The Sword and the Sheath: Three Notes on Kaccāyana 1

A. Ruiz-Falqués

Introduction

The Kaccāyanabyākaraṇa has been used over the centuries as a beginner’s manual for the study of the Pāli Buddhist literature. It opens with a fascinating and mysterious aphorism: attho akkharasaññāto (meaning is understood by means of speech-sounds). Although there seems to be no doubt in modern editions that Kaccāyana 1 is the first rule of the grammar, its status as a sutta (grammatical rule) has been debated among traditional scholars. The different views of the ancient commentaries and the possible motivation behind them will be discussed in Sect. 1 of the present paper. Sect. 2 focuses on the formulation of the sutta. It hypothesizes that it could have been adapted from a Sanskrit sūtra that originally had a different meaning. In Sect.3 I will discuss the convention of the divine revelation of speech-sounds in the Sanskrit grammatical tradition and I will draw parallels with Kaccāyana 1. One theme runs through these three notes is the rarity of Kaccāyana 1 as an opening statement that emphasizes meaning (attha), unlike the first rules of most grammars that open with a statement about speech-sounds (akkhara). My point of departure in this article is the work of other scholars in the field, notably Subhūti (cf. Gornall and Gunasena (2018), Pind (1996, 2012), Gornall (2014) and more recently Watanabe (2019)).

§1. The Status of Kaccāyana 1 in the Pāli Tradition

1.1 The first problem that we encounter with Kacc 1 has to do with its role. Some interpreters believe that it is a pubbavākya, a sort of introductory remark, not a sutta (grammatical rule). Others are of the opinion that it is a sutta and it should be read as a part of the sutta-pāṭha, that is to say the sequence of rules that constitute the grammatical system. According to the Mukhamattadīpanī (Mmd), the oldest extant commentary on Kacc/Kacc-v, Kacc 1 is defined as a sutta, but an alternative opinion is also reported:¹

vuttiṁ kubbatā vuttādo gāthādvayaṁ vuttaṁ | sutte kubbatā suttass’ ādo pubbavākyam āraddhaṁ attho akkharasaññāto ty apare | – Mmd 7.15-19

The author of the vutti composed the two stanzas at the beginning of the vutti and the author of the suttas started at the beginning of the [first] sutta with the preliminary statement, namely, “meaning is conveyed by means of speech-sounds” (attho akkharasaññāto). This is what others maintain.

According to Watanabe (2019: 49) we can draw two conclusions from the quoted passage:

The authors of the Kacc and Kacc-v might have been different individuals, and
Vimalabuddhi (or Vajirabuddhi), author of the Mukhamattadīpanī, most probably did not support this position.

Watanabe’s hypothesis is plausible given the fact that Vimalabuddhi uses the term suttakāra/-ka only twice in the entire commentary,² and he never uses the term vuttikāra. It seems, however, that the main point of controversy is whether the Kacc 1 is a sutta or a pubbavākya. Vimalabuddhi makes clear that, for all purposes, Kacc 1 will be treated as a sutta. But he complicates this matter by first denying its status as saññāsutta (definition rule),³ and then later on treating it as a saññāsutta.⁴ Indeed, he seems to understand Kacc 1 as a paribhāsā (meta-rule), but that is never explicitly stated.

1.2 The anonymous Mukhamattadīpanī-porāṇaṭīkā (Mmd-pṭ), the oldest commentary upon the Mukhamattadīpanī, explains apare (others) as those who believe that the same thera, presumably Mahā Kaccāyana, wrote the treatise “twice” (dikkhattuṁ): he first wrote the sutta with the examples, opening it with a preliminary statement and afterwards he wrote the vutti in order to elucidate the cryptic text of the sutta. According to Watanabe (2019: 10-92), the view of the “others” is rejected by the Mukhamattadīpanī and Mukhamattadīpanī-porāṇaṭīkā. But here, again, the nature of the disagreement is not clear. What the Mukhamattadīpanī and Mukhamattadīpanī-porāṇaṭīkā seem to be defending, in my opinion, is that Kacc 1 is a sutta, i.e. it is part of the sutta-pāṭha and not an independent statement. It is interesting to note, in this respect, that Vimalabuddhi only begins his full-fledged commentary in Kacc 2, for he treats Kacc 1 as a general principle of word comprehension that applies from the Kacc 2 onwards.⁵

1.3 More importantly, according to the Mukhamattadīpanī-porāṇaṭīkā, the function of Kacc 1 is to represent the Buddha’s principle that meaning (attha) is paramount and linguistic sound (sadda, akkhara) is subordinate to meaning. It does so by quoting some verses of unknown origin:

[One could object:] “But is it not true that speech-sounds are primary because it is when there are speech-sounds that meaning is produced?” That is not so. Only meaning is primary because, if there is nothing to be named, the applying (nyāsa) of speech-sounds does not occur. And furthermore, it has been stated:

When the meaning is not known, that [speech-sound] cannot be established, but when there is a meaning, it is possible to establish the speech-sound.

The word akkhaṁ here means akkharaṁ (speech-sound). And, furthermore, it has been stated:

“Only the sword cuts the creeper, not the sheath;

Only water removes thirst, not the jugs;

Only the lion kills the enemy, not the cave” – thus thinking,

the Thera [Kaccāyana] mentions “meaning” in the first rule of his grammar (nirutti),

“Only the wish-fulfilling jewel (cintāmaṇi) is charming, not the thread;

only clothes beautify people, not the baskets;

only food removes hunger, not the plate” – thus thinking,

the Thera [Kaccāyana] mentions “meaning” in the first rule of his grammar (nirutti),

and having first attained these four types of discrimination,

following what was stated by the Conqueror, he refers to “meaning” (attho).⁶

According to these stanzas, probably quoted from a lost commentary on Kacc, Kaccāyana Thera composed the first rule “following what had been stated by the Conqueror” (jinavuttānusarinā). This expression could be interpreted in two ways:

Following a similar statement made by the Buddha on this topic; and
following the principles of the dhamma explained by the Buddha.

Option (1) fits well into the context, for, as Pind has pointed out, Kacc 1 could be based on a canonical passage that highlights the importance of phonetics as a requisite for the correct inference of meaning (see §2.1). The passage from (Mukhamattadīpanī-porāṇaṭīkā) could perhaps be read as a veiled critique of Buddhappiya’s Rūpasiddhi and other grammarians who favoured phonetics over meaning (cf. Pind 1996: 69).

1.4 The debate on the status of Kacc 1 is carried over in the Mukhamattadīpanīsāra (Mmd-sāra) of Guṇasāgara (thirteenth-fourteenth century ce, Pagan, Myanmar).⁷ Guṇasāgara repeats the aforementioned sword and sheath simile from the Mukhamattadīpanī-porāṇaṭīkā and adds a verse that is later on quoted in the Kaccāyana-vaṇṇanā (see §1.7):

[14.] Other teachers call this sutta a pubbavākya, but due to its similarity to suttas, we think that it is a meta-rule (paribhāsā).

