Moksopaya, Yogavasishtha and Related Texts, J. Hanneder

T HE M OKS. OP ĀYA , Y OGAV ĀSIS. T.
HA
AND R ELATED T EXTS
E DITED BY
J ÜRGEN H ANNEDER
Vakatseite
INDOLOGICA HALENSIS
GEISTESKULTUR INDIENS. TEXTE UND STUDIEN.
HERAUSGEGEBEN VON WALTER SLAJE

UNTER MITWIRKUNG VON
JÜRGEN HANNEDER UND ANDREAS POHLUS
BAND 7
SHAKER VERLAG
AACHEN 2005
THE MOKṢOPĀYA, YOGAVĀSIṢṬHA
AND RELATED TEXTS
EDITED BY
JÜRGEN HANNEDER
SHAKER VERLAG
AACHEN 2005
Bibliographische Information Der Deutschen Bibliothek
Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen
Nationalbibliographie; detaillierte bibliographische Daten sind im Internet
über http://dnb.ddb.de abrufbar.
Copyright Shaker Verlag 2005

Alle Rechte, auch das des auszugsweisen Nachdruckes, der aus-
zugsweisen oder vollständigen Wiedergabe, der Speicherung
in Datenverarbeitungsanlagen und der Übersetzung, vorbehalten.
Printed in Germany.
ISBN 3-8322-4265-1
ISSN 1619-4470
Shaker Verlag GmbH • Postfach 101818 • 52018 Aachen

Telefon: 02407 / 95 96 - 0 • Telefax: 02407 / 95 96 - 9
Internet: www.shaker.de • eMail: info@shaker.de
Editor’s Preface
The Yogavāsis.t.ha or – as it is called in its oldest version – Moks.opāya is a

monument of Sanskrit literature, not merely by its epic size – it is larger than
the Rāmāyan.a – but especially through its unique style and contents. Most
scholars agreed that it can not be categorized easily, especially since its blend
of narrative, poetry and philosophy escapes the traditional boundaries of the
genres of Classical Sanskrit, but the work’s philosophy has puzzled many.
From its oldest version (“Moks.opāya”), which is without the thin but persua-
sive layer of Advaita Vedānta, it now appears that the work propounds an
idiosyncratic philosophy that is almost unique in the Indian sphere – by its
contents, but also because it rejects the authority of any scripture or given
philosophical framework. The fact that it nevertheless uses other doctrines in
an inclusivistic way has tricked scholars into reading the most diverse philoso-
phies into the text. In fact, the author of the Moks.opāya seems to have con-
ceived a non-dualistic system of his own – and elaborated it on an enormous
scale. This is, however, only one aspect of the Moks.opāya. Some readers more
interested in Indian poetical theories have found it to be equally unique and
interesting,1 especially for its integration of philosophy and poetry.
The Moks.opāya Research Group is a coordinated effort of several academi-
cal projects at a comprehensive investigation of the Moks.opāya, ranging from
a critical edition of the text as well as the fragments of the commentary
of Bhāskarakan.t.ha, an assessment of the abridged versions, to translations
and topical studies. Two projects, funded by the German Research Foun-
dation, are located in the Indological Institute at the University of Halle-
Wittenberg,2 the complete critical edition is under the patronage of the
Mainzer Akademie der Wissenschaften und der Literatur. A critical edition of the
1 “The philosophical stories are replete with lyric descriptions of great beauty, and the very
language itself gives evidence of a highly literary mind.” (M ASSON and PATWARDHAN (1985),
p. 30) “There is no finer example in world literature of a profound philosophical mind with a
genius for artistic description, even though many of the verses betray a certain lack of traditional
literary education (odd syntax, unorthodox similes etc.). There is a fullness and an overflowing
of the creative spirit in this work such as we have never come across in any other Sanskrit text.”
(M ASSON and PATWARDHAN (1985), p. 30, fn. 3.)
2 “Kritische Edition des Utpattiprakaran a” (P ETER S TEPHAN and J ÜRGEN H ANNEDER) and
.
“Indo-Persische Übersetzungsliteratur aus der Mogulzeit (16./17. Jhd)” (H EIKE F RANKE and S U -
SANNE S TINNER), the latter being a collaboration of the Indological and the Oriental Institute
(Prof. WALTER S LAJE and Prof. J ÜRGEN PAUL).
4
fragment of Bhāskarakan.t.ha’s commentary on the Nirvān.aprakaran.a is being

prepared by B RUNO L O T URCO at the University of Rome, a translation of the
Utpattiprakaran.a has been recently started by M ARTIN G ANSTEN at the Uni-
versity of Lund and is funded by the Swedish Research Council.
The present volume is a collection of articles resulting from a panel on
the Moks.opāya, Yogavāsis.t.ha and related texts, which was held on the 24th of
September at the 29th Deutscher Orientalistentag in Halle, where the interna-
tional Moks.opāya Research Group took the opportunity to present results of re-
cent research as well as ongoing editorial and other projects to a wider Indo-
logical public. The publication has been made possible through a generous
grant by the Helmuth von Glasenapp-Stiftung.
An introduction by the editor is followed by four articles that deal with
the text of the Moks.opāya and its philology. First WALTER S LAJE analyses the
evidence for a localization of the text in Kashmir, then he presents a pre-
liminary description of Delhi and Śrı̄nagar manuscripts, which have only
recently become accessible. Another tour in search of manuscripts in Ma-
harashtra and Gujarat has resulted in a further survey by P ETER S TEPHAN
and S USANNE S TINNER, which has been included in this volume almost in
the last minute. Next P ETER S TEPHAN introduces the critical edition of the
third book of the Moks.opāya, the Utpattiprakaran.a, which will be completed
in the near future. In the following three articles the focus is on abridged
versions of the Moks.opāya and Yogavāsis.t.ha. S USANNE S TINNER presents re-
sults of her studies on the Laghuyogavāsis.t.ha as well as previous unknown ver-
sions, J ÜRGEN H ANNEDER presents a brief analysis of the Moks.opāyasam . graha,
whereas H EIKE F RANKE traces the history of the earliest Persian translations
of the Laghuyogavāsis.t.ha produced at the Mogul court. The volume is com-
pleted by B RUNO L O T URCO’s study of the deep structure of the Moks.opāya,
namely its use of ākhyānas as an integral part of philosophical instruction.
At present the publications by the Moks.opāya Research Group are often
based on materials, mostly preliminary editions, that are shared among the
group, but have not yet appeared in print. In quoting from the Moks.opāya
we therefore resort to the following guidelines: Where available, quota-
tions from the Moks.opāya are based on the preliminary version of the forth-
coming critical edition, which applies only to the Utpattiprakaran.a. In that
case all readings are given in the apparatus. In the case of the Vairāgya-,
Mumuks.u- and Sthitiprakaran.a the text as contained in Bhāskarakan.t.ha’s com-
5
mentary3 is quoted. In all other cases, notably that of the Nirvān.aprakaran.a

the text is cited according to the manuscripts, which are then identified.4
Quotations from other versions, i.e. the Yogavāsis.t.ha, Laghuyogavāsis.t.ha and
(Yoga-)Vāsis.t.hasāra are based on the printed editions,5 the other abridged ver-
sions remain unedited and are therefore quoted from manuscript.
3 Published in H ANNEDER and S LAJE (2002), S LAJE (1993) and S LAJE (2002).
4 See below, p. 139f., for a brief list of sources.
5 For this and other primary texts, see the bibliography on page 144.
Vakatseite
Contents
The Moks.opāya: An Introduction

J ÜRGEN H ANNEDER 9
Locating the Moks.opāya

WALTER S LAJE 21
The Moks.opāya Project III:

Manuscripts from the Delhi and Srinagar Collections
WALTER S LAJE 37
The Moks.opāya Project IV:

Manuscripts from Pune, Wai and Baroda
P ETER S TEPHAN and S USANNE S TINNER 55
The Critical Edition of the Utpatti-Prakaran.a: A Research Report

P ETER S TEPHAN 75
Sāras, Sam
. grahas und ‘Laghus’: Kurzfassungen des Moks.opāya
S USANNE S TINNER 91
The Moks.opāyasaṅgraha
J ÜRGEN H ANNEDER 105
Die persischen Übersetzungen des Laghuyogavāsis..tha

H EIKE F RANKE 113
The Metaphorical Logic of the Moks.opāya

B RUNO L O T URCO 131
Appendix
A list of manuscript sources for the Moks.opāya 139
Illustrations 141
Bibliography 144
Index 154
Vakatseite
The Moks.opāya: An Introduction
J ÜRGEN H ANNEDER
The research activities on the Moks.opāya (MU) and related texts in the last
decade have produced interesting results: a large number of manuscripts
could be examined through which the history of the transmission became
more transparent; new versions were found and the date and localization of
the earliest text, the Moks.opāya, could be settled. Some of these results are for
the first time presented in the subsequent articles. But also for those not di-
rectly involved in the field, these results when placed in a wider context can be
of value, since the processes involved are not untypical for the development
of research in historiography of Indian literature.
In the case of the Yogavāsis.t.ha (YV) scholarly research commenced soon
after the editio princeps of the text in 1880. With minor alterations and in var-
ious reprints1 this edition has become the received text, a sort of vulgate ver-
sion accompanied by the commentary Vāsis.t.hatātparyaprakāśa composed by
Ānandabodhendra in 1710.2 Few years later a shorter version of the text ap-
peared, the so-called Laghuyogavāsis.t.ha (LYV).3 It contained complementary
halves of two commentaries by Ātmasukha and Mummad.ideva. Both edi-
tions were based only on very few manuscripts, a fraction of the surviving
sources. The criteria for the selection of sources were not implausible; in both
instances it was attempted to provide the reader with a complete commentary,
and in the case of the Laghuyogavāsis.t.ha to produce the most complete version
of the conclusion of the text.
It may have to do with the high esteem of the printed word in our culture
that with the publication of a text the investigation of further sources suddenly
comes to a halt. Once in printed form, the text, although it may be hardly more
than the transcript of a single manuscript, acquires an undeserved persuasive-
ness. But being content with an edition based on two or three manuscripts,
while dozens of unchecked mss. are lying in various libraries, is as absurd as
if archaeologists had limited their excavation of a suspected site of a town to
the suburbs without trying to find the old town.
1 The Yogavāsist.ha of Vālmı̄ki with the Commentary Vāsist.hamahārāmāyan.atātparya-
prakāsha, ed. Wāsudeva Laxman.a Śāstrı̄ Pan.sı̄kar, Bombay 1911, 2 1918, 3 1937.
2 See K ARL -H EINZ G OLZIO ’s calculation of the date given in the text in G OLZIO (∗ 2005).
3 Bombay 1888, no copy of this edition could be located.
10 Jürgen Hanneder: Introduction
In the case of the Moks.opāya literature it was for a long time only
P.C. D IVANJI , who tried to return to the sources; no other scholar seemed in-
clined to go beyond the printed version. But his analysis of only a few more
manuscripts was sufficient to question the basis of all previous secondary
studies. For D IVANJI had concluded in 1939 that the LYV could not have been
based on the YV, but on a different version.4 Returning to our analogy: we
find evidence that the old town was in the north-east of the suburb, but the
archaeologists would not try to investigate the suspected new site, but argue
that this will be in vain. M AINKAR writes in 1955 that any attempt to produce
a critical edition “is not likely to give any satisfactory results”.5 The next at-
tempt to tackle the history of this unwieldy text was made by P ETER T HOMI,
which for reasons explained elsewhere,6 was unsuccessful.
Further manuscript sightings as well as the discovery of large fragments
of the commentary of Bhāskarakan.t.ha by WALTER S LAJE marked a break-
through. In 1994 he could show that, as D IVANJI had postulated, the LYV
was an abstract not of the YV, but of its older Kashmirian recension. The YV,
on the contrary, was a redacted version that presupposes both the MU and
the LYV. This Kashmirian recension was then called Moks.opāya, which is the
original title of the text.
The YV differs from the MU, apart from a large number of variant read-
ings, in that it has added a set of further frame stories in the first and last Sarga,
and that it substitutes a number of Sargas from the MU with their counterpart
from the LYV. Thus D IVANJI’s thesis proved correct. The retrospective analy-
sis shows that in some cases the inclusion of one or few further manuscripts
can devalue the printed edition of a text to such an extent that the majority
of secondary literature including the description of its religious and literary
history has to be fundamentally revised. While this may seem obvious to the
philologist, the reaction of the scientific community may not only be favorable.
4D IVANJI (1939).
5His argumentation runs as follows: “The manuscript material is scanty and is not likely
to throw any light on the evaluation of the text. Further the present Nirn.aya Sāgara text in two
volumes appears to have a certain unified character about it. The same excessively poetical style is
to be met with in all the six Prakaran.as. Similarly the same diction saturated with the Bhagavadgı̄tā
and Gaud.apāda is to be met with throughout. Finally, the same metaphysical and ethical views
are taught with a remarkable consistency. [. . . ] Thus, whatever may be the phases through which
the text has passed, the text as it is now, is a homogenous one and an attempt to have a critical
edition of the same is not likely to give any satisfactory results.” M AINKAR (1977), p. 247f.
6 See H ANNEDER and S LAJE (∗ 2005).
Jürgen Hanneder: Introduction 11
Although it is certainly wise not to reject established knowledge prematurely,

the evidence in the case of the MU is overwhelming and the consequences for
previous secondary literature are considerable.
As an example, we may mention two articles published in 1951, one by
B HATTACHARYA,7 the other by D IVANJI,8 both of which diagnose a proximity
of the YV to the monistic Śaivism of Kashmir. Other scholars since followed
this thesis, as for instance most recently F RANCOIS C HENET.9 It is indeed
possible to demonstrate that the author of the MU was aware of “Kashmir
Śaivism”, but in view of his provenance and times this is not too astonish-
ing.10 But the author quite obviously quotes or adapts what he later chooses
to dismiss or reinterpret inclusivistically. What matters most are that in those
passages where he describes his own philosophy we do not find him influ-
enced by the philosophy of monistic Kashmir-Śaivism. For instance, he uses
words like cit and prakāśa which coincide with the terminology of this philoso-
phy, but not its distinctive elements as the concept of vimarśa. The only excep-
tion, much quoted in secondary literature, are a few verses in YV 6.128, where
we read about the “threefold impurity” (malatraya), the “grace of Śiva” and the
“fall of the Śakti”. These verses, and the presence of a so-called “Śivākhyāna”,
were taken as collective evidence for an influence of Śaivism.
However, in reviewing this evidence we find that chapter 6.128 is not part
of the MU, but among those chapters inserted from the LYV, whereas the MU
has nothing to correspond with this. Also the next argument, the Śivākhyāna,
if read together with its subsequent interpretation in the text, does not support
an influence of Śaivism, because there the Śaiva elements are reduced to the
author’s acosmistic philosophy: Śiva appears in the form of Ākāśabhairava,
which from the background of the MU means that he is no more than an unreal
appearance within empty space.11
The two results of this observation are that we have to study the older
version of the text and that we cannot rely on compilations of philosophical
7 B HATTACHARYYA (1967).
8 D IVANJI (1951).
9 C HENET (1998-99).
10 See below, p. 21ff. We may add that he quotes the Spandakārikās and the Vijñānabhairava,
quotations which remained undetected by the truly astonishing efforts to find references to Indian
literature by ATREYA, R AGHAVAN and M AINKAR, but this only shows that standard quotations
from Kashmirian works were not as much on the mind of these Pandits as, for instance, Advaita
Vedānta.
11 Compare H ANNEDER (2003).
passages12 – truly tempting in view of the length of this text – but have to read
in context. As an example why only a critical edition of the MU grants access
to the thought world of the author we may introduce verse 3.66.14, which runs
as follows in the YV:
cidghanaikaprapātasya rūd.hasya parame pade

nairātmyaśūnyavedyādyaih. paryāyaih. kathanam
. bhavet (3.66.14)
The commentator Ānandabodhendra explains that the genitives refer to

the mind (citta), which when concentrating only on the “cidghana” attains to
the highest state and that one may describe “such a mind with synonyms as
‘emptiness from an own-being’ (nairātmyam . = svarūpaśūnyatā) or ‘free from
objects’ (śūnyavedya = nirvis.ayatā)”.13 However, when reading the passage in
context, one is at a loss to see why the commentator suddenly introduces the
“mind”. The two preceding verses read as follows:
cidghanenaikatām etya yadā tis.t.hasi niścalah.

śāmyan vyavaharan vāpi tadā sam . śānta ucyase (3.66.12)
tanv¯ı cetayate cetyam. ghanā cin nāṅga cetati
alpaks.¯ıvah ks.obham eti ghanaks.¯ıvo hi śāmyati (3.66.13)

12a naikatām Ś1 Ś3 Ś7 ŚSam . ] na katām Ś9 12b tis.t.hasi Ś1 Ś3 Ś7 ŚSam
. ] tis.t.hati Ś9
12d ucyase Ś1 Ś3 Ś7 ŚSam . ] ucyate Ś9
If you, having become one with the mass of consciousness, remain

motionless whether you are pacified or even active, then you [may be]
called ‘completely pacified’.
Consciousness causes the cognition of objects [only when] subtle;
[when] dense, it does not cognize – for when half-drunk, [a person] be-
comes agitated, fully drunk he becomes silent.
Here the topics are the pacified person (sam . śānta) and consciousness in
a dense state, but not the mind. The problem for Ānandabodhendra seems
to be that the description as nairātmya and śūnyaveda does neither fit natural
persons, nor consciousness, which for the Vedāntin is hardly ‘nirātman’. An-
other oddity in his interpretation is that as synonyms for the description of
12 As for instance the Vāsis.t.hadarśana by ATREYA (1936).
13 tathāvidhasya cittasya nairātmyam . svarūpaśūnyatā śūnyavedyam
. nirvis.ayatetyādiparyāyaih. .
the mind, we have an abstract noun nairātmya compounded with the adjec-
tive śūnyaveda. The mind may be called śūnyavedya, which is however not a
synonym (paryāya), but a description of its state; the case of nairātmya is differ-
ent, for surely the mind is not referred to with the synonym nairātmya, what
the commentator means is that it is inactive and in this sense nirātman.
The problems can be solved by a glance at the Moks.opāya, where the verse
appears – in the mss. available at present – without any variant readings in
the following form:
cidghanaikyam . prayātasya rūd.hasya parame pade
nairātmyaśūnyavādādyaih paryāyaih kathanam. bhavet (3.66.14)

We may now interpret the first part of the sentence in connection with
verse 12: “For a person [as yourself],14 who has attained to the unity of the
mass of consciousness and is [thus] grounded in the highest state [. . . ]”. The
minimal difference in readings permits a plausible contextual interpretation.
In the next line the variant readings seem again inconspicuous, but have
more far-reaching consequences: we merely have to read vāda for vedya. Then
the compound in pāda c has to be understood as “through nairātmyavāda,
śūnyavāda and other [doctrines]”. Since we seem to be talking about a person,
kathanam with the genitive may denote either a description, or an instruction
of a person. The doctrines mentioned suggest a Buddhist context, thus the
following paryāya may also be interpreted as “[mode of] instruction” in a doc-
trine, as in dharmaparyāya.15 The main question, which can not be solved in the
narrow context, is whether a person is more likely to be described, or taught,
with the help of the nairātmyavāda. In general one would expect the latter,
but then it would imply that someone who has already attained the highest
state, in which he has become one with pure consciousness, is still in need of
instruction. We shall return to this question at the end of this article.
By extent the variants between the two versions are quite unspectacular,
but especially when taken together with the tendentious interpretation by
Ānandabodhendra the difference in meaning is hardly trivial – a constellation
not untypical for the Moks.opāya. When editing portions of the text, one is often
curious to see how the commentator manages to defuse too obviously hetero-
dox passages and expressions. Of course, Sanskrit being a flexible language,
14
Rāma is addressed in verse 12 and again in vss. 18–19.
15
Compare F RANKLIN E DGERTON: Buddhist Hybrid Sanskrit Grammar and Dictionary. Vol. 2,
New Haven 1953, sub voce.
he may explain to the astonished reader that a ‘bhiks.u’ in a ‘vihāra’16 is not a

17
Buddhist monk in a monastery, but a parivrāj, i.e. a sam . nyāsin in a garden.
We can infer a strong motivation on the part of the redactors to reinterpret
passages like the one just described, perhaps not even consciously, because
it was surely unthinkable that the R.s.i Vasis.t.ha could have taught heterodox
doctrines in a text that was already quoted by Vidyāran.ya as authoritative.
But, to be exact, Vidyāran.ya quoted merely the Laghuyogavāsis.t.ha, in which
many of the problematic passages had not been included.
After this excursus we return to the history of the text and its versions.
The Moks.opāya was composed in Kashmir near 950 A.D., apparently in the
Ks.atriya rather than the Brahminical milieu, since its professed aim was to
provide a secret lore for kings (rājavidyā) that would enable them to attain a
liberation in life (j¯ıvanmukti) amidst their duties. There are some testimonies
of such an instruction of Indian royals, both “Hindu” and Muslim, with the
help of the Moks.opāya.18
The further transmission and reception of the text is subject to several ten-
dencies. Firstly a variety of abridged versions were produced, some that re-
tain the original structure and character of the work with its blend of narra-
tives and philosophical discourses, as for instance the LYV, and others that
are extracts mainly of the philosophical portions. Of the latter there is the
19
Moks.opāyasam . graha, which transmits the philosophical discourses almost
completely, but leaves out almost all narratives. One extreme case is the brief
Vāsis.t.hasāra,20 of slightly more than 200 verses, which one regularly encoun-
ters in mss. catalogues. During the search for manuscripts of the MU a few
other versions were accidentally uncovered.
16 bhiks.uh. vihārasthah. (6.67.37).
17 Here Ānandabodhendra’s interpretation is simply ahistorical: While for a Vedāntin of the
18th century the title “bhiks.u” usually refered to sam
. nyāsins – and this preconceived interpretation
excluded the possibility that the word vihāra could refer to a Buddhist monastery – in tenth cen-
tury Kashmir the combination of the two words obviously referred to a Buddhist monk. These
interpretations can of course only be gained by ignoring both the narrow and the wider context.
Any careful reader not sharing the same preconceptions, or subscribing to the Vedāntic method of
reinterpretation, cannot fail to note that even the YV-version does not fit well with Vedānta. For
instance A RJUNWADKAR has noted with amazement that the author “never once utters the word
sam. nyāsa, renunciation, in the vast expanse of the work, although the concept is inseparable from
Upanis.adic thought.” A RJUNWADKAR (2001), p. 217.
18 See H ANNEDER (2003).
19 Only a single manuscript (Göttingen Ms. Sanscr. Vish. 126) is known, see below, p. 105ff.
20 Edition: T HOMI (1999). For the title, see below, p. 39.
Two charts in the appendix (p. 141f) give a chronological overview and
depict the relationship between the versions according to the present state
of knowledge. We can see that the YV presupposes the MU and the LYV,
21
but many details concerning the minor abstracts, as the Vāsis.t.hasam . graha or
22 23
the Moks.opāyasāra still need to be determined; the voluminous Vāsis.t.ha-
candrikā24 still needs to be analysed in detail.25 The later history of the
Moks.opāya literature26 is therefore more varied and complex than was previ-
ously known. In addition to this, the history of the early Persian translations
opens up a new area of research.27
Furthermore the ascetic tradition of the Advaita-Vedānta incorporated
the LYV by quoting it as a source: In the fourteenth century Vidyāran.ya
utilizes the text as a crucial source for the idea of a liberation in life in
his J¯ıvanmuktiviveka,28 but with the considerable change in the concept of
j¯ıvanmukti from an active to an ascetic one.29 Placed within this context the
LYV seemed to be Vedāntic and also the longer version was reworked partly
along these lines: Put into the ‘right’ perspective through an additional set of
frame stories, which lift the initial problem of how a Ks.atriya, who has insight
into the futility and even inexistence of the world, can still do his duty and
fight the enemy, into a Brahminical discourse of knowledge versus (ritual) ac-
tion, the text then seemed to address the Brahmin householder. In this process
Buddhist associations were removed, references to Śruti or Vedānta carefully
added, and difficult passages simplified. The end result is a sometimes so-
called Yogavāsis.t.ha-Mahārāmāyan.a accompanied by a Vedāntic commentary.
The selection of this version for publication at the close of the 19th century
has conserved this stage in the development of the text. The printed version
circulated through the subcontinent, into the text’s homeland Kashmir and
21 This text survives in one manuscript (Bodleian Library, CSS d559) of 165 fols., and is divided
into Sargas.
22 There are two known manuscripts of this text, which is divided into adhyāyas. The more
complete manuscript has 116 fols.

23 See also below, p. 91ff.
24 Described in M ITRA (1871ff), p. 268f.
25 With Ātmasukha’s commentary on the LYV (“Vāsisthacandrikā”) and another text, called here
..
Śrı̄vāsis.t.hacandrikā (see below, p. 92), there are apparently three texts of the same name.
26 See the article by S USANNE S TINNER in this volume.
27 Compare the article by H EIKE F RANKE below (p. 113ff).
28 Compare R AGHAVAN (1939b).
29 S LAJE (1998).
was there even transcribed into Śāradā script.30 With the triumphant progress
of the YV with Ānandabodhendra’s commentary the text was perceived as
Vedāntic – naturally the more critical readers remained puzzled about the ab-
sence of Vedāntic terminology.31
In Kashmir, where the MU is still transmitted intact, although some-
times with readings added from the YV version, the well-known Śaiva au-
thor Bhāskarakan.t.ha wrote a commentary on the MU, of which large frag-
ments survive.32 This commentary is neither Vedāntic, nor Śaiva, as PANDEY
thought,33 but testifies to a Kashmirian tradition, which considered the work
written by a human author. This, together with the data collected during the
editorial work, opened a new view on the text.
WALTER S LAJE will deal with the localization of the MU below, here it
may suffice to summarize the present state of knowledge on the date of the
text.34 Many arguments based on the printed edition that were previously
brought into play by various scholars have turned out to be irrelevant, because
they were not contained, or not in the form necessary for the argument, in the
MU. For instance R AGHAVAN’s deliberations on the type of recension of the
Bhagavadg¯ıtā used in the YV turned out to be inapplicable to the MU, because
there almost all relevant passages read with the Kashmirian recension of the
Bhagavadg¯ıtā. Furthermore, early references to the text as “Moks.opāya” had to
remain undetected as long as this was not considered the name of the text.
If one leaves out the irrelevant arguments, the following picture ensues:
The terminus post quem can be determined by the reference to king Yaśaskara,
who ruled Kashmir from 939 to 948.35 Even if we regard, if only for testing
the argument, this episode as a later insertion, we cannot place the terminus
post quem much lower, since the MU quotes Ānandavardhana and the Span-
dakārikās.
30 See ms. Ś17 described below, p. 46, in the article of WALTER S LAJE.
31 This, by the way, has hardly changed. In 2001 A RJUNWADKAR is formulating a critique of
the YV, which to his mind as a Vedāntic work has failed to remain in the right track: “He has
equated concepts from odd sources, e.g. Śūnya from Buddhist philosophy with Brahman from
the Upanis.ads, so that protagonists of these concepts would shudder if they knew whom they are
bracketted with.” A RJUNWADKAR (2001), p. 217.
32 See bibliography and page 5 (fn. 2) for publications of these fragments.
33 See PANDEY (1963), p. 265.
34 For details, see H ANNEDER (2003).
35 See below, p. 24.
The determination of the terminus ante quem is based on quotations and

references that were not yet discussed in secondary literature. The first oc-
curs in Ks.emendra’s (ca. 980–1060)36 Kavikan.t.hābharan.a, where he enumerates
works and topics a good poet should be acquainted with. Here the Moks.opāya
follows upon the Mahābhārata and the Rāmāyan.a. One might assume that this
refers to the “ways to liberation” in general, but the example verse is in accor-
dance with the MU and the placement after the epic from which it derives its
frame story suggestive.
An even more narrow time span for the composition results from a quota-
tion of verses from the MU in Rāmakan.t.has Sarvatobhadra, a commentary to
the Bhagavadg¯ıtā.37 There are at least three Kashmirian authors of that name
before the eleventh century, two in a well-known family of Śaiva Siddhāntins;
then the author of the Spandavivr.ti,38 also called Rājānaka Rāma, as well as the
author of the Sarvatobhadra. The last two are according to the editors identical
because of parallels in the two texts.39
If we now try to interpret the biographical data contained in the two works,
we arrive at the following: Rāmakan.t.ha mentions Utpaladeva as his teacher
and the poet Muktākan.a as his older brother. Assuming that Utpaladeva lived
between 900 and 970,40 Rāmakan.t.ha’s reference to him should not have been
made too early in his suspected life-span, perhaps not before 940. The date
of Muktākan.a, who is associated with the Kashmirian king Avantivarman
(855-883) can be brought into agreement, although with some difficulties. If
the Sarvatobhadra was a work written by Rāmakan.t.ha at an advanced age of
sixty in 940, and if his brother was twenty years older – he would have been
born in 860 – then Muktākan.a would have been 23 at the end of Avantivar-
man’s time. The alternative to this model calculation, which is perhaps un-
usual but not impossible, would be to lower the date of Utpaladeva, or deny
the identity of the author of the Spandavivr.ti and the Sarvatobhadra.
For the date of the MU the difference is only slight. The testimony
of Ks.emendra brings us to a terminus ante quem of around 980, unless we
assume that the work was written during Ks.emendra’s lifetime, whereas
36 See K ANE (1987), p. 265f.
37 The Bhagavadgı̄tā with the commentary called Sarvatobhadra by Bhat.t.a Rāmakan.t.ha. Ed.
M ADHUSŪDAN K AUL. Bombay 1943 (KSTS 64).
38 The Spanda Kārikās with the Vivrti of Rāmakantha. Ed. J.C. C HATTERJI . Srinagar 1911
. ..
(KSTS 6).
39 Introduction to the edition of the Sarvatobhadra, p. 10.
40 See T ORELLA (1994), p. xx.
Rāmakan.t.ha’s quotation pushes down the terminus ante quem even nearer to
Yaśaskara. In this case the episode in the Sthitiprakaran.a, in which Yaśaskara’s
minister Narasim . ha is mentioned as reciting one episode from the text, be-
comes crucial for determining the localization and authorship of the work.41
The implications for a general view of this text are interesting: Before these
findings the length of the text – the MU is larger than the Rāmāyan.a – its rep-
etitiousness, and the double end did not favour the assumption of a single
authorship. As VON G LASENAPP said:
“Am Ende des ersten Teils des VI. Buches, Kap. 127, hat Rāma aus den
‘den Vedānta zusammenfassenden’ (vedāntasam . graha VIa.127.3) Vorträgen
Vasis.t.has so viel gelernt, daß er in tiefer Meditation der Wonne der All-
Einheit teilhaftig wird. [. . . ] Man sollte erwarten, daß damit die Unter-
weisungen Vasis.t.has ein Ende gefunden hätten und nur noch der Ab-
schluß der Rahmenerzählung bevorstehe. Dies ist aber keineswegs der
Fall. Der redselige Vasis.t.ha setzt vielmehr seine Darlegungen in dersel-
ben Weise noch die 214 Kapitel des 2. Teils des VI. Buches hindurch fort
. . . ”.42
With the new data the picture has changed. The end of the pūrvārdha of the
Nirvān.aprakaran.a in the YV is an addition, whereas in the MU the last Prakaran.a
forms one continuous text. But when VON G LASENAPP notes that Rāma has
already reached his religious goal this is also wrong in another respect: Rāma’s
awakening takes place, undetected by previous studies, at the beginning of the
Nirvān.aprakaran.a, which is approximately the middle of the whole MU. And
this is clearly intended. According to the author this awakening is the pre-
requisite and marks the “time of the Siddhānta” (siddhāntakala), where Rāma
is able to understand his philosophical instructions and is thereby liberated.
Before this point in the text Vasis.t.ha even postpones questions, because the
disciple is not yet able to understand their answer. In one case this didactic
structure could be verified by tracing the rephrased repetition of the ques-
tion again later in the text.43 The reference to the earlier passage, removed by
many thousand verses, suggests that the didactic plan involved in this work
41 See below, p. 24.

42 G LASENAPP (1951), p. 263f.
43 See H ANNEDER (2003).
is elaborate and is more likely caused by a single author than in a long phase
of textual growth.44
This observation may also explain the verse analysed above. It is now clear
that a person who has attained to the supreme state, still needs to be taught
through a final doctrine, there called nairātmya- or śūnyavāda.
In the end not much remains from the wide-spread picture of a “philo-
sophical Rāmāyan.a”, containing a hotchpotch of ideas. The plan of the work
and its use of others’ ideas seem well calculated and this tight construction
and the narrow time frame for its production suggests that it was written or
at least substantially redacted by a single author. His ideas were so unusual
that he was received only in the garb of Advaita Vedānta, but this makes him
all the more interesting for research in the history of Indian philosophy.
44 Pending further studies this cannot be applied to the first two Prakaranas, which are – at least
.
in some parts – introductions that were composed after the completion of the main work, i.e.
Prakaran.as 3–6.
Vakatseite
Locating the Moks.opāya
WALTER S LAJE
Since the early days of Yogavāsis.t.ha studies many features have become
known, which closely connect this text with the Himalayan region, more pre-
cisely with the Valley of Kashmir, and with one of the country’s medieval
rulers.1 It was obviously there that the text must have been composed and
been given its earliest shape that can be reconstructed at present. By ‘re-
construction’ I mean to say, by way of preparing a critical edition based on
manuscript evidence of direct or indirect Kashmirian provenance. Interest-
ingly, local Kashmirian texts that refer to the work under consideration here
do so by the title of Moks.opāya or Moks.opāya-Śāstra, not by Yogavāsis.t.ha.2 This
title of Moks.opāya fully agrees with the self-referential usage in the oldest lay-
ers of the text.
The arguments brought forward so far in favour of a Kashmirian origin
were based on scattered statements in the Moks.opāya. They are mostly of
botanic and climatic facts quite typical of Kashmir. Further testimony dis-
plays unambiguous geographic and historical knowledge of the region. Nu-
merous quotes from the Rāmāyan.a (in the Vairāgyaprakaran.a) and from the Bha-
gavadg¯ıtā (in the Arjunopākhyāna), which clearly reflect readings characteristic
of the Kashmir recensions of both these works, also point to such a local ori-
gin.3 In summary, it may be said that frequent mention is made of birch-bark
(bhūrjatvac), of snow and snowflakes (himakan.a), and that many similes refer
to freezing water and excruciating icy cold. Shared experiences of that sort
were therefore clearly presupposed as common on the part of his audience by
the author of the Moks.opāya.
In one passage, the author shows his acquaintance with the outward ap-
pearance and the interior of a Buddhist monastery in Central Asia. From
1 See, e.g., D IVANJI (1935); S CHRADER (1934), p. 643; H ACKER (1951), p. 162.
2 Except for Sadānanda Yati (17th century) in his Advaitabrahmasiddhi (Ed.: Bibliotheca Indica.
New Series, Calcutta 1890). He seems to be the first author from, not necessarily in, Kashmir
(‘Kāśmı̄raka’) quoting passages under the title of ‘Vāsis.t.ha’ and ‘Yogavāsis. t.ha’, which however, re-
flect the author’s acquaintance with the LYV abridgement. See S. S TINNER’s contribution, below,
p. 91ff., for more details.
3 Collected evidence of that sort will be found treated more elaborately elsewhere, see S LAJE
(1990), p. 151, fn. 32; S LAJE (1994), p. 172ff.

22 Walter Slaje: Locating the Moks.opāya
his position, it was situated in “the northern direction” (uttarāśā), in the “big
Country, called Cı̄na”:
aham . yāto [. . . ] (6.70.7bc Ś1 , ∼ YV 6.66.7bc)
. [. . . ] uttarāśāntaram
“I betook myself to another direction, situated in the north [. . . ]”
c¯ınanāmātha tatrāsti śr¯ımāñ janapado mahān (6.70.8ab Ś1 , ∼ YV 6.66.8ab)
“Now, in this [northern direction] there is the big [and] splendid

Country, going by the name of Cı̄na.4 ”
A Central Asian itinerary,5 drawn up during king Abhimanyu’s rule (22nd

December 958 until 13th October 972),6 and thus almost contemporary with
the composition of the Moks.opāya (during or soon after Yaśaskaradeva’s reign,
AD 939–948), contains an interesting description of the way from the ’Coun-
try’ (janapada) in Central Asia to Śrı̄nagar (Adhis.t.hāna), and of Śrı̄nagar in
particular.7 The information contained in this itinerary explains the route con-
necting Śrı̄nagar with Central Asia as used in the 10th century.8
4 ‘Cı̄na’ is the reading of the Kashmirian recension only. The two occurrences of cı̄na in the
Moks.opāya have suffered corruption in the vulgate to the extent that they are no longer dis-
cernible as a geographical reference (∼ NEd 6.66.8ab and 6.67.1d / Ś1 6.70.8ab; 71.1d). ’Cı̄na’ was
used not only for ’China’, but also for Central Asia in a more general way. For further references
cp. S LAJE (1994), p. 174ff.
5 S KJÆRVØ (2002), p. 524–526 (text and translation).
6 Calculated by K.-H. G OLZIO .
7 S KJÆRVØ (2002), p. 524–526, ll. 30–35. It mentions Abhimanyu (‘Abimanyagaupta’), Śrı̄nagar
(‘Adis.t.am . ’) and the Vitastā river (‘Vı̄ttasa’) by name. I am very grateful to Professor VON
H INÜBER for having drawn my attention to this contemporary witness of the area and for having
kindly provided me with an analysis, from which ("Nr. 36: Das sakische Itinerar") the following
quote is taken: "Die Beschreibung des Weges aus "dem Lande" (janivi: janapada) in Zentralasien
nach Śrı̄nagara entstand zur Zeit des Königs Abhimanyugupta († 972). Sie hat sich in einer einzi-
gen Handschrift erhalten, die in Dun Huang in der Bibliothek in Ch‘ien Fo Tung gefunden wurde.
[. . . ] Verlauf des Weges etwa aus dem Gebiet von Kashgar (?) über Sarikol [. . . ], durch den
Wachan über Pässe in das Ishkoman-Tal und schließlich entlang Ishkoman(?)- und Gilgit-Fluß
[. . . ] nach Gilgit [. . . ] Von dort nach Süden führt der Weg nach Indien. Der nächste [. . . ] Ort ist
Chilas [. . . ] am Ufer des "Goldenen Flusses" [. . . ] also des Indus." (H INÜBER (2004), p. 74ff).
8 It should perhaps be recalled here that D IVANJI (1935), p. 28, fn. 2 has made the following
observation with reference to the Moks.opāya: “The descriptions of some of the places occurring in
some of these episodes, particularly those relating to Kashmir and those near Mt. Kailāsa in Tibet
are so minute and elaborate that they can reasonably be inferred to have been written only by one
residing at a place from where he could easily have gone to any of them and from where he could
easily have gathered the necessary information. Such a place is none other than the province of
Kashmir."
Walter Slaje: Locating the Moks.opāya 23
That the detailed descriptions of the author of the Moks.opāya cannot be

explained but by visual perceptions made by himself, will become clear also
from the following examples. In particular, he speaks of a monastery (vihāra)
there, built on the upper side of a ’termites’ nest’ (valm¯ıkopari), giving shelter
to many people:
. śrayah. (6.70.8cd Ś1 , ∼ YV 6.66.8cd)

valm¯ıkopari tatrāsti vihāro janasam
“In this [country] there is a Buddhist monastery on a termites’ nest,

with many people living in.”
Since the term valm¯ıka stands also for ‘mountain’ and ‘stūpa’,9 the word
may quite well connote architectural components of a rock monastery carved
out of the sandy rocks of Eastern Turkestan (modern Sinkiang), resembling a
termites’ nest both in colour and in form.10
An ‘upright’ Buddhist monk (subhiks.uka), who used to meditate there in
a cell locked up by a wooden bolt (argala), is depicted by the author of the
Moks.opāya as red-haired (kapilamūrdhaja11):
tasmin vihāre svakut.¯ıkośe kapilamūrdhajah. | bhiks.ur [. . . ] (6.70.9a–c Ś1 , YV

6.66.99a–c)
“In this monastery, inside his own cell [. . . ] the red-haired monk [. . . ]"
9 Cp. K ÖNIG (1984), p. 19ff; 89.
10 “Some [termites’ nests] have chimneys and pinnacles. Longitudinal and horizontal chambers
and galleries comprise the interior. Generally the outer wall is constructed of hard soil material,
[. . . ] three to four metres [. . . ] high, 2.5 metres [. . . ] wide, and one metre [. . . ] thick at the base.”
(Encyclopedia Britannica, 2001 CD-ROM Edition, s.v. Termites). See D URKIN -M EISTERERNST
ET AL . (2004) (pl. 10 [Yarkhoto] and pl. 14 [Subashi]) for illustrations of remains of monasteries
and Stūpas, which clearly resemble termites’ nests. Although I am ready to admit a methodologi-
cal flaw in comparing the appearance of contemporary termites’ nests to dilapidated monasteries,
the general impression the architectural structure and sandy colour of a Turfan Vihāra may have
made on a Sanskritic visitor could quite reasonably have resulted in a designation as ‘valmı̄ka’.
11 It is not likely that mere use of Henna could have induced the author to qualify the monk as
√
"red-haired". In such a case, we would expect a reference to dyed (e.g. raj) hair. In its absence,
it is rather an emphatic hint at a rare, though natural colour, which the author has put to the fore.
Redheads are testified to also elsewhere, as Professor VON H INÜBER pointed out to me in a letter
dt. October 8, 2004: "Denn ich glaube nicht, daß man tief in Zentralasien suchen muß. Auch in
‘meinem’ Nordwesten gibt es gelegentlich rote Haare und sogar Bān.a sagt in seiner Kādambarı̄
von einem Asketen: uttaptalohinı̄nām [. . . ] jat.ānām, Kād 78,5 (NSP 9 1948 = R IDDING p. 35 "matted
locks [. . . ] red as heated iron")." Cp. also Mahābhārata 1.100.5ab, the kapilā jat.ā of Vyāsa.
dr.d.hārgalam. gr.ham. dhyānabhaṅgabh¯ıtā viśanti no

bhr.tyāh., prāyah. kila tathā sa tis.t.hati subhiks.ukah. (6.70.10c–11b Ś1 , ∼ YV
6.66.10c–11b)
“Being afraid of interrupting [his] contemplation, the servants would

never enter the [monk’s] cell, [locked] with a firm wooden bolt.12 Indeed,
this virtuous monk remains mainly in such a condition [of contempla-
tion].”
Mural paintings with portraitures of red-haired and green- or blue-eyed

persons were found in Turfan by L E C OQ, who has published facsimile plates
prepared from them.13 In this connection, let it be noted that until today Kash-
mir seems to be one of the few regions of India where blue-eyed natives can
be met with, and not infrequently at that.14
Apart from that, it has not escaped the attention of scholars that the
Moks.opāya mentions the Kashmirian king Yaśaskaradeva,15 who was a Brah-
min and reigned the country for 9 years and one month in the first half of
the tenth century, precisely from June 26, AD 939 to July 27, AD 948.16 P. C.
D IVANJI was the first to point out the textual reference to Yaśaskaradeva and
to draw some conclusions that would follow from it.17 D IVANJI was thus lead
to the opinion that “the work under consideration must have been composed
by a poet and philosopher living in Kashmir for a long time, whether born or
settled there.”18
This, however, is certainly not the place to discuss a closely related ques-
tion, namely whether the Yaśaskaradeva-passage19 under consideration might
have been interpolated only later. From the contextual point of view, if we just
consider the complex parabolic episode, where Yaśaskaradeva is connected
12 On the nature and the use of keys and wooden bolts in Buddhist monasteries, cp. H INÜBER
(1992), p. 18f; 22ff; 30–34.

13 L E C OQ (1913). p. 4; cp. plate 17 (Bäzäklik) and note 9; L E C OQ (1928), plate 20A (comment
on p. 84).
14 In September 2003, I myself chanced upon at least three light blue-eyed persons in Śrı̄nagar.
All of them identified themselves as of Kashmirian stock. One was the keeper of Zayn al-Ābidı̄n’s
tomb and its surrounding graveyard, the second was a member of the security staff in the Śrı̄nagar
airport, and of the last, I failed to note down his profession.
15 Cp. Rājataraṅgin ı̄ 5.469–6.114 (see S TEIN (1900)).
.
16 Calculated by K.-H. G OLZIO .
17 D IVANJI (1935), p. 21f; D IVANJI (1938), p. 29; 38f; 44.
18 D IVANJI (1935), p. 28.
19 MU 4.32.16.
with a prophecy, in its entirety,20 it becomes difficult to advance objective

reasons in support of such an assumption. In fact, today we not only have
access to the Kashmirian recension of this lengthy episode, but we are also
in a favourable position to take on such an investigation, since we can avail
ourselves of a commentary on the Moks.opāya written by a Kashmiri author
(Bhāskarakan.t.ha’s T.¯ıkā). Incidentally, this passage contains unique traces of
also the authorship of the Moks.opāya and therefore deserves a thorough anal-
ysis and evaluation in a future publication.
For the time being, however, I shall content myself with a few observations
on only some topographical references made in this chapter, which prove be-
yond doubt its author’s intimate knowledge of Kashmir and in particular his
familiarity with the locality of ancient Śrı̄nagar. With the exception of those in
D IVANJI ’s articles as mentioned above, the toponymic references to be dealt
with here have never been recognized as local names by scholars concerned
with Yogavāsis.t.ha studies, and still less have those scholars made any attempts
at identifying or localizing them. Although they are not many in number, they
are decisive enough for narrowing the Moks.opāya’s place of origin – or at least
of its early recitation – down to Śrı̄nagar. For achieving the present aim, it
will – and, because of the space reserved for it, must – suffice to consider the
place names as such, without making them the subject of an investigation of
the context of the entire parable they appear in.
In short, the fate of three demons (Dāma, Vyāla, Kat.a)21 is placed in the
narrator’s contemporary (adya), and in the local context of a lake north of
Kashmir (4.30.14; 17). Thereupon their fate is transferred to a future envi-
ronment, which coincides with king Yaśaskaradeva of Śrı̄nagar: So, “today”
(adya), the three demons live reborn as fish “in a marsh in the wilderness
of Kashmir.”22 The exact locality is specified elsewhere in two passages as
“in a marsh on the bank of the Mahāpadmasaras, in Kashmir.”23 Now, the
Mahāpadmasaras is the largest lake of Kashmir and goes today by the name
of Vular/Volur, probably derived from skt. ullola (‘waving’). It has always
20 MU 4.25.4–32.31. Passages of importance in terms of prosopographic and topographic facts
of Kashmir have left Ānandabodhendra Sarasvatı̄, the commentator on the vulgate, virtually
“speech-less”, see VTP on YV 4.32.5; 16f; 21; 25f. He finished his commentary on (AD) March
7, 1710 (G OLZIO (∗ 2005)).
21 Vasistha refers to the whole situation (4.32.2) as one having been determined in advance by
..
Yama (4.30.4–9).
22 kaśmı̄rāran yapalvale (MU 4.30.14d).
.
23 kaśmı̄resu mahāpadmasarası̄tı̄rapalvale (MU 4.31.10ab = 32.5ab).
.
been very famous in Kashmir, and a number of local texts refer to it by its
ancient Sanskrit name of Mahāpadmasaras, such as the N¯ılamatapurān.a, the
Rājataraṅgin.¯ıs, the Śr¯ıkan.t.hacarita, and many Māhātmyas. Its verbatim mean-
ing is “lake of Mahāpadma”. Mahāpadma is the name of a Nāga, traditionally
believed as residing there as the tutelary deity.24
The interesting point is that Al-Bı̄rūnı̄, in his account (AD 1030) based on
local informants, also refers to the marshy banks of the Mahāpadmasaras by
saying that “the people have their plantations on the borders of this swamp,
and on such parts of it as they manage to reclaim.”25 A UREL S TEIN made a
similar comment: “The marshes and peaty meadows merge almost impercep-
tibly into its area.”26 If we compare the three statements made on the lake’s
swamps and marshes in the Moks.opāya (X1 ), by Al-Bı̄rūnı̄ (XI1 ) and by A U -
REL S TEIN (XIX2 ), to each other, it becomes clear that they all must have been
based on independent direct perception.
The demons’ future forms of existence and their places of residence are
specified by way of a description of certain sites, which can be located in
present-day Śrı̄nagar. Thus, the demon Vyāla (4.32.15) is said to experience
his final rebirth as a sparrow (kalaviṅka), Dāma (4.32.17) as a gnat (maśaka),27
and Kat.a (4.32.20) as a young partridge (krakara). For achieving final libera-
tion, they have to live separated from each other and must each get a chance
to listen to the narrative of their previous existences, by which their true na-
ture would reveal itself.28
The topographic references are as follows:
kaśm¯ıraman.d.alasyāntar nagaram
. [. . . ] nāmnādhis.t.hānam [. . . ] (4.32.11)
“A town in the country of Kashmir bearing the name of Adhis.t.hāna.”
Adhis.t.hāna, meaning ‘residence’ or ‘capital’, is one among different des-
ignations used for ancient Śrı̄nagar since it had become the new capital.29 It
was founded by Pravarasena II by the end of the sixth century (Rājataraṅgin.¯ı
3.336–349),30 and is therefore referred to as Pravarapura or Pravarasenapura
24 For more details, see S TEIN (1900) (I), note on Rājataraṅgin. ı̄ 4.593.
25 S TEIN (1900) (II), p. 363.
26 S TEIN (1900) (II), p. 423.
27 Spelt ‘masaka’ (4.2.17b etc.).
.
28 Consisting in a total lack of any latent psychic impressions, which had previously caused
inadequate notions of possessing an individual identity (ahaṅkāravāsanā).

