Page #1
--------------------------------------------------------------------------
________________
EDITING OF ANCIENT ARDHAMĂGADHI TEXTS IN VIEW OF THE TEXT OF
VISEŞĀVAŠYAKA-BHĀŞYA?
K. R. Chandra
On the basis of a consideration of subject-matter, style, and metrical form, scholars have formed an almost unanimous opinion that certain portions of the Jaina canonical literature in Ardhamāgadhi possess very old strata. On the same grounds these ancient portions bear similarity with the oldest portions of the Pali Tripitaka. However, the concerned phonological, and to some extent the morphological forms of the language are not so archaic in these portions and in this respect the editions of these works differ among themselves.
In the Western scholars' editions, we notice that the intervocalic stops generally lost their conclusion as per the rules framed by Prakrit grammarians which, strictly speaking, were meant for the Maltārāstrī Prakrit. As a result, in these editions, not only the consonantal spelling of the words but occasionally the morphological forins of the later period also are encountered. A preference accorded to later forms consequently neglects the archaic, in point of fact original, and hence authentic, variants available in some manuscripts.
The same method is adopted by Indian editors though usually with lesser rigour and hence on a smaller scale. Therefore, noticeable disparity is discernible between the age of the work and its language. Hence the current form of the language available in the so far published editions is not authentic even when they do not affect the sense intended to be conveyed. This situation impels one to find out the reasons behind it.
For explaining the reasons behind this kind of disparity, in this paper, an attempt has been made in its proper perspective based on the analysis of the language of the text of the Višeşāvasyaka-bhāsya in its different editions and its variants in MSS. on the one hand and that of the language of older inscriptions and the Isibhāsiyain? edited by W. Schubring on the other hand.
The Visesavasyaka-bhāsya is a versified exposition of the Sāmāyika-āvasyaka and its corresponding niryukti (c. A. D. 525) by the learned caityavasi or abbatial pontiff Jinabhadra gani kşamāśramana (ob. A. D. 588 or 594). He had also composed its coinmentary in Sanskrit which had remained unfinished since he did not live as has been reported by Kotřārya vādi gani (c. A. D. 675-700) who completed the coinmentary
Page #2
--------------------------------------------------------------------------
________________
K. R. Chandra
Nirgrantha
1. Phonological Analysis of the VABh.
(a) Phonological analysis of the initial word/words of every verse (No.1 to 100 only) quoted by the author himself in his auto-commentary.
Medial consonants
Dropping
Voicing
Retention
(i) Unaspirates
09 (13%)
09 (13%) 49 (74%) (ii) Aspirates
00 (00%)
06 (35%) 11 (65%) i.e. changing
into that Total
09 (10%)
15 (18 %) 60 (72%) (b) The same words (i. e. of 1 (a) above) in the text of MS. J. Medial consonants Dropping
Voicing
Retention (i) Unaspirates
07 (10%)
12 (18%) 48 (72%) (ii) Aspirates
00 (00%)
09 (53%) 08 (47 % i.e. changing
into ha Total
07 (08.5 %) 21 (25%) 56 (66.5%) This analysis reveals that there is very little advance in the phonological changes of the consonantal spelling of the words of the text of the MS-J, it is 5.5% only (72?--66.5). 2. A phonological analysis of the words of all the initial 100 verses according to MS-J (as
given below) reveals that the percentage of dropping is 11.5% and that of retention is 70% i. e. only a difference of 2% as compared to that of the auto-commentary. The phonological changes occuring in two MSS, and two editions can be tabulated as follows: (a) Non-aspirate Medials
Conso- Dropping =D Voicing = V
Retention - R Total nants J. T. H. K. IJ. T. H. K. J. T. H. K. K 34 35 36 36 22 21
04 04 02 02 6 00 00 00 00 00 00 00 00 28 28 28 28 C 05 05 05 05 00 00 00 00 00 00 00 00 J 03 03 03 03 00 00 00 00 | 00 00 00 00 T 13 15 185 185 oi 01 00 00 184 182 13 13 198 D 06 29 62 31 33 08 00 00 39 41 16 1778 P 00 00 01 01 50 50 49 49 09 09 09 09 59 Y 17 20 38 40 00 00 00 00 79 76 56 56 96
v 06 06 06 06 00 00 00 00 147 147 147 147 153 Total 84 113 336 337 106 80 71 71 490 487 273 272 680
Page #3
--------------------------------------------------------------------------
________________
Vol. I-1995
Editing of ancient..