[15.] “And because it is stated first, it is said to be a preliminary meta-rule, therefore, this sutta is a paribhāsā”, thus is it stated in the Sandhivinicchaya.⁸

By using the pronoun “we”, Guṇasāgara is probably trying to establish the standard interpretation of his school: Kacc 1 is a sutta; and among the different types of suttas, it is a paribhāsā, not a saññā. Furthermore, the name pubbavākya is compatible with being a paribhāsāsutta. Stanza 15 seems to name a text called the Sandhivinicchaya (Elucidation on the Sandhi [Book of Kaccāyana]?). To the best of my knowledge, this work has not survived. It is possible that the verses quoted by the Mukhamattadīpanī-porāṇaṭīkā and repeated in the Mukhamattadīpanī-sāra 11ff. are from the Sandhivinicchaya itself. From this reference and the previous one in the Mukhamattadīpanī-porāṇaṭīkā we may surmise that the status of Kacc 1 was a subject of discussion even in other grammars that are now lost.

1.5 The implicit background of this controversy is probably the old debate concerning the relationship between sound (sadda, Skt: śabda) and meaning (attha, Skt: artha). If Kacc 1 is read as a saññāsutta, that is to say as a sutta giving the definition of the term akkhara (speech-sound), the sense of the rule is “akkharas are those things that convey meaning”. The focus would be on speech-sounds, as if they were the agents in the process of communication – here akkharehi (cf. Kacc-v 1.11) would be analysed as kattusādhana (a means to express the agent). Conversely, if Kacc 1 is read as a paribhāsā, the rule states that meaning is accessed through akkharas and meaning is the primary element – here akkharehi would be analysed as karaṇasādhana (a means to express the instrument). It is a matter of detail, but the detail matters because it reflects two opposite views on language: one sees sound and meaning as forming an inseparable, eternal unit and the other sees sound, or words, as mere conventions that make communication possible. In the Pāli tradition, Vinaya scholarship has tended to adopt the first position and Sutta scholarship the second.⁹ The idea that one should take meaning only as a “teacher” and not the letters or speech-sounds and that one “should not take delight in speech-sounds” is echoed in the first pages of Sri Rahula’s Padasādhanaṭīkā (Pds-ṭ), which includes a statement attributed to the Buddha: “Take refuge in the meaning; bhikkhus, do not take refuge in the letter.”¹⁰

1.6 In later commentaries the focus of the discussion about Kacc 1 shifts to authorship. The question, here, is whether the rule comes from the Buddha or from the grammarian Kaccāyana. This controversy inspired a story narrated in Chapaṭa Saddhammajotipāla’s Kaccāyanasuttaniddesa (Kacc-nidd), written around 1446 ce in Pagan, Myanmar. The story can be summarized as follows: In seeing an old monk near the lake Anotatta, wrongly reciting the mantra udaya-(b)baya (arising-passing away) as udaka-baka (water heron), the Buddha admonished him that he should pronounce words correctly because “meaning is conveyed by speech-sounds” (attho akkharasaññāto).¹¹ As Pind has pointed out, this story is reminiscent of a passage in the Mūlasarvāstivādin Vinaya, which is also found in other texts of Southeast Asia, such as the Saddavimala (Pind 1996: 70). It is noteworthy that the original context of attho akkharasaññāto is not a normal communication process, in which there is a sender and a receiver but a monk’s recitation of a mantra, by and for himself. What the Buddha seems to emphasize is not only that correct pronunciation is necessary in order to convey the right meaning to others but also to convey the right meaning to oneself. The passage seems to suggest that the efficacy of certain texts is dependent on correct recitation, a notion that is not foreign to Pāli grammarians (Ruiz-Falqués 2017: 75). After telling this story, Chapaṭa adds:

Furthermore, the Thera Kaccāyana, having known the intention of the Blessed One, placed this statement (vākyaṁ) attho akkharasaññāto at the beginning, and then he produced this treatise [on grammar]. They also say that this sutta was composed by Kaccāyana.¹²

1.7 A similar version of the story is found in Mahā Vijitāvī’s sixteenth-century commentary called the Kaccāyana-vaṇṇanā (Kacc-vaṇṇ). It is difficult to know if this version is a formerly existing story or created by Mahā Vijitāvī. Two elements reveal some level of corruption. First of all, as Pind has observed, the original mantra for “meditation exercises” (kammaṭṭhāna) is not udayabbaya, but khayavaya, and yet the mispronounced mantra after seeing the heron in the water is still udakabaka; second, after mispronouncing the mantra: khayavaya as udakabaka, one of those brāhmaṇas produces an even stranger variation: “after seeing a cloth (paṭa) in a pot (ghaṭa), he distorted [the mantra again] as ghaṭapaṭa (pot-cloth)” (eko ghaṭe paṭaṁ disvā ghaṭapaṭo ti virajjhati).¹³

This second pseudo-mantra could have been introduced on the basis of the following passage of Mukhamattadīpanī upon Kacc 1:

This is the meaning herein: whatever meaning/object of a word, such as pot (ghaṭa), cloth (paṭa), and so on, is to be expressed, it can only be conveyed by means of speech-sounds.¹⁴

After telling the story of the two brāhmaṇas, Vijitāvī gives a second interpretation:

It is also stated that, because [this statement/this rule] was established by the Thera, it is said to be a paribhāsā (meta-rule).¹⁵

He substantiates this claim by quoting the Mukhamattadīpanīsāra verse 14 (§1.4), which Watanabe has interpreted as the final opinion of the Kaccāyana-vaṇṇanā (Watanabe 2019: 1093). Finally, the story of the old recluse at the bank of the Anotatta lake is cited, in a clear reference to the preface of Kaccāyanasutta-niddosa.

1.8 After this short survey, the interpretation of the Mukhamattadīpanī seems to be that Kacc 1 is not a pubbavākya, but a sutta by Kaccāyana; the Mukhamattadīpanī-porāṇaṭīkā and Mukhamattadīpanī-sāra clarify that it is a paribhāsāsutta by Kaccāyana; the Kaccāyanasutta-niddosa is of the opinion that it is a pubbavākya by the Buddha originally addressed to an old monk who mispronounced a mantra, and that it was later on adopted by Kaccāyana the grammarian; the Kaccāyana-vaṇṇanā, too, maintains that Kacc 1 is a pubbavākya by the Buddha but that it was originally addressed to two brahmaṇas who mispronounced a mantra. Whether authors believe that Kacc 1 was uttered by the Buddha himself or not, they all seem to acknowledge the fact that this rule expresses an idea that had been taught by the Buddha.¹⁶ Furthermore, it is possible to identify two different interpretations of Kacc 1, which are not necessarily incompatible: (1) Kacc 1 is about the importance of mastering speech-sounds (akkharas), that is to say, phonetics and grammar, and (2) Kacc 1 is about the primacy of meaning (attha) over speech-sounds (akkharas): “Take refuge in the meaning, bhikkhus, do not take refuge in the expression” [Sn-a Ee II 398.7-8].