29 S TEIN (1900) (II), p. 362. See also above, n. 7.
30 According to W ITZEL (1994a), p. 253, “Pravarasena II is to be dated around 580 A.D.”
as well. Pravarasena’s foundation, first described by the Chinese pilgrim Xu-

anzang who visited the city and stayed there in a monastery from 631 to 633
during Durlabhavardhana’s reign (ca. 625–661),31 must be distinguished from
the ‘old capital’ in the south-east (Purān.ādhis.t.hāna, the modern village of
Pāndret.hān), a few miles away only from present-day Śrı̄nagar. It is impor-
tant to note that the new city remained for long confined to the right bank of
the Vitastā (modern Jhelum), since Pravarapura was originally built round the
foot of the Śārikā hill. The site of the royal palace also remained unchanged
until king Ananta (r. AD 1028–1063) abandoned it in the eleventh century and
transferred it to the left bank.32 Thus, Yaśaskaradeva’s residence, as he had
been reigning in the tenth century, must still have been located on the right
bank, in close vicinity to the Śārikā hill. From the evidence of various other
sources, S TEIN has clearly determined the location of the old city of Śrı̄nagar
between the sixth and eleventh centuries as “round the foot of the Śārikā”
hill.33
In the words of the Moks.opāya, then unknown to S TEIN:
pradyumnaśikharam
. nāma tasya madhye [. . . ] śr.ṅgam
. [. . . ] (4.32.12ab)
“In the middle of this [town] [. . . ] a lofty peak called Pradyum-

naśikhara.”
Pradyumnaśikhara is another name for Śārikāparvata. In the tenth cen-

tury, during Yaśaskaradeva’s reign, the relative position of the hill was there-
fore precisely the centre of the city. In the Moks.opāya, the hill’s position is
explicitly depicted as such, namely central to the town (tasya madhye). This
means, the author must have made his observations at a time when the city
did not yet extend to the left bank of the Jhelum, an extension that was
achieved only later, under Ananta’s reign in the eleventh century. From such
a later perspective – and this resembles much that of today – the Pradyum-
naśikhara would have clearly risen to the northeast.
The mountain peak is referred to by a variety of names, such as
Pradyumnaśr.ṅga (4.32.25), Pradyumnagiri (30), or Ars.yaśr.ṅga (31). Kalhan.a
connects the site fundamentally with Pravarasena’s II foundation of Śrı̄nagar,
31 Cp. also F UNAYAMA (1994), p. 370ff.
32 S TEIN (1900) (II), p. 439–451.
33 S TEIN (1900) (II), p. 445.
in about AD 580. He gives a legendary account (RT 3.339–349), accord-

ing to which the place had been pointed to by a demon, who encouraged
Pravarasena to build his new capital there at its southern slope, where a vil-
lage called Śārı̄t.aka once existed34 (RT 3.348f), which is no longer extant. Thus,
the hill initially constituted also the historic – not only the topographic – cen-
tre of Śrı̄nagar. The area walled in around its foot still bears the name of
Nāgarnagar. Different legends handed down elsewhere35 have it that the god-
dess Durgā once took on the shape of a Śārikā bird and carried the hill in her
beak from Mount Meru to its present position. She is believed having taken
her abode on the hill36 and has ever since received worship there. The peak
is therefore found referred to also by the name of Śārikā in many a textual
source from Kashmir,37 the most ancient of which was considered to be So-
madeva’s Kathāsaritsāgara (12.6.105ff) so far. Since, however, the Moks.opāya
can be assigned more precisely to the middle of the tenth century, it has be-
come our earliest source for toponymic and topographic details of the sort
just pointed out. The Moks.opāya predates Somadeva (1063/1081) by c. one
hundred, and Kalhan.a (1149/50) by c. two hundred years. Somadeva clearly
identifies the location called Pradyumnaśikhara and Śārikāparvata as identi-
cal (nāmadvaya, 111ab) and gives a slightly different reason as to just why the
goddess was dwelling exactly there. It was, according to Somadeva, with a
34 S TEIN (1900) (II), p. 442f.
35 Śārikāmāhātmya, KSS 12.6.105ff; see S TEIN (1900) (I), p. 99–103; note on Rājataraṅgin. ı̄ 3.339–
349) for a detailed representation of what is given here in a summarized form only.
36 Cp. also: [. . . ] śārikārūpādhārin ı̄ [. . . ] pradyumnaśikhare sthitā (Śārikāsahasranāma (ŚSN) 1, edited
.
as part of the Śārikāpañcāṅga as an appendix to the Devı̄rahasya, ed. R AMCHANDRA K AK and
H ARABHATTA S HASTRI , Delhi 1993 [reprint]); śilāyāh. śārikākhyāyāh. [. . . ] (ŚSN 3a); [. . . ] śilārūpāsti
śārikā (ŚSN 5d); [. . . ] śilārūpāsti śārikā [. . . ] pradyumnapı̄t.ham āśritā (ŚPA 4); etc. I am grateful
to Dr J ÜRGEN H ANNEDER for having drawn my attention to this collection of texts on the local
Śārikā tradition. Cp., moreover, Somadeva, Kathāsaritsāgara (ed. PAN. D. IT J AGAD Ī ŚA L ĀLA Ś ĀSTR Ī ,
New Delhi 1970) 12.6.116 (tatra [. . . ] sam . pūjya śārikām
. devı̄m
. [. . . ]) and the (modern) inscriptions
on the archway leading to the Pradyumnaśikhara: pradyumnaśikharāsı̄nām . matr.cakropaśobhitām |
pı̄t.heśvarı̄m śilārūpām. śārikām. pran.amāmy aham. There is another line in Persian characters be-
low the Sanskrit. The language (‘Persian’), which contains elements of Sanskrit and Urdū, is
somewhat doubtful in terms of grammar, and therefore not fully intelligible. Anyhow, it omits
the Śākta connotation inherent in the Sanskrit! I am grateful to Dr A DVAITAVĀDIN Ī K AUL (New
Delhi) and Dr H EIKE F RANKE (Halle) for their help in transliterating the line and for a paraphrase
of its probable meaning: Čakrešwarat hāǧat rawā sazō gedā-rā bādšāh | wāh wāh če lakšmı̄ thāpanā šrı̄
šārikā dewı̄ namah. “Slave and king are made (supposed?) to require (hāǧat rawā) Cakreśvar. Oh,
oh, what a place of Laks.mı̄! Homage to Śrı̄ Śārikā Devı̄.”
37 See, e.g., S TEIN (1900) (I), p. 113; note on Rājataraṅgin ı̄ 3.460.
.
view to having the hill guarded, as it was considered an important entrance

gate to the underworld (pātāla), that Pradyumna asked Durgā to reside per-
manently there. Today, the hill goes by the name of Hariparbat. Contrary to
the popular etymology, which sometimes tends to connect the present name
with Vis.n.u (Hari), hāra represents the regular Kashmirı̄ phonetic derivative of
skt. Śārikā.38 S TEIN has given the following description of the site of the Devı̄
worship: “The goddess Śārikā, which has given to the hill its name, has been
worshipped since ancient times on the north-west side of the hill. Certain
natural markings on a large perpendicular rock are taken by the pious as rep-
resenting that kind of mystical diagram, which in the Tantraśāstra is known
as Śrı̄cakra”.39 The red colour of the stone may be linked with the garments of
the Devı̄, the colour of which is described – and accordingly found painted –
40
as red: [. . . ] raktavastrām . bhaje “This [. . . ]
. raktābharan.abhūs.itām [. . . ] śārikām
is still a much-frequented pilgrimage-place for the Brahmans of the City, and
has been so probably since early times”, wrote S TEIN in 1900.41 Today, offici-
ating Pan.d.its (Kauls and Dhars) are still present at this shrine. The Pradyum-
naśikhara, praised by Kalhan.a (RT 3.361) as “the pleasure-hill from which the
splendour of all the houses is visible as if from the sky”,42 was not fortified in
pre-Mogul times and seems to have been a quite built-up area. To quote again
from S TEIN: “The eastern slopes [. . . ] are now occupied by extensive build-
ings [. . . ]. It is probable that Muhammadan shrines have taken here the place
of Hindu religious buildings [. . . ]”.43 Kalhan.a speaks of one Mat.ha that was
built for Pāśupatas on the hill under king Ran.āditya (d. c. 623?), and – most
probably on the same site there – of also two temples.44 In the Moks.opāya, the
existence of a Vihāra and of royal palaces is referred to, directly or indirectly
related to Yaśaskaradeva. In this connection, it should perhaps be mentioned
38 See S TEIN (1900) (I), p. 102; note on Rājataraṅgin ı̄ 3.339–349; S TEIN (1900) (II), p. 443 and
.
note 20.
39 It is therefore that the temple goes today by the name of "[Śrı̄] Chakreshwarı̄ Temple", too. A
Cakreśvarı̄ temple is mentioned also in the Nı̄lamatapurān.a (1015), see T OKUNAGA (1994), p. 404.
For local texts dealing with the present Yantra, of which an inscription on the wall represents an
abridged description, cp. Devı̄rahasya 12.32f and Śārikāpañcāṅga 19ff (appendix to the Devı̄rahasya,
p. 408f).
40 Śārikāsahasranāma, p. 425, v. 2.
41 S TEIN (1900) (II), p. 446.
42 S TEIN (1900) (II), p. 444.
43 S TEIN (1900) (II), p. 446.
44 Rājataraṅgin ı̄ 3.460. On Ranāditya’s date, see W ITZEL (1990), p. 35 (probable misprint 523
. .
A.D. for 623 A.D.).
that Yaśaskara was the son of one influential Brahmin named Prabhākaradeva,
treasurer and minister of a previous ruler, Gopālavarman (902–904). An as-
sembly of Brahmans elected Yaśaskara, who was renowned for his learning
(vidvān) and his eloquence (vāgmin), as Brahman king on June 26, AD 939. His
election had been under debate for several days.45
The buildings on the Pradyumnaśikhara testified to in the Moks.opāya are
the following:
tasya mūrdhni girer geham . ko ’pi rājā karis.yati

abhraṅkas.amahāsālam
. śr.ṅge śr.ṅgam ivāparam (4.32.13)
“On top of this mountain, a certain king will erect a mansion, a sky
scraping, gigantic building (/sky scraping, with a gigantic rampart), as if
it were another peak piled up on [this mountain’s] peak.”
This cannot but refer to a royal palace, clearly visible from below as over-
towering everything else on the hill. The simile of a skyscraper (abhraṅkas.a),
also used by Kalhan.a (abhram . liha, Rājataraṅgin.¯ı, 3.359a), has to be taken in
its verbatim meaning. It is well known and is found referred to by many
chroniclers from Kalhan.a to Muslim and Mogul writers that the mansions in
Śrı̄nagar were lofty, “at least five stories high, and each storey contains apart-
ments, halls, galleries, and towers.”46 Were it not for the position of the palace
mentioned, which seems to have been rather in the north-eastern direction,
and were it not for the building material, which invariantly was pine wood,
the sight of the palace from below may have presented itself to the viewer
from the west or the south similar to that of the fort on its top at present. The
text locates Vyāla, the sparrow, “in a nest of that mansion, [which is] inside a
wall fissure of the north-eastern mountain peak”.47 Elsewhere, the sparrow is
said to have dwelled “on the far edge of the Pradyumna peak”.48 From this,
we may conclude that the palace on top of the Pradyumna, erected by an un-
named king, was directly attached to a rockface in the northeast. Furthermore,
it may be assumed that at the time of Yaśaskara this one had already been de-
serted, since Yaśaskara himself dwelled in a palace different from the mansion
45 S TEIN (1900) (I), p. 103. See Rājataraṅgin. ı̄ 5.473–477.
46 S TEIN (1900) (II), p. 444, note 26.
47 grhasyeśānakon ādriśirobhittivran odare (4.32.14ab); cp. also 4.32.30cd: [. . . ] pradyumnagirau gehe
. . .
bhittivran. avihaṅgatā.
48 pradyumnaśikharaprāntavāstavyah [. . . ] (4.32.23ab).
.
just referred to, and also because the nest in the wall fissure of Yaśaskara’s
palace is described as “of straw, ruffled by incessantly blowing winds”.49
The three demons occupied different, however nearby locations.50
tasminn eva tadā kāle tatra rājā bhavis.yati

śr¯ıyaśaskaradevākhyah. śakrah. svarga ivāparah. (4.32.16)
“Then, at the same time, a king of royal dignity (śr¯ı), Yaśaskaradeva

by name, will reside there [on the hill], as another Indra in [his] heaven,
as it were.”
Yaśaskaradeva, we know from Kalhan.a, apart from his royal residence

(mandira, rājadhān¯ı) owned also a college (mat.ha), which he had built else-
where for having students from Āryadeśa (Uttar Pradesh) educated there,
and whereto he resorted in vain when his violent death was imminent.51 His
palace was equipped with a “hall of eight columns” (as.t.astambhaman.d.apa).52 It
was most likely with reference to this very hall that Dāma is said to dwell as
a mosquito “in the king’s palace (sadman), softly humming inside a crack on
the surface of a mighty column (br.hatstambha)”.53 In a subsequent passage, the
building material of the palace and its columns is even specified as “consisting
of wood, in a fissure of which” the mosquito would have been living.54 This
statement in particular justifies the assumption that we are indeed concerned
here with an eyewitness’ account of topographical facts, reported in detail by
this contemporary of Yaśaskaradeva.
Apart from the two buildings on the hill just dealt with, one on the north-
eastern peak and presumably already deserted at the time, and the other being
Yaśaskaradeva’s palace without any detailed indication of landmark particu-
lars, there is yet another topographic description made in the text, related to
one of Yaśaskara’s ministers and his place of residence:
49 [. . . ] aviśrāntavātoddhūtatr. n.āṅkite (4.32.14cd).
50 4.32.3; the three locations are 1) Pradyumnaśr.ṅga, 2) king Yaśaskaradeva’s palace, 3) the
Ratnāvalı̄-Vihāra (4.32.25).
51 See Rājataraṅgin ı̄ 6.87f; 99; 104ff.
.
52 Rājataraṅgin ı̄ 6.96b.
.
53 dānavo dāmanāmā tu masakas tasya sadmani bhavisyati brhatstambhaprsthacchidre mrdudhvanih
. . . . .. . .
(4.32.17).
54 rājamandiradārvantar vranavāstavyatām gatah (4.32.24ab). Cp. also 4.32.29cd: [. . . ] rājagrha-
. . . .
dārvantar vran.e mas.akarūpatā.
adhis.t.hānābhidhe tasminn evogranagare tadā

ratnāval¯ıvihārākhyo vihāro ’pi bhavis.yati (4.32.18)
“Then, in the same mighty town called Adhis.t.hāna, there will be a

monastery also, going by the name of Ratnāvalı̄vihāra.”
Kalhan.a, too, mentions the Ratnāvalı̄-Vihāra of Śrı̄nagar (Rājataraṅgin.¯ı

3.476). It had been constructed by Galūna, who named it after his wife. He
had been a minister of king Vikramāditya, the great-grandson of Pravarasena
II, founder of Śrı̄nagar, and son of Ran.āditya, who had built a Pāśupata-Mat.ha
on the Pradyumnaśikhara (Rājataraṅgin.¯ı 3.460). We may therefore safely as-
sume that the Vihāra under consideration, too, must have been confined to
the city. Kalhan.a informs us about another monastery there, called after its
founder Jayendra, an uncle of Pravarasena II, the ’Jayendra-Vihāra’. Xuan-
zang, who stayed in Śrı̄nagar for two years (AD 631–633), living and studying
in this latter Vihāra, confirms its existence at that time.55 In the case of the
Ratnāvalı̄vihāra, however, the Moks.opāya clearly specifies its position as on
the Pradyumna hill:
tasmim
. s tadbhūmipāmātyo narasim
. ha iti śrutah. [. . . ] bhavis.yati (4.32.19)
“In this [Vihāra] there will live a companion (/councillor)56 of that

king [Yaśaskaradeva], known as Narasim . ha.”
The mansion (as part of the Vihāra)57 owned by him58 is explicitly located
on the Pradyumnaśikhara:
[. . . ] ars.yaśr.ṅge nr.sim
. hasya gehe [. . . ] (4.32.31c)
59
“In Nr.sim
. ha’s house on the Pradyumna hill [. . . ]”.
55 Cp. S TEIN (1900) (I), p. 103, note on Rājataraṅgin ı̄ III 355; 1900, 2: 439; cp. also Rājataraṅgin ı̄
. .
5.428. The Khotanese Itinerary (10th century, see above, note 7) contains the following reference to
Vihāras in Śrı̄nagar at the time: "[. . . ] there is a large monastery with a dharmarāja (stupa) (and)
500 rock cells. Smaller monasteries are countless." S KJÆRVØ (2002), p. 526, ll. 33–34.
56 Characterised as Amātya also in 4.32.21a: sa nrsimho nrpāmātyah .
. . . .
57 It is perhaps worthy of note that the names of particular religious buildings such as Vihāras
or Mat.has would also designate the surrounding city quarters. See S TEIN (1900) (II), p. 339.
58 [. . . ] grhe tasya [. . . ] (4.32.20a).
.
59 arsyaśrṅge [=] pradyumnaśrṅge (Moksopāya-T ı̄kā ad 4.32.31).
. . . . .
As a result, we now have come to know of altogether four medieval ed-

ifices on the Śārikā hill, which are testified to as historical by the Moks.opāya
(X1 ) and by Kalhan.a (XII1 ): Two royal palaces, one of them Yaśaskaradeva’s,
the other a deserted one, furthermore the Ratnāvalı̄ Vihāra, as also a Pāśupata
Mat.ha. From this collective evidence for the existence of palaces, Mat.has and
Vihāras on the hill – not to forget the Śārikā- and other shrines60 –, it may
be imagined as a busy area, quite comparable to the Palatino crowning the
Forum Romanum in Rome. Kings, ministers, philosophers and poets – some-
times holding more than only one office – had chosen the Pradyumnaśikhara
as their residence, living there in close neighbourhood.
Exactly this, then, brings us to no one else than to Narasim . ha, who is likely
to have been involved some way or other in the Moks.opāya’s authorship.61
There are at least three learned and approximately contemporaneous individ-
uals of Kashmir known to have borne the name of Narasim . ha or Nr.sim. ha.
1
One is the Narasim . ha under consideration here (X century), the second was
the father of Abhinavagupta (X2 /XI1 century),62 called Narasim . hagupta or
Cukhala,63 and the third was a certain Ācārya Narasim . ha, an exponent of a
particular “non-difference from perception” doctrine (pratyaks.ādvaya), again
referred to by Abhinavagupta.64 Now it may be argued that Narasim . ha
of the Moks.opāya was no historical figure since Kalhan.a has not enumer-
ated him among Yaśaskaradeva’s ministers. Kalhan.a, it is true, mentions
only one “out of five ministers” (mantrin) by name, and this is Parvagupta
(Rājataraṅgin.¯ı 6.103ab). However, the latter was a politically important figure.
An exceptionally ill-disposed and deceitful person, he had risen to an influ-
ential position already under king Unmattāvanti (Avanti, the maniac), cruelly
60 S TEIN describes also another place of worship “close to the foot of the southern extremity of
the hill”, a “rock which has from ancient times received worship as an embodiment of Gan.eśa,
under the name of Bhı̄masvāmin.” This, as well as the fort on the summit of the hill, I had not
been able to visit on account of the fact that these areas were closed to the public due to military
restrictions. The fort was built only after Akbar, whereas the wall enclosing the hill was built by
the order of the latter. S TEIN (1900) (II), p. 447.
61 Cp. 4.32.21.
62 Ca. AD 950-1020 according to K ANE (1987), p. 243.
63 Cp. H ANNEDER (1998a), p. 58f; 127f; 131f.
64 See Mālinı̄vārttika (ed. M ADHUSUDAN K AUL S HASTRI , Srinagar 1921 [KSTS 31])
1.762c-764b: ittham . pratyaks. am evedam. nih.sapatnam

. vijr.mbhate [cp. 752ab] tato na bhidyate
cārthah. pratyaks.ādvaitam ı̄dr.śam | idam
. sandhānakalikāparinis. t.hitabuddhinā | ācāryanarasim
. hena
pratyaks. ādvayam ucyate. I owe this reference to Dr S OMDEV VASUDEV (letter to Dr Hanneder,
May 17, 2002).
murdered the crippled successor of Yaśaskara, and thereupon usurped the

throne.65 Furthermore, we must not expect Kalhan.a to have recorded every
official of the royal court by name. Thus, basing himself on the evidence pro-
vided by Abhinavagupta in his Parātrim . śikāvivaran.a, K ANE pointed out that
Yaśakara must also have had a minister called Vallabha.66 Next, Jayaratha in
his Tantrālokaviveka has handed down the name of yet another of Yaśaskara’s
ministers, Pūrn.amanoratha.67
So, from among altogether four names we dispose of evidence for,
Narasim . ha (MU), Parvagupta (Rājataraṅgin.¯ı), Vallabha (PTV), and Pūrn.a-
manoratha (TĀV), only one (Parvagupta) has been mentioned explicitly by
name by Kalhan.a. An objection against Narasim . ha as being historical, by way
of argumentum e silentio Kalhan.ae, would thus obviously miss the mark.68
Furthermore, it is not entirely clear whether the designation of mantrin (Par-
vagupta), saciva (Pūrn.amanoratha) or amātya (Narasim . ha, Vallabha) might
point to different areas of responsibility or to different executive posts at the
court.69 Over a period of almost ten years of Yaśaskaradeva’s continuous
reign, it is plausible that the refilling of councillor or court Pan.d.it positions
may have occurred repeatedly.
In conclusion, the historicity of the local sites and personal names pre-
served in the Moks.opāya passages as treated here may safely be considered
65 See Rājataraṅgin. ı̄ 5.420ff; 6.115–149.

66 K ANE (1987), p. 237. Cp. Parātrı̄śikāvivaran. a (ed. in: G NOLI (1985)), p. 284, vs. 5ab: kaśmı̄res.u
yaśaskarasya nr.pater āsı̄d amātyāgran. ı̄h. śrı̄mān vallabha ity [. . . ] dvijah.. I am grateful to Dr
H ANNEDER for having drawn my attention to this passage as also to the one quoted in note 67.
67 śrı̄mān yaśaskaranr pah sacivam samastadharmyasthitis v akrta pūrnamanorathākhyam. See Tantrā-
. . . . . .
lokaviveka, concluding verse 8cd. In: The Tantrāloka of Abhinavagupta with the Commentary of
Jayaratha. [Reprint] Ed. by R. C. D VIVEDI and N AVJIVAN R ASTOGI , Delhi 1987, vol. 8, p. 3720.
68 Cp. also D IVANJI (1935), p. 23, according to whom “in the familiarity which he shows in
describing the places situated [. . . ] there is sufficient warrant for the inference that he must either
be writing this account at the time when Yaśaskara was ruling [. . . ] and Nr.sim . ha was one of his
ministers or at a time when any successor of that king [. . . ] had been ruling there. I therefore
conclude that such was really the case.”
69 On the functional difference between a mantrin and an amātya according to the Arthaśāstra
cp. K ANGLE (2000), p. 133ff. Let it be noted that Bhat.t.a Jayanta, writer and philosopher, does
not seem to differentiate between these denotations. Thus, he speaks of himself as an amātya or
as a mantrin in the service of king Śaṅkaravarman (883–902): dālun.e khu lāe śaṅkalavamme. tado
viśame śe bamhan.e taśśa amacce dulāālajayam . te. Āgamad.ambara (Āgamad.ambara otherwise called
S.an.matanāt.aka of Jayanta Bhat.t.a. Ed. V. R AGHAVAN and A NANTALAL T HAKUR. Darbhanga:
Mithila Institute 1964) 3rd Aṅka, p. 46, line 18f. mantrı̄ śāstramahāt.avı̄viharan. aśrānto jayanto [. . . ]
Āgamad.ambara 3rd Aṅka, 8b, p. 54, line 20.
as established. The origin of its composition must be searched for inside the
city limits of ancient (10th century) Śrı̄nagar, on the slopes or at the top of the
Pradyumna hill.
Vakatseite
The Moks.opāya Project (III):
Manuscripts from the Delhi and Śrı̄nagar Collections
WALTER S LAJE
In continuation of two reports on ‘The Moks.opāya Project’, published else-

where1 and, as far as that goes, preceding the present one, this is to deal briefly
with some additional manuscripts transmitting the text of the Moks.opāya.
These manuscripts are kept in collections, which still are almost inaccessible to
the public.2 During September and October 2003, by permission of the respec-
tive authorities, it became possible to consult the materials on the spot and to
take notes on them. Having lain behind closed doors for a considerable pe-
riod, the manuscripts under consideration had increasingly become shrouded
in mystery. The mystery surrounding them was further enhanced by various
rumours and some title-lists put in circulation discreetly. That the hurdles
were finally taken and that transparency could be achieved, goes above all
to the credit of Professor G. C. T RIPATHI, Head of the Kalākośa Division in
the IGNCA. He spared no efforts to this end, brought his influence to bear on
wherever necessary and presided over the whole campaign as a vighnajit in
the true sense of the word.
National Archives, New Delhi

The Manuscripts Department in the National Archives keeps a small collec-
tion of 212 Sanskrit manuscripts3 from Śrı̄nagar.4 Their titles are enumerated
in List No. 298: Gilgit Manuscripts and Sanskrit Manuscripts: “Manuscripts
belonging to the Archaeology and Research Department, Jammu & Kashmir
Government, Srinagar [. . . ] consisting of 212 items [. . . ] temporarily trans-
ferred to the National Archives of India for safe custody and preservation on
Oct., 19, 1948 [. . . ]”.
1 S LAJE (1997), S LAJE (2000).
2 Cp. W ITZEL (1994b), p. 17, for an enumeration of collections keeping Kashmirian
manuscripts.
3 Note that in the following all quotations from manuscripts strictly document the wording
exactly as found. Scribal mistakes are not marked.

4 Cp. W ITZEL (1994b), p. 17: “The first 212 numbers of this” [i.e. of the Research Library of the
University of Kashmir at Srinagar] “collection have been transported to Delhi and they have re-
mained there ever since, in the National Archives of India where they are kept almost inaccessible,
as they are classified with actual government documents.”
38 Walter Slaje: Moks.opāya Project III
Catalogue entry: “Moks.opāya, 90 folios.”
No. 178 Paper, cardboard cover, Śāradā.

Varying colophons, e.g.: moks.opāye . . . ; śr¯ımoks.opāyasāre . . . ; moks.opāya-
sam. hitāyām
. . . . nāma sargah.. Title abbreviation: mo sā. Incomplete, 95 numbered
folios. Page numbering in the margins on the top right in Roman numbers:
190. No Prakaran.a structure discernible, only Sargas are indicated.
Moks.opāya-Sāra, incomplete.
Begins:
. | divi bhūmau . . . || 1|| (= MU 1.1.1)
om
. baddho . . . || 2|| (= MU 1.1.2)
aham
yāvan nānugrahah. sāks.āj jāyate parameśvarāt |

tāvan na sadgurum . kaścit sacchāstram . vāpi no labhet || 3|| (= VS 1.3, not
in MU)
śr¯ıvālm¯ıkih.
kathopāyān vicāryādau moks.opān imān atha |

yo vicārayati prājño na sa bhūyo bhijāyate || 4|| (= MU 1.1.3)
asmin rāmāyan.e . . . || 5|| (= MU 1.1.4)

śis.yāyāsmai vinitāya bharadvājāya . . . || 6|| (= MT. (I) 1.5)
Breaks off on folio 95v :
jad.am
. cittādi duh.khasya bhājanam . dehatām gatam |
na caitasmin ks.ate ks.¯ın.e kiñcid evātmanah. ks.atam || 59|| (= NEd 6.54.10)
This abridgement in its initial part (Vairāgya) appears to be dependent on,

or, at any rate related to, manuscripts of the Vairāgyaprakaran.a (Ś12 and Ś13 ).
As Sarga 1.1 of the vulgate is lacking there, they all doubtlessly transmit the
characteristic beginning of the Moks.opāya recension (MU), with the notable ex-
ception of Śloka no. 3 of the manuscript described. This Śloka occurs normally
only in the small Vāsis.t.hasāra5 consisting of ten Prakaran.as. The volume of the
present version exceeds Sāras of the latter kind. Closer investigation required.
5 Ed. by T HOMI (1999) by the title of Yogavāsis. t.hasāra.

Walter Slaje: Moks.opāya Project III 39
National Museum, New Delhi6

Catalogue entry: “Yogavasis.t.ha-Sāra-Vivaran.am”
No. 57.106/44 Collective paper codex, cardboard cover, Śāradā.

Vāsis.t.hasāra in 10 Prakaran.as.
Contrary to the catalogue entry the manuscript contains no commentary;
a few glosses are added in the margins. As clearly indicated in Mahı̄dhara’s
commentary (Vivr.tti), written in A.D. 1597 on this abridgement,7 its original
title must have been ’Vāsis.t.hasāra’.8 Despite this, the text was re-edited in 1999
as ’Yogavāsis.t.hasāra’.9 For the sake of transparency, the very title ’Vāsis.t.hasāra’
(VS) will henceforth be consequently applied to all manuscripts of this ver-
sion, disregarding the various and often misleading designations occurring in
the colophons.
Begins 33v :
om
. namah. śivāya | om
.
dikkālādyanavacchinnānantacinmātramūrttaye |
svānubhūtyekamānāya namah. śāntāya tejase || 1|| (= VS 1.1)
aham . baddho vimuktah. syām iti yasyāsti niścayah. |
nātyantam ajño no tajjñah. so smiñ chāstre dhikāravān || 2||
yāvan nānugrahah. sāks.āj jāyate parameśvarāt |
tāvan na sadgurum . kaścit sacchāstram . vāpi no labhet || 3||
mahānubhāvāsam . parkāt sam
. sārārn.avalaṅghane |
. prāpyate rāma dr.d.hā naur iva nāvikāt || 4||
yuktih. sam
6 Cp. W ITZEL (1994b), p. 17: “Another larger collection is in the National Museum at New
Delhi; of this, there exists a card catalogue that equally remains unpublished.” I am grateful to
my friend R AFFAELE T ORELLA for having placed a digitized copy of a typewritten title-list at my
disposal in November 2002. Access to the collection was granted in a most liberal manner by
the Keeper of the Manuscripts Department, Dr N ASIM A KHTAR. I should also like to thank the
Deputy Keeper Dr S ATYA V RATA T RIPATHI for his friendly assistance on the spot.
7 Calculation according to A UFRECHT , quoted in T HOMI (1999), p. 22.
8 For a copy of Mahı̄dhara’s Vivrtti, with an excerpt of the passages dealing with the title, see
.
below, Śrı̄nagar Ms No 4813/1968.05.
9 Cp. above, note 5.
Ends 41r :
saumyāmbhasi yathā v¯ıcir na cāsti nanu nāsti ca |
tathā jagad brahman.¯ıdam . janam || 36|| (∼ VS 10.34)
. śūnyāśūnyapadam
iti śr¯ıyogavasis.t.hasāraviracite paranāmni tatvanirūpan.am . prakaran.am ||

. nāma daśakam
Catalogue entry: “Yogavāsis.t.ha, Big Grantha, 704 folios”
No. 57.106/100 [Ś16 ]10

Paper codex, Śāradā.
Moks.opāya, Nirvān.aprakaran.a (complete) and Khilas (incomplete, 1.1–
14.2a).
Obvious affinities with Ms N15 and N25 (= Viśrām II/24 and Viśrām I/419 in
the B.O.R.I collection) in extension and readings.
Begins on folio 1v :
śr¯ıvālm¯ıkir uvāca
upaśāntiprakaran.ād anantaram imam . śr.n.u |

tvam . nirvān.aprakaran.am. jñātam. nirvān.akāri yat ||
kathayaty evam uddāmavacanam . munināyake |
śravan.aikarase maunam . sthite rājakumārake ||
munivāgarthaniks.iptamanasy astetarakriye |
rājaloke gataspandam . cittrārpita iva sthite ||
vasis..thavacasām artham. vicārayati sādaram |
lasadaṅgulibhaṅge ca munisārthe sphuradbhūvi ||
End of Nirvān.a on folio 670v :
śrutvaitac ciranirvr.ttim
. bhaja bhr.śam
. j¯ıvanvimuktāśayo
laks.m¯ım
. jñānatapah.kriyākramayutām . bhuṅks.vāks.ayām aks.ayah. || (∼ NEd
7.215.17; MU 6.374.17)
ity ārs.oparacite śr¯ımahārāmāyan.e devadūtokte śatasāhasryām

. sam . hitāyām
. bālakān.d.e
moks.opāyes.u nirvān.aprakaran.am. samāptam || samāptāh. moks.opāyāh. || 371||
End of Khilas on folio 696v :
10 The numbers in square brackets correspond with the sigla now in use by the Moksopāya
.
edition project.
cetanādityatapane dr.śye smin vyomarūpin.i |

. sadasadātmani ||
ālokavedaśāstrādi svāṅgam
bhavān aham. dr.śyam idam. jagat kham ityādi sarvam . sadasatsvarūpam |
yathāsthitam
. sad vyavahāry ap¯
ı ttham
. brahmes t
.. hikaikopalakośat ūs.m¯ım ||
. nāma sargah. || 13
nānāpraśnes.u parabrahmasvarūpavarn. anam
vasis.t.hah. ||
sarvārthātmana evāsya sarvārthābhigatātmanah. |
jagac cinnabhaso brūhi kāran.am. kvopayujyate || (= Khila 14.1)
śaśaśr.ṅgakhapus.pān.ām
. va
The Research and Publications Department, Jammu & Kashmir

Government, Śrı̄nagar11
Despite many applications directed to and numberless telephone conversa-
tions held with the authorities of the “Public Libraries, Kashmir” and of the
“State Libraries, J & K”, no permission for digitizing the mss. on the spot could
be achieved at the time. It is particularly in this regard that I feel very much
indebted to Dr. A DVAITAV ĀDIN Ī K AUL (IGNCA, New Delhi) who through
her good relations with the Śrı̄nagar authorities, by negotiating for weeks on
end, advanced the process to the extent that all required mss. were eventu-
ally put at my disposal for consulting them in Śrı̄nagar. One year and a half
later, the IGNCA has now been entrusted with the task of digitizing the valu-
able Śrı̄nagar mss. collection, and their work is admirably progressing with
rapid strides. Back in 2003, the awaited decision was promised not to come
up before a couple of weeks – or even months.12 Therefore, by the shortage
of time caused by not a few delays of that sort, all that could be done at the
11 Cp. W ITZEL (1994b), p. 17: “[. . . ] the Research Library of the Department of Archaeology
at Srinagar has collected some 8000 texts (contained in a lesser number of actual MSS). They are
now housed in the Research Library of the University of Kashmir at Srinagar.”
12 Under the circumstances, one felt somewhat reminded of M. W ITZEL ’s (loc. cit.) assess-
ment that “fortunately he (scil. M. A. S TEIN) sent most of them (scil. of the manuscripts he had
bought in Kashmir) to the libraries of Vienna, Tübingen, Berlin, Paris and especially to Oxford
[. . . ] where they are kept and are accessible in original and microfilm [. . . ].” Elsewhere, W ITZEL
(W ITZEL (1990), 54f, fn. 133) had stated, “libraries and archives in South Asia are notoriously dif-
ficult of access, due to bureaucratic and other restraints. Instead of complaining about the “theft”
of mss. during the colonial period, action should be taken to save the many private collections
time was consulting the mss. quite hastily by taking rough notes only. The
remaining time had to be used for quickly identifying all of those manuscripts
that might belong to the Moks.opāya recension. Testing the Prakaran.as against
certain formal criteria allows in general for a quick identification as follows:13
– Vairāgya beginning with NEd I 2.

– Utpatti extending to NEd IV 18.
– Sthiti beginning with NEd IV 19.
– Nirvān.a undivided into two halves of a Pūrva- and an Uttarārdha.
– Khilas (nānāpraśnāh.) appended to the Nirvān.aprakaran.a.
It should be noted, however, that the following identifications as “Yo-

gavāsis.t.ha recension (NEd )” are frequently based on such superficial criteria
as are e.g. the presence of NEd 1.1 (= frame-story E) or of the commentary
of Ānandabodhendra (Vāsis.t.hatātparyaprakāśa [VTP]), both doubtlessly char-
acteristic of the vulgate Yogavāsis.t.ha (NEd ) only. From this, however, it does
not follow with any necessity that the Śrı̄nagar manuscripts under considera-
tion would not have retained particular Moks.opāya readings. It may generally
be stated that in quite a number of cases the conflation with VTP seems to be
restricted predominantly to the Vairāgyaprakaran.a, in the case of which VTP
was copied out in full. Influence of that sort decreases sometimes perceptibly
in the subsequent Prakaran.as, where excerpts from VTP occur rather occasion-
ally, in the margins only.
The following presents a complete overview of Moks.opāya/Yogavāsis.t.ha re-
lated manuscripts kept in the Research and Publications Department, arranged
according to their catalogue entries.14
Catalogue entries: “Yogavāsis.t.ha”

No 4787/361 Papercodex, leather binding, Kaśmı̄r-Devanāgarı̄.
Yogavāsis.t.ha recension (NEd ), Vairāgyaprakaran.a.
Sanskrit and vernacular translation (bhās.ā)
of manuscripts and documents in the subcontinent from destruction by the forces of nature and
their all too often ignorant proprietors. Those mss. that had been brought to European and other
non-Indian libraries have survived just because of this fact and are easily accessible to research.”
13 Cp. also S LAJE (1997), p. 211ff.
14 See "The Research and Publication Department, Jammu and Kashmir Government, Srinagar
(Kmr): A Catalogue of Sanskrit Manuscripts. Srinagar 1989."

No 4788/827 [Ś14 ] Old birch-bark codex, leather binding, Śāradā.

Moks.opāya, Vairāgya- to Upaśamaprakaran.a.
Beginning and end totally crumbled. The bark has become fragile to the
extent that turning over the folios would damage the bark and cause its layers
to coming off. Professional restoration urgently required.
Judging from my faint memory of the handwriting of an otherwise in-
complete birch-bark ms (Ś5 , Oxford), which contains the latter half (Nirvān.a-
prakaran.a) only of the complete work, the present manuscript could originally
have made up the former part. Ś14 abounds throughout in glosses written
interlinear or in the margins by different hands. First consultable folio:
mahārāmāyan.e viśvāmitravākyam
. nāma sargah. |
vālm¯ıkir uvāca |
tac chrutvā rājaśārdūlo viśvāmitrasya bhās.itam | (= MU 1.7.1ab)
1st legible colophon of the Utpattiprakaran.a:
utpattiprakaran.e ādyasr.s.t.ivarn.anam
. nāma sargah. | (= MU 3.2)
rāmah. |
evam etan manaś śuddham
. pr.thvyādirahitam. nabhah. |
mune brahmeti kathitam. sabhyam . pr.thvyādivarjitam | (∼ MU 3.3.1)
The Utpattiprakaran.a contains the Bhārgavopākhyāna. No intercalary Sthiti-
colophon. Last Sarga preceding the Bhārgavopākhyāna:
tayor dvayor manasi nirantaram . ks.ate
. jagan na tu jagati ks.ate manah. || (= MU 4.4.16cd)
ks.atam
. nāma sargah. ||
sthityaṅkurakathanam
Beginning of Bhārgavopākhyāna within Utpatti:
rāma |
bhagavan sarvadharmajña pūrvāparavidām. vara |
ayam
. manasi sam. sāras sphārah

katham iva sthitah. || (= MU 3.127.1 NEd
4.5.1)
The above quoted colophon preceding the actual beginning of the

Bhārgavopākhyāna and containing the element ’sthiti’ could – through the im-
pact of also the immediately following first Śloka with ’sthitah.’ in Pāda d) –
have led some redactor or scribe to assume that this should indeed form part
of the ’Sthitiprakaran.a’. It would have contributed to an uncertainty about the

Prakaran.a boundaries and eventually resulted in shifting the conclusion of Ut-
patti to exactly before this Ākhyāna, which, by way of an intervention of that
sort, became itself part of the Sthitiprakaran.a.15
Colophon of Utpatti:
nd
samāptam
. cedam utpattiprakaran.am (2 hand:) tr.t¯ıyam |
vasis.t.hah. |
athotpattiprakaran.ād . . . śr.n.u . . . || (= MU 4.1.1.)

j¯ıvab¯ıjam
. param
. brahma sarvatra kham iva sthiram | (= MU 4.19.1ab)
Final legible colophon:
upaśamaprakaran.e sam
. sr.tib¯ıjanirāharan.akramopadeśo nāma sargah. | (∼ NEd
5.92)
The following Sarga begins:

manāg api vicāren.a ceta — |
manāg api ks.ato yena — || (∼ NEd 5.93.1)
Final legible fragment:

j¯ıvanmuktamatir maun¯ı nigr.h¯ıtendriya — |
—madamātsarya āryas tis.t.hati vijvarah. || (∼ NEd 5.93.95)
No 4789/1149 3 independent manuscripts registered under one entry.
1 Paper, Pot.hi, Śāradā. Folios in a mess. Yogavāsis.t.ha recension (NEd ),

Vairāgya- and Mumuks.uprakaran.a. VTP glosses in the margins.
2 Paper, Pot.hi, Kaśmı̄r-Nāgarı̄ and Śāradā (glosses). Yogavāsis.t.ha recension

(NEd ), Upaśāntiprakaran.a. VTP glosses (Śāradā) in the margins.
3 Paper, Pot.hi, Śāradā. Yogavāsis.t.ha recension (NEd ), incomplete Nirvān.a-

prakaran.a. VTP glosses in the margins.
15 Compare P. S TEPHAN’s contribution below, p. 85.
No 4790/1155 [Ś21 ] Paper codex, leather binding, Śāradā. Yogavāsis.t.ha recen-

sion (NEd ), Vairāgya- to Sthitiprakaran.a. (Frame story E plus VTP introduction
to the Vairāgyaprakaran.a, VTP glosses in the margins). Bhārgavopākhyāna con-
tained in Utpatti.
No 4791/1212.06 [Ś18 ] Collective paper codex, leather binding, Śāradā.

Moks.opāya (?), Vairāgyaprakaran.a (incomplete in the beginning). Interlinear
glosses and glosses in the margins.
No 4792/1238 [Ś19 ] Country paper, Pot.hi, Śāradā.

Folios arranged back to front.
Moks.opāya (?), Nirvān.aprakaran.a (incomplete).
Breaks off with:
svatah. citi ghane cittvāc cidbhāva— (∼ NEd 7.188.16ab)
No 4793/1274 [Ś20 ] Paper codex, cardboard cover, Śāradā.

Two separate Prakaran.as bound in one:
1 Yogavāsis.t.ha recension (NEd ), Vairāgyaprakaran.a (complete, frame story E

plus VTP introduction)
2 Moks.opāya (?), Nirvān.aprakaran.a (incomplete, different hand)
Final colophon of the Nirvān.a:

moks.opāyes.u nirvān.aprakaran.e bhiks.usam
. sārodāharan.am
. nāma sargah. | (=
NEd 6.62)
śr¯ıvasis.t.hah. |
sa kadācid dadarśātha rudram . . . . | (= NEd 6.63.1; MU 6.67.1)
Breaks off with:
kvāpi j¯ıvat.asam
. sāram
. cidākāśaikakon.agam || 41|| (= NEd 6.63.41; MU
6.67.40)
No 4794/1311.04 Collective paper codex, cardboard cover, Śāradā.

Yogavāsis.t.ha recension (NEd ), parts from the Nirvān.aprakaran.a, uttarārdha
(VTP).
No 4795/1562 [Ś15 ] Old paper codex, leather binding, Śāradā.

Moks.opāya, Vairāgya- to the beginning of Nirvān.aprakaran.a (incomplete, a few
folios only).
Folio no 1 of Vairāgya has been lost. Folio no 1 is preceded by a substi-
tute enclosed in the form of an unbound quire of several folios containing
Ānandabodhendra’s VTP-introduction together with a beginning correspond-
ing to the Yogavāsis.t.ha recension. Cp. also below, "Conclusion".
End of Utpatti:
iti śr¯ımahārāmāyan.e moks.opāyes.u utpattiprakaran.e j¯ıvas.an.d.akāvatārah. nāma
sargah. | [ = MU 4.18] utpattiprakaran.am . samāptah. sam . pūrn.am
Sthiti begins with:
j¯ıvab¯ıjam
. . . . (= MU 4.19.1)
No intercalary Sthiti-colophon.
No 4796/1821.10 [Ś13 ] Collective paper codex, cardboard cover, Śāradā.

Moks.opāya, Vairāgyaprakaran.a.
Final Śloka:
phalati no tad ime vayam eva hi
. munayo hatabuddhayah. || (= MU 1.32.43cd)
sphut.aram
iti śr¯ımahārāmāyan.e moks.opāye vairāgyaprakaran. am
. sampūrn.am
. samāptam iti
śivam
No 4797/2281 [Ś17 ] Paper codex, Pot.hi, cardboard cover, Śāradā, 4 volumes.

