Pāṇini1 composed the Aṣṭādhyāyī around 350 BC2 in North-Western South Asia.3 The Aṣṭādhyāyī is a samāhāra ‘collection’ of aṣṭa(n) ‘eight’ adhyāyas ‘books’, hence the name Aṣṭa-adhyāy(a)-ī. Each book of the Aṣṭādhyāyī has four pādas ‘chapters’ that are made up of sūtras ‘rules’. In all, the Aṣṭādhyāyī comprises about 4000 rules. The Aṣṭādhyāyī is a comprehensive grammar of the Sanskrit language as known to its author Pāṇini. It stands out for doing more than merely describing its object language: the Aṣṭādhyāyī is a full-fledged machine which helps construct grammatically correct Sanskrit words and sentences through a step-by-step derivation4 process. In the Aṣṭādhyāyī, Pāṇini does not give us a general introduction to his work, nor does he discuss the theoretical principles that have been used to construct his sūtras. He conveys whatever has to be said, through his sūtras alone.
The first two books are mainly composed of saṁjñā sūtras ‘definition rules’ and paribhāṣā sūtras ‘metarules’5. The remaining books mainly consist of vidhi sūtras ‘operational rules’. Books three to five teach the addition of both inflectional and derivational affixes to bases. Book three teaches the addition of various affixes to verbal roots and stems, and books four and five teach the addition of different affixes to nominal stems.
8
Books six, seven and eight teach various morpho-phonological operations that should be performed on both bases and affixes. Different kinds of rules from multiple books are required to derive a word using Pāṇini’s method.
To truly understand the Aṣṭādhyāyī, one needs to familiarize oneself with the methodology used by Pāṇini to compose and arrange rules in his work. Pāṇini’s style is not entirely self evident, and one faces challenges at multiple levels when attempting to unravel the enigma that is the Aṣṭādhyāyī. Firstly, it is not easy to determine the exact meanings of Pāṇini’s rules because the sūtra style in which they are composed is very concise and compact. Much information is often packed into a few words, thereby making it considerably difficult to comprehend their exact purport. Consider 6.1.9 sanyaṅoḥ, which teaches that a verbal base6, which has not undergone reduplication, undergoes reduplication in the presence of affixes saN6 and yaṄ, the desiderative and intensive markers, respectively.7 The question about whether sanyaṅoḥ is a genitive dual or a locative dual is a crucial one, and has important implications for how we conceptualize prakriyā ‘the (derivational) procedure’.8
Secondly, to make sense of any given rule, it is essential to take into account the contents of preceding rules. This is because Pāṇini uses a device called anuvr̥tti ‘continuation into the following rules’ to economically express his observations: to understand the complete and correct meaning of a rule, certain words from preceding sūtras may need to be borrowed into that rule by anuvr̥tti. But there is no universal convention as to which terms are supposed to or can become anuvr̥tta ‘continued’ into a certain rule. For example, consider 1.1.33 prathamacaramatayālpārdhakatipayanemāś ca, which teaches that certain words are called sarvanāma. But it is difficult to determine whether or not the words from the previous rule 1.1.32 vibhāṣā jasi should be continued into this rule.
9
Books six, seven and eight teach various morpho-phonological operations that should be performed on both bases and this would restrict 1.1.33 only to those cases where these stems are followed by the nominative plural affix Jas, and would also make 1.1.33 optional.9
Thirdly, even after the meaning of the rule has been understood, it does not become patently obvious how to use it. This is because Pāṇini’s rules are placed together on the basis of topical and functional categories, and not according to the derivations in which they participate.10 Thus, one cannot easily ascertain the order in which rules apply or select the step at which they become applicable. For example, consider the rule 3.1.33 syatāsī lr̥luṭoḥ, which teaches that the affixes sya and tāsI should be added to the left of LR̥ (LR̥Ṭ and LR̥Ṅ) and LUṬ respectively. But the question that has troubled both traditional and modern scholars is: should and can this rule apply before the lakāras are replaced with finite verb endings (3.4.77 lasya; 3.4.78 tip-tas jhi…11)?[^13]
Fourthly, after one has come to a conclusion about where to apply a given rule, one is often faced with situations in which two rules become applicable at the same step. In many such cases, one rule blocks the other, or both rules block each other. This is called ‘rule conflict’. According to the tradition, a metarule taught by Pāṇini, namely 1.4.2 vipratiṣedhe paraṁ kāryam addresses this issue. However, it seems unable to give the right answer when applied to certain cases of conflict.
