Chapter

ch6_ratie_other_streams

10966 words · vendored research asset

source →

me Hitherto Unknown Fragments of Utpaladeva’s Vivr̥ti (I)

the Buddhist Controversy Over the

istence of Other Conscious Streams*

abelle Ratié

Raffaele Torella,

ns qui les œuvres d’Utpaladeva

raient demeurées à maints égards apratyabhijñāta.

me Old Sources for New Vivr̥ti Fragments

dology owes to Raffaele Torella one of the most exciting manuscript
scoveries made in the last decades, namely, that of the only extensive
agment thus far known of Utpaladeva’s Vivr̥ti on his ĪPK. The edition
d translation of this fragmentary codex unicus by its discoverer1 has
d several important consequences. First of all, it has confirmed
ffaele Torella’s initial hypothesis that the great innovator as
gards the Pratyabhijñā system is not Abhinavagupta but Utpaladeva:2 a
ick comparison between the extant fragment of the Vivr̥ti and the
rallel passages in Abhinavagupta’s ĪPV and ĪPVV shows that at least in
is part of the treatise, Abhinavagupta is extremely faithful to
paladeva’s thought and closely follows his argumentation. But there is
other great advantage to this discovery: we are now able to examine
e marginal annotations contained in the manuscripts of Abhinavagupta’s
V3 and to check whether some of them contain quotations from the
vr̥ti. And this is a task of the utmost importance because, so far,
art from the fragment discovered by Raffaele Torella, we could only
ly on Abhinavagupta’s commentaries so as to try and reconstruct the
st of the arguments in the Vivr̥ti.

mittedly, Abhinavagupta’s ĪPVV in particular constitutes a very rich
urce of information regarding Utpaladeva’s detailed commentary on his
n verses: although the ĪPVV most often quotes only one or two words of
e Vivr̥ti while commenting on it, not only does Abhinavagupta expound
Utpaladeva’s reasonings: he sometimes gives lengthier quotations as
ll, although these are not always found where we would expect them to
. Thus as I have shown elsewhere, the first concluding verse of
paladeva’s Vivr̥ti can be retrieved from Abhinavagupta’s ĪPV ad ĪPK
4.3.4

t now that we are in possession of a lengthy fragment of the Vivr̥ti,
can be certain that we have at our disposal another crucial source of
formation on this text besides Abhinavagupta’s commentaries, since a
mparison between the text edited by Raffaele Torella and the
notations found in the margins of ĪPV manuscripts shows that some of
ese annotations are in fact quotations of Utpaladeva’s Vivr̥ti.

is is the case of some of the annotations in manuscript D2, which, as
ffaele Torella has already noticed, was in all probability one of the
ur manuscripts used by the editors of the Kashmir Series of Texts and
udies (KSTS).5 But D2 is not an exception: Yohei Kawajiri has been
udying a Devanāgarī manuscript of the ĪPV preserved in Lucknow that
so contains such passages in its margins,6 and among the manuscripts
at I have been able to consult, it is also the case of the ĪPV
nuscript preserved in the School of Oriental and African Studies
brary (SOAS), and to a greater extent, of one of the manuscripts
eserved in the Srinagar Oriental Research Library (S3). Thus D2, SOAS
d S3 share at least two quotations belonging to the part of the Vivr̥ti
ited by Raffaele Torella (and appearing as well in footnotes of the
shmir Text and Studies edition of the ĪPV).7 I could not find in S3 a
mber of quotations that are present in the margins of both D2 and
AS,8 while S3 gives quotations (sometimes significantly lengthier than
ose found in D2 and SOAS) that are not found in the first two
nuscripts,9 which seems to indicate that if D2 and SOAS are very
obably related, the annotations in S3 might have an independent
igin.

w, since we are sure that these manuscripts bear quotations of the
xt found in the incomplete manuscript of the Vivr̥ti discovered by
ffaele Torella, we may quite safely surmise that they also contain
otations of hitherto unknown parts of the Vivr̥ti. I think that it is
doubtedly the case of S3. Of course, for now – that is, as long as no
mplete manuscript of the Vivr̥ti is discovered –, we cannot always be
solutely sure that the passages quoted do belong to Utpaladeva’s lost
xt: the author of the marginal annotations in S3 sometimes specifies
at he is indeed quoting the Vivr̥ti,10 but alas he does not do so
stematically. However, here Abhinavagupta’s commentary is once more an
tremely precious source: while commenting on Utpaladeva’s lost text,
regularly quotes some of its words, so that the probability for the
rginal annotations to be taken from the Vivr̥ti is high enough to
nfine to certainty if the words quoted in the ĪPVV appear in the
rginal annotations in the same sequence as that found in
hinavagupta’s commentary. This means that we now have at our disposal
mine (or rather several mines) of fragments of Utpaladeva’s lost work,
d that thanks to Raffaele Torella’s pioneering works, we may be able
retrieve many passages of the Vivr̥ti even if no complete manuscript
the text comes to light in the near future. The following pages,
ich present the first results of an ongoing investigation of the
rginal annotations found in S3,11 offer an edition and translation of
passage of the Vivr̥ti quoted in the margins of S3 and devoted to the
planation of ĪPK 1.5.4-5. Both the edition and the translation are
ntative and I still have many doubts as to the meaning (or adequate
ading) of a number of passages. Nonetheless it seems to me that given
e present extent of our knowledge regarding the Vivr̥ti, an imperfect
tempt at editing and translating any fragment of Utpaladeva’s lost
rk is better than no attempt at all, and I am confident that my
stakes will soon be corrected by scholars more learned than me.

e Context of the Fragment ad ĪPK 1.5.4-5:

e Debate between a Vijñānavādin and a Sautrāntika

d the Problem of the Other Conscious Streams

ĪPK 1.5.4-5,12 Utpaladeva lets a Sautrāntika criticize the
jñānavādins’ idealistic thesis that the universe is merely an internal
rm taken on by consciousness and that phenomenal variety is the result
an internal mechanism of latent traces or imprints (vāsanā). In the
urse of this criticism, not only does the Sautrāntika attack the
jñānavādins’ position regarding the status of the external object:13
also argues that the Vijñānavāda can only lead to solipsism, since in
e Vijñānavādin’s perspective, other subjects must be mere internal
pects of my own stream or series (santāna) of momentary cognitive
ents. I have examined elsewhere the details of this controversy,14 so
re I shall only give a basic summary of it.