Dated Vikramasam . vat 1991–1995 [by digits and chronograms, e.g. for the
year 1995: prān.a-nidhāna-graha-bhūmi virājamāne śr¯ıvikramāṅkaśubhavatsara . . . ].
Neatly and meticulously written over a period of four years (= A.D. 1934/35–
1938/39) by K AULA T. H ĀKURA (lipir iyam . kaulat.hākurasya), giving the impres-
sion of a ’collector’s edition’. A table of contents indicates the respective page
numbers. The number of Sargas and Ślokas are also given in a chart for each
Prakaran.a. Appears as being a Śāradā transcript from the printed vulgate
(NEd ), with the notable exception of the Khilas which are separately appended
to the end.
Yogavāsis.t.ha recension (NEd ), Vairāgya- to Nirvān.aprakaran.a (uttarārdha, Sarga
218) with excerpts from VTP. Incomplete, the Pūrvārdha-volumes are missing.
Complete number of Ślokas indicated as: 29.365.
Moks.opāya, Khilas (complete). Number of Ślokas indicated as: 624.

Volume 1 contains Vairāgya to Utpatti. Utpatti ends: j¯ıvas.an.d.akāvatāro
. śadadhikaśatatamah. sargah. || 140|| sampūrn.am utpatti-
nāma catvārim
prakaran.am.
The table of contents gives 140 Sargas for Utpatti, calculating 9123 (cor-
rected from 9035) Ślokas.
Volume 2 contains Sthiti to Upaśānti. Sthiti begins: j¯ıvab¯ıjam
. . . . (= MU
4.19.1)
Volume 3 contains Nirvān.a, Uttarārdha (Sarga 1–107).
Volume 4 contains Nirvān.a, Uttarārdha (Sargas 108–218). Colophon of the
above: sampūrn.am . ca śr¯ımanmoks.opāyaśāstram ||
End of excerpts from the VTP:
r.turasaturagamahat¯ı (1766) śākavikāriśubhavatsarasya śiśirortuh. | . . . || (=
A.D. March 7, 1710)16
iti .t¯ıkākārakr.tāh. ślokāh. ||
No 4798/2305 [Ś22 ] Paper codex, leather binding, Śāradā.

Yogavāsis.t.ha recension (NEd ), Vairāgya- (complete, frame story E plus VTP
introduction) to Upaśamaprakaran.a.
Utpatti contains Bhr.gu, no intercalary colophon, Sthiti begins with j¯ıvab¯ıjam
. (=
MU 4.19.1)
Colophon (in a different handwriting):
pan.d.itacandrikena bhed.haragrāmavāsakena17 ayam . pustakam . moks.opāyam.
nāma mumuks.ūn.ām . vicāran.ārtham
. ātmanā moks.aphalaprāptyartham
. kr.tah.
. . . kaularāmahastena nikhyātavān.
No 4799/1073 Country paper, Pot.hi, Devanāgarı̄.

Gigantic bundle.
Yogavāsis.t.ha recension (NEd ) with VTP.
16 Calculated from the printed edition (N ∗
Ed ) by Dr K ARL -H EINZ G OLZIO (G OLZIO ( 2005)).
TA ṄGASVĀMI ŚARMAN (TA ṄGASVĀMI ŚARMAN (1980), s.v. Ānandabodhendra) calculated the
date wrongly, converted it into A.D. 1842 and established an untenable floruit of Ānanda-
bodhendra as 1780–1850 A.D.
17 This is Bhedara of Kalhana in the southern district. For references, see I KARI (1994), p. 436;
. .
for the exact location, see p. 443, No. 12.
Catalogue entries: “Yogavāsis.t.ha-T.ı̄kā”

No 4827/1633
No 4828/2306 Paper, Pot.hı̄, Śāradā.

Yogavāsis.t.ha recension (NEd ) with Vāsis.t.hatātparyaprakāśa (VTP), parts of the
Nirvān.aprakaran.a.
Catalogue entry: “Yogavāsis.t.ha-Sam

. graha”
No 4800/954 Paper, Pot.hi, cardboard cover, Devanāgarı̄, spidery handwriting,
26 foll.
Moks.opāyasaṅgraha (complete).
Condensed version, unknown so far. Although complete in itself, it represents
nothing more than a very small excerpt from the Nirvān.aprakaran.a.
Begins:
śr¯ıvasis.t.ha uvāca |
rāma samyak prabuddho si svātmānam asi labdhavān |
etam evābalabyātas tis.t.ha meha madam
. kr.thāh. | (∼ NEd 6.29.4c-5b; MU
6.30.10)
Ends:
aham eva jagan naikadeśe me paramān.ukah. || (unidentified)
iti nirvān.arūpātmā śām. to vyaharann api yathācaram asadrūpo j¯ıvanmukto vatis.t.hate | iti
śr¯ınirvān.aprakaran.e mahārāmāyan.e moks.opāye sadehodahamuktavarn. anam . nāma sargah. 29
|| moks . opāyasam. graha saptārta sam
. vat 1864 [= A.D. 1806/07]
Catalogue entries: “Yogavāsis.t.ha-Sāra”
No 4801/1095.02 Collective paper codex, cardboard cover, Śāradā. Minia-

ture paintings, one painting illustrating the Vāsis.t.hasāra by depicting Vasis.t.ha
and Rāma. Vāsis.t.hasāra in 10 Prakaran.as (incomplete, up to VS 10.1). See
No 57.106/44 of the National Museum, above, and No 4813/1968.05, below,
for the reason of using the title as indicated here. Breaks off: . . . grāhyam . gr.h¯ıteti
mr.s.ā vikalpah. || 28|| (= VS 9.32d) iti śr¯ıyogavāsis.t.hasāre ātmanirūpan.am. nāma
navamam . prakaran.am |9|
svapnendrajālavat paśya d¯ınāni tr¯ın.i pam
. ca vā |
mittraks.ettradhanā— (unidentified)

Vāsis.t.hasāra in 10 Prakaran.as (complete).
No 4803/1261 Paper codex, Śāradā.

Vāsis.t.hasāra in 10 Prakaran.as (incomplete, up to Prakaran.a 9), glosses in the
margins.
No 4804/1272.04 Collective paper codex, cardboard cover, Śāradā. Miniature

paintings.

paintings.
No 4806/1311.01 Collective paper manuscript, Pot.hi format, bound, Śāradā.

Vāsis.t.hasāra in 10 Prakaran.as (complete, foll. 20r–35v). Numerous interlinear
glosses in red ink.
No 4807/1583.06 Collective paper codex, leather binding, Śāradā.


Vāsis.t.hasāra in 10 Prakaran.as (complete). Isolated glosses.


No 4811/1803.02 Collective paper manuscript, Pot.hi format, bound, Śāradā.

No 4812/1821.08 [Ś12 ] Collective paper codex, cardboard cover, Śāradā. 128

foll. Title abbreviation in the margins: mo vai.
Moks.opāya, Vairāgyaprakaran.a (complete).
Begins:
. . . om. divi bhūmau . . . || 1|| (= MU 1.1.1)
aham . baddho . . . || 2||
vālm¯ıkir uvāca
kathopāyān vicāryādau moks.opāyān imān atha |. . . || 3||
Ends:
phalati no tad ime vayam eva hi
sphut.ataram. munayo hatabuddhayah. || (= MU 1.32.43cd)
iti śr¯ımahārāmāyan.e moks.opāye vairāgyaprakaran. am
. sam
. pūrn.am
. samāptam iti śivam |

paintings.
Vāsis.t.hasāra in 10 Prakaran.as with the commentary Vāsis.t.hasāravivr.tti of
Mahı̄dhara,18 54 foll.
Comm.:
om. laks.m¯ıkāntam . namaskr.tya yathāmati viracyate |
vāsis.t.hasāravivr.ttih. paropakr.taye mayā || 1||
yatas sarvān.i bhūtāni . . . || 2|| . . .
tatrādau vāsis.t.hasārākhyam
. grantham āripsus tatpratipādyes.t.adevatānatirūpam
. maṅga-
lam ācarati | om
.
Text:
dikkālādyanavacchinnānantacinmātramūrttaye | . . . (= VS 1.1)
Comm.:
evam. vidhāya śāntāya gun.āt¯ıtāya tejase brahmasvarūpāya namah. | tejaśśabdenātra brah-
maiva na bhautike tejas tasya vaks.amān.alaks.an.ānupapatteh. | kim . bhūtāya | tejase . . .
Colophon on folio 8v :
iti śr¯ıyogavāsis.t.hasāravivaran.e vairāgyaprakaran. am |
Text ends:
18 Cp. T HOMI (1999), p. 22f.
saumyāmbhasi yathā v¯ıcir asti na cāsti na ca nāsti ca | . . . gatam || (= VS 10.34)

Comm.:
saumye sthire jale yathā v¯ıcir asti nāsti ca tahedam
. ...
Colophon: iti śr¯ıvāsis.t.hasāre brahmasvarūpam. nāma daśamam. prakaran.am ||
Catalogue entries: “Mahı̄dhara”
No 4823/1140 Collective paper manuscript, Pot.hi format, Śāradā.

Vāsis.t.hasāra in 10 Prakaran.as with vernacular translation (Bhās.āyogavāsis.t.ha-
sāra).
Colophon:
. . . bhās.āyogavāsis.t.hasāre jñānasāre tattvanirūpan.am
. nāma daśamam
. prakaran.am
.
samāptam |
The remaining 12 catalogue entries registered under ’Yogavāsis.t.hasāra’ and un-
der ’Mahı̄dhara’ (Ser. Nos 4814–4822; 4824–4826) could not be consulted due
to lack of time. Judging from the comparatively few folios each of these
manuscripts contains it is extremely unlikely that they would substantially
differ from the Sāras already referred to.
Preliminary Results
From the above, a list of additional Moks.opāya manuscripts may be drawn up,
which, however, must not be regarded as a finalized one. Second, a few pre-
liminary conclusions may also be drawn, although likewise with due reserva-
tion only.
Assorted list of manuscripts
Moks.opāya recension (certain)

Ś12 Vairāgyaprakaran.a. Śrı̄nagar, No 4812/1821.08.
Ś13 Vairāgyaprakaran.a. Śrı̄nagar, No 4796/1821.10.
Ś14 Vairāgya- to Upaśamaprakaran.a. Śrı̄nagar, No 4788/827.
Ś15 Vairāgya- to Nirvān.aprakaran.a (incomplete). Śrı̄nagar, No
4795/1562.
Ś16 Nirvān.aprakaran.a and Khilas. National Museum New Delhi, No.
57.106/100.
Ś17 Khilaprakaran.a. Śrı̄nagar, No 4797/2281 (Volume 4).
Moks.opāya recension (likely)

Ś18 Vairāgyaprakaran.a (incomplete in the beginning). Śrı̄nagar, No
4791/1212.06.
Ś19 Nirvān.aprakaran.a (incomplete). Śrı̄nagar, No 4792/1238.
Ś20 Nirvān.aprakaran.a (incomplete). Śrı̄nagar, No 4793/1274 (2nd
part).
Ś21 Utpatti- and Sthitiprakaran.a. Srı̄nagar, No 4790/1155 (2nd half).
Ś22 Utpatti- to Upaśamaprakaran.a. Srı̄nagar, No 4798/2305 (2nd half).
Moks.opāya recension (VTP-conflated)

– Vairāgya- to Mumuks.uprakaran.a Srı̄nagar, No 4790/1155 (1st half, see
Ś21 ).
– Vairāgya- to Mumuks.uvyavahāraprakaran.a Srı̄nagar, No 4798/2305 (1st
half, see Ś22 ).
– Vairāgya- and Mumuks.uprakaran.a Srı̄nagar, No 4789/1149 [bundle 1].
– Upaśāntiprakaran.a Srı̄nagar, No 4789/1149 [bundle 2].
– Nirvān.aprakaran.a (incomplete) Srı̄nagar, No 4789/1149 [bundle 3].
Yogavāsis.t.ha recension
– Vairāgyaprakaran.a in Sanskrit and vernacular (bhās.ā) Śrı̄nagar, No
4787/361.
– Vairāgyaprakaran.a Śrı̄nagar, No 4793/1274, (1st part, see Ś20 ).
– Nirvān.aprakaran.a, uttarārdha Śrı̄nagar, No 4794/1311.04
– Vairāgya- to Nirvān.aprakaran.a (uttarārdha, Sarga 218) Śrı̄nagar, No
4797/2281, (see Ś17 , Khilaprakaran.a).
– Yogavāsis.t.ha (NEd ) with VTP Śrı̄nagar, No 4799/1073.
– Nirvān.aprakaran.a (incomplete) with VTP Śrı̄nagar, No 4827/1633.
– Nirvān.aprakaran.a (incomplete) with VTP Śrı̄nagar, No 4828/2306.
Abridgements:
Moks.opāya-Sāra
– Moks.opāyasāra, incomplete. National Archives New Delhi, No. 178.
Moks.opāya-Saṅgraha
– Moks.opāyasaṅgraha, Nirvān.aprakaran.a (complete). Śrı̄nagar, No 4800/954.
Vāsis.t.ha-Sāra
– Vāsis.t.hasāra (10 Prakaran.as). National Museum New Delhi, No
57.106/44.
– Vāsis.t.hasāra (10 Prakaran.as). Śrı̄nagar, 11 manuscripts, Nos 4801-4811.
– Vāsis.t.hasāra (10 Prakaran.as) with Bhās.ā translation. No 4823/1140
Vāsis.t.hasāravivr.tti of Mahı̄dhara.
– Vāsis.t.hasāravivr.tti (10 Prakaran.as). Śrı̄nagar, No 4813/1968.05.
Conclusion
With reference to the complete versions, it is clear that the vulgate Yogavāsis.t.ha
recension (NEd ) as commented upon by Ānandabodhendra in A.D. 1710 had
made its way into the valley. In this regard, the incorporation of the Kash-
mir territory into the Mogul Empire (until 1752), and in particular the Sikh
(1818–1846) and Hindu rules (1846–1947) subsequent to Afghan domination,
must not be underestimated in terms of widening the political borders for
cultural and intellectual exchange. Thus, apart from copies transmitting only
the vulgate Yogavāsis.t.ha- or the Moks.opāya recension respectively, quite a num-
ber of the preserved manuscripts are doubtlessly the result of a conflation of
the Moks.opāya with the vulgate. The process by which the Vairāgyaprakaran.a
may have received its various shapes, ranging from a ’pure’ Moks.opāya- to
the vulgate recension can perhaps best be reconstructed from Ś15 : There, the
original folio No 1 of the Vairāgyaprakaran.a got lost some time. Today, folio
no 1 is preceded by a substitute enclosed in the form of an unbound quire of
several folios containing Ānandabodhendra’s VTP-introduction together with
the first Sarga characteristic of the Yogavāsis.t.ha recension. This is an absolutely
interesting find since it physically displays the way by which the conflation
with the VTP-introduction so frequently met with might originally have hap-
pened. Worn off by their rough leather binding the first and last folios of
Kashmir manuscripts are quite often missing. This applies in particular to old
manuscripts. For completing his old manuscript the owner would have sup-
plemented it by a number of folios taken from the VTP, which after its impor-
tation into the valley and its subsequent spread in comparatively recent times,
had obviously become easily accessible. A complete transcript made from
such a compilation would exhibit exactly the features as observed in many a
manuscript of the conflated Moks.opāya recension: The Vairāgya contains the
VTP-introduction (frame-story E) in its beginning, but in the later parts the
tendency to copy out the readings characteristic of the vulgate decreases sig-
nificantly. Excerpts from Ānandabodhendra’s commentary are only occasion-
ally found in these parts, written interlinear or in the margins, not rarely by
a different scribe’s hand. With reference to the abridgements, the most strik-
ing fact is that there is not even a trace of the so-called Laghuyogavāsis.t.ha ver-
sion among the Kashmir-related mss. The picture that now emerges points
to mainly a regional distribution of the various abridged versions. From my
present perspective,19 a hypothetical outline would result in the following:
The Jñānavāsis.t.ha version, still to be closer investigated, has clearly prevailed
only in the south. The Laghuyogavāsis.t.ha, the most reliable manuscripts of
which – according to our present state of knowledge – also hail from the south,
has reached, and was frequently copied in, the region where Nāgarı̄ writ-
ing prevailed, probably as far as the Hindı̄ belt from Delhi to Benares. The
Vāsis.t.hasāra, composed by Mahı̄dhara in Benares in 1597 A.D.,20 may eventu-
ally have spread from this place as far as Kashmir, after its annexation in 1589
by Akbar. Peculiar local Kashmirian abridgements have been preserved as
unique manuscripts, such as two Moks.opāyasaṅgrahas,21 or the Moks.opāyasāra,
all of them in need of closer investigation. Manuscripts of the Chandra Shum
Shere Collection in the Bodleian Library, Oxford, could represent local Nepal
abridgements.22
19 Again, attention is to be drawn to S. S TINNER’s contribution to this volume.

20 Cp. TA ṄGASVĀMI ŚARMAN (1980), No. 145: Mahı̄dhara; T HOMI (1999), p. 22f, and above,
p. 39.
21 1) Niedersächsische Staats- und Universitätsbibliothek Göttingen (Cod. Ms. Sanscr.
Vish. 126). By its huge size and removal of the Ākhyānas it totally differs from all other ab-
stracts that have ever come to notice. It depends doubtlessly on the Kashmirian recension of
the Yogavāsis.t.ha. For details, see the article by J. H ANNEDER below (p. 105ff.). 2) Śrı̄nagar, No
4800/954. Its size is a very small percentage only of the first Saṅgraha.
22 Cp. S LAJE (1996), p. 16, n. 1. Also in their case, a thorough investigation is required.
The Moks.opāya Project (IV)
Manuscripts from Pune, Wai, Baroda
P ETER S TEPHAN and S USANNE S TINNER
As part of the pursuit for mss. relating to the projects underway at the Insti-
tute of Indology, Martin-Luther-University Halle-Wittenberg1 various manuscript
libraries were contacted for arranging the digitization of mss. from their col-
lections that belonged in some way or another to the Moks.opāya (MU) liter-
ature. Aided by funds of the DFG project (“Critical Edition of the Utpatti-
prakaran.a of the Moksopāya”) and with kind permission from the respective
authorities a number of mss. from libraries in Maharashtra and Gujarat could
be checked and digitized during February and March 2005.2
This report contains merely a listing, classification and brief description of
all the mss. that could be consulted and were considered relevant. Prefixed
to each are quotations of the entries from the printed manuscript catalogues
in order to facilitate identification. Since these entries are often superficial,
they are regularly contradicted by the subsequent description. Wherever the
entries are correct, the description merely adds the extent of the text. All quo-
tations are diplomatic transliterations. A more detailed description and eval-
uation, especially of the mss. of the Laghuyogavāsis.t.ha (LYV), will be part of
future publications.
For understanding the descriptions of Yogavāsis.t.ha (YV) and MU mss. be-
low one might recall that the distinction between the mss. of the Kashmirian
MU-tradition and those of the pan-Indian YV recension is based on certain
formal criteria.3 Since the majority of mss. consulted do not contain the entire
text, they cannot be tested on the basis of all of these criteria. Absolute cer-
tainty about whether a fragmentary mss. that lacks those passages that allow
a quick distinction between recensions belongs to one or the other can only be
achieved by examining their readings. Mss. in Devanāgarı̄ script accompanied
by the Vāsis.t.hatātparyaprakāśa (VTP) by Ānandabodhendra Sarasvatı̄ as well as
1 See above, p. 3.
2 P ETER S TEPHAN is responsible for the selection and digitization of all the mss. described
below and for the description of Yogavāsis. t.ha and Moks.opāya mss. S USANNE S TINNER undertook
the task of describing all the remaining mss., i.e. those related to the Laghuyogavāsis. t.ha and other
abridged versions.
3 Compare, p. 42 and p. 75f.
56 Stephan/Stinner: Moks.opāya Project IV
those mss. transmitting the YV frame can, nevertheless, be clearly identified

as belonging to the YV recension. Hence, the description of such mss. focuses
merely on establishing their identity. Since a vast majority of Devanāgarı̄-mss.
belongs to the more widespread YV recension, any non Kashmirian ms. must
be carefully checked for strong evidence, as an undivided Nirvān.aprakaran.a or
the absence of an introductory frame story.4
Given the complicated nature of the textual transmission and the disparity
of names used to designate the abridged versions of the YV and their com-
mentaries, most of the catalogue entries do not provide sufficient information
even for a most basic identification of the text. Thus the identity of the text,
i.e. whether a ms. contains a part of the MU, YV, LYV, Vāsis.t.hasāra, or even an
unknown version, as well as the identity of the commentaries had to be estab-
lished. For pragmatic reasons Prakaran.as, Sarga and verse numbers are quoted
according to the printed editions.
Nevertheless, it should be noted that apart from the printed LYV (LNEd )
only one ms. – namely Wai 6926 – counts the Sargas contained in each
Prakaran.a separately. All other mss. count Sargas successively from the begin-
ning to the end, to a certain extent. In the same way the two complementary
commentaries to the LYV, namely the Vāsis.t.hacandrikā (VC) of Ātmasukha as
well as the Sam . sārataran.i (ST) of Mummad.ideva Allād.asūnu, transmitted in
the respective sections (VC ad prakaran.a 1–3 and ST ad prakaran.a 4–6.15), were
identified with the help of LNEd . The unpublished first half of the ST, that
is the commentary on Prakaran.a 1–3, was identified by comparing it with the
mss. LG1 and LG3 .5 Meticulous attention was given to the structure of the final
sections of all those mss. containing the Nirvān.aprakaran.a. Moreover the same
textual varieties of transmission could be observed as described in the article
below.6
4The Utpattiprakaran. a extending to NEd 4.18 alone is not a sufficient indicator.

5See below, p. 100ff. It is still unclear how the peculiar segregation of these two commentaries
arose. In some mss. the diverging association of sections of the ST (designated in colophons
as -vāsis.t.havivaran. a) and VC (designated as -t.ı̄kā and -dı̄pikā in some colophons) as compared to
LNEd is observed in Ānandāśrama S12(1)-4-40 and Wai 6923. It is likely that mss. transmitting
this irregularity might have caused the confusion observed in some catalogues as well as in the
first and second edition of LNEd .
6 A detailed description of these four major varieties is given in p. 98ff. The first of these is
the shortest version extending up to LNEd 6.15.113. The second version is characterised by the
Gurupraśasti (LNEd 6.16.1–4) frequently transmitted independent of the mūla-text in the form of
a concluding addition to ST. The third version concludes with the Sārasarga (LNEd 6.16) and the
Stephan/Stinner: Moks.opāya Project IV 57
Finally we should mention that for the success of this venture the support
of numerous colleagues, employees and in-charges of the mss. libraries – too
many to be mentioned here – was crucial.
————
Two institutions in Pune were approached: the Ānandāśrama Sam . sthā, 22

Budhwar Peth, 411002 Pune, and the Bhārat Itihās Sam. śodhak Man.d.al.-Mandir
(next to Bharat Nat.yam Man.d.al.), Sadashiv Peth, 411030 Pune. All mss. of
both institutes, which are described below, are made of paper and written in
Devanāgarı̄ script with ink.
Ānandāśrama Sam
. sthā, Pune
The library of the Ānandāśrama Sam
. sthā has not yet published its catalogue, but
due to the kind support of Prof. K. S. A RJUNWADKAR and the collaboration of
the Managing Trustee VASANT A. A PTE a list of YV related mss. – henceforth
called catalogue – was forwarded to the author. Out of the 30 mss. consulted
19 were selected as related to the projects and are thus described below.7
1. Moks.opāya
Catalogue entry “yogavāsis.t.ha(upaśama prakaran.a) F[olios]. 63 C[omplete].
33x17cms. S12(1)-4-38”.
Description: 413 folios, 16 lines per page, first folio illustrated with a minia-
ture painting in the centre of the folio encircled by the text. The siglum “mo
ks.o” is given in the left upper margin of each folio accompanied by the siglum
of the respective Prakaran.a and folio number.
Moks.opāya (N30 ) Nirvān.aprakaran.a and Khilas (nānāpraśnātmikāh. khilāh.).
Without pūrva- and uttarārdha partition, without the final YV-frame; the Khilas
follow immediately upon the last Sarga. The final colophon gives an account
of the textual structure of the MU.8
fourth version is characterised by the final frame LNEd 6.17–18 transmitted only in Nāgarı̄ mss.
Conclusions on these data, also focusing on the final colophons, are given in the same article.
7 Another 4 mss. were reported missing: S12(1)-3-23, S12(1)-4-31, S12(1)-5-41, and S12(1)-6-56.
8 The Bhārgavopakhyāna is nevertheless placed in the Sthitiprakaran a, as its first upākhyāna:
.
sthitiprakaran. e bhārgavadāmavyālakat. a . . . upākhyānacatus.t.ayam
..
2. Br.had-Yogavāsis.t.ha
Catalogue entry "mahārāmāyan.a sthiti prakaran.a F. 59 C. 32x18 cms. S12(1)-
4-32".
Description: Frequently added verses in the margin by a second hand, 14
lines per page. Text: NEd 4.19–4.62. Penultimate Sarga’s colophon, which in
NEd gives the name of the respective Sarga as jananamaran.asthitivarn.anam,
runs as follows: iti śr¯ımahārāmāyan.e moks.opāyes.u sthitiprakaran.e mohavicāro
nāma sargah.. Final colophon: ity śr¯ımahārāmāyan.e ārs.opacarite devadūtokte
śatasāhasryām
. sam . hitāyām. śr¯ımoks.opāyes.u sthitiprakaran.am . ||
. samāptam
sthitiprakaran.am. sam . pūrn.asamāptam.
Catalogue entry "yogavāsis.t.ha F. 112 C. 33x17cms. S12(1)-4-33".

Description: Upaśamaprakaran.a, complete, 17 lines per page. In-
cipit as well as explicit like NEd . Date in the final Prakaran.a-
colophon: ity ārs.opacite mahārāmāyan.e devadūtokte śatasahasryām . sam
. hitāyām.
vāsis.t.he brahmadarśane vālm¯ık¯ıye moks.opāyes.ūpaśamaprakaran.am . pam . camam .
. || 93|| les.akapāt.hakayo mam . || sam
9
samāptam . galam
. kalpān.am . bhūyāstām . va 18
vars.e 11 || lis.atam. daśapūragnāt¯ıvyāsago kala || śr¯ı rāmacam. drāya namah. || śr¯ı
kr.s.n.āya nama. Annotation added below the colophon with ball pen: sam . vat –
1811.
Catalogue entry "yogavāsis.t.ha (utpattiprakaran.a) F. 203 C. 32x18 cms. S12(1)-

4-35".
Description: Frequent corrections and additions by a second hand, 12 lines
per page. Text: NEd 3.1–4.18.
Catalogue entry "yogavāsis.t.ha(upaśama prakaran.a) com[mentator] Ar [=Au-

thor] ānam. da bodha F. 453 I.C. [=incomplete] 36x15cms. S12(1)-4-36".
Description: Additions in the margin, corrections by another hand, 8-12
lines per page. Text: Upaśamaprakaran.a with the VTP of Ānandabodhendra
Sarasvatı̄, complete. Same paper, same scribal and corrector’s hand, same
style of conveying information on the extent (pattra, sarga, grantha) on
cover folio as in S12(1)-5-49A and S12(1)-5-50 (see below). Author’s param-
parā mentioned in the final colophon: iti śr¯ımatparamaham. saparivrājakācārya-
9 Figure 8 doubtful.
varyaśr¯ımatsarvajñasarasvat¯ıpūjyapādaśis.yasya śr¯ımadrāmacam . drasarasvat¯ıpūjya-

pādaśis.yaśr¯ıgam
. gādharendrasarasvatyākhyabhiks oh
. . śis yen
. . a śr¯
ımadānandabodhen-
drasarasvatyākhyabhiks.un.ā viracite śr¯ıvāsis.t.hātparyaprakāśe upaśamaprakaran.am .
sam. pūrn.am ..
Catalogue entry "yogavāsis.t.ha(upaśama prakaran.a) F.141 C. 31x17 cms.

S12(1)-4-37"
Description: 12–14 lines per page. Text: NEd 5.1–93. Final colophon:
ity ārs.opacarite devadūtokte mahārāmāyan.e moks.opāyes.ūpaśām. taprakaran.am.
punarupadeśavaran.am . nāma sargah. || 94 || samāptam . cedam mupaśām . ti-
prakaran.am. m || ata ūrdhvam . bhavis.yati || yasyāyam
. nirvān.aprakaran.am .
pratisam. dhiślokah. . . . The scribe adds the pratisandhiśloka of the following
Prakaran.a.
Catalogue entry “yogavasis.t.ha(nirvān.a prakaran.a-pūrvārdha) F.401 C. 36x15

cms. Sk 1672 S12(1)-5-42”.
Description: No annotations in the margin, first and final folios in
poor condition, 10-14 lines per page. Text: NEd 6.1–6.128 with the VTP
of Ānandabodhendra Sarasvatı̄. Date in the final colophon: śakem . 1672
pramodasam. vatsare bhādrapada krs n
.. . a ekādaś¯
ı idam
. pustakam
. samāptam
. .
Catalogue entry “yogavāsis.t.ha(nirvān.a prakaran.a-pūrvārdha) F.594 C. 32x14

cms. S12(1)-5-43”.
Description: Occasional annotations in the margin, 694 folios, 11 lines per
page. Text: NEd 6.1-6.128 with the VTP of Ānandabodhendra Sarasvatı̄.
Colophon nearly identical with S12(1)-4-36, S12(1)-5-49A, and S12(1)-5-50.
Catalogue entry “yogavāsis.t.ha(nirvān.a prakaran.a-uttarārdha) F.117 I.C.

33x14 cms. S12(1)-5-44”.
Description: Ms. consists of two parts. First part ranges from folio 5r until
13v with first scribal hand on old paper in poor condition. Ms. breaks off and
continues on folio 96r until 204v with different scribal hand on different paper
and in better condition. First part (5r–13v) counted as verse 12–86 is LNEd
3.1.12–88 with commentary VC; second part (96r–204v) is YV 7.29.47–7.57.23
with commentary VTP.
Catalogue entry “yogavasis.t.ha(mumuks.u. utpatti prakaran.a) F.126 C. 32x17

cms. S12(1)-5-45”.
Description: Text: NEd 2.1–3.122, 15 lines per page.
Catalogue entry “yogavāsis.t.ha(mumuks.u. vyavahāra prakaran.a) F.24 C.

32x18 cms. S12(1)-5-47”.
Description: Additions – some referring to a“t.¯ ıkā” – by a different hand in
the margin, 15 lines per page. Incipit verse corrected in accordance with NEd ,
verse 2.1.2 to 2.1.4 attributed to Viśvāmitra as speaker. Designations of Sargas
in inspected colophons differ slightly from the ones used in NEd .
Catalogue entry “yogavāsis.t.ha(vairāgya prakaran.a) F.36 C. 32x17 cms. S12(1)-

5-49”.
Description: 15 lines per page. Text: NEd 1.1–1.33.
Catalogue entry (uncatalogued)

Description: Designation on the cover: “Yogavāsis.t.ha vairāgya prakaran.a
F.374 C. (36x15) cms. Sk1706 S12(1)-5-49A”
Description: Ms. part of S12(1)-4-36 (see above) and S12(1)-5-
50 (see below), same paper, same scribal and corrector’s hand,
same style used for the cover folio and identical colophons.
Text: NEd 6.1–6.128 with commentary VTP. Text designated on
cover folio: . bhah. || 6||
yogavāsis.t.hanirvān.aprakaran.apūrvārdhaprāram
s.as.t.hanirvān.aprakaran.am
. pūrvārdham . || 6 patre 590 sarga 128 gram . tha 12940.
Catalogue entry “yogavāsis.t.ha(sthiti prakaran.a) com Ar ānam

. da bodha F.256
C. 36x15 cms. S12(1)-5-50”.
Description: Occasional marginal additions by the same hand, corrections
by a second hand. Ms. part of S12(1)-4-36 and S12(1)-5-49A (see above), same
paper, same scribal and corrector’s hand, same type of cover folio and identi-
cal colophons. Text: NEd 4.1–4.62 with VTP.
Catalogue entry "yogavāsis.t.ha(sthiti prakaran.a) F.59 C. 32x18 cms. S12(1)-5-

51".
Description: 17 lines per page. The text and the inspected colophons cor-
respond exactly with the vulgata edition (NEd ). Date in the final colophon:
. bat 1812 rābaisās.a(?)v¯ıda 6 budhe || l¯ıkh¯ıtam
sam . dasapurajñāt¯ım¯ısras¯ıtārāmam
. m.
3. Laghuyogavāsis.t.ha
Catalogue entry "yogavāsis.t.ha sat.ı̄ka(upaśama prakaran.a) com Ar ānam . da

bodha F.453 C. 36x15 cms. S12(1)-4-39".
Description: Last folio no. 117. Ms. contains the final part (Nirvān.aprakaran.a)
of S12(1)-4-40 YV (see below): same scribal hand, same style of cover fo-
lio and pagination as in S12(1)-4-40. Text: Nirvān.aprakaran.a of the LYV
with the commentary Sam . sārataran.i of Mummad.ideva Allād.asūnu (and
not the VTP of Ānandabodhendra Sarasvatı̄ as indicated in the catalogue).
Explicit mūla text on folio 116–117: LNEd 6.15.105 (numbered 103) to
6.15.113 (numbered 110). Half verse added at the end of the text: bhāvayan
. karma samācara || 111|| .
10
śaśvadam . tah. svam . kāryam Final colophon mūla
text: iti śr¯ıvālm¯ık¯ıye yogavāsis.t.he nirvān.aprakaran.e saptabhūmikopākhyānam . nāma
tricatvārarim . śatamah. sargah. || 43 || iti yogavāsis.t.ham . samāptam . || śr¯ı. Date
and scribe’s name in the final colophon of the commentary: iti śr¯ıyogavāsis.t.he
vālm¯ık¯ıye moks.opāye nirvān.aprakaran.avivaran.e saptabhūmikopākhyānam . nāma tri-
catvārim . śatamah . sargah . || 43 [. . . ] śake 1750 vikārināmasam. vatsare māgha-
kr.s.n.advit¯ıyāyām . guruvāre śr¯
ıkāśyām. śr¯
ıgam . gāviśveśvarasannidhau idam. pustakam .
gurjaśepanānmāgam . gādharen.a nr.sim . haputren.a likhitam . svopakārāya paropakārāya
ca likhitam ..
Catalogue entry “yogavāsistha(t.ı̄kā-vasis.t.ha cam

. drikā) F.120 C. 33x16 cms.
S12(1)-4-40 YV”.
Description: 120 folios, 15 to 17 lines per page. Frequent additions in the
margin by the same hand, occasional corrections with yellow pigment, very
good state of preservation. Text: Vairāgya-, Mumuks.u-, and Utpattiprakaran.a of
the LYV with the commentary Vāsis.t.hacandrikā of Ātmasukha. Sthitiprakaran.a
with the commentary Sam . sārataran.i of Mummad.ideva Allād.asūnu. The fi-
nal part of this ms. (Nirvān.aprakaran.a) is contained in S12(1)-4-39 (see above),
Upaśamaprakaran.a missing. New Prakaran.as indicated by cover folio, some
Prakaran.as with closing folio. Page sigla in the right margin: “pra◦ ” count the
entire ms., sigla in the left margin count each Prakaran.a separately. The text
of the two commentaries runs in accordance with LNEd , but from the second
10 This last verse is not part of the LYV as represented in LN

Ed or in other mss. In the “Br.had-
Yogavāsis.t.ha Pāda-Index” (T HOMI (1990)) this additional verse can be traced as NEd 6.106.101e-f,
but is not contained in NEd 6.106, which comprises only of 70 verses.
sarga in the Sthitiprakaran.a onwards the name of the commentary is confused

in the colophons.11
Catalogue entry “yogavāsistha(mumuks.u. utpatti, sthitı̄, upaśama, vairāgya

va nirvān.a prakaran.a) F.374 C. 36x15 cms. Sk 1706 S12(1)-5-46”.
Description: 375 folios, 9 to 13 lines per page. Additions in the margin by
the same hand, occasionally by a second hand. Corrections with yellow pig-
ment mainly in the Nirvān.aprakaran.a. Very good state of preservation. Text:
Vairāgya-, Mumuks.u-, and Utpattiprakaran.a of the LYV with the commentary
Vāsis.t.hacandrikā of Ātmasukha. Sthiti-, Upaśama-, and Nirvān.aprakaran.a of the
LYV with the commentary Sam . sārataran.i by Mummad.ideva Allād.asūnu. Sep-
arate pagination in each Prakaran.a, sigla in the left margin: vā◦ vai◦ , mu◦ vā◦
t.¯ı, vā◦ u◦ , vā◦ sthi◦ , vā◦ upa◦ t.¯ı◦ or upa◦ vā◦ t.¯ı◦ and vā◦ upa◦ , vā◦ ni◦ . Each
Prakaran.a has a front and back cover folio stating merely the name of the
text. Explicit mūla text on folio 136 as in LNEd 6.15. with LNEd 6.15.113. Fi-
nal colophon of the mūla text: iti śr¯ıvālm¯ık¯ıye vāsis.t.havivaran.e nirvān.aprakaran.e
saptabhūmikopākhyānam . nāma tricatvāriśattamah. sargah. sam . pūrn.ah.. Date in
final colophon of the commentary on folio 137: iti sam . sārataran.ināmni
vāśis.t.avivaran.e yogasaptabhūmikopākhyānam . tricatvāriśattamah. sargah. followed
immediately by the gurupraśasti as part of the commentary, which is desig-
nated in LNEd as 6.16.1–4. Final dedication: śr¯ıgopālakr.s.n.ārpan.am astu, date in
red ink: śake 1706 krodh¯ınāmās.t.e.
4. Yogavāsis.t.hasāra
Catalogue entry “yogavāsis.t.a vivaran.a Ar śam . karācārya F.20 I.C. 25x10cms.
Sk 1700 S12(1)-5-48”.
Description: 31 folios, 10 lines per page. Text: Yogavāsis.t.hasāra with
commentary of Mahı̄dhara (starting with 3.9), incomplete. Author’s name
in the last Prakaran.a’s colophon: bhagavatpādapūjyaśam . karācārya. Date in
the final colophon: śakke 1700 vilam . banāmasam . vatsare phalgunabahulāyām .
ekādaśyām
. tithayah
. mam
. davāsare midam. dines. u naks
. atram. 12 uttarās. ād.hāyām
. yoga
11 For instance on folio 8 / 104 the colophon of the second sarga in the Sthitiprakaran a runs as
.
follows: “iti śrı̄vāsis.t.hacam
. drikāyām
. dāmādyupākhyānam. (...)” but on folio 10/106 the colophon of
the third sarga in the Sthitiprakaran. a reads “iti śrı̄yogavāsis. t.hadı̄pikāyām
. moks.opāye (...)”. However,
the text of the commentary runs as in LNEd and is usually attributed to Allād.asūnu (Sam . sārataran.i
or -taraṅgin. i).
19 paripa(?)ttat dines.u sūryodayānam . taram. dinam. ghat.i 5 tatsamayes.u samāptam

idam
. pustakam. ram
. ganāthas ūnunā rāyan . am idam
. nāmnena likhyate parārtham
. ....
Bhārat Itihās Sam

. śodhak Man.d.al.-Mandir, Pune
One of the authors was kindly informed about the existence of this library
by Dr. S IDDHARTH Y. WAKANKAR of the Oriental Institute M.S. University of
Baroda. Access to the catalogue and a permission to consult the required mss.
was obtained with the help of private acquaintances.12 However, on the basis
of the catalogue listings and an inspection on the spot, it could be ascertained
that no ms. kept in the library shows the formal criteria to classify it as belong-
ing to the MU recension, therefore no attempt was made to produce copies of
them.
The library’s mss. catalogue (K HARE (1960), p. 230–231) lists 13 YV-mss.
(three Marāt.hı̄ mss), 1 YV-t.ı̄kā, 7 YV-sāra (4 with Mahı̄dhara’s commen-
tary, 2 in Marāt.hı̄, 1 in Sanskrit accompanied by a Marāt.hı̄ commentary or
translation by Mādhavadās), 2 YV-sāra-t.ı̄kā, 2 YV-sāra-vivaran.am, 1 YV-sāra-
samuccaya, and 2 Laghu-YV-sāra (p. 248). Ten Sanskrit-mss. are listed as
“Yogavāsis.t.ham” in the catalogue. The following four mss. were reported
missing by the librarian Mrs. M EENA J OSHI: 7.7 containing Prakaran.a 1–3,
50.250 named “Yogavāsis.t.ham” as well as 51.10 and 25.222 – both desig-
nated as “Yogavāsis.t.ham (sat.ı̄kam)”. One further ms. could be identified
by its catalogue entry and was therefore not further inspected: “59.314 yo-
gavāsis.t.ham (vāsis.t.hatātparyaprakāśah.) | (Ānandabodhendra Sarasvatı̄) vi
1870 vaiśākhah. va 3 ravih.”.
Catalogue entry “7.8 yogavāsis.t.ham”.

Description: 264 folios, Vairāgyaprakaran.a to Utpattiprakaran.a. The colophons
designate the text as “mahārāmāyan.e moks.opāya”. The first Prakaran.a con-
tains the YV-frame.
Catalogue entry “7.9 yogavāsis.t.ham | pra 2”.

Description: 258 folios, Nirvān.aprakaran.a pūrvārdha, but no mention of
pūrvārdha in colophons or cover folio. Final colophon: samāptaś cāyam
.
12 Thanks to Mr. S ANJAY G ODBOLE for conveying the name and telephone no. of the director
Dr. G OKHALE and for her co-operation.

trit¯ıyakhan.d.ah.. In the final sarga 127, designated in the colophon as

rāmasamādhivyuthānam. , the verses 100–110 are missing although the pagi-
nation continues without a gap. The preceding penultimate sarga – also
counted as 127 – is designated as bharadvajānuśāsanam . and follows sarga 126
paramārthasvarūpavarn.anam . . Like in N Ed , sarga 124 called mr.gavyādh¯ıyam
. is
followed by sarga 125 which has no designation other than nirvān.e.

Description: 419 folios, Nirvān.aprakaran.a uttarārdha, but no mention of ut-
tarārdha in colophons or cover folio. Text of the final sarga 215: NEd 215 + 216,
i. e.: sarga 215 includes the final YV frame and continues the verse numbering.
The final colophon contains the following dating: mārgasaramāse śuklapaks.e
tithi 4 somavāre sam. pūrn.am punanam
. va tu kalyān.am astu. This colophon is fol-
lowed by two lines in Persian script.

Description: 8 folios each consisting of four pages. The unidentified text
seems to contain or be a fragment of the L¯ılākhyāna of the Utpattiprakaran.a.
Incipit: śriganeśāya namah. | utpatyākhye prakaran.e navākhyānavirājite | ādyam
ākāśajākhyānam. l¯ılākhyānam atah. param || 1|| Explicit: iti te varavarn.inyau tato
brahmām . taman.d.alāt.
Catalogue entry “48.2 k yogavāsis.t.ham”.

Description: Folios completely out of order and in poor condition. Some
of the folios contain different paginations in the right and left margins, des-
ignated as sauptika. The text seems to consist of four Prakaran.as separately
paged. One legible colophon states: iti śrisārala saupt¯ıkaparvan.i mukteśvara 6313
ta .t¯ıkāyām
..
Catalogue entry “41,262 yogavāsis.t.hat.ı̄kā | darśanam”.

Description: The text is in fact the Vāsis.t.hasāra with the commentary of
Mahı̄dhara as noted in Marāt.hı̄ by an unidentified scholar on a separate sheet
of paper kept at the beginning of the ms.
13 Identification of the numbers doubtful.
Three mss. are listed in the catalogue as Marāt.hı̄ Yogavāsis.t.ha mss:

29.132, 29.1566 and 47.217. Out of these only 47.217 was inspected.
The text is incomplete starting on folio 12r until 32v, final colophon: śr¯ı
vāsanopaśamanasampurn.am astu. The exact source of the text could not be as-
certained. The other two were not ready for inspection.
All further mss. listed in the catalogue as “Yogavāsis.t.hasāra” and the like
(see above) were not consulted due to lack of time.
Wai
In the Pradnya Granthalaya,14 Madhali Ali, Wai 412 803, which was also ap-
proached for digitizing YV-/MU-related mss., out of the 15 mss. consulted 10
were selected as being related to the above mentioned projects. As in Pune all
these mss. are made of paper and written in Devanāgarı̄ script with ink.
1. Br.had-Yogavāsis.t.ha
Catalogue entry “6918 | 8-6/453 | Yogavāsis.t.ha | Vālmı̄ki | 46.3x20.6 cm |
800 folios | 11 lines per page | 46 letters per line | incomplete | old | śaka
1595”.
Description: At least two scribal hands each in Marāthı̄-ductus („l.a“ fre-
quently used for “la”). Text designation on cover folio: br.hadvāsis.t.ha. First
sarga YV-frame designated as śr¯ısarga. Text: Utpattiprakaran.a: NEd 3.1 to
3.66.12a, incomplete due to missing folios; then Sthitiprakaran.a: NEd 4.1 – 4.62.
Nirvān.aprakaran.a divided as in NEd , but no mention of pūrva- nor uttarārdha.
Text in final sarga: NEd 7.215–7.216.25. The concluding YV-frame does not
constitute a separate Sarga, as in ms. 7.10 of the Bhārat Itihās Sam . śodhak Man.d.al.-
Mandir described above.
Prakaran.a-colophons: Mumuks.u: śake 1595 prasthānadanāmasam . vatsare . . . ,
Sthiti: śake 1596 vars.e ānam . danāmasam . vatsare vaiśākha vadi daśimi bhaume
sthitiprakaren.e sam . pūrn.am. , Upaśama: svasti śr¯ı śake 1596 ānam . dasam . vatsare
jyes.t.aśuddhapratipadā taddine likhitam . , Nirvāna first part: svasti śr¯ı śake 1596
ānam . dasam. vatsare vaiśākha vadi trayodaśyām . nirvān. aprakaranam
. . gram . thoyam.
samāptah., Nirvān.a second part, i.e. final colophon: sam . vat 1596 vars.e
ānam . dasam. vatsare vaiśākha vadi 3 tr.t¯ıyābhaumavāsare taddine sam . pūrn.am ..
14 See J OSHI (no date).
Catalogue entry “6919 | 8-6/453 | Yogavāsis.t.ha with Tātparyaprakāśa |

Vālmiki | Ānandabodhendrasarasvatı̄ | 25x19.5 cm | 55 folios | 28 lines per
page | 42 letters per line | incomplete | good | śaka 1778”.
Description: Two different, well legible hands. Text starts with NEd 1.1 and
ends after NEd 3.11, lacunae within this range.

Vālmiki | Ānandabodhendrasarasvatı̄ | 32x20 cm | 345 folios | 24 lines per
page | 48 letters per line | incomplete | good | śaka 1775”.
Description: Text NEd 2.1–7.83.

