Kārake u prvoj pālijskoj gramatici i njezinu komentaru

Abstract

This dissertation is a result of a research into kārakas as they are presented in the third kappa of the Kaccāyana grammar of the Pāli language and in the corresponding third textual portion of the commentarial grammatical work known as Kaccāyanavutti. In addition, it is also a result of a research into the general sigificance of kārakas not only in Pāli but also, to a lesser extent, in Sanskrit. It can be said that a certain dualistic complexity typifies the objects and goals of the research. Firstly, it is the duality of languages and their grammatical structures that characterises the primary objects of the research (Pāli and Sanskrit). Secondly, it is the duality of grammatical traditions that characterises the secondary objects of the research (non-orthodox Buddhist/Jainist tradition and orthodox Hinduist tradition). Thirdly, it is the duality of scientific approaches that characterises the preliminary goals of the research (Western and Asian). Fourthly, it is the duality of pre-linguistic assumptions with regard to temporality that characterises the final goals of the research (cyclic temporality and linear temporality). The dissertation is divided into two parts. The first part deals with Sanskrit grammar only. Actually, it is a general study of Sanskrit grammar with special reference to two different Sanskrit grammatical systems (Pāṇini's and Śarvavarman's) and the roles of kārakas played within the domain of their innermost layers. On the other side, the second part comprises a ground plan of the first grammar of the Pāli language as well as a study concerning the peculiarities of its system of kārakas. At the beginning of the first part one can find a short introduction that offers a historical sketch of Pāṇini's linguistic predecesors and their ideas. In fact, this first rubric is preparatory in some regard and is to be seen as the first chapter of the dissertation. The chapter is further subdivided into seven sections. The first three sections are non-thematically comprehensive and synthetical to a certain extent. On the other hand, the remaining sections are much narrower in their scope and somewhat analytical. In brief, the first section is dedicated to Pāṇini and his grammatical system of Sanskrit (bhāṣā). The second section touches upon Kātyāyana's vārttikas and their relation to Patañjali's commentarial work entitled Mahābhāṣya. The third section is concerned with Śarvavarman's grammatical system of Sanskrit and its commentarial traditions. The fourth section introduces the topic of kārakas and their functions within the context of the Aṣṭādhyāyī grammar and its first chapter in particular. The fifth section depicts Kātyāyana's vārttikas pertaining to the set of kārakas as well as Patañjali's commentarial attitude towards Kātyāyana's and Pāṇini's treatment of kārakas. The sixth section lays bare Śarvavarman's unique system of kārakas within the framework of Durgasiṃha's grammatical commentary. Finally, the seventh section gives a condensed sketch containing the systems of kārakas as understood and interpreted by almost all the other significant Sanskrit grammarians (e.g. Bhartṛhari, Jayāditya, Vāmana, Nāgeśa etc. ) Although many Sanskrit grammatical works have hitherto been presented in the sections of the chapter, this part of the research into the general significance of kārakas nevertheless has the strongest emphasis on the Aṣṭādhyāyī grammar and the Kātantra grammar. For it is fundamentally presupposed that these two grammatical works are of very special concern, when it comes to the first grammar of the Pāli language. The second part, which represents the core of the dissertation, discloses the Kaccāyana grammar of the Pāli language in light of the aforementioned Sanskrit grammatical works and their influences on the Pāli system of syntactic categories (kārakas). A research into the Pāli system of syntactic categories has been carried out in a multi-featured manner. It means that several different viewpoints have been taken into account and categorised as pertaining to three significant perspectives. The first perspective is philological. The second one is linguistic. The last one is semiotic. Philological perspective on the Kaccāyana grammar and its chapter on kārakas The sutta 3.30 (302) of the Kaccāyana grammar of the Pāli language contains a segment of the Aṣṭādhyāyī grammar (A 1.4.52). The segment is interwoven with a fragment taken from Jayāditya's and Vāmana's Kāśikāvṛtti. The fragment is as follows: “bhuja-paṭha“. It is what Jayāditya and Vāmana added to the sūtra 1.4.52 of the Aṣṭādhyāyī grammar while commenting on it. The fragment might have been interpolated from the Kāśikāvṛtti to the Kaccāyana grammar between the 7th century A.D. and the beginning of the 9th century. The terminus ante quem non is the year 650 A.D. For it was the year around which the aforementioned Sanskrit