e Vijñānavādin denies the accusation of solipsism by using
armakīrti’s line of argument in the Santānāntarasiddhi.15 According to
armakīrti, just as in a system that admits the reality of entities
ternal to consciousness, in the idealistic system of the Vijñānavāda,
e other streams of consciousness cannot be directly perceived (since a
nsciousness cannot be directly apprehended by another consciousness
t only by itself); yet, just as in a system that accepts the reality
external entities, in the Vijñānavāda, the other streams of
nsciousness can be inferred. This inference of the existence of other
nscious entities or streams of momentary cognitions rests on the fact
at we experience in ourselves that when we act (and for instance when
speak), our action is preceded by a will to act. So when we witness
action that has not been preceded by our own will to act, we can
fer that this action was preceded by a desire to act that does not
long to our own stream of cognitions – and this other willing entity
st be another conscious subject.

this argument the Sautrāntika replies that the Vijñānavādin’s own
esuppositions prevent him from making the very inference that he
fends. He thus shows that the Vijñānavādin has no means of
tablishing the invariable concomitance on which his inference of other
nscious streams rests. For indeed, I can observe an invariable
ncomitance between my own speech and my previous desire to speak, so
at the logical pervasion (vyāpti) between “I want to speak” and “I am
eaking” is established; but in the case where I hear a speech that is
t uttered by me, my thought does not take the form “I am speaking” but
e form “he/she is speaking”. Now, the Vijñānavādin has no way of
tablishing any invariable concomitance between the thought “I want to
eak” (in someone else’s conscious stream) and the thought “X is
eaking” (in my conscious stream when I hear someone else speaking).
e two propositions “I want to speak” and “I am speaking” are
sentially linked to the subject, or they are not “separated”
icchinna) from the “I”, whereas the proposition “X is speaking” is
parated from the “I” since it is expressed by using the third
ammatical person, and we have no way of determining that there is a
usal relationship (kāryakāraṇabhāva) between somebody else
bjectively thinking “I want to speak” and our own objective16 thought
is speaking”, so that the cause of the phenomenon “X is speaking”
mains perfectly unknown.

e Vijñānavādin replies that in fact this necessary relation between
e subjective cognition “I want to speak” and the objective cognition
is speaking” is constantly apprehended in the ordinary world, because
r actions are always related to our awareness that they can be
tnessed by others: when we act, we act while desiring either to be
en acting or not to be seen acting, but whether we conceal our actions
whether we act ostentatiously, we are aware that our desire to act
es not merely result in the cognition “I am acting” (which occurs in
conscious stream) but also in the cognition “X is acting” (which
curs in the other conscious streams witnessing my action). And
cording to the Vijñānavādin, it is thanks to this understanding of the
variable concomitance between my subjective desire to act and the fact
at my action becomes an object of cognition for the others that the
ference of other conscious streams is possible.

e following passage of the Vivr̥ti begins with the Vijñānavādin’s
planation of this very argument, after which Utpaladeva (or, as
hinavagupta says, the Īśvarādvayavādin, the “proponent of the thesis
the non-duality [of the universe] with the Lord”) helps the
utrāntika criticize the Vijñānavādin’s position. Utpaladeva points out
at this position entails a logical circle, because our awareness of
e existence of other conscious entities is supposed to stem from our
areness that the others can witness our actions, while our awareness
at the others can witness our actions must derive from our awareness
the existence of other conscious entities. He also points out that in
ct there is no such thing as an invariable concomitance between the
bjective thought “I want to be heard by others” and the objective
ought “X is speaking”, because we sometimes experience that we can
eak without being heard whether we want it or not: thus the causality
lationship invoked by the Vijñānavādin between a subjective will to
t and the objective awareness that X is acting cannot be established.

e Śaiva non-dualist concludes from this that the other subject can
ly be established when the other’s action is perceived both by him/her
n the form “I want to speak/I am speaking”) and by me (in the form
e/she is speaking”), from the supposition that the same action is then
prehended in these two different forms. For we have to make this
sumption in order to justify the unity of the sphere of human practice
yavahāra): if we do not accept it, then we cannot justify our belief
at the other subjects share with us the same spatio-temporal
terminations and therefore live in a common world, and the notion of
tersubjectivity collapses.

Abhinavagupta makes clear in his commentary, the Śaiva non-dualists
ve a solution to this problem of intersubjectivity:17 according to
em, the unity of the sphere of human practice is to be explained as
e result of the constant activity of division and synthesis through
ich the universal consciousness appears to itself in the form of
veral distinct subjects and yet is capable of unifying these subjects
terms of spatial and temporal determinations, because although it
nifests itself as plural, it remains fundamentally one, just as a
eaming consciousness is capable of dividing itself into a multiplicity
characters without losing its unity. Recognizing the others is
erefore understood by Utpaladeva as a partial “recognition of the
rd” (īśvarapratyabhijñā), i.e. as the incomplete realization of the
namism inherent in consciousness: when I witness an activity that is
t mine in the world, I recognize in that activity a freedom that is
thing but the essence of my own subjectivity, so that my awareness of
e existence of other subjects is not the result of an inference based
causality (“another conscious entity must have produced this
tion”): rather, it arises from an inference based on the identity of
ture (svabhāva) between the freedom to act and consciousness.

netheless, in the Vivr̥ti fragment below, this solution remains mostly
plicit,18 because the Śaiva non-dualist is almost immediately
terrupted (at least if we follow Abhinavagupta’s commentary)19 by the
utrāntika, who takes advantage (or at least thinks he does so) of
paladeva’s criticism of the Vijñānavāda so as to put forward his own
sition. He agrees with the Śaiva non-dualist that the Vijñānavādin
ils to explain the spatio-temporal determination (avadhāra) of the
her subjects (a reproach that Vasubandhu had already faced in his
ṁśikā)20 and he points out that the only option left to the
jñānavādin trying to infer the existence of other subjects is to admit
at the thought formulated as “I want to act/I am acting” in one given
nscious stream projects an image of itself in another conscious stream
n image that takes the form “he/she is acting”): the other subject
st be inferred as the reflected object (bimba) producing in my
nscious stream the reflection (pratibimba) “he/she is acting”, just as
particular reflection in a mirror presupposes the existence of a
flected object which exists outside of the mirror and gives it its
rticular shape.

first sight, this theory must be appealing to the Vijñānavādin
ponent as it seems to fit his idealism: if he adopts it, he can admit
e existence of several conscious streams projecting reflections of
emselves onto each other without conceding the existence of external
jects. As a matter of fact, there is evidence in several Yogācāra
xts that at least some Buddhist idealists held this thesis and used it
as to explain the fact that several subjects seem to perceive the
me objects: according to those texts, the objects apprehended by the
nses must be reflecting an “original” object in the ālayavijñāna,21
d this “original” object is in turn a reflection of images occurring
other conscious streams.22 This idea that the phenomenal variety
perienced in one conscious stream is the result of the influence of
other conscious stream rather than of an external object is found in
rticular in Vasubandhu’s Viṁśikā23 (demonstrably known of Utpaladeva
d Abhinavagupta).24

course, such an explanation is not without danger for the
jñānavādin, since as the Sautrāntika points out in Abhinavagupta’s
V, acknowledging the existence of entities distinct from my own
gnitive stream (be they other conscious streams) seems to pave the way
r externalism.25 But whereas in the ĪPV, the Sautrāntika apparently
rgets in a general way any idealistic thesis that concedes the
istence of a plurality of conscious streams while denying the reality
anything external to them, in the fragment presented here, he seems
have in mind, more specifically, the idealistic contention that
enomenal variety can be explained as a result of the mutual influence
various conscious streams; and he shows that such a thesis leads to
absurd consequence, since each conscious stream should then perceive
infinite number of objects, just as (according to Abhinavagupta’s
age) several mirrors reflecting each other produce an infinite number
reflections.