Vālmiki | Ānandabodhendrasarasvatı̄ | 36.6x15.5 cm | 365 folios | 6 lines
per page | 30 letters per line | incomplete | good”.
Description: Text NEd 1.1–1.33.
2. Laghuyogavāsis.t.ha
Catalogue entry “6922 | 8-6/453 | Laghuyogavāsis.t.ha with Vāsis.t.hacandrikā
Pra 1-3 & Sam . sārataran.i Pra. 4-6 | Gaud.abhinandayati | Ātmasukha &
Mummad.ideva | 44.4x21.8 cm | 268 folios | 17 lines per page | 46 letters
per line | complete | good | important”.
Description: 286 folios, 11 to 18 lines per page. Frequent addi-
tions and corrections in the margin by the scribe’s and other hands,
corrections with yellow pigment, composed out of at least three dif-
ferent mss., first and ultimate cover folios indicating the transition of
Prakaran.as. Text: Vairāgya-, Mumuks.u-, and Utpattiprakaran.a with the
commentary Vāsis.t.hacandrikā by Ātmasukha. Sthiti-, Upaśama-, and
Nirvān.aprakaran.a with the commentary Sam . sārataran.i by Mummad.ideva
Allād.asūnu. Nirvān.aprakaran.a up to LNEd 6.18. Sārasarga without gurupraśasti.
Colophon of the mūla text after LNEd 6.15: iti śr¯ıvālm¯ık¯ıye vāsis.t.harāmāyan.e
nirvān.aprakaran.e saptabhūmikopākhyānam . nāma tricatvārim . śah. sargah. 43. Fi-
nal colophon of the commentary after LNEd 6.15 on folio 71r in the mar-
◦
gin: iti sam . sārataran.ināmni 43. Colophon of the mūla text after LNEd
6.16: iti śr¯ıvālm¯ık¯ıye moks.opāye nirvān.aprakaran.e śāstragrahan.aphalanirdeśo nāma
catuh.catvārim . śah. sargah. || śr¯ı|| 44. Colophon of the mūla text after LNEd
6.17: iti śr¯ıvālmik¯ıye moks.opāye bhāradvājānuśāsanam . nāma pam . cacatvārim
. śah.
sargah. 45. Final colophon of the mūla text with author’s name after
6.18: iti śr¯ıtarkavād¯ıśvarasāhityācāryagaud.amam
. d.alālam
. kāraśr¯ımatabhinam
. dayati-
samudhr.te vāsis.t.harāmāyan.e moks.opāye sāroddhāre bharadvājopākhyāne śr¯ıvāsis.t.ha-
rāmasam. vāde s.at.catvārim. śah. sargah. 46.

Pra 1-3 & Sam . sārataran.i Pra. 4-6 | Gaud.abhinandayati | Ātmasukha &
per line | complete | good | important”.
Description: 412 folios, 5 to 12 lines per page. Additions and cor-
rections in the margin by the scribe’s and other hands, corrections with
yellow pigment, first and ultimate cover folio indicating the transition
of Prakaran.as, Prakaran.as counted separately. Text: Vairāgya-, Mumuks.u-,
and Utpattiprakaran.a with the commentary Vāsis.t.hacandrikā by Ātmasukha.
Sthiti-, Upaśama-, and Nirvān.aprakaran.a with the commentary Sam . sārataran.i
by Mummad.ideva Allād.asūnu. Nirvān.aprakaran.a up to LNEd 6.15. Fi-
nal colophon of the commentary VC in the Utpattiprakaran.a: iti śr¯ımat-
paramaham . saparivrājakācāryottamapūjyapādaśis.yaśr¯ımahātmasukhakr.tau vāsis.t.ha-
cadrikāyām . lavan.opākhyānam . nāma trayodaśah. sargah.. Colophon of the com-
mentary ST in the first sarga of the Sthitiprakaran.a: iti śr¯ımadallālasūri-
sūnunā padavākyapramān.apārāvāradr.śvanā śr¯ımammum . d.idevavidvadācāryaviracite
sam . sārataran.¯ınāmni vāsis.t.havivaran.e śukropākhyānam . nāma caturdaśah. sargah..
Colophon of the commentary in the second sarga of the Sthitiprakaran.a: iti
◦
śr¯ımatparamaham . saparivrājakācārya śr¯ımahātmasukhakr.tau vāsis.t.hacam . drikāyām
.
15
dāmādyupākhyānam . nāma pam . cadaśah. sargah.. Explicit mūla text on folio 135
in accordance with LNEd 6.15 but the last verse LNEd 6.15.113 numbered as
100. Final colophon of the mūla text: iti śr¯ıvālm¯ıkiye vāsis.t.harāmāyan.e nirvān.a-
prakaran.e saptabhūmikopākhyānam . nāma tricatvārim . śah. sargah.. Final colophon
of the commentary: iti sam . sārataran.¯ınāmn¯ı vāsis.t.havivaran.e nirvān.aprakaran.e
yogasaptabhūmikopākhyānam . nāma tricatvārim . śatamah. sargah., followed by the
gurupraśasti within the commentary (!), but only up to LNEd 6.16.1–3ab and
counted separately as verse 1 to 2, finally followed by: samāptoyam . granthah..
Catalogue entry“6924 | 8-6/455 | Laghuyogavāsis.t.ha with Vāsis.t.hacandrikā

Pra. 1-3 & Sam
. sārataran.i Pra. 4-6 | Gaud.abhinandayati | Ātmasukha &
15 The commentary is in fact the ST of Allād.asūnu as designated in LNEd .

per line | complete | good, worm eaten| śaka 1747 | important”.
Description: 407 folios, 9 to 16 lines per page, occasional additions and cor-
rections in the margin, some corrections with yellow pigment. Page sigla
in the left margin: yo◦ vā◦ , in the right margin: pra◦ . Text: Vairāgya-,
Mumuks.u-, and Utpattiprakaran.a with the commentary Vāsis.t.hacandrikā by
Ātmasukha. Sthiti-, Upaśama-, and Nirvān.aprakaran.a with the commentary
Sam . sārataran.i by Mummad.ideva Allād.asūnu. Nirvān.aprakaran.a up to LNEd
6.15. Colophon of the commentary VC in the last sarga of the Utpattiprakaran.a
on folio 146: iti śr¯ıparamaham . saparivrājakācāryottamapūjyapādaśis.yaśr¯ımadātma-
sukhakr.tau vās.is.t.acam. drikāyām. lavan.opākhyānam . nāma trayodaśah. sargah. || 13.
Colophon of the commentary ST in the first sarga of the Sthitiprakaran.a:
iti śr¯ımadallālasūrisūnunā padavākyapramān.apārāvāradr.śvanā śr¯ımammud.ideva-
vidvadācāryaviracite sam . sārataran.¯ınāmni vāsis.t.avivaran.e śukropākhyānam . nāma
caturdaśah. sargah. || 14. Colophon of the commentary in the second sarga
of the Sthitiprakaran.a: iti śr¯ıparamaham . saparivrājakācāryaśr¯ımaccham . karācārya-
śr¯ımadātmasukhakr.tau vāsis.t.hacam . drikāyām . dāmādyupākhyānam . nāma pam . cadaśah.
sarggah. || 15.16 Colophon of the commentary of the last sarga of the
Upaśamaprakaran.a on folio 272: iti samsārataran.¯ınamn¯ıvāsis.t.hacam . drikāyām. ākā-
śagatyabhāvādinirūpan.am . nāmās t
.. havim . śatih . sargah
. || 28. Final colophon of
the mūla text on folio 407: iti vālm¯ıkiye vāsis.t.harāmāyan.e nirvān.aprakaran.e
saptabhūmikopākhyānam . nāma tricatvārim . śah. sargah.. Final colophon of the
commentary iti sam . sārataran . ¯
ınāmn¯ ı vāsis . t.havivaran.e nirvān.aprakaran.e yoga-
saptabhūmikopākhyānam . nāma tricatvārim . śattamah. sargah. followed by the gu-
rupraśasti, but only up to LNEd 6.16.1–3ab and counted separately as
verse 1 to 2. Date in the final colophon: idam . granthalekhan¯ıyam . tat
samāptilekhanam . || svastiśr¯ımannr.paśālivāhana || śake 1747 pārthiva nāma
sam . vatsare || jyes.t.aśuddhapratipat tithau saumyavāsare śr¯ıvā¯ıks.etre kr.s.n.ām . sani-
dhau samāptam ..

(Mumukśuvyavahārapra) | Gaud.abhinandayati | Ātmasukha | 36.1x15.9 cm
| 10 folios | 12 lines per page | 58 letters per line | complete | old”.
Description: 11 lines per page on an average, lower resp. upper margin
partly damaged, corrections with yellow pigment.
16 The commentary is the ST of Allād.asūnu as edited in LNEd .
Catalogue entry “6926 | 8-6/455 | Laghuyogavāsis.t.ha with Sam . sārataran.i

(Pra. 4-6) | Gaud.abhinandayati | Mummad.ideva | 35.8x16.1 cm | 257 fo-
lios | 13 lines per page | 34 letters per line | complete | good | śaka 1734”.
Description: 260 folios, lines per page vary from 5 to 13, occasional ad-
ditions in the margin by the scribe’s hand, occasional corrections with yel-
low pigment. First and ultimate cover folio indicating the transition of
Prakaran.as, Prakaran.as counted separately as well as the sargas are counted
separately in each Prakaran.a! This is a unique feature for a LYV ms., since
this type of enumeration was invented in the second edition of LNEd by
the editor. Text: Sthiti-, Upaśama-, and Nirvān.aprakaran.a with the commen-
tary Sam . sārataran.i by Mummad.ideva Allād.asūnu. Nirvān.aprakaran.a up to
LNEd 6.15 with the gurupraśasti. Colophon of the mūla text on LNEd 6.15.113
on folio 129: iti śr¯ıvālmik¯ıye moks.opāye vāsis.t.harāmāya yogasapta nirvān.a◦
bhūmikopā◦ pam . cadaśodhyāyah. || 15|| (correction in the margin: śah. sargah.)
Final colophon of the commentary: iti sam . sār.ataran.ināmni vāsis.t.havivaran.e
moks.opāye nirvān.apra◦ yogayākhyānam . pam. cadaśah. sargah. || 15|| samāptoyam .,
followed by the gurupraśasti counted separately, finally followed by: iti yo-
gavāsis.t.he nirvān.aprakaran.e samāptah. || śr¯ıkr.s.n.ārpan.am astu || śake 1734
am . girānāmasam . vatsare śrāvan.aśuddhapratipadyām. idam . pustakam . samāptam . ||
śr¯ıbhārgavarāmāya namah.
Catalogue entry “6927 | 8-6/455 | Laghuyogavāsis.t.ha with Sam . sārataran.i

(Pra. 6) | Gaud.abhinandayati | Mummad.ideva | 30x13.7 cm | 50 folios |
11 lines per page | 35 letters per line | incomplete | good”.
Description: 50 folios, 11 lines per page on an average. Occasional additions
and corrections in the margin. Text: Sarga 9 of the Nirvān.aprakaran.a of the
LYV with commentary Sam . sārataran.i of Mummad.ideva Allād.asūnu. Verses
in each upākhyāna counted separately. The ms. breaks off with folio 50 at LNEd
6.10.13.
Baroda
The catalogue of the renowned Oriental Institute, Palace Road, Baroda 390001
affiliated to the Maharaja Sayajirao (MS) University of Baroda lists 19 mss.
related to Yogavāsis.t.ha in general.17 Out of these, eight mss.18 were selected as

being of interest to the projects and thus digitized entirely.
1. Laghuyogavāsis.t.ha19
Catalogue entry “serial No 253 | accession No 6614 | yogavāsis.t.ham .
savyākhyam | mū. vālmı̄kih., vyā ? | 180 leaves | 4800 granthas | palm
leaf grantha up to udyālakāpākhyāna in upaśamaprakaran.a.”
Description: 186 palm leaf folios, 10 lines per page on an average. Text: LYV
with the commentary Sam . sārataran.i of Mummad.ideva Allād.asūnu. The ms.
breaks off on folio 186 with LNEd 5.6.166. Last colophon: iti vāsis.t.harāmāyan.e
uddālakopākhyānan nāma caturvim . śas sarga.

savyākhyam | mū. vālmı̄kih., vyā ? | 158 leaves | 3150 granthas | palm
leaf grantha sthiti, upaśama and 1st sarga of utpattiprakaran.a.”
Description: 158 palm leaf folios, 5 to 12 lines per page. At least two differ-
ent scribal hands, folios of the last part of the ms. appear to be more recent.
Text: Utpattiprakaran.a from 3.9.90 onwards, Sthiti-, and Upaśamaprakaran.a of
the LYV with the commentary Sam . sārataran.i of Mummad.ideva Allād.asūnu.
Last colophon on folio 158v: iti jñānavāsis.t.he upaśamaprakaran.e as.t.āvim
. śatis
sargah..

savyākhyam sam . sārataran.i | vyā. vedavidācāryah. | 252 leaves | 10000
granthas | palm leaf grantha injured.”
Description: 239 palm leaf folios, 10 lines per page on an average.
Partly damaged, mainly in the margin, worm eaten. Page sigla hardly
legible. Due to the poor state of preservation only a preliminary de-
scription could be made so far. Interlinear corrections, occasional ad-
17 Serial nos. 252–269 (Vol. II, pp. 846–848) and serial no. 344 (Vol. I, pp. 548).
18 The ms. listed in the catalogue as “serial No 256 | accession No 10747 | yogavāsis.t.ham .
savyākhyam | vyā. ānam . dabodhayattih. | 21 leaves | 1500 granthas | Telugu 16 sargas of
vairāgyaprakaran.a” was digitized, but could not yet be described. According to D IVANJI (1938),
p. 34, this ms. contains the first 16 Sargas of the Vairāgyaprakaran. a of the YV and a portion of the
17th together with the commentary VTP.
19 All the following mss. except No. 6614 have been described by D IVANJI (1938), p. 32f., and
D IVANJI (1939), p. 705f., and by S LAJE (1994), p. 127ff.

ditions in the margin, occasionally inserted leaves with less old appear-
ance. Text: LYV with the commentary Sam . sārataran.i of Mummad.ideva
Allād.asūnu, Utpattiprakaran.a apparently incomplete. Nirvān.aprakaran.a up to
LNEd 6.15.113. No final colophon of the mūla text. Final colophon of the com-
mentary: †āl.asūnunā padavākyapramān.apārāvāraśr¯ımadvedavidācāryen.a viracite
sam. sārataran.ināmni vāsis.t.havivaran.e †n.aprakaran.e yogasaptabhūmikopa†nāma tri-
catvārim. śat sargah.

savyākhyam vyā. vāsis.t.hacandrikā and sam . sārataran.ih. | vyā. ātmasukhah.
vyā. munmad.idevah. son of allālasūrih. | 806 leaves | 13000 granthas | Sam.
1739 vāsis.t.hacandrikā is up to the end of utpattiprakaran.a and then begins
sam . sārataran.ı̄.”
Description: 806 folios, 7 to 9 lines. Textual corpus of the com-
mentary not separated from the mūla text. Text: Vairāgya-, Mumuks.u-,
and Utpattiprakaran.a of the LYV with the commentary Vāsis.t.hacandrikā of
Ātmasukha and Sthiti-, Upaśama-, and Nirvān.aprakaran.a with the com-
mentary Sam . sārataran.i of Mummad.ideva Allād.asūnu. Explicit mūla text
identical with LNEd 6.15. Author’s name in the final colophon of the
mūla text on folio 802: iti śr¯ıtarkkavād¯ıśvarasāhityācāryasam . vidvivekavācaspati-
gaud.amam d
.. alālam. kārapam d
. . itaśr¯ıabhinam. danasamudhr . te vālmik¯ ıye mahārāmāyan.e
moks.opāye nirvān.aprakaran.e yogasaptabhūmikopākhyānam . nāma dvicatvārim. śat-
sargah. || 42. Commentary without final colophon followed by the 4 verses
of the gurupraśasti counted separately from 1 to 4 and the Sārasarga as printed
in LNEd (6.16), but with the first verse counted as number 1 again.20 One
additional “concluding” verse after LNEd 6.16.34 (numbered 31) almost iden-
tical with the final verse LNEd 6.18.84: idam . vairāgyamumuks.uvyavahārotpatti-
sthitayah. | upaśamanirvān.āv iti vasis.t.e [free space marked by the scribe]
ran.āni || 32. Scribe’s name and dating on the final folio 806: sam . vat 1739 vars.e
bhat.ajagannāthasutanat.abhān.ajikena likhitam idam . pustakam . | | phālguna śudi 8
śanivāsare samāptam . || śubham . bhavatu
Catalogue entry “serial No 258 | accession No 6394 | yogavāsis.t.hasaṅgraha-

sārah. | gaud.a-abhinandah. | 13-116 leaves | 5100 granthas | palm leaf
20 Since mūla text and commentary are not separated, it is impossible to ascertain whether the
scribe assumed the gurupraśasti to be part of the mūla text or of the commentary. However, the
numbering suggests that is was not considered part of the Sārasarga as printed in LNEd .
grantha. Incomplete. Injured. Otherwise known as vāsis.t.harāmāyan.a-

sam . ks.epah.”
Description: 129 palm leaf folios, 8 lines per page on an average. Since the
page sigla (counting from 13 to 116) do not maintain the ductus of the text
(see below) they apparently were added later. Folios partly damaged, mainly
in the margin, and worm eaten; interlinear notes and corrections, frequent
additions and corrections in the margin, occasionally inserted leaves by a sec-
ond scribe of a more recent appearance without page sigla. Since a satisfying
description would require a closer investigation of the contents of the ms.,
which is in a poor state of preservation, only a preliminary description could
be made so far.
Text: A hitherto unknown Sāra called “Vāsis..tharāmāyan.asam . grahasāra”,
which was most likely extracted from the YV/MU. It is not a commented
LYV as stated by others.21 This assessment is based on the following two
observations: 1. a high quantity of verses are not contained in the LYV,
but in the YV/MU. 2. no part of the text belonging exclusively to the LYV
could be detected so far – not even the Sārasarga LNEd 6.16, since the ex-
tracted text ranges from NEd 6.125/126 (=LNEd 6.15) to NEd 6.127 (=LNEd
6.17). Nevertheless it should be noted that the final colophon giving the
name of Abhinanda suggests some influence of the LYV tradition. The orig-
inal order of verses seems to be retained for the most part. Verses are ex-
tracted from all six Prakaran.as of the YV/MU including the uttarārdha of the
Nirvān.aprakaran.a. Text proceeds roughly as follows: Folios 13 to 45v con-
tain extracts from NEd 7.32 to NEd 7.214; the following folios, starting with
NEd 1.3.10, comprise extracts from the entire YV/MU up to NEd 7.215.17,
on folio 116 followed by the supposedly final colophon, which contains
the name of the supposed author: iti tarkkavād¯ıśvarasāhityācāryasam . vidviveka-
vācaspatigaurmman.d.alālam . kārapan.d.itaśr¯ıabhinandasamuddhr.te vālm¯ık¯ıye vāsis.t.ha-
rāmāyan.e moks.opāyasāre sam . grahavāsis.t.ha (corr. illeg.) rāmāyan.asam
. grahasāram .
sam ..
. pūrn.n.am
Catalogue entry “serial No 259 | accession No 12810 | yogavāsis.t.ha-

saṅgrahasārah. | gaud.a-abhinandah. | 204 leaves | 5400 granthas | Sam.
1656”
21 Compare D IVANJI (1939), p. 705, and S LAJE (1994), p. 127f.

Description: 204 folios, 10 lines per page. Corrections in the mar-

gins by the scribe’s and a second hand. Corrections with yellow pig-
ment. Text: LYV with the Nirvān.aprakaran.a up to LNEd 6.18 without gu-
rupraśasti. Page sigla in the left margin: vā śi. Colophon after LNEd
6.15: iti śr¯ıvālm¯ık¯ıye vāsis.t.harāmāyan.e nirvān.aprakaran.e saptabhūmikopākhyānam .
nāma tricatvārim . śattamah. sargah. sam. pūrn.ah.. Concluding verse after LNEd 6.16
(Sārasarga without gurupraśasti) almost identical with LNEd 18.84: iha vairāgya-
mumuks.uvyavahārotpattisthitayah. | upaśamanirvānāv iti vāsis.t.he s.at.prakaran.āni.
Colophon after LNEd 6.16: iti tarkavād¯ıśvarasāhityācaryyasam . siddhiviveka-
vācaspatiman.d.alākārapan.d.itaśr¯ıānandasamuddhr.te vālm¯ık¯ıye rāmāyan.e moks.opāya-
sāras sam . pūrn.ah.. The last two sārgas run as in LNEd 6.17 and 6.18 but
colophon of LNEd 6.17 after LNEd 6.18.5: iti vālm¯ık¯ıye moks.opāye bharadvājānu-
śāsanam . . Final colophon with date after LNEd 6.18 starts on folio 203: iti
tarkavād¯ıśvarasāhityācāryagaud.aman.d.alālaṅkāraśr¯ıabhinam . dasamuddhr.te śr¯ıvālm¯ı-
kiye vāsis.t.harāmāyan.e moks.opāyasāre bharadvājopākhyāne (left margin of folio
204 destroyed) †masam . vādah. sargah. || samāptaś cedam . ||
. vāsis.t.harāmāyan.am
sam . vat 1656 samaya †pam . cam¯ıbhaumavāsare likhitam idam . paramānam . dena || śr¯ı-
kāś¯ınāthāya namah..
2. One ms. wrongly associated with the Yogavāsis.t.ha

Catalogue entry “serial No 252 | accession No 13039 | mahārāmāyan.am | 14
leaves | 300 granthas | Sam. 197? 48-52 sargas.”
Description: 10 lines per page on an average. Text: Unidentified, styled
as a dialogue between Śiva and Pārvatı̄, and in the colophon allocated to
the Mahārāmāyan.a. The text refers frequently to Rāma, but no overlap with
any known version of the MU/YV could be made out. Final Colophon:
iti śr¯ımanmahārāmāyan.e umāmaheśvarasam . vāde dvipam
. cāsattamah. sargah. sam
. vat
197ke || śr¯ıjānak¯ıvallabho vij¯ıyate śr¯ırāmāya namah..
Finally one should add that we also endeavoured to trace MU, YV, and
LYV related mss. in the library of the Lalbhai Dalpatbhai Institute of Indology,
Navarangpur, Ahmedabad 380 009, having been alerted through a list of
supposed YV manuscripts that came to the attention of WALTER S LAJE two
decades ago, but could not be backtraced. Due to the kind support of the Di-
rector S HRI J ITUBHAI S HAH permission to consult a pre-selected list of mss.
and to digitize them, if necessary, was granted without complications. How-
ever, only two mss. transmitting the Yogavāsis.t.hasāra with the commentary of
Mahı̄dhāra22 are kept in the library according to the catalogue. Since cities
like Pune, Baroda and Ahmadabad house a multitude of mss. libraries – some
of them privately funded – it is not unlikely that the unidentified list refers
to one of these hardly known institutions. It is therefore difficult to predict
when the search for MU mss., if ever, will come to a close, unless of course the
IGNCA is successful in carrying out its large scale microfilming project.
22 Accession no. 22374 (25 folios) and 186361 (27 folios)

The Critical Edition of the Utpattiprakaran.a
A Research Report
P ETER S TEPHAN
The Moks.opāya (MU), an anonymous Sanskrit work composed in the 10th

century A.D. in Kashmir, was redactionally modified under the influence of
the Advaita School of Vedānta and became famous as the “Yogavāsis.t.ha” (YV).1
Although both versions still contain more or less the same text, the process of
revision has distorted its original philosophy and in the course of this process
the textual quality has suffered substantially. Therefore the published text of
the YV (NEd ) can not be used as a source unless one wishes to study specifi-
cally the text as used later by Advaita Vedānta. Only recently, with the help of
mss. not used before, the history of this text and its earliest available version
could be recovered.2 Thus, the current project of editing the MU will have
an impact on the comprehension of the philosophy of this text, which is still
based almost entirely on the YV-edition (=NEd ).
The aim of the present edition of the third “book” of the MU, namely the
Utpattiprakaran.a, is thus not a revision of the published text of the YV, but an
attempt to recover the philosophy of the text by rescuing as far as possible
the earliest version in the form of the Kashmirian MU-tradition. Addition-
ally, more information will be gained from the collected data, regarding the
regional variation of Sanskrit, its peculiar linguistic, vocabulary, syntax, and
semantic.
The sources
The identification of the MU-mss. is based upon variants characteristic of the
Kashmirian strand and the following formal criteria:3
1. The Vairāgyaprakaran. a begins with the second Sarga (1.2.1) of the YV and hence
omits the introductory narrative YV-frame.
2. Omission of the corresponding closing YV-frame.
3. The Utpattiprakaran.a extends to 4.18 of the YV, the Sthitiprakaran.a starts with 4.19
of the YV.
1 For further reference, compare: S LAJE (2001) and H ANNEDER (2003).
2 See above, p. 10.
3 Compare also, above, p. 42.
76 Peter Stephan: Critical Edition of the Utpattiprakaran.a
4. The Nirvān.aprakaran.a is not divided into Pūrva- and Uttarārdha.

5. Khilas (nānāpraśnāh.) are appended to the Nirvān.aprakaran.a.
The mss. listed below4 containing in fragments or as a whole the
Utpattiprakaran.a were initially considered as sources for the critical edition.
Nevertheless, their validity as textual sources had still to be estimated during
the process of editing the text.
Ś1 Mūla complete Facsimile ed. in C HANDRA (1984)

Ś3 Mūla complete Sri Pratap Singh Library, Srinagar
Ś6 T.ı̄kā 123-127 SBPK Berlin Hs. or. 12704
Ś7 Mūla complete IGNCA New Delhi RAR/181.045/MOK
Ś9 Mūla complete SBPK Berlin Hs. or. 12869
ŚSam
. Mūla extracts Göttingen Vish 126
N8 Mūla excluded BORI Pune Visram II/450
N10 Mūla excluded SBPK Berlin
N17 T.ı̄kā 123-140 SBPK Berlin Hs. or. 12704
N20 Mūla 101-133 BORI Pune Visram II/232
N21 Mūla 61-64 / 114-115 BORI Pune Visram II/224
The material of all the MU-mss. containing the Utpattiprakaran.a5 consists

of paper which is written on with ink. None of them is older than 300 years –
except maybe the birch-bark ms. described above6 which is not yet accessible
and hence not listed among the sources as above.
Ś1 was published in the form of a facsimile-version by L OKESH C HANDRA
in the series “Sanskrit Texts from Kashmir”. It seems to be a composition
of different mss., since various scribal hands can be traced within it.7 The
Utpattiprakaran.a was copied by a scribe different from the one who wrote the
Vairāgya- and the Mumuks.uvyavahāraprakaran.a in a comparatively easily legi-
ble hand and maintains a rather conservative textual transmission of peculiar
MU-readings. Additions in the margin are rare, but minor scribal mistakes
are frequent. Ś1 opens with the 24 introductory verses of Ānandabodhendra’s
4
H ANNEDER (2003), p. 289–291.
5
The mss. mentioned below have already been described more extensively in: S LAJE (1997)
and H ANNEDER (2003), p. 160–167.
6 See p. 43.
7 As described in S LAJE (1994) p. 39.
Peter Stephan: Critical Edition of the Utpattiprakaran.a 77
commentary on the YV (VTP)8 and his comment on the YV-maṅgala. Ś1 con-
tains only the introductory, but not the final part of the narrative YV-frame,
since the ms. is incomplete ending at 7.212.16a (=NEd ), i. e. sarga 212 of the
Uttarārdha of the Nirvān.aprakaran.a without any colophon. The addition of this
concluding frame is regarded as secondary: it meets the other criteria and
the readings are otherwise clearly that of the MU. Ś1 is therefore considered a
valuable source with regard to the content and condition of the oldest avail-
able MU-Version – at least as far as the Utpattiprakaran.a is concerned.
The scribal hand of Ś3 , of which we merely possess a microfilm taken in
the 1960s, is hard to make out and is even partly illegible. Ś3 contains very
few mistakes in terms of grammar and sandhi. It was obviously written un-
der the influence of the YV-recension since the introductory narrative frame
of the YV is included at the very beginning of the text, but the concluding
narrative YV-frame was skipped. The khilas, also called nānāpraśna, follow
immediately upon sarga 215 of the Nirvān.aprakaran.a. Furthermore Ś3 – as com-
pared to Ś1 – exhibits a stronger tendency to transmit textual variants in ac-
cordance with NEd . Nevertheless, considering its accuracy, Ś3 is estimated
to be a reliable source regarding the transmission of syntactically challenging
constructions and semantically peculiar readings. It seems that Ś3 was later
partly revised and affixed with the same glosses, interlinear notes, and post-
corrections which were added also to Ś7 . These additions produce evidence
of interpolation between Ś3 and Ś7 or their archetypes by a later redactor. Fur-
thermore verses absent from other mss. but contained in Ś7 and NEd are quite
often added in the margin of Ś3 .
Ś7 is well legible, with occasional scribal mistakes, and is influenced even
more than Ś3 by the YV-tradition regarding its readings, added corrections,
and verses contained in no other Śāradā-ms. but in the NEd .
Ś9 , discovered in the Janert-Collection, is not comfortable to read but main-
tains like Ś1 a conservative textual transmission in general quite independant
from the YV-recension.
The tendency of Ś1 to read in accordance with Ś9 and the tendency of Ś3
to read like Ś7 was observed over long passages. Conversely, a reverse ac-
cordance in some sequences was also noticed. This circumstance of a con-
taminated transmission prohibits the reconstruction of a stemma of the four
main mss.-sources. It seems rather likely that the scribes of the present or pre-
8 Vāsistha[mahārāmāyan a]tātparyaprakāśa (1710 A. D. ). Compare G OLZIO (∗ 2005) for further
.. .
information on the date of this commentary.
vious apographs were equipped with several ādarśa-pustakas, perhaps com-

posed already of independent fragments, from which they eclectically copied
the present mss. Some problems of restoring a text on the basis of contamined
mss. will be discussed below.
Besides these four major Śāradā-mss., four Nāgarı̄-mss. containing the
complete or parts of the Utpattiprakaran.a have been taken into considera-
tion for the critical edition. They are all – without exception – transcrip-
tions of Śāradā-mss. which becomes evident through peculiar writing mis-
takes. Two Nāgarı̄-mss., namely N20 and N21 , which contain fragments of the
Utpattiprakaran.a, were considered as relevant for the edition. However, N20
is not regarded as a source of strong evidence. Many scribal mistakes, most
of which can be attributed to wrong interpretations of Śāradā letters, prove
the dependence of N20 on Śāradā-mss. beyond any doubt. The number of
meaningless readings peculiar to this ms. show a lack of accuracy in N20 .
N21 contains but a few scribal mistakes. Its readings show a strong ten-
dency to agree with Ś1 , without sharing its unique writing mistakes. Some-
times N21 agrees with Ś1 even where the transmitted Ś1 readings differ from
those contained in a l l other sources available (61.21a, 22d, 24c, 30d, 34a;
62.6a, 8a, 10b, 14b, 15c, to mention just a few such incidences).
N8 , which contains Prakaran.as I–V, and N10 , which contains Prakaran.as I–
III, provide an abundance of strange readings in nearly every verse and thus
were ultimately eliminated. The type of mistakes raises doubts about whether
the scribes did understand at all what they were expected to write. Since again
there is no inclination of these mss. to read in accordance with a particular
ms. it seems likely that also N8 and N10 might have been produced out of
various Śāradā-mss. The introductory narrative YV-frame was nevertheless
not incorporated in N8 and N10 .
Like the T.ı̄kā-mss. Ś6 and N17 9 the ŚSam. has to be considered separate from
those mss. that contain the mūla-text. ŚSam. is a summary of the MU which con-
siderably shortens the text by excluding most of the ākhyānas.10 The remaining
parts are further trimmed by occasionally omitting single verses, groups of
verses, or even smaller sections. Although ŚSam. contains many variants char-
acteristic of the Kashmirian strand, it sets the end of the Utpattiprakaran.a in
accordance with NEd . Since it omits verses quite liberally, no claims about the
authenticity of single omitted verses can be made on the basis of ŚSam. . How-
9 Ś6 and N17 are described in S LAJE (2002), p. 12–13.
10 On this text, see below, p. 105ff.
ever, due to its considerably high textual quality, it is regarded as a source and
hence its readings are incorporated in the edition.
Editing the MU
In the following few sections, the more technical aspects of the production of
the edition, as well as methods and guidelines for establishing the critical text
will be presented.
Editing a text requires settling some standards, as for arranging the critical
apparatus or for dealing with orthography. Since the MU is of Kashmirian
origin, the respective regional orthography, which is likely to reflect pronoun-
ciation, is restored in the MU-edition. Thus visarga-sandhis are treated in ac-
cordance with Pān.ini 8.3.37: Assimilation of visarga (h.) to the sibilants “ś” “s.”
“s”, jihvāmūl¯ıya (h) before surd gutturals (“k” and “kh”), and upadhmān¯ıya (h)

before the surd labials (“p” and “ph”).
Visarga-sandhis and assimilated class-nasals instead of anusvāra are ap-
plied in the critical text without documenting the actual variants of each ms.
in the critical apparatus. The orthography of variants entered in the apparatus
is nevertheless preserved. In case the same reading is provided by more than
one ms. each using different orthography, the orthographic variation that is
closest to the defined standard is entered in the critical apparatus.
The MU-edition separates auxiliaries from derivative noun-stems, when
periphrastic tenses are made following As.t.ādhyāy¯ı 3.1.35, 40 (anuprayukta), e.g.
¯ıks.ām. cakre or cintayām āsa.
Particles conveying an indefinite meaning such as cid are deliberately not
printed seperately. Since the noun cit is a frequently occurring keyword in the
MU it might confuse the reader if the particle “cit” (from cid through sandhi),
cannot be distinguished at first sight from the noun “cit”.
Since the syllabic units “s..th” and “s..t” are homograph in Śāradā-mss., the
Nāgarı̄-mss. N20 and N21 transcribe “s.t.h” as “s.t.”. These mistakes were not
recorded in the critical apparatus. Additionally, N20 does not distinguish
clearly between “ch” and “cch” (only twice, namely in 111.16 and 126.4cd)
and transcribes “cca” usually as “śca”. Mistakes of this kind are noted only
where they provide a semantically meaningful variant.
In the absence of variant readings the orthographic standardization is tac-
itly applied to the critical text, whereas the variants recorded in the apparatus
inform about the actual orthography in the mss.
Arrangement of the critical apparatus The critical apparatus provides all

data available in the mss. The accepted reading, rejected variants (printed
in italics) as well as glosses, interlinear notes etc., are given in the apparatus.
In order to render the apparatus comprehensible and to enable the reader to
grasp each variant in its particular context, full syntactic unities are presented.
The following sample shows the main features of the critical edition de-
scribed above, like orthographic standardisation and arrangement of the crit-
ical apparatus.
yato yato viyujyate tatas tato vimucyate

ato ’ham ity asam
. vidah ka eti janmasam
. vidam (61.35)

35a viyujyate Ś1 Ś3 Ś9 ] virajyete Ś7 , viyucyeta N21 35b tatas tato
Ś1 Ś3 Ś7 N21 ] tatatas tato Ś9 35c asam . vidah. Ś1 Ś7 Ś9 N21 ] asam
. vidan Ś3 (=NEd )
35d ka eti Ś1 a.c. Ś3 Ś9 Ś7 N21 ] kacati Ś1 p.c.
The jihvāmūl¯ıya (h) in pāda 35c asam. vidah is used in the critically edited
text even though the mss. use visarga instead. If all mss. had read asamvidah
. .
no variant would have been reported and the critical text would simply read
asam. vidah. In pāda 35b the lemma is given as tatas tato, which means that at
least one manuscript reads tatas tato, the others could well have the ortho-
graphical variant tatah. tato, but this is not reported in order to relieve the ap-
paratus of divergent sandhis.
Philological Commentary The critical edition of the text is accompanied by a

philological commentary which provides syntactic and semantic explanations
to make the editors’ understanding and choice of variant transparent. In view
of the extent of the text the explanations are reserved for instances where the
establishment of the text and its underlying comprehension require further
reasoning, justification or interpretation. Grammatical and morphological di-
vergences from the pān.inian standards, as well as lexically striking items are
also recorded here. The underlying idea of the internal structure of the com-
mentary is to move from textual criticism to exegesis. Thus, a commentary
referring to verse and pāda starts with the discussions of variants, proceeds to
explanations on syntax, grammar and lexicography, and concludes with ex-
egetical notes as well as further references from other parts within the text11
or such references to other texts.
11 Parallels from the unedited parts of the MU are given by reference to the mss. consulted.
Methods of text-constitution The fact that the MU manuscript transmis-

sion is contaminated with readings characteristic of the later YV-tradition has
been mentioned. Apart from the usual scribal mistakes, additions and textual
changes were applied to the text in order to render it more consistent with
some concepts of the Śaṅkara school of Advaita.12 The following examples
serve to illustrate the state of contamination of the textual transmission in the
MU-mss. and the stratification of variants.
. cet parin.amyate
niyatir brahmatattvābhā tasyām
nūnam. paramaśuddhāyām. tat prāptaiva parā gatih. (62.34)
34a brahmatattvā Ś1 Ś9 N21 ] brahmasattā Ś3 Ś7 (=NEd ) 34b cet Ś3 Ś7 Ś9 ] cit Ś1
aham . tvam ityādi jagatsvarūpā viśuddhabodhaikavibhā vibhāti

ākāśakośe nijaśūnyataiva dvaitaikyasaṅkalpavikalpanāh kva (60.63)

63b interlinear note: d¯ıptih. Ś9 63c śūnyataiva Ś1 Ś7 ] śūnyateva Ś3 Ś9 (=NEd )
63d dvaitaikya Ś3 Ś7 Ś9 ] dvaitye ’pi Ś1
yena yena yadā yad yad yathā sam . vedyate ’nagha

tena tena tadā tat tat tathā samanubhūyate (60.16)
16a yadā Ś3 Ś7 ] yathā Ś1 Ś9 (=NEd ) 16c tadā Ś3 Ś7 ] tathā Ś1 Ś9 (=NEd )
16d tathā Ś3 Ś7 ] tadā Ś1 Ś9 (=NEd ) 16d samanubhūyate Ś3 Ś7
Ś9 ] sanamanubhūyate Ś1
cidan.oh paramasyāntas sad evāsad iva sthitam

sattāpy evam asatteva sphurat¯ıdam
. jagat sthitam (80.5)
5b sad evāsad Ś3 ] sad ivāsad Ś1 Ś7 Ś9 (=NEd ) 5b iva sthitam Ś1 Ś9 ] ivāpi vā
Ś3 (=NEd ) , avasthitam Ś7 5c sattāpy evam asatteva Ś1 Ś9 ] b¯ıje ’ntar drumasat-
teva Ś3 Ś7 (=NEd )
The first sample (62.34) shows the general tendency of Ś1 , Ś9 , and N21 to
read in accordance and against Ś3 and Ś7 , which on the other hand tend to read
in line with NEd . The second sample (60.63) shows the reverse accordance in
63c. The third sample (60.16) demonstrates an instance where the readings
of Ś1 and Ś9 are distinctively in line with NEd against Ś3 and Ś7 , while the
12 Refer to H ANNEDER (2000), H ANNEDER (2003) and S LAJE (2001) for a recapitulation of this
examination.
fourth sample (80.5) illustrates the actual contamination over major parts of
the MU-mss. It is obvious that the stratification of the readings in the MU-
mss. does not allow the production of a stemma. It seems most likely that
some sections of the mss. were compiled by using a set of ms.-sources of a
certain line of the transmission while other sections used ādarśa-pustakas of
another line of transmission. This could explain the changing accordance of
readings diagnosed among the mss.-groups. Later scribes or redactors must
have used manuscript sources of the YV-tradition and thus have contamined
the MU-sources with YV-readings. These mss. might have again provided
the sources for the compilation of the next generation of MU- mss. This way
of producing apographs on the basis of contamined sources resulted in the
present contamination of MU-mss. with readings of YV-mss.
It is particularly challenging to attempt a critical edition of a text which is
contaminated in the way described. Since in this situation of contaminated
transmission no stemma and thus no group of mss. more or less free of con-
tamination can be traced and hence the argumentation for the authenticity of
a certain variant solely based on the observation that it is transmitted by an
older group of mss. apparently “closer” to the original is not possible. The
editor has n o t o n l y to use well established criteria of textual criticism,
b u t has also to draft further criteria to prove the authenticity of a certain
variant. Above all, the editor has to understand the motive for the scribe’s
preference of a certain reading from among the variants in the mss. at his dis-
posal. As S RINIVASAN has shown with his critical edition of Vācaspatimiśras
Tattvakaumud¯ı,13 the genesis of variants in conflated texts can be determined
and hence a critical edition of a text from contaminated sources can indeed be
accomplished. The criteria developed by S RINIVASAN have been adopted for
editing the MU.
The Importance of the lectio difficilior The lectio difficilior is still one of the
most valuable criteria for establishing a given reading as authentic, since gen-
erally no scribe is expected to complicate the text wilfully, but rather tends to
simplify it. Scribal alterations may occur undeliberately due to misinterpre-
tation of letters, eye-skip, subconscious contextual expectations and the like,
or due to wrong comprehension of an expression or of the syntactic structure
of a certain sequence, which could cause the scribe to reproduce the misap-
prehended section according to his own understanding. Most of such alter-
13 Compare: S RINIVASAN (1967).
ations can be traced by investigating the plausible generation of the variants,

given the authentic reading still exists as one among the variants. Consider-
ing semantics and paleography one has to determine whether variant A was
more likely the source of variant B, or vice versa. Deliberate alterations of the
text, induced by an educated scribe’s urge to “correct” something, are more
difficult to trace. In this case the lectio difficilior might not necessarily be the
authentic reading.
For this reason both the lectio difficilior and the reading, which one suspects
to be a scribal simplification, have to be further examined regarding their con-
textual consistency, as will be described below. In case none of the variants
is the obvious lectio difficilior, the primary reading can only be determined by
contextual consistency, or the question has to be left open, as in the case of
synonymous expressions.
Textual Consistency The author of the MU exhibits a unique style in his oeu-
vre, aspects of which have been discussed elsewhere.14 His preference for
onomatopoeia, unusual constructions of compounds, and his tendency to con-
strue syntactic units extending beyond pāda-limits provide chances to redis-
cover his genuine style among variants generated by later scribes or redactors.
Besides these features there are also criteria for distinguishing MU-variants
from alterations inspired by the YV-tradition through contents.15 Variants in
accordance with the supposed author’s literary style and consistent with re-
gard to the contents are accepted as authentic.
In dubio contra NEd pro Ś-MU As shown above, contaminations with read-
ings of the YV-version are omnipresent also in the MU-mss. Thus, wherever
none of the methods discussed above produces evidence for or against a cer-
tain variant, the variants also transmitted by NEd are rejected. This principle is
applied to exclude the interpolation of those readings in MU-mss.,16 because
YV-variants are in general suspected to be less authentic.
14 Compare: S HASTRI (1968), S HASTRI (1980–81), S HASTRI (1963), S HASTRI (1975), S HASTRI
(1979) H ANNEDER (2000), p. 205.