commentary (Kāśikāvṛtti) was brought to complition. It can be concluded that the sutta 3.30 (302) of the Kaccāyana grammar is of importance to anyone tackling the problem of how to date the Kaccāyana grammar. It goes without saying that the sutta 3.30 (302) belongs to the chapter on kārakas. For almost over a century, the Kaccāyana grammar of the Pāli language has been seen as an adaptation of the Kātantra grammar in Europe. Charpentier (1929) and recently Pind(1989) have opined that the Kaccāyana grammar was patterned after one Sanskrit grammatical work only. They have tacitly implied that something like an imitation of Kātantra had undergone the process of changing to suit the conditions of the particular enviroment of Pāli and its Buddhist background before it surfaced as a well-formed system of grammatical suttas. Their opinion is called into question in the dissertation. As a matter of fact, the Kaccāyanabyākaraṇa is not a miniature version of the Kātantra grammar expressed in the Pāli language. Based upon the Aṣṭādhyāyī grammar and the Kātantra grammar, it is rather a hybrid work that contains very few traces of originality. Systematically, it is partly modelled after the Aṣṭādhyāyī grammar. Its body of suttas is partially axiomatised with the first sutta of the Kaccāyanabyākaraṇa being seen as the most important axiom of the system (attho akkharasaññāto). This method must have been taken from Pāṇini. On the other side, terminologically, the Kaccāyana grammar of the Pāli language relies on the Kātantra grammar and its specific technicalities to a very large extent. It is also the Kātantra grammar that served as the basis of general grammatical definitions, after which so many rules of the Kaccāyana grammar were patterned. Not surprisingly, Saṅghānandi's commentarial work on the Kaccāyana grammar, also known as Kaccāyanavutti, is far from being an adaptation of Durgasiṃha's commentary on the Kātantra grammar. Actually, it is a complex grammatical commentary that is in some of its aspects closer to Jayāditya's and Vāmana's Kāśikāvṛtti than to Durgasiṃha's commentary. Linguistic perspective on the kārakas and case endings of the Kaccāyana grammar It can be interpreted as a perspective that is bound up with the inquiry into the nature of kārakas carried out by Blake B. J. (2004). His argument is that kārakas are not like the deep cases and semantic roles of modern theories because they do not remain constant under paraphrase. In a manner of speaking, this phase of the research into the sigificance of kārakas could be understood as a continuation of the aforementioned inquiry. To make the inquiry more sucessful, the two different sets of concepts have been taken into account. The first one comes from the linguistic tradition of south Asia. It means that the kārakas of the Kaccāyana grammar have been fully presented and numerically distributed with the case endings of Pāli (vibhattis) being seen as their counterparts within the dualistic conceptual framework consisting of sadda (sk.śabda) and attha (sk.artha). The second one has been taken from the linguistic tradition of the West. The case system of Pāli have been contrasted with the similar case system of Sanskrit within the context of the concept of “abstract base“ which is actually another term for the concept of deep structure. According to Householder F.W.(1981) Apollonius Dyscolus was the inventor of the concept of “abstract base“ (or deep structure). The Greek scholar invented the concept during the 2nd century A.D., but it feel into oblivion as time passed. Fortunately, it was rediscovered and has even become popular with some western generative linguists since the beginning of the second half of the 20th century right up to the present moment. As a matter of fact, the Kaccāyanabyākaraṇa is a generative grammar. What is more, it is generative in the fullest sense of the idea of generativity. But it differs from any western type of generative grammars mainly in that it is based upon a unique and indivisible semantic/syntactic core. By no means can its syntax be autonomous. For it is totally inseparable from its semantic component. Hence, it follows that the Kaccāyanabyākaraṇa is not a phrase structure grammar in the sense that the syntactic elements of the NPs can become the semantic arguments of any verbs belonging to the VPs. Nor does it appear to be a dependency grammar in the sense of Tesnière's grammatical model. Generally speaking, the vyākaraṇa types of grammars differ significantly from any modern case grammar types despite the fact that they have something in common with some of these western case grammar types. Namely, the vyākaraṇa types of grammars are all rooted in generativity. The Kaccāyanabyākaraṇa is no exception to that general rule. Although it is generative, it bears no similarity to any modern case grammar whatsoever. As to the cases of the Kaccāyana grammar, they are as follows: pathamā (the nominative case), dutiyā (the accusative case), the instrumental case (tatiyā), the dative case (catuthā), the ablative case (pañcamā), the genitive case (chaṭṭhā), the locative case (sattamā). On the other side, the kārakas of the Kaccāyana grammar are as follows: kattu (doer), kamma (deed), karaṇa (instrument), apādāna (ablation), sampadāna (bestowal), okāsa (locational occasion). To discover the deep structure of any Pāli sentence and its case endings, it is sufficient for one to translate it from Pāli into Sanskrit. The result of the act of translating will unveil what lies behind the given Pāli sentence and its case endings in terms of deep structure (e.g. the sentence “na tesaṃ koci sarati.