§2. The Formulation of Kaccāyana 1 and Its Parallels

2.1 Kacc 1 is a unique sutta in the history of Pāli grammar. Neither Aggavaṁsa’s Saddanīti, nor Moggallāna’s grammar,¹⁷ nor derivates such as the Payogasiddhi and Padasādhana, have a similar pubbavākya. Even the Bālāvatāra, a work of the Kaccāyana school, excludes Kacc 1. The rule has no parallel in Pāṇini either. In his A Comparative and Critical Study of Kātantra and Kaccāyana Grammars, Dwivedi and Kumar does not give any Sanskrit parallel for Kacc 1. Instead, he establishes the correspondence between the first rule of the Kātantra (Kā 1.1.1) and Kacc 2 (2004: 16). Pind points to a Pāli canonical passage that reassembles Kacc-v upon Kacc 1 as a possible source for the formulation of Kacc 1,¹⁸ and he seems to dismiss the possibility that it is based on an older grammatical rule.

2.2 One may observe, however, that Kacc 1 bears a certain resemblance to the initial sūtras of works belonging to the so-called “Aindra School”,¹⁹ such as Kātantra 1.1.1: siddho varṇasamāmnāyaḥ (the list of sounds has been established [by tradition]); Yāska’s Nirukta (Nir) I.1: samāmnāyaḥ samāmnāta (a traditional list [of words] has been handed down)²⁰; the Taittirīya Prātiśākhya (TPr) I.1: atha varṇasamāmnāyaḥ (now the list of sounds) (Whitney 1868).

	(1)	(2)	(3)
TPr I.1	atha	varṇa-	samāmnāyaḥ
Nir I.1	samāmnātaḥ	[pada-]	samāmnāyaḥ
Kā 1.1.1	siddho	varṇa-	samāmnāyaḥ
Kacc 1	attho	akkhara-saññāto

The possibility that Kacc was using as one of its models some grammar similar to the Taittirīya Prātiśākhya should not be dismissed.²¹ If we compare these two grammars we see that atha / atho > attho is not problematic from the phonetic point of view (cf. CPD s.v. atho). The beginning of a śāstra with the “auspicious” (maṅgala) word atha is a well-established convention (cf. MW s.v. atha). As to the words vaṇṇa/varṇa and akkhara/akṣara, they are synonyms (cf. Deokar 2008: 78) and they occupy the same position in the sūtra. Finally, samāmnāyaḥ and saññāta also present some superficial similarities in terms of sound, perhaps saññāta < samaññāta/-ya (?). Senart (1871: 11) pointed out a similar confusion between saññā/samaññā with regard to Kacc 9 parasamaññā payoge. Commentaries such as Kacc-v, the Mukhamattadīpanī and Rūpasiddhi seem to understand the word samaññā in Kacc 9 as equivalent to saññā (technical term),²² thus interpreting the rule as “the technical terms (sa[ma]ññās) of others [may be used] when applicable (payoge)”. As Senart (1871: 11) has observed, one would rather expect a derivation based on Skt: sāmānya (common), i.e. “those [technical terms] which are in common [i.e. shared] with others [i.e. with Sanskrit grammars, may be used] when applicable”.

2.3 What could be the reason for the author of Kacc to modify the formulation of the rule? If we assume that the model of Kacc 1 was a treatise similar to the Taittirīya Prātiśākhya, the original model could be represented in Pāli as follows: *atho akkharasamannāyo (now, the list of speech-sounds). Indeed, the list of sounds is given in Kacc 2 akkharā p’ ādayo ekacattālīsaṁ (and letters, starting with a, are forty-one). The purpose of Kacc 2 is not only to establish the number of speech-sounds in Pāli, but also to make clear that they are different from the Sanskrit sounds:

Or rather there is an objection (codanā) that, since there is mention of [the sounds] a, etc. in the expression “the sounds beginning with «a»”, why is “forty-one” mentioned? Here, the answer (parihāra) is that [the enumeration of sounds] indicates that the forty-one [sounds] are a help [in understanding] the suttantas, even though there are also other sounds existing outside [of the suttantas].²³

It is possible, therefore, that a rule such as *atho akkharasamannāyo was considered redundant and the author of Kacc decided to replace it with a general statement about the purpose (payojana) of grammar, especially if that was meant to be an opening statement not only for the section on phonetics (sandhi), but for the entire compilation of the four major books of Kacc.²⁴ The meaning of the new formulation, attho akkharasaññāto, is glossed in Kacc-v as expressing the purpose of grammar:

The meaning of any verbal expression is conveyed only by means of speech-sounds. For, if there is any failure in [pronouncing/writing] the speech-sounds, the meaning is difficult to understand. Therefore, skilfulness in speech-sounds is of great help in [properly reciting, or properly understanding] the suttantas.²⁵

2.4 In addition to possible Sanskrit parallels to Kacc 1, we have testimonies of what could be the opening suttas of two lost Pāli grammars,²⁶ one by Bodhisatta Mahāthera and the other by Sabbaguṇākara Mahāthera.²⁷ The former is mentioned in the following passage of the Padasādhanaṭīkā:

The Venerable Mahāthera Bodhisatta says “the forty-one speech-sounds are helpful in [understanding] the speech of the best among men [i.e. the Buddha]”.²⁸

The grammar of Sabbaguṇākara is quoted also in the Padasādhanaṭīkā in a passage that is illuminating in several ways. Sri Rahula (fifteenth century ce) refers to the work when commenting upon the word sutta in another stanza that he cites:²⁹

Furthermore, Master Sabbaguṇākara, in [his work] Māgadhikasaddakalikā (Flower Bud (?) of Magadhan Words) has stated: “The sutta [is] vyākaraṇa (grammar):³⁰ only the sutta is the embodiment (sarīra) of vyākaraṇa because it [i.e. grammar] cannot be grasped by words that go beyond the sutta (ussutta, Skt: utsūtra³¹). When an explanation (anvākhyāna) is being made with regard to an error (vippaṭipatti) occurring in language (sadda), the statement (vacana) will not be grasped as long as the sutta [i.e. grammatical treatise] is not taught (dassita) – in the same way that a loose bundle of flowers remains impossible to grasp without a string (sutta) [to bind it].³² For that reason [the sutta is vyākaraṇa]. This is the meaning. Another [meaning is] that lakkhiya (that which is characterized) [i.e. language/words] and lakkhaṇa (that which characterizes) [i.e. the grammatical rules] are grammar.³³ Kātyāyana thinks that grammar is both [of these] together.”³⁴

Sabbaguṇākara seems to adhere to Patañjali’s view, which somehow excludes language from the definition of the term vyākaraṇa. From this passage we also understand that Sabbaguṇākara’s grammatical framework is based on the pāṇinīya tradition.