e fragment ends on an objection made by the Vijñānavādin. The latter
adily accepts that all the other manifestations contained in other
nscious streams do not project reflections of themselves onto my
nscious stream (he agrees that this would lead to an absurd
nsequence), but he argues that some of these manifestations in other
nscious streams awaken a particular imprint (vāsanā) already contained
my own conscious stream and left by previous experiences. He thus
es the theory of the vāsanā mechanism to explain why we only perceive
limited number of objects.

fortunately, the end of the discussion (won by the Sautrāntika, at
ast in this part of the treatise) is missing. This is all the more
grettable since in the corresponding passage of the ĪPVV,
hinavagupta alludes to a very interesting discussion on ālayavijñāna26
an important topic for the historian of Buddhist doctrines, and one
at is not found in Utpaladeva’s ĪPK and Vr̥tti, nor in Abhinavagupta’s
V, but that may well have been treated in some detail in Utpaladeva’s
vr̥ti.

netheless, the fragment as it stands is of great interest, not only
cause it confirms once again how faithful Abhinavagupta’s ĪPVV is to
e text on which it comments (all the ideas mentioned in the
rresponding passage of the ĪPVV are to be found in the Vivr̥ti), or
cause it confirms interpretations of the ĪPV that were so far only
sed on the ĪPVV,27 but also because it illustrates Utpaladeva’s
timate acquaintance with Buddhist theories in general,28 and his
terest for the Buddhist debate over the epistemological (and/or
tological)29 status of other streams of consciousness in particular.
e reason for this interest probably lies in the fact that Utpaladeva’s
stem is a variety of absolute idealism: according to him, the universe
nothing but a manifestation of a single and all-powerful
nsciousness out of which nothing exists. He therefore shares with the
jñānavāda the burden of dealing with the accusation of solipsism, and
is all the more attentive to this problem since he wants to offer a
lution capable of superseding both that of the Vijñānavādins and that
their externalist opponents, the Sautrāntikas. His strategy, here as
ten, consists in letting the Sautrāntika triumph over the Vijñānavādin
fore showing (as he will in the following verses) that the Sautrāntika
ew regarding both objects and other subjects is untenable.30 His
urce regarding the Vijñānavādins’ position in this fragment is in all
obability Dharmakīrti’s Santānāntarasiddhi (a treatise that was read
the Śaiva philosophers, whether dualist or non-dualist).31 As for his
urce regarding the Sautrāntika opponent, it is difficult to judge to
at extent it is borrowed from externalist objections answered by
armakīrti in the Santānāntarasiddhi, and one might also suspect that
paladeva is inventing a Sautrāntika position rather than depicting a
al historical position.32 However, it seems to me that Utpaladeva’s
rvapakṣas are hardly ever (if at all) figments of his imagination, and
at on the contrary, they show a precise knowledge of his opponents’
eories, as can be verified in all instances where his particular
urce can be determined and has come down to us. In this regard, it is
rth noting that Abhinavagupta mentions in his commentary on the
ssing end of this fragment a lost work by Dharmottara, the
hyārthasiddhi:33 for now at least, it seems to me that we cannot rule
t the possibility that the latter text was one of Utpaladeva’s sources
f not his main source) for his depiction of the Sautrāntikas’ position
this debate.

Fragment of Utpaladeva’s Vivr̥ti ad ĪPK 1.5.4-534

nu yatra parajñānaviṣayābhāvāpattivāraṇena nijaṣṭhīvanādiceṣṭām
chet, tatra paṭādyapavāraṇādinaiva †. . .†35, yatra punar anyathaiva
thā ceṣṭate, tatra parajñānaviṣayabhāvagamanābhisandhir asty eva.

d etan mandamatibhāṣitaṁ parapramātr̥pravibhāgapratyayasya vyutpatteḥ.
thābhūtavyāhārādy-36avabhāsenaiva hi pramātrantaraniścayaḥ syāt, tad
ad itaretarāśrayatvam, vyāhārādy-37avabhāsaviśeṣaiḥ
amātrantaragatiḥ, tadgatyā ca
jjñānagatavyāhārādyavabhāsajananāyābhisandher viśiṣṭād
āhārādyavabhāsaviśeṣāḥ. yadi hi38 paratrotpādanābhisandheḥ
ratrotpannanr̥ttaṣṭhīvanodgrāhaṇādyavabhāsānāṁ svātmano vicchinnānāṁ
amāṇāntarāt kāryakāraṇabhāvo’vagataḥ syāt, tat svātmany api
cchinnaṣṭhīvanādyavabhāsair anyair apy anyadā pramātrantaram
umīyetāpi, yāvatānyataḥ parapramātary avijñāte
dgataṣṭhīvanādyābhāseṣv anavagateṣu kathaṁ svātmagatair api tair
cchinnair ābhāsaiḥ parasantānānumānam.

ca paro māṁ paśyatv ity abhisandhānāt parasantāne
cchinnavyāhārādyavabhāsotpattir yuktā, tathābhisandhāne mā drākṣīd iti
viparītābhisandhāne tadanutpattiprasaṅgāt.

d eva[ṁ] tad buddhisamīhāyuktapramātravicchinnāt
tpratyakṣatānuṣaṅgina eva yathāvadhāritotpādakra[ma]m avasthitād
āhārādyavabhāsāt siddhyet pramātrantaraṁ yathāsmābhir vakṣyate.

aṁ ca viśiṣṭadeśakālatāpi tasya tadāveśāt sidhyati, [a]nyathā39
thāvabhātadeśasthenaiva pramātrantareṇa paratrāsau janyeteti kuto
ścayaḥ.