15 The MU-ontology advocates an a-cosmic (ajāti) subjective illusionism, in its soteriology an
active life and no inclination to ascetism, in its epistemology it shows a favour for rationalism and
is contextually situated in a rather heterodox environment (cp. H ANNEDER (1998b), H ANNEDER
(2003) and S LAJE (2001)).
16 For the explanation of this process, see above, p. 53.
Results
This concluding section summarizes some preliminary results concerning the
extension of the Utpattiprakaran.a and illustrates the textual quality of the edi-
tion in comparison to the YV-edition.
The Extension of the Utpattiprakaran.a in the MU There are no substantial

criteria which help ascertain the end of the Utpattiprakaran.a and the begin-
ning of the Sthitiprakaran.a. The MU-T.ı̄kā-mss. consider the Utpattiprakaran.a to
end with sarga 122 (paramārthanirūpan.am . ), and the Sthitiprakaran.a to start with
the next sarga called janyajanakanirākaran.am . as does the NEd -version. Conse-
quently they insert a pratisandhi-śloka17at the beginning of this sarga. This tra-
dition of terminating the Utpattiprakaran.a after sarga 122 is transmitted also by
ŚSam. .18 The remaining mss. of the mūla-text do not set the end of the Utpatti-
prakaran.a after sarga 122, but after sarga 140. Thus, later redactors added their
justification in mss.19 Ś3 and Ś7 proposing that the Utpattiprakaran.a should end
here, at the end of sarga 122.20 The scribe of Ś1 seems to have been confused
about the end of the Utpattiprakaran.a as he first determined it at the end of
21
sarga 126, which is called sthityaṅkurakathanam . . This confusion might have
been caused by the name of sarga 126 (sthityaṅkurakathanam . ) as well as by the
first verse of the next sarga introducing a new topic with Rāma’s rhetorical
question about the ontological status of the world:22
bhagavan sarvadharmajña pūrvāparavidām
. vara |
17 Serves to indicate the transition to a new Prakaran.a:
athotpattiprakaran. ād anantaram idam . śr.n.u
sthitiprakaran. am . rāma jñātam. nirvān.akāri yat
18 Ś
. adds in the colophon of sarga 122 (ŚSam
Sam . folio 157v): samāptam idam utpattiprakaran. am.
19 All the colophons and scribal annotations found in the mss. are henceforth reproduced in
diplomatic transliteration.
20 Ś adds in the margin (Ś p. 171): iti vāsisthabrahmadarśanam moksupāy[e]su utpattiprakaran am
3 3 .. . . . . .
samāptam atah. param . sthitiprakaran. am. †py ati †yam ādiśloka evam. tāvad idam . ityā †.
. dr.śyam
Ś7 adds in the margin (Ś7 folio 240v): iti vāsis.t.he brahmadarśane moks.opāyes.u utpattiprakaran. am .
samāptam | atah. param . sthitiprakaran. am. bhavis.yati | yasyāyam ādiślokah. evam. tāvad idam
. dr.śyam ityādi
and “corrects” the marginal pagination of the ms. which also indicates the Prakaran.a at issue by
the annotation: atah. param . sthitiprakaran. am . . The nearly identical formulation of this annotation in
both mss. is just another evidence for a later redaction of Ś3 and Ś7 , perhaps by the same person.
N20 on the other hand just mentions the name of the sarga: utpattiprakaran. e paramārthanirūpan. am .
(N20 folio 240r) without any notice about a possible end of the Utpattiprakaran. a.
21 samāptaś cāyam utpattiprakaran am | om namo bhagavate vāsudevāya (Ś p. 325).
. . . 1
22 Mentioned above, p. 43
ayam manasi sam

. sāras sphārah katham iva sthitah. (3.127.1)

The scribe of Ś1 thus considered this verse to be the first of the
Sthitiprakaran.a and thus indicated the conclusion of the Utpattiprakaran.a and
the start of the new Sthitiprakaran.a, but did not add the pratisandhi-śloka. Fi-
nally, at the end of the bhārgavopākhyāna in sarga 140 called j¯ıvas.an.d.akāvatāra,
the scribe of Ś1 adds: samāptaś cāyam utpattiprakaran.am
. sampūrn.om | śubham
astu sarvajagatām
..
Although the tradition which ends the Utpattiprakaran.a after sarga 122
seems to have left traces in the colophon of sarga 140 in Ś3 , the writer finally
decides to terminate the Utpattiprakaran.a after sarga 140.23 Also in Ś7 24 and Ś9 25
the Utpattiprakaran.a ends with sarga 140. No further comments concerning the
end of the Utpattiprakaran.a are added in the margins of Ś7 and Ś9 at the end of
sarga 140.
Since the tendency to end the Utpattiprakaran.a with sarga 140 and thus to
include the bhārgavopākhyāna is obviously much stronger – at least in those
Śāradā-mss. which contain the mūla-text – than to conclude with sarga 122 this
division was also maintained in the critical edition of the Utpattiprakaran.a.
23 Colophon sarga 140 (Ś3 p. 190): iti śrı̄moks.opāyes.u vāsis.t.he brahmadarśane mahārā[māyān. e]
bālakān. d.akhiles. u devadūtoktau śrı̄vasis.t.harāmasam. vāde śrı̄madādikavivālmı̄kimahars. iviracite [deleted
sequence] sthitiprakaran.e jı̄vas.an.d.akāvatārah. sargah. 18 | samāptam . cedam utpattiprakaran. am .
tritı̄yam . cintitam . cedam . pan.d.itabhat.t.ottamasatkr. tipūrn.ahr.dayena śrı̄rāmacandren.a || atah param .
sthitiprakaran. am
. bhavis.yati | yasyāyam ādiślokah. āditah. sargah. 193 |
athotpattiprakaran. ād anantaram idam . śr.n.u |
sthitiprakaran. am . rāma jñātam . nirvān.akāri yat ||
svanirvr.tisamullāsasaccamatkr. tighūrn.itam |
kam . vā na mādayel loke śrı̄vāsis.t.hāmr.tāsavah. |
kr.tih. pan.d.itabhat..taśrı̄rāmacandrasya.
24 Colophon sarga 140, which is actually counted as sarga 139 in Ś (folio 269r): samāpta cedam
7 .
utpattiprakaran. am | śubhāya bho bhavı̄tarām iti lekhakapākayoh. |
bhagnapr.s..thakat.igrı̄vas tus..tadr.s..tir adhomukhah. [|]
kas.t.ena lipitam. grantham . putravat paripālyatām
25 Colophon sarga 140 (Ś folio 400r-v): samāptas utpattiprakaran am | śrı̄rāmabhadrāya namah |
9 . . .
kūjantam . rāma rāmeti madhuram . madhurāks . aram |
āruhya kavitāśākhām . vande vālmı̄kikokilam ||
śrı̄rāmacandras sa punātu nityam . yan nāma madhyendraman.im
. vidhāya |
śrı̄candramuktāphalayor amāyāś (?) cakāra kan.t.hābharan.am
. girı̄śah. ||
atha sthitiprakaran. am . likhyate.
Two examples of the restored MU-Text will be presented below to compare

their textual quality to the YV-version transmitted in the NEd . The following
samples were selected as paradigms since they demonstrate how the original
ideas of the MU have been distorted in the NEd by seemingly small textual
alterations.
Example 1 A living creature comes into existence by mere imagination within

the absolute, which is Brahman consisting of pure consciousness. Its creation
comes into being as purely accidental, just as meaningless babble originates
in a child’s mind. This is one comparison used in the MU to teach Rāma how
a creature comes into existence. It implies the absense of a particular purpose
or plan for creation. Creation happens spontaneously due to imaginative ac-
tivity, which is the nature of the absolute.
hullas sabhullo ghullāṅgha iti bālahr.di sphut.am

yathodeti tathodeti pare brahman.i j¯ıvatā (67.40)
40ab sabhullo ghullāṅgha Ś1 Ś9 ] satullaghulāṅga Ś7 a.c. , sabhullaghullāṅga Ś3
Ś7 p.c. 40b sphut.am Ś1 Ś7 Ś9 ] illegible Ś3
Although three variants of the babble are transmitted in the MU-mss. each
of them maintains more or less the genuine intention of the upamāna, which is
an alliterative utterance consisting of deliberately meaningless phonetic com-
ponents. With just a few “corrections” the NEd -text replaces these utterances
by seemingly meaningful words and distorts the intent of the upamāna used
in the MU:
ullāsaphullo phullāṅga iti bālahr.di sphut.am

yathodeti tathodeti pare brahman.i j¯ıvatā (67.41)
Without knowledge of the MU-version of this verse, the reader might won-
der why exactly a rather poetic sentence combining ullāsa “shining” and phulla
“blossoming, sprouting” – in a way syntactically open to a wide range of in-
terpretation – should be so unique to a child’s or any other “weakminded”
person’s (bāla) mind that it serves Vasis.t.ha as comparison for Brahman’s cre-
ative activity. In the commentary of Ānandabodhendra on the YV (VTP), a
rather unconvincing exegesis of this “corrected” verse is offered. To add more
content to this originally meaningless phrase the commentator superimposes
on it the well known topic of a child being afraid of an imagined demon.26 Ob-
viously, the idea of an absolute which creates planlessly and without intention
was either not understood by later scribes or redactors, or considered disre-
spectful. In the same way Vasis.t.ha’s untranslatable sentence in verse 3.11.7 of
the MU, which explains to Rāma the ontological status of the world as a mere
illusion operating with non-existing entities, was distorted in the YV and fur-
nished with significance by Ānandabodhendra’s exegesis, as already depicted
elsewhere.27
Example 2 Vasis.t.ha instructs Rāma about the purely cognitive character of

the phenomenal world by adducing the example of the travellers’ experience
of a boat trip: The boat seems to move only in the perception of the travellers.
If they stopped perceiving, the boat would not move any longer, although the
travellers would still take part in the shaky trip. Perception or sensation is the
cause of their affliction. The verse runs like this:
naur yāyinām . vedanāt tu vivartate

. bhramārtānām
avedanād bhramārtānām api nais.ā vivartate (60.30)
30a naur yāyinām . Ś1 Ś7 Ś9 ] nauyāyinām

. Ś3 (=NEd ) 30b vedanāt tu Ś1 Ś7 a.c.
Ś9 ] vedanād bhūr Ś3 Ś7 p.c. (=NEd ) 30c avedanād Ś3 Ś7 Ś9 ] avedanā Ś1 (=NEd )
30c bhramārtānām Ś1 Ś3 Ś7 p.c. Ś9 ] bhūmārtānām Ś7 a.c.
For [nautic] travellers, affected by the whirling movement, the boat

rocks because they are conscious [about it]. It does not sway, even for
those affected by [its] whirling, because they are not conscious [about it].
The YV-variant (=NEd 60.30cd–31ab) differs only marginally from the MU-
version:
nauyāyinām . vedanād bhūr vivartate

. bhramārtānām
avedanābhramārtānām api nais.ā vivartate
Here two small changes triggered a entirely new interpretation, the loss
of the ending in naur, which then formed a compound with the next word,
26 . . . aphullāṅgo (the mūla-text should read correctly ’phullāṅgo according to Ānandabodhendra)
vastutah. phullāṅgaśūnyo ’pi phullāṅgo vetālo yathā bālahr. di sphut.am udeti tathodetı̄ty anvayah. (NEd
Vol. I, p. 294).
27 Compare: H ANNEDER (2000), p. 193–195.
and – in the absence of a subject – the reading bhūr replaced tu. But the idea
in the MU is quite different from that of the commonly known phenomenon
expressed in the YV-Version. While the MU-Version describes the situation of
being involved in the movement of a boat – maybe even of being sea-sick –
as being solely dependant on the traveller’s conscious experience (vedanāt),
the YV-version emphasizes that the particular experience of the outer world
is dependent on the point of observation. The MU argues that the movement
of the boat would stop if nobody experienced it consciously (avedanāt), which
implies that this might be the way to liberation from world experience itself.
Ānandabodhendra, on the other hand, holds that for those at another point
of observation, i.e. for those who are on the bank of the river and are not sub-
ject to the shaky boat-trip, the world appears different, i.e. for them the earth
does not move.28 More precisely, according to the MU this phenomenon origi-
nates due to the individual’s cognition, i.e. the phenomena’s ontological status
is a subjective one. The YV-version, on the other hand, does not question the
ontological status of the phenomenon, but their erroneous interpretation due
to the individual’s cognitive misconception (avidyā).
It is difficult to decide which reasons might have caused this change
in the YV-tradition29 but doubtlessly the genuine idea of the MU-version
expressed in this verse was completely distorted by the YV-version and
Ānandabodhendra’s interpretation.
Conclusion
The work on the critical edition of the Utpattiprakaran.a underlines on the ba-
sis of more material the necessity of a complete critical edition of the MU,
which has already been advocated by others.30 Since the first critical edition
of the Utpattiprakaran.a is taking on shape, it seemed appropriate to inform the
participants of the Moks.opāya panel about its aim and scope, the underlying
methods of textual criticism and its intended production. The editors hope
28 avedanābhramārtānām [=] vedanābhramārtiśūnyānām esām tı̄rasthānām tu drstyā [bhūr] na vivar-
. . . . ..
tate (NEd Vol. I, p. 278–279).
29 Paleographic factors as well as subconscious expectations of the scribe towards the content
could probably have caused the generation of this YV-variant. First the omission of “r” might
have brought about the compound nauyāyinām . . The ensuing lack of a subject, led to the replace-
ment of tu by bhūr, most likely in order to produce a meaningful sentence, which was plausible
to the scribe, since it matched conventional concepts.
30 Compare: S LAJE (2001) and H ANNEDER (2000).
to initiate a process of further studies on the history and the philosophy of

this text as well as on its peculiarities in linguistics, vocabulary, syntax, and
semantics.
Vakatseite
Sāras, Sam
. grahas und ‘Laghus’: Kurzfassungen des
Moks.opāya1
S USANNE S TINNER
Angesichts der Textmasse und Komplexität des Moks.opāya (MU) bzw. sei-
ner sekundären Version, des Yogavāsis.t.ha (YV),2 verwundert es nicht, daß ge-
lehrte Rezeptoren in ihrer Auseinandersetzung mit dem Werk das Bedürfnis
empfanden, das jeweils als essentiell Betrachtete zu definieren und festzu-
halten: Es entstanden Auszugswerke, die zum Teil ihrerseits Grundlage wei-
terer Zusammenfassungen wurden. So setzt beispielsweise der mit zahlrei-
chen Mss. überlieferte Vāsis.t.hasāra (VS, vor 1597)3 die wohl bekannteste al-
ler Kurzfassungen, das sog. Laghuyogavāsis.t.ha (LYV, nach 950) voraus, wäh-
rend die (bislang unbekannte) Śr¯ıvāsis.t.hacandrikā (Ms dat. 1660) auf Material
aus beiden genannten Texten beruht. Zwei umfangreichere Werke hingegen –
das bislang in mindestens zwei Mss. identifizierte Jñānavāsis.t.ha (JV)4 und ein
erst kürzlich gefundener “Vāsis.t.harāmāyan.asam. grahasāra” – gehen weit über
den im LYV enthaltenen Textumfang hinaus und dürften eine Langversion
(MU/YV) zur Vorlage gehabt haben.
Als erstes Kriterium zur Kategorisierung der Auszugswerke bietet sich ih-
re beträchtlich differierende Länge an: Kürzungen größeren Umfangs, die in
Aufbau und Inhalt noch wesentliche Elemente der langen Versionen bewah-
ren, stehen regelrechte Kurzkompilationen von nur einigen hundert Versen
gegenüber.5
Inhaltliche Tendenzen bei der Textauswahl werden schon in dem auf etwa
ein Sechstel des Gesamttextes reduzierten LYV an der Auslassung unortho-
1 Der vorliegende Beitrag ist ein Teilergebnis des von der Deutschen Forschungsgemeinschaft
(DFG) geförderten, an den Instituten für Indologie und Islamwissenschaft der Martin-Luther-
Universität Halle-Wittenberg lokalisierten Projektes “Indo-persische Übersetzungsliteratur aus
der Mogulzeit (16./17. Jh.)”.
2 Zum Verhältnis der beiden Versionen siehe z.B. den Beitrag von J. H ANNEDER auf S. 9ff. in
diesem Band sowie S LAJE (1994), S. 131–134.

3 Hrsg. und übersetzt von T HOMI (1999). Zusätzliche Mss. sind aufgeführt in dem Artikel von
W. S LAJE in diesem Band auf S. 37ff., siehe dort (S. 39) auch zum Werktitel „Vāsis.t.hasāra“.
4 Mss. beschrieben bei S LAJE (1994), S. 41f.
5 Der vorläufig jüngste Text dieser Art ist der in den 70er Jahren des 20. Jhds. von
J Ñ ĀN ĀNANDA B HĀRAT Ī kompilierte Vāsis.t.hasam

. graha (herausgegeben und übersetzt von P ETER
T HOMI ; vgl. T HOMI (1988b) und T HOMI (1988a)).
92 Susanne Stinner: Kurzfassungen des Moks.opāya
doxer Standpunkte bei gleichzeitiger Betonung traditioneller Elemente deut-

lich.6
Von der überlieferten Struktur des Langtextes behält das LYV die Eintei-
lung in die sechs Hauptkapitel (prakaran.a) und in entsprechend wenigere,
verkürzte Unterkapitel (sarga) ebenso bei wie einen Teil der gleichnishaften
Erzählungen (ākhyāna), die im Kolophon („upākhyāna“) zumeist den jeweili-
gen Sarga betiteln.7 Hingegen sind die zehn Sargas bzw. neun Prakaran.as der
beiden als „Kurzkompilationen“ charakterisierten Texte – des VS und der
Śr¯ıvāsis.t.hacandrikā – als rein formale Anlehnung zu betrachten. Beide Texte
epitomieren (oft) isolierte Verse aus allen Prakaran.as und stellen diese in einen
neuen thematischen Zusammenhang. Die ursprüngliche Abfolge des Textes
geht dabei verloren – sowohl innerhalb, als auch zwischen den einzelnen
Kapiteln. Die Śr¯ıvāsis.t.hacandrikā8 ähnelt in den Kapiteln zwei bis neun hin-
sichtlich dieser Struktur dem VS. Ausschließlich VS-spezifische (und YV/MU-
fremde) Verse beschränken sich auf lediglich drei.9 Sonst beruht diese Kurz-
fassung durchgängig auf Material des LYV und bewahrt nicht nur die durch
häufige Sprecherzuschreibungen markierte Dialogsituation zwischen Rāma
und Vasis.t.ha, sondern übernimmt auch in dem 80 Verse umfassenden ersten
Kapitel („ [. . . ]vairāgyotpattih. prathamam. prakaran.am
. “) beinahe den gesamten
ersten Sarga des Vairāgyaprakaran.a aus dem LYV. Dadurch bleibt die Einbin-
dung des Lehrtextes in die Handlung des Rāmāyan.a prominent erhalten,10 was
angesichts der starken Reduzierung des übrigen Textes sicher intendiert war.
6 Eine solche Tendenz ist die Unterdrückung schicksals(daiva)-kritischer Aussagen im Hin-
blick auf die Rolle der menschlichen Entschluß- und Tatkraft (paurus.a) für die (Selbst-)Erlösung
(vgl. S LAJE (1998)). Traditionelle Rāma-Bhakti hingegen wird besonders deutlich in den beiden
sekundären Schlußrahmen vertreten.
7 Bei D IVANJI (1938), S. 34f., sind die Fälle aufgeführt, in denen einzelne Verse in anderen
Sargas bzw. Ākhyānas erscheinen.

8 Diese auf Samvat 1716 / Śāka 1582 [=AD 1660] datierte, nur 14 Folios enthaltende Hs. wur-
.
de im Zuge der Forschungsarbeiten an dem DFG-geförderten Projekt „Kritische Edition des
Utpattiprakaran. a“ in den Beständen des Indira Gandhi National Centre for the Arts gefunden.
Der Text ist weder identisch mit dem gleichnamigen Kommentar des Ātmasukha zum LYV (dazu
s.u., S. 96), noch, wie auf der Registrierkarte irrtümlich angegeben, der „Jaina Philosophy“ zuzu-
ordnen.
9 ŚVC 1.1=VS 1.1; ŚVC1.5=VS 1.3; ŚVC 1.8=VS 1.9.
10 Die Textgeschichte des MU ist durch das schrittweise Hinzufügen mythischer Rahmenhand-
lungen gekennzeichnet (vgl. S LAJE (1994)). Bearbeitungsstufe „D“, eingeleitet durch den Vālmı̄ki-
Bharadvāja-Mythos, bewirkt die Inkorporation des Textes in das epische Rāmāyan. a (S LAJE (1994),
S. 106–117). Das LYV setzt diese Stufe bereits voraus (S LAJE (1994), S. 125–134) und führt in den
mythischen Plot mit dem LYV-spezifischen Vers 1.1.3 ein.
Susanne Stinner: Kurzfassungen des Moks.opāya 93
Im Gegensatz zum VS11 schöpft die Śr¯ıvāsis.t.hacandrikā ausschließlich aus

dem LYV. Das im VS enthaltene MU/YV-fremde Material umfaßt beinahe ein
Drittel des Gesamtumfangs.12 Ein weiterer Unterschied besteht vor allem in
der historischen Verbreitung beider Texte: Während vom VS Dutzende von
Abschriften existieren, zeigt sich die bislang einzige Hs. der Śr¯ıvāsis.t.hacandrikā
als Zufallsfund. Ein solcher ist auch der auf einem Palmblatt-Ms.13 erhalte-
ne Text, der in einem Kolophon als ◦ vāsis.t.harāmāyan.asam
. grahasāra bezeichnet
wird. Dieser Sāra besteht zu großen Teilen aus nicht im LYV enthaltenem Vers-
material einer YV-Version. Umgekehrt konnten keine LYV-spezifischen Pas-
sagen gefunden werden: Es handelt sich also um eine weitere, unbekannte
Kurzversion.14
Zusammenfassend zeichnet sich in der handschriftlichen Überlieferung
von auf den Langversionen (MU/YV) beruhenden Texten eine regelrechte
Kürzungsliteratur ab, in der sich eine Fortsetzung der umfassenden Umdeu-
tung und Redaktion des MU15 andeutet und deren innere Erschließung noch
im Beginn begriffen ist. Den bekannteren Teil stellen die mit zahlreichen Mss.
überlieferten und dann durch Drucklegung verbreiteten Werke VS und vor
allem LYV dar. Der Rückgriff auf das mehrheitlich mit dem LYV assoziier-
ten Handschriftenmaterial enthüllt die Existenz bislang unbekannter Kurz-
texte, die vorläufig als Zufallsfunde (oder Exzerpte ohne weitere Verbrei-
tung) gelten können.16 Aber auch die Überlieferungsverhältnisse der – rezep-
tionsgeschichtlich gesehen – einflußreichsten Kurzversion, des sog. Laghuyo-
11 Nichtsdestotrotz deuten die Übernahme spezifischer Lesarten sowie das Vorhandensein nur
weniger außschließlich aus einer YV-Version stammender Verse auf das LYV als Hauptquelle des
VS hin (T HOMI (1999), S. 8).
12 Genau sind von den wenig mehr als 200 Versen (vgl. D IVANJI (1939), S. 697) 61 „Fremd“-
material.
13 Baroda No 6394, siehe oben, S. 72.
14 D IVANJI (1939) , S. 705 und D IVANJI (1938) , S. 31 hatte im Rahmen seiner Studien zur Textge-
stalt des LYV das Ms. (Baroda) “No Ic. 6394” als das LYV bis zum 15. Sarga des Nirvān.aprakaran. a
enthaltend beschrieben. Höchstwahrscheinlich war dieses Urteil aufgrund des Kolophons zustan-
de gekommen, der den Text nach LNEd 6.15 (= NEd 125/126) abzuschließen scheint. Es folgen
jedoch umfangreiche, dem uttarārdha des Nirvān.aprakaran. a entnommene Abschnitte. (Dieses und
die drei anderen von D IVANJI verwendeten Mss. finden sich neu beschrieben in dem Beitrag oben
auf S.55ff.)
15 Für den MU ausführlich dargestellt v.a. in S LAJE (1994).
16 Für einen Überblick aller derzeit bekannten Kurzfassungen vgl. die Abbildung auf S. 141f.
gavāsis.t.ha, haben sich als weitaus komplizierter erwiesen, als es die 1888 er-
schienene unkritische Edition der Nirn.aya Sāgara Press (LNEd ) suggeriert.17
Rezeptionsgeschichte des LYV
Das LYV ist nach bisherigem Stand die einzige der umfangreicheren Kurzfas-
sungen, der eine pan-indische Rezeption beschieden war.18 Bekanntes Beispiel
hierfür ist die Tatsache, daß es – ebenso wie die Br.had-Version – als Quelle ei-
19
niger der sog. “Minor” bzw. Sam . nyāsa-Upanis.ads identifiziert wurde.
Weniger bekannt hingegen ist das Zeugnis der persischen Übersetzungen
der Mogul-Zeit. Die indo-persischen Versionen des LYV20 sind Teil einer um-
fangreichen Übersetzungsliteratur wissenschaftlicher, epischer sowie philo-
sophischer und religiöser Texte, die zwar in Überblicksdarstellungen oft er-
wähnt wird, zu der es aber bisher kaum Detailstudien gibt.21 Dabei sind diese
Adaptionen altindischen Schrifttums nicht nur als primäre Quellen zur indo-
persischen Kultur in ihren historischen und (vor allem im Falle Akbars) ge-
sellschaftspolitischen Bezügen von Bedeutung. Wie anhand des LYV deutlich
werden wird, stellen die Übersetzungen u.U. auch eine von der indologischen
Textkritik nutzbare „Nebenüberlieferung“ dar, die das bisherige Bild um eini-
ge Aspekte zu erweitern vermag.
Die Entstehungszeit des LYV läßt sich bisher nur auf die drei Jahrhunder-
te zwischen der Abfassung des MU um 950 und vor dem ersten sicher aus
dem LYV zitierenden Zeugnis, der Sūktimuktāval¯ı (um 1258),22 festlegen. Die
Frage der Autorschaft muß im Moment noch – oder besser wieder – als un-
17 Zu LNEd s.u., S. 98, und Bibliography, S. 144 unter Laghuyogavāsis. t.ha.

18 Der folgende kursorische Überblick der Rezeptionsgeschichte des LYV beruht hauptsäch-
lich auf den Studien S PROCKHOFF (1963), S PROCKHOFF (1964), S PROCKHOFF (1970), S PROCKHOFF
(1976), S PROCKHOFF (1990) und S LAJE (1990), S LAJE (1994), S LAJE (1995–6) und S LAJE (1998) so-
wie H ANNEDER (2003), aber auch auf den älteren und in wichtigen Punkten noch immer gültigen
Beiträgen D IVANJI s [D IVANJI (1938); D IVANJI (1939)].
19 S PROCKHOFF (1976), S. 16f.
20 Derzeit bekannte Versionen und Hss. sind aufgeführt und besprochen in dem Beitrag von H.
F RANKE auf S. 113ff. in diesem Band.

21 Erwähnt seien hier die Beiträge von C ARL W. E RNST (vgl. z.B. E RNST (2003b)) sowie die
Untersuchung zur Übersetzung der Praśna-Upan.is.ad für Dārā Šukoh von G ÖBEL -G ROSS (1962)
1961). Für einen Überblick zum Forschungsstand einschließlich der relevanten Literatur, s. S. 113f.
22 R AGHAVAN (1939a), S. 128 und S LAJE (1990), S. 147. Auf den Umstand, daß ausschließlich
aus dem Bestand des LYV zitiert wird, weist H ANNEDER (2003), S. 44, hin.
geklärt gelten.23 Grund hierfür ist einerseits, daß der von Kolophonen über-
lieferte und als Autor in die Tradition eingegangene Abhinanda mit keinem
der beiden historischen Abhinandas überzeugend identifiziert werden kann,
nachdem das LYV den MU oder eine darauf zurückgehende Version voraus-
setzt: Abhinanda, Sohn des berühmten Bhat.t.a Jayanta, lebte in der 2. Hälfte
des 9. Jhd., also um die Zeit und in der Region der Abfassung des MU selbst.
Das LYV beruht jedoch auf einer Textstufe des MU, die bereits die vorletzte der
später hinzugekommenen Rahmenhandlungen enthalten haben muß. Daß ei-
ne solche Version bereits einem Zeitgenossen des Originalautors vorgelegen
haben kann, ist unwahrscheinlich. Der Poet namens Abhinanda hingegen, aus
dessen Rāmacarita Teile in das LYV aufgenommen wurden, und der schon
allein deswegen als Autor festzustehen schien, ist in seiner Datierung allzu
unsicher.24 Problematisch ist hier nicht allein die, wie in H ANNEDER (2003)
gezeigt, kaum haltbare und außerdem zu frühe Datierung, sondern auch Wi-
dersprüche innerhalb der Überlieferung sowie die Chronologie der entspre-
chenden Textzeugen. Der später folgende Blick in die Textgeschichte des LYV
ermöglicht hier zwar noch keine Lösung, aber doch eine Neuformulierung der
Problematik.
Die eigentliche Aneignung und Verbreitung des Werkes durch Vertre-
ter religiöser und philosophischer Traditionen erfolgte – analog den für den
MU/das YV gefundenen Abläufen – durch Vertreter des (späteren) Advaita-
Vedānta. Bekanntestes Beispiel ist (Mādhava-)Vidyāran.ya (1296–1386), der für
seinen J¯ıvanmuktiviveka (um 1380) aus einer auf das LYV zurückgehenden
Textversion zitiert,25 und der diese seine Hauptquelle mit dem Status einer au-
toritativen Überlieferung (smr.ti) versieht.26 In umfangreichen, kommentierten
Zitaten erzeugt er den Anschein von mit Śaṅkaras Advaitavedānta überein-
stimmenden Grundpositionen.27 Die im LYV noch vorhandenen Elemente der
Soteriologie des MU – kurz: Erlösung im (aktiven) Leben durch Welterkennt-
nis – rückt er nicht nur in die Nähe einer mit dem As.t.āṅgayoga Patañjalis kor-
respondierenden Rückzugslehre,28 er interpretiert zentrale Termini mit Bezug
23 Die sich seit beinahe 150 Jahren mit dem Autor des LYV auseinandersetzende Literatur ist
vollständig aufgeführt in S LAJE (1994), S. 125 (n. 72).

24 H ANNEDER (2003), S. 50f.
25 S LAJE (1998), S. 104. Die Argumentation geht zurück auf R AGHAVAN (1939b).
26 S LAJE (1995–6), S. 388.
27 S LAJE (1998), S. 116.
28 S LAJE (1998), S. 114 und S LAJE (1995–6), S. 390. Tatsächlich wendet der MU sich sogar aus-
drücklich gegen die Praktiken des As.t.āṅgayoga (S LAJE (1995–6), S. 392). Hinsichtlich ihrer Er-
auf die normativen Konzepte der Dharmaśāstras.29 Diese bewußte Eingliede-

rung in die orthodoxen Systeme ist exemplarisch für die Vereinnahmung des
MU – hier am LYV vollzogen – in die vedāntischen Entsagertraditionen.
Anhand der bisher untersuchten Passagen der z.T. unveröffentlichten
Kommentare läßt sich die eigentliche exegetische Tradition zum LYV dieser
Entwicklung zu-, wahrscheinlich auch zeitlich nachordnen: Die Kommentar-
literatur umfaßt: (1) Die (entgegen den Informationen verschiedener Kata-
loge) nur für die ersten drei Prakaran.as überhaupt überlieferte und edierte
Vāsis.t.hacandrikā (VC), verfaßt von Ātmasukha, einem wahrscheinlich nordin-
dischen30 sam . nyāsin. Die andereren beiden Kommentare stammen von südin-
dischen Autoren: (1) die vollständig überlieferte, jedoch nur für die letzten
drei Prakaran.as edierte Sam . sārataran.i (ST) des Mummad.ideva Allād.asūnu und
(2) die nicht edierte, ebenfalls unvollendete, d.h. im 3. Prakaran.a abbrechende
Vāsis.t.hatattvabodhin¯ı (VTB) eines Rāmabrahmendra, letzterer eventuell in Ta-
milnadu, in der Gegend des Mı̄nāks.ı̄-Tempels (Madurai) zu lokalisieren.31
Die drei Kommentare binden den Text sowohl in die traditionelle Auto-
rität von Śruti und Smr.ti als auch in vedāntische Theoreme ein.32 Besonders
in den südindischen Kommentaren verleiht die starke Betonung der Vis.n.u-
Bhakti mit den entsprechenden mythologischen Bezügen dem Text Offenba-
rungscharakter und verortet ihn in der nivr.tti-Ideologie.33
lösungsmittel seien die yogischen Traditionen der adäquaten Wirklichkeitserkenntnis (jñāna) un-
terlegen (S LAJE (1995–6), S. 397–400 und S LAJE (1998), S. 111–113).
29 S LAJE (1998), S. 109–111.
30 D IVANJI (1938), S. 706. Die Annahme geht auf Ātmasukhas Erwähnung des Viśvanātha-
Tempels zurück (VC 1.6).

31 Ausgangspunkt dieser Vermutung ist die Beschreibung der Mı̄nāksı̄ im Incipit des Kom-
.
mentars (LG2, Fol.2r, 5): sabhartr.kām . sumı̄nāks.ı̄m
. śukapadmalasatkarām. | bodhaikarūpinı̄m
. vande suk-
hasantānasiddhaye ||
32 Vgl. z.B. VC 2,5–6 (=BAUBh-VĀ I 4, 342) u. 17–18 (=BAU-VĀ I.3.13 und I.4.1335) ad
LNEd 1.1.1. Nebenbei scheint sich die Vāsis.t.hacandrikā des Ātmasukha auf deutlich komplexerem
sprachlichen und argumentativen Niveau zu bewegen. Dies gilt zumindest für die eingehender
verglichenen Einleitungen.
33 So wird der Text von Visnu hergeleitet, der, als Rāma dāśarathi zunächst das Rückzugside-
..
al verwirklicht, dann aber aus Mitleid der in der Aktivität verstrickten Welt den “Pfad zur Er-
lösung” (apavargamārga) in dem “Lehrwerk, das als Vāsis.t.ha bekannt ist”(śāstre vāsis.t.asam . jñake)
verkündet: sa ca svata eva niravadyātmavidyayā samullasitamānaso ’pi “ācāryavān purus.o vedeti” śruti-
prāmān.yāya gurukule vasan svagurumukhād abhyasanı̄yasamastaśāstrārthatattvam. niravartis.t.a | tasmād
api nyavartis.t.a ca | atha punah. pravr.ttimārgavr. ttyā mahāmohamaye mahati sam . sāre magnonmagnajagad
avalokya paramakārun. ikatayā tad ujjihı̄rs. uh. pañcadaśe vayasi svamukhena svasvarūpam . gurumukhena ca
praśnottaralı̄layā svayam antaryāmı̄ bhagavān ādigurur apavargamārgam upadideśa| (LG2, Fol.3v, 2ff.).
Im Gegensatz zu dieser eben beschriebenen (spät-)vedāntischen forma-

len und gedanklichen Inkorporation des Textes in den eigenen Kanon lassen
die persischen Übersetzungen ein völlig anders gelagertes Interesse vermu-
ten. Nach dem Zeugnis der Rājataraṅgin.¯ı des Śrı̄vara wurde bereits für den
König Zayn al-Ābidı̄n (r. 1418/20–1470) aus einer „Moks.opāyasam . hitā“, also
einer Langversion, vermutlich dem MU rezitiert.34 Im Jhd. der Mogulherr-
scher hingegen griff man auf das LYV zurück: die 1602 für Akbar angefer-
tigte Übersetzung des Farmulı̄ geht ebenso eindeutig auf eine solche Version
zurück wie die wenige Jahre früher (1597–98) entstandene Übersetzung des
Pānı̄patı̄. Vor allem letztere scheint – nach der Menge der derzeit bekannten
Hss. (ca. 15) zu schließen – weitere Verbreitung gefunden zu haben. In den
bisher untersuchten Textabschnitten der sich eng an den Text des Originals
haltenden Prosaübersetzung des Farmulı̄ finden sich keine Anhaltspunkte für
ein Vorhandensein der sonst so dominanten vedāntischen Exegese, obgleich
– wahrscheinlich mündlich über ein neuindisches Idiom vermittelte – Kom-
mentierungen aufgenommen wurden. In den Erklärungen zu poetischen Kon-
ventionen, mythologischen Hintergründen oder Details der Sachkultur steht
Rāma als epischer Held und Herrscherfigur im Mittelpunkt.35
Die von Akbar initiierte Übersetzungstätigkeit sollte offiziell auch der För-
derung kultureller Kontakte zwischen Hindus und Muslimen dienen, moti-
viert durch Akbars Bestreben, sich als uneingeschränkter, über religiösen und
kulturellen Konventionen stehender Herrscher zu profilieren.36
Ob die deutliche Nähe im gedanklichen und formalen Aufbau der Ausführungen bis hin zu (bei-
nahe) identischen Wendungen und Zitaten auf eine unmittelbare Vertrautheit eines der beiden
Kommentatoren mit dem anderen schließen läßt, wird an einer erweiterten Materialbasis zu klä-
ren sein.
34 S LAJE (1990), S. 147 (n. 1).
35 Die untersuchten Textpassagen stammen vorwiegend aus dem Vairāgyaprakaran a, welches
.
zwar den (im LYV nur noch bedingt) philosophischen Kapiteln (ab Utpattiprakaran. a) vorangeht,
jedoch genügend philosophische und religiöse Konzepte und Schlüsselterminologie bietet, die
vor einem islamischen Hintergrund zur Auseinandersetzung hätten reizen können. Ob und wie
dies später im Text oder in anderen Übersetzungen noch stattfindet, wird weiter zu untersuchen
sein.
36 Durch die Einbindung des MU in das Rāmāyana (s.o. S. 92) wurde gewissermaßen eine Rāma-
.
Figur geschaffen, die die im MU propagierte Erlösung während und mit Verbleib im (aktiven)
Leben (jı̄vanmukti) mit dem epischen Abenteurer und idealen Herrscher des Rāmāyan.a vereint.
Daher wäre einzelnen Hinweisen sowohl in der historiographischen als auch in der persischen
LYV-Übersetzungsliteratur nachzugehen, inwiefern besonders Akbar sich ein entsprechend um-
gedeutetes Konzept eines zugleich in der Welt herrschenden und über weltliche Konventionen
erhabenen Königs zunutze machen wollte. Unabhägig davon, ob sich derartige Motive anhand
Dies läßt darauf schließen, daß das LYV im nördlichen Indien des 16. Jhd.s
eine gewisse Bedeutung erlangt hatte, wurde es doch gemeinsam mit Wer-
ken wie dem Mahābhārata, dem Rāmāyan.a und der Bhagavadg¯ıtā zu den Texten
gerechnet, von denen man sich die Kenntnis zentralen Gedankengutes der
Hindus erhoffte.37
Anhand der oben erwähnte Zeugnisse ist die Rezeptionsgeschichte des
LYV zeitlich und geographisch wie folgt charakterisierbar: 1) Der früheste
Schwerpunkt der Rezeption durch Exponenten philosophisch-religiöser Tra-
ditionen38 zeichnet sich ab der zweiten Hälfte des 14. Jhd.s im Süden des Kon-
tinents ab, wo der Text bereits als autoritative Überlieferung betrachtet wird.
2) Zum Ende des 16. Jhd.s ist das LYV an den (muslimischen) Adelshöfen im
Norden verbreitet, während zu einem früheren Zeitpunkt und weiter nördlich
in Kaschmir diese Kreise die MU-Version rezipierten.
Zur handschriftlichen Überlieferungslage des LYV
Schon in den 1930er Jahren hatte D IVANJI auf die grundlegenden Mängel der
Edition von 1888 hingewiesen und auf das divergierende Zeugnis der ihm
vorliegenden Handschriften aufmerksam gemacht.39 Seine Kritik an der Edi-
tion bezog sich noch vornehmlich auf die dürftige Materialbasis von nur zwei
Hss., die unklaren Prinzipien der Textkonstitution sowie auf offenkundige
Mißgeschicke (z.B. bei der Zählung der Sargas und der falschen Identifizie-
rung des Kommentars ab dem 4. Prakaran.a). Die für jedes Prakaran.a neu ein-
setzende Zählung der Sargas hingegen wurde offensichtlich erst vom Heraus-
geber der zweiten Auflage (LNEd , 1937) eingeführt. All jene Mss. die D IVANJI
vorlagen (und mit einer einzigen Ausnahme40 auch alle hier untersuchten),
strukturieren jedoch den Text in fortlaufend gezählte Sargas,41 wodurch die
der Texte als nachvollziehbar erweisen werden, darf an einem durchschlagenden Erfolg vorerst
gezweifelt werden, da sich die Überlieferungslage für die Übersetzungen der Akbar-Zeit bislang
sehr dürftig darstellt (vgl. unten S. 124ff.).
37 Vgl. L EACH (1995).
38 Die Charakterisierung dieser Kreise als solche von religiösen Spezialisten im religionssozio-
logischen Sinne soll selbstverständlich nicht unterstellen, daß – wie im Falle des Vidyāran.yas be-
schrieben – politisch-gesellschaftliche Ambitionen keine Rolle gespielt hätten (vgl. S LAJE (1995–
6), S. 406 und S LAJE (1998), S. 115f.)
39 Vgl. hierfür und für die folgenden Ausführungen: D IVANJI (1938) und D IVANJI (1939).
40 Nämlich Ms. Wai 6926. Siehe auch oben, S. 69.
41 D IVANJI (1938), S. 31, und D IVANJI (1939), S. 700.
Zuordnung zu den Prakaran.as sekundär anmutet.42 Außerdem befand D I -

VANJI, daß der Abschluß des Nirvān. aprakaran.a von keiner einzigen der insge-
samt zehn von ihm konsultierten Hss. so überliefert ist, wie er ediert wurde.
S LAJE43 hat dann diesen Befund nicht nur bestätigt, sondern, insofern er
die Hss. nach verschiedenen Fassungen kategorisieren konnte, auch grundle-
gende Schlüsse zur Textgeschichte gezogen: Die auf 17 Hss. erweiterte Quel-
lenlage umfaßte alle bekannten und verfügbaren Hss., die entweder aus dem
Nāgarı̄- oder aus dem südindischen44 Schriftkreis stammen.
Es ließen sich am Umfang des Textendes insgesamt vier Überlieferungsty-
pen unterscheiden; die diese repräsentierenden Mss. wurden jeweils zu einer
signierten Gruppe zusammengefasst:
1. Die kürzeste Fassung, überliefert von Hss. in Grantha-Schrift, erstreckt
sich bis Sarga 15 des Nirvān.aprakaran.a.45
2. Die nächst längere Fassung, bezeugt von vier Hss. ausschließlich in
Nāgarı̄-Schrift und kommentiert, enthält noch vier Verse einer Guru-
praśasti. In der Edition umfaßt diese Gurupraśasti die Eingangsverse des
nächsten, 16. Sarga. Inhaltlich ist die Vasis.t.ha in den Mund gelegte Lob-
preisung eines Hariharārya dort fehl am Platze.46
3. Den darauf folgenden Sarga Nirvān.a 16.5–34 überliefern acht Hss. (drei
Nāgarı̄, drei Grantha, zwei Telugu). Sarga Nirvān.a 16 kommt nur im LYV
vor und bezeugt die traditionelle Ansicht, daß es sich bei dem LYV um
einen Auszug (sāra) handelt.47 Daher wird dieser im folgenden als Sāra-
Sarga bezeichnet.48
4. Der längste Abschluß umfaßt den Sāra-Sarga Nirvān.a 16 plus zwei zu-
sätzliche Sargas Nirvān.a 17–18. Diese stellen eine formale Ergänzung
42 Da die erste Auflage der Edition kaum erhältlich ist und sich Indices, Konkordanzen etc. an
der zweiten Auflage orientieren, wird auch hier aus praktischen Gründen nach dieser zitiert.
43 Siehe v.a. S LAJE (1994), S. 125–134.
44 Bislang sind Hss. in Tamil-Grantha- und Telugu-Schrift bekannt.
45 Sarga 15 in Nirvāna entspricht also Sarga 43, die folgenden 44–46.
.
46 D IVANJI (1939), S. 107: “Even if ‘Hariharārya’ was another name of Valmı̄ki, . . . still Gauda
.
Abhinanda, who in that case could be the author of those stanzas, could not have placed them
in the mouth of Vasis.t.ha because he cannot be believed to have forgotten that it was not Vasis.t.ha
who had been reciting a work of Vālmı̄ki but it was Vālmı̄ki who had been re-producing . . . a
discourse which was alleged to have occured between Vasis.t.ha and Rāmacandra.”
47 LN
Ed 6.16.31cd und 6.16.24.
48 S LAJE (1994), S. 129.
zum Eröffnungsrahmen (Vālmı̄ki-Bharadvāja-Mythos, s.o.) des Werkes

dar, inhaltlich überwiegt orthodoxe Rāma-Bhakti. Zwei Nāgarı̄-Mss.
enthalten diese Version.
Die wichtigsten Schlüsse, die aus diesem Befund gezogen wurden, seien
kurz zusammengefaßt: Der am Ende unvollendete Text Nirvān.a 15 (überlie-
fert in beiden Schriftkreisen sowie durch den [südindischen] Kommentar, der
in allen Hss. an dieser Stelle endet) wurde in Stufen überarbeitet, deren Funk-
tion klar erkennbar ist, nämlich die, den Text formal abzuschließen. Die Gu-
rupraśasti sowie die längste Version finden sich ausschließlich in der Nāgarı̄-
Überlieferung. Die gewissermaßen dazwischen liegende Version, die nur den
Sāra-Sarga als Abschluß enthält, ist im Vergleich dazu mit mehr Hss. in beiden
Schriftkreisen stärker belegt.
Somit hat – analog zu den Verhältnissen der YV-Edition (NEd ) – eine nach
nicht offengelegten Kriterien hergestellte unkritische Edition gewissermaßen
auf Basis des LYV eine „rezente“ Kurzversion geschaffen, die als Grundlage
kritischer Forschung und insbesondere auch der Klärung historischer Verhält-
nisse denkbar ungeeignet ist.49
Zum Schluß sollen diese Ergebnisse noch durch einige zusätzliche Beob-
achtungen ergänzt werden, die sich bislang sowohl durch das Zeugnis der
persischen Nebenüberlieferung als auch aufgrund der erneuten Aufnahme
des Studiums von nunmehr insgesamt 25 für die Fragestellung relevanten
Hss. ergeben haben. Die folgenden Ausführungen sind als vorläufige Arbeits-
hypothesen zu verstehen, die illustrieren sollen, wie eine erweitere Material-
basis das Bild noch zu differenzieren vermag.
Zu 1 Die kürzeste Fassung wird nunmehr außer von zwei Grantha-Mss. auch
von zwei Nāgarı̄-Mss. (LN1 , ĀĀ S12(1)-4-39) repräsentiert. Werke, die
aus einem nicht über Nirvān.a 15 hinausgehenden Textumfang zitieren,
sind der J¯ıvanmuktiviveka (um 1380), der VS sowie die Śr¯ıvāsis.t.hacandrikā.
Zu den externen Zeugen gehört auch die persische Übersetzung des
Farmulı̄ (1602) sowie eine mit mehreren Mss. überlieferte Version des
49 An dieser Stelle sei auch angemerkt, daß im Zuge der Arbeiten an den persischen Überset-
zungen einzelnen Lesarten aus den Hss. nachgegangen wurde. Ein Vergleich mit der Edition hat
den Verdacht ergeben, daß am Text stillschweigend geändert und Lesarten in Kommentar und
mūla-Text vertauscht wurden. Kurz: der gedruckte Text repräsentiert weder im Umfang noch im
Wortlaut eine einzige der bisher bekannten und zugänglichen Hss.
Pānı̄pātı̄ (1590–1598), die mit dem „43 Kapitel“, mithin bei Nirvān.a 15
enden.50
Zu 2 Wie bereits herausgestellt, ist die Gurupraśasti (Nirvān.a 16.1–4 in LNEd )
nur einem Teil der Nāgarı̄-Überlieferung eigentümlich. In den entspre-
chenden Hss. aber wird sie entweder eindeutig im Kommentar (ST)
überliefert, oder, wo Kommentar und mūla-Text optisch nicht abgeho-
ben sind, zumindest separat gezählt, nicht etwa als Vers 1—4 wie in
LNEd . Die südindischen Hss. die den gleichen Kommentar überliefern
– der schließlich von einem Südinder verfaßt wurde – kennen diese Ver-
se nicht. Es wird sich also, wie schon S LAJE 1994 vermutete, um einen
Nāgarı̄-spezifischen Schreiberspruch handeln, der nicht in den mūla-
Text gehört. In dieser Hinsicht überliefern immerhin sieben kommen-
tierten Nāgarı̄-Hss. den gleichen Textumfang wie die erste Gruppe.
Zu 3 Der Sāra-Sarga als erster formaler Abschluß ist weiterhin mit vier
Nāgari-Hss. und immerhin fünf südindischen Hss. belegt. Von zwei
südindischen Hss., die eingehender untersucht werden konnten, wies
eine genau am Übergang von Nirvān.a 15 zu Nirvān.a 16 einen Hand- und
Ms.-Wechsel auf, die andere wechselte schon zuvor offensichtlich Vor-
lage mitsamt Paginierung und änderte die Kolophone. In Zukunft wä-
re damit die Suche nach Anhaltspunkten für eine mögliche allmähliche
Durchsetzung dieser Version durch Kontamination verschiedener Hss.
bei erneuten Abschriften zu bedenken.
Zu 5 Die längste Version mit dem Sāra-Sarga Nirvān.a 16 und 17–18 ist zwar
lediglich von drei Nāgarı̄-Hss. (außer von LN/Web 643 und LN/Bar
12810 auch von Wai 6922) überliefert, jedoch von einer Version der per-
sischen Übersetzung des Pānı̄pātı̄ (1597–98) bestätigt. Durch letztere so-
wie die Datierung einer Hss. auf 1600 zeichnet sich als terminus ante
quem für das Vorhandensein der drei abschließenden Sargas das Ende
des 16. Jhd.s ab.
50 Siehe unten, S. 120 und S. 124.