2.5 In another passage of the Padasādhanaṭīkā we find a second quotation from the same work. This one bears a striking similarity to Kātantra 1.1.1 and presents the list of speech-sounds as something already established, much in the same way as Sanskrit grammarians do:

Mahāthera Sabbaguṇākara has said: siddhakkamādādayo vaṇṇākkharā titālisā, “There are forty-three sounds established (siddha) in an order (kama) that begins with a.”³⁵

This could be the opening sutta of Sabbaguṇākara’ grammar or a sutta at the very beginning of the treatise. It appears as a combination of Kacc 1 and 2:

Kaccāyana	(1) attho akkharasaññāto	(2) akkharā p ’ādayo ekacattālīsaṁ
Sabbaguṇākara	siddhakkama-vaṇṇa-	a-d-ādayo akkharā titālisā

The opening suttas of Bodhisatta and Sabbaguṇākara resemble what the reading of Kacc 1 might have been if a model such as Taittirīya Prātiśākhya I.1 had been faithfully adapted:

Actual	(1) attho akkharasaññāto (Meaning is understood by means of speech-sounds)	(2) akkarā p’ ādayo ekacattālīsaṁ (and speech-sounds are forty-one, beginning with a)
Reconstructed model	(1) atho akkharasamannāyo (Now the list of speech-sounds:)	(2) akkharā ādayo (speech-sounds, beginning with a, are forty-one)

This is not meant to reconstruct a supposedly lost proto-Kacc, but rather to show what the model on which Kacc was based probably looked like and how the author of Kacc decided to adapt it to serve new purposes. If this hypothesis is true, it could help to explain why some commentators fail to recognize attho akkharasaññāto as a grammatical rule.

§3. Kaccāyana 1 and Legends of Grammatical Revelation

3.1 The story about the Buddha inspiring Kacc 1 represents an archetypal tale about the divine revelation of grammatical knowledge. It has abundant parallels in the Sanskrit tradition. According to a well-known legend in Somadeva’s Ocean of Rivers of Stories (Kathāsaritsāgara I.7.1-13), the Kātantra grammar was originally inspired by the war god Kumāra Kārttikeya, who conferred its first sūtra of Kātantra to the grammarian Śarvavarman. According to this legend, Śarvavarman was on a mission to compose the shortest grammar of Sanskrit. For this purpose, he retreated to the forest and undertook great austerities, “a vow of fast and silence”.³⁶ Eventually he fell to the ground unconscious. Śarvavarman himself recounts the rest of the story to the king who is the main recipient of his new grammatical method:

After that I remember a man with a spear in his hand arriving and saying to me in a clear voice: “Get up, my son, everything will turn out well for you.” Then, as if I had been showered by a downpour of the nectar of immortality, I awoke feeling well, free from hunger and thirst. Next I arrived in the vicinity of the lord, overcome by the burden of my devotion. After bathing, I excitedly entered his inner sanctum. Inside Lord Kārttikeya granted me his darshan and then Sarasvatī took bodily form and entered my mouth. Immediately afterwards the blessed lord recited with his six lotus-mouths a sūtra that was a perfected form of the alphabet. As soon as I heard it, with the impertinence that, alas, comes so easily to mankind, I guessed the next sūtra and said it myself. The lord then said to me, “If you had not spoken it yourself, this treatise would have wiped out that of Pāṇini. Because it is now so concise, it shall be called the Ka Tantra and also the Kalāpaka, after the name of my vehicle.” On saying this, he revealed that new concise grammar. …³⁷

Variations of this legend occur, but what is important here is that the revelation does not go beyond the first aphorism: siddho varṇasamāmnāyaḥ (the collection (samāmnāya) of speech-sounds (varṇa) has been established (siddhaḥ) [by tradition]) (Kā I.1.1).³⁸ This narrative seems to imply that the gift of the god is the complete list of sounds or “letters” (akṣarasamāmnāyaḥ), and not the grammatical rules, which are described as a by-product, or rather as a consequence, of the discovery of the akṣarasamāmnāya. The legend reflects a pattern that appears in other grammatical traditions as well. As Saini (1987: viii) has pointed out:

It is difficult to say whether the origin of the Kātantra-vyākaraṇa, as given in the Kathāsaritsāgara, is correct or not, because most of the post-Pāṇinian systems claim their origin from some god.

The reports and stories attached to Kacc 1 follow the same convention of tracing the origin of the treatise to a divine or great being.

3.2 The same idea applies to Pāṇini’s pratyāhāra sūtras, traditionally known as Śivasūtras, that is to say, the sūtras given or revealed by Lord Śiva. This tradition is relatively late, and there are reasons to believe that the pratyāhāra sūtras were actually authored by Pāṇini or someone belonging to the same grammatical school.³⁹ But conventions are strong and they adapt well across religious affiliations. In Sanskrit Buddhist traditions, for instance, the Hindu god Śiva is replaced by the bodhisattva Avalokiteśvara. According to the Tibetan chronicler Tāranātha, the bodhisattva is none other than Pāṇini and his initiation to grammar consists in receiving the revelation of the list of speech-sounds from a deity:

The brāhmaṇa Pāṇini was a friend of King Nanda. He was born in the Bhirukavana, in the west. He asked the palmist whether he was going to be an expert in grammar. The prediction was in the negative. With a sharp knife, he changed the lines of his own palm, studied grammar under all the grammarians of the world, worked hard and acquired great proficiency. Yet he remained dissatisfied. By intense propitiation, he received the vision of the tutelary deity. The deity appeared before him and uttered a, i, u, and he acquired knowledge of all the words in the three worlds.

The “outsiders” [bāhyas or tīrthikas] consider him as the Īśvara [= Śiva (?)]. But the “outsiders” have no basis for this. The “insiders” [= Buddhists] consider him as Avalokiteśvara. This is based on the prophecy of the Mañjuśrīmūlatantra: “Pāṇini, the son of a brāhmaṇa, will certainly attain the śrāvakabodhi. I have predicted that he would be the great lokeśvara [= Avalokiteśvara] by his own words.”

– Quoted in Deshpande 1998: 453-54

Note that the sūtra “a i u” is most probably a reference to aiuṆ, the first śivasūtra, where Ṇ is simply a “metalinguistic marker” (anubandha). Tāranātha relates a similar story about Candragomin, the fifth-century Buddhist Sanskrit grammarian of Nālandā. According to this narrative, Candragomin was secretly instructed by a statue of the Bodhisattva Avalokiteśvara, known in a previous life as Pāṇini:

From the outside door, he overheard the stone image of Ārya Avalokiteśvara teaching the Doctrine to Candragomin, much in the manner in which an Ācārya teaches his disciples. … – Ibid.: 457

These few examples are suffice to show that the stories behind Kacc 1 should not be taken in isolation. Reports that Kacc 1 was not the work of Kaccāyana, but a “revelation” from the Buddha, need to be understood in the context of this same convention.