āvat punaḥ syāt, buddhisamīhājanita-svapratyakṣavyāhārādyavabhāsād
padyate parapramātr̥ṣu vicchinnas tathāvabhāsaḥ pratibimbam iva, tataś
tadutpādakabimbabhūtavyāhārāvabhāsānumānamukhena parapramātānumīyeta.

y evaṁ ca ghaṭapaṭādyavabhāsā api kvacana krimisarvajñādipramātrantare
hitās tatsadr̥śābhāsa-40 sahabhāvinaḥ pramātrantarasyāpy avabhāsāñ
nayeyuḥ, tathā ca bāhyārthavādaprakāra eva syāt, jñānabāhyānām
hāsabhede hetutvābhyupagamaviśeṣaḥ, yato’vabhāsānāṁ41
asadr̥śāvabhāsāntarajanakatve pramātr̥ṣv ābhāsāparyavasānaṁ42 syāt, tair
y anya[syā]nyasya jananād iti.

hāpi parapramātr̥gatāvabhāsās tathāvabhāsajanikāṁ vāsanām eva
abodhayanti, na tu svato janayanti yenābhāsānantyaṁ syāt. vāsanā hy
abhāsāntaraśate nādhipatitvocitaikaiveti43 pratipramātrekam eva
davabhāsaṁ janayatīti nātiprasaṅgaḥ.

Tentative Translation

– Objection from the Vijñānavādin:] But in the [case] where one wants
accomplish one’s own activity – such as spitting etc. – while hiding
so as to prevent it from becoming an object for [someone] else’s
gnition, it is only while concealing [it] for instance behind a screen
hat one acts?],44 whereas in the [case] where one performs the same
tion [of spitting] but in a different way[, i.e. without concealing
],45 there must be an intention (abhisandhi) [of accomplishing the
tion while] becoming the object of someone else’s cognition.46

– Answer from the Śaiva non-dualist momentarily siding with the
utrāntika:]47 This [statement can only be] uttered by some simpleton,
cause of [the way in which it] explains the origin (vyutpatti) of the
owledge that the other subject is distinct [from myself]; for
ccording to this thesis,] it is only thanks to the manifestation of
g. [my action] of speaking, which is such [, i.e. which involves the
tention of being heard or not by someone else,]48 that one can
certain [the existence of] another subject. Therefore [your reasoning
volves] this logical circle (itaretarāśrayatva): the knowledge of [the
istence of] another subject [supposedly occurs] thanks to the
rticular manifestations [of my actions] such as speaking, [i.e. this
owledge supposedly occurs thanks to my awareness that my own speech is
rticularized by an intention to be heard by the other;] and it is
anks to this [very] knowledge [, i.e. the knowledge that the other
bject exists,] that the particular manifestations of [my actions] such
speaking [can arise] from an intention that is particularized
nasmuch as it] aims at producing the manifestation [of my own action]
e.g. speaking inside that [other] consciousness [in the form “this
rson is speaking”].49 For if [we] could understand through some other
ans of knowledge [i.e. perception] the relation of cause and effect
tween [my] intention of making [these manifestations of my actions]
ise in someone else [and] the manifestations [of my actions] such as
ncing, spitting or arguing that arise in another [conscious stream
en I am acting while being] distinct from myself [i.e. expressed in
e other conscious stream as “he/she is acting” and not as “I am
ting”,] then the other subject could also be inferred at some other
me [, i.e. when this relation is not perceived,] from the
nifestations [in my own conscious stream] of [actions of which I am
t the agent] such as spitting and so on [that] are separated [from me
so far as they occur objectively in myself as “X is spitting”, etc.,]
though [these manifestations] which [occur] in myself are different
rom what the other subject must be thinking in the subjective form “I
nt to spit” – just as a fire is inferred from smoke although smoke is
fferent from fire].50 But [I] cannot know the manifestations such as
itting and so on that reside in the [other subject, e.g. in the form
e/she is spitting” when I am spitting,] because the other subject is
t known through any other [means of knowledge besides inference, i.e.
rough perception,51 and since therefore the relation of cause and
fect between the subjective will “I want to spit” and the objective
ought “he/she is spitting” remains unknown,] how could the inference
another conscious stream [take place] through manifestations which,
though existing in myself, are separated [from myself, in the form “X
spitting”]?

reover,52 it is not [true] that the arising in some other [conscious]
ream of a manifestation of [an action of] e.g. speaking separated
rom the subject, i.e. taking the form “X is speaking”,] is [only]
ssible due to the [previous] intention [in my own conscious stream]
ay the other see me”; because whether [the acting subject has] such an
tention [of being seen acting] or whether [he has] the contrary
tention “may [the other] not see [me acting]”, an absence of arising
the [manifestation of this action in the form “X is acting”] can
sue.53

erefore [since it is] so, as we shall explain [later], another subject
n be established from the manifestation of [his/her action of], e.g.
eaking, [a manifestation which is] not separated from [this] subject
dowed with a will [to speak consisting in] a cognition, [i.e. a
nifestation which is expressed in that subject as “I want to speak”,]
ly (eva) [when this manifestation] involves (anuṣaṅgin) its
rceptibility for the [hearer as well, i.e. only if this manifestation
perceived by me54 too and not just by the other subject, and] comes
existence according to the [specific] order of [phonological]
oduction determined (avadhārita) [by the various causes that are the
ints of articulation and the phonatory organs of the speaker].55

The Sautrāntika:]56 And [it is] thus [only – i.e. it is only if this
nifestation of the action of speaking is common to both subjects and
termined by the same conditions – that we can] establish as well the
ct that this [other subject] has the particular place and time
ssociated with his/her manifestation in my conscious stream,] because
hen this fact] belongs to this [manifestation of speech shared by both
gnitive streams]. Otherwise how could [we] be certain that this
anifestation of an activity of speech] must have been produced in some
her [stream of consciousness, i.e. mine,] by another subject who
sides precisely in [the particular] place where it is manifested [to
]?57

t this only can be [admitted by the Vijñānavādin]58: from a
nifestation such as [my action of] speaking that is directly perceived
myself [and] produced by a will [consisting in] a cognition, such a
nifestation arises in the other subjects [as being] separated [from
eir own subjectivity, in the form “he/she is acting”,] as a reflection
ratibimba) [arises in a mirror]; and therefore, the other subject must
inferred through an inference [based on] this manifestation of
eech, as being the reflected object (bimba) that produces this
eflection in my own conscious stream].

d [if the Vijñānavādin admits that this is] so, [i.e. if he accepts
at there is a plurality of conscious streams reflecting their images
to each other, but no external object,] even the manifestations of
bjects such as] a pot, a cloth and so on which reside in any other
bject – whether a worm or an omniscient [being], etc. – should
oduce manifestations occurring simultaneously [and] with aspects
milar to [what the worm or the omniscient being, etc., perceive] for
ny] other subject as well; and thus [this inference of the other
nscious streams as entities reflected in my own conscious stream that
e Vijñānavādin is ready to accept] is [in fact] nothing but a variety
externalism. [That is to say, it is] the particular acknowledgement
at [entities] external to cognitions are the causes of the phenomenal
fference[s], since if the manifestations [in other conscious streams]
re what produces other manifestations similar to them [in my conscious
ream,] there would be an infinite number (aparyavasāna) of
nifestations in [each of] the subjects, because these [reflections in
conscious stream] would in turn produce another [reflection in
other conscious stream, and] another[, and so on ad infinitum].59