(b) Aspirate Medials Conso- Dropping =D Voicing = V
Retention : R Total nants d. T. H. K. J. T. H. K. J. T. H. K. Kh | 01 01 01 01 | 00 00 00 00 | 00 00 00 00 01 Gh 00 00 00 00 00 00 00 00 01 010101 01 Th | 02 03 30 30 28 27 00 00 00 00 00 00 30 Dh 02 03 04 05 00 00 00 00 36 35 34 33 38 Ph 00 00 00 00 00 00 00 00 01 010101 01 Bh 00 00 04 04 00 00 00 00 36 36 32 32 36 Total 05 07 39 40 28 27 00 00 74 73 68 67 107
Grand Total Total 89 120 375 377 134 107 71 71 564 560 341 339 787
D
(c) Percentage (1) M.
Unaspirates Aspirates Total
84 (12.5)
OS (05.0) 89 (11.5%)
106 (16.0)
28 (26.0) 134 (18.5%)
R 490 (72.0)
74 (70.0) 564 (70%)
(ii)
M. S.
Unaspirates Aspirates
113 (16.5)
07 (07.0) 120 (15%)
80 (12.0)
27 (25.0) 107 (14 %)
R 487 (70.0)
73 (70.0) 560 (71%)
Total
(iii)
R
Edition
Unaspirates Aspirates Total
336 (50.0)
39 (36.0) 375 (48 %)
71 (10.5) 00 (00.0) 71 (9%)
273 (40.0)
58 (53.0) 331 (43 %)
(iv)
D
V
Edition
Unaspirates Aspirates Total
337 (50.0)
40 (36.0) 377 (48 %)
71 (10.5) 00 (00.0) 71 (09 %)
272 (40.0)
57 (53.0) 329 (43%)
Page #4
--------------------------------------------------------------------------
________________
K. R. Chandra
Nirgrantha
3. A high percentage of retention and very low percentage of dropping of medial consonants in the MS-J is questionable in view of the evolution of MIA language and the nature of the Mahārāstri Prakrit in the 6th century A. D., the date of the composition of the VHBh can be assumed between A. D. 580-585. If the author made use of large the number of Sanskrit equivalents (tatsama words), it seems to be the consequence of the influence of Sanskrit during its golden age. In the MS-J., medial dental - t- is often retained in the words, namely ablative singular suffix -to, verbal termination -ti of the present tense of III singular, and in the past participle -ta. Out of 198 instances, -t-is voiced on a single occasion e. g. disadi for Sanskrit drsyate (verse No. 53) and it is dropped 13 times only whereas retained 184 times. In the MS-T the dropping is 15% and the retention 70%. When we compare these two MSS. with the two editions H and K., the dropping of medial consonants rise up to 48% and the retention goes down to 43%. The percentage of voicing goes on decreasing with the passage of time (i.e. with the later dates of MSS.) e. g. 18.5% in J, 14% in T, and 9% in H as well as in K. This fact makes it abundantly clear that the copyists and the reciter-teacher-monks (preachers) brought about this kind of modification either inadvertantly or intentionally on account of the influence of the Prakrit language of their own times and also under the sway of the Prakrit grammar of very late dates.
If this was the condition of the original language of the text of VAB), within a few centuries after its composition, then what would be the state of the original Ardhamāgadhi language of the Jaina agama texts in their MSS. copied a thousand or 1500 years late. We are now becoming increasingly aware that there has been a continual change in the language of a text in Prakrit MSS. which were copied in different periods by different copyists in western India.
4. Position of the Language of Inscriptions :
A study of the phonological changes in the language of inscriptions reveals that the dropping of medial consonants goes on advancing with the passage of time. This tendency of dropping has travelled from the North-West and North to other regions of india. The use of -n- (cerebral nasal) for -n- (dental nasal) has been a trait of the Northwest, North, and South. One of the oldest available Prakrit inscription, now regarded of c. 200 from Barli?, there is no dropping of medial consonant (e. g. there is no -e in place of -te and -ye, the inflectional suffixes.)