Tabelle: Überlieferung des Nirvān.aprakaran.a nach den (hinsichtlich des

Textabschlusses relevanten) Hss. (25)51
Textbestandteil Schriftkreis: Nāgarı̄ Schriftkreis: Südindisch
(Grantha, Telugu)
bis Nirvān.a 15 LN1 LG1
ĀĀ S12(1)-4-39 LG/Bar 9809
Gurupraśasti LN3
LN/P(oona)786
LN10
ĀĀ S12(1)-5-46
Wai 6923
Wai 6924
Wai 6926
Sāra-Sarga Nirvān.a 16 LN5
LN6
LG2
LG3
[LG/Mad 40H18]
[LT2 ]
LT3
LN8
Sāra-Sarga Nirvān.a 16 LN/Bar 10561
und Gurupraśasti
Sāra-Sarga Nirvān.a 16 LN/Web 643
und 17-18 LN/Bar 12810
Wai 6922
Zum Abschluß dieses Beitrags soll das Problem der Autorschaft des LYV
nochmals aufgenommen und zu den Handschriftenbefunden in Beziehung
gesetzt werden. Überliefert ist von Kolophonen namentlich ein mit dem
51 Hss. in eckigen Klammern werden derzeit noch beschafft, sind aber in der Literatur zuverläs-
sig beschrieben worden. Mit Ausnahme der (neu hinzugekommenen) Hss. LN10 und LN8 finden
sich alle Mss. aufgeführt bei S LAJE (1994) sowie in dem Beitrag von S TEPHAN /S TINNER, S.55ff.
Gaud.aman.d.ala in Verbindung stehender Abhinanda o d e r ein (anonymer)

Kāśmı̄rapan.d.ita. Dabei zeigen sich folgende Zusammenhänge: Wenn nach
dem kürzesten Abschluß ein Schlußkolophon vorhanden ist, berichtet dies
von dem besagten Pan.d.it aus Kaschmir. Ein Gaud.a-Abhinanda ist erst ab dem
Rahmen Nirvān.a 16 von den Kolophonen überliefert.52 Damit korrespondiert,
daß allein der sekundäre Schlußrahmen (ab Nirvān.a 17.11) Teile des von Abhi-
nanda verfaßten Rāmacarita enthält. Wie zu Beginn bereits bemerkt, läßt sich
dieser Autor hingegen weder mit hinreichender Sicherheit in die Entstehungs-
zeit des LYV datieren, noch überhaupt mit Kaschmir in Verbindung bringen.53
Nach dem geteilten Zeugnis der Kolophone ist diese Verbindung allerdings
auch gar nicht notwendig: Aus seiner direkten oder indirekten Beteiligung
am Schlußrahmen muß nicht zwingend hervorgehen, daß Abhinanda der Au-
tor des gesamten LYV war. Der „anonyme kaschmirische Pan.d.it“ hingegen
birgt ein anderes Problem: wenn er seinen Auszug tatsächlich in Kaschmir
verfaßt hätte, müßte man erklären, warum angesichts der ansonsten so rei-
chen pan-indischen Handschriftenlage für das LYV einerseits und der für den
MU vorhandenen Śāradā-Überlieferung andererseits, sich keine handschrift-
liche Überlieferung in der Region der Herkunft seines angeblichen Autors er-
halten hat? Vielleicht, weil das LYV weder in Kaschmir verfaßt wurde, noch
jemals im äußersten Norden seiner sonstigen Verbreitung vergleichbar rezi-
piert wurde.54 Neben der Möglichkeit, daß der Urheber ein aus Kaschmir
stammender Emigrant war, könnte der „Gelehrte aus Kaschmir“ auch eine
vage Reminiszenz an den ursprünglichen Autor des MU sein, weitergegeben
in einer Tradition, die den Gesamtkontext nicht vor Augen hatte. So, wie z.B.
Vidyāran.ya, als er aus dem LYV zitierte, nicht zu wissen oder zu beachten
schien, daß ihm eine Kurzfassung vorlag – auch der südindische Kommen-
tator Rāmabrahmendra erwähnt diesen Umstand in seiner Einleitung nicht –
mag man zeitweise neben dem Verfasser des MU keinen zweiten Urheber ver-
mutet haben – zumindest solange nicht, bis der sekundäre Rahmen dies nicht
nahelegte. Das Zeugnis Br.had-Version hatte man jedenfalls im Süden nicht:
Bis heute sind keine südindischen YV- oder MU-Hss. bekannt geworden. Wo
52 S. auch S LAJE (1994), S. 131.
53 H ANNEDER (2003).
54 Schließlich wird ein auf die regionale Herkunft eines Autors verweisendes Epitheton wohl
nur in der Fremde verliehen, wo es den Benannten von Einheimischen unterscheidbar macht.
Möglicherweise trifft dies auch auf den Kāśmı̄raka Sadānanda Yati zu, der in seiner Advaita-
brahmasiddhi aus einer Laghu-Version zu zitieren scheint (LNEd 4.5.34ab [YV 4.57.55ab] und LNEd
5.10.9ab [5.89.33ab].)
dieser Kontext hingegen vorhanden war, wurde er auch einbezogen (wie an

den Bemerkungen des Kommentators Ātmasukha ersichtlich ist).
Schlußbemerkungen
Die hier vertretene Hypothese zur Textgeschichte des sog. Laghuyogavāsis.t.ha
besagt, daß sich in der Handschriftenlage anhand der entsprechenden Schrift-
kreise zwei Überlieferungsschwerpunkte abzeichnen, nämlich der Norden
und der Süden des Kontinents. Sowohl die exegetischen Traditionen als auch
Zeugen externer Rezeption bieten Anhaltspunkte für eine Chronologie der
einzelnen Textstufen, welche zusätzlich mit der internen Struktur des Textes
begründbar sind. Demnach liegt die Vermutung nahe, daß das LYV – zumin-
dest ab dem uns überlieferten Zeitpunkt – in einer im Vergleich zum MU ent-
gegengesetzten Richtung weiter- bzw. zurückgegeben wurde, mithin aus dem
Süden in den Norden kam.
Selbstverständlich würde erst eine gründliche Untersuchung der einzel-
nen Varianten und ihrer Abhängigkeiten klären helfen, ob sich diese Hypo-
these im Mikrokontext erhärten läßt und inwieweit sich daraus Kriterien zur
Konstitution eines kritischen Textes entwickeln ließen.
The Moks.opāyasaṅgraha
J ÜRGEN H ANNEDER
During the course of editing the Utpattiprakaran.a the present author inves-
tigated one paper manuscript written in Śāradā which is kept in the Staats- und
Universitätsbibliothek, Göttingen, as Cod. Ms. Sanscr. Vish. 126. The beginning
of the text is missing, the manuscript starts on folio 3r with:
uttamo moks.a ucyate
brahma†sa eva vimalakramo jñānaprakāśakah. [= MU 1.2.8]
The text ends on folio 563r with verse 6.374.17 [= NEd 7.215.17], i.e. the last
verse of the MU. Thereupon follow the colophon and the scribe’s concluding
verse on folio 563v:
. samāptam
iti śr¯ımahārāmāyan.e moks.opāyasaṅgrahe nirvān.aprakaran.am
samāptam . cedam . moks.opāyasaṅgraham
śāke gate śivanandavidhau vikramabhūpateh.
itim . śivāyate
. n¯ıtā gan.eśena moks.asāram
The essence of [the way to] Liberation (=Moks.opāya), completed (?)
by Gan.eśa in the year 1911 of the Vikrama era [i.e. 1854/55], becomes
auspicious.
The interpretation of this verse assumes that moks.asāra is brief for

moks.opāyasāra, although the name of the text according to the colophon is
moks.opāyasaṅgraha. √We also have to read n¯ıtam. and assume from the context
that the phrase itim . n¯ı means “to complete”.
When I first received copies of this manuscript during editing the Utpatti-
prakaran.a, it looked at first sight very promising, because it seemed to transmit
the text of the MU, merely lacking some of the doubtful and even problematic
verses. For instance, the awkward set of verses that start this Prakaran.a (3.1.1–
4) is missing and the Prakaran.a is, as would be expected, introduced by the
pratisandhiśloka. Further comparison showed that the manuscript contained in
the beginning of the Utpattiprakaran.a an almost complete text and one was left
wondering about its self-designation as saṅgraha.
Upon reading more it became clear why the name was in fact fully justi-
fied, but in an unexpected way: After reproducing a fairly complete beginning
106 Jürgen Hanneder: Moks.opāyasaṅgraha
of the third Prakaran.a, we suddenly find that MU 3.13.54 is followed by 3.64.1

(fol. 120). The large part of the text omitted is the story of Lı̄lā and it took
only few more checks to arrive at the diagnosis that this “Moks.opāyasaṅgraha”
(ŚSam. ) was a version that was condensed in a unique way, namely through
removing the ākhyānas!
One could surmise that someone interested in the “philosophy” of the
work wanted to produce a version that contained just the philosophical parts
without the ākhyānas. This approach would not only be unusual, but, espe-
cially in the light of the MU’s views about the use of “yukti”1 , a grave misun-
derstanding of the intention of the text. But as we shall see below, the method
of abridgement used in the Moks.opāyasaṅgraha is even more unexpected.
The Mumuks.uvyavahāraprakaran.a
The second Prakaran.a commences as in the MU: The first verse refers back to
the Vairāgyaprakaran.a; in vss. 2 and 3 Rāma is addressed by Viśvāmitra, who
says that his insight into the futility of the world, which was expressed in the
long poetical lamentations in the first Prakaran.a, is fundamentally correct, but
that he would still need to purify his mind. Verse 4 introduces the story of
Śuka and Janaka, which depicts a person in a similar condition. The story,
related in the MU from 2.1.4 up to 2.2.1, is carefully removed in the Saṅgraha.
The last omitted verse marks the return to the main topic:
tasmāt prakr.tam evedam. śr.n.u śravan.abhūs.an.am|

mayopadiśyamānam . tvam. jñānam ajñāndhyanāśanam (= MU 2.4.7)
Therefore listen [now] as I expound the main topic itself, an ornament

for the ears, [namely] the knowledge that destroys the darkness of igno-
rance.
The next verse which continues the instruction is consequently contained

in the Saṅgraha. Did the redactor of the summary think that the prakr.ta
could be separated from the aprakr.ta, whereas in fact both are necessary for
a dr.s.t.ānta?
For investigating the character of the ŚSam. and its method of abridgement
a full list of verses in the Mumuks.uvyavahāraprakaran.a follows. The verse num-
1 See B RUNO L O T URCO ’s article in this volume.
Jürgen Hanneder: Moks.opāyasaṅgraha 107
bers refer to the MU as edited,2 no attempt was made to provide the ŚSam.
with a verse numbering of its own.
2.1.1 2.9.26 2.11.29 2.13.10 2.14.28 2.18.19
2.1.2 2.9.27 2.11.36 2.13.11 2.14.41 2.18.23
2.1.3 2.9.28 2.11.39 2.13.15 2.14.46 2.18.25
2.2.13–28 2.9.29 2.11.40 2.13.16 2.14.53 2.18.26
2.3.1 2.9.43 2.11.42 2.13.19 2.14.54 2.18.28
2.4.8 2.10.1 2.11.43 2.13.20 2.15.1 2.18.29
2.4.11–18 2.10.2 2.11.44 2.13.21 2.15.6 2.18.30
2.5.4 2.10.6 2.11.47 2.13.22 2.15.8 2.18.31
2.5.9 2.10.7 2.11.48 2.13.24 2.15.9 2.18.35
2.5.11 2.10.8 2.11.50 2.13.28 2.15.10 2.18.42
2.5.12 2.10.9 2.11.51 2.13.31 2.15.16 2.18.43
2.5.14 2.10.10 2.11.53 2.13.32 2.15.17 2.18.44
2.5.15 2.10.11 2.11.54 2.13.34 (NEd 2.15.19) 2.18.45
2.5.18 2.10.12 2.11.55 2.13.35 2.15.19 2.18.46
2.5.19 2.10.13 2.11.57 2.13.36 2.16.1 2.18.47
2.5.20 2.10.14 2.11.58 2.13.37 2.16.3 2.18.50
2.5.25 2.10.16 2.11.59 2.13.38 2.16.5 2.18.51
2.6.29 2.10.17 2.11.60 2.13.40 2.16.7 2.18.52
2.6.31 2.10.18 2.11.67 2.13.41 2.16.8 2.18.54
2.6.36 2.10.19 2.11.68 2.13.43 2.16.10 2.18.55
2.6.38 2.10.20 2.11.69 2.13.45 2.16.12 2.18.56
2.7.2 2.10.23 2.11.72 2.13.46 2.16.15 2.18.58
2.7.4 2.10.24 2.12.1 2.13.48 2.16.16 2.18.61
2.7.22 2.10.27cd 2.12.2 2.13.50 2.16.17 2.19.2
2.7.32 2.10.28 2.12.7 2.13.55 2.16.19 2.19.9
2.8.1 2.10.29 2.12.8 2.13.56 2.16.20 2.19.10
2.8.5 2.10.30 2.12.10 2.13.57 2.16.21 2.19.11
2.8.17 2.10.32 2.12.11 2.13.58 2.16.27 2.19.13
2.9.1 2.10.33 2.12.12 2.13.59 2.16.31 2.19.14
2.9.6 2.10.34 2.12.13 2.13.61 2.16.32ad 2.19.16
2.9.8 2.10.35 2.12.14 2.13.70 2.16.33 2.19.17
2.9.10 2.10.36 2.12.16 2.13.72 2.16.34 2.19.19
2.9.11 2.10.37 2.12.17 2.13.74 2.16.35 2.19.20
2.9.12 2.10.38 2.12.18 2.13.80 2.17.1 2.19.23
2.9.13 2.10.39 2.12.19 2.13.82 2.17.3 2.19.24
2.9.14 2.10.40 2.12.20 2.14.1 2.17.4 2.19.35
2.9.16 2.10.41 2.12.21 2.14.2 2.17.6ab 2.20.10
2.9.17 2.10.42 2.13.1 2.14.4 2.17.8ab 2.20.11
2.9.21 2.11.1 2.13.2 2.14.7 2.17.9 2.20.12
2.9.18 2.11.2 2.13.3 2.14.10 2.18.1 2.20.13
2.9.32 2.11.22 2.13.6 2.14.14 2.18.5
2.9.24 2.11.23 2.13.7 2.14.18 2.18.12
2.9.25 2.11.27 2.13.8 2.14.22 2.18.15
2.9.25 2.11.28 2.13.9 2.14.23 2.18.18
The passages counted as 2.12.12–18 and 2.13.9 are in prose. With 2.15.19 a
verse from NEd has slipped into the text, although this observation is, in the
absence of a critical edition of the mūla text of the second Prakaran.a, prelim-
2 See S LAJE (1993).
inary. There is a Sarga colophon after 2.12.21: tattvajñamāhātmyapratipādanam .

nāma sargah.; other colophons were ignored and the concluding verses in non-
anus.t.ubh-metres that are found at the end of each Sarga are more often than
not omitted.
A second hand can be discerned in the beginning of the second Prakaran.a,
which is responsible for adding the abbreviated Prakaran.a marks (“mu pra”) in
the margin next to the folio number and the “mo sā”3 of the first hand. The sec-
ond hand has also changed the first words of the second Prakaran.a “nāradeneti”
to “iti nādena”, which is the reading of NEd . Similarly, in 2.10.30 dvijah. is cor-
rected to the NEd -reading kila. Apart from these few secondary influences
of the Nāgarı̄ recension, not untypical for Kashmirian manuscripts,4 the text
version is that of the MU, and with very few scribal errors.
From the above list we see that the compiler has a tendency to adopt com-
plete verses. The case of 2.17.6cd is an exception, because this half-verse5 indi-
cates the total number of verses of the MU, which does not apply to the ŚSam. .
Similarly 2.17.10–51, which is a description of the contents of the MU and
refers to the ākhyānas, is omitted, as also 2.18.62, which explains the function
of the stories as dr.s.t.āntas. The above list suggests that the ŚSam. is indepen-
dent of the LYV, which has extracted only verse 2.5.4 of Sarga 5 and nothing
of Sarga 6. The LYV has, with 180 verses, adopted a smaller number than the
ŚSam. with 250.
The story of Lı̄lā
We have seen that the author of the ŚSam. has systematically omitted passages
that were unnecessary or even contradictory to his aim. For instance, the ta-
ble of contents contained in 2.17 was carefully removed by joining 2.17.9 with
2.18.1. The numbers given in these verses and especially the characterization
of the text as being equipped with dr.s.t.āntas would of course be inappropriate
for the Saṅgraha’s presentation. A more radical example is that of the omis-
sion of one of the most voluminous stories in the MU, the L¯ılopākhyāna, alias
Man.d.apākhyāna. This story is introduced in the MU with verse 3.15.17 and
ends with 3.60.1:
3 According to this hand the text seems to be rather a Moksopāyasāra as in the scribe’s verse
.
quoted above, but there it may be due to the constraints of metre.
4 See also above, p. 53, for this type of contamination.
5 moksopāyābhidhāneyam samhitā sārasammitā | trimśad dve ca sahasrāni jñātā nirvānadāyinı̄
. . . . . . ||
atredam. man.d.apākhyānam. śr.n.u śravan.abhūs.an.am

nih.sandeho yathais.o ’rthaś citte viśrāntim es.yati (3.15.17)
...
etat te kathitam. rāma dr.śyados.anivr.ttaye
l¯ılopākhyānam anagha ghanatām . jagatas tyaja (3.60.1)
The ŚSam. reads the Utpattiprakaran.a only up to 3.13.54 and then jumps
ahead to 3.64.1, thereby omitting also the explanation of the story, as well as
some prose passages.
The Nirvān.aprakaran.a
A reading of the last Prakaran.a brought another surprise. Upon the concluding
colophon of the Upaśamaprakaran.a follows the pratisandhiśloka introducing the
new Prakaran.a (6.1.1),6 then the following verses:
6.2.19–32 6.5.8 6.11.37 6.11.84 6.11.123
6.2.35 6.5.11 6.11.39 6.11.85 6.11.124–129
6.2.37 6.5.12 6.11.40 6.11.86 6.12.1
6.2.40 6.5.13 6.11.44cd 6.11.87 6.12.2
6.2.41 6.5.14 6.11.45ab 6.11.90 6.12.13
6.2.42 6.5.15 6.11.46cd 6.11.65ab 6.12.14
6.2.44 6.6.1 6.11.47ab 6.11.66cd 6.12.15
6.2.46 6.11.1 6.11.48cd 6.11.67ab 6.12.16
6.2.47 6.11.2ab 6.11.49 6.11.67cd 6.12.17
6.2.48 6.11.2cd (=NEd ) 6.11.50cd 6.11.69ab 6.12.21
6.2.49 6.11.3cd (=NEd ) 6.11.51ab 6.11.69cd 6.12.22
6.2.52 6.11.2cd (MU) 6.11.51cd 6.11.94 6.12.24
6.2.53ab (=NEd ) 6.11.3–6ab 6.11.52ab 6.11.95 6.12.25
6.2.53cd 6.11.7cd 6.11.55cd 6.11.96 6.13.1
6.2.54 6.11.8ab 6.11.57 6.11.97 6.13.2
6.2.55 6.11.6cd 6.11.58 6.11.98 6.13.3
6.2.56 6.11.7ab 6.11.59ab 6.11.99 6.13.4
6.2.57 6.11.8cd 6.11.62cd 6.11.100 6.13.7
6.2.58ab 6.11.8ab 6.11.63 6.11.101 6.13.8
6.3.1 6.11.9 6.11.64ab 6.11.102 6.13.9
6.3.4 6.11.10ab 6.11.74 NEd 6.11.90ab 6.13.10abc
6.2.59 (2 pādas untra- 6.11.75 6.11.104–114 6.13.11d
6.4.15 ced) 6.11.76 6.11.116 6.13.12
6.5.1 6.11.13ab 6.11.78 6.11.117 6.14.1
6.5.2 6.11.15cd–20 6.11.79 6.11.118 6.14.2
6.5.3 6.11.26–27 6.11.80 6.11.119 6.14.3
6.5.5 6.11.31 6.11.81 6.11.120
6.5.6 6.11.32 6.11.82 6.11.121ab
6.5.7 6.11.36 6.11.83 6.11.122cd
6 The numbers refer to Ś ; for a rough concordance with N

1 Ed , see the appendix to my forth-
coming Studies on the Moks.opāya.
With this we enter the story of Bhusun.d.a, which is given in the ŚSam. in an
only slightly shortened form. Also the explanation of this story in Sarga 6.29
is given fairly completely:
6.29.2 6.29.25cd 6.29.43cd 6.29.57 6.29.72cd
6.29.7 6.29.26 6.29.44 6.29.58ab 6.29.73ab
6.29.8–13 6.29.27ab 6.29.45 6.29.60cd 6.29.75ab
6.29.19 6.29.28cd 6.29.48 6.29.61 6.29.76ab
6.29.20ab 6.29.29 6.29.49ab 6.29.62ab 6.29.81cd
6.29.24cd 6.29.30ab 6.29.50cd 6.29.65cd 6.29.82ab
6.29.25ab 6.29.36cd 6.29.51 6.29.66 6.29.84
6.29.22cd 6.29.37 6.29.52ab 6.29.67 6.29.88–95
6.29.23ab 6.29.38ab 6.29.56cd 6.29.68ab
Sarga 6.30 and 31 at the end of the story of Bhusun.d.a are also summarized,
then follows a condensed version of the Śivākhyāna (MU 6.31–46). The subse-
quent ākhyānas are partly excised, as for instance the Arjunākhyāna, others as
the story of the mithyāpurus.a (6.116–117) and Bhr.ṅg¯ıśa appear in an abridged
version. Before we try to understand the rationale behind this type of abridge-
ment, we shall deal with a passage in the last Prakaran.a, which is crucial for
the later textual history of the MU literature.
The bipartite Nirvān.aprakaran.a
The most significant test for establishing the relationship between the ŚSam. ,
the MU and the LYV, is a comparison of that passage in the Nirvān.aprakaran.a
which has been lost in NEd at its juncture between the pūrva- and uttarārdha.
As S LAJE has shown,7 NEd lacks MU 6.122–157 (more than 500 verses) and
reads instead merely 70 verses taken from the LYV:
MU NEd
6.120–121 6.116.1–12 (= LYV 6.13.1–12)
6.122–157 6.117–128 (= LYV 6.13.13–6.18.83)
If we can show that the ŚSam. selects verses not contained in the LYV or the
YV, its direct dependence on the MU is proven. On folio 456r the ŚSam. reads
MU 6.138.14, that is, the concluding verse of chapter 14, and its colophon.
Then follow a number of verses from Sarga 153, one from Sarga 155, then the
ŚSam. jumps to 159. This passage is given below with a concordance to the YV,
which is in this passage more or less identical with the LYV:
7 See S LAJE (1994), further details in H ANNEDER (∗ 2006).
ŚSam
. YV
6.153.1 (NEd 6.126.58)
6.153.2 (NEd 6.126.59)
6.153.3 (NEd 6.126.60)
6.153.8ab (NEd 6.126.61ab)
6.153.10–14
6.153.15
6.153.18–20
6.153.22–26
6.153.28
6.153.30–31
6.153.45
6.154.1 (NEd 6.126.61cd)
6.154.2 (NEd 6.126.62cd–63ab)
6.154.7 (NEd 6.126.63cd–64ab)
6.154.20 (NEd 6.126.64cd–65ab)
6.155.1 (NEd 6.126.65cd–66ab)
6.155.2 (NEd 6.126.66cd–67ab)
6.155.3ab/4cd (NEd 6.126.67cd–68ab)
6.155.25 (NEd 6.126.68cd–69a)
6.155.32
6.155.34ab (NEd 6.126.69cd)
(NEd 6.126.70ab untraced)
6.156.2cd/3ab (NEd 6.126.70cd)
6.156.3ab (NEd 6.126.71ab)
6.156.4 (NEd 6.126.71cd–72ab)
6.156.6 (NEd 6.126.72cd–73ab)
6.156.14 (NEd 6.126.73cd–74ab)
6.157.1–6 (NEd 6.126.74cd–80ab)
(NEd 6.126.80cd–81 untraced)
6.157.14cd (NEd 6.126.82ab)
6.157.15–17 (NEd 6.126.82cd–85ab)
6.157.19 (NEd 6.126.85cd–86ab)
6.157.22 (NEd 6.126.86cd–87ab)
6.157.27ab (NEd 6.126.87cd)
6.157.23 (NEd 6.126.88)
6.157.24ab (NEd 6.126.89ab)
6.157.27cd (NEd 6.126.89cd)
(NEd 6.126.90–91ab untraced)
6.158.3cd (NEd 6.126.91cd)
6.158.4 (NEd 6.126.92)
Here follow several verses from 6.158 in NEd , whereas the ŚSam. contin-
ues with 6.159.6. We see from the list that there is not even a single overlap
between the ŚSam. and the YV/LYV, which proves beyond any doubt that the
ŚSam. and the LYV are independent. Further proof of this is that while the LYV
breaks off after this passage, the ŚSam. continues its summary until the end of
the Nirvān.aprakaran.a. The ŚSam. is therefore a direct extract from the MU.
In the second half of the Nirvān.aprakaran.a most ākhyānas have again been
omitted, as for instance the Vidyādhara story; later the text leaps from
Sarga 205 to 255, thereby omitting the voluminous Pās.ān.ākhyāna, then from
263.32 to 330.1, cutting out the Vipaścit- and Śavākhyāna. It is not necessary for
the sake of this preliminary analysis of the ŚSam. to complete the list. In any
case the impression that larger ākhyānas are removed and only very few small
ākhyānas are retained is certainly confirmed.
The method of abridgement
We have seen that the compiler of the ŚSam. has in some places excised
ākhyānas, references to the size of the text and has even removed doubtful pas-
sages in a way that suggests that the abbreviation was not executed haphaz-
ardly, but methodically. Since not all ākhyānas were removed this plan cannot
have meant a lopsided assemblage of the philosophical discourses. The ques-
tion is rather: what could have distinguished the stories that appear in the
Nirvān.aprakaran.a, especially the Bhusun.d.a- and the Śivākhyāna, to merit their
inclusion. And why would the author retain two succeeding ākhyānas, thereby
shifting the balance in this part of his text from philosophy to narrative.
If we rule out accident the most likely reason for this is the internal struc-
ture of the MU. In brief,8 the turning point in the text, as far as the devel-
opment of Rāma is concerned, is of course his awakening to the truth. The
instructions given after this passage, which lies in the middle of the whole
work are apparently on a different didactic level than those that lie before this
incident. Vasis.t.ha once explicitly refuses to answer a question and asks Rāma
to ask again during the time of the siddhānta.
Now Rāma’s enlightenment takes place between the Bhusun.d.a- and the
Śivākhyāna and the author’s singling out of these stories among all emphasizes
this crucial passage in the whole work. Of course there can be no definite proof
that the compiler of the ŚSam. had this larger structure in mind, but we should
add that the cross-referential passages are contained in the ŚSam. , as is another
important passage where Vasis.t.ha gives the ultimate answer to a question of
Rāma by remaining silent.
If this impression of a careful and thoughtful redaction, which seems, quite
unlike many other abbreviated versions, guided by the original spirit of the
work is not shaken by contradicting findings, we have in the ŚSam. – as in
Bhāskarakan.t.ha’s commentary on the MU – instances of an understanding of
the MU that is far removed from its wide-spread Vedāntic reinterpretation.
8 Compare above, p.18; for details, see H ANNEDER (2003).
Die persischen Übersetzungen des Laghuyogavāsis..tha1
H EIKE F RANKE
Betrachtet man die frühe Geschichte der europäischen Beschäftigung mit

der Geisteswelt des Alten Indien, so fällt auf, dass den persischen Überset-
zungen von Sanskrit-Texten zeitweilig eine nicht unwichtige Rolle bei der
gedanklichen Erschließung des fremden Kulturraumes zukam. Die von dem
französischen Gelehrten A NQUETIL D UPERRON im Jahre 1801 veröffentlich-
te lateinische Übersetzung der Upanis.aden, die A RTHUR S CHOPENHAUER so
tief beeindruckten,2 fußte nicht auf einer Sanskrit-Version des Textes, sondern
auf einer persischen Übersetzung, die im 17. Jahrhundert im Auftrag des Mo-
gulprinzen Dārā Šikōh (st. 1659), eines Sohnes Šāh Ǧahāns (reg. 1628–1657),
angefertigt worden war.
Zur Zeit, als D UPERRONs Upanis.adenübersetzung erschien, war die Be-
deutung des Sanskrit aber bereits erkannt, und mit der Etablierung des
Sanskrit-Unterrichtes an den Universitäten wurde der unmittelbare Zugriff
auf die indischen Quellen selbstverständlich. Da man den persischen Über-
setzungen aus dem Sanskrit auch innerhalb der Geschichte der persischen Li-
teratur keine größere Bedeutung zumaß, gerieten sie aus dem Blick der For-
schung und fielen nahezu vollkommen dem Vergessen anheim.
Erst in den vergangenen Jahrzehnten haben sich wieder einige wenige
Fachleute mit den persischen Übersetzungen der Sanskrit-Literatur befasst.
Die nachfolgend aufgelisteten Namen sind in unserem Kontext besonders er-
wähnenswert:
1. E RHARD G ÖBEL -G ROSS hat in seiner Dissertation aus dem Jahre 1962
die Übersetzungsmethode der persischen Upanis.aden-Übertragung des
Mogulprinzen Dārā Šikōh untersucht.
2. FATHULLĀH M UJTAB Ā ’ Ī bietet, ebenfalls in seiner Dissertation, Edition

und englische Übersetzung des Muntahab („Auswahl“, „Kurzfassung“)

zur persischen Übersetzung des Laghuyogavāsis . t.ha von Niz.ām ud-Dı̄n
Pānı̄patı̄.
1 Der vorliegende Artikel ist ein Teilergebnis des interdisziplinären DFG-Projektes „Indo-
Persische Übersetzungsliteratur“.
2 G ÉRARD (1963): 215ff.
114 Heike Franke: Die persischen Übersetzungen des LYV
3. C ARL W. E RNST, Professor of Religious Studies an der University of

North Carolina, hat sich des Phänomens der Sanskrit-Persischen und
Sanskrit-Arabischen Übersetzungsliteratur aus einer allgemeineren Per-
spektive angenommen und anhand von Fallbeispielen eine geistesge-
schichtliche Einordnung unternommen.3 Vor allem die arabische und
persische Übersetzung des Amr.takunda wurde ausführlich von ihm be-
arbeitet.4
4. Nicht unerwähnt bleiben dürfen die ausschließlich in Iran und Indi-

en publizierenden Gelehrten, die im Laufe der Jahrzehnte eine ganze
Anzahl mogulzeitlicher Übersetzungen von Sanskrit-Texten fachkundig
ediert haben. Allen voran ist M UH. AMMAD R I Ż Ā Ǧ AL ĀL Ī N Ā ’ ĪN Ī zu
nennen, der sich seit den 50er Jahren der Übersetzungsliteratur widmet
und bis in die unmittelbare Gegenwart hinein Studien hierzu veröffent-
licht.5 Weitere wichtige Fachleute auf diesem Feld sind T ĀR Ā C HAND
. ASAN Ā BID Ī.
6
und S AYYID A M ĪR H
Die Übersetzung von Sanskrit-Schrifttum ins Persische ist seit frühester

Zeit ein fester Bestandteil der indisch-persischen Kulturbegegnung. Bekannt
ist die im 6. Jahrhundert angefertigte mittelpersische Übersetzung des Pañca-
tantra,7 der in den nachkommenden Jahrhunderten eine Reihe weiterer Über-
setzungen aus dem Sanskrit folgten. Ein deutlicher Höhepunkt der Überset-
zungsaktivitäten lässt sich unter dem Mogulherrscher Akbar ausmachen; aber
auch Akbars Nachfolger Ǧahāngı̄r und einzelne spätere Angehörige des Mo-
gulhofs, wie der schon genannte Dārā Šikōh, waren an der Übersetzung von
Sanskrit-Werken interessiert.
Ein DFG-Projekt zur „Indo-persischen Übersetzungsliteratur“ bietet der-
zeit die Möglichkeit, eines der aus dem Sanskrit ins Persische übertragenen
Werke intensiv zu bearbeiten. Die Erkenntnisziele, mit denen wir uns heute
den persischen Sanskrit-Übersetzungen nähern, sind freilich nicht mehr die-
selben wie zur Zeit A NQUETIL D UPERRONs: Es kann nicht mehr darum gehen,
3 E RNST (2003b), S. 173–195.
4 Siehe E RNST (2003a); vgl. auch: E RNST (1996) und E RNST (∗ 2005).
5 Vgl. Ǧ AL ĀL Ī N Ā ’ ĪN Ī s Texteditionen des Mahābhārata, der Bhagavadgı̄tā, des Sirr-i Akbar und
des Āyyin-i tarı̄qat wa hudāyābı̄ im Literaturverzeichnis, S. 145.

6 Vgl. ihre Texteditionen Daryā-yi asmār und Pañcākhyāna im Literaturverzeichnis.
7 Vgl. H ELMUT H OFFMANN : „Pañcatantra.“ In: Kindlers Neues Literatur Lexikon. München 1998,
Bd. 19, S. 233f.

Heike Franke: Die persischen Übersetzungen des LYV 115
Aufschlüsse über die indische Religiosität zu erhalten; Ziel ist es vielmehr, die
Art der muslimischen Annäherung an die religiös-philosophische Geisteswelt
der Hindus zu untersuchen. Im Vordergrund steht deshalb die Frage, auf wel-
che Weise Konzepte aus dem hinduistischen in den islamischen Kontext trans-
feriert wurden, und – damit zusammenhängend – inwieweit die Übersetzer
die Intention des Ausgangstextes verstanden und auch entsprechend übertra-
gen haben. Wichtig ist dabei also nicht nur, ob sie den Sanskrittext für den
persophonen Leser verständlich übersetzt haben, sondern auch, in welchem
Ausmaß sie möglicherweise Uminterpretationen ihrer Vorlage vorgenommen
haben, die Rückschlüsse auf ihre eigene Zielsetzung oder auf die ihres Auf-
traggebers zulassen. Im Zuge dessen gilt unsere Aufmerksamkeit besonders
der Übersetzungstechnik, also der konkreten praktischen Vorgehensweise bei
der Übertragung vom Sanskrit ins Persische.
Als Textgrundlage für das oben erwähnte DFG-Projekt wurde die persi-
sche Übersetzung des Laghuyogavāsis.t.ha (LYV) ausgewählt. Das Interesse der
Mogulherrscher an diesem Werk war offenbar besonders groß, denn inner-
halb eines recht kurzen Zeitraumes ließen sie mehrfach Übersetzungen anfer-
tigen. Gerade dieser Umstand macht die Bearbeitung des Materials besonders
reizvoll, ist doch für jede der Übersetzungen eine unterschiedliche Herange-
hensweise an die sprachliche Transformation festzustellen. Dies wird Thema
zukünftiger Veröffentlichungen sein. Aufgabe des vorliegenden Artikels ist
es zunächst, Ordnung in die äußerst verworrene Handschriftensituation zu
bringen und die genaue Anzahl und Entstehungszeit der persischen LYV-
Übersetzungen zu bestimmen.
Wenn wir alle sekundären Weiterbearbeitungen des Stoffes sowie die
Kurzfassung Vāsis.t.hasāra8 aus der Betrachtung ausschließen, kommen wir bei
einem Vergleich der einschlägigen Handschriftenkataloge und Publikationen
auf eine unterschiedliche Zahl von Übertragungen ins Persische. Die diver-
sen Einträge in den Handschriftenkatalogen der größeren europäischen, per-
sischen und indischen Bibliotheken und Museen – Dublin, Oxford, London,
Paris, Berlin, Teheran, Qum, Mašhad, Patna, Kalkutta – ließen zunächst auf
insgesamt fünf Übersetzungen schließen:
1. Eine für Akbar, die C HARLES R IEU in seinem Katalog des British Muse-
um unter Add. 5637 nennt,9
8 Siehe unten, Fn. 20 u. 21, und S. 39.
9 R IEU (1879–1881), S. 61.
2. eine illustrierte Handschrift in der Chester Beatty Library (Dublin), nach

Ausweis des Kataloges von „Farmulı̄, Schüler des Kabı̄r“ für Prinz Salı̄m
übersetzt,10
3. eine Übersetzung von Niz.ām ud-Dı̄n Pānı̄patı̄ für Prinz Salı̄m,11
4. eine, die im Auftrag von Dārā Šikōh erstellt wurde,12
5. und schließlich schien eine unbestimmte Anzahl von anonymen Über-

tragungen ins Persische zu existieren.13
D.N. M ARSHALL zählt in seiner 1967 erschienenen Bibliographie der

Mogul-Handschriften vier Fassungen: Die von R IEU genannte Übersetzung
für Akbar, eine weitere für Dārā Šikōh, eine dritte Übersetzung für Prinz
Salı̄m, von Niz.ām ud-Dı̄n Pānı̄patı̄ angefertigt, der er auch die meisten An-
onyma zuordnet, und schließlich eine vierte, von unbekanntem Übersetzer,
die nach seinen Worten „auf der gekürzten Version des kaschmirischen Pan-
dit Anandan basiert“ (M ARSHALL Nr. 251). Die illustrierte Handschrift in der
Chester Beatty Library hat M ARSHALL nicht berücksichtigt.
Der Umstand, dass FATHULLĀH M UJTAB Ā ’ Ī in seiner Dissertation von
1976 wieder auf fünf Übersetzungen kommt, schien zunächst darauf hinzu-
weisen, dass er die verschiedenen Handschriften genauer untersucht und
nach feineren Unterscheidungsmerkmalen eingeordnet hat. Er nahm jeweils
eine Lang- und eine Kurzfassung für Akbar und für Prinz Salı̄m an. Die für
Dārā Šikōh erstellte Übersetzung war nach seiner Berechnung bereits die fünf-
te.14
Im Jahre 1981 erwähnt Ǧ AL ĀL Ī N A’ ĪN Ī in seiner Edition der Übersetzung
von Niz.ām ud-Dı̄n Pānı̄patı̄ (s.u.) nur noch eine zweite Übersetzung, nämlich
10 L EACH (1995), S. 155–189.
11 (a) London, BM Or. 8443, vgl. M EREDITH -O WENS (1968), S. 37. (b) Qum, Bibliothek des
Āyatullāh Az.amı̄ Naǧafı̄ Marašı̄, Nr. 6999, vgl. H
. USAYN Ī (1989), S. 170f. In dieser Hs. auf S. 361v-
398v eine Abschrift der Kurzfassung Muntahab-i Ǧōg bašist. (c) Qum, Bibliothek des Āyatullāh
Az.amı̄ Naǧafı̄ Marašı̄, Nr. 5252. Vgl. H
. USAYN Ī (1987), S. 47f. (d) Kalkutta, Asiatic Society of
Bengal, Nr. 1699. Vgl. I VANOW (1924).
12 Vgl. z.B. E THÉ (1903), Mss. Nr. 1185, 1859, 1355, 3165. P ERTSCH (1888), Nr. 1077/4.
13 Als Anonyma waren zunächst folgende Handschriften einzordnen: (a) Oxford, Bodleian Li-
brary, Ms. Walker 117, vgl. S ACHAU (1889), Nr. 1328. (b) London, India Office, Ms. Nr. 806. Vgl.
E THÉ (1903), Nr. 1971. (c) Paris, Bibiothèque Nationale. Vgl. B LOCHET (1905), Bd. 1, S. 183f., Ein-
trag Nr. 223. (d) London, British Museum Add. 5644. Vgl. R IEU (1879–1881), S. 61f.
14 M UJTABĀ ’ Ī (1976), S. xiii-xv.
diejenige für Dārā Šikōh. Obwohl Ǧ AL ĀL Ī N A’ ĪN Ī zu dieser Zeit schon seit
über 20 Jahren mit der Übersetzungsliteratur befasst war, hatte er offenbar
kein Ǧōg bašist für Akbar entdecken können.
S HRIRAM S HARMA, der 1982 eine kleine Übersicht über die persischen
Sanskritübersetzungen veröffentlicht hat, konnte insgesamt drei verschiede-
ne Versionen ausmachen: erstens eine anonyme, die M ARSHALL bereits Prinz
Salı̄m zugeordnet hatte (BM, Add. 5644); zweitens eine Kurzfassung des Tex-
tes für Akbar, bei der es sich wieder um die oben genannte Handschrift aus
dem British Museum (Add. 5637) handelt, und drittens eine Übersetzung für
Dārā Šikōh.15
Zweifelsfrei und sicher lässt sich zunächst die Übersetzung für Dārā Šikōh
identifizieren. Sie existiert in mindestens zwölf Handschriften und seit 1968
liegt eine Edition des Textes durch T ĀR Ā C HAND und S.A.H . . Ā BID Ī vor. In
der Einleitung wird das Datum der Übersetzung, das Jahr 1066H, entprechend
dem Jahr 1655/56 u. Z., sowie der Name des Auftraggebers, Muh.ammad Dārā
Šikōh ibn Šāhǧahān Bādšāh, genannt. Im weiteren Verlauf dieser Vorrede be-
richtet Dārā Šikōh selbst, dass er eines Nachts bei einer „in der Realität (dar
wāqi)“ stattfindenden Begegnung – nicht im Traum! – mit Vasis.t.ha und Rāma
zusammengetroffen sei und er gemeinsam mit Rāma von Vasis.t.ha gereichte
Süßigkeiten gegessen habe. Sowohl Rāma als auch Dārā sind, das soll der Le-
ser vermutlich aus dieser rituellen Speisung schließen, gleichermaßen Schüler
Vasis.t.has.
Wer die Übersetzung für Dārā Šikōh angefertigt hat, geht aus dem Text
nicht hervor. E RNST16 sowie C HAND und Ā BID Ī in ihrer Edition der Dārā
Šikōh-Übersetzung favorisieren die Zuschreibung an Banwālı̄ Dās, der auch
als Walı̄ Rām (st. 1674/75) bekannt ist.17 Banwālı̄ Dās war Dārā Šikōhs Sekre-
tär (munš¯ı), hat sich als Dichter und Historiker hervorgetan und Kr.s.n.adāsas
Prabodhacandrodaya18 ins Persische übersetzt.
Im Katalog des India Office wird zum Manuskript Nr. 3165, einer Ab-
schrift der Übersetzung für Dārā Šikōh, angegeben, es handele sich hierbei
um die einzige Handschrift, in der der Übersetzer identifiziert werde.19 Dies
sei ein gewisser H . abı̄bullāh, dessen Name auf Fol. 1b, Zeile 6 erwähnt werde.
15 S HARMA (1982), S. 9–13.
16 E RNST (2003b), S. 184.
17 S TOREY (1927-1939), S. 450–452.
18 Vgl. Gulzar-i h.āl im Literaturverzeichnis.
19 Vgl. E THÉ (1903), Bd. 2, Nr. 2927.
M ARSHALL führt die Übersetzung für Dārā Šikōh deshalb unter dem Namen
„H. abı̄bullāh“ auf.
Die Angabe von E THÉ /E DWARDS bezüglich des Übersetzers ist bei einge-
hender Betrachtung des Textes jedoch so nicht haltbar: Der Ausdruck, der in
dem Katalog als Eigenname interpretiert wurde, steht in der betreffenden Pas-
sage im Kontext einer tas.liya, einer Propheteneulogie, wie sie bei Vorworten
zu islamischen Texten an dieser Stelle nach der h.amdala, dem Gotteslob, üb-
lich ist. Während die tas.liya gewöhnlich lautet „s.allā Llāhu alayhi wa-sallama“ –
„Möge Gott das Gebet über ihn sprechen und ihm Heil spenden“, und manch-
mal durch Nennung weiterer Personengruppen, etwa der Prophetengefähr-
ten, der Familie des Propheten oder der Gottesfreunde erweitert ist, findet
sich in unserer Handschrift die ungewöhnliche Form „s.allā Llāhu alayhi wa
alā man h.ab¯ıbu Llāh“ – „Möge Gott das Gebet über ihn sprechen und über den,
der der Geliebte Gottes (h.ab¯ıb Allāh) ist.“ Wer genau mit „Geliebter Gottes“
gemeint ist, lässt sich schwer sagen. Es ist anzunehmen, dass es sich dabei um
einen allgemeinen Segenswunsch handelt, der jeden meint, der ein h.ab¯ıb Allāh
ist. Nicht ganz von der Hand zu weisen wäre auch, dass der Autor speziell
eine Ehrung Dārā Šikōhs damit im Sinn hatte. Mit einiger Sicherheit ist aber
auszuschließen, dass sich der Übersetzer an dieser besonders hervorgehobe-
nen Stelle selbst eingeführt hat, wäre seine Selbstvorstellung doch vielmehr
im Rahmen angemessener Bescheidenheitsformeln zu erwarten und überdies
mit Worten, die ihn eindeutig als Übersetzer identifizieren. Auf diese Weise
präsentieren sich zumindest die Übersetzer in anderen Texten.20 Die Frage,
wen Dārā Šikōh mit der Übertragung des LYV betraut hat, muss deshalb wei-
terhin unbeantwortet bleiben.
Im Vorwort zum Ǧōg Bašist für Dārā Šikōh erfährt der Leser, dass der Prinz
mit den bisher erstellten Übersetzungen nicht zufrieden war. Man darf folg-
lich davon ausgehen, dass der Text vor 1656 zumindest mehr als einmal ins
Persische übertragen worden ist. Dārā erwähnt, dass er die Tuh.fah-i maǧlis von
Šayh S.ūfı̄ Šarı̄f Hūbǧahānı̄ gelesen habe21 – eine nur rund ein Dutzend Folios

20 Vgl. z.B. Chester Beatty Library, Ms. 5, Fol. 3a: „der geringste der Muriden von Kabir, der mit
vielen Fehlern [behaftet] und als Farmulı̄ bekannt ist [...] Dieser Geringste (= Farmulı̄) sollte sich
an die Übersetzung machen [...] Wie befohlen hat er (Farmulı̄) mit der Übersetzung jenes [Werkes]
begonnen.“ Vgl. auch die Handschrift im British Museum, Or. 8443, Fol. 2a: „Gemäß dem hohen
Befehl machte sich der geringste der Knechte des Palastes, Niz.ām ud-Dı̄n Pānı̄patı̄, ans Werk.“
21 Ǧōg Bašist für Dārā Šikōh, ed. C HAND / Ā BID Ī , S. 4: „die kurzgefasste Übersetzung dieses
Buches, die Šayh S.ūfı̄ zugeschrieben wird, habe ich studiert.“

umfassende persische Übersetzung des Vāsis.t.hasāra,22 das für unsere Unter-

suchung keine Rolle spielt. Von großem Interesse ist jedoch die Übersetzung
des LYV für Akbar, die unter dem Eintrag „Add. 5637“ im Katalog der persi-
schen Handschriften im British Museum aufgeführt ist. Das Manuskript trägt
das Datum 1006H, und macht daher den Anschein, aus der Regierungszeit
Akbars zu stammen. Dessen Name, „Muh.ammad Akbar Bādšāh“, wird auch
in der Tat in der Einleitung zum Text genannt, die im weiteren Verlauf überra-
schenderweise ebenfalls von einer Begegnung mit Rāma und Vasis.t.ha berich-
tet. Es war naheliegend, diese Handschrift mit der LYV-Übersetzung für Dārā
Šikōh zu vergleichen, und dabei stellte sich heraus, dass der Text Add. 5637
exakt mit der Übersetzung für den Mogulprinzen übereinstimmt. Lediglich
das in der Vorrede genannte Datum wird in Add. 5637 mit 1006H anstatt mit
1066H/1655-56AD angegeben, und der Name des Auftraggebers, Dārā Šikōh,
ist durch „Akbar Bādšāh“ ersetzt. Bei Add. 5637 handelt es sich also nicht um
eine Übersetzung für Akbar, sondern die bereits bekannte Fassung für Dārā
Šikōh.
Wie oben erwähnt, lässt sich ebenfalls sicher eine Übersetzung des Gelehr-
ten Niz.ām ud-Dı̄n Pānı̄patı̄ identifizieren. Sein Name sowie der seines Auf-
traggebers, Prinz Salı̄m, des späteren Ǧahāngı̄r, ist aus dem Vorwort zu seiner
Übersetzung bekannt, das allerdings nicht in allen Manuskripten überliefert
ist.23
Da sich diejenigen Handschriften, in denen diese Vorrede fehlt, in ihrem
Incipit von denjenigen mit Vorwort unterscheiden, schien es sich auf den er-
sten Blick um zwei verschiedene Versionen zu handeln.24 Diesbezüglich war
aber die bereits von M ARSHALL vorgenommene Zuordnung zu Niz.ām ud-
Dı̄n Pānı̄patı̄ richtig.
Dennoch sah es bei der Durchsicht der Handschriften zunächst so aus, als
gäbe es eine modifizierte Pānı̄patı̄-Fassung: Das Manuskript BM Add. 5644
glich zwar größtenteils den anderen Handschriften derselben Übersetzung,
22 Z.B.: P ERTSCH (1888), Nr. 1077, S. 4.
23 Das Vorwort von Niz.ām ud-Dı̄n Pānı̄patı̄ ist in folgenden Hss. enthalten: (a) London, BM
Or. 8443 (M EREDITH -O WENS (1968)). (b) Qum, Bibliothek des Āyatullāh Az.amı̄ Naǧafı̄ Marašı̄,
Nr. 6999. (H . USAYN Ī (1989)). (c) Qum, Bibliothek des Āyatullāh Az.amı̄ Naǧafı̄ Marašı̄, Nr. 5252
(H. USAYN Ī (1987)). (d) Kalkutta, Asiatic Society of Bengal, Nr. 1699 (I VANOW (1924)).
24 In den folgenden Hss. fehlt das Vorwort Niżām ud-Dı̄n Pānı̄patı̄s: (a) Oxford, Bodleian Libra-
ry, Ms. Walker 117 (S ACHAU (1889)) (b) London, India Office, Ms. Nr. 806 (E THÉ (1903)) (c) Paris,
Bibiothèque Nationale, B LOCHET (1905), Eintrag Nr. 223 (d) London, British Museum Add. 5644
(R IEU (1879–1881)).
die letzten sargas (skt. „Unterkapitel“) des Textes wiesen aber erhebliche Ab-
weichungen auf. Bei einer genaueren Prüfung stellte sich heraus, dass es sich
hierbei offenbar um eigenmächtige Kürzungen und Zusammenfassungen von
Seiten des Kopisten handelt. Nach dem Ende der Geschichte von Kača auf
S. 943 der Handschrift, im Text gleichzeitig als der Abschluss des 38. sarga be-
zeichnet, ist der Text durch Streichungen und Zusammenfassungen um mehr
als zwei Drittel gekürzt (in der Edition entspricht dies den Seiten 410 bis 484).
Ab S. 1007 folgen auf die kurzgefasste Geschichte vom Muni und dem Jäger
die sargas 44 bis 46 wieder in voller Länge. Man darf in diesem Text also keine
zweite Version von Pānı̄patı̄ sehen, sondern eine auf den Schreiber zurück-
gehende Eigenheit nur dieser einen Abschrift. Zu bemerken ist jedoch, dass
einige Kopien der Pānı̄patı̄-Übersetzung mit dem 43. sarga enden,25 während
andere, wie etwa die soeben genannte BM Add. 5644, auch die sargas 44 bis 46
einschließen.26
Das Datum der Fertigstellung der Pānı̄patı̄-Übersetzung geht weder aus
dieser selbst noch aus ihrer Einleitung hervor. S.A. H . USAYN Ī nennt im Kata-
log der Bibliothek von Qum das Jahr 1006H, gibt die Quelle für diese Informa-
tion jedoch nicht an.27 Wir haben aber insofern eine gewisse Bestätigung die-
ser Datierung, als Ǧahāngı̄r selbst in einer Notiz am Rande des Manuskriptes
in der Chester Beatty Library anmerkt, die Übersetzung des Werkes in seinem
„ [. . . -]zwanzigsten Lebensjahr“ – die Einer sind nicht lesbar – befohlen zu ha-
ben. Wenn wir der Einfachheit halber grob in Sonnenjahren rechnen, so ergibt
sich ein Zeitrahmen von 1590–1598, in dem das Werk entstanden sein muss,
denn der Prinz ist im Jahre 1569 geboren und hat 1599 sein dreißigstes Lebens-
jahr vollendet. Die Datierung auf das Jahr 1006H/ca. 1598 liegt also durchaus
im Rahmen des Möglichen und Wahrscheinlichen.
Rund 15 Kopien von Pānı̄patı̄s persischer Übersetzung des LYV werden
in europäischen, iranischen und indischen Bibliotheken und Museen sowie in
Privatsammlungen verwahrt. Im Jahre 1981 hat Ǧ AL ĀL Ī N Ā ‘ ĪN Ī eine kritische
Edition des Textes in Teheran veröffentlicht.28 Im Vorwort weist er darauf hin,
25 Die sargas 44–46 sind beispielsweise nicht enthalten in den Hss. Oxford, Bodleian Library,
Ms. Walker 117 und BM Or. 8443.