– Objection from the Vijñānavādin:] But still, the manifestations
curring in another subject only awaken an imprint (vāsanā) that [in
rn] produces manifestations of the same sort, but they do not produce
hese manifestations] by themselves, which would [indeed] lead to an
finite number of manifestations. For one imprint only is not capable
influencing (adhipatitva) the innumerable other manifestations;
erefore it [merely] produces a manifestation of the [previously
perienced object] that is only one for each subject – therefore there
no absurd consequence [involved in our theory].60

This article was written thanks to the generous financial help of the
utsche Forschungsgemeinschaft, as part of project FR 2531/3-1. I am
s usual) greatly indebted to Alexis Sanderson, who generously shared
th me some of his copies of manuscripts of the ĪPV as well as his
thusiasm for the historical study of the Pratyabhijñā texts, and to
ncent Eltschinger, Eli Franco and Harunaga Isaacson, who read a
evious version of this paper, corrected a number of mistakes, engaged
lengthy discussions on several difficult points and provided many
sightful remarks.

bliographical References

nuscripts

= Īśvarapratyabhijñā[-Vr̥tti,] Vivr̥ti [and Vimarśinī], Delhi, National
chives of India (Manuscripts belonging to the Archaeology and Research
partment, Jammu and Kashmir Government, Srinagar), no. 30, vol. 9
Pratyabhijñāvivr̥tiḥ by Utpaladeva”, śāradā script].

= Īśvarapratyabhijñāvimarśinī by Abhinavagupta, Delhi, National
chives of India (Manuscripts belonging to the Archaeology and Research
partment, Jammu and Kashmir Government, Srinagar), no. 5, vol. 2,
āradā script].

= Īśvarapratyabhijñāvimarśinī, Śrinagar, Oriental Research Library,
. 838 (Īśvarapratyabhijñākaumudi) [śāradā script].

AS = Īśvarapratyabhijñāvimarśinī, London, School of Oriental and
rican Studies Library, no. 207 in R.C Dogra’s 1978 catalogue / MS no.
255 [“Pratyabhijñāsūtra with Abhinavagupta’s Sūtrārthavimarśinī”,
radā script].

itions

āskarī = Īśvarapratyabhijñāvimarśinī of Abhinavagupta, Doctrine of
vine Recognition, vols. I-II: Sanskrit text with the commentary
āskarī edited by K.A.S. Iyer and K.C. Pandey [Allahabad, 1938, 1950],
l. III: English translation by K.C. Pandey [Allahabad, 1954], Delhi:
tilal Banarsidass, 1986.

K = Īśvarapratyabhijñākārikā: See Torella [1994] 2002.

V= Īśvarapratyabhijñāvimarśinī, edited with notes by M.R. Shāstrī and
K. Shāstrī, KSTS 22 & 33, 2 vols., Srinagar: Nirnaya Sagar Press,
18-21.

VV = Īśvarapratyabhijñāvivr̥tivimarśinī by Abhinavagupta, ed. M.K.
āstrī, KSTS 60, 62 & 65, 3 vols., Bombay: Nirnaya Sagar Press,
38-43.

in = Dharmakīrti’s Pramāṇaviniścaya, Chapters 1 and 2, Critically
ited by E. Steinkellner, Beijing/Vienna: China Tibetology Publishing
use/Austrian Academy of Sciences Press, 2007.

antānāntaradūṣaṇa, in] Ratnakīrtinibandhāvaliḥ, ed. A. Thakur, Patna:
shi Prasad Jayaswal Research Institute, 2nd edn., 1975, pp. 145-49.

ntānāntarasiddhi (rec.) = Santānāntarasiddhiḥ of Ācārya Dharmakīrti
d Santānāntarasiddhiḥ Ṭīkā of Ācārya Vinītadeva, restored and edited
J.S. Negi, Bibliotheca Indo Tibetica Series 37, Sarnath–Varanasi:
ntral Institute of Higher Tibetan Studies, 1997.

ṁśikā, Viṁśikāvr̥tti = Vijñaptimātratāsiddhi. Deux traités de
subandhu. Viṁśatikā accompagnée d’une explication en prose et Triṁśikā
ec le commentaire de Sthiramati, publié par S. Lévi, Paris:
bliothèque de l’École des Hautes Études 245, 1925.

vr̥ti (fragments) = See Torella 1988; 2007a-d; Ratié (forthcoming).

anslations and studies

u, J., 2011, “Sanskrit Fragments of Dharmakīrti’s
ntānāntarasiddhi*”, in H. Krasser et al. (eds.) Religion and Logic in
ddhist Philosophical Analysis: Proceedings of the Fourth International
armakīrti Conference, Vienna, 23-27 August 2005, Wien: Österreichische
ademie der Wissenschaften, pp. 33-42.

rmigatti, C.A., 2011, Sanskrit Annotated Manuscripts from Northern
dia and Nepal, Unpublished Ph.D. Thesis, Universität Hamburg.

jiyama, Y., 1965, “Buddhist Solipsism: A Free Translation of
tnakīrti’s Saṁtānāntaradūṣaṇa”, Journal of Indian and Buddhist
udies, 13(1): 9-24.

no, K., 2008, “Two Short Glosses on Yogācāra Texts by
irocanarakṣita: Viṁśikāṭīkāvivr̥ti and *Dharmadharmatāvibhāgavivr̥ti”,
F. Sferra (ed.) Manuscripta Buddhica 1: Sanskrit Texts from Giuseppe
cci’s Collection, part I, Roma: Istituto Italiano per l’Africa e
Oriente, pp. 343-80.

tsura, S., 2007, “Dharmakīrti’s Proof of the Existence of Other
nds”, in B. Kellner, H. Krasser et al. (eds.) Pramāṇakīrtiḥ: Papers
dicated to Ernst Steinkellner on the Occasion of His 70th Birthday,
ener Studien zur Tibetologie und Buddhismuskunde 70.2, Wien:
beitskreis für tibetische und buddhistische Studien Universität Wien,
. 407-22.

tagawa, H., 1955, “A Refutation of Solipsism (Annotated Translation of
e Santānāntarasiddhi)”, Journal of the Greater India Society, 14(1-2):
32.

Crea, L.J. and P. Patil, 2010, Buddhist Philosophy of Language in
dia – Jñānaśrīmitra on Exclusion, New York: Columbia University Press.

tié, I., 2007, “Otherness in the Pratyabhijñā Philosophy”, Journal of
dian Philosophy, 35(4): 313-70.

