upayogavacanaṃ labbhati’ (Pj. II.139). Similarly:

rattiyā paccūsasamayaṃ paccuṭṭhāya D II.203
“having risen up early in the morning”;

imaṃ bhikkhave rattiṃ aññatarā devatā (maṃ) ... upasaṅkami A IV.28
“this night, monks, a certain deity approached me ...”;

yañ ca rattiṃ parinibbāyati D II.134
“whichever night.. (the Tathāgata) ... passes away”.

The day on which something happens is also denoted by the acc. e.g.,

atha kho sattamaṃ divasaṃ Kosinārakānaṃ Mallakānaṃ etad ahosi D II.159
“on the seventh day it occurred to the Mallas of Kusinārā”;

Acela Korakkhattiyo sattamaṃ divasaṃ ... kālaṃ karissati D III.7
“A.K. will die on the seventh day”

(Comy. glosses it by ‘sattame divase’).

b. The acc. of certain compounds in which the first member is a past (passive or middle) participle and the second is pubbaṃ (acc.) occurs frequently in the Nikāyas denoting the time when. This is mostly found in the instance bhūtapubbaṃ which appears as an introductory remark in narration, e.g.,

Bhūtapubbaṃ bhikkhave rājā ahosi Pacetano nāma A I.111
“(It happened) in the past, there was a king by name Pacetana”;

Bhūtapubbaṃ Pūraṇo Kassapo anekasatāya parisāya dhammaṃ desesi M II.3
“It [58] happened in the past that Pūraṇa Kassapa was preaching his doctrine to a crowd of many hundreds” (cp. D II.130,169);

sannisinna-pubbaṃ sallapita-pubbaṃ D II.109
“met in the past, conversed in the past”.

In Skr. pūrvaṃ occurs in narration in the same sense instead of atīte. e.g., Vārānasyāṃ abhūt pūrvaṃ Brahmadattābhido nṛpaḥ, Kathā-saritsāgara I.7 (cp. I.3). The compound bhūtapūrvaṃ too is met with in the Mahābhārata and Kāvya literature (vide Monier Williams Dict. s.v.). In such instances the acc. -pubbaṃ is adverbially referring to what has been before.

c. There are also a few temporal accusatives of an archaic character used in the Nikāyas as adverbs. e.g.,

āyatiṃ pi evarūpena paṇītena piṇḍapātena pariviseyyâti M I.369
“would treat again (lit. in the future) too with such excellent alms as this”;

sāyapātaṃ upaṭṭhānaṃ āgacchanti D II.188
“they come for the purpose of attending (on him) morning and evening”.

§47. The Adverbial Accusative.

The acc. singular of substantives and neuter adjectives is copiously employed in the Nikāyas in adverbial sense. As in the other languages adverbs formed from adjectives predominate and it is only rarely that substantives are so used, most of them being treated by local grammarians as particles (nipāta), especially those stereotyped adverbial accusatives inherited from Vedic like nāma etc. (cp. §2). In their particular functions these adverbs admit of the usual classification into local, temporal, modal, causal and so on. On the whole Brugmann’s division as found in his Greek Grammar (§441) tallies with the distinctions that appear in the Nikāya prose. In the preceding paragraphs we have already referred to a few such uses. The following is a more exhaustive treatment:

§48. The Adv. Acc. of Time and Space.

The acc. as adverb of time and space is mostly found of adjectives and participles that imply these conceptions. The following are temporal uses: sīghaṃ A I.45 “quickly” (mostly [59] found in compounds where the ending -aṃ is dropped for euphony); sanikaṃ D II.333; M I.120; S I.82,203; khippaṃ A II.118;III.164 (cp. Sn 413,682,998) “soon”; tuvaṭaṃ (cp. Skr. tvaritaṃ) A V.342 “hurriedly”. The adverbial acc. of nominal stems is only found in one or two archaic instances. e.g.,

yena samantā yojanaṃ passati divā c’eva rattiñca D II.20
“sees a yojana all around by day as well as by night”,

where the use of the inst. divā side by side with acc. rattiṃ proves beyond doubt that the latter is an adverbial usage (cp. Vedic naktaṃ “by night”; vide §2). The local use of this kind of adverb is not so frequent as the above in the Nikāyas since it is more liable to fall in with the adverb of manner. e.g.,

rassaṃ assasāmîti pajānāti ... dīghaṃ assasāmîti pajānāti M I.56
lit. he knows that he breathes short ... he knows that he breathes long”.

A similar acc. is found of nouns denoting the cardinal points. e.g.,

uttariṃ Vesāliyaṃ D III.10
“to the north (lit. northly) of Vesāli”.

In the commonly found idiom

yena.. tad avasari D II.126,156
“whither ... thither went”

(§81.b.ii), it is almost certain that the acc. tad is a local adverb though the verb ava-sari is capable of taking the acc. of the place entered (§40). Another local use seems to be contained in the phrase (taṃ) mūlaṃ chindeyya A II.199, which is admitted as a textual variation of mūle chindeyya. Here, if the former reading is authentic, which is not unlikely, mūlaṃ would be an adv. acc. meaning “if one were to cut (it) by the root”.

§49. Adv. Acc. of Contents (internal object).

a. Some cognate accusatives are adverbially used parallel to the construction with katvā (§35.b.) as in the idiom visuṃ karoti “makes apart, scatters” etc.; almost all are adverbs of manner, e.g.,

so taṃ thūṇaṃ khaṇḍâkhaṇḍikaṃ chindeyya A II.199
“he would cut the post into pieces (lit. piecemeal)”.

