Pāli Grammatical Studies in Bengali Language: Dr. B. M. Barua’s Thoughts on Pāli Grammar – Edited by Upali Sramon

Introduction

Dr. B.M. Barua (full name Benimadhav Barua, 1888-1948, born in Bangladesh) is one of the early Indian scholars who initiated Buddhist and ancient inscriptional studies according to the research methodology followed by Western scholars like Max Muller. Among his books, A Prolegomena to Early History of Buddhism, Pre-Buddhist Philosophy of India, and Prākŗt Dharmapada (co-editor Sailendranath Mitra) are great contributions to the indological studies . He also produced many academic papers in prestigious journals, and prefaces and introductions to many books. The collection of Ceylon Lectures, exhibit Dr. Barua’s eloquence elucidating historical, philosophical, and doctrinal points. He was also an active educationist and social reformist engaged in many social and academic institutions. During his service in Kolkata University he edited the prominent journal jagajyoti for many years.

This is not Dr. Barua’s extensive study on the subject. This article (written in 1937) appears as the ‘Introduction’ in Pāli Prabeṡ of Ven. Jñānisvar Mahasthavīr. I have decided to translate it because in it I have noticed some brief but important observations of Dr. Barua on some Pāli grammatical problems. Some conclusions he has arrived at, such as even to assert the invalidity of the Sandhi rule – yavamadanataralacāgama are thought-provoking, and need to be revisited by Pāli scholars. Discussions on some of the issues that Dr. Barua has raised in the study of Pāli grammar, comparative study of Pāli and Sanskrit for example, have already been started by modern Pāli scholars. Dr. Barua mainly wrote in English; this article if written in his idiomatic English, certainly, would have been something very different, a fact for which I accept my deficiency. However, I believe the essence of the original essay has not been harmed in the following translation.

Pāli Studies in Bengali

Ven. Srimat Jñāniṡvar Mahasthavir is a prominent monk of Chittagong. He is the abbot of the Laṅkārāma Vihāra, Unainpura Village. He is knowledgeable and an active promoter of education. Throughout his lifetime he has been courageously and enthusiastically contributing for the promotion of Pāli language and literature. His compilation of the three books – ‘Pāli prabeṡ’[1], ‘Vuttoday’, and ‘Dhātokoş’ in one cover is a great contribution for the Bengali students of Pāli. Pāli prabeṡ is a pertinent compilation for Bengali readers. This grammar is relevant for all categories of readers. In this, Ven. Jñāniṡvar has made a convivial compilation accumulating and aptly situating essential aphorisms from ‘Kaccāyana-vutti’, ‘Mahārūpasiddhi’, ‘Bālāvatāra’ and etc. ancient Pāli grammars.

The ‘Vuttodaya’ of Ven. Saṁgharakkhita is the only text on Pāli prosody. In ‘Vuttodaya’, the sūtras (aphorisms) of ‘Vŗttaratnākara’[2] by Kedārnāthbhaṭṭa of 12^th century A.D. are emulated in concurrence to Pāli language. In this, the Vedic rules of prosody are not taken into consideration; but, it follows the rules of Classical Sanskrit (idaṁ vuttodayaṁ nāma lokiyacchandaṁ nissitaṁ). Publishing this text into Bengali Ven. Jñāniṡvar has done a valuable service by uniting the Vŗttaratnākara and Pāli Vuttodaya.

In the Dhātukoş Pāli roots have been compiled according to the classes (gaṇa) as bhvādi, rudhādi, divādi, etc. and arranged according to the Pāli alphabetical order (ākārādi). Compiling the Dhātukoş the author has used the Dhātumañjūsa and the Dhātumāla section of Saddanīti. Although not exactly as ‘dhātupāṭha’ of Dhātumālā the Dhātukoş answers a great need of students.