—, 2010, “The Dreamer and the Yogin: On the Relationship between
ddhist and Śaiva Idealisms”, Bulletin of the School of Oriental and
rican Studies, 73(3): 437-78.

—, 2011a, Le Soi et l’Autre. Identité, différence et altérité dans la
ilosophie de la Pratyabhijñā, Jerusalem Studies in Religion and
lture 13, Leiden-Boston: Brill.

—, 2011b, “Can One Prove that Something Exists Beyond Consciousness? A
iva Criticism of the Sautrāntika Inference of External Objects”,
urnal of Indian Philosophy, 39(4-5): 479-501.

—, forthcoming, “Some Hitherto Unknown Fragments of Utpaladeva’s
vr̥ti (II): Against the Existence of External Objects”, in D. Goodall
d P.S. Filliozat (eds.) Mélanges tantriques à la mémoire de N.
macandra Bhatt.

hmithausen, L., 1987, Ālayavijñāna. On the Origin and the Early
velopment of a Central Concept of Yogācāra Philosophy, vol. I: Text,
l. II: Notes, Bibliography and Indices, Tokyo: The International
stitute for Buddhist Studies.

—, 2005, On the Problem of the External World in the Ch’eng wei shih
n, Studia Philologica Buddhica (Occasional Paper Series) 13, Tokyo:
e International Institute for Buddhist Studies.

rella, R., 1988, “A Fragment of Utpaladeva’s
varapratyabhijñā-vivr̥ti”, East and West, 38: 137-74.

—, 1992, “The Pratyabhijñā and the Logical-Epistemological School of
ddhism”, in T. Goudriaan (ed.) Ritual and Speculation in Early
ntrism, Studies in Honor of André Padoux, SUNY Series in Tantric
udies, Albany: State University of New York Press, pp. 327-45.

— [1994] 2002: Īśvarapratyabhijñākārikā of Utpaladeva with the
thor’s Vr̥tti, critical edition and annotated translation [Roma, 1994],
rrected Edition, Delhi: Motilal Banarsidass.

—, 2007a, “Studies on Utpaladeva’s Īśvarapratyabhijñā-vivr̥ti. Part I:
upalabdhi and apoha in a Śaiva Garb”, in K. Preisendanz (ed.)
panding and Merging Horizons: Contributions to South Asian and
oss-Cultural Studies in Commemoration of Wilhelm Halbfass, Wien:
rlag der Österreichischen Akademie der Wissenschaften, pp. 473-90.

—, 2007b, “Studies on Utpaladeva’s Īśvarapratyabhijñā-vivr̥ti. Part II:
at Is Memory?”, in K. Klaus and J.-U. Hartmann (eds.) Indica et
betica. Festschrift für Michael Hahn zum 65. Geburtstag von Freunden
d Schülern überreicht, Wiener Studien zur Tibetologie und
ddhismuskunde 66, Wien: Arbeitskreis für Tibetische und Buddhistische
udien Universität Wien, pp. 539-63.

—, 2007c, “Studies on Utpaladeva’s Īśvarapratyabhijñā-vivr̥ti. Part
I: Can a Cognition Become the Object of Another Cognition?”, in D.
odall and A. Padoux (eds.) Mélanges tantriques à la mémoire d’Hélène
unner, Collection Indologie 106, Pondichéry: Institut Français de
ndichéry/École Française d’Extrême-Orient, pp. 475-84.

—, 2007d, “Studies on Utpaladeva’s Īśvarapratyabhijñā-vivr̥ti. Part IV:
ght of the Subject, Light of the Object”, in B. Kellner, H. Krasser et
. (eds.), Pramāṇakīrtiḥ: Papers Dedicated to Ernst Steinkellner on the
casion of His 70th Birthday, Wiener Studien zur Tibetologie und
ddhismuskunde 70.2, Wien: Arbeitskreis für tibetische und
ddhistische Studien Universität Wien, pp. 925-40.

----------------------------------------------------------------------

See Torella 1988 (this first part of the fragment was in fact edited
om a transcript of the original manuscript D later found by Raffaele
rella in the National Archives of India and used for the edition of
e remaining parts of the fragment); 2007a-d.

See Torella 1988: 140; [1994] 2002: xliii.

For a study of some marginal annotations in ĪPV manuscripts (which
es not, however, bear on fragments of the Vivr̥ti), see Formigatti
11.

See Ratié 2011a: 250. Abhinavagupta gives a complete quotation of the
rse in ĪPV I: 130 while explicitly ascribing it to Utpaladeva (yad āha
anthakāra eva), and from ĪPVV, vol. III, p. 404, it is obvious that
e verse was the first of a series concluding the Vivr̥ti: vivr̥tikāraḥ
ṁkṣipya sarvam āhettham ity
iślokaiḥ. anenāsmadukteneśvaratāprakāreṇa yā svasaṁvittiḥ* sthitā tām.
svasaṁvittiḥ corr.: susaṁvittiḥ ĪPVV]. According to the ĪPV the verse
ns as follows: itthaṁ svasaṁvittim apahnuvānair yat tad vadadbhiḥ
luṣīkr̥taṁ yat । pramātr̥tattvaṁ sphuṭayuktibhis tān mūkān vidhāya
akaṭīkr̥taṁ tat ।। “After having thus reduced to silence through clear
guments those who, denying their [own] self-consciousness, hold this
that thesis, [I] have made obvious the true nature of the subject [so
r] stained by them.”

See Torella 2007b: 544, n 14; 2007c: 479, n 8; 2007d: 929, n 11.

See Yohei Kawajiri’s contribution to this volume.

D2, SOAS and S3 all contain the passage from pūrvāvabhāsitavastunaḥ. .
to vicitrecchābodhitāvabhāsabhedavāsanāmātropapāditavikalparū-
vijñānāpohanaśaktiḥ (as noted in Torella 2007a: 480, n 38, so does
0, n 71, in the Kashmir Series of Texts and Studies edition of the ĪPV
the three manuscripts as well as the ĪPV footnote invert the two
ntences) and the passage from smr̥tyādīnām idānīṁ. . . to sambhavam āha
Torella 2007b: 544 (which is also quoted in ĪPV I: 116, n 17).