Here the verb chindeyya contains the notion of “bits” or “pieces” which is the sense of the noun khaṇḍa. Local grammarians would paraphrase it by ‘so taṃ thūṇaṃ khaṇḍâkhaṇḍikaṃ katvā chindeyya’.

b. Closely related to the above is a class of adjectival compounds used adverbially, its second member being the etymological [60] object of the verb following. The origin of this construction can be traced back to the Vedic dialect, where it is found with the privative particle a-. e.g.,

‘... lokn anapajayyáṃ abhy ájayan’
“unconquerably (i.e. irrevocably) conquered these worlds”

(vide VGS §196.5.b.). In Pāli it occurs mostly with the particle su- “well”. e.g.,

samaṃ suvibhattaṃ vibhaji D II.166,235
“he divided (it) equally and well (lit. well-dividedly)”,

where the parallel use of samaṃ, a pure adverb, shows that the latter too is employed in the same manner;

suvinītaṃ vineti D III.189
“he trains (him) well (lit. well-trainedly)”.

There are a few instances where the first member is an adjective or a participle, e.g.,

chinnapapātaṃ papatanti D II.140
“falls prostrate on the ground (lit. falls a-fall- as-if-it-were-cut)”;

gāḷha-bandhanaṃ baddho D I.245
“bound tightly (lit. by a strong binding)”.

A similar cognate use is found with yathā-. e.g.,

yathābhuttañca bhuñjathâti D II.173
“may ye eat as ye have eaten (before)”.

In all these examples the inst. can be substituted for the acc. without any alteration of the meaning. In later Pāli the inst. is actually found in such places showing clearly how the original adverbial sense of the acc. is being gradually lost, the latter being superseded by the former (inst.) which becomes the commoner adverbial case after the model of Classical Sanskrit, though in Vedic, as mentioned above, the adverbial employ of the acc. is not unusual.

§50. Neuter Pronoun (acc. sg.) as Adverb.

a. Within this class of adverbial accusatives Brugmann includes the acc. sg. of neuter pronouns (interrogative, relative and demonstrative) used adverbially (Greek Grammar, §441.2.b.). This is widely used in Pāli and Skr.. In the case of those derived from interrogatives, by virtue of their origin, they come to mean “how, why?” etc. and are therefore adverbs of manner or of reason. e.g.,

kin ti me sāvakā dhammadāyādā bhaveyyum no āmisadāyādā ti M I.12
“How now, do my disciples become the heirs of my Law and not of my possessions”;

kin ti te sutaṃ D I.104
“why, have you not heard?”;

kin ti te D II.174
“how then?”;

kin nu kho āvuso D II.8,131
“why, friends?”

(but why, why in the world etc. vide [61] P.T.S. Dict. s.v.). This seems to be a development from the acc. of relation, implying originally “as to what.” (cp. Latin ‘Quid ille me castigat’? “Why (lit. as to what) does he lecture me?”).

b. In similar contexts we find the acc. taṃ and yaṃ used to denote “therefore” and “wherefore” respectively, yaṃ is more frequent in this connection. e.g.,

yaṃ sukho bhavaṃ taṃ sukhā mayaṃ, yaṃ dukkho bhavaṃ taṃ dukkhā mayaṃ D II.233
“since (lit. wherefore) you are happy therefore we are happy, since you are sorrowful therefore we are sorrowful”

[cp. Latin ‘Quod non venisti, timebam’ “because (lit. as to the fact that) you did not come, I was anxious”].

c. Related to this is the employ of the relative neuter pronoun yaṃ as connective between the principal sentence and subordinate clauses. This corresponds to yad in Skr. (cp. Monier Williams Dict, s.v.), which can stand either for the nom. or for the acc. according to the context, e.g.,

yampi Bhoto Soṇadaṇḍassa yaso hāyissati samaṇassa Gotamassa yaso abhivaḍḍhissati iminā p’ aṅgena na arahati bhavaṃ Soṇadaṇḍo ... Gotamaṃ dassanāya upasaṅkamituṃ D I.113
“since the fame of venerable Soṇadaṇḍa will diminish and that of the recluse Gotama will increase, because of this very fact it is not fitting that the venerable Soṇadaṇḍa should go to see him”;

kim-atthi-yaṃ āvuso samaṇe Gotame brahmacariyaṃ vussatîti S IV.51
“what (profit) is there, friend, now that the Holy Life is lived under the recluse Gotama?”;

aṭṭhānaṃ kho etaṃ mārisa anavakāso yaṃ ekissā loka-dhātuyā dve arahanto sammāsambuddhā apubbaṃ acarimaṃ uppajjeyyuṃ D II.225
“It is not possible, friend, there is no chance, that in the same world-system two Holy Ones, perfectly enlightened Buddhas can arise at the same time”;

Buddhaghosa calls yaṃ in this context ‘the nom. used in the sense of the inst.’ (kāraṇatthe paccattaṃ) and glosses it by ‘yena kāraṇena’ (Sum. II. p.659). It is however more likely that the form yaṃ here represents an acc. after the type of the adverbial accusatives discussed above. The following examples support the latter view:

Tasmā-t-iha Ānanda tuyh’ ev’etaṃ dukkaṭaṃ ... yaṃ tvaṃ ... na Tathāgataṃ yāci D II.115
“Therefore, Ānanda, this mistake has been committed [62] by you ... that you ... did not implore the Tathāgata”;

nâhaṃ bhante etaṃ rodāmi yaṃ maṃ Bhagavā evaṃ āha M I.389
“I do not grieve, Sir, over this, that the Blessed One has spoken to me thus”;

na hi sādhu yaṃ uttama-puggalassa sarīra-bhaṅge siyā sampahāro D II.166
“it is not good that there should be a quarrel over the distribution of the relics (lit. the body) of the noblest man”;

It may be remarked here that sometimes instead of the connective (nt. acc. sg.) the actual relative pronoun agreeing with the subject of the clause in number and gender is employed in similar constructions. e.g.,

na kho me taṃ paṭirūpaṃ yo’haṃ ākiṇṇo viharāmi D II.30
“it is not fitting for me that I should live crowded”;

ovadatu maṃ bhante Bhagavā anusāsatu maṃ bhante Bhagavā yaṃ mam’ assa dīgharattaṃ hitāya sukhāyâti S III.1
“may the Blessed One advise me and admonish me that (sci. it = taṃ) may conduce to my benefit and weal for a long time”.

