Do Counterexamples on the Kāraka Rules A. 1.4.33-36 in the Kāśikāv r̥ tti Serve Any Purpose?

Ajotikar et al. (2016) claim that most of the counterexamples provided in the Kāśikāvr ̥ tti conform to the distinctive feature of a counterexample, namely, having all the conditions stated in the rule except one ( ekāṅgavikalatā ). Ajotikar (2021) discusses how a variant reading for a counterexample helps understand the relation between two operational rules. This article adds one more aspect to the importance of counter-examples. However, there are some cases where the purpose of the counterexample of a complex semantic condition is not clear. In this article, I study counterexamples provided on the sūtras, A. 1.4.33-36, in the kāraka section, on which Patañjali did not comment. These sūtras are chosen for discussion because the counterexamples available on these sūtras are first provided in the Kāśikāvr ̥ tti . When it comes to the issue of complex semantic conditions ( priyamāṇa , jñīpsyamāna , īpsita or uttamarṇa ) stated in A. 1.4.33-36, it is difficult to justify the usefulness of the available counterexamples. After carefully examining Bhatr̥hari’s views along with Helārāja’s explanation, it is evident that these counterexamples must have been included in order to fulfil the criteria of a vr ̥ tti . A vr ̥ tti typically includes an example, a counterexample and a supplementary word, which are necessary to complete the meaning of the rule. However, these counterexamples fail to justify the significance of the semantic conditions stated in the rule for which they are provided. Hence they do not serve any purpose.


Introduction
Counterexamples are an important part of commentaries on rules in the Aṣṭādhyāyī.Ajotikar et. al. (2016) conclude that the essential feature of a counterexample is that it have all the conditions stated in the rule except one (ekāṅgavikalatā).Ajotikar (2021) adds one more aspect to the importance of counterexamples, namely that they help determine the scope of the operation provided by the sūtra.The Author demonstrates that a variant reading for a counterexample provided in the Kāśikāvr̥ tti on A.7.2.8 neḍvaśi kr̥ ti helps understand the relation between the two operational rules A.7.2.8 and A.7.2.35 ārdhadhātukasyeḍvalādeḥ.However, it is difficult to comprehend the purpose of a counterexample of a complex semantic condition in a sūtra that introduces a technical term.Several of this type of counterexample occur in the kāraka section, for example, counterexamples on the sūtras A. 1.4.33-36.These sūtras are chosen for discussion because Patañjali did not comment on them, and the counterexamples are first provided in the Kāśikāvr̥ tti.This article is divided into five sections.After this introduction, the second section of the article deals with the technical difficulty regarding the counterexample given on A. 1.4.33 along with a survey of counterexamples provided in almost all the commentaries available in print form on the Aṣṭādhyāyī.Then, in section 3, are discussed Bhartr̥ hari's views on A. 1.4.33-36,where it is examined whether the counterexamples provided in the Kāśikāvr̥ tti comply with Bhartr̥ hari's discussion.In section 4, I discuss counterexamples provided by pre-Kāśikāvr̥ tti non-pāṇinian grammars, namely, the Kātantra and Cāndra, where they have a sūtra equivalent to one among A. 1.4.33-36. 2 Technical Difficulty in the Counterexample on A. 1.4.33 Since Patañjali did not comment on rules A.1.4.33-41, the Kāśikāvr̥ tti is the source of all discussion on these rules in the Pāṇinian tradition.A glance at the subsequent transmission of the counterexamples on these rules will demonstrate this.Table 1 illustrates beyond any doubt that all the subsequent commentaries relied on the Kāśikāvr̥ tti for the choice of counterexamples.Hence analysis of counterexamples in the Kāśikāvr̥ tti is sufficient to settle the issue regarding their significance [tab.1].Kāśikāvr̥ tti (Sharma et al. 1969) Bhāṣāvr̥ tti (Chakravarti 1918) Rūpāvatāra (Rangacharya 1916) Prakriyākaumudī (Trivedi 1925) No counterexample īpsita iti kim.puṣpebhyo spr̥ hayati vane.