Ven. Jñāniṡvar’s is not the first attempt in composing Pāli grammar in Bengali language. In this matter Paṇḍit Nabarāj Barua is the path finder for us. With the encouragements and financial assistants of Upendralāl Barua, Paṇḍit Nabarāj Barua has published the first Pāli grammar. But, his work did not gain prominence among students. My respectable professor Dr. Satişcandra Vidyābhūşan has published the Kaccāyanavutti in Devanāgari script and thus opened the door of Pāli studies in this country.[3] Mahāmahopadhyay Prof. Vidhusekhar Bhaṭṭacaryā Mahāṡaya has attempted to compose a Pāli grammar in Bengali following modern linguistic disciplines in his ‘Pāli prakāṡ’. Mr. Lokendranāth Barua[4] has composed the first Pāli grammar in Bengali language appropriating it for students. Ven. Baṁşadīpa Mahasthabīr Mahoday (lecturer, Nālanda Vidhyābhavan) has edited and translated the Bālāvatāra into Bengali and made the approach to Pāli studies easier.

Not exactly in accordance with modern scholastic research methodology or linguistic disciplines, but, as an apposite companion for all categories of readers Pāli prabeṡ ranks after the Pāli Prakāṡ. With less labour but easy pursuit for obtaining proficiency in Pāli aphorisms I do not remember of a better text in Bengali than Pāli prabeṡ. Pāli prosodic text Vuttodaya has not yet been published in Bengali, far from being translated. The Dhātukoş is also very new for Bengali readers.

On Pāli Grammatical Studies[5]

However, I do not agree that until now a satisfactory Pāli grammar has been composed. Among the western scholars who have been studying and researching this area, names of Edward Muller and Dr. Geiger come to the forefront. Prof. Charles Duroiselle’s Pāli grammar has been accepted by students, there is no doubt. In the second edition of Bālāvatāra published from Kolkata University Prof. Srīzuta Sailendranāth Mitra has established some new arguments.

Now, what is the reason for asserting that there is no Perfect grammar composed for Pāli studies yet? Kaccāyana-vutti, and Moggallāna-vutti are the two main texts on Pāli grammar. Both the texts were composed by scholar monks of Sri Lanka. Following the aphorisms of Kaccāyanavutti, Mahārūpasiddhi, and Bālāvatāra were composed later. Mukhamattadīpanī and Saddanīti are also great Pāli grammars. Saddanīti composed by Ven. Aggavaṁsa of Burma is the best contribution among all the Pāli grammars in Pāli. Recently, this was edited in three volumes in Roman script by Helmer Smith.

Both the Kaccāyana-vutti and Moggallāna-vutti cannot be said as were composed before the 7^th/8^th century A.D. In Kaccāyana-vutti the aphorisms (sūtras) of Kātantra or Kalāpa[6] are exactly employed in Pāli or the pure Māgadhī.[7] New aphorisms crept in only when the direct application of the Kalāpa sūtras proved inapt. It is, by no means, acceptable that the Buddha’s contemporary disciple Ven. Kaccāyana authored the Kaccāyana-vutti. Long before the compilation of Kaccāyanavutti, even in 6^th/7^th centuries Ven. Buddhaghosa and the great commentators had taken helps from the Pāṇini’s aphorisms. They did never use the rules of Kaccāyana. Therefore, we can see that the Pāli grammar evolved only after the Pāli literature was fully established. Even for that cause the main inspiration did not come from Pāli. The initiation of Kaccāyana-vutti was in imitation of the Kalāpa and that, in fact, provided the basis for whole tradition of Pāli grammar. Therefore, to make it appropriate for students of this country a grammar has to be composed aptly comparing with the Sanskrit and Prākŗt grammars. It is also important to compare Pāli with its contemporary languages of the rock-edicts, and Buddhist Hybrid Sanskrit. Moreover it is important to observe the differences between the Pāli prose and verse compositions. The form of Pāli cherished during the oral tradition worthy of reciting and in what manner and how much it diverges from the time the texts were put into writing also must be considered.