The origin of this idiom may have been due to such contact as afforded by instances of the type:

yaṃ rūpaṃ aniccaṃ dukkhaṃ vipariṇāmadhammaṃ ayaṃ rūpassa ādīnavo S III.62
“that the form is impermanent, subject to grief and change, this is the danger of form”,

where the proper connectiveyaṃ’ accidentally agrees in gender and number with the subject of the subordinate clause, viz., rūpaṃ.

d. The acc. yaṃ is also used as a temporal adverb meaning “when”, usually introducing adverbial clauses or even in the temporal function of “since”, often to be rendered by English ‘that’, e.g.,

ito so bhikkhave eka-navuto kappo yaṃ Vipassi Bhagavā ... loke udapādi D II.2
“it was the ninety first aeon hence that the Blessed One Vipassi ... was born in the world”.

Commenting on this Buddhaghosa says that yaṃ is found in four distinct uses: 1. paccatta-vacane “in the sense of the nom.”; e.g., yaṃ me bhante devānaṃ ... sammukhā sutaṃ ... ārocemi taṃ bhante Bhagavato ti, where it is only the relative pronoun agreeing with the subject sutaṃ. 2. upayoga-vacane “in the sense of the acc.”; e.g., appucchimha akkittayī no aññaṃ, yaṃ pucchāma tad iṅgha brūhîti, where too it is the relative pronoun acc. sg. agreeing with the object tad. 3. karaṇa-vacane “in the sense of the inst.”; e.g., aṭṭhānaṃ etaṃ bhikkhave anavakāso yaṃ ekissā loka-dhātuya ti, [63] where, as has been shown in the foregoing (c), it is the proper connective adverb. 4. bhumm’ atthe “in the sense of the loc.”, as in the context under discussion (idha pana; Sum. II. p.410). The commentator is therefore aware of the syntactical distinctions of the uses of yaṃ. Its purely temporal application is attested by many other examples, too. e.g.,

Atha kho Bharaṇḍu Kālāmo Kapilavatthumhā pakkāmi, yaṃ Kapilavatthumhā pakkāmi tadā pakkanto va ahosi na puna pacchāgañchi A I.278
“Then B.K. set out from Kapilavatthu and when he left it (then) it was never to return”,

where the parallel use of tadā is proof of the temporal sense of yaṃ;

So kho Ānanda samayo yaṃ mahāvātā vāyanti D II.107
“That is the time, Ānanda, when great gales blow”;

bhavissati bhikkhave so samayo, yaṃ imesaṃ manussānaṃ dasavassâyukā puttā bhavissanti D III.71
“there will be that time when these men will have sons who will (only) live up to ten years”,

where the Comy. has ‘yaṃ imesanti yasmiṃ samaye imesaṃ ... ’;

hoti kho āvuso samayo yaṃ kadāci karahaci dīghassa addhuno accayena ayaṃ loko saṃvaṭṭati D II.28
“there is a time, brethren, when at some time or other after the lapse of a long period this world will evolve ...”.

All these examples go to show that the acc. in ekaṃ samayaṃ occurring at the beginning of all suttas has its origin in the above-discussed temporal adverbial use.

§51. Adverbial Use of Acc. (nt. sg.) of Adjectives.

a. Adjectives denoting quantity or degree form corresponding adverbs. e.g.,

atibāḷhaṃ paridevesi D II.232
“he lamented very much”;

suññatāvihārenâhaṃ, Ānanda, etarahi bahulaṃ viharāmîti M II.104
“mostly I spend my time now, Ānanda, in the (lit. by the) dwelling on Emptiness”;

etad eva bahulaṃ bhikkhūnaṃ dhammiṃ kathaṃ karoti D II.123
“he speaks much to the monks on doctrinal matters”;

yo ciraṃ jīvati so vassa-sataṃ appaṃ vā bhiyyo D II.4
“he who lives long lives a hundred years or a little more”.

b. Several acc. adverbs are formed from comparatives of adjectives in -tara (cp. VGS §196.5.). e.g.,

yathā ahaṃ cirataraṃ [64] passeyyaṃ D II.178
“that I may see longer”.

The suffix may be added even to adverbial case-forms like divā. e.g.,

Kāḷīdāsī divātaraṃ yeva uṭṭhāsi M I.125
“the maid Kāḷī rose up later-on-in-the-day”.

c. Adverbs are also formed from adjectives denoting priority and posteriority of time. e.g.,

sammāsambuddhā apubbaṃ acarimaṃ uppajjeyyuṃ D II.225
“All-Enlightened Ones are born neither before nor after (the other, i.e. simultaneously)”;

tadanantaraṃ paṭisallīnā D II.265
“meditating the meanwhile”;

na ciraṃ Tathāgatassa parinibbānaṃ bhavissati D II.119
“ere long (lit. not long after) will take place the passing away of the Tathāgata”.

d. The acc. sg. nt. of the numeral adjective eka “one” is found in a peculiar adverbal usage in a passage of philosophic importance, viz., ye te samaṇabrāhmaṇā ... ekaṃ attānaṃ damenti ekaṃ attānaṃ samenti ekaṃ attānaṃ parinibbāpenti A II.68. To the student of syntax ekaṃ here clearly appears as an adverb meaning “once” (i.e. one time) and not by any means as an adjective qualifying attānaṃ. Accordingly the sense would be: “whichever recluses and brahmins ... once train themselves, again (lit. once) calm themselves and still again (lit. once) tranquillize themselves”; cp. eka- standing for ekaṃ (adv.) in

ekajaṃ vā dijaṃ vā pi Sn p.21
“once-born or twice-born”;


devo ca ekamekaṃ phusāyati Ud 5
“and it rains drop by drop (lit. one by one)”.