Table 1 shows that the counterexamples provided by the Kāśikāvr̥ tti on A. 1.4.33-36 are repeated by other commentators without any change wherever they are included.All the counterexamples in the Kāśikāvr̥ tti have a definite pattern.The sentence that constitutes one of the examples for the sūtra is used to form a counterexample with the addition of a locative singular form at the end.For example, the Kāśikāvr̥ tti provides a pair of examples on A. 1.4.33:devadattāya rocate modakaḥ "Devadatta likes sweat-meats" 1 and yajñadattāya svadate 'pūpaḥ "Yajñadatta likes cake".A counterexample is provided by adding just one word, pathi, in devadattāya rocate modakaḥ pathi "Devadatta likes sweat-meats on the path", in order to explain the significance of the condition pri ̄yamāṇaḥ 'being pleased'.The word pathi is a locative singular of the word pathin 'path'.As is evident from Table 1, the same counterexample is repeated by the post-Kāśikāvr̥ tti commentators even if the commentator provides a different example for the sūtra.For example, the Siddhāntakaumudi ̄ provides the example haraye rocate bhaktiḥ "Hari likes devotion" (Chaturveda, Vidyabhaskara 1961, 644) whereas it does not change the counterexample sentence.This pattern of counterexample with a locative singular form is continued in the subsequent three sūtras, A. 1.4.34-36 2 (by addition of a locative singular word at the end such as grāme, vane).
We observe one change in the Kāśikāvr̥ tti on A. 1.4.36,namely that the word vane is placed before and not after the verb (puṣpebhyo vane Bhasha e-ISSN 2785-5953 2, 2, 2023, 1-14 4 spr̥ hayati).This sequence is copied in the Siddhāntakaumudi ̄; however, in the Rūpāvatāra, the regular pattern with the locative at the end is restored (puṣpebhyo spr̥ hayati vane).
All commentators follow the same pattern, with the exception of one, where we find variant readings.A. 1.4.34.There is a variant reading found in three different editions of the Kāśikāvr̥ tti: devadattaḥ ślāghate "Devadatta praises" (Mishra 1985, 552;Tripathi, Malaviya 1986, 148;Vidyavaridhi 1997, 74).In the Osmania edition of the Kāśikāvr̥ tti (Sharma et al. 1969-70, 82 fn. 10), one additional variant reading is mentioned in the critical apparatus, i.e. devadattaṁ ślāghate "he/she praises Devadatta" along with devadattaḥ ślāghate.These two variant readings deviate from the pattern of having a locative singular word added at the end of the example sentence.These two variant readings are known to Haradatta (Mishra 1985, 552).In the Padamañjarī, he quotes devadattaḥ ślāghate as the main reading and notes the variant devadattaṁ ślāghate.Interestingly, Haradatta does not show any awareness of the commonly available reading devadattāya ślāghate pathi.He discusses the counterexample devadattaḥ ślāghate as follows: devadattaḥ ślāghate iti.jñipsyamānavacanāt karmasaṁjñaiva bādhyate, na kartr̥ saṁjñetyarthaḥ "The purpose of mentioning the counterexample, devadattaḥ ślāghate, is to show that the fact that the condition jñipsyamāna is mentioned in sūtra, the rule blocks only the term karman, not the term kartr̥ ̥ ".He further states: kvacit tu devadattaṁ ślāghate iti pāṭhaḥ "in some sources, there is a variant reading devadataṁ ślāghate".These two counterexamples available to Haradatta are not recorded in any other post-Kāśikāvr̥ tti commentary.

2.1
Is the Given Counterexample Correct?