Here, I attempt to establish the reasons for my above assertions with some examples.

Expressing the infinitives like – to go, to do, to eat etc. –tuṁ ending verbs are generally used in Pāli irrespective of prose or verse. On the other hand, Kaccāyana has prescribed the three endings – –tave, –tuye, and –tāye (pā.pra.p.239).[8] Instead of –tuṁ, –i is also used sometimes – e.g. etasi (to go). We know that –tave, –tuye, –tāye, and –i were used only in Vedic Sanskrit. These could be observed in Pāli only in metrical compositions used for the sake of prosody. In the Pāli verses hessati instead of bhavissati, kāhasi instead of karissati are used; but, these terms are not found in prose passages. In the inscriptions of Asoka also these usages can be observed.

In declension of Pāli nouns the Dative and Genitive plurals are same. In Sanskrit the plurals of these two cases have different forms. In this instance, Māgadhī, Ardha-Māgadhī and such Prākŗt languages have similarities with Pāli. In Pāli declensions the Instrumental and Ablative plurals have same forms. In Prākŗt also it is the same where influences of the Dative case are noticeable. In Pāli and Prākŗt there are no dual forms, but, the Sanskrit dual forms have occasionally crept into the plural forms (in these languages). It has yet to be researched whether there is application of dual forms anywhere in Pāli. Where the Pāli grammar accords Sanskrit the rules are pleasant, but where previous Prākŗt words furtively manipulate, alterations have been increased. During the oral transmission of Pāli texts the language was rhythmical and onomatopoeic (dhvanyātmaka) appropriating it for recitation. Even the prose maintained some prosodic/metrical and phonemic harmony. There was a system of punctuating while reciting a prose passage also. For example, eveṁ me sutaṁ ekaṁ samayaṁ bhagavā viharati jetavane anāthapiñiḍikassa ārāme. In this expression, following phonetic harmony can be observed – evaṁ/ me sutaṁ/ ekaṁ samayaṁ/ bhagavā viharati jetavane/ anāthapinḍikassa ārāme.

The Sandhi-prakaraṇa, is full of substitutive rules, erroneous at many occasions. The main reason is that, lexical and phonemic changes of words/sounds cannot always be explained satisfactorily according to grammar. These changes have been made according to the Philology.[9] For examples, in iti + ādi = iccādi, adhi + ārāma = ajjhārāma, we cannot say that icca in place of iti, and ajjha in place of adhi are substituted. In these instances, iti + ā =ityā, and adhi + ā = adhyā according to rules of vowel combination (sara-sandhi) and changed tyā as cca, and dhyā as jjhā. According to phonetics if any letter of ta vagga is combined with ya then both change according to the corresponding letters of ca vagga; examples, satya – sacca, rathyā – racchā, madya – majja, madhyā – majjha, kanyā – kaññā.[10]

The Sandhi rule ‘ya–va–ma–da–na–ta–ra–la–cāgama’ is grammatically invalid (pa.pra.p13). Example, in pari + antaṁ = pariyantaṁ, ya is not (just) an insertion. According to Sandhi, pari + antaṁ = paryantaṁ and according to phonetic extension paryantaṁ becomes pariyantaṁ. In bhū + ādayo = bhuvādayo, ‘va’ is not (just) an insertion. According to Sandhi, bhu + ādayo = bhādayo, and later extending to bhuvādayo. In ekaṁ+ekaṁ = ekamekaṁ m is not insertion, because we disjoin ekamekaṁ as ekaṁ + ekaṁ.[11] tāva + eva = tāvadeva, here d is not an insertion, because Pāli tāva = Sanskrit tāvat, and tāvat + eva = tāvadeva. In the same way, yasmā + iha = yasmātiha, where t is not an insertion, because Pāli yasmā = Sanskrit yasmāt. In ni + antaraṁ = nirantaraṁ, r is not an insertion, because Pāli ni = Sanskrit nir. In cha + abhiññā = chalabhiññā, here, la is not inserted, because Pāli cha = Sanskrit şaḍ, thus şaḍ + abhiññā = şaḍabhiññā. In these instances the lost consonants have reappeared.