In the Aśokan inscriptions, the average phonological change is only 1 1/2 to 3% only. In the inscription of king Khäravela (c. B. C. 50-25), the voicing of medial consonants advances (i. e. 6 out of 8 instances are voiced). There -tho- becomes -dh-, and the phonological changes on the whole are 5 to 6%. It is in the inscriptions of Pañjatār, Kalwān and Takşaśilā (of the 1st century A. D.), located in the North-West, that the
Page #5
--------------------------------------------------------------------------
________________
Vol. 1-1995
Editing of Ancient....
dropping of medial consonants has gone high, indeed up to 30% to 35% and the voicing of unvoiced consonants is near about 17%. Dental nasal n- is initially cerebralised into n- up to 100% and medially -n- into -n- up to 75%. This tendency of cerebralisation of -n- into -n- is traced in the South and West. And in the later centuries it spreads to other regions of India 10.
5. Phonological changes in the Isibhästyäiri
Isibhāsiyaim" is regarded as one of the oldest of the Ardhamāgadhi texts along with the Acaränga, the Sütrakṛtänga, the Uttaradhyayana and the Daśavaikälika. The state of phonological changes in this work is quite different from that encountered in other older Jaina canonical texts in the Ardhamägadhi. An analysis of the data from some select chapters No. 1, 2, 3, 5, 11, 29 and 31 reveals that an average of dropping, voicing. unvoicing, and retention is 11 to 35%, 7 to 22% and 45 to 81% respectively. The chapter-wise analysis is as follows:
chs
D
V
R
27
12
61
Total 100
01
chs
D
V
R
01
222
27
12
61
02
05
03
37
45
Percen- D tage R
02
The percentage of phonological changes in these chapters is as follows:
29 31 Average
11
07
81
03
27
32
87
146
32
68
03
18
22
60
05
07
01
19
27
00
100
05
27
04
69
11
22
02
53
77
15
85
11
The position of medial t- in these chapters is as follows:
27
03
70
29
28
72
49
28
62
139
35
20
45
31
27
73
71
208
15
93
140
459
226 760
Total
31
07
62
48
52
27.4%
12.2%
60.4%
6. The position of phonological changes in the Acaränga edited by W. Schubring, the oldest book of the Ardhamagadhi canonical literature, is quite different from the Isibhāsiyaim. In it the dropping of medial consonants is more than 50%, a state which
21%
79%
5
Page #6
--------------------------------------------------------------------------
________________
K. R. Chandra
Nirgrantha
cannot be relied upon. The palm-leaf Ms., the oldest one in date that was used by Schubring in editing the text of the Acāranga, part I, is dated V. S. 1348/ A.D. 1292 and there in it the verbal termination -ti of present tense III singular is retained up to the extent of 50%. Schubring has not noted from the available variant -t- and he has dropped it wholesale. In comparision of it the edition of the Acārāńga by Muni Jambūvijayaji " reveals quite a different state of the affairs. In it the dropping of medial consonants is 24% only. Linguistically, it is a better edition; yet it needs emendation on the basis of the variants available in various MSS. of the text and its cūrmi. commentary (c. late 7th cent. A. D.) because they preserve the archaic traits 14.
7. It can now be surmised on the basis of the above comparative study that the Ardhamāgadhi language of older Jaina canonical texts went on changing from century to century and generation to generation as the time progressed. Therefore, in editing older agamic works and the older portions of the Ardhamāgadhi canon one must discard younger words and forms. Older words and forms that are available in the MSS. of the text itself, or its cūrni, or in the quotations from the text in other works, should be adopted even if they do not agree with the rules of Prakrit grammarians as the treatises on grammar are of much younger origin and they do not present any historical or comparative grammar. It is now clear that the original language of the canonical works went on changing to a certain extent as the work of copying multiplied in different periods and at different places : otherwise how could there have been so much variation in the phonology of words in the text of Isibhásiyāin on the one hand and that of the Acarariga on the other hand".