This adverbial use of ekaṃ is only a logical development from the adverbial acc. sg. of neuter ordinals such as paṭhamaṃ “firstly”, dutiyaṃ “secondly” and tatiyaṃ “thirdly” (cp. Latin primum etc.). e.g.,

dutiyam pi ... tatiyam pi kho .. D II.103
“secondly ... thirdly ...”.

e. The acc. sg. of certain nouns implying divisions of time like khaṇaṃ “a moment” is used adverbially to denote frequency or continuity with the prefix abhi-. e.g.,

so abhikkhaṇaṃ ... dukkhaṃ patisaṃvedi M I.308
“he experienced ... sorrow continually (i.e. moment by moment)”.

The numerical adverbs in -khattu (Skr. kṛtvaḥ) denoting the number of times are also formed on the model of adverbial accusatives. e.g.,

tikkhattuṃ padakkhiṇaṃ [65] katvā D II.163
“having walked round to the right (i.e. paid respects) three times”.

§52. Acc. as Adv. of Manner.

a. A large class of adverbial accusatives play the part of adverbs of manner. These seem to have developed from the original oppositional accusatives (cp. VG §642.b.). They are either the acc. sg. of substantives, neuter pronouns, adjectives or adjectival compounds. e.g., from substantives:

dukkhaṃ sessati S I.83
“will sleep miserably”

(cp. A I.137; M I.192);

sukhaṃ seti S I.212
“he dwells at ease (i.e. happily)”

(cp. A I.136). A similar acc. is found in the following examples where it has the sense of ‘after the manner of’ or ‘like’:

udumbarakhādikaṃ vâyaṃ kulaputto bhoge khādatīti A IV.283
“this clansman eats his wealth like a fig tree-glutton”;

ajaddhumārikaṃ vâyaṃ kulaputto marissatīti (ibid)
“this clansman will die like a starveling”

(cp. E.M. Hare Gradual Sayings IV, 189); from adjectives: samaṃ vibhaji D II.166,235 “divided equally”;

sādhukaṃ manasikarotha D I.63; II.2,255
“reflect well”;

sādhukaṃ uggahetvā D II.119
“having learnt well”;

from adjectival compounds:

puṇḍarikāni ... samodakaṃ ṭhitāni D II.38
“white-lotuses ... standing at a level with the water”;

ime dhamme anavasesaṃ samādāya D I.165
“having taken up these doctrines completely (lit. remainderlessly)”;

pasannacittaṃ anussareyya M I.210,211
“would reflect over with a delighted heart (lit. delighted-heartedly)”.

A class of compounds with -upamaṃ as the second member is used adverbially implying accordance. e.g.,

kullûpamaṃ vo bhikkhave ājānantehi dhammā pi vo pahātabbā M I.135
“according to the parable of the raft even the dhammas should be discarded by you when you, brethren, attain to realization”.

b. To this category belongs a class of acc. adverbs formed from descriptive-determinatives where the first member is an adverb (inclusive of particles and prepositions) and the second is a noun or a past passive participle with substantival significance. Such compounds when employed thus as adverbs are treated by local grammarians as a special class called abbayībhāva (Skr. avyayībhāva [66] samāsa; cp. Kac. 321). e.g.,

ajjhattaṃ arūpasaññi D II.112
“internally conscious of the formless”;

ajjhattaṃ kāye kāyânupassī D II.216
“internally he sees the body in terms of the body”;

paccattaṃ yeva parinibbāyati D II.68
“attains parinibbāna individually of himself”;

(Comy. attanā);

paccattaṃ veditabbo D II.217 (cp D I.24)
“should be known personally or individually”;

nivātañca vata ayaṃ ca Migāramātupāsādo ... asaṅkampi S V.270
“even in the absence of a gale (without a storm) this mansion of Migāramātā trembled”;

cp. archaic acc.

anuraho maṃ ... codeyyuṃ M I.27
“they would accuse me ... secretly”

(vide §2);

yathābhūtaṃ pajānāti D I.162
“he knows according to reality”

(cp. M I.91);

yathābhirantaṃ viharitvā D II.94,126
“having lived as long as he desired (lit. according to fancy)”;

yathā-kāmaṃ S I.227
“according to inclinations”;

yathā-sattiṃ yathā-balaṃ D I.102
“according to energy, according to strength”;

yathāninnaṃ pavattamānaṃ A II.140
“flowing downward according to the bent”;

yāva-jīvaṃ anatikkamanīyā D III.133
“not to be transgressed as long as life (lasts)”;

to this adv. acc. is related the peculiar use of taṃ in the stock phrase yathā- taṃ. e.g.,

yathā-taṃ Mārena pariyuṭṭhita-citto D II.103
“because (lit. according to the fact that) his mind is obsessed by the Evil One”;

the Comy. has the characteristic remark ettha tanti nipāta mattaṃ’; cp. yathā-taṃ D II.264,269; III.8; M I.249; II.142 “because, accordingly as”;

tiro-kuḍḍaṃ tiro-pākāraṃ tiro-pabbataṃ asajjamāno gacchati M I.69
“he goes without getting stuck through wall and mountain”.

There are also to be found a few similar adverbial formations with the particle sa- which is the contracted form of either the sociative adv. saha or the prefix saṃ-. e.g.,

sanidānaṃ samaṇo Gotamo dhammaṃ deseti M II.9
“the recluse Gotama preaches the doctrine reasonably”;

sasīsaṃ pārupitvā M II.16
“covering himself head and all (lit. with the head)”;


sapadānaṃ piṇḍāya caramāno Sn p.21
“going for alms (from house to house) uninterruptedly”.

c. The acc. sg. of certain other compound formations is used as adverbs of reason. e.g.,

na ca maṃ dhammâdhikaraṇaṃ viheṭhesi M III.270
“he did not give me trouble by way of any [67] doctrinal point”;

tato-nidānaṃ hi so bhikkhave maraṇaṃ vā nigaccheyya A IV.130
“by reason of that, brethren, he would either meet death ...”;


tato-nidānaṃ labhetha pāmujjaṃ D I.72
“by reason of that he would obtain joy”.

The inst. or the abl. is the more usual in this function.

§53. Acc. with Adverbs.