“ becomes “na tān kaścit smarati.“ if translated into Sanskrit, disclosing the genitive case of Pāli (chaṭṭhā/ṣaṣṭhaḥ) as a surface element and the accusative case of Sanskrit (dutiyā/dvitīyaḥ) as a deep element). It is clear that the level of kārakas has not been reached yet. For kārakas operate on the level that is much deeper than any strata of the deep structure of the Pāli language (e.g. if any sentence has the genitive case of Pāli as a surface element and the accusative case as a deep element, it does not follow that the kāraka known as kamma (Sk.karman) lies hidden somewhere within the deep structure of the given sentence, no matter how strongly the accusative case might be bound up with the concept of kamma). In other words, the case endings of the Pāli language do belong to the surface structure of Pāli in some ancient sense, but not in the sense of any modern generative case grammars and their theoretical frameworks. Neither the case system elements of the Pāli language nor their peculiarities can be forced into the mould of Fillmore's model, if their grammatical nature is to be truly revealed. Neither can the vibhattis of the Pāli language be compared to the surface structure cases of Fillmore's case grammar model, nor do the kārakas of the Kaccāyanabyākaraṇa bear any similarity to the deep structure case concepts of his model. Actually, kārakas belong to some other region. It can be shown where they fundamentally belong to, if one takes steps leading towards the sphere of semiotics. Semiotic perspective on the Kaccāyana grammar and its kāraka system According to Al-George S.(1968), old Indian linguistics has taken from a universal semiotics a system of categories which is applied to a particular semiotic system. Surely, it is Sanskrit that is to be seen and interpreted as that particular semiotic system. The abstract form of the Sanskrit language is to be conceived in the same way the background of mythical reality was conceived in ancient India. In addition to that, the phenomenal reality is to be understood and made valid by its analogy with the ultra-phenomenal reality, more real than real, according to the Rumanian scholar. As for the concept of kāraka, he has focused his attention on the function of that concept within the Pāṇinian system of Sanskrit grammar only. More precisely, his article limits itself to a segment of the concept of kāraka. Of course, it is the segment that deals with the linguistic accuracy of the concept of kāraka against its extra-linguistic origin. The main objective at this semiotic stage of the research into the general sigificance of kārakas is to extend the scope of Al-George' s presumption to those Sanskrit gramatical works which has been known as post-pāṇinian. Not only that. It should be extended to the Pāli language and its grammatical tradition. To put it simply, it is Pāli that is to be seen and understood as the particular semiotic system. In addition, the linguistic accuracy of the concept of kārakas of the Kaccāyana grammar is to be contrasted with its extra-linguistic correlates. In order to achieve this objective, one should depict some distinguishing features of Pāli and its grammatical tradition. They can serve as reference points in the course of the exposition of distinctions between the Vedic tradition interwoven with Sanskrit and the non-Vedic tradition which Pāli primordially belongs to. Firstly, Pāli lacks any ritualistic background. Being linked to the Theravāda branch of Buddhism to a very large extent, it could be affiliated with the categories of symbolic expression which are much more closely related to the mythical line of thinking than to the ritualistic one. Secondly, the auxiliary science of the vedāṅgas might not have been the only source of the linguistic terminology of its grammatical tradition. It is this second point that deserves special attention. To realise its importance, one should take a closer look at Buddhaghosa's linguistic terms associated with the syntax of Pāli. Buddhaghosa's syntactic categories are as follows: paccattavacana, upayogavacana, karaṇavacana, sampadānavacana, nissakkavacana, sāmivacana, bhumma vacana and bhummālapana. Although it is clear that the concept of sāmivacana is equivalent