Let us discuss the correctness of one of the counterexamples with a locative singular.For example, consider the counterexample on A. 1.4.33 rucyarthānāṁ pri ̄yamāṇaḥ: devadattāya rocate modakaḥ pathi "Devadatta likes sweetmeats on the way".A. 1.4.33 means: "the technical term sampradana denotes one who is pleased (pri ̄yamāṇa) in relation to the action denoted by verbal roots meaning 'to please' (rucyrtha)".The counterexample in question is provided for the condition to explain the need for stating pri ̄yamāṇa (see tab. 1, raw 1).We noted at the beginning of our introduction that the essential feature of a counterexample is that having all the conditions stated in the rule except one (ekāṅgavikalatā) (Ajotikar et al. 2016).There is no other recurring term available in this rule other than kāraka (from A. 1.4.23).Hence, when we apply this essential feature, it implies that in the absence of the term pri ̄yamāṇa, the rule (rucyarthānāṁ) would mean that any participant in the action denoted by a verb having the same meaning as the verb ruc 'to please' would be termed sampradana.Thus, pathin 'path' would be termed sampradana.If not denoted (A.2.3.1 anabhihite) by a verbal affix, kr̥ t or taddhita affix or a compound, the item termed sampradana would get a fourth-triplet nominal termination by A. 2.3.13 caturthi ̄ sampradāne.Thus, in the sentence, devadattāya rocate modakaḥ pathi, the word pathi would not be termed adhikaraṇa by A. 1.4.45ādhāro 'dhikaraṇam and would not get a seventh-triplet nominal termination by A. 2.3.36 saptamy adhikaraṇe ca.
There is a technical difficulty in this counterexample.The kārakasection comes under the scope of the heading A. 1.4.1 ā kaḍārād ekā sañjñā: "Beginning with this sūtra and ending with A. 2.2.38 káḍārāḥ karmadhāraye, only one (ekā) technical term (saṁjñā) applies (to a given item)".A. 1.4.2vipratiṣedhe paraṁ kāryam is a conflict resolution metarule which means: "When there is a conflict (vipratiṣedha) (between two rules which can equally apply in a given domain), the operation provided by the subsequent rule (para) alone applies".Hence, in any situation of a conflict between two rules, the term stated by the later rule overrides the one provided by the previous rule.Thus, any kāraka term provided after the term sampradana would override the term sampradana.Therefore, it is not proper to posit that, in the absence of the condition pri ̄yamāṇa, pathin 'path' would not be termed adhikaraṇa and would not get seventh-triplet nominal termination.On the contrary, the word pathin 'path' should be termed adhikaraṇa and should get seventh-triplet nominal termination because the term adhikaraṇa is stated by A. 1.4.45ādhāro 'dhikaraṇam which is a subsequent to the rules (A.1.4.32-41) that provide the term sampradana.Thus, the counterexample involving the adhikaraṇa (devadattāya rocate modakaḥ pathi) fails to explain the real purpose of the condition pri ̄yamāṇa in A. 1.4.33.The same is true for the counterexamples on the subsequent rules.

3
Bhartr̥ hari on A. 1.4.33-36 In the quest for the origin of these counterexamples, it is necessary to study Bhartr̥ hari's views on A. the convention is to express an agent who prompts something to be the object of his desire, such as Devadatta, by the dative case rather than the nominative case in an active clause or instrumental in a passive clause.The term sampradāna is provided by A. 1.4.33 to such an agent who prompts something to be the object of his desire, Devadatta in this case; and, when not denoted by any verbal termination, such an agent gets the fourth-triplet nominal termination.Thus, devadattāya rocate modakaḥ "a sweetmeat is pleasing to Devadatta" is the valid expression; not *devadattaḥ modakaṃ rocayate "Devadatta prompts sweet-meat to become the object of his desire".Thus A. 1.4.33 is an exception to tatprayojako hetuś ca A. 1.4.55.
If the meaning of devadattāya rocate modakaḥ is "a sweetmeat causes Devadatta to desire it", and being pleased (pri ̄yamāṇa) qualifies the direct object in the action of causing the desire then Devadatta is termed karman by A. 1.4.52 gatibuddhipratyavasānārthaś abdakarmākarmakāṇāmaṇikartā sa ṇau.In that case A. 1.4.33 is an exception to A. 1.4.52. 4 Thus devadattāya rocate modakaḥ is a valid expression; not *devadattaṁ rocate modakaḥ "Modaka makes Devadatta to desire it".It is important to note that there is no reference to the adhikaraṇa in this discussion.So, the question concerning the source for the counterexample devadattāya rocate modakaḥ pathi remains unanswered.