Certain adverbial formations, mostly the inst. or abl. sg. of older (Vedic) nominal stems (cp. VGS §197), take an acc. of the noun-concept which they qualify or ‘govern’. Here the abl. and the gen. are concurrent idioms (§§130 & 150). e.g.,

uttarena Opasādaṃ M II.167
“to the north of Opasāda”;

antarena Campaṃ M I.340
“this side of Campā”;

antarā ca Sāvatthiṃ antarā ca Sāketaṃ M I.149
“between Sāvatthi and Sāketa”;

antarā ca Rājagahaṃ antarā ca Nālandaṃ D I.1
“between Rājagaha and Nālandā”,

where Buddhaghosa observes that the acc. is used because it is construed with the word antarā (‘antarā saddena pana yuttattā upayogavacanaṃ kataṃ’ Sum. I. p.35);

samantā Vesāliṃ D II.98
“around Vesāli”;

santike nibbānām S IV.74(V.)
“near nibbāna

Bhagavantaṃ sammukhā D II.155; Sn p.100
“before the Blessed One”;

pacchā bhattaṃ D II.102,122
“after the meal”;

kiṃ paccayā D II.31
“due to (lit. depending on) what?”;

ime ... dve paccayā D II.207
“owing to these two (reasons)”.

In the phrase kiṃ kāraṇā ti? D III.65 “due to what?”, it is clear, therefore, that the acc. (kiṃ) is due to treating kāraṇā as an adverb on the analogy of paccayā above, although a contamination of the two idioms kiṃ kāraṇaṃ and kasmā kāraṇā has been suggested (vide P.T.S. Dict. -kāraṇā).

§54. Acc. with Prepositions.

The acc. is ‘governed’ by more prepositions than any other case in Pāli just as in the older language (VGS §197.c.). Most of these are adverbial prepositions. e.g., ati, “beyond”, anu “after”, abhi “towards”, pati or paṭi “against, near”, and tiro “across” (cp. VG §176.1.1). [68]

a. The only prepositions used adnominally in the Nikāyas are pati or paṭi, always found as postposition in gāthā literature, tiro and anu which are difficult to distinguish from those employed as first member of adverbial compounds of the type discussed in the preceding paragraph. e.g.,

suriyass’ uggamanaṃ pati Th 1.517,628
“about, near, sunrise”;

utuveramaṇiṃ pati Sn 291
“(the time) about the cessation of menstruation”;

nadiṃ Nerañjaraṃ pati Sn 74
“near the river Nerañjarā”;

tiro dussaṃ tena manteti D I.103
“converses with him through a curtain”;

anvaddhamāsaṃ saṅghamajjhe osaranti M II.8
“come into the midst of the Order every half-month (lit. after every half-month)”.

There is a similar uncertainty with regard to the syntactical function of the adverbial prepositions adho, paro, anu etc. which too apparently occur only as the first member of avyayībhāva compounds. e.g.,

adho-mukhaṃ M I.132,134
“with head downwards, i,e., headlong”;

paro sahassaṃ D II.16; S I.192
“over a thousand”;

cp. anuraho M I.127.

b. Most of these prepositions, however, are only found in the Nikāyas adverbally i.e. as prefixes to verbs which in their normal form do not take an acc. of the object (cp. §37). e.g.,


te aññe deve atirocanti D II.208
“they outshine the other gods”;


padhānaṃ anuyuñja D II.144 lit.
“strive after exertion”;

cp. anuyogaṃ anuyutto D I.167; II.223 ;

bhāsitaṃ anumodi D II.279
“rejoiced in (lit. after) what was said”;


anujāto Tathāgataṃ Sn 557
“born after the Tathāgata”;


Bhagavato bhāsitaṃ abhinandi D II.71,277 cp. 54,69
“rejoice at (lit. after) the speech of the Blessed One”;

this verb nandati, however, could take the acc. even in its primitive form in the older language (cp. SS §42);

pa: savanaṃ pamajjati A IV.24
“neglects the hearing.”;


agāraṃ ajjhāvasati D I.63; II.16
“inhabits the house”;


(taṃ) upaṭṭhāti M III.25; S I.167; A III.94
“waits upon him”;


Vipassiṃ Bhagavantaṃ ... imā ... gāthā paṭibhaṃsu D II.36 (cp. M I.79)
“these verses dawned upon ... the Blessed One Vipassi”;

sādhu vata bhavantaṃ yeva bho Gotamaṃ paṭibhātu etassa bhāsitassa attho D I.124
“well, may the meaning of what is said dawn upon the venerable Gotama himself”.

The acc. in such examples, as has been mentioned earlier, is regarded by Indian [69] grammarians as being due to the karmapravacanīya or ‘adjuncts to sambandha or relation between kriyā and kārakavide Philosophy of Sanskrit Grammar p. 166. (vide §42). The gen. is the alternate construction in this special instance. In mamaṃ (v.1. mama)

yeva sāvakā anusāsaniṃ paccāsiṃsanti M II.10
“the disciples look to me for advice”

the pronoun mamaṃ may be either acc. or gen..

§55. Acc. with Indeclinables.

A class of words which are treated as indeclinables by Pāli grammarians but which were originally the gerunds of certain verbs, usually with the prefixes pa-, ud-, ā- and ni(s)-, by virtue of being verbal formations take an acc. of the thing ‘governed’ i.e. the original object. Such are:

paṭicca (prati + itya) “on account of”; e.g.,

etaṃ paṭicca M I.265
“on account of this”;

macchariyaṃ paṭicca ārakkho D II.58
“on account of avariciousness care”;

katamaṃ ... atthavasaṃ paṭicca D II.143
“on account of what ... matter?”;