§4. Concluding Remarks

4.1 In this paper I have tried to expand the analysis of Kacc 1 attho akkharasaññāto in three notes that examine different aspects of its exceptional character. We have seen how Kacc 1 is not exactly a grammatical rule, but rather a philosophical statement. Some scholars trace it back to the Buddha, whose grammatical insight, a manifestation of his paṭisambhidā (analytic skill), is thought to permeate the entire treatise of Kaccāyana (see §1.3). The first part of this paper shows that the controversy about the status of Kacc 1 involves a plurality of opinions, some of which partly intersect. By surveying the main commentaries on the Kaccāyana one may observe how the underlying problem of this controversy is whether Kacc 1 should be integrated in the sequence of suttas or not.

4.2 The discussion on the authorship of Kacc 1 is rooted in the same problem. Those who proposed the Buddha as the original author of Kacc 1 ultimately tried to substantiate the idea that this rule is not part of the sutta-pāṭha, but a theoretical principle that surrounds it. The general opinion of the Kaccāyana commentators seems to be that Kacc 1 is a paribhāsā-sutta and it was written by the grammarian Kaccāyana. Some of them, moreover, accept the possibility that Kacc 1 could have been originally uttered by the Buddha or that Kacc 1 represents an idea that had already been taught by the Buddha. Beneath the technical discussion, however, we found a more philosophical debate about the conception of language, grammar and Buddhist literature. The tradition is ambivalent with regard to the actual purport of Kacc 1: to some, it emphasizes correct pronunciation, to others it emphasizes the importance of meaning over expression.

4.3 This leads us to the second part of the paper, in which we have seen how the formulation of the rule attho akkharasaññāto reveals an original model that would follow the expected convention, viz. *atho akkharasamannāyo. The reason why the author of Kacc, or perhaps a compiler of Kacc and Kacc-v, may have adapted the rule, is that it was seen as redundant. Moreover, by changing its formulation, the idea that meaning is more important than expression was introduced as a programmatic statement. It also became a sort of justification for studying a worldly science such as grammar. This point ties in with the third and final note, where legends around Kacc 1 are re-assessed in the context of an ancient Indian tradition that presents grammatical knowledge as a revelation. Kacc 1 is presented as a revelation from a great being (the Buddha), in the same way that the first rule of the Kātantra was given by Lord Kumāra, or that the pratyāhāra sūtras of Pāṇini are known as a revelation from Lord Śiva. Such narratives dramatize the idea that grammatical knowledge, being a human creation, is only possible when something that is not created by humans, the alphabet, is acquired. In the case of Kacc 1, the revelation aspect is maintained, but we find an important variation. Here, meaning is more important than the sounds or letters that convey it. It is not the alphabet that comes from a superior being, but some “meaning”. Sounds or letters are only useful insofar as they convey something that is useful or true. That is why the author of the Mukhamattadīpanī-porāṇaṭīkā quotes the simile of the sword and the sheath. The usefulness of grammar is not denied, but what really “cuts the creeper” is the meaning of the Tipiṭaka, which shall be kept safe in the “sheath” of grammar. It is a minor variation, but it nevertheless explains why the author of Kaccāyana decided not to begin with a sutta that would be too similar to Sanskrit models. Those models assume that the alphabet is revealed independently from any meaning and that language itself (i.e. Sanskrit) is sacred. In the Sanskrit tradition, speech is divine. In the Kaccāyana tradition, conversely, the Tipiṭaka is the only interesting repository of speech-sounds, and these sounds (or written letters) are important as long as they convey the Buddha’s teaching, which is what ultimately needs to be studied.

Acknowledgements

I would like to thank William Pruitt, Petra Kieffer-Pülz and Alastair Gornall for their most valuable help in revising the draft of this article. All errors remain my own responsibility.

Abbreviations

If not stated otherwise, abbreviations follow the Bibliography of the Critical Pali Dictionary (https://cpd.uni-koeln.de):

CPD	=	Critical Pali Dictionary (https://cpd.uni-koeln.de)
DOP	=	A Dictionary of Pali, see Cone 2001–2020.
Ee	=	European edition
Kacc	=	Kaccāyanabyākaraṇa
Kacc-nidd	=	Kaccāyanasuttaniddesa
Kacc-v	=	Kaccāyanavutti
Kacc-vaṇṇ	=	Kaccāyanavaṇṇanā
Kā	=	Kātantra
MBh	=	Mahābhāṣya
Mmd	=	Mukhamattadīpanī
Mmd-pṭ	=	Mukhamattadīpanī-porāṇaṭikā
Mmd-sāra	=	Mukhamattasāra
MW	=	Monier-Williams (https://www.sanskrit-lexicon.uni-koeln.de/scans/MWScan/2014/web/webtc2/index.php)
Nir	=	Nirukta
Psd-ṭ	=	Padasādhana-ṭīkā
Rūp	=	Rūpasiddhi
Skt	=	Sanskrit
Sn-a	=	Suttanipāta-aṭṭhakathā
TPr	=	Taittirīya Prātiśākhya

References

PRIMARY SOURCES

Aṣṭādhyāyī, see Katre 1987.
Kaccāyana and Kaccāyanavutti, see Pind 2013.
Kaccāyanasuttaniddesa, ed. Medhankara, Colombo, 1915.
Kathāsaritsāgara, see Mallinson 2007.
Kātantra, see Eggeling 1874.
Mahābhāṣya = Vyākaraṇa-mahābhāṣya, ed. Kielhorn/Abhyankar, Poona, 1985.
Mukhamattadīpanī = Nyāsapāṭh, Yangon, Sudhammavati Press, 1933.
Mukhamattadīpanī-porāṇaṭīkā = Saṁpyaṅ-ṭīkā-pāṭh, Yangon: Kavi Myat Hman Press, 1914.
Mukhamattasāra, forthcoming edition by A. Ruiz-Falqués, Pune: Pune Indological Series.
Nirukta, see Swarup 1984.
Padasādhana-ṭīkā, see Dhirananda and Vachissara 1908.
Rūpasiddhi = Padarūpasiddhi, Jambu Meiswe Pitaka Press, Yangon, 1940.
Saddanīti, see Smith 1928–66.
Taittirīya Prātiśākhya, see Whitney 1868.

