________________
114
Pravacanasāra
post-canonical literature, only because JM. inherits many features of AMg. Forms like pagada, pudhatta etc. smack of Sauraseni.
The treatment accorded to intervocalic consonants, especially stops, is of a very uncertain character; the general tendency is towards softening or retention, a phenomenon quite usual in the early text of the AMg. canon and in Sauraseni; in Bhāsa's Sauraseni retention or softening of intervocalic consonants is still facultative, while in that of Kālidāsa the tendency is towards omission. It must be remembered that the Svetāmbara canon has been, in later days, subjected to a strong Māhārāstri influence, because, since the days of Valabhi redaction, 980 years after Mahāvira, the canon came to be shaped, nourished, nurtured, copied and studied, in Western India, especially in Gujarat and Kathiawar. Still words like anega, udahī, logāloge, vibhāga etc. can be seen on any page of the canonical texts, and they indicate the basic nonMāhārāştri element of AMg. The softening of k to g is merely an extension of Saurasens phenomenon, generalised by grammarians, of softening t and th to d and dh. The retention of g is quite normal in AMg. of the canon. The retention of c and j exhibit a strong contrast with normal Māhārāştri. The softening of t, which is almost universally taken recourse to in Pravacanasūra is peculiarly Sauraseni and extended by grammarians to Māgadhi and other dialects; as a result of this the Present 3rd p. sg. termination is necessarily di, which, according to critical editors is i in AMg. texts, though some editors of the orthodox school would preserve ti. The treatment accorded to dental nasal in Pravacanasāra is worthy of note: it is universally cerebralised whether initial, medial or conjunct; and this is in agreement with Sauraseni. Turning to Prakrit grammarians on this point, Vararuci wants unexceptional cerebralisation, while Hemacandra says that dental may be retained if initial. Coming to the practice in AMg., the MSS. are never in agreement; but critical scholars, from Weber to Vaidya, 1 have created a substantial convention of retaining a dental at the beginning of a word and of changing it to cerebral elsewhere. Exception is, however, made in the case of nam.2 With regard to conjunct nnor nn, Dr. Jacobi's rule3 is that nn should be preferred, if there is n in the Sk. original, otherwise nn; but one must say that (p. 120:) Dr. Jacobi has over-extended his convention, when he prints savvannā, which, as a matter of phonetic necessity, should be savvannū.4 Our MSS. of Pravacanasāra are almost in agreement in preferring » everywhere, and that has been uniformly followed. The treatment of aspirated consonants th, dh, bh at times agrees with AMg. and at times with Sauraseni. The phenomenon of ya-śruti is practically the same as in AMg. canon, though in JM., as seen in modern editions, there is a tendency to observe strictly the rules of Hemacandra. In AMg. there are many cases where r is changed to l; but in our text the tendency is towards retaining r, though we have a few cases of r changing to l; this is more in agreement with Sauraseni.
1 Weber: Uber ein Fragment der Bhagavati, Berlin, 1866-7; Dr. Vaidya: his editions of
Nirayāvaliyão, Antagadadasāo etc., Vivāgasuya and Paesikahānaya etc. Poona. 2 Leumann retains forms like nūnam, no etc., see his Das Aupapātika-sutra, sections 132-3,
137; Dr. Vaidya: A Manual of Amg. Grammar, pp. 21-2. 3 Dr. Jacobi: Ayāramga sutta, PTS. London 1882, p. xv. 4 Hema. Grammar, VIII, i, 56, ii, 83.
Jain Education International
For Private & Personal Use Only
www.jainelibrary.org