ārabbha (ā + rabhya or rambh) “beginning with, taking into consideration, (hence) referring to, about”; e.g.,

bhikkhusaṅghaṃ ārabbha D II.100
“concerning the Order of monks”;

paricārake ārabbha D II.204
“about the followers”;

santiṃ ārabbha D II.157
“about peace”;

also at D I.180; A II.27,301; āgamma (< ā + gamya of gam) lit. “coming to, (hence) on account of”; e.g., D I.229; It 79; used as synonym for the above;

uddissa (ud + diśya of diś)
“pointing out, (hence) with reference to, concerning”;


bhavaṃ Gotamaṃ uddissa M I.16
“concerning venerable Gotama”,

where bhavaṃ should be corrected to bhavantaṃ; it is mostly found in the developed sense of “for the sake of”. e.g., (ahataṃ)

devaṃ yeva uddissa D II.80
“brought especially for the sake of the lord”;

kaṃ si tvaṃ āvuso uddissa pabbajito M I.170
“on account of whom, friend, have you left home?”;

where it comes to mean “following whom ... etc”. nissāya (< ni+ śrāya of śri) “leaning on (cp. bhittiṃ nissāya D III.208), (hence) depending on, owing to”; e.g.,

dhammaṃ yeva nissāya A I.109
“owing to the doctrine”;

rājānaṃ nissāya M II.185
“owing to the king”;

its sense is further developed in [70]

Āḷārakālāmaṃ nissāya nissāya atikkamiṃsu D II.130
“they went on keeping close to Āḷārakālāma”

i.e. “followed Ā. closely”. sandhāya (< saṃ + dhāya of dhā) “putting together, (hence) considering, on account of”; e.g.,

na kho āvuso Bhagavā sukhaṃ yeva vedanaṃ sandhāya sukhasmiṃ paññāpeti M I.400
“the Blessed One, friends, does not lay down (a thing) as (part of) happiness just on account of the pleasant feeling”.

§56. The Accusative Absolute.

a. The absolute use of cases is generally connected with, if not directly inherited from, their temporal application. We have in the earlier paragraphs observed many temporal uses of the acc., which were syntactically parallel to the loc. of time. Similarly a few instances are found in the Nikāyas of the accusatives of substantives agreeing with some participle, making up an obviously absolute construction. The participle mostly found in this connection is santa- “being”, which is involved in a similar acc. absolute in Greek (Greek on = Skr. san-), the origin of which Brugmann traces to the appositional use of the acc. “this being so ...” (Greek Gr. §582). The construction being thus quite historical in I.E., the possibility of the ending -aṃ, in such instances as are found in the Nikāyas, being a restoration of the normal loc. sg. -e as a result of the confusion of -e and -o in Prākṛt, seems to be precluded. Buddhaghosa too regards such accusatives as used in the sense of the loc. The following therefore may be reasonably regarded as acc. absolutes expressing the conditions during the permanence of which another event occurs. e.g.,

santaṃ yeva kho pana paraṃ lokaṃ atthi paro loko ti ’ssa diṭṭhi hoti M I.403
“there being a further world, he gets the notion that there is another world”;

santaṃ yeva kho pana kiriyaṃ atthi kiriyā ti paraṃ saññāpeti M I.406
“there being action, he convinces others that there is action”.

Here, just as in the case of the loc. absolute (§186), the adverb evaṃ can sometimes stand for the nominal member put in the loc.. e.g.,

evaṃ santaṃ kho te Poṭṭhapāda aññā vā saññā bhavissati añño attā D I.186,187
“it being so, Poṭṭhapāda, is perception one (thing) and the soul another”. Tattha evam santan to evam sante, bhummatthe hi etaṃ upayoga-vacanaṃ ...” Budg. cp Sum. II. p. 376.

On [71] the analogy of these constructions it seems not impossible that the acc. in the following is also absolutely employed though the verb pajānāti can take an external object when used in the sense of “perceive”. e.g.,

santaṃ vā ajjhattaṃ byāpādaṃ atthi me ajjhatto byāpādo ti pajānāti M I.60
“there being inward ill-will, he knows that there is inward ill-will”;

santaṃ vā ajjhattaṃ rāgadosamohaṃ atthi me ajjhatto rāgadosamoho ti pajānāti S IV.140
“there being inward passion, hatred and delusion, he knows that he has them”.

b. There are also a few other constructions with different participles. It is however not quite certain whether they are legitimate absolute uses. e.g.,

na bhikkhave mātā puttaṃ jīramānaṃ evaṃ labhati: ahaṃ jīrāmi mā me putto jītîti A I.179
“when the son is aging, monks, the mother does not find it possible (to say): let me decay but not my son!”.

That the acc. here is not necessarily meant to be the object of labhati is shown by the similar ‘intransitive’ use of the latter in:

so ’haṃ na labhāmi: mā me bhonto atthakaraṇe nisinnassa antarākathaṃ opātentu M II.122
“As such I do not find it possible (to say): do not interrupt me, friends, when I am presiding over the administration of justice”.

Similarly, in the following, though Buddhaghosa regards the construction as elliptical, supplying “ñatvā” after the participial clause, the acc. may still be considered as used absolutely:

Atha kho Suppavāsā Koliyadhītā Bhagavantaṃ bhuttāviṃ onītapattapāniṃ ekamantaṃ nisīdi A II.63; cp. D II.93; M I.393
“then the Koliya maiden Suppavāsā, when the Blessed One had eaten and taken out his hand from the bowl, sat on a side”.

The nom. ‘bhuktāvī ...’ in the BSk. version (cp. Mhvs. III. p.142) further supports the suggestion that it might be an acc. absolute.

§57. The Acc. of Purpose.