8. The authenticity of this conclusion can be verified with the help of variants that are noted in different edition of the VABh. A glance at thiem 16 makes it clear that the ** change in the phonology of the type that is given below had taken place at different stages of the evolution of Prakrit dialects :
Stage
(i)
(ii)
.
k
=
dh = 1
dh
b
dy
dh
th = d =
0
According to the historical evolution of the Middle Indo-Aryan languages, the voicing of unvoiced and the unvoicing of voiced medial consonants takes place at an earlier stage than the dropping of medial consonants (i. e. stops). In this respect two published editions of the VABh., e. g. K and H are highly influenced by the later stage of Prakrit
Page #7
--------------------------------------------------------------------------
________________
Vol. 1-1995
Editing of Ancient....
i. e. the Mahārāstrī Prakrit, whereas the MSS. J and T preserve older forms and so the earlier stage of the Prakrit language and the MS. J preserves still earlier stage (i. e. older spellings of the words). 9. The language of the editions H and K had been regarded as the genuine language of the author until the palm-leaf MSS. J and Sam were discovered (i. e. the language having medial consonants generally dropped). Further non-dropping of medial - t- cannot be admitted as - t- śruti because it is preserved in the MS-J also and, therefore, it constitutes a feature of the original language of the author. The way the language underwent change at the hands of copyists and teacher-reciters can be discerned from the editions H and K which are highly influenced by popular language as well as by the rules of later Prakrit grammarians. Hence the principle of editing ancient Ardhamāgadhi texts needs wholesale modification, namely adopting older words and forms from the available source-material, discarding the younger words and forms, and thus being faithful and closer to the original language. The comparative table of phonological variants in different editions and MSS. of VABh is appended as the sequel. (See phonological analysis of VABh Table for serial No. 8)
Notes and References :
1. L. D. Series 10, ed. Dalsukhbhai D. Malvania, Ahmedabad 1966. 2. L. D. Series 45, ed. W. Schubring, Ahmedabad 1974. 3. The MS. San. having auto-commentary is dated A. D. 1434, vide VABH, part I, Preface,
L. D. Institute of Indology, Ahmedabad 1996, p.4. 4. Palm-leaf MS. from Jaisalmer, datable to c. A. D. 950, vide ibid., p. 1. 5. Ibid., pp. 1, 3, 4. 6. Ibid., p. 1. 7. M. A. Mehendale, Historical Grammar of Inscriptional Prakrit, Deccan college, Poona 1948,
pp. 326-7. 8. This change is higher in the North and North-West and lower in other regions. 9. D. C. Sircar, Select Inscriptions, Vol. 1, No. 32-34, (total instances 89; dropping 27, and
voicing 13, unvoicing 2 and retention 47). 10. Mehendale, Historical., p. 276. 11. L. D. Series No. 45, Ed. W. Schubring (of the original German edition), Ahmedabad 1974. 12. On the basis of the first chapter. 13. Äyaranga-suttam, Jaina-Agama-Series 2, Bombay 1977. 14. Some of them (i. e. the older forms) have been given in the footnotes of the text by
the editor.
Page #8
--------------------------------------------------------------------------
________________
K. R. Chandra
Nirgrantha
15. Both works had been edited by W. Schubring, the former on the basis of two MSS. and one
printed work and the later on the basis of several MSS., the curni, etc. Further, the Acārānga has been published in several editions and all of them differ from one another in this respect.
16. See the table given below.
Abbreviations and References :
8
D =
H =
J =
K =
MS =
R =
T =
V =
Voicing or un-voicing of inter-vocal stops.
VABh. The text of the Viseṣavasyaka-bhāṣya edited by Pt. D. Malvania, L. D. Series No. 10,
1966, Ahmedabad.
Note: See the above-noted work for the details of the editions and Mss. H, K, and J,
Dropping Loss of occlusion of inter-vocal stops.
Edition of Maladhari Hemacandra.
Palm-leaf manuscript from Jaisalmer.
Edition of Kottyācārya.
Manuscript.
Retention Retention of inter-vocal stops.