SECONDARY SOURCES

Burnell, A.C., 1875, On the Aindra School of Sanskrit Grammarians, Mangalore.
Cardona, G., 1969, “Studies in Indian Grammarians: I. The Method of Description Reflected in the Śivasūtras”, Transactions of the American Philosophical Society, 59(1): 3-48.
Cone, M., 2001-20, A Dictionary of Pāli, vols I–III, Oxford/Bristol: Pali Text Society.
Crosby, K., 2004, “The Origin of Pāli as a Language Name in Medieval Theravāda Literature”, Journal of the Centre for Buddhist Studies, 2: 70-116.
D’Alwis, J., 1863, An Introduction to Kachchayana’s Grammar of the Pali Language, Colombo.
Deokar, M., 2008, Technical Terms and Technique of the Pali and the Sanskrit Grammars, Varanasi: Central Institute of Higher Tibetan Studies.
Deshpande, M., 1998, “Evolution of the Notion of Authority of the Pāṇinian Tradition”, Histoire Epistémologie Langage, 20(1): 5-28.
Deshpande, M., 1997, “Who Inspired Pāṇini?: Reconstructing the Hindu and Buddhist Counter-Claims”, Journal of the American Oriental Society, 117(3): 444-65.
Dhirananda, S.S. and S. Vachissara (eds), 1908, Buddhippasadani: A Commentary on Padasadhana, Grammar of the Pali Language by Sangharaja Sri Rahula, Principal of the College Srisanghabodhi-Srivijayabahu Parivena, Totagamuwa, Colombo: Vidyasagara Press.
Dwivedi, J.P. and S. Kumar, 2004, A Comparative and Critical Study of Kātantra and Kaccāyana Grammars = Kātantra Tathā Kaccāyana Vyākaraṇa kā Samīkṣātmaka Adhyayana, Varanasi: Central Institute of Higher Tibetan Studies.
Eggeling, J., 1874, The Kātantra with the Commentary of Durgasiṁha, Calcutta: Royal Asiatic Society.
Franke, R.O., 1902, Geschichte und Kritik der Einheimischen Pāli-Grammatik und-Lexicographie, Strassburg: K.J. Trübner.
Gornall, A.M., 2014, “How Many Sounds Are in Pāli?: Schism, Identity and Ritual in Theravāda Saṅgha”, Journal of Indian Philosophy, 42: 511-50.
Gornall, A. and A. Gunasena, 2018, “A History of the Pali Grammatical Traditions of South and Southeast Asia by Vaskaḍuvē Subhūti (1876), Part One: The Kaccāyana-Vyākaraṇa, Its Commentaries and Major Handbooks”, Journal of the Pali Text Society, 33: 1-53.
Joshi S.D. and J.A.F. Roodbergen, 1986, Patañjali’s Vyākaraṇa-Mahābhāṣya: Paspaśāhṇika, Introduction, Text, Translation and Notes, Pune: University of Poona.
Katre, S.M., 1987, Aṣṭādhyāyī of Pāṇini, Austin: University of Texas Press.
Kielhorn, F., 1985, The Vyākaraṇa-Mahābhāṣya of Patañjali, revised by K.V. Abhyankar and Furnished with Additional readings, References and Select Critical Notes, vol. 1: Adhyāyas I, II, 4th edn, Poona: Bhandarkar Oriental Research Institute.
Lüders, H., 1930, Kātantra und Kaumāralāta, Berlin: Verlag der Akademie der Wissenschaften. [Republished in H. Lüders, 1940, Philological Indica. Ausgewalte Kleine Schriften von Heinrich Lüders, pp. 659-720, Göttingen: Vandernhoeck & Ruprecht.]
Malai, P.Th, 1997, “Kaccāyana-Vyākaraṇa: A Critical Study”, Unpublished PhD thesis, Department of Sanskrit and Prakrit Languages, University of Pune.
Mallinson, J. (tr.), 2007, The Ocean of Rivers of Story by Somadeva, New York: New York University Press/JJC Foundation.
Mason, F., 1868, A Pali Grammar on the Basis of Kachchayano with Chrestomathy and Vocabulary, Toungoo: Institute Press.
Pind, O.H., 1996, “Saddavimala 12.1-11 and Its Mūlasārvāstivādin Origin”, in Saddavimala – La Pureté Par Les Mots, ed. F. Bizot and F. Lagirarde, pp. 67-72, Paris: EFEO, 67-72.
———, 2012, “Pāli Grammar and Grammarians from Buddhaghosa to Vajirabuddhi: A Survey”, Journal of the Pali Text Society, 31: 57-124.
———. (ed.), 2013, Kaccāyana and Kaccāyana-Vutti, Bristol: Pali Text Society.
Ruiz-Falqués, A., 2014, “Two Treasures of Pāli Literature from the U Pho Thi Library in Thaton: The Saddanīti-Ṭīkā and the Mukhamattasāra”, in Philosophica Asiatica Monograph Series 1: Manuscripts in the U Pho Thi Library, Saddhammajotika Monastery, Thaton, Myanmar, ed. William Pruitt, Sunao Kasamatsu, Aleix Ruiz-Falqués, Yumi Ousaka, Tokyo.
———, 2017, “The Role of Pāli Grammar in Burmese Buddhism”, Journal of Burma Studies, 21(1): 1-96.
———, A., 2019, “Purifying the Pātimokkha: Pali Grammar and Buddhist Law in 17th-century Haṁsāvatī”, Buddhism, Law and Society, 4: 93-128.
Saini, R.S. (ed.), 1987, Kātantravyākaraṇa, Delhi: Bharatiya Vidya Prakashan.
Sarup, L., 1984, The Nighaṇṭu and the Nirukta, Delhi: Motilal Banarsidass.
Scharfe, H., 1977, Grammatical Literature: A History of Indian Literature, vol. 5, fasc. 2, ed. Jan Gonda, Wiesbaden: Harrassowitz.
Senart, Émile, 1871, Kaccāyana et la Littérature Grammaticale du Pāli, Paris: À L’imprimerie Nationale.
Smith, H. (ed.), 1928-66, Saddanīti: La Grammaire Palie D’aggavaṁsa, vols I-V, Lund: Gleerup.
von Hinüber, O., 1994, “Buddhist Law and the Phonetics of Pāli”, in Selected Papers on Pāli Studies, pp. 198-232, Oxford: Pali Text Society.
Watanabe, Y., 2019, “Is Kacc: attho akkharasaññāto a pubbavākya?”, Journal of Indian and Buddhist Studies, 67(3): 10-92.
Whitney, W.D., 1868, “The Tâittirîya-Prâtiçâkhya, with Its Commentary, the Tribhâshyaratna: Text, Translation, and Notes”, Journal of the American Oriental Society, 9: 1-469.