Similarly, Bhartr̥ hari elaborates the meanings of the verbal roots listed in A. 1.4.34,namely, ślāgh 'to praise', hnu 'to hide from', sthā 'to stand' and śap 'to curse'.Helārāja clarifies Bhartr̥ hari's position only on the verbal root ślāgh, not on the others. 5With the help of this explanation we understand that A. 1.4.34 is an exception to A. 1.4.49(which assignes the technical term karman) or A. 1.4.55 (which assignes the technical term hetu).
5 dhātorarthāntare hi vr̥ ttau sakarmakatvam.evaṁ devadattāya ślāghate iti guṇotkarṣeṇa devadattaḥ śasyamāno guṇavattayā tatsamarthācaraṇād adhyaropitaprayojakabhāvo hetusañjñāṁ prāpto guṇ ākhyānena vā jñāpayitum iṣṭo jñāpanenāpyamānatvāt karmasañjñāḥ iti ślāghhnuṅ ityādinā sampradanasañjñāḥ kathyate "For the verbal root becomes transitive when the meaning of the verbal root changes.Thus in the sentence, 'Someone praises Devadatta,' The technical term sampradāna is provided by A. 1.4.34either in exception to the term hetu which Devadatta, who is being praised because of his virtues, would obtain because the status of a prompter is superimposed on him because he behaves virtuously because of the fact that he possesses virtues, or in exception to the term karman which would obtain by virtue of the fact that he is the object of the desire to make him known by the means of description of his virtues" (Tripathi 1979, 319).2785-5953  2, 2, 2023, 1-14   8 The essence of his discussion is as follows: the verbal root ślāgh is transitive (sakarmaka) here.There are two possibilities as far as the expression devadattāya ślāghate is concerned.(1) Yajñadatta gets encouraged to praise Devadatta because of Devadatta's extraordinary merits.In such a situation Devadatta's merits cause Yajñadatta to praise him.Thus, Devadatta would be termed hetu by A. 1.4.55 tatprayojako hetuś ca, and *devadattaḥ ślāghayate "Devadatta causes someone to praise him" would be the expression.However, such a construction is blocked by A. 1.4.34 that provides the term sampradana for Devadatta.(2) When Yajñadatta wants Devadatta's merits to be known then Devadatta would be termed karman by A. 1.4.49kartur i ̄psitatamaṁ karma, and *devadattaṁ ślāghate "someone praises Devadatta" would be the expression.However, it is blocked by A. 1.4.34 that terms Devadatta sampradana.Thus A. 1.4.34 is an exception to A. 1.4.55 and A . 1.4.49.Here also we do not find any discussion related to the locus that would be termed adhikaraṇa by A. 1.4.45 and condition a seventh-triplet nominal termination by A. 2.3.36.Hence the source for the counterexample devadattāya ślāghate pathi is not the Vākyapadi ̄ya.

Bhasha e-ISSN
This discussion is incomplete without the reading devadattaḥ ślāghate "Devadatta praises" which seems to be accepted by Haradatta as a valid reading.This reading merely points out that Devadatta is an agent of the action of praising.However, Haradatta (Mishra 1985, 552) explains this counterexample saying jñīpsyamānavacanāt karmasaṁjñaiva bādhyate, na kartr̥ saṁjñetyarthaḥ "A.1.4.34blocks only [the assignment of] the term karman by A. 1.4.49not the term kartr̥ by A. 1.4.54 because the condition jñīpsyamāna 'desired to be made known to' is added".The statement that the term karman alone is blocked agrees with what Bhartr̥ hari and Helārāja argued, namely that A. 1.4.34 is an exception to A. 1.4.49.However, the example devadattaḥ ślāghate as read by Haradatta does not show this.Instead, it implies the opposite, namely that the term kartr̥ would be blocked if the condition jñīpsyamāna is not stated in A. 1.4.34.Haradatta goes on to provide another counterexample gārgikayā ślāgate sabhāyām "He or she boasts of belonging to the family of Garga in the court".And he claims this example shows that the terms karaṇa and adhikaraṇa are not blocked by A. 1.4.34.However, this counterexample too shows just the opposite, i.e. these terms would be blocked by A. 1.4.34 in the absence of the condition jñīpsyamāna.