In idha + pamādo = idhappamādo, su + pasanno = suppasanno, hetu + kamo = hetukkamo etc. the aphorism ‘paradvebhāvo ṭhāne’ has been applied grammatically. It is not that the first consonant of the second word has been doubled following the last vowel of the previous word. Because, Pāli pamāda = Sanskrit pramāda, pasanno = prasannaḥ , kamo = kramaḥ. Therefore, pamādo is, in fact, ppamādo; pasanno is ppasanno, kamo is kkamo. The pronunciation of –pp(a)– and –kk(a)– are possible as they follow vowel sounds.

As there is no şa sound in Pāli there is no need of application of the şatva-bidhān.[12] But, ṇatva-bidhān[13], bākāra and batva-bidhān[14] are important. In Kaccāyana-vutti there is only one sūtra for ṇatva bidhān – rahādito ṇa (ka.su.4-1-26)[15]; alternatively read as rahādito no ṇa. In Mahārūpasiddhi also the same aphorism is seen. According to Mahāmahopadhyay Srīzuta Vidhusekhar Bhaṭṭacaryā, the correct reading/version should be – rahādito ṇo, or rahādito no ṇo. For examples, ramaṇīyo, āpaṇa, ñāṇa, tāṇa, tīṇi, etc. In Saddanīti another aphorism has been added, viz. ṇo nassa pa-pari ādito. Examples, paṇidhāna (praṇidhāna),[16] paṇipāta (praṇipāta), pariṇāma (pariṇāma). According to another rule in Saddanīti, ṇa becomes na (ṇassa ca no). Examples, taruṇa = taluna, karuṇa = kaluna. I do not accept that taluna and kaluna are Pāli words anyway. taruna and karuna are the correct forms.

Sāstrī Mahāṡay has rightly said that generally ‘ṇa’ in Sanskrit originals remains same in their corresponding Pāli words. Examples, varṇa = vaṇṇa, tŗṇa = tiṇa, lakşaṇa = lakkhaṇa, grahaṇa = gahaṇa. Again, where ṇa is used in Sanskrit according to inflection into cases, in Pāli na is used instead. Examples, cireṇa = cirena, ākāreṇa = ākārena, dharmeṇa = dharmena, rūpāṇi = rūpāni.

In the verb-conjugations, Pāli is subservient to Sankrit rules. Examples, gŗhnāti = gaṇhāti, ṡŗṇoti = suṇoti. Sāstrī Mahāṡay suggests that, rare examples of ñāṇa (jñāṇa) and oṇata (avanata) still remain due to some furtive influences from Prākŗt. (Indian Historical Quarterly, Vol. II, pp. 192-5, vide, ‘The Cerebralization of the Dental Nasal in Pali’ by H.P. Sāstrī). We may discuss on batva-bidhān shortly here. When joined with ya, both ba and va are used in Pāli. Example, abyākata, avyākata are the same. ba doubled becomes bba; if joined with bha and ma also ba remains. Examples, caturvidha = catubbidha, sarvadā = sabbadā, garbhā = gabbhā, ambu = ambu.

In ākhyāta-prakaraṇa of Pāli prabeṡ, pañcamī, sattamī and kālātipatti are classed under Tenses. Pañcamī, sattamī, and kālātipatti however, are not tenses but moods. To classify them under Tenses following justification from Saddanīti is predicable.