26 Siehe oben, S. 100.
27 Vgl. H USAYN Ī (1987) zu Ms. Nr. 5252 in der Bibliothek des Āyatullāh Azamı̄ Naǧafı̄ Marašı̄
. .
in Qum.
28 Ǧōg Bāsišt, ed. Ǧ AL ĀL Ī N Ā ‘ ĪN Ī . Diese Edition der Pānı̄patı̄-Übersetzung ist außerordentlich
schwer erhältlich und steht dem Projekt nur durch die freundliche Hilfe von Herrn Dr. David
Durand-Guédy zur Verfügung, dem die Verfasserin zu großem Dank verpflichtet ist.
dass Abschriften dieser Übersetzung nicht nur in den großen Bibliotheken

Irans zu finden sind, sondern auch in zahlreichen öffentlichen und privaten
Büchersammlungen in den Provinzen des Landes.29 Man muss folglich von
einer noch größeren Verbreitung des Textes ausgehen, als der durch Kataloge
ersichtliche Bestand vermuten lässt.
Besondere Bedeutung erhielt die Übersetzung Niz.ām ud-Dı̄n Pānı̄patı̄s
durch den Kommentar des bekannten persischen Philosophen Mı̄r Abū ‘l-
Qāsim Findiriskı̄ (geb. 970H/1562). Dieser gehörte zusammen mit Mı̄r Dāmād
und Šayh Bahā’ ud-Dı̄n Āmilı̄ zu den großen Meistern der Schule von Isfahan
und wurde in religiösen Kreisen wie am Hof gleichermaßen hoch verehrt. Ei-
ner seiner Schüler, Mullā Raǧab Alı̄ Tabrı̄zı̄ (st. 1080H/1669-70), gilt als ei-
ner der hervorragendsten Lehrer für Philosophie in der zweiten Hälfte des
17. Jahrhunderts. Mı̄r Findiriskı̄, selbst ein Anhänger Ibn Sı̄nas (Avicennas),
hatte seinen Blick auch für die indische Philosophie geöffnet. Zu seinen we-
nigen Schriften wird ein nicht erhaltenes Werk über indische Weisheit gerech-
net.30 Er unternahm Reisen nach Indien und kam durch die Empfehlung Ās.af
Hāns (st. 1641), des Schwiegervaters von Šāh Ǧahān, im Jahre 1037H/1627
ein weiteres Mal im Jahre 1046H/1636 in den Genuss einer Audienz bei
und
dem Herrscher des Mogulreiches.31 Mı̄r Findiriskı̄ starb im Jahre 1050H/1640-
41 in Isfahan.
Ein weiterer Schüler Mı̄r Findiriskı̄s, der Dichter und Mystiker Sarmad
Kāšānı̄ (st. 1070H/1659), begab sich ebenfalls nach Indien. Nach langen Reisen
traf er mit Dārā Šikōh zusammen und wurde zu dessen spirituellem Mentor.
Daher liegt die Annahme nahe, dass der Prinz nicht nur durch die Lektüre des
Tuh.fah-i maǧlis von Šayh S.ūfı̄ auf das LYV aufmerksam wurde, sondern auch
durch Sarmad, der mit dem Kommentar Mı̄r Findiriskı̄s zum Ǧōg Bāsišt sicher-
lich vertraut war. – Sarmad Kāšānı̄ wurde, wie Dārā Šikōh unter dem Vor-
wand, ein Häretiker zu sein, auf Betreiben des Mogulherrschers Aurangzēb
hingerichtet.32
Der Kommentar Mı̄r Findiriskı̄s, der vermutlich in den 1620er oder 1630er
Jahren entstand, ist mehreren Abschriften der Pānı̄patı̄-Übersetzung, gewöhn-
29Ǧ ALĀL Ī N Ā ’ ĪN Ī in seiner Edition des Ǧōg Bāsišt, Seite lām.
30N ASR (1986), S. 675-677.
31 S AF Ā (1994), S. 310–314.
.
32 K.A. N IZAMI : „Muhamad Saı̄d Sarmad“, in Encyclopedia of Islam 2. (Encyclopaedia of Islam.
.
New edition. Bisher 11 Bde. Leiden 1954ff.)
lich in Form von Randglossen, beigefügt.33 Dieser Kommentar wurde bislang

noch nicht ediert, ebensowenig das von ihm zusammengestellte Glossar, das
sich im Anhang einiger Abschriften der Pānı̄patı̄-Fassung findet. Es führt in
alphabetischer Reihenfolge die in der Übersetzung verwendeten Sanskritbe-
griffe und die dazugehörigen, meist aus dem Text selbst extrahierten persi-
schen Erklärungen dazu auf.34
Von der häufigen Rezeption der Pānı̄patı̄-Übersetzung zeugt schließlich
eine Kurzfassung des Textes unter der Bezeichnung Muntahab („Auswahl“),
oder Hulāsa („Resümee“). M UJTAB Ā ’ Ī konnte nur zwei Manuskripte dieser

Kurzfassung lokalisieren,35 auf die er auch seine Edition stützt. Laut Ǧ AL ĀL Ī
N Ā ’ ĪN Ī, der den Text nach eigenen Angaben bereits im Jahre 1970 ediert und
in Teheran veröffentlicht hat, finden sich „von dieser Übersetzung Hand-
schriften in den öffentlichen und privaten Bibliotheken in Teheran und in In-
dien.“36
Während Ǧ AL ĀL Ī N Ā ’ ĪN Ī den Muntahab für die Übersetzung einer ge-
kürzten indischen Version des Yogavasis.t.ha hält, kommt M UJTAB Ā ’ Ī zu dem
Ergebnis, dass der Text dieser Kurzfassung einen Auszug aus der Pānı̄patı̄-
Übersetzung darstellt, der stellenweise durch eingefügte Verse persischer my-
stischer Dichtung kommentierend erweitert wurde. M UJTABĀ ’ Ī liefert auch ei-
ne grobe Aufschlüsselung, welche Teile des Pānı̄patı̄-Textes für den Muntahab
verwendet wurden.37
Autorschaft und Entstehungszeit des Muntahab-i Ǧōg Bāsišt sind bislang
noch umstritten. Aufgrund der darin enthaltenen Verse des Dichters Fānı̄
Is.fahānı̄ (st. 1222H/1807) geht Ǧ AL ĀL Ī N Ā ’ ĪN Ī davon aus, dass nur eine Da-
33 Z.B.: (a) Fihrist-i kutub-i kitābh āna-yi Āstān-i Quds-i Radawı̄. Mašhad 1305-1346Š/1927-1968.
.

Bd. IV, S. 339f.; (b) Fihrist-i kitāb-hāna-yi Maǧlis-i Šurā-yi Millı̄. Teheran 1318-1321Š/1940-1943. Nr.
651, mit Kommentar von Mı̄r Findiriskı̄ am Rand. Diese Hs. ist möglicherweise identisch mit der
von Ǧ ALĀL Ī N Ā ’ ĪN Ī genannten Hs. Nr. 7243 derselben Bibliothek, die 1090H in Mašhad kopiert
wurde und einen Umfang von 185 Folios hat. (c) Teheran, Nationalbibliothek, A NWAR (1976),
Bd. VI, S. 207f., Nr. 2646/F. (d) Drei Handschriften im Privatbesitz (nach Angaben bei M UJTABĀ ’ Ī
(1976), S. xxvi und Ǧ AL ĀL Ī N Ā ’ ĪN Ī in seiner Edition des Ǧōg Bāsišt, S. t.ā’-lām). Ǧ ALĀL Ī N Ā ’ ĪN Ī
erwähnt im Vorwort zum Ǧōg Bāsišt, S. kāf-lām, zwei Kopien der Pānı̄patı̄-Übersetzung mit dem
Kommentar von Mı̄r Findiriskı̄ in seinem privaten Besitz. Die Nationalbibliothek in Teheran be-
sitzt ebenfalls eine Pānı̄patı̄-Übersetzung mit Kommentar, vgl. A NWAR (1976), Bd. VI, S. 207f.
34 M UJTABĀ ’ Ī (1976), S. xxviii.
35 M UJTABĀ ’ Ī (1976), S. lxi.
36 Ǧ AL ĀL Ī N Ā ’ ĪN Ī in Ǧōg Bāsišt, S. tā’.
.
37 M UJTABĀ ’ Ī (1976), S. xl-xli
tierung während oder nach der Lebenszeit des genannten Dichters in Frage
kommt.38
M UJTABĀ ’ I hingegen schreibt das Werk Mı̄r Findiriskı̄ zu, weil beide von
ihm verwendeten Handschriften (kopiert 1816 und 1845) diesen im Titel als
Autor nennen39 Die Verse Fānı̄ Is.fahānı̄s sind seiner Ansicht nach erst spä-
ter von einem Kopisten hinzugefügt worden, und zwar zu irgendeinem Zeit-
punkt vor 1816, dem Datum des älteren seiner beiden Manuskripte. Er be-
gründet seine Annahme einer nachträglichen Interpolation mit dem Hinweis
auf den merkwürdigen Umstand, dass bis auf Fānı̄ Is.fahānı̄ sämtliche Dichter,
deren Verse in den Text eingefügt wurden – Farı̄d ud-Dı̄n At.t.ār (st. ca. 1230),
Ǧalāl ud-Dı̄n Rūmı̄ (st. 1273), Mah.mūd Šabistarı̄ (st. 1320), H
. āfı̄z. (st. 1391) und
Nimat-allāh Walı̄ (st. 1431) – aus den zwei Jahrhunderten zwischen 1230 und
1430 stammen. Rund 350 Jahre liegen zwischen dem jüngsten Vertreter dieser
Gruppe von Mystikern und Fānı̄ Is.fahānı̄. M UJTAB Ā ’ Ī wertet dies als sicheres
Indiz, dass der in seinem Hauptteil von Mı̄r Findiriskı̄ geschriebene Text nach
langer Zeit wieder hervorgeholt und durch die Zufügung von Versen Fānı̄
Is.fahānı̄s von unbekannter Hand aktualisiert wurde.40
Problematisch an M UJTAB Ā ’ Īs These ist, dass kein Text bekannt ist, der
in seiner vorliegenden Gestalt des Werk Mı̄r Findiriskı̄s sein kann. Die bei-
den Manuskripte, die Mujtabā’ı̄ seiner Edition zugrunde gelegt hat, kommen
durch den Einschluss der späten Fānı̄ Is.fahānı̄-Verse jedenfalls nicht als solche
in Betracht. Sie können lediglich spätere Bearbeitungen eines Muntahab-i Ǧōg
Bāsišt von Mı̄r Findiriskı̄ sein.
Unabhängig von der Frage, ob die Kurzfassung des Ǧōg Bāsišt ganz oder
vielleicht nur zum Teil im 19. Jahrhundert entstanden ist, zeigt die Existenz
dieser Fassung nicht nur die langanhaltende Popularität des Werkes, sondern
auch, dass man es, wie die eingeschobenen Verse islamischer Mystiker zeigen,
für übereinstimmend mit der eigenen religiösen Tradition hielt.
Um nun auf unsere Ausgangsfrage nach der Anzahl der Übersetzungen
des LYV zurückzukommen, so ist es wichtig, festzuhalten, dass der Muntahab-
i Ǧōg Bāsišt in keinem Fall als eine eigenständige Übersetzung betrachtet wer-
den kann, wie von Ǧ AL ĀL Ī N Ā ’ ĪN Ī und C.W. E RNST angenommen wurde,41
38 Ǧ ALĀL Ī N Ā ’ ĪN Ī in Ǧōg Bāsišt, S. t.ā’.

39 M UJTABĀ ’ Ī (1976), S. lxi.
40 M UJTABĀ ’ Ī (1976), S. xxxix; lix.
41 Ǧ ALĀL Ī N Ā ’ ĪN Ī in Ǧōg Bāsišt, S. .tā’. E RNST (2003b), Fn. 46.
sondern eine Auswahl verschiedener Textpassagen der Pānı̄patı̄-Übersetzung

darstellt.42
Bisher haben wir also eine Übersetzung für Dārā Šikōh aus dem Jahre 1656,
und wir wissen ebenfalls sicher, dass es eine weitere Übersetzung für Prinz
Salı̄m gibt, die etwa 1598 von Niz.ām ud-Dı̄n Pānı̄patı̄ angefertigt, vor 1640 von
Mı̄r Findiriskı̄ kommentiert und zu einem noch unbekanntem Zeitpunkt von
anonymer Hand zu einem Muntahab zusammengestellt wurde. Ist also für Ak-

bar kein Ǧōg Bāsišt übersetzt worden? Nachdem der Katalogeintrag von R IEU
(Add. 5637) der für Dārā Šikōh erstellten Übersetzung zugewiesen und alle
anonymen Handschriften auf die Version von Niz.ām ud-Dı̄n Pānı̄patı̄ verteilt
waren, blieb keine Fassung übrig, die im Auftrag Akbars hätte erstellt worden
sein können. Da auch Ǧ AL ĀL Ī N Ā ‘ ĪN Ī, einer der besten Kenner auf dem Ge-
biet der Sanskrit-Persischen Übersetzungsliteratur, während seiner jahrzehn-
telangen Tätigkeit offenbar auf keine für Akbar erstellte Übersetzung gesto-
ßen ist, lag der Schluss nahe, dass schlichtweg keine existiert.
Per Zufall stellte sich gleich die erste Handschrift, die im Rahmen des er-
wähnten DFG-Projektes näher untersucht wurde, als das Ǧōg Bāšištha43 für
Akbar heraus: Es handelt sich um ein illustriertes Manuskript aus der Chester
Beatty Library in Dublin, das im Katalog der genannten Bibliothek allerdings
nicht etwa als eine Übersetzung für Akbar, sondern als eine für seinen Sohn,
Prinz Salı̄m, beschrieben ist. Ausschlaggebend für diese Zuschreibung war
offenbar eine von dem Prinzen selbst stammende Anmerkung am Rand der
ersten Seite, in der er sich als den Auftraggeber der vorliegenden Übersetzung
bezeichnet, die er in seinem „[. . . -]zwanzigsten Lebensjahr“ (s.o.) habe anfer-
tigen lassen. Liest man jedoch den eigentlichen Text der ersten Folios, so wird
deutlich, dass nicht Salı̄m, sondern Akbar diese Übersetzung befohlen hat. So
wird schon auf Fol. 2a „Seine Majestät, Schatten Gottes, Salomo-Siegel, Schutz
des Kalifats, Herr der Glückskonjunktion, Ǧalāl ad-Dı̄n Akbar Pādišāh“ als
Auftraggeber genannt. Auf Fol. 2b, Zeile 5 lesen wir: „Abū’l Muz.affar Sult.ān
Salı̄m Šāh Akbar Ġāzı̄ – Gott möge seine Herrschaft verewigen! “ Und schließ-
lich findet sich der Name Akbars nochmals auf Fol. 2b, letzte Zeile:
tā buwad bar āsmān hurš¯ıd-i anwar bādšāh

42 M UJTABĀ ’ Ī (1976), S. xlf.
43 In den drei Übersetzungen wird „Yogavāsis. t.ha“ je verschieden geschrieben: In derjenigen für
Akbar als Ǧōg Bāšištha, in der für Prinz Salı̄m als Ǧōg Bāsišt und schließlich in der Übersetzung
für Dārā Šikōh als Ǧōg Bašist.
bādšāh-i haft kišwar bād Akbar bādšāh

„Solange der Padischah am Himmel die leuchtende Sonne sein
wird, möge der Padischah der sieben Klimata Akbar Padischah
sein.“
Die Übersetzung ist im Kolophon datiert:
„43. Abschnitt des s¯ıt-bhūmikā ayā-khyāna aus dem Kapitel Nirbāna
prakarana. Abgeschlossen wurde das Buch Ǧōg Bāšištha am Sonn-
tag, entsprechend dem 15. des Monats Āz- ar des Ilāhı̄-Jahres 47
[1602/03 u.Z.].“
Der Übersetzer des Werkes stellt sich auf Fol. 3a vor als „der geringste der
Muriden von Kabir, der mit vielen Fehlern [behaftet] und als Farmulı̄ bekannt
ist“.
Es kann kein Zweifel daran bestehen, dass es sich hier um eine Überset-
zung für Akbar handelt und nicht für Salı̄m. Dieser hat seine Notiz an der
falschen Handschrift angebracht. Rätselhaft ist nur, warum die insgesamt 41
Textillustrationen offenbar in den Ateliers des Prinzen entstanden sind,44 der
ab Mitte des Jahres 1600 gegen seinen Vater rebelliert hat und, mit einer län-
geren Unterbrechung im Jahre 1603, bis 1604 in der Festung von Allāhābād
Hof hielt. Der Text des Chester-Beatty-Manuskriptes scheint ebenfalls erst in
Allāhābād geschrieben worden zu sein. Ein Vergleich der Schrift mit dem et-
wa zeitgleich entstandenen illustrierten Rāǧ Kunvar, das von einen gewissen
„Burhān aus Allāhābād“ niedergeschrieben wurde,45 legt den Schluss nahe,
dass dieser auch der Schreiber des Ǧōg Bāšištha gewesen sein dürfte.
Eine mögliche Erklärung für den merkwürdigen Umstand, dass in den
Ateliers von „Šāh Salı̄m“ nicht etwa die für ihn selbst erstellte Übersetzung
kunstvoll kopiert und illustriert wurde, sondern diejenige für Akbar, wäre ei-
ne schlichte Verwechslung der beiden Texte. Geplant war wahrscheinlich, der
Übersetzung für Salı̄m eine würdige Form zu verleihen; nur der Zufall hat ver-
mutlich dazu geführt, dass stattdessen heute die Übertragung von Farmulı̄ in
dieser Form vorliegt.
Gemäß einem Eintrag im Katalog der Punjab Public Library von Lahore46
wird dort noch ein Manuskript der Farmulı̄-Übersetzung verwahrt, bei dem
44 Vgl. dazu die Ausführungen von L EACH (1995), S. 158f; siehe auch S. 143.
45 Vgl. L EACH (1995), S. 189.
46 A BBĀS Ī (1963), Nr. 22.
es sich um die einzige existierende Abschrift des Textes zu handeln scheint.

Gemessen an den zahlreichen Kopien, die von den anderen Übersetzungen
existieren, hat die Farmulı̄-Fassung also kaum Verbreitung gefunden.
Die äußerst geringe Rezeption scheint ein allgemeines Phänomen der
Akbar-Übersetzungen zu sein, wie wir es auch von den beiden nahezu
vergessenen Neuübersetzungen des Pañcatantra47 oder des Kathāsaritsāgara48
kennen. Möglicherweise waren auch sprachliche Gründe für das Vergessen
der Akbar-Übersetzungen ausschlaggebend. Der geplante Vergleich der drei
Übersetzungen des LYV – derjenigen für Prinz Salı̄m von 1597, einer weiteren
für Akbar von 1602 und schließlich einer dritten für Dārā Šikōh von 1656 –
wird hoffentlich auch darüber Aufschluss geben.
ANHANG
Die persischen Übersetzungen des Laghuyogavāsis.t.ha im Überblick
1) Ǧōg Bāsišt
Übersetzung für Šāhzāda Sult.ān Salı̄m aus dem Jahre 1006/1597–98 von
Niz.ām ad-Dı̄n Pānı̄patı̄
Incipit der Vorrede: šukr u sipās b¯ı-qiyās sazāwār-i h.ażrat-i dādār
Incipit des Textbeginns: brahmanān-i hind-rā dar wah.dat-i z- āt-i h.aqq
subh.āna tacāla
Abschriften mit Vorrede:
1. London, British Museum, Or. 8443. 259 Folios; kopiert 1238/1822 endet
mit dem 43. Sarga (M EREDITH -O WENS 1968: 37).
2. Kalkutta, Asiatic Society of Bengal, Nr. 1699 386 Folios, Nastaclı̄q, ko-
piert 1151 HŠ/1773 AD (I VANOW 1924).
3. Qum, Az.ami Bd. 14: 47f., Nr. 5252 291 Folios, kopiert 1244/1828
(H USAYN Ī 1366Š/1987, Bd. 14).
4. Qum, Az.ami, Bd. 18: 170f., Nr. 6999 Fol. 14v–354v (Sammelhandschrift);
kopiert 1258/1841 (H USAYN Ī 1368Š/1989, Bd. 18).
47 Siehe Pančākyāna im Literaturverzeichnis und Iyār-i dāniš in L EACH (1995), S. 74–105.
48 Siehe Daryā-yi asmār im Literaturverzeichnis.
Abschriften ohne Vorrede:
1. Oxford, Bodleian Library, Ms. Walker 117 131 Folios; kopiert 1108/1697
endet mit dem 43. Sarga. (S ACHAU /E THÉ 1889, Nr. 1328).
2. London, India Office, Ms. Nr. 806 (Ethé 1971) 224 Folios; kopiert
1177/1764 (E THÉ 1903).
3. Paris, Bibliothèque Nationale; 327 Folios; kopiert 1184/1770 (B LOCHET
1905, Bd. 1, 223).
4. London, British Museum, Add. 5644; 513 Folios; kopiert etwa Ende des
18. Jh.s endet mit dem 46. Sarga. (R IEU 1879–1881: 61f).
Abschriften, die von der Verfasserin nicht eingesehen wurden und bei
denen daher nicht bekannt ist, ob die Vorrede eingeschlossen ist:
1. Mašhad, Āstān-i Quds enthält den Kommentar von Mı̄r Findiriskı̄

Fihrist-i kutub-i kitābhāna-yi Āstān-i Quds-i Rad.aw¯ı. Mašhad 1305–
1346Š/1927–1968. Bd.IV, S. 339f.
2. Teheran, Bibliothek der Maǧlis Kommentar von Mı̄r Findiriskı̄ am Rand
Fihrist-i kitāb-hāna-yi Maǧlis-i Šurā-yi Mill¯ı. Teheran 1318–1321Š/1940–
1943. Nr. 651.
3. Teheran, Universitätsbibliothek Fihrist-i kitāb-hāna-yi Dānišgāh-i. Tehran.
(Bibliothèque de L’Université de Tehran.) Teheran 1332Š/1954. Bd. III,
Teil 1, S. 443.
4. Teheran, Nationalbibliothek 241 Folios; Übersetzung ab Seite 32; schließt
den Kommentar von Mı̄r Abū ’l-Qāsim Findiriskı̄ ein. Vermutl. 19. Jh.
Anwar 1354Š/1976, Bd. VI, S. 207f., Nr. 2646/F.
Edition: Ǧōg Bāsišt (Yoga Vasistha): dar falsafa wa cirfān-i hind.

1360HŠ/1981.
2. Ǧōg Bāšištha
Übersetzung für Akbar, vollendet im Dezember 1602 von Farmulı̄, Schüler
des Kabı̄r.
Incipit der Vorrede: h.amd wa sipās mar-mubdic¯ı-rā ki cālam-rā

Incipit des Textbeginns: ba-h.aqq al-h.aqq wa ahlihi sitāyiš
1. Dublin, Chester Beatty Library, Ms. 5, 323 Folios, 41 Illustrationen;

„vollendet am Sonntag, entsprechend dem 15. Tag des Monats Āz- ar des
Ilāhı̄-Jahres 47“. (L EACH 1995), Bd. 1, S. 155–189.
2. Lahore, Panjab Public Library. A BB ĀS Ī 1963, Nr. 22.
3. Ǧōg Bašist
Übersetzung im Auftrag Dārā Šikōhs aus dem Jahre 1066/1655–56 von un-
bekanntem Übersetzer.
Incipit der Vorrede: sipās wa sitāyš tamām niyāyiš
Incipit des Textbeginns: Bālm¯ık mus.naf-i kitāb-i ōg Bašist m¯ı-farmāyad
1. Berlin, Staatsbibliothek, Pertsch (1077(1), bis Fol. 105a. (P ERTSCH 1888:

1021).
2. London, India Office, Ms. Nr. 1185; (E THÉ 1972) 115 Foll.
3. London, India Office, Ms. Nr. 1859; (E THÉ 1973) 213 Foll.
4. London, India Office, Ms. Nr. 1355; (E THÉ 1974) Fol. 33–147.
5. London, India Ofice, Ms. Nr. 3165; (E THÉ 2927) Ethé 1903/Edwards
1937.
6. London, British Museum, Add. 5637; 155 Foll; kopiert 1192/1784 (R IEU
1879–1881: 61).
7. Cambridge, Library of the University Catalogue of the Persian Mss. in
the Library of the University of Cambridge. Cambridge 1896. Nr. 35(2).
8. Kalkutta, Asiatic Society of Bengal Nr. 1700 (I VANOW 1924).
9. Kalkutta, Curzon Collection der ASB, Nr. 680 (I VANOW 1926).
10. Patna, Oriental Public Library. M AULVI A BU ’ L M UQTAQDIR: Supple-
ment to the Catalogue of the Persian Manuscripts in the Oriental Public Libra-
ry, Patna. Patna 1932. (II 2080)
11. Lahōr, Panjab Public Library
12. Edinburgh, University Library. M OHAMMED A SHRAFUL H UKK: A

descriptive Catalogue of the Arabic and Persian Mss in the Edinburgh Uni-
versity. Edinburgh 1925. Nr. 327.
Editionen: Lithographische Publikation: Cawnpore 1883.

Yogavāsis.t.ha (Ǧōg bašist) 1968.
Vakatseite
The Metaphorical Logic of the Moks.opāya
B RUNO L O T URCO
My paper takes as its starting point an article by J OHANNES B RONKHORST

entitled ‘Pour comprendre la philosophie indienne’.1 In this article, among
other things, B RONKHORST deals briefly with the Moks.opāya2 or Yogavāsis.t.ha.
B RONKHORST reports3 the philosophical position of the MU on the basis of
S LAJE’s book Vom Moks.opāya-Śāstra zum Yogavāsis.t.ha-Mahārāmāyan.a:4 “The
author of the Moks.opāya teaches a subjective illusionism, which denies the
existence of a world that is real and has an objective existence. The world
is merely imaginary; also the god-creator Brahmā is no more real than the
model in the mind of a painter. The only thing that exists is consciousness.”
B RONKHORST wonders what the arguments deemed to support this position
are. According to him, S LAJE5 answers this question explicitly: “The method
of argumentation (yukti) that prepares the correct understanding of the philo-
sophical point of view (siddhānta) of the YV [Yogavāsis.t.ha] consists above all
in examples (dr.s.t.ānta) in the form of stories (ākhyāna, kathā).” In other words,
S LAJE is seeking here to demonstrate – I am still following B RONKHORST –
an equivalence, though a partial one, between method of argumentation
(yukti) and story (ākhyāna, kathā) that acts as an example (dr.s.t.ānta). Now,
B RONKHORST adds that as far as this equivalence is concerned one remains
skeptical for two reasons: the first reason is that usually yukti (method of argu-
mentation) does not mean dr.s.t.ānta (example); the second reason is that ‘true’
arguments (namely rational arguments) in favour of idealism exist, and are
contained both in the text of the MU and elsewhere. Therefore, if we conclude,
with B RONKHORST, that the stories in the MU are not arguments in favour of
idealism, then what are they in relation to the whole work? As a matter of fact
B RONKHORST does not reply to this question, which now demands our atten-
tion. What is the nature of the stories in the MU? If they cannot be considered
arguments in favour of idealism, and therefore they do not have philosophical
or rational relevance, can we say that they constitute only a sort of propaedeu-
1 B RONKHORST (2001).
2 Henceforth MU.
3 B RONKHORST (2001), p. 208.
4 S LAJE (1994).
5 Cit. in B RONKHORST (2001), p. 208.
132 Bruno Lo Turco: Metaphorical Logic
tic appendix, which is superfluous and extraneous to the essential core of the
philosophical work?
Our main thesis is that, in reality, the stories in the MU are totally consub-
stantial to the work, and that they cannot be removed, or passed over, or put
in brackets, without our understanding of the MU being seriously impaired.
Why therefore are the stories in the MU consubstantial to the work? If
we pose questions about a text (as B RONKHORST does, for instance, and as
we do now), this happens because we wish to understand that text. And we
wish to understand it because “it seems to us that, in some way, we do not
understand”, as W ITTGENSTEIN would say. And why does it seem to us that
we do not understand? The reason is that the text before our eyes is not a
shopping list, neither is it anything similar. In other words, the meaning of
the text is not immediately obvious, since we are separated from it by a dis-
tance that is historical, linguistic, psychological etc.6 This means that the text
before us, if we wish to understand it, must be interpreted. Nevertheless, as
G ARRONI7 maintains, “the interpretation of a text is not an understanding of
it subsequent to the establishment and the reading of the text [. . . ] A reading
is already an interpretation, and a text does not exist at all outside an interpre-
tation or an understanding, except as a material object.” The interpretation
or the understanding of a text is not something that follows the reading of it,
but something that already always accompanies the reading. The interpreta-
tion is the “condition of observability” of the text as such, and therefore it is
“unobservable in principle”. But if this is so, if the interpretative function can-
not be isolated and structured at will, is there any escape from interpretative
arbitrariness?
What we can say – we are still following G ARRONI8 – is that the interpreta-
tion cannot be a set of already existing conceptual instruments that is applied
to the text, but something that “is revealed through” the text itself, since the
text is read and at the same time inevitably already interpreted. The text es-
tablishes itself as such always and only through an interpretation (there is no
text before its reading/interpretation, but only a material thing). Therefore
the interpretation has, at least, the task of establishing “precisely that text”. In
other words, the interpretation must dwell, in order not to be totally arbitrary,
on the very letter of the text.
6 I paraphrase here the clear definition of ‘interpretation’ given in VATTIMO (1981), p. 451.
7 G ARRONI (1996), p. 247–248.
8 G ARRONI (1996), p. 254.
Bruno Lo Turco: Metaphorical Logic 133
And what do we find, if we dwell on the letter of the text of the MU? We
find of course that, in fact, the MU has both philosophical sections and narra-
tive sections, and that neither are presented as mere appendixes. But this is not
enough. The MU has much to say about itself and opens up consciously and
clearly particular lines of interpretation. The work defines itself (or sections of
itself) as śāstra,9 kāvya,10 ākhyāyikā,11 itihāsa12 etc. at the same time. Therefore
it is explicitly stated that the text combines different literary genres.13 Could
this be merely a secondary aspect? In other words, is the presence of narra-
tive sections unessential to an understanding of the text, in spite of the way in
which the text presents itself?
At this point, by examining more closely every single ākhyāna, each nar-
rative section, perhaps it will be possible to add something about the deep
structure of the entire MU. On the one hand the narrative sections carry out, if
we consider each ākhyāna as a whole, the function of dr.s.t.ānta, that is they are
illustrations of the philosophical reasoning or ‘parables’.
On the other hand it is also true that often, in the course of each ākhyāna,
the MU describes, or rather, shapes precisely that universe, the escape from
which it teaches through its philosophical sections and the symbolic meaning
of every ākhyāna as a whole (‘the moral of the story’).
Now, when the MU opens up and establishes a world, it shows itself in the
guise of a kāvya. In other words, as a kāvya, it creates the Hindu cosmos, a cos-
mos that at the time when the MU was written was actually going through a
process of determination. In effect the treatment of the descriptive and narra-
tive materials takes place in the style of a kāvya. Therefore the MU is by rights
śāstra and kāvya at the same time.
From the typically Indian standpoint, the establishment of a world implies
in general also the problem of escaping from that world. This escape, accord-
ing to the MU, is not literally an escape: it is equivalent, in fact, to the discov-
ery that there is not any world from which one can extricate oneself, that one
has always been free. The world is understood as a subjective projection (even
if this projection is shared by a great number of individuals).
9 See, for example, MU 1.1.2 (S LAJE (1996), p. 28); MU 2.13.13 (S LAJE (1993), p. 99).
10 See, for example, MU 2.18.33 (S LAJE (1993), p. 143).
13 Cf. B OISTARD (1994).
Therefore the MU is composed of a pars construens and a pars destruens:

the pars construens consists in establishing, through the imagination, a uni-
verse around oneself; the pars destruens consists in the destruction of the uni-
verse, which follows the individual’s realization that he himself established
the universe by imagining it. Roughly, the pars construens coincides with the
descriptive and narrative discourse of the MU in the style of a kāvya, while the
pars destruens coincides with the large number of philosophical sections and
the final symbolic meaning of the stories. In other words, the MU alternately
creates and destroys the Hindu universe, which in this respect is similar to
Penelope’s weaving.
It is useful to remember that this universe is natural and cultural at the
same time, without these two aspects being distinguishable, so that its creation
and dissolution regards both nature and human institutions. Therefore the
MU supplies us, in its pars construens, with a rich series of descriptions of
elements pertaining to both aspects; these descriptions say both what it is like
and what it should be like at the same time; they are both statements and
instructions or rules. They are, according to the well-known terminology of
C. G EERTZ “model of” and “model for”. The description of a monarch given
in Yogavāsis.t.ha 3.19.5ff. can be cited as an example of the representation of a
human institution that must be included in the pars construens of the poem.
It is obvious that the purpose of such a description is not only artistic, and
functional in relation to the story as a whole, but also regulative, in the sense
that it makes clear and confirms what the Hindus mean and what they should
mean by royalty, and consequently the way a monarch should look and act.
At this point we could ask ourselves why the MU manifests two conflicting
trends; in other words, why it establishes a world (through kāvya) in order to
deny it (through śāstra and the symbolic meaning of each ākhyāna), or why it
denies a world in order to establish it.
If we wish to reply to this question, perhaps we can resort to narratology,
since the MU states that it belongs, among other things, to the literary genre
of the ākhyāyikā, the short narrative. In a narrative we must distinguish two
planes: fabula and story, namely what is told and how it is told. This distinc-
tion is virtually useless as long as the two planes develop in harmony, which
is supposed to happen in the works of the followers of naturalism and real-
ism. In these works the way in which the story is told depends totally on what
is told (assuming that this is possible). On the contrary this distinction is jus-
tified when there is no perfect mutual confirmation between the two planes,
but a dialectic relationship. There is no doubt that in the MU the fabula (in
harmony with the philosophical discourse, with śāstra) points out the non-
existence of the world. On the contrary the story, in the style of a kāvya, opens
up a magnificent world. Therefore the originality of the MU consists in the
fruitful dialectic of these two; it is only the juxtaposition of the two planes that
is able to generate a meaning that is much more complex (and much less ex-
pressible in a philosophical-rational language) than the simple meaning of the
fabula.
Consequently, in addition to the typical polysemy of poetical language,
the MU possesses polysemy on a more abstract, more general level, namely
the level of literary genres, of the structure consisting in an interlocking of
the two divergent planes of śāstra and kāvya, while the planes of śāstra and
ākhyāyikā (or fabula) tend to converge. Śāstra and ākhyāyikā converge because
in the MU the ākhyāyikās are in effect parables, as B OISTARD14 has shown,
namely narrations that by virtue of their symbolic general meaning are set
within a theoretical context (śāstra in this case) that substantially confirms it.
As things stand, the interpretations of the MU that tend to sacrifice ei-
ther the plane of kāvya or the twofold level fabula/story by departing from
the letter and the very structure of the text also sacrifice the polysemy result-
ing from the coexistence of the literary genres. Therefore, in spite of the fact
that the stories of the MU are not arguments, in the strict sense of the word,
in favour of subjective illusionism, they can never be considered secondary
or auxiliary, without our understanding of this specific text suffering serious
consequences.
Nevertheless a very important question still remains unsolved: if the
planes of śāstra and ākhyāyikā (or fabula) tend to converge, what is the pre-
cise function of the fabula in the MU? Is it only a repetition of the śāstric plane
with a mere propaedeutic function?
In order to answer this question, I wish to refer to the distinction made by
D. D AVIDSON15 between literal and metaphoric, or rather, to the use that R.
R ORTY16 makes of such a distinction, which in my opinion exemplifies very
well the distinction between śāstra and fabula in the MU. A metaphor is not
something that has an actual meaning and more precisely it is not, as one
would expect, something that has a meaning that is different from the literal
14 B OISTARD (1994).
15 D AVIDSON (1979).
16 R ORTY (1989), p. 18–19.
one. To have a meaning means to be part of a language game and a metaphor

is not part of it by definition. A metaphor cannot be broken down into the
elements of a language game; it represents a break with that game. In fact, if
one had wanted to say something, namely to formulate an utterance with a
meaning, one would have done so. Introducing a metaphor in a conversation,
or an ākhyāyikā in the course of a śāstra, is like “making a face”, or “pointing
at” something, or “slapping your interlocutor”. All these are primarily ways
of making an impact on one’s interlocutor or reader, and not ways of commu-
nicating a message.
This is exactly the case of the stories in the MU. They are meant above all to
make an impact on the listener or reader. In this sense they have a metaphoric
value. They are not meant to support through rational argumentation so-
called subjective illusionism, but to point directly at the world of subjective
illusionism for the benefit of the reader. The stories in the MU, since they pos-
sess a metaphoric value, do not have a definite place in the śāstric language
game; in other words, exactly as B RONKHORST has noted, they do not con-
stitute a yukti, a mode of argumentation in the strict sense of the word (for
accuracy’s sake, we must add that yukti does not necessarily mean “argumen-
tation”; it can mean also “method” and “skill”). In fact, within that language
game, within the sphere of the śāstric language, their truth or falsity cannot
be determined. They can only be “savored or spat out”. But that does not
mean that they cannot be truth-value candidates. When, in fact, metaphors
become habitual, perhaps giving rise to a new language game, they cease to
be metaphors and become arguments, that is dead metaphors, sets of utter-
ances in which truth or falsity can be distinguished.
Therefore, the coexistence of śāstra and ākhyāyikā in the MU is evidence of
the fact that the author held a position, that of so-called subjective illusion-
ism, that was not easily tenable within the sphere of the śāstric language game
known at that time. And in fact B RONKHORST17 finds the argumentation of
the MU in the śāstric sphere particularly inadequate. Thus the author of the
MU corroborates his own arguments with stories, which are deliberately situ-
ated outside the śāstric language game.
The argumentative and rational function to which B RONKHORST refers,
the śāstric function, from which he quite rightly excludes the stories of the MU,
makes sense as a check of the coherence of the inferences within the narrow
17 B RONKHORST (2001), p. 212.

sphere of a particular socio-linguistic system.18 But the aim of the author of

the MU is precisely to induce the listener or reader of the work to place himself
outside that socio-linguistic system.
That particular socio-linguistic system probably consisted, at least partly,
in orthodox Brahmanism (which in fact is sometimes attacked explicitly). Had
it not been so, no one would have felt the need to rework the text along the
lines of orthodox Brahmanism (a reworking that undoubtedly took place, as
S LAJE19 has shown). Moreover, the MU explicitly attacks and devalues the
language of the philosophical schools known at that time. These schools,
through their language, multiplied entities unnecessarily. On various occa-
sions several cosmological and psychological principles established by the
philosophers of the different traditions are interpreted as traceable actually
to a single principle. For example, according to the MU the terms buddhi,
ahaṅkāra, citta, karman, kalpanā, smr.ti etc. are in reality equivalent to manas,
‘mind’. It is the mind alone that carries out all those functions (cf. Yogavāsis.t.ha
3.96.44). This is not a mere anti-intellectualistic polemic, as is shown by the
fact that the author of the MU greatly emphasizes vicāra, namely reflection,
investigation. He attempts to create a new paradigm. The author of the MU
was the kind of philosopher who aspired to create what we would describe
as a new language game; he was concerned about dissolving the problems in-
herited from tradition, rather than solving them, as R ORTY20 would say. And
in fact he never declares that he follows in the wake of any tradition.
It is necessary to specify that what we have outlined here, for the sake
of clarity, as a perfect incommensurability between the śāstric language and
the metaphoric language is not actually such. Perfect incommensurability is
merely an ideal situation. Incommensurability regards in reality subgroups of
terms and the sentences that contain them. Moreover, we know that histori-
cally the language of subjective illusionism was in effect demetaphorized, in
the sense that it was reabsorbed by the śāstric language, and began to be fully a
part of it. Of course the śāstric language was in its turn irreversibly changed by
this process, since every language is by nature a coordinated whole, a system.
Moreover the non-śāstric language of the MU is metaphoric also in an-
other remarkable sense. M. B LACK21 has emphasized that the most important
18 Cf. R ORTY (1989), p. 3ff.
19 S LAJE (1994).
20 R ORTY (1989), p. 20.
21 B LACK (1962).
metaphors from the philosophical point of view are the interaction metaphors,
namely those metaphors that are not confined to a comparison between two
contexts, because they modify both. For example, one of the central metaphors
of the MU consists in associating the world, which is the original context, and
dream, which is the applied context. To say that the world is a dream does
not only modify our understanding of the world, but also modifies our un-
derstanding of dreaming, since to dream thus means to create a world. There-
fore replacing the metaphor of the world as a dream with a literal explana-
tion would imply a serious impoverishment of the cognitive capacity of the
metaphor. According to B LACK this kind of metaphor is philosophically im-
portant because it explains, at least partly, modelization, and what the author
of the MU wants to provide through his stories is precisely a model. As a re-
sult, once more, what is metaphoric in the MU cannot simply be removed, or
translated into non-metaphoric terms, without our understanding of the text
suffering serious consequences.
What is important, here, is the conclusion that the stories in the MU must
not be put in brackets, or passed over because they are irrational, in spite of
B RONKHORST’s thesis that they are not arguments in favour of idealism. On
the contrary, the stories in the MU play a pivotal role in the comprehension of
the text, since they are effectively arguments, except for the fact that the logic
that supports them is metaphoric, in the sense that instead of producing infer-
ences within an already known language game or paradigm, it heralds a new
language game or a new paradigm. The embarrassment of the western inter-
preter before a work that makes extensive use of metaphoric language within
a śāstric framework derives from a philosophical position, adopted more or
less consciously, which disqualifies metaphor from the philosophical-rational
sphere and turns it into a mere stylistic tool, while in reality metaphor is a
means of increasing knowledge. That is a typical old positivist position, but it
goes back to Hobbes, who maintained in the Leviathan that the source of the
absurd assertions of philosophers is the use of metaphors instead of proper
words. It is hardly necessary to note, here, that this position can be abandoned
nowadays. The importance and, indeed, the need for the use of metaphor
within the philosophical-rational sphere (and of course within science too),
have been sufficiently emphasized since the middle of the last century, thanks
to the efforts of N ELSON G OODMAN, the above-mentioned D ONALD D AVID -
SON and M AX B LACK, H ERBERT PAUL G RICE , M ARY H ESSE , and others.
Appendix 139
Moks.opāya Manuscripts
Ś1 Complete Facsimile ed. by L. C HANDRA (S LAJE 1994, p. 38f)