Since the notion of purpose is only the metaphorical application of the idea of direction, both implying the aim in a general sense, it is not surprising to find the acc. employed as a case denoting purpose, a use which is evidently derived from the acc. of direction (§37). This is to be particularly seen in the case of the dat. in - āya which is preserved in Pāli only in its functions [72] of direction and purpose (§4). Hence many examples of the acc. of direction with verbs of motion have the implication of purpose at the same time and consequently appear to be used parallel to the dat. or infinitive of purpose. e.g.,

upaṭṭhānaṃ gacchati D III.188
“he goes for the purpose of (lit. to) attending on.”;

cp. upaṭṭhātuṃ “to attend” A V.72;

Sakkassa ... anucariyaṃ upāgami M II.264
“he came for the purpose of waiting upon Sakka”,

lit. “he came to the waiting upon Sakka”;

divā-seyyaṃ upagato D I.112; M II.164
“gone for the day’s rest”;

Rājagahaṃ vassâvāsaṃ osaṭā M II.2 et. seq.
“entered Rājagaha for the passing of the rainy season”;

vāsaṃ upagacchi A V.29
“went for residence”

i.e. “took up abode”;

janapadacārikaṃ pakkamiṃsu D II.48
“they set forth on their mission of (lit. for the purpose of) tramping the country”;


methunaṃ dhammaṃ nâssu gacchanti brāhmaṇā Sn 292
“the brahmins did not go for sexual intercourse”,

where the Comy. has: ‘methunaṃ dhamman ti methunāya dhammāya; sampadānavacanappattiyā kir’ etaṃ upayogavacanaṃ’ (Pj. II. p.317). The acc. of purpose occurs only rarely with verbs other than those of motion. e.g.,

divā-vihāraṃ nisinno Ud 5
“seated down for the day’s rest”.

The postposition -atthaṃ (acc.) is the usual periphrasis, often appearing in compound, like its dat. -atthāya, which takes the place of this construction. e.g.,

parikkhāratthaṃ dānaṃ deti D III.258
“he gives alms for the purpose of (obtaining) requisites”.

§58. Double Accusative.

a. Direct + Appositional Object

Just as in Vedic (VGS §198) and Classical Sanskrit (SS §46.a) so in Pāli a second acc. appears in apposition to that of the direct object with certain verbs. It usually defines or qualifies the latter and is placed side by side with it in the sentence. e.g.,

maṃ hi bhante aññatitthiyā sāvakaṃ labhitvā M I.379
“those of other sects having received me, Sir, as a disciple”;

upāsakaṃ maṃ bhavaṃ Gotamo dhāretu D I.110; Sn p.25
“may the venerable Gotama accept me as a disciple”;

Sakyā kho pana Ambaṭṭha rājānaṃ [73] Okkākaṃ pitāmahaṃ dahanti (v.1. dissanti) D I.92
“the Sākyans, Ambaṭṭha, claim (lit. put up, place; cp. Comy. ‘ṭhapenti’, Sum. I.258) king Okkāka as their ancestor”;

here the abl. is the parallel idiom. e.g.,

mittato daheyya S III.113
“would consider as a friend”.

On the analogy of these the acc. in the stock phrase saraṇaṃ gacchati, to which reference has already been made in another connection (§37.a), can also be regarded as an acc. used in apposition to the other one. e.g.,

bhavantaṃ Gotamaṃ saraṇaṃ gacchāma M I.290
“we go to the venerable Gotama as refuge”;

Bhagavantaṃ yeva tāṇaṃ gavesi D I.95
“sought the Blessed One as protection”.

b. Direct + Predicative Object.

An acc. other than the direct object of the simple verb appears predicatively with verbs of speaking (in the sense of “calling”), thinking, knowing, perceiving, making and the like. This is closely related to the above appositional use. With verbs of speaking: This, however, is only found in the gāthā literature. e.g.,

taṃ ahaṃ brūmi brāhmaṇaṃ Sn 620
“him I call a brahmin”;

tatiyaṃ bhikkhunaṃ āhu maggajīviṃ Sn 88
“the third monk they call ‘one living in the way’”;

etad āhu vasuttamaṃ Sn 274
“this they call the best gem”.

In the Nikāya prose this construction has been superseded by that of the nom. with iti (§ 21). With verbs of thinking: e.g.,

taṃ kiṃ maññasi Pukkusa D II.131
“what do you think it (is), Pukkusa?”;

taṃ kiṃ maññatha bhonto devā Tāvatiṃsā D II.213
“what do the honourable Tāvatiṃsa gods think it (is)?”.

With verbs of knowing: e.g.,

yadā te Bhagavā aññāsi kallacitte ... D II.41
“when the Blessed One knew them to be (lit. as) of suitable disposition ...”;

bhāsamānañca maṃ na jānanti D II.109
“they do not know me to be speaking”;

petaṃ maṃ jānāhi Pv II.9
“know me as a departed (spirit)”;

cp. also

taṃ vā pi dhīrā muniṃ vedayanti Sn 212
“him the wise know as a sage”.

With verbs of making


cittaṃ attano ujukaṃ akaṃsu D II.254
“they made their minds straight’’;

ekaṃsaṃ uttarāsaṅgaṃ karitvā D II.172
“(lit.) having made the upper robe one-shoulder-covering”

i.e. “having put the upper robe over the left shoulder”; similarly, [74] ekaṃsaṃ cīvaraṃ katvā D II.163. When the expressions such as ekaṃsaṃ karoti are regarded as compound verbs (cp. §35.b), however, the construction loses its original significance.

c. Direct object + acc. of person indirectly affected.

i. Beside the acc. of the thing said, which is the direct object, another is found denoting the person to whom the statement is made with verbs of speaking. e.g.,

āyasmantaṃ Sāriputtaṃ etad avoca M I.31
“he told the venerable S. this”;

Kāḷiṃ dāsiṃ etad avoca M I.125
“she told this (to) the maid-servant Kāḷi”;

te nighaṇṭe etad avoca M I.92
“he told the naked ascetics this”;

as seen from these examples the idiom is only frequent when the direct object is etad (cp. also D II.102,165). With other verbs the dat. is employed.

ii. A similar double acc. construction occurs with the verb pucchati, where the person questioned is also put in the acc. e.g.,

Bhagavantaṃ imaṃ paṭhamaṃ pañhaṃ pucchi D II.76
“he asked the Blessed One this first question”;

samaṇaṃ Ānandaṃ ... phāsuvihāraṃ puccha D I.204
“inquire (from) the recluse Ānanda his ease and comfort”;

yaṃ kiñci maṃ Subhaddo pucchissati D II.150
“whatever Subhadda shall ask me”.

