Book Reviews

Author(s): J W De Jong
Publisher: J W De Jong

Page 30

216
REVIEWS
these remains of New Ardhamāgadhi support Lüders's hypothesis regarding Old Ardhamāgadhi as the language of the oldest Buddhist tradition.
K. R. Norman (The dialects in which the Buddha preached, pp. 61-77) discusses the two following questions: (a) Were the sources which Buddhaghosa followed correct in calling the Buddha's language Māgadhi? (b) Why was the language of the canon, and of Buddhaghosa's commentaries theron, called Māgadhi? According to Norman Buddha taught in the Old Māgadhi dialect which was spoken in a small area of Magadha with all three typical characteristics, viz. e, 1 and ś, while elsewhere in Magadha Old Ardha-Māgadhi was spoken with only two of these features, viz. -e and I. Pāli, the language of the Theravāda canon, akin to Māgadhi and spoken somewhere in Magadha, was called 'Māgadhi' when Buddhism was introduced into Ceylon. It is perhaps not surprising to see that Norman's attempt to justify calling the language of the canon Magadhi did not meet with universal approval (cf. p. 185).
In her paper (La langue primitive du bouddhisme, pp. 43-60) Colette Caillat emphasises "that what is termed 'langue primitive du bouddhisme' is really but the result of our reconstruction schemes; moreover, that only fragmentary aspects are generally considered, and conclusions are drawn mostly from partial analysis" (p. 57). She draws attention to several facts relating to morphology and stylistics which deserve to be taken into consideration in studying the language of the oldest Buddhist tradition.
Gustav Roth examines the "Particular Features of the Language of the Arya-MahāsāmghikaLokottaravādins and their Importance for Early Buddhist Tradition' (pp. 78-93). He points out several correspondences between the Prākrit of the Mathurā inscriptions (e.g. forms in are for any oblique singular case-form) and the Prakrit underlying the scriptures of the Mahāsāmghika-Lokottaravādins. Roth also finds traces of Sauraseni (the particle dāni, absolutives in-ya) and of Māgadhi/Ardhamāgadhi (absolutives in -yānam, the particle of address hamgho).
In a supplement Roth gives an edition of the complete text of the Patna Dharmapada (pp. 93-135). He points out the specific features of the language of the text and remarks that this language is a western type of Prakrit very close to Pāli. Recently the text of the Patna Dharmapada has also been published in India, by N. S. Shukla, as The Buddhist Hybrid Sanskrit Dharmapada (Patna, 1979). It is very useful to have two different editions of the same text in order to discover which akşaras are difficult to distinguish in the manuscript. Roth's promised study of the script of the manuscript of the Bhiksuni-Vinaya has not yet appeared and no facsimile editions of manuscripts of the Patna collection have been published sofar. A fascimile edition of the manuscript of the Patna Dharmapada would certainly be very desirable because in both editions there remain a number of difficulties. In the first folio, three lines in Tibetan are written in dBu-med. Champa Thubten Zongtse reads: (1) chos-Idan zab[s] kyis gsuns-pa dan (2) chos kyi cho-ga bdud-rci'i cho-ga (3) żes bya ba głun chad brgyad grags-pa. Roth gives the following tentative translation: "(1) That which is said by Dharmika-pāda and (2) the way of the Dharma (which is] the way of the nectar (i.e. of the doctrine of the Buddha), (3) so it is known as a work that comprises) eight principles." Shukla's edition has bcom-dandas instead of chos-ldan zab(s), chig instead of cho-ga and brgya-phag Ina'o instead of brgyad grags-pa. This certainly makes more sense: “Said by the Bhagavat and called the Dharmapada, the Amộtapada. The extent of the text is five hundred (stanzas)." In Shukla's edition verse 120 begins with the following line which is not found in Roth's edition: tathavidha samaņā prabhūtapramña. In Shukla's edition there are two verses beginning with sukhāmāni bhūtāni. The second is missing in Roth's edition: sukhakāmāni bhūtāni / yo dandena na vihimsati // ättano sukham esano / precca so labhate sukham (cf. Roth No. 204). In 121b Shukla has: na candanam tagaram vählikam vā; Roth: na candanam vähnikam vā.
It is easy to confuse bh and t. In 12b Roth has: yo dhammam abhivattati; Shukla: yo dhammam ativattati (cf. Dīghanikāya, vol. II, p. 182, 1.14). According to Wayman's tables (Analysis of the Srāvakabhūmi manuscript, 1961, pp. 5-8) there is no difference at all between ba and va. In the language of the text dy-> b-, cf. 11b bitiyatā. Roth mostly prefers a v, cf. 141a: tahna-vitiyo puruso, 333a: sraddha-vitiyam. Rather similar are ya and pa. In 41a Roth