D2 and SOAS both contain the passage from arthitvāc ca . . . (the two
nuscripts bear the reading arthitāvaśāt . . .) to avabhāso bhavati in
rella 2007b: 547 (Torella notes in n 44 that the same passage is found
n 62, p. 124 of ĪPV I, and that the text of the footnote shows some
ignificant divergence” from the text he has established. It is worth
ting in this regard that while D2 bears exactly the same reading as
at found in the ĪPV n, the text in SOAS differs from it in several
stances, among which upajātāyāṁ, also found in manuscript D used by
ffaele Torella for his edition, but for which D2 and the ĪPV n have
ayāyamānāyāṁ); the passage from kadācit punar . . . to na kṣatiḥ in
rella 2007b: 548 (as noted in Torella 2007b: 548, n 51, this passage
also quoted in ĪPV I: 125, n 67; the text of the footnote has the
me variant readings as the two manuscripts); the passage from aham
vabhūvam iti . . . to anubhāvyadharmopacāreṇa in Torella 2007c: 481
n fact this passage is also found in ĪPV I: 129-30, n 69; SOAS bears
e obviously mistaken reading iti vā pūrvapakṣānubhavasya instead of
i vā pūrvānubhavasya); and the passage from pratyakṣapr̥ṣṭapātināpi . .
to parāmarśa ity āha in Torella 2007d: 930 (the sentence is also
oted in ĪPV I: 142, n 141).

S3 quotes the passage from viṣayopalabdhr̥tāpy arthasya . . . to
tpūrvakāloparaktānubhūtabhāvāvamarśanāvasānā in Torella 2007b: 544-45;
e passage from tena ca sūtre liṅartho . . . to prakāśanād bhavati in
rella 2007b: 546-47; the passage from nanu prakāśānupraveśaṁ vinā . .
to sahaikyaṁ syād iti in Torella 2007c: 481-82; the passage from
ginām api parabodhaḥ . . . to na tv etannidarśane’nubhavasmr̥tyor
āhyagrāhakabhāva iti in Torella 2007c: 482 (as noted in Torella 2007c:
2, n 42, this passage is also reproduced in ĪPV I: 137, n 121, from
āhyabhūmi . . . onwards. Note that both S3 and the ĪPV footnote bear
e reading ’pi vā tasya parabodhasya instead of ’pi vā tasya); the
ssage from kadācid dr̥kparāmarśapūrvatvena . . . to dr̥kparāmarśa iti in
rella 2007d: 931-32; the passage from athāpi yogyadeśasthatayā . . .
pītasyāpi syāt in Torella 2007d: 937 (in S3 a gloss is interpolated
ter vā pratyāsattyā, and the manuscript shows several variant
adings, among which pītasyāpi na syāt instead of pītasyāpi syāt at the
d of the passage).

Some of the quotations end with iti ṭīkāyām or iti ṭīkākāraḥ; some
e preceded by tad āha ṭīkākāraḥ or iti darśayati vivr̥tikāraḥ.

See also Ratié (forthcoming).

tattadākasmikābhāso bāhyaṁ ced anumāpayet । na hy abhinnasya bodhasya
citrābhāsahetutā ।। na vāsanāprabodho’tra vicitro hetutām iyāt ।
syāpi tatprabodhasya vaicitrye kiṁ nibandhanam ।। “[A Sautrāntika
ght] object (cet) [the following]: the manifestation of this or that
articular object,] which is [apparently] devoid of cause (ākasmika),
st lead [us] to infer an external [reality], for consciousness, which
[in itself] undifferentiated, cannot be the cause of manifestations
at are varied. [And] the varied awakening of imprints (vāsanā) cannot
the cause of this [phenomenal variety]; for what would be in turn the
use of the variety of the awakening of these [imprints]?”

On this criticism of the Vijñānavāda by a Sautrāntika as explained in
hinavagupta’s commentaries, see Ratié 2010: 453-60.

See Ratié 2007 and 2011a (chap. 8).

As far as I know the text is entirely preserved in its Tibetan
rsion only. For an English translation from the Tibetan, see Kitagawa
55; for a tentative reconstruction of the Sanskrit text, see
ntānāntarasiddhi (rec.), and for a short bibliography of scholarly
rks on this text, see Katsura 2007: 408-09. For newly-published
nskrit fragments, see also Chu 2011. On Abhinavagupta’s quotation and
mmary of Dharmakīrti’s treatise, see Ratié 2007: 323, n 20; 2011a:
1.

Objective in the sense that the agent of the action is not
prehended by him- or herself, subjectively, but as an object distinct
om the “I”.

See below, n 55.

For an analysis of Utpaladeva’s position in this debate see Ratié
07: 336-68; 2011a: 606-42.

See below, n 56.

See below, n 57 (on the probability that Vasubandhu’s work originally
re the title Viṁśikā rather than Viṁśatikā, see Kano 2008: 345).

See Schmithausen 1987, vol. II: 415-16 n 769.

See Schmithausen 2005: 39-47.

Viṁśikā 18ab: anyonyādhipatitvena vijñaptiniyamo mithaḥ । “There is a
tual determination of representations (vijñapti) due to the fact that
he various representations in various conscious streams] influence
dhipatitva) each other.” Cf. Viṁśikāvr̥tti, p. 9: sarveṣāṁ hi sattvānām
yonyavijñaptyādhipatyena mitho vijñapter niyamo bhavati yathāyogam.
tha iti parasparataḥ. ataḥ santānāntaravijñaptiviśeṣāt santānāntare
jñaptiviśeṣa utpadyate nārthaviśeṣāt. “Thus for all living beings
attva), due to the fact that representations influence each other,
ere is accordingly a determination of representation[s] that is
tual, i.e. [exerted] by each other. Therefore a particular
presentation arises in one [conscious] stream due to the particular
presentation of another [conscious] stream[, and] not due to a
rticular [external] object.” Note that Vasubandhu, as well as
hiramati (see Schmithausen 2005: 44) and Dharmakīrti in the
ntānāntarasiddhi, call this influence of one conscious stream on the
hers an adhipati/adhipatipratyaya an – expression used by the
jñānavādin portrayed by Utpaladeva in the last paragraph of the
agment below. And whereas Vasubandhu seems to consider that thanks to
is influence the occurrence of specific representations can be
plained away without having recourse to an external object, Sthiramati
ecifies that this influence does not produce the representational
ject itself but only functions as some “controlling or directing
ndition” (ibid.), and Dharmakīrti contrasts this influencing condition
a conscious stream “A” of another conscious stream “B” with the
terial cause (upādānakāraṇa) that the previous cognition in the
nscious stream “A” constitutes for the next cognition in that same
ream (Kitagawa 1955: 20-21).

See e.g. ĪPVV II: 126 (quoted below n 57) and ĪPVV: II: 144 (quoted
Ratié 2010: 446; 2011a: 391).