This can be compared with the similar use of the double acc. with yācati often found in verse e.g.,

pabbajjaṃ ayāci Buddhaṃ Th 1.869
“asked ordination (from) the Enlightened One”;

pabbajjaṃ ahaṃ ayāciṃ sabbasattāna uttamaṃ Th l.624
“I implored the highest of all men for ordination”.

iii. A double acc. construction occurs also with verbs of doing, the second standing in place of the dat. of the person indirectly affected (§101). e.g.,

kinti naṃ karosîti A II.112
“what do you do (to) him?”;

taṃ enaṃ bhikkhave nirayapālā pañcavidhabandhanaṃ nāma kāraṇaṃ karonti A I.141
“to him, brethren, the warders of Hell do the punishment called the ‘five-fold-binding’”;

Sakuludāyiṃ paribbājakaṃ antarāyaṃ akāsi M I.39
“they did harm to (opposed) the wandering ascetic Sakuludāyi”.


Yet with none of the said verbs is the double acc. of necessity, as Speyer points out (SS §47). In Pāli especially the variety of idiom and expression permits of other constructions which are quite as usual and sometimes even preferable. Thus verbs of speaking and teaching may take a dat. of the person spoken to; verbs of asking may take an abl. of the person questioned (vide §93.b).

d. Direct object + (original) agent of causative verb.

i. A second acc. occurs with the causative verb, if in its simple state it was ‘transitive’ to express the agent, i.e. the subject of the verb in the primitive or non-causative state, beside the acc. of the original affected object. This agent or the original subject can also be denoted by the inst. (vide §59). e.g.,

ekamekaṃ Bhagavantaṃ vandāpessāmi D II.148
“I shall make each in turn worship the Blessed One”.

ii. When the verb is ‘intransitive’ and denotes motion, the acc. of the aim (which, however, is never a person, in which case the loc. is used, cp. VGS §198.3f.n.2.) is preserved and the agent as before is put in the acc. e.g.,

uppannuppanne pāpake akusale dhamme ... anabhāvaṃ gameti M I.11 lit.
“the constantly arising, evil, unmeritorious thoughts ... he causes to go to non-existence”.

iii. The difference between these actual causative verbs and the faded causatives such as vāceti “teach” in the following example is an essential. For verbs such as those of teaching take the double acc. even in the older languages (cp. SS §47). e.g.,

tīṇi māṇavakasatāni mante vāceti M II.166
“he teaches the mantras to 300 young men”;

satta ca brāhmaṇamahāsāle satta ca nahātakasatāni mante vācesi D II.236
“he taught the mantras to seven noble brahmins and 700 initiated students”.

The person taught may also be expressed by the dat. (§93.e). We have probably a similar faded causative in the following where the acc. kāyaṃ may be alternately put in the loc.:

te imehi kāyaṃ balaṃ gāhenti M I.238
“with these they infuse strength into their bodies”.


§59. Accusative with Causative Verbs.

As has been shown in the preceding paragraph, the subject of the simple verb which would be expressed by the nom. in the original sentence is put in the acc. case denoting the agent with the causative. This is invariably the case if the simple verb was ‘intransitive’. e.g.,

Atha kho Sakko ca devānamindo Vessavano ca mahārājā āyasmantaṃ Mahāmoggallānaṃ Vejayante pāsāde anucaṅkamāpenti anuvicarāpenti M I.253
“Then Sakka, the lord of the gods, and the Great Regent Vessavana caused the venerable Mahāmoggallāna to walk and move to and fro in the mansion Vejayanta”.

But if the simple verb be a so-called ‘transitive’ there is diversity of construction. In that case the inst. is also permissible to denote the agent (vide §88.d). This alternate idiom, though frequent in the dialect of the Brāhmaṇas (VGS §198.3) and in Classical Sanskrit (SS §49), is, however, not so popular in the Nikāyas. Still it is enjoined by Pāli grammarians. Cp. ‘gatibodhāhārasaddatthâkammakabhajjâdīnaṃ payojje’ Mog II.4 et. seq. e.g.,

Kosinārake Malle Bhagavantaṃ vandāpesi D II.148
“he made the Mallas of Kusinārā worship (pay their respects to) the Blessed One’’.

The original object of the simple verb is naturally preserved in the acc. e.g., (taṃ) aññā khīraṃ pāyenti D II.19 “others make (him) drink milk” i.e. “others suckle him”.

§60. Acc. with Passive Verbs.

In the case of verbs that admit of two accusatives, like pucchati etc., if in the passive construction the person is put in the nom. case as the grammatical subject of the passive verb, the logical subject being denoted by the inst. as the agent of the action, the thing questioned is preserved in the acc. case (cp. §43). e.g.,

samaṇa-brāhmaṇā ... pañhaṃ puṭṭhā D I.24
“the recluses and brahmins ... being asked a question”.

In a similar way the passive past participle used actively (or the reflexive participle) may take an acc. of the thing or person that was the object in the primitive state. e.g.,

tam enaṃ Paṇḍuputto ... paccupaṭṭhito hoti M I.31
“Paṇḍuputta ... was attending on him (or was present by his side)”.

The passive potential [77] participle (-tabba) used in the acc. nt. sg. as predicate of the clause in an indirect statement takes an acc. of its original object. e.g.,

Tathāgate arahante sammāsambuddhe āsādetabbaṃ maññati D II.24
“thinks that the Tathāgatas, the holy and perfectly enlightened ones should be appeased”.

Buddhaghosa, surprised by this peculiar employment of the acc. with the passive potential participle and ignoring the possibility of a legitimate construction with the impersonal “-tabbaṃ”, says the ending -e stands for the plural and -tabbaṃ for -tabbe. The suggestion is at any rate syntactically plausible since the verb maññati is capable of taking the double acc. (§58.b.). But the reading need not be altered as it appears also in other places. e.g., amhe ovaditabbaṃ ... maññati M I.460.