The variant reading devadattaṁ ślāghate, as Haradatta states, is not in conformity with Bhartr̥ hari's and Helārāja's conclusion.They state that the term karman is blocked by A. 1.4.34.It seems that, according to Bhartr̥ hari and Helārāja, Pāṇini does not attest the expression *devadattaṁ ślāghate when the verbal root ślāgh means 'to praise'.Haradatta (Mishra, 1985, 552) explains this counterexample as follows: jñi ̄psyamāna means the one who is made known so it is proper to use the second-triplet nominal termination (after the word which is an object of the action of making know).Those, however, who explain that "jñīpsyamāna" = "ākhyāyamāna" (what is being related/ spoken about), propose to add sixth-triplet nominal termination after the receiver of the information as is observed in devadattāya ślāghate yajñadatto viṣṇumitrasya (Yajñadatta praises Devadatta to Viṣṇumitra).
Here Haradatta points out that, in the sentence devadattaṁ ślāghate, Devadatta is merely an object of the praise but neither him nor anyone else are made known of that praise.Moreover, when someone (Viṣṇumitra) other than the direct object (Devadatta) is made known of the praise, he gets a sixth-triplet nominal termination, not a fourthtriplet one.This discussion indicates that some grammarians allow the expression *devadattaṁ ślāghate when that praise is not intended to be known by Devadatta or anyone else.The probable source for this reading will be discussed in Section 4.2.Interestingly, at least five manuscripts of the Kāśikāvr̥ tti support the reading devadattaṁ ślāghate which was known to Haradatta. 6It is surprising that the editors of the Kāśikāvr̥ tti (Sharma et. al. 1969-70) never considered this reading seriously even though it had the support of a commentator as well as from the manuscript tradition.
In the case of A. 1.4.36 spr̥ her i ̄psitaḥ also, Helārāja, elaborating Bhartr̥ hari's argument, states that it is an exception to A. 1.4.49. 7 In the sentence puṣpebhyaḥ spr̥ hayati "He desires flowers", flowers are the most desired objects so there is the possibility of being termed karman by A. 1.4.49.They are not so termed because A. 1.4.36 provides the term sampradāna for the object of desire.Thus A. 1.4.36 is an exception to A. 1.4.49.Here as well, according to Bhartr̥ hari as Helārāja explains, Pāṇini does not provide for the expression *puṣpāṇi spr̥ hayati.Joshi and Roodbergen (1995, 111 fn. 11) also opine that *puṣpāṇi spr̥ hayati is not allowed.Furthermore, in fn. 12, they say that there is no technical difference between i ̄psita 6 Manuscripts C4452 at the Banares Hindu University; IOL 4087 at the India Office London; 145-1K-145-2K at Shri Ranbir Prasad Research Institute, Jammu; VI 863 at Rajasthan Oriental Research Institute, Jodhpur; 37926 at Sampurnananda Sanskrit Vishvavidyalaya, Varanasi.
Bhasha e-ISSN 2785-5953 2, 2, 2023, 1-14 10 and i ̄psitatama.So one should use the fourth-triplet nominal termination after the object of desire in the case of the verbal root spr̥ h.However later grammarians, 8 disagreeing with Bhartr̥ hari, unanimously say that when the speaker intends to express that the desire for something is excessive, then A. 1.4.49overrides A. 1.4.36 by the principle that the subsequent rule (para) alone applies.Here none of the commentaries provides *puṣpāṇi spr̥ hayati as a counterexample.Instead they create optionality between A. 1.4.36 and A. 1.4.49.There is a lot of emphasis on the relation between A. 1.4.36 and A . 1.4.49;yet there is no reference to the relation of A. 1.4.36 to rules that provide any other kāraka terms like adhikaraṇa.The expression puṣpebhyaḥ spr̥ hayati and other such expressions in which the object of desire appears in a case other than accusative are idiosyncratic.