to the concept of genitive case and that the the concept of bhummālapana is almost identical to the concept of the vocative case, it nonetheless seems that the nature of the remaining six syntactic and morphological categories is far from being clear-cut. Not unexpectedly, the origin of these categories is far from being expressed in precise terms. And then some philologists have speculated about the existence of an unknown grammatical work which might have been the source of Buddhaghosa's syntactic categories. What is more, they have even established a conjecture about the unknown grammatical work which might have predated the Kaccāyana grammar. The speculation has been rejected as groundless in the dissertation and thus the Kaccāyana grammar has been recognised as the first Pāli linguistic work dealing with the grammatical rules of the Pāli language in a sophisticated and systematic manner. However, the question about the real source of Buddhaghosa's syntactic categories has arisen again: an issue requiring resolution afresh. In order to answer the question, it has been hypothesised that Buddhaghosa's syntactic categories originate from a lost linguistic tradition of south Asia. It is likely that the Vedic linguistic tradition bound up with the vedāṅgas was not the only source of the concepts and terms pertaining to the science of language. It is almost tantamount to saying that an alternative linguistic tradition might have run parallel to that of the Vedas and the vedāṅgas. It is possible, without being demonstrable, that Buddhaghosa was familiar with the lost linguistic tradition of the śrāmaṇa movement. Stated differently, he might have taken the aforementioned syntactic categories from the grammatical tradition that had no strong ties to the Vedic lore of the brāhmaṇa tradition. Though the hypothesis lacks firm evidence, it has nonetheless the potential to become a point of departure for some deliberations in the future. Namely, Buddhaghosa's syntactic categories do not seem to be exclusively related to the case endings of Pāli. They indeed have something in common with the kārakas of the first Pāli grammar. For example, they share the same references with some of the kārakas of the first Pāli grammar in spite of the fact that their modes of presentation are different. But the relationship between references and modes of presentation is overturned, when it comes to the aforementioned extra-linguistic correlates. For the semiotic correlates share the same modes of presentation with the six kārakas of the first Pāli grammar, although their references are quite different. Apropos of the representational model of syntax advocated by Al-George, the six kārakas of the Kaccāyana grammar can be seen as the indexical signs corresponding to the six spokes of the wheel of temporality. On the other side, the wheel itself functions as an iconic sign within the representational model of grammar that is interpreted in such a manner. Of course, it is an archaic picture of cyclic temporality that underlies the extra-linguistic core of the particular semiotic system known as the Pāli language. As a technical and functional unit, the term “cyclic temporality“ can be meaningfully incorporated into the semiotic context of the research into the Kaccāyana grammar and its chapter on kārakas because a semiotically similar term can be found in the technical vocabulary of Indian linguistics. Actually, the term “kārakacakka“ (Sk.“kārakacakra“) indeed belongs to the terminology of old Indian linguistics, although it is not mentioned in the suttas of the first Pāli grammar explicitly. The archaic picture of cyclic temporality does appertain to the extra-linguistic and even prelinguistic strata that make Pāli grammar possible in a semiotic sense. The same holds likewise for Sanskrit grammar. Although the vyākaraṇa types of grammars(Pāli/Sanskrit) and various western types of generative grammars are all rooted in the similar concepts of generativity, they nevertheless differ radically, when it is a question of the aforementioned extra-linguistic and prelinguistic strata of semiotics. Actually, what makes any western generative grammar types possible appears to be a differrent conceptual framework of temporality. Surely, it is much closer to the thinking pattern of which linear temporality is a basis than to that pertaining to the archaic picture of cyclic temporality that was typical of old India. Of course, some western linguists have realised that the linearity of speaking does not imply the linearity of the internal grammatical processing of speech. But practically none of them has tried to reach the extra-linguistic and pre-linguistic strata of linguistics in such a semiotic manner. After all, the theoretical no less than pre-theoretical presuppositions of the modern sciences of the West have almost always been more or less opposed to any kind of the mythical and symbolical line of thinking.