Palm-leaf manuscript (undated)
Page #9
--------------------------------------------------------------------------
________________
Comparative Table of phonological Variants of some words from different editions and MSS. of
Visesavasyakbhasya (initial words of verses No. 1 to 100)
Year of Publication
|
1966 A.D. L.D.Serles
| 1936 A.D. | 1914 A.D. | MS. of C. 10th |
Undated MS.
Consonant
Cent.
Verse No.
Sanskrit
L.D. Text |
Auto.Comm.
I
K
H
T
अतीत अधवा
अतीतं अहवा अहवा अहवा अहवा
अधवा
.
अधवा
अहवा
अतीतम् अथवा अथवा अथवा अथवा अथवा अनुमतम् अभिधानम् अभिधानम् अवधीयते आगमतः आदि आदि
अतीतं अधवा अधवा अधवा अधवा अधवा अणुमतं अभिधाणं अभिधाणं अवधीयते आगमतो आदि -आइ
अतीतं अधवा अधवा अधवा अधवा अधवा अणुमतं अभिधाणं अभिधाणं अवधीयते आगमतो आदि आदि
अतीय अहवा अहवा अहवा अहवा अहवा अणुमयं अभिहाणं अभिहाणं अवहीयए आगमओ
अधवा अधवा अणुमतं अभिधाणं अभिधाणं अवधीयते आगमतो आदि आति
अणुमयं अभिहाणं अभिहाणं अवहीयए आगमओ आइ
अतीतं अधवा अधवा अधवा अधवा अधवा अणुमतं अभिधाणं अभिधाणं अवधीयते आगमतो आदि आई
dh
dh&t
आइ
आइ
इध
इध
स्थ
इह
* य अ अ ३ ॐ
ऋजुसूत्री
उज्जुसुया
उज्जुसुया
उज्जुसुता कत गालयति
छ अ अ अ
कृत गालयति
कय
उज्जुसुता कत गालयति चूतो
उज्जुसुता कत गालयति
उजुसुता कत गालयति चूतो
गालय
गालयइ
चूओ
चूतो
Page #10
--------------------------------------------------------------------------
________________ 10 K. R. Chandra Nirgrantha ___L.D. Text | Auto. Comm.| K Consonant Verse No. | Sanskrit चूतादिभ्यः 36 चूताईए. चूताईएहितो णाण णाण चूता. णाण नाण चूताईए. णाण णाणं ज्ञानम् णाण चूयाईए. नाण नाणं जाणय तओ नाण 46 ज्ञायक जाणग जाणग जाणय 81 ततः ततो जाणग ततो | विधा तिहा तिधा प. द्रवति ततो तिधा दवते दूयति णमोकारो णिवातणातो -णिबुज्झति द्रूयते नमस्कार निपातात् -निबुध्यते -रिवृत तिधा दवते दूयति णमोकारो णिवातणातो -णिबुज्झति -णिब्बुत तओ तिहा दवए दुयए नमोकारो निवायणाओ -निबुज्झइ -निब्य भूय मंगिज्जए | चूयातीए. णाण णाणं जाणग ततो तिधा दवते दूयति णमोकारो | णिवातणातो -णिबुज्झति -णिबुत भूत मंगिजते दवए दुयए नमोक्कारो निवायणाओ -णिबुज्झइ -निव्वुय 97 दवते दुयए णमोकारो णिवातणातो -णिबुन्झति -निव्वुय भूत मंगिज्जते 58 -णिज्बुत भूय भूत मंगिजते 22 मङ्ग्यते मंगिज्जते मंगिज्जए मइ / 86 मइ मति जह जध जध जति जति जह जति 30 जति जति जइ जइ 100 जइ ཝཾ ཝཱ ཝཱ མི @ # # མ जाता जति 72 विवदन्ति जति विवदंति सुत विवदंति विवदंति विवदंति सुत विवयंति सुय सुय विवयंति सुय सुत सुत श्रुत सुत सुत सुय सुत सुत 100 सुत सुत सुतस्स सुय सुअस्स सुत सूत्रस्य सुतस्स सुअस्स सुतस्स हेतू हेऊ हेक सुतस्स हेतू हेतू