We can conclude that the Aṣṭādhyāyī is a very sophisticated grammar, and that to operate its grammatical machine, we have to understand it at multiple levels. What would an early grammarian or linguist have done in order to interpret the Aṣṭādhyāyī independently? With negligible access to any commentary on the text, and with limited or no guidance of a teacher well-versed in the Aṣṭādhyāyī, a scholar would have taken notes for himself in order to comprehend, analyse and corroborate the teachings of the Aṣṭādhyāyī. He would have started by paraphrasing the contents of the Aṣṭādhyāyī to establish what they exactly mean, both independently and in the context of the preceding rules.
10
To ensure that he had understood such a complex grammar correctly, or to confirm that the grammar accurately describes the structure of the language, a scholar would have tried to verify the validity and accuracy of different rules against spoken language or attested literature. He would have gradually developed his own ideas about where rules should apply, and how derivations should proceed. He would have noticed how rules interact amongst themselves and would have come up with ways to classify and deal with such interactions. He would also have suggested certain changes to these rules to make them more precise, to help them better characterize their object language and/or to help them function more consistently with other rules within the Pāṇinian system.
This is presumably what happened in the Indian grammatical tradition when Kātyāyana understood the meanings and functions of Pāṇinian rules on the basis of his independent study of the Aṣṭādhyāyī.12 Then as a teacher, he also taught them to his pupils using his notes on the Aṣṭādhyāyī as pedagogical aid. His students taught the Aṣṭādhyāyī to their students using Kātyāyana’s work and also commented on Kātyāyana’s writings, thereby sharing their own opinions, interpretations and analyses with their students and readers. Successive generations participated in this process of knowledge processing, production and transmission, thereby giving birth to the Pāṇinian grammatical tradition.
The texts of the Pāṇinian grammatical tradition have played a dominant role in influencing and shaping our understanding of, and opinions about the Aṣṭādhyāyī. They also give us significant insights into the evolution of different ideas in the Pāṇinian tradition. Below I introduce the texts that I shall refer to in the rest of the thesis and briefly discuss the history of the Pāṇinian tradition with special reference to metarules.
-
There are many disagreements about the dates, and what I mention here are the dates agreed upon by much recent scholarship. ↩︎
-
Cardona 1976: 267-268. ↩︎
-
I say ‘composed’ and not ‘wrote’ because scholars disagree on whether he used the aid of writing to create his grammar. In recent times, Vergiani (2020) has present strong arguments in favour of the proposition that Pāṇini did use written means to put together his magnum opus. Writing or not, it is known that, just as happened with the Vedas, the Aṣṭādhyāyī too was orally transmitted from one generation to the next. ↩︎
-
In the modern literature on the Pāṇinian grammatical tradition, it is customary to use the verb ‘to derive’ and its derivatives (e.g., derivation) to simply mean ‘to construct’. The verb ‘to derive’ is used in the context of not only derivational but also inflectional morphology. I shall abide by this convention in this thesis. ↩︎
-
Metarules teach us how rules should be interpreted, how certain operations should be undertaken, and how rules interact with one another. ↩︎
-
In this thesis, I use capital letters in Pāṇinian morphemes to represent itsaṁjñakas (taught in 1.3.2 upadeśe’j anunāsika it and following sūtras). Such its (commonly called anubandhas in post-Pāṇinian grammatical literature) are used to mark certain properties of the item to which they are added, and are not actually part of the item. Their unconditional deletion is taught by 1.3.9 tasya lopaḥ. ↩︎
-
Note that, in this thesis, I have used English translations of Pāṇini’s rules by Sharma (1987-2003) and Katre (1987), for many but not all rules. I have taken the liberty to edit their translations as required. For the remaining rules, I have presented my own translations. ↩︎
-
Cardona 1997: xvii, Kiparsky 1982: 85-86. ↩︎
-
Bloomfield 1927: 61-70. ↩︎
-
Besides, it is not possible to arrange rules on the basis of the derivations in which they participate because most rules participate in umpteen different derivations. ↩︎
-
Tip-tas-jhi-sip-thas-tha-mib-vas-mas-t(a)-ātāṁ-jha-thās-āthāṁ-dhvam-iḍ-vahi-mahiṅ. 13 Roodbergen 1991: 293-299. ↩︎
-
I think that there was a break in the transmission of the Aṣṭādhyāyī between Pāṇini and Kātyāyana, since Kātyāyana seems to be in the process of understanding the Aṣṭādhyāyī without much help from anyone else. I shall furnish evidence to support this statement in chapter 6. ↩︎