ĪPV I: 174-75: anumīyamānam api ca bodhāntaram anumātr̥saṁmatād bodhād
di bhinnaṁ tad asti tāvat saṁbhavo yat prameyaṁ bodhād bhinnam astīti.
hopalambhaniyamāder anaikāntikatvān nīlapītādināpi prameyarāśinā kim
akr̥taṁ yenāsya svarūpaviśrāntir na sahyate. tasmāt pramātrantarāṇām
y asiddhir eva. “Moreover, if the other consciousness being inferred
distinct (bhinna) from the consciousness considered as the inferring
bject, then obviously, the possibility arises that the object of
nsciousness [too] may be distinct from consciousness! Since [as a
nsequence, Dharmakīrti’s idealistic arguments] such as ‘the necessity
or the object] of being perceived along with [the cognition]’
ahopalambhaniyama) are inconclusive, what crime have all objects – the
ue, the yellow, etc. – committed that would prevent them from resting
their own nature [independently of the consciousness apprehending
em]? Therefore, [just as there can be no correct account of phenomenal
riety in the Vijñānavāda,] there can be no demonstration of [the
istence of] other subjects either.” Cf. Ratnakīrti’s remark in the
ntānāntaradūṣaṇa, p. 148: if other streams are admitted to exist
dependently of the cognitive stream apprehending them, “then how does
e demonstrate that the external object too does not exist? For [then]
e can cast a doubt on this [non-existence of external objects] as
ll” (tarhi bahirarthasyāpi katham abhāvaḥ sidhyati, śakyam hi tatrāpi
ndeham avatārayitum). On the Pratyabhijñā’s exploitation of
armakīrti’s sahopalambhaniyama as formulated in PVin,
atyakṣapariccheda 54ab, see Ratié 2011b: 484-86.

See ĪPVV II: 128.

See, for example, the problem of the interpretation of ĪPV I: 173 and
divergence with K.C. Pandey’s interpretation in particular (see
āskarī III: 62; Ratié 2011a: 596-97, n 61): it seems to me that the
agment definitely settles the matter of what the vyāpti attacked by
e Sautrāntika really is.

See Torella 1992.

As explained elsewhere (see Ratié 2011a: 644, n 133) I believe that
armakīrti’s Santānāntarasiddhi mainly focuses on an epistemological
oblem (how do we become aware of the existence of other conscious
reams?) and answers this problem by saying that the other conscious
reams are mere conceptual objects for us, i.e. that the assumption of
eir existence is epistemologically valid in the sphere of human
actice (vyavahāra), but cannot be considered as an ultimate truth
nce the existence of other streams remains impossible to perceive.
netheless, later Buddhist authors claiming to belong to Dharmakīrti’s
adition explicitly deny the ontological reality of other conscious
reams, as Ratnakīrti does in his Santānāntaradūṣaṇa (see Kajiyama
65: 13).

On Utpaladeva’s refutation of the Sautrāntikas’ externalism, see
tié 2011b.

On Abhinavagupta’s and Rāmakaṇṭha’s quotations of this treatise see
tié 2011a: 591, nn 50-51.

See Ratié 2011b: 481, n 5, on the problematic identity of these
utrāntikas.

See ĪPVV II: 128. On the hypothesis that Dharmottara may have held a
eory similar to that ascribed by Abhinavagupta to the Sautrāntikas,
d on the ongoing debate among scholars regarding this issue, see e.g.
Crea and Patil 2010: 141-42, n 71.

The passage is quoted on the margins of S3 ad ĪPK 1.5.4-5 (the folio
ars the number 40); it starts on the top of the page, extends on the
ght margin down to the bottom of the page, and goes on in the left
rgin down to the bottom of the page. The words quoted in
hinavagupta’s ĪPVV are underlined and in bold type. I have taken the
berty of adding punctuation marks and of standardizing the use of the
ndhi whenever needed. Sanskrit words or syllables are in brackets
enever the akṣaras are barely legible or illegible but obviously
quired by the context. All my conjectural emendations are indicated in
otnotes.

A few akṣaras (four?) are illegible.

-vyāhārādy- conj.: -vyavahārādy- S3.

vyāhārādy- conj.: vyavahārādy- S3.

yadi hi conj.: yad iha S3.

From here on the text is noted on the left margin.

tatsadr̥śābhāsa- conj.: tattadr̥śābhāsa- S3.

yato’vabhāsānāṁ conj.: yathāvabhāsānāṁ S3.

pramātr̥ṣv ābhāsāparyavasānaṁ conj.: pramātr̥prābhā-saparyavasānaṁ S3.

nādhipatitvocitaikaiveti conj.: nādhipratitvocitaikaiveti S3.

See above, n 35.

Cf. ĪPVV II: 122: anyathaivety anapavāryety arthaḥ.

This is the reason why, according to the Vijñānavādin, we do know the
variable concomitance between “I want to act” and “X is acting”: when
act we are aware that our subjective will to act must produce the
areness “he/she is acting” in other conscious streams – otherwise we
uld not try to hide our actions, nor would we act while wanting the
hers to see us. And since the invariable concomitance between “I want
act” and “X is acting” is known, we can legitimately infer the
istence of a will to act in some other conscious subject when we have
e cognition “he/she is acting” – so our awareness that all our actions
e potential objects for other conscious subjects is the basis of our
ferential knowledge that other subjects exist.

See ĪPVV II: 122, quoted and translated in Ratié 2011a: 598-99, n 64.

See ĪPVV II: 123: dr̥śyatāṁ hi yadā pāmarakadāserakāv anyonyaṁ
hīvato vādināv udgrāhayato yuvānau nr̥tyataḥ, tadāyaṁ para etaj jānātv
i yad abhisandhānaṁ tat parapramātr̥jñānapūrvakam, parajñānaṁ ca
thābhisandhijanitanr̥ttābhāsajñānapūrvakam ity anyonyāśrayam. tad āha
thābhūteti. paratrotpadyatām ayam ābhāsa ity
adabhisandhānajanito’yam vyāhārābhāsa iti. “For see: when two
w-caste men are spitting on each other, when two philosophers are
guing, when two young persons are dancing, the intention [occurring]
n the form]: ‘may this other [person] know this [action of mine]’ must
preceded by the knowledge of [the existence of] the other subject;
d the knowledge of [the existence of] the other [subject] must be
eceded by the knowledge that the manifestation [in the other conscious
ream] of [my action of] dancing [etc.], results from such an intention
f being perceived by others] – therefore there is a logical circle.
is is what [Utpaladeva] expresses [by saying] ‘[the manifestation of
e action of speaking,] which is such’ – [i.e.], the manifestation of
he action of] speaking which is the result of this intention: ‘may
is manifestation [of my action] arise in [someone] else’.”

That is, if I understand correctly, the Vijñānavādin is accused here
grounding his inference of the existence of other subjects on the
ct that my actions are particularized by a will to be perceived by the
hers (which is why, according to him, the relation of cause and effect
tween the subjective will “I want to speak” and the objective thought
e/she is speaking” is already ascertained, and can be used as a
nsequence to prove that the other subject exists in the form of a
bjective will causing my objective thought “he/she is speaking”),
ereas the very fact that my will is thus particularized already
esupposes the knowledge of the existence of other subjects. Cf. e.g.
V I: 173: vicchinna ābhāsaḥ paratrotpadyatām iti