Ś3 Complete Sri Pratap Singh Library, Srinagar (S LAJE 1994, p. 39f)
Ś4 T.ı̄kā: Vairāgya BHU 328118 (S LAJE 1994, p. 40f)
Ś5 Nirvān.a/Khilas Ms. Sansk. c.89, Bodleian Library, Oxford (S LAJE
1994, p. 41)
Ś6 T.ı̄kā: Sthiti 1–5 Hs. or. 12704 SBPK Berlin
Ś7 Utpatti (–4.18) No. RAR/181.045/MOK IGNCA New Delhi
Ś8 1.25.31d–32.32, 4.21.53– Ms Indic (beta) 494 Wellcome Institute for the His-
5.12.4 tory of Medicine (London). foll. 4r–57v
Ś9 Vai to Upa 93.62 Hs. or. 12511 SBPK Berlin. 687 fols.
Ś10 Mumuks.u; Sthiti [19.1–] Hs.or.12869 SBPK Berlin
Ś11 Mumuks.u RAR 181.045 IGNCA New Delhi
Ś13 Vairāgya 4796/1821.10 Research & Publ. Dep., Śrı̄nagar
Ś14 Vairāgya–Upaśama 4788/827 Research & Publ. Dep., Śrı̄nagar, birch bark
Ś15 Vairāgya–Upaśama 4795/1562 Research & Publ. Dep., Śrı̄nagar.
Ś16 Nirvān.a/Khilas 57.106/100 National Museum (New Delhi)
Ś17 (NEd ) Khilas 4797/2281 Research & Publ. Dep., Śrı̄nagar. Tran-
script of NEd , but with addition of Khilas
Ś18 Vairāgya 4791/1212.06 Research & Publ. Dep., Śrı̄nagar
Ś19 Nirvān.a 4793/1274 Research & Publ. Dep., Śrı̄nagar
Ś20 Nirvān.a 4793/1274 Research & Publ. Dep., Śrı̄nagar
Ś21 Utpatti–Sthiti 4790/1155 Research & Publ. Dep., Śrı̄nagar
Ś22 Utpatti–Upaśama 4798/2305 Research & Publ. Dep., Śrı̄nagar
N6 Mumuks.u BORI 778 Visram II/218 (R 583) (S LAJE 1994, p.33)
N8 Vairāgya to Upaśānti BORI 772 Visram I/450 (R 579). A.D. 1666.
N10 Vairāgya to Upaśānti NGMPP A 95/12 (S LAJE 1994, p. 34)
N11 T.ı̄kā: Vairāgya BHU 328648 (S LAJE 1994, p. 34ff)
N12 T.ı̄kā: Mumuks.., Utp. BHU 328148 (S LAJE 1994, p. 36ff)
140 Moks.opāya Manuscripts
N13 T.ı̄kā: Utpatti BHU 331122 (S LAJE 1994, p. 38)

N14 Incomplete NGMPP B 72/10 [= A 899/10]. ‘Sammelhandschrift’
of very poor scribal quality.
N15 Nirvān.a/Khilas BORI 775 Visrama II/24
N16 Khilas IOL Nr.2423 (Catalogue of the Sanskrit Manuscripts
in the Library of the India Office (London 1894), Pt
IV, A.VII)
N17 T.ı̄kā: Sthiti 1–33 SBPK Berlin Hs.or.12704
N19 Vairāgya BORI 779 Viśrāma II/217 (R 583). Same scribe as N6.
38 fols.
N20 3.1.1–57, 20; 3.101.28c– BORI 776 Visrama II/224.
4.11.68a
N21 3.61.18–64.21; 3.114.26– BORI 773 Viśrāma II/232 (R.581). Folios 124–127 and
115.16; 4.19.31-57.1; 223–225 of the Utpatti erroneously embedded in the
5.48.31c-53.35; 5.53.77– Upaśānti (= foll. 108–124; 127–222). 121 fols.
93.53
N22 Sthiti 22.26–61.8c BORI 777 Viśrāma II/225 (R 583). Seems to belong
originally to N20 , 2nd hand.
N23 Nirvān.a 2.49c–98.1a BORI Viśrāma II/116 (nicht in Delhi). 206 fols.
N24 Nirvān.a 132 to Khilas BORI Visrama II/117. Nirvān.a 132 to Khilas XII
XII
N25 Nirvān.a/Khilas BORI Viśrāma I/419. Dated sam
. 1800, 913 fols.
N26 T.ı̄kā: Nirvān.a 195–271 BORI Viśrāma I/623. 208 fols.
N27 Nirvān.a 115.43b–215.17 SBPK Berlin Hs. or. 11354. contains a Khila-
Pratisandhi-śloka. fols. 172–322.
N28 Sthiti 19.1–53.8c SBPK Berlin: Hs. or. 11724
N29 Khilas NGMPP Reel No. E 719/2. Text scattered. Khila-
ślokas appear as inserted after Sarga 2.51 of the
Nirvān.aprakaran.a
N30 Nirvān.a/Khilas Ānandāśrama Sam
. sthā, Pune S12(1)-4-38
Appendix 141
An Overview of the Moks.opāya Literature
A.D. Longer Versions Shorter Versions Commentaries
1000 Moks.opāya
1200 Laghuyogavāsis.t.ha Ātmasukha
Yogavāsis.t.ha Jñānavāsis.t.ha Rāmabrahmendra
1400 Moks.opāyasam
. graha Mummad.ideva
Vāsis.t.harāmāyan.asāra
Vāsis.t.hasāra Mahı̄dhara
1600 Śrı̄vāsis.t.hacandrikā
Moks.opāyasāra Ānandabodhendra
1800 Bhāskarakan.t.ha
YV-Edition LYV-Edition
1900 (Vāsis.t.hadarśana)
(Yogavāsis.t.hasam . graha)
The above chart lists all known texts and commentators in a rough chrono-
logical grid, although the position of most of the works is merely a guess:
apart from the date of the MU, only a terminus ad quem can be given for some
of the works. In this chart colours indicate close relationship between texts, as
also between texts and commentaries, gray indicates that no clear affiliation
has as yet been established. Two recent compilations22 have been listed for
completeness, but these are solely dependent on the printed editions.
22 For the Vāsis.t.hadarśana, see ATREYA (1936); the Yogavāsis. t.hasam

. graha was complied by
J Ñ ĀN ĀNANDA B HĀRAT Ī and published Śrı̄raṅga 1973.
142 The Moks.opāya Literature
The Moks.opāya Literature: Dependencies
The following chart gives a resume of the present state of knowledge about
the dependencies of the different versions. Except for the MU, where the date
can be considered as established, the dates in brackets are merely the terminus
ad quem derived from manuscript datings, or – in the case of the LYV – the
date of the first quotation. The position of the Jñānavāsis.t.ha still needs to be
established.
Yogavāsis.t.ha Vāsis.t.harāmāyan.asam
. grahasāra
Moks.opāyasāra
Moks.opāya (∼ 950)
Laghuyogavāsis.t.ha (-1258) Vāsis.t.hasāra (-1597)
Śrı̄vāsis.t.hacandrikā (-1660)
Moks.opāyasam
. graha (-1854)
Appendix 143
Als König Janaka in seinem Garten lustwandelt, hört er Gesänge von Sid-
dhas, die ihn zum Nachdenken über die illusorische Natur der Welt und
des Selbst veranlassen. – Abbildung aus dem Ǧōg Bašista für Akbar (A.D.
1602/03). Chester Beatty Library, Ms. 5, Fol. 128b. © The Trustees of the Chester
Beatty Library, Dublin.
144 Bibliography
Bibliography
Sanskrit Texts
Moks.opāya(-t.¯ıkā) Vairāgyaprakaran. a: H ANNEDER and S LAJE (2002)
Mumuks.uvyavahāraprakaran. a: S LAJE (1993)
Sthitiprakaran.a: S LAJE (2002)
Moks.opāyasam . graha 1 Ms., see p. 105
Moks.opāyasāra 2 Mss.:
(1) (662/1887-91) Roll 1723 Ś-Script
(2) Vishvesvaranand Vishva Bandhu Institute Panjab Uni-
versity Hoshiarpur (Acc. No. 1537)
Yogavāsis.t.ha The Yogavāsist.ha of Vālmı̄ki with the Commentary
Vāsist.hamahārāmāyan. atātparyaprakāsha, ed. Wāsudeva
Laxman.a Śāstrı̄ Pan.sı̄kar, Bombay 1911, 2 1918, 3 1937.
Yogavāsis.t.hasāra → Vāsis.t.hasāra
Laghuyogavāsis..tha Laghuyogavāsis.t.ha. Ed. V ĀSUDEVA Ś ARMAN . Bombay:
Nirn.aya Sagar Press 1937
Vāsis.t.hacandrikā Commentary on the → Laghuyogavāsis.t.ha by Ātmasukha
Vāsis..thatattvabodhin¯ı by Rāmabrahmendra.
Vāsis.t.hatātparyaprakāśa Commentary on the → Yogavāsis.t.ha by Ānandabodhen-
dra.
Vāsis.t.hasam . graha 1 Mss. Bodleian Library CSS d.559
Vāsis.t.hasāra Ed.: T HOMI (1999)
Vāsis..thasāravivr.tti Commentary on the → Vāsis.t.hasāra by Mahı̄dhara.
Śr¯ıvāsis.t.hacandrikā 1 Ms.: Indira Gandhi National Centre for the Arts. Dated
1660.
Sam
. sārataran.i Commentary on the → Laghuyogavāsis.t.ha by Mummad.i-
deva Allād.asūnu
Appendix 145
Persian Texts
Bhagawadg¯ıtā Bhagawadg¯ıtā, tarǧama-yi Muh.ammad Dārā Šikōh. Ed.

und mit einem Vorwort von M UH. AMMAD R I Ż Ā Ǧ AL ĀL Ī
N Ā ’ ĪN Ī. Ohne Ort, 1357 HŠ/1979.
Daryā-yi asmār Daryā-yi asmār. Tarǧuma-yi Katāsaritsāgara. Übersetzt von
Mus.t.afā Hāliqdād
Abbāsı̄. Hrsg., mit einer Einleitung

und Wortverzeichnis versehen von T ĀR Ā Č AND und
S AYYID A M ĪR H. ASAN
A BID Ī. New Delhi 1375 HŠ/1997
AD.
Ǧōg Bašist — Ǧōg Bašist. [Yogavāsis.t.ha] Hrsg. von TARA C HAND und
S.A.H. A BIDI. Aligarh Muslim University. Calcutta: The
Baptist Mission Press 1968. [Übersetzung für Dārā Šikōh].
— Ǧōg Bāsišt. Dar falsafa wa
irfan-i hind. Tarǧuma-
yi Niz.ām Pānı̄patı̄. Be-tas.h.ı̄h. wa tah.qı̄q-i S AYYID
M UH. AMMAD R I Ż Ā Ǧ AL ĀL Ī N Ā ’ ĪN Ī wa N( ARAYAN )
S( HANKER ) Š UKLA . Teheran, Iqbāl 1360 HŠ/1981. (Hind
šināsı̄ 18) [Übersetzung für Prinz Salı̄m].
Gulzar-i h.āl yā .tulū
-i Gulzar-i h.āl yā .tulū
-i qamar-i ma
rifat / Prabodhacandrodaya.
qamar-i ma
rifat Aus dem Sanskrit ins Persische von B ANWALI D AS. Hrsg.
von T ĀR Ā C HAND und S AYYID A M ĪR H . ASAN
A BID Ī. Ali-
garh 1962.
Mahābhārata (persisch) Übersetzt von Mı̄r Ġiyāsad-Dı̄n ‘Alı̄ Qazwı̄nı̄, bekannt
als Naqı̄b Hān. Hrsg. von S AYYID M UH. AMMAD R I Ż Ā

Ǧ AL ĀL Ī N Ā ’ ĪN Ī und N.S. Š ŪKL Ā . 4 Bände. Teheran 1380
HŠ/2001.
Pančākyāna Pančākyāna yā Panǧ dāstān. Übersetzt von Mus.t.afā
Hāliqdād Hašimı̄
Abbāsı̄. Herausgegeben und mit einer

Einleitung, Anmerkungen und Glossar versehen von
T ĀR Ā Č AND, S.A.H . .
Ā BID Ī und M UH. AMMAD R I Ż Ā
Ǧ AL ĀL Ī N Ā ’ ĪN Ī. Teheran 1984.
Sirr-i-Akbar Sirr-i-Akbar (Sirr-ul-Asrar). The oldest Translation of the
Upanishads from Sanskrit into Persian by Dara-Shukoh, Son
of Emperor Shah-Jahan. Edited, with introduction, notes,
commentary and comparison with the old manuscripts
and concordance with the original Sanskrit by Dr. TARA
C HAND and S.M. R EZA J ALALI N AINI. Iran, Taban Print-
ing Press 1957.
146 Bibliography
References
A BB ĀS Ī , M AN ŻŪR A H. SAN : Tafs.¯ıl¯ı fihrist-i maht.ūt.āt-i fārs¯ıya. Lahore: Panjab Public
Library 1963.
A NWAR , S EYYED A BDOLLAH: A Catalogue of the Manuscripts in the National Library. Bd.
VI. Teheran: Ministry of Culture and Art, National Library 1976 (1354 HŠ).
A RJUNWADKAR , K.S.: “Yoga-Vāsis.t.ha — A Frank Appraisal”. In: ABORI LXXXII

(2001): 213–231 [appeared 2002].
ATREYA , B.L.: The Yoga-Vasis.t.ha-darśanam. Allahabad: Government Press 1936 (The

Princess of Wales Sarasvati Bhavana Texts. 64).
B HATTACHARYYA , S IVAPRASAD: “The Bhuśun.d.opākhyāna and its Esoteric Drift”. In:

Kavirāj Abhinandana Grantha Lucknow 1967, pp. 79–89.
B LACK , M AX: Models and Metaphors: Studies in Language and Philosophy. Ithaca: Cornell
University Press 1962.
B LOCHET, E.: Cataloge des Manuscrits Persans. Paris 1905.
B OISTARD , M ARIE -J EANNE: “Les ambiguïtés génériques du Yogavāsis.t.ha”. In: Genres

littéraires en Inde. Ed. N ALINI B ALBIR. Paris: Presses de la Sorbonne nouvelle 1994,
pp. 287–324.
B RONKHORST, J OHANNES: “Pour comprendre la philosophie indienne”. In: La Ratio-

nalité en Asie. Ed. J OHANNES B RONKHORST . Lausanne: Études de Lettres 2001, pp.
195–223.
C HANDRA , L OKESH (Ed.): Sanskrit texts from Kashmir. Vols. 8–9. New Delhi 1984 (Śata-
Pit.aka Series. 334–335).
C HENET, F RANÇOIS: Psychogenèse et Cosmogonie selon la Yoga-Vāsis.t.ha. Paris: Édition-

Diffusion De Boccard 1998-99 (Publications de l’Institut de Civilisation Indienne.
67).
D AVIDSON , D.: “What Metaphors Mean”. In: On Metaphor. Ed. S HELDON S ACKS.
Chicago: University of Chicago Press 1979, pp. 29–45.
D IVANJI , P RAHLAD C ANDRASHEKARA : “The Date and Place of Origin of the Yo-
gavāsis.t.ha”. In: Proceedings and Transactions of the Seventh All-India Oriental
Conference. Baroda 1933 (1935): 15–28.
Appendix 147
D IVANJI , P RAHLAD C ANDRASHEKARA : “Further Light on the Date of the Yo-

gavāsis.t.ha”. In: Poona Orientalist 3 (1938): 29–44.
D IVANJI , P RAHLAD C ANDRASHEKARA : “The Text of the Laghu Yogavāsis.t.ha”. In:

New Indian Antiquary 1 (1939): 697–715.
D IVANJI , P RAHLAD C ANDRASHEKARA : “Yogavāsis.t.ha, a Vārttika on the Upanis.ads

by a Kaśmir Śaivite”. In: Bhāratı̄ya Vidyā 12 (1951): 26–49.
D URKIN -M EISTERERNST, D ESMOND, R ASCHMANN , S IMONE -C HRISTIANE, W ILKINS ,

J ENS and D URKIN -M EISTERERNST, D ESMOND: Turfan Revisited – The First Century
of Research into the Arts and Cultures of the Silk Road. Berlin: Reimer 2004 (Monogra-
phien zur indischen Archäologie und Kunst).
E RNST, C ARL W.: “Sufism and Yoga according to Muh.ammad Ghawth”. In: Sufi 29
(1996): 9–13.
E RNST, C ARL W.: “The Islamization of Yoga in the Amrtakunda Translations”. In: Jour-
nal of the Royal Asiatic Society. Series 3 13.2 (2003a): 199–226.
E RNST, C ARL W.: “Muslim Studies of Hinduism? A Reconsideration of Arabic and

Persian Translations from Indian Languages”. In: Iranian Studies 36.2 (2003b): 173–
19.
E RNST, C ARL W.: “The Pool of Nectar: Islamic Interpretation of Yoga”. ∗ 2005 [forth-
coming].
E THÉ , H ERMAN : Catalogue of Persian Manuscripts in the Library of the India Office. Bd. I.
Bd. II revised and completed by E. Edwards. Oxford 1903.
F UNAYAMA , T ORU : “Remarks on Religious Predominance in Kashmir: Hindu or Bud-

dhist?”. In: A Study of the N¯ılamata. Aspects of Hinduism in Ancient Kashmir. Ed. YA -
SUKE I KARI . Kyoto: Institute for Research in Humanities 1994, pp. 367–375.
G ARRONI , E MILIO: “Interpretare”. In: Il testo letterario. Istruzioni per l’uso. Ed. M ARIO
L AVAGETTO. Roma: Laterza (Manuali Laterza. 197) 1996, pp. 247–284.
G ÉRARD , R ENÉ : L’Orient et la Pensée Romantique Allemande. Paris: Didier 1963.
G LASENAPP, H ELMUTH VON : Zwei philosophische Rāmāyan.as. Wiesbaden: Franz

Steiner Verlag 1951 (Akademie der Wissenschaften und der Literatur. Abhandlun-
gen der Geistes- und Sozialwissenschaftlichen Klasse. 6).
G NOLI , R ANIERO: Il commento di Abhinavagupta alla Parātrim . śikā. (Parātrim

. śikā-
tattvavivaran.am) Traduzione e Testo. Roma: IsMEO 1985 (Serie Orientale Roma. LVIII).
148 Bibliography
G ÖBEL -G ROSS , E RHARD: Sirr-i akbar. Die persische Upanis.adenübersetzung des Mogul-
prinzen Dārā Šukoh. Eine Untersuchung der Übersetzungsmethode und Textauswahl nebst
Text der Praśna-Upanis.ad Sanskrit – Persisch – Deutsch. Marburg 1962 [Unveröf-
fentlichte Dissertation].
G OLZIO , K ARL -H EINZ: “Zur Datierung eines Kommentars zum Yogavāsis.t.ha von
Ānandabodhendra”. In: IIJ (∗ 2005) [forthcoming].
H ACKER , PAUL : “Jayantabhat. t.a und Vācaspatimiśra, ihre Zeit und ihre Bedeutung
für die Chronologie des Vedānta”. In: Beiträge zur indischen Philologie und Alter-
tumskunde. Walter Schubring zum 70. Geburtstag dargebracht von der deutschen
IndologieHamburg: Cram, de Gruyter (Alt- u. neuindische Studien. 7) 1951, pp.
160–169 [=Kleine Schriften, p. 110–119].
H ANNEDER , J ÜRGEN : Abhinavagupta’s Philosophy of Revelation. An Edition and Anno-

tated Translation of Mālinı̄ślokavārttika I, 1–399. Groningen: Egbert Forsten 1998a
(Groningen Oriental Series. 14).
H ANNEDER , J ÜRGEN : “Śaiva Tantric Material in the Yogavāsis.t.ha”. In: WZKS 42

(1998b): 67–76.
H ANNEDER , J ÜRGEN : “The Yogavāsis.t.ha and its Kashmirian recension, the Moks.opāya.
Notes on their Textual Quality”. In: WZKS 44 (2000): 183–210.
H ANNEDER , J ÜRGEN : “Studies in the Moks.opāya”. 2003 [Habilitation thesis, Ham-

burg].
H ANNEDER , J ÜRGEN : “Dreams and other States of Consciousness in the Moks.opāya”.

In: The Indian Night, Sleep and Dream in Indian Culture. Ed. N ALINI B ALBIR. Delhi:
Rupa and Co. ∗ 2006, [forthcoming].
H ANNEDER , J ÜRGEN and S LAJE , WALTER: Bhāskarakan.t.ha’s Moks.opāya-T.¯ıkā. A Com-

mentary on the Earliest Available Recension of the Yogavāsis.t.ha I. Vairāgyaprakaran. am.
Aachen: Shaker Verlag 2002 (Indologica Halensis. Geisteskultur Indiens. 1).
H ANNEDER , J ÜRGEN and S LAJE , WALTER: “Noch einmal zur längeren und kürzeren
Fassung des Yogavāsis.t.ha in ihrem Verhältnis zur Moks.opāya-Rezension”. In: AS
(∗ 2005) [forthcoming].
H INÜBER , O SKAR VON: Sprachentwicklung, und Kulturgeschichte. Ein Beitrag zur ma-
teriellen Kultur des buddhistischen Klosterlebens. Stuttgart: Steiner 1992 (Akademie der
Wissenschaften und der Literatur, Mainz. Abhandlungen der geistes- und sozialwis-
senschaftlichen Kl. 1992.6).
Appendix 149
H INÜBER , O SKAR VON: Die Palola-S.āhis, ihre Steininschriften, Inschriften auf Bronzen,
Handschriftenkolophone und Schutzzauber. Mainz 2004 (Antiquities of Northern Pak-
istan. 5).
. USAYN Ī , S AYYID A H. MAD: Fihrist-i nushahā-yi hat.t.¯ı – kitāb-hāna-yi

umūm¯ı h.azrat-i
H
Āyatullāh l-
Āz.am¯ı Naǧaf¯ı Mar
aš¯ı. Vol.14. Qum 1987

(1366 HŠ).
. USAYN Ī , S AYYID A H. MAD: Fihrist-i nushahā-yi hat.t.¯ı – kitāb-hāna-yi

umūm¯ı h.azrat-i
H

Āyatullāh l-
Āz.am¯ı Naǧaf¯ı Mar
aš¯ı. Vol.18. Qum 1989 (1368 HŠ).
I KARI , YASUKE : “Map of Ancient Tı̄rthas in Kashmir Valley”. In: A Study of the
N¯ılamata. Aspects of Hinduism in Ancient Kashmir. Ed. YASUKE I KARI. Kyoto: Institute
for Research in Humanities 1994, pp. 425–442.
I VANOW, W.: Concise descriptive Catalogue of the Persian Mss. in the Collection of the ASB.
Calcutta: Baptist Mission Press 1924.
J OSHI , L AXMANSHASTRI (Ed.): Descriptive Catalogue of Sanskrit Manuscripts [Prājña

Pāt.haśālā Man.d.ala Collection] Part II. Wai: Prājña Pāt.haśālā Man.d.ala no date.
K ANE , P.V.: History of Sanskrit Poetics. Delhi: Motilal Banarsidas 1987 [=4 1971].
K ANGLE , R.P.: The Kaut.il¯ıya Arthaśāstra. Part III. A Study. Delhi: Motilal Banarsidass
2000 reprint.
K HARE , G AN. E ŚA H ARIS ŪNU (Ed.): Bhārata Itihāsa Sam

. śodhaka Man.d.alastha Hastalikhita-
granthānukraman.ikā. Pune: Āryabhūs.an.a 1960.
K ÖNIG , D ITTE : Das Tor zur Unterwelt: Mythologie und Kult des Termitenhügels in der
schriftlichen und mündlichen Tradition Indiens. Stuttgart: Steiner-Verl.-Wiesbaden 1984
(Beiträge zur Südasienforschung. 97).
L E C OQ , A LBERT V.: Chotscho. Faksimile-Wiedergaben der wichtigeren Funde [. . . ]. Berlin:

D. Reimer 1913 (Ergebnisse der Kgl. Preussischen Turfan-Expeditionen).
L E C OQ , A LBERT V.: Die buddhistische Spätantike in Mittelasien. VI Neue [=] Bildwerke II

[. . . ] Wandgemälde aus Qyzil bei Kutscha. Berlin: D. Reimer 1928 (Ergebnisse der Kgl.
Preussischen Turfan-Expeditionen).
L EACH , L INDA Y ORK: Mughal and Other Indian Paintings from the Chester Beatty Library.
2 Bde. Dublin: Scorpion Cavendish 1995.
M AINKAR , T.G.: The Vāsis.t.ha Rāmāyan.a: A Study. New Delhi: Meharchand Lachman-
das 1977 [=1 1955].
150 Bibliography
M ASSON , J.L. and PATWARDHAN , M.V.: Śāntarasa and Abhinavagupta’s Philosophy of

Aesthetics. Poona: BORI 1985 (Bhandarkar Oriental Series. 9).
M EREDITH -O WENS , G.M.: Handlist of Persian Manuscripts 1895–1966 in the British Mu-
seum. London 1968.
M ITRA , R AJENDRA L AL : Notices of Sanskrit mss. Calcutta: Baptist Mission Press 1871ff.
M UJTABĀ ’ Ī , FATHULLĀH : “Muntakhab-i Jug-Basasht or Selections from the Yoga-

Vāsis.t.h.a, Attributed to Mı̄r Abu’l Qāsim Findiriskı̄. Critical edition of the Persian
text and translations into English with indroductory studies, notes, glossary and
index”. 1976 [Unpublished ph.D. thesis, Harvard University].
N ASR , S.H.: “Spiritual Movements, Philosophy and Theology in the Safavid Period”.
In: The Cambridge History of Iran. Ed. P ETER J ACKSON . Cambridge: Cambridge Uni-
versity Press 1986, pp. 675–677.
PANDEY, K ANTI C HANDRA : Abhinavagupta. An Historical and Philosophical Study.

Varanasi: Chowkhamba Sanskrit series Office 1963 (Chowkhamba Sanskrit Stud-
ies. 1).
P ERTSCH , W ILHELM : Die Handschriften-Verzeichnisse der königlichen Bibliothek zu Berlin.

Berlin: A. Asher & Co. 1888.
R AGHAVAN , V.: “The Date of the Yogavāsis.t.ha”. In: Journal of Oriental Research 13
(1939a): 110–128.
R AGHAVAN , V.: “The Yogavasistha Quotations in the Jivanmuktiviveka of Vid-

yaranya”. In: Journal of Andhra Historical Research Society 12 (1939b): 149–156.
R IEU , C HARLES: Catalogue of the Persian Manuscripts in the British Museum. London
1879–1881.
R ORTY, R ICHARD: Contingency, Irony, and Solidarity. Cambridge: University Press 1989.
S ACHAU , E.: Catalogue of the Persian, Turkish, Hindustani and Pushtu Mss. in the Bodleian
Library. Part I – Persian Manuscripts. Oxford 1889.
¯
S. AF Ā , ZAB ĪH. OLL ĀH : Tār¯ıh-i adabiyyāt dar Irān. Bd. V.1. Tehran: Entesharat-e Ferdous
–
8
1994 (1372 HŠ).
S CHRADER , F. O TTO: “[Collective Review of Atreya’s and Divanji’s Articles]”. In: OLZ
37.10 (1934): 641–644.
Appendix 151
S HARMA , S HRIRAM : A Descriptive Bibliography of Sanskrit Works in Persian. Hyderabad:

Abul Kalam Azad Oriental Research Institute 1982 [S. 9–13].
S HASTRI , S ATYA V RAT : “Un-Pān.inian Forms in the Yogavāsis.t.ha”. In: Vishveshvarand

Indological Journal 1 (1963): 247–266.
S HASTRI , S ATYA V RAT : “Notes on the Language of the Yogavāsis.t.ha”. In: ABORI 48–
49 (1968): 313–323.
S HASTRI , S ATYA V RAT : “Some Anomalies in the Language of the Yogavāsis.t.ha”. In:
Sanskrit and Indological Studies. Dr. V. Raghavan Felicitation Volume Delhi 1975,
pp. 325–329.
S HASTRI , S ATYA V RAT : “The Yogavāsis.t.ha — A study in Vocabulary”. In: Indologica

Taurinensia 7 (1979): 361–368.
S HASTRI , S ATYA V RAT : “Taddhita Formations in the Yogavāsis.t.ha”. In: Adyar Library
Bulletin 44–45 (1980–81): 320–332.
S KJÆRVØ , P RODS O KTOR: Khotanese Manuscripts from Chinese Turkestan in the British
Library. A Complete Catalogue with Texts and Translations. London: The British Li-
brary 2002 (Corpus Inscriptionum Iranicarum Part II: Inscriptions of the Seleucid
and Parthian Periods and of Eastern Iran and Central Asia. Vol. V: Saka. Texts VI).
S LAJE , WALTER: “A Guide to the Philosophical and Religious Terms in the

(Laghu-)Yogavāsis.t.ha”. In: WZKS 34 (1990): 147–177.
S LAJE , WALTER: Bhāskarakan.t.has Moks.opāyat.¯ıkā. 2. (Mumuks.uvyavahāra). Graz: Leykam

1993.
S LAJE , WALTER: Vom Moks.opāya-Śāstra zum Yogavāsis.t.ha-Mahārāmāyan.a. Philologische

Untersuchungen zur Entwicklung- und Überlieferungsgeschichte eines indischen Lehr-
werks mit Anspruch auf Heilsrelevanz. Wien: Verlag der Österreichischen Akademie
der Wissenschaften 1994 (Veröffentlichungen der Kommission für Sprachen und
Kulturen Südasiens. 27).
S LAJE , WALTER: Bhāskarakan.t.has Moks.opāya-T.¯ıkā. Die Fragmente des 3. (Utpatti-)

Prakaran.a. Graz: EWS Fachverlag 1995.
S LAJE , WALTER: “Zur Traditionsgeschichte der Vorstellung von einer ‘Erlösung noch
im Leben’ (j¯ıvanmukti)”. In: BEI 13–14 (1995–6): 387–413.
S LAJE , WALTER: Bhāskarakan.t.has Moks.opāya-T.¯ıkā. 1. (Vairāgya-Prakaran.a). Graz: EWS

Fachverlag 1996.
152 Bibliography
S LAJE , WALTER: “The Moks.opāya Project”. In: Annals of the Bhandarkar Oriental
Research Institute (1997): 209–221.
S LAJE , WALTER: “On Changing Others’ Ideas: The Case of Vidyāran.ya and the Yo-
gavāsis.t.ha”. In: IIJ 41 (1998): 103–124.
S LAJE , WALTER: “The Moks.opāya Project (II): New manuscripts and the present state
of things”. In: 125 Jahre Indogermanistik in Graz. Ed. M ICHAELA O FITSCH and C HRIS -
TIAN Z INKO . Leykam 2000, pp. 459–467.
S LAJE , WALTER: “Observations on the making of the Yogavāsis.t.ha (caitta, nañartha and
vah.)”. In: La Parole e i marmi. Ed. R AFFAELE T ORELLA . Rome: Istituto Italiano per
l’Africa e l’Oriente (Serie Orientale Roma. CXII, 1/2) 2001, pp. 771–796.
S LAJE , WALTER: Bhāskarakan.t.has Moks.opāya-T.¯ıkā. Die Fragmente des 4. (Sthiti-)

Prakaran.a. Aachen: Shaker Verlag 2002 (Indologica Halensis. Geisteskultur Indiens.
2).
S PROCKHOFF , J OACHIM F RIEDRICH: “Die Idee der Jı̄vanmukti in den späteren

Upanis.ads”. In: WZKS 7 (1963): 190–208.
S PROCKHOFF , J OACHIM F RIEDRICH: “Der Weg zur Erlösung bei Lebzeiten, ihr Wesen
und ihr Wert, nach dem Jı̄vanmuktiviveka des Vidyāran.ya”. In: WZKS 8 (1964):
224–262.
S PROCKHOFF , J OACHIM F RIEDRICH: “Der Weg zur Erlösung bei Lebzeiten, ihr We-
sen und ihr Wert, nach dem Jı̄vanmuktiviveka des Vidyāran.ya (Fortsetzung)”. In:
WZKS 14 (1970): 131–159.
S PROCKHOFF , J OACHIM F RIEDRICH: Sam . nyāsa. Quellenstudium zur Askese im Hinduis-

mus. I. Untersuchung über die Sam
. nyāsa-Upanis.ads. Wiesbaden: Franz Steiner 1976
(AKM. XLII,1).
S PROCKHOFF , J OACHIM F RIEDRICH: “Vom Umgang mit den Sam

. nyāsa-Upanis.ads”.
In: WZKS 34 (1990): 5–48.
S RINIVASAN , S RINIVASA AYYA : Vācaspatimiśras Tattvakaumud¯ı. Ein Beitrag zur Text-

kritik bei kontaminierter Überlieferung. Hamburg: De Gruyter & Co. 1967 (Alt- und
Neuindische Studien. 12).
S TEIN , M ARCUS A UREL : Kalhan.a’s Rājataraṅgin¯ı. A Chronicle of the Kings of Kaśm¯ır.

Transl., with an introd., comm. and app., 2 vols.. Westminster 1900.
Appendix 153
S TOREY, C.A.: Persian Literature. A Bio-Bibliographical Survey. Vol. 1. Qur’ānic Literature;

History. London: Luzac 1927-1939 [Repr. 1989].
TA ṄGASVĀMI ŚARMAN : Advaitavedāntasiddhāntetihāsakośah. . Madras 1980.
T HOMI , P ETER: Yogavāsis.t.ha. Die Lehren des Weisen Vāsis.t.ha. Nach der Version des
Vāsis.t.ha-Sam
. graha aus dem Sanskrit übersetzt von Peter Thomi. Interlaken: Ansata Ver-
lag 1988a.
T HOMI , P ETER: Yogavāsis.t.ha. Die Lehren des Weisen Vāsis.t.ha. Nach der Version des
Vāsis.t.ha-Sam
. graha aus dem Sanskrit übersetzt von Peter Thomi. Band 2. Text und Kon-
kordanz. Wichtrach: Institut für Indologie 1988b.
T HOMI , P ETER: Br.had-Yogavāsis.t.ha. Pāda-Index. Wichtrach: Institut für Indologie 1990.
T HOMI , P ETER: Yogavāsis.t.hasāra. Die Quintessenz des Yogavāsis.t.ha. I: Sanskrit-Text. II:

Übersetzung und Kommentar. Wichtrach: Institut für Indologie 1999.
T OKUNAGA , M UNEO: “Description of Temples and Tı̄rthas in the Nı̄lamatapurān.a.

Vss. 989–1356a”. In: A Study of the N¯ılamata. Aspects of Hinduism in Ancient Kashmir.
Ed. YASUKE I KARI. Kyoto: Institute for Research in Humanities 1994, pp. 399–421.
¯
T ORELLA , R AFFAELE : The Iśvarapratyabhijñākārikā of Utpaladeva with the Author’s Vr.tti.
Roma: Is. M. E. O. 1994 (Serie Orientale Rome. Vol. LXXI).
VATTIMO , G IANNI: “Interpretazione”. In: Enciclopedia di filosofia. Milano: Garzanti

1981, pp. 450–451.
W ITZEL , M ICHAEL : “On Indian Historical Writing: The Role of Vam . śāvalı̄s”. In: Jour-
nal of the Japanese Association of South Asian Studies 2 (1990): 1–57.
W ITZEL , M ICHAEL : “The Brahmins of Kashmir”. In: A Study of the N¯ılamata. Aspects
of Hinduism in Ancient Kashmir. Ed. YASUKE I KARI. Kyoto: Institute for Research in
Humanities 1994a, pp. 237–294.
W ITZEL , M ICHAEL : “Kashmiri Manuscripts and Pronounciation”. In: A Study of the

N¯ılamata. Aspects of Hinduism in Ancient Kashmir. Ed. YASUKE I KARI. Kyoto: Institute
for Research in Humanities 1994b, pp. 1–53.
Index
Abhimanyu, 22 Farmulı̄, 97, 116, 125
Abhinanda, 95
abhraṅkas.a, 30 Ǧahāngı̄r, 119
Adhis.t.hāna, 22 Galūna, 32
Kashmirian capital, 26 Gaud.a-Abhinanda, 103
Advaitabrahmasiddhi, 21 Göttingen ms., 105
ahaṅkāravāsanā, 26 Ǧōg Bāšištha, 124
Al-Bı̄rūnı̄, 26 Ǧōg Bašist, 118
Amr.takunda, 114 Gopālavarman, 30
Ānandabodhendra, 9 Gurupraśasti, 99
date, 25
Arjunopākhyāna, 21 H. abı̄bullāh, 117
Ars.yaśr.ṅga, 27 hāra (Kashmiri), 29
As.t.āṅgayoga, 95 Hariharārya, 99
as.t.astambhaman.d.apa Hariparbat, 29
Yaśakara’s palace, 31 himakan.a, 21
Ātmasukha, 96 Hulāsa, 122

Janaka
Banwālı̄ Dās, 117
image of, 143
Bhagavadg¯ıtā, 21
story of, 106
Bhās.āyogavāsis.t.hasāra, 50
Jayendra-Vihāra, 32
Bhāskarakan.t.ha, 16, 25
Jhelum, 27
Bhr.ṅg¯ıśa, 110
J¯ıvanmuktiviveka, 95, 100
bhūrjatvac, 21
Jñānavāsis.t.ha
Bhusun.d.a, 110
area of distribution, 54
birch bark, 21
Buddhist monk, 23 Kat.a
reborn as a partridge, 26
Cı̄na, 22 Kathāsaritsāgara, 28
Dāma Laghuyogavāsis.t.ha
reborn as a gnat, 26 indo-persian versions, 94
Dāma, Vyāla, Kat.a, 25 mss. of Persian translations, 115
Dārā Šikōh, 116 area of distribution, 54
D IVANJI , P.C., 10 author, 94, 102
Duperron, Anquetil, 113 commentaries, 9, 96
154
Appendix 155
date, 94 method of abridgement, 106

manuscript transmission, 98 Moks.opāyasāra, 15
manuscripts, 99 Mummad.ideva Allād.asūnu, 96
quality of the printed text, 98 Muntahab, 122
L¯ılopākhyāna, 108
Nāgarnagar, 28
Mahāpadma (nāga), 26 Narasim . ha, 33
Mahāpadmasaras (lake), 25 N¯ılamatapurān.a, 26
Mahı̄dhara, 39 Nirvān.aprakaran.a, 56, 110
Mı̄r Abū ‘l-Qāsim Findiriskı̄, 121 Nr.sim. ha, 33
mithyāpurus.a, 110
Pāndret.hān, 27
Moks.opāya
Pānı̄patı̄, 97, 116, 119
3.127.1, 85
Parvagupta
3.60.16, 81
minister of Yaśakara, 33
3.60.30, 87
Pās.ān.ākhyāna, 112
3.60.63, 81
Pāśupatas, 29
3.61.35, 80
Prabhākaradeva
3.62.34, 81
father of Yaśaskara, 29
3.66.12–14, 12
Prabodhacandrodaya, 117
3.67.40, 86
Pradyumnaśikhara, 27
3.80.5, 81
Pravara(sena)pura, 26
4.32.11, 26
Pravarasena II, 26
4.32.12ab, 27 Purān.ādhis.t.hāna, 27
4.32.13, 30 Pūrn.amanoratha
4.32.16, 31 minister of Yaśaskara, 34
4.32.18, 32
4.32.19, 32 Rājataraṅgin.¯ı
4.32.31c, 32 of Śrı̄vara, 96
6.66.8ab, 22 Rāma
6.70.10c–11b, 24 his enlightenment, 112
6.70.7bc, 22 Rāmabrahmendra, 96
6.70.8cd, 23 Rāmacarita, 95, 103
6.70.9a–c, 23 Rāmāyan.a, 21
condensed versions, 14 Ran.āditya, 29
criteria for identification, 42, 75 Ratnāvalı̄vihāra, 31, 32
extension of the Sthitipr., 84 red-haired, 23
manuscripts, 76, 139
regional orthography, 79 Sadānanda Yati, 21
style, 83 Śaivism of Kashmir, 11
Moks.opāyasaṅgraha, 48, 54, 105 Salı̄m, 116
156 Index
Sam . sārataran.i, 96 Vipaścit

Sanskrit-Persian Translations story of, 112
research on, 113 Vitastā, 27
Sāra-Sarga, 99 Vular/Volur, 25
Śārikā Vyāla
embodied in a red stone, 29 reborn as a sparrow, 26
her worship, 29
hill, 27 Walı̄ Rām, 117
story of the bird, 28
the goddess, 28 Xuanzang, 27
Śārikā hill
= Pradyumnaśikhara, 28 Yaśaskaradeva, 22
entrance to the underworld, 29 reign, 24
Śārı̄t.aka, 28 royal residence, 27
Sarmad Kāšānı̄, 121 Yogavāsis.t.ha
Śavākhyāna, 112 commentary on, 9
Śivākhyāna, 110 editio princeps, 9
Somadeva, 28 influence on MU-mss, 53, 77
Śr¯ıkan.t.hacarita, 26 Vedāntic influence, 15
Śrı̄nagar, 22 Yogavāsis.t.hasāra
Śr¯ıvara, 97 original title, 39
Śr¯ıvāsis.t.hacandrikā, 91, 92, 100 yukti, 131
Śuka
story of, 106
termite’s nest, 23
three demons
story of, 25
Tuh.fah-i maǧlis, 121
Vairāgyaprakaran. a
development, 53
Vallabha
minister of Yaśaskara, 34
valm¯ıka, 23
Vāsis.t.hacandrikā, 15, 96
Vāsis.t.hasaṅgraha, 15, 91
Vāsis.t.hasāra, 14, 39, 91
Vāsis.t.hatātparyaprakāśa, 9, 55
Vāsis.t.hatattvabodhin¯ı, 96
Vidyāran.ya, 95

Geisteskultur Indiens
Texte und Studien
GI 1 H ANNEDER , J ÜRGEN and S LAJE , WALTER: Bhāskarakan.t.ha’s

Moks.opāyat.ı̄kā I. Vairāgyaprakaran.a. Revised Edition in Devanāgarı̄
Script. 2002. ISBN 3-8322-0217-X. 35,80 €
GI 2 S LAJE , WALTER: Bhāskarakan.t.has Moks.opāya-T.ı̄kā. Die Fragmente des
4. (Sthiti-) Prakaran.a. 2002. ISBN 3-8322-0691-4. 45,80 €
GI 3 S TEPHAN , P ETER: Erlösung im Spannungsfeld von aktivem Leben und
Entsagung. Eine Studie zu Śaṅkaras Exegese der Bhagavadgı̄tā. 2002.
ISBN 3-8322-0884-4. 30,80 €
GI 6 H INÜBER , O SKAR VON: Indiens Weg in die Moderne. Geschichte und
Kultur im 19. und 20. Jahrhundert. 2005. ISBN 3-8322-3647-3. 35,80 €
GI 7 The Moks.opāya, Yogavāsis.t.ha and Related Texts. Ed. by J ÜRGEN
H ANNEDER 2005. ISBN 3-8322-4265-1.
Klassiker der Indologie
GI 4.1 F RAUWALLNER , E RICH: Geschichte der indischen Philosophie. I.

Band. Die Philosophie des Veda und des Epos. Der Buddha und der
Jina. Das Sām
. khya und das klassische Yoga-System. Herausgegeben
von A NDREAS P OHLUS. 2003. ISBN 3-8322-1076-8. 25,00 €
GI 4.2 F RAUWALLNER , E RICH: Geschichte der indischen Philosophie. II.
Band. Die naturphilosophischen Schulen und das Vaiśes.ika-System.
Das System der Jaina. Der Materialismus. Herausgegeben von A N -
DREAS P OHLUS . 2003. ISBN 3-8322-2226-X. 22,80 €
GI 5 S TRAUSS , O TTO: Indische Philosophie. Herausgegeben von A NDREAS

P OHLUS. 2004. ISBN 3-8322-3107-2. 22,80 €

Moksopaya, Yogavasishtha and Related Texts, J. Hanneder

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Moksopaya, Yogavasishtha and Related Texts, J. Hanneder

Загружено:

Авторское право:

Доступные форматы

T HE M OKS. OP ĀYA , Y OGAV ĀSIS. T.

HERAUSGEGEBEN VON WALTER SLAJE

Copyright Shaker Verlag 2005

Shaker Verlag GmbH • Postfach 101818 • 52018 Aachen

The Yogavāsis.t.ha or – as it is called in its oldest version – Moks.opāya is a

fragment of Bhāskarakan.t.ha’s commentary on the Nirvān.aprakaran.a is being

mentary3 is quoted. In all other cases, notably that of the Nirvān.aprakaran.a

The Moks.opāya: An Introduction

Locating the Moks.opāya

The Moks.opāya Project III:

The Moks.opāya Project IV:

The Critical Edition of the Utpatti-Prakaran.a: A Research Report

Die persischen Übersetzungen des Laghuyogavāsis..tha

The Metaphorical Logic of the Moks.opāya

Although it is certainly wise not to reject established knowledge prematurely,

cidghanaikaprapātasya rūd.hasya parame pade

The commentator Ānandabodhendra explains that the genitives refer to

cidghanenaikatām etya yadā tis.t.hasi niścalah.

If you, having become one with the mass of consciousness, remain

he may explain to the astonished reader that a ‘bhiks.u’ in a ‘vihāra’16 is not a

complete manuscript has 116 fols.

The determination of the terminus ante quem is based on quotations and

41 See below, p. 24.

(1990), p. 151, fn. 32; S LAJE (1994), p. 172ff.

c¯ınanāmātha tatrāsti śr¯ımāñ janapado mahān (6.70.8ab Ś1 , ∼ YV 6.66.8ab)

“Now, in this [northern direction] there is the big [and] splendid

A Central Asian itinerary,5 drawn up during king Abhimanyu’s rule (22nd

That the detailed descriptions of the author of the Moks.opāya cannot be

. śrayah. (6.70.8cd Ś1 , ∼ YV 6.66.8cd)

“In this [country] there is a Buddhist monastery on a termites’ nest,

tasmin vihāre svakut.¯ıkośe kapilamūrdhajah. | bhiks.ur [. . . ] (6.70.9a–c Ś1 , YV

dr.d.hārgalam. gr.ham. dhyānabhaṅgabh¯ıtā viśanti no

“Being afraid of interrupting [his] contemplation, the servants would

Mural paintings with portraitures of red-haired and green- or blue-eyed

(1992), p. 18f; 22ff; 30–34.

with a prophecy, in its entirety,20 it becomes difficult to advance objective

inadequate notions of possessing an individual identity (ahaṅkāravāsanā).

as well. Pravarasena’s foundation, first described by the Chinese pilgrim Xu-

“In the middle of this [town] [. . . ] a lofty peak called Pradyum-

Pradyumnaśikhara is another name for Śārikāparvata. In the tenth cen-

in about AD 580. He gives a legendary account (RT 3.339–349), accord-

view to having the hill guarded, as it was considered an important entrance

tasya mūrdhni girer geham . ko ’pi rājā karis.yati

tasminn eva tadā kāle tatra rājā bhavis.yati

“Then, at the same time, a king of royal dignity (śr¯ı), Yaśaskaradeva

Yaśaskaradeva, we know from Kalhan.a, apart from his royal residence

adhis.t.hānābhidhe tasminn evogranagare tadā

“Then, in the same mighty town called Adhis.t.hāna, there will be a

Kalhan.a, too, mentions the Ratnāvalı̄-Vihāra of Śrı̄nagar (Rājataraṅgin.¯ı

“In this [Vihāra] there will live a companion (/councillor)56 of that

As a result, we now have come to know of altogether four medieval ed-

1.762c-764b: ittham . pratyaks. am evedam. nih.sapatnam

murdered the crippled successor of Yaśaskara, and thereupon usurped the

65 See Rājataraṅgin. ı̄ 5.420ff; 6.115–149.

In continuation of two reports on ‘The Moks.opāya Project’, published else-

National Archives, New Delhi

exactly as found. Scribal mistakes are not marked.

Catalogue entry: “Moks.opāya, 90 folios.”

No. 178 Paper, cardboard cover, Śāradā.

yāvan nānugrahah. sāks.āj jāyate parameśvarāt |

kathopāyān vicāryādau moks.opān imān atha |

asmin rāmāyan.e . . . || 5|| (= MU 1.1.4)

Breaks off on folio 95v :

This abridgement in its initial part (Vairāgya) appears to be dependent on,