Bhartr̥ hari states that A. 1.4.35dhārer uttamarṇaḥ is an exception to A. 2.3.50 ṣaṣṭhi ̄ śeṣe.Helārāja explains the example devadattāya śataṁ dhārayati "He or she owes a hundred to Devadatta", as follows: Devadatta, who is the creditor, is the cause of the action of owing because he is the one who gives a hundred.However, there is no explicit mention of the action of giving.In relation to the implicit action of giving, the agent would get a sixth-triplet nominal termination.But this is blocked by A. 1.4.35 by providing the term sampradāna.In short, *devadattasya śataṁ dhārayati "He or she owes hundred of Devadatta" is not an accepted usage.Here also, we can observe that there is no mention of any other kāraka term like adhikaraṇa.
It remains unanswered why the counterexamples on all of these sūtras include the adhikaraṇa in locative case.Bhartr̥ hari's discussion revolves around clarifying how the term sampradāna provided by A. 1.4.33-36 is an exception to the two kāraka terms hetu and karman, and to the sixth-triplet termination (ṣaṣṭhi ̄).It is obvious that we cannot expect any counterexample that includes a word that denotes an object designated by either of these two kāraka terms.It seems the commentators avoid using any hypothetical counterexample that does not actually occur in correct usage.Hence, they provide a counterexample that uses a word denoting an object that is termed adhikaraṇa.The rules that provide the term adhikaraṇa (A.1.4.45)occur before the rules that provide the terms hetu (A.1.4.55) and karman (A. 1.4.49-52).Because these rules occur in the section governed by the ekasañjñā adhikāra in which the subsequent rule applies (A.1.4.1-2), the subsequent kāraka term prevails over the previous one when there is a conflict.The term adhikaraṇa is provided 8 Jinendrabuddhi: yadā tu puṣpādi ̄nām i ̄psitatamatvaṁ vivakṣyate, tadā paratvāt karmasaṁjñaiva bhavati -puṣpāṇi spr̥ hayati ̄ti (Mishra 1985, 553).Haradatta: prakarṣavivakṣāyāṁ tu paratvāt karmasaṁjñaiva bhavati -puṣpāṇi spr̥ hayati ̄ti (Mishra 1985, 553).Bhaṭṭojī Dikṣita: prakarṣavivakṣāyāṁ tu paratvāt karmasaṁjā (Chaturveda, Vidyabhaskara 1961, 647-8).after the term sampradāna and before the term karman.Therefore, sentences that include a word that denotes an item termed adhikaraṇa easily present themselves as uncontroversial candidates to be counterexamples that are neither hypothetical nor blocked by the term sampradāna.There is no other explanation we can think of for these counterexamples.Whether these serve as suitable counterexamples is a different question.In fact, they do not serve any purpose whatsoever, nor do they help one understand the relation of these special rules to other rules.

4
The Kātantra and the Cāndravyākaraṇa: Since the Kātantra (first century CE) and the Cāndra (fifth century CE) are pre-Kāśikāvr̥ tti, it is interesting to see how they deal with these special cases. 9Thereby we may find some trace of counterexamples with the counterexample that uses a word denoting an object that is termed adhikaraṇa.

The Kātantra vyākaraṇa
The Kātantra grammar states only one rule that deals with the term sampradāna: K. 2.4.10 yasmai ditsā rocate dhārayate vā tat sampradānam "the participants in the action to which one desires to give, the one to whom something is pleasing, and the one to whom one owes are termed sampradāna".The Kātantra grammar deals here with only three of the several constructions treated by Pāṇini and covers all of them in one rule.This rule corresponds to A. 1.4.33, A. 1.4.34 and A. 1.4.35.Durgasiṁha (ninth-tenth century C.E), a wellknown commentator of the Kātantra grammar, provides three examples (Eggeling 1874, 79-80): 1. brāhmaṇāya gāṁ dadāti "he/she donates a cow to a brahmin".2. devadattāya rocate modakaḥ "sweetmeats are pleasing to Devadatta". 3. viṣṇumitrāya gāṁ dhārayate "he or she owes a cow to Viṣṇumitra".
The remaining expressions like devadattāya ślāghate "He/she praises Devadatta" etc. dealt with by Pāṇini are covered under the category 9 As the chronology was recently summarised by Ben-Dor (2019), the Jainendramahāvṛtti and Kāśikāvr̥ tti were composed in the same time period.However the conclusion of this section equally applies to the Jainendramahāvṛtti.
Bhasha e-ISSN 2785-5953 2, 2, 2023, 1-14 12 of purpose (tādarthya). 10The Kātantra grammar does not term them sampradāna.So, it avoids any kind of detailed semantic treatment of the topic.It focuses merely on accounting for the caturthi ̄ vibhakti.Durgasiṁha does not discuss here any counterexample, nor does he mention any alternative expressions like *devadattaṁ ślāghate or *puṣpāṇi spr̥ hayati.Although the Kātantra sūtra-pāṭha is dated no later than first century CE, the available commentaries are dated after the ninth century CE.So, there is a gap of at least eight hundred years between the composition of the sūtra-pāṭha and the commentaries.Hence the discussion we find in Durgasiṁha's commentary or later commentaries can be traced to other commentaries like the Cāndravr̥ tti or Kāśikāvr̥ tti.It is difficult to state with certainty whether the author of the Kātantra sūtra-pāṭha accounted for both usages: devadattāya ślāghate and *devadattaṁ ślāghate, or merely for the former.Moreover, there is no counterexample provided that matches with any counterexample in the Kāśikāvr̥ tti.
12 vyāpyavivakṣāyāṁ tu devadattaṁ ślāghate puṣpāṇi spr̥ hayati ̄ti "When the speaker has the desire to express (Devadatta, or the flower [puṣpa]) as the direct object, then devadattaṁ ślāghate, puṣpāṇi spr̥ hayatīti, etc. are valid expressions" (Chatterji 1953, 176).the fifth century CE.This change was also accepted by the Pāṇiniȳas (footnote 6 and 8).The variant devadattaṁ ślāghate for the counterexample devadattāya ślāghate pathi on A. 1.4.34mentioned by Haradatta seems to account for the language change that is clearly accepted by the Cāndravr̥ tti.Alternatively, it is possible that devadattaṁ ślāghate is an interpolation in the text of the Kāśikāvr̥ tti under the influence of the Cāndravr̥ tti.Even though there is no counterexample given in the Cāndravr̥ tti on C. 2.1.74,there is a counterexample on C. 2.1.75uttamarṇa iti kim? gāṁ dhārayati devadattaḥ "why the condition uttamarṇa 'creditor'?'Devadatta owes a cow'".Devadatta who is a debtor is not termed sampradāna.This counterexample is similar to the variant devadattaḥ ślāghate "Devadatta praises" in A. 1.4.34,which is noted by Haradatta and many printed editions of the Kāśikāvr̥ tti.In short, we can observe that there is no counterexample provided that matches any counterexample given in the Kāśikāvr̥ tti that includes a word denoting an object termed adhikaraṇa.Thus there is no historical trace of these counterexamples.

Conclusion
Ajotikar et al. (2016) claim that most of the counterexamples provided in the Kāśikāvr̥ tti conform to the distinctive feature of a counterexample, namely having all the conditions stated in the rule except one (ekāṅgavikalatā).But the counterexamples discussed in this paper do not comply with this general claim.When it comes to the issue of complex semantic conditions (prīyamāṇa, jñi ̄psyamāna, i ̄psita or uttamarṇa) stated in A. 1.4.33-36, it is difficult to justify the usefulness of the available counterexamples.These must have been included in order to fulfil the criteria of a vr̥ tti.A vr̥ tti typically includes an example, a counterexample and supplementary words necessary to complete the meaning of a rule.However, these counterexamples fail to justify the significance of the semantic conditions stated in the rule for which they are provided.Hence, they do not serve any purpose.
Kāraka Rules A. 1.4.33-36 in the Kāśikāvr̥ tti Serve Any Purpose? the Kāraka Rules A. 1.4.33-36 in the Kāśikāvr̥ tti Serve Any Purpose? 11 Kāraka Rules A. 1.4.33-36 in the Kāśikāvr̥ tti Serve Any Purpose? the Kāraka Rules A. 1.4.33-36 in the Kāśikāvr̥ tti Serve Any Purpose? 13

Tanuja Ajotikar Do Counterexamples on the Kāraka Rules A. 1.4.33-36 in the Kāśikāvr̥ tti Serve Any Purpose? 3Table 1
given in pāṇinian commentaries

Table 2
Sūtra and its exceptions according to Bhartr̥ hari 1.4.33-36.In the Sādhanasamuddeśa (verse 130) of his Vākyapadi ̄ya, Bhartṛhari explains that these rules are exceptions to the provision of the term hetu(A.1.4.55), karman(A.1.4.49)and the provision of the sixth-triplet nominal termination(ṣaṣṭhi ̄ vibhakti A. 2.3.50).Helārāja elaborates on Bhartr̥ hari's views.On the basis of their discussion, I present a list of those rules to which A.1.4.33-36are exceptions in table 2 [tab.2].Bhasha e-ISSN 2785-5953 2, 2, 2023, 1-14 6In the sūtras beginning with rucyarthānāṁ pri ̄yamāṇaḥ(A.1.4.33-41), it is the śāstra which gives the name of Recipient (sampradāna) to what would otherwise have been Hetu, Karman or S éṣa.(Iyer  1971, 223)Bhartr̥ hari focuses on explaining the semantics of the verbal roots listed in the sūtras that extend the provision of the term sampradāna to items other than the recipient in the action of giving(A.1.4.32).According to Bhartr̥ hari, in a sentence like devadattāya rocate modakaḥ, Devadatta prompts the sweet-meat to be the object of his desire and is therefore termed hetu (agent of a causative action) by A. 1.4.55 tatprayojako hetuś ca.The meaning of the verbal root ruc is such that its agent is something other than the kāraka who is pleased, that is, who is the substrate of desire.The sweet-meat (modaka) is the object of Devadatta's desire.Devadatta prompts the sweet-meat (modaka) to be the object of his desire.Because he is the agent who prompts (prayojaka), Devadatta would be termed hetu, and that would force the affix ṇic to apply after the verbal root ruc by A. 3.1.26. 3 However, yate "For in this way in the sentence 'Devadatta likes sweet-meats' the meaning of the verbal root ruc is the desire by another agent.Thus (the sentence means), 'Devadatta prompts the sweet-meat to become the object of his desire, that is, because of his desire, acts in accordance with his desire.'Sothe technical term sampradāna is provided by A. 1.4.33 in exception to the technical term hetu which obtains to Devadatta because he is a prompter"(Tripathi 1979, 318)

Tanuja Ajotikar Do Counterexamples on the Kāraka Rules A. 1.4.33-36 in the Kāśikāvr̥ tti Serve Any Purpose?
11Further, the Cāndravr̥ tti elaborates as follows: when desired as a direct object of the actions denoted by the verbal roots ślāgh or spr̥ h, the expressions devadattaṁ ślāghate "he/ she praises Devadatta" and puṣpāṇi spr̥ hayati "he/she desires flowers" are valid. 12This is very interesting because what Pāṇini seems not to approve of is accepted as a valid expression.Historically we can say that these expressions not accepted earlier were allowed by Cāndra grammar states five different rules (C.2.1.73-77)to cover the term sampradāna.Among them, C. 2.1.74rucimati and C. 2.1.75dhārer uttamarṇe correspond to A. 1.4.33 and A. 1.4.35respectively.Interestingly, like the Kātantra, there are no rules that correspond to A. 1.4.34 and A. 1.4.36.While commenting on C. 2.1.74rucimati, it is asked: "how do we provide for devadattāya ślāghate, chātrāya hnute, chātrāya āśr̥ ṇoti etc.?" It is answered: "Those [usages] will be taken care by tādarthya-caturthi ̄".