This passage is discussed in Watanabe 2019: 49, see also Mason 1868: 3 and Malai 1997: 92. ↩
The term suttakāra appears in Mmd 12.26; suttakāraka, with the same meaning, appears in Mmd 448.29. ↩
Mmd 8.12-17: keci pana attho akkharasaññāto ti iminā va saññā vihitā | idaṁ [i.e. Kacc 2] pana tabbivaraṇam iti vadanti | taṁ na gahetabbaṁ | evañ hi sati attho akkharasaññāto ti suttavuttiyaṁ yeva akkhara icc anena kv attho | akkharāpādayo ekacattālīsan ti vattabbaṁ bhaveyya | idha ca na vattabbaṁ | akkhara icc anena kv attho | attho akkharasaññāto ti na vuttañ ca tattha, vuttañ c ’eha | ↩
Mmd 14.26: tesu pan’ idaṁ saññāsuttan ti parihāro. ↩
Mmd 14.27ff. ↩
Mmd-pṭ 46.1-16: nanv akkharaṁ padhānaṁ akkhare saty atthādhigamassa sambhavato ti | n ’evaṁ | attho va padhāno | asaty ābhidheyye akkharanyāsāsambhavato | vuttañ ca: avijjamānake atthe na taṁ sakkā thapetave | vijjamāne pan’ atthamhi sakkā akkhaṁ thapetave ti || akkhan ti akkharaṁ | aparam pi vuttaṁ: khaggo va chindati lataṁ na ca kosako ti | āpo va kaḍḍhati tasaṁ na ca bhājanāni || sīho va jayyati ripuṁ na guhā ti mantā | thero niruttividhim ādi avoca atthaṁ || cintāmaṇī manaharo na pilotikāni | vatthaṁ vibhūsati naraṁ na ca peṭakāni || annaṁ jighacchavinodaṁ na ca ukkhalī ti | thero niruttividhim ādi avoca atthaṁ || paṭisambhidānam etāsaṁ catunnañ cādisambhavā | attho so gahito tena jinavuttānusārinā ti ca || ↩
I am currently editing this text. So far it has only been found in palm-leaf manuscripts, cf. Ruiz-Falqués 2014. ↩
Mmd-sāra §§14–15: pubbavākyan t ’idaṁ suttaṁ vadant ’ācariyā pare | suttānam ānurūpena paribhāsā ti no mati || pubbattā pubbavākyañ ca paribhāsāya bhāsato | paribhāsā ti suttan taṁ vuttaṁ sandhivinicchaye || ↩
Cf. von Hinüber 1994; Crosby 2004: 78: “By the time of the commentaries [i.e. c.fifth century ce], the attitude towards language held by preservers of the Vinaya differed from the attitude held by the preservers of the Sutta Piṭaka. The latter emphasized the preservation of meaning, attha, while the former also emphasized the preservation of correct phonetics, vyañjana/ akkhara. This is because in Vinaya correct pronunciation, wording and word order were regarded as essential for the correct performance of the liturgy required in legal procedures, kammavācā.” For pronunciation and Pāli grammar specifically, see Gornall 2014. For an example of the Vinaya approach to speech-sounds, see Ruiz-Falqués 2019. ↩
Pds-ṭ 8.16-23: atthappadhānaṁ hi buddhavacanaṁ na vyañjanappadhānaṁ; ten’ āhu ācariyā: atthaṁ hi nātho saraṇaṁ avoca na byañjanaṁ lokahito mahesī | tasmā akatvā ratim akkharesu atthe niveseyya matiṁ mutimā ti || aññatthā pi vuttaṁ: atthānurakkhaṇatthāya vuttaṁ sabbam idam ato | attham eva guruṁ katvā gaṇhe na vyañjanaṁ vidū ti || bhagavatā pi vuttaṁ: atthappaṭisaraṇā bhikkhave bhavatha mā byañjanapaṭisaraṇā ti | “For the words of the Buddha have the meaning (attha), not the expression (vyañjana) as its primary (padhāna) [element]. That is why the masters stated: “The Lord, the great seer who benefits the world, declared meaning to be the refuge, therefore the intelligent person should cause his mind to dwell on meaning, not taking pleasure in speech-sounds.” And elsewhere it has also been stated: “All this has been stated for the sake of protecting the meaning, therefore wise ones should take the meaning only as a teacher, not the expression.” And even the Blessed one has stated: “Take refuge in the meaning, bhikkhus, do not take refuge in the expression” [cf. Sn-a Ee II.398.7-8]. ↩
Kacc-nidd 3.23-32; cf. D’Alwis 1863: xxi; Pind 1996: 68. ↩
Kacc-nidd 3.31-33: kaccāyanattherena pi bhagavato adhippāyaṁ jānitvā attho akkharasaññāto ti vākyaṁ pubbe ṭhapetvā idaṁ pakaraṇaṁ katan ti | kaccāyanena katasuttan ti pi vadanti | ↩
Kacc-vaṇṇ 7.28: eko ghaṭe paṭaṁ disvā ghaṭapaṭo ti virajjhati | ↩
Mmd 7.10-12: ayañ h’ etth’ attho | yo koci ghaṭapaṭādivacanattho so sabbo akkhareh’ eva saññāto | ↩
Kacc-vaṇṇ 8.3-4: therena thapitattā paribhāsā ti pi vuttaṁ | ↩
Cf. §2.1. See also Psd-ṭ 12.31-36: vuttaṁ hi bhagavatā: dve me bhikkhave dhammā saddhammassa ṭhitiyā asammosāya anantaradhānāya saṁvattanti, katame dve? sunikkhittaṁ ca padabyañjanaṁ attho ca sunikkhitto, sunikkhittassa bhikkhave padabyañjanassa attho pi sunayo hoti. ime kho bhikkhave dve dhammā saddhammassa ṭhitiyā asammosāya anantaradhanāya saṁvattantī ti | “And it has been stated by the Blessed One: «These two dhammas, bhikkhus, result in the preservation of the True Dhamma, in its not falling into oblivion, in its non-disappearance. Which two? An expression (padabyañjana) that has been correctly laid down and a meaning (attha) that has been correctly laid down. When the expression is correctly laid down, bhikkhus, the meaning, too, becomes easy to infer. These are the two dhammas, bhikkhus, that result in the preservation of the True Dhamma, in its not failing into oblivion, in its non-disappearance»,” see Gornall 2014: 513-14. ↩
In the case of Moggallāna, it is understandable, because no such a rule appears in the Cāndravyākaraṇa. ↩
- | * II.147.20: dunnikkhittassa … padabyañjanassa attho pi dunnayo hoti | (if the expression of a word has been incorrectly laid down, its meaning, too, is difficult to infer); cf. Pind 1996: 69; 2013: 1 n. 1; Gornall 2014: 513.
↩