“pañcamī sattamīvhitā āṇatti-parikappikā,

saṅgayhamānā tā yanti paccuvappannamhi saṅgaṁ”

“atītānāgate cāpi kālātipatti-sambhavā”

Translation – ‘pañcamī and sattamī are indicative of command, wish, (suppositions, instructions, advice etc). To classify them under Tenses they have to be included in Present Tense. kālātipatti can be included in Past and Future Tenses.’ (Saddanīti, padamālā, 3)

Conclusion

In Conclusion, Dr. Barua has expressed the need of examination and proper application of the metrical aphorisms of ‘Vuttodaya’ to Pāli literature. Also, he asserts that Ven. Sangharakkhita did not concern or relate the Vedic metrical compositions whereas grammatically and prosodically Pāli poems show affinities to Vedic more than Classical Sanskrit. Ven. Jñānisvar is also silent in this issue. Dr. Barua urges Pāli students[17] also to examine and research the Pāli grammars not always relying upon the ancient ācaryas and encourages exploring new perspectives in Pāli studies. Finally, he expresses his gratitude to Ven. Jñāniṡvar particularly for editing and translating Vuttodaya into Bengali, and mentions that the University of Kolkata has honoured Ven. Jñāniṡvar by prescribing Vuttodaya for M.A. students.[18]

---

[1] Reading the transliterated Bengali words – a to be pronounced as au in audio; ā to be pronounced as ā in father. The Bengali words used here are – Pāli prabeṡ, ‘Vuttoday’, ‘Dhātokoş’, dhvanyātmaka, Satyadarṡan, pāli bā ṡuddha Māgadhī, ṇatva-bidhān, bākāra, batva-bidhān, şatva-bidhān

and. This applies to pronunciation of the names of Bengali authors also.

[2] Sanskrit

[3] Bangladesh was included in the then undivided India.

[4] Dr. Barua has mentioned Mr. Lokendralal as Pāli teacher in Higher English school, Nayapara, Gujra.

[5] Dr. Barua has mainly referred to the Pāli prabeṡ.

[6] Also from Pāṇini’s Aşṭadhyāyī

[7] From this expression – pāli bā ṡuddha māgadhī – it seems that Dr. B.M. Barua recognises Pāli as pure Māgadhī (ṡuddha Māgadhī). p.12

[8] Pāli prabeṡ, p. 239

[9] The words which have their origins in Sanskrit or other languages have to be understood from that perspective. In this instance Pāli philological studies should be done in comparison to Sanskrit, Prākŗt and other available resources.

[10] The comparison made here is with Sanskrit. The first word in each pair is Sanskrit.

[11] From this example and those that follow Dr. Barua seems to suggest that the words are direct liftings from Sanskrit in their combined forms. And, only later Pāli grammarians formulated new rules for them while they already had some grammatical explanation. Therefore, the new aphorism pertaining to such already sandhified forms do not seem valid. To explain these terms according to grammar the Sanskrit Sandhi rule in which they fall seem more appropriate.

[12] Also, read as şatva-vidhi, refers to the grammatical rules pertaining to the use of letter şa, especially substituting in place of ṡa, and sa.

[13] The rules for changing the dental na into cerebral ṇa, and vis-à-vis.

[14] In Bengali, the sound va has been lost in ba

[15] The abbreviation indicates – Kaccāyana sūtra.

[16] The words in brackets are Sanskrit.

[17] By etaddeṡiya he probably means ‘(students) of this country’ may be referring to then India.

[18] Many Bengali Buddhist monks were not trained academically in the universities. But, their works have been venerated by academics for high scholastic values. Mention must be made of Ven. Bisuddhananda’s ‘Satyadarṡan’ which like the Visuddhimagga of Ven. Buddhaghoşa is an exegesis to whole Buddhist philosophy. In fact, the text could be duly regarded as an updated Visuddhimagga, because, in it Ven. Bisuddhananda has not only referred to the Tipiṭaka and its commentaries but also has shown their application to the present world. His expressions and interpretations of philosophical points are very lucid and deep. Sadly, the text has not yet been translated into English. Ven. Jyotipala Mahathera’s (the tenth Sangharāja of Bangladesh) karmatava is worth mentioning in this regard beside others works by eminent scholar monks.