I follow the denomination of Burnell (1875: 1-37), which is still useful in order to understand the “philogenetic” relationship between Kacc and other grammatical treatises. ↩
The technical terminology of Yāska’s Nirukta has to be considered as belonging to the Aindra system. ↩
Some rules in Kacc are closer to the Taittirīya Prātiśākhya than they are to Pāṇini or the Kātantra, e.g. Taittirīya Prātiśākhya I.6: śeṣo vyañjanāni and Kacc 6: sesā byañjanā. For an example of a Buddhist Sanskrit grammar based on, or following the same model as, the Kātantra, see Lüders 1930. ↩
The words samaññā and saññā are clearly synonyms in Mukhamattadīpanī 19.23-26: attano samaññaṁ vatvā parasamaññāya pacchā vattabbato ih’ edaṁ vuttaṁ. tividhā hi saññā anvatthasaññā sakasaññā parasaññā ti | tā pana akkharavaggaghosasaññādivasena dīpetabbā | (This is stated because, after stating one’s (his?) own samaññā, the samaññā of others should be stated next (pacchā). Indeed, saññā (!) is of three types: expressing its meaning, one’s own saññā and a saññā of others. As to these (tā pana) [saññās of others,] they are to be illustrated by means of terms such as akkhara (speech-sound), vagga (consonantal group), ghosa (aspirate) and so on.) ↩
Mmd 10.2-5: atha vā akkharāpādayo ti vutte yeva akārādīnaṁ gahaṇe sati pi ekacattālīsan ti kasmā vuttan ti codanā. idha suttantopakārā ekacattālīsaṁ | bahiddhā pana aññe pi akkharā santī ti ñāpanatthan ti parihāro | – tr. Alastair Gornall (2014: 528) ↩
On the original structure of Kacc in four books, see Pind 1996: 71. ↩
Kacc-v ad Kacc 1: sabbavacanānam attho akkhareh’ eva saññāyate | akkharavipattiyaṁ hi atthassa dunnayatā hoti, tasmā akkharakosallaṁ bahūpakāraṁ suttantesu | Note that the vutti only mentions the suttantas, i.e. the Sutta Piṭaka. It does not mention the Vinaya, nor the Abhidhamma. See also Mmd 6.15ff. The Rūpasiddhi states that the first aphorism is in order to state the “purpose” (payojana) and the “subject” (abhidheyya) of grammar: tattha jinasāsanādhigamassa akkharakosallamūlakattā taṁ sampādetabban ti dassetuṁ abhidheyyappayojanavākyam idam uccate. ↩
Both are probably later than Kacc, cf. Pind 2012: 71 n. 76. ↩
Franke 1902: 2; Pind 2012: 71. ↩
Psd-ṭ 12.9-10: badhantabodhisattamahāthero naravaravacanopakārāni cattālīs’ akkharānī ti āha | ↩
Psd-ṭ 6.21-23: vuttaṁ hi: suttesv eva hi taṁ sabbaṁ yaṁ vuttyaṁ [read v.l. vuttaṁ] pañcikāya yaṁ suttam uppatti atthānaṁ sabbaṁ sutte patiṭṭhitan ti | ↩
Cf. MBh I.11.15: atha vyākaraṇam iti asya śabdasya kaḥ padārthaḥ? sūtram | ↩
Cf. Joshi and Roodbergen 1986: 184 n. 768: “Utsūtram. The compound is formed in the sense of sūtrād udgatam (what has gone beyond the rules) by Saunāgavārttika 9 on P. 2.2.18.” ↩
There is a “pun” (silesa) here with the word sutta “thread/ treatise” and the verb ādiyati “to seize, to grasp” (cf. DOP s.v. ādiyati), literally, like grasping a flower, or figuratively, like grasping the meaning of a word. ↩
Cf. MBh I.12.15: [Vārttika §14] lakṣyalakṣaṇe vyākaraṇam | ↩
Pds-ṭ 6.24-30 [see also Pind 2012: 71 n. 76]: vuttañ c’ ācariyasabbaguṇākarena māgadhikasaddakalikāyaṁ suttaṁ vyākaraṇam. ussuttābhidhānen’ ādeyattā suttam eva vyākaraṇasarīraṁ | saddavippaṭipattiyaṁ pavattā yaṁ [read pavattāyaṁ] suttaṁ vinā agathitapuppharāsimhi viya anvākhyāne [Pind atthākhyāne] karīyamāne yāva suttaṁ na dassitaṁ, tāvad anuppādeyaṁ vacanaṁ bhavati | tasmā kāraṇā ti attho | lakkhiyalakkhaṇāni vyākaraṇam añño [em. ti añño] | ubhayāni samuditāni vyākaraṇaṁ nāma ti kāccāyano maññate iti | Cf. MBh I.11.15-12.27. The problem discussed in the Mahābhāṣya is whether vyākaraṇa is śabda, i.e. language, or sūtra, the grammatical rule. Patañjali is of the opinion that vyākaraṇa is the sūtra only. Kātyāyana, however, postulates that vyākaraṇa is the sum of both words and grammatical rules; cf. Scharfe 1977: 83; Joshi and Roodbergen 1986: 158ff. I thank Alastair Gornall and Petra Kieffer-Pülz for their crucial assistance in clarifying this passage. ↩
Psd-ṭ 12.11-12: sabbaguṇākaramahāthero siddhakkamādādayo vaṇṇākkharā titālisā ti āha. I follow Gornall 2014: 532-33; see also Pind 2012: 71 n. 76. ↩
Mallinson’s translation, cf. Kathāsaritsāgara 1.7.4: ito rājan nirāhāro maunastho ’haṁ tadā gataḥ | ↩
Kathāsaritsāgara 1.7.6-14, tr. by Mallinson (2007: 163-65). The vehicle of Kumāra is the kalāpin (peacock). The following is the Sanskrit text followed by Mallinson: uttiṣṭha putra sarvaṁ te saṁpatsyata iti sphuṭam śaktihastaḥ pumān etya jāne mām abravīt tadā | tenāham amr̥tāsārasaṁsikta iva tatkṣanam prabuddhaḥ kṣutpipāsādihīnaḥ svastha ivābhavam | atha devasya nikaṭaṁ prāpya bhaktibharākulaḥ snātvā garbhagr̥haṁ tasya praviṣṭo ’bhūvam unmanāḥ | tato ’ntaḥ prabhuṇā tena skandena mama darśanam dattaṁ tataḥ praviṣṭā me mukhe mūrtā sarasvatī | athāsau bhagavān sākṣāt ṣaḍbhir ānanapaṅkajaiḥ siddho varṇasamāmnāya iti sūtram udairayat | tac chrutvāiva manuṣyatvasulabhāc cāpalād bata uttaraṁ sūtram abhūhya svayam eva mayoditam | athābravīt sa devo māṁ nāvadiṣyaḥ svayaṁ yadi abhaviṣyad idaṁ śāstraṁ pāṇinīyopamardakam | adhunā svalpatantratvāt kātantrākhyaṁ bhaviṣyati madvāhanakalāpasya nāmnā kālāpakaṁ tathā | ity uktvā śabdaśāstraṁ tat prakāśyābhinavaṁ laghu | ↩
Saini 1987: viii: “When Kārttikeya uttered the first sūtra: siddho varṇasamāmnāyaḥ of the proposed system and was about to utter the second sūtra: Śarvavarman himself spoke the sūtra tatra caturdaśau svarāḥ.” ↩
Cf. Deshpande 1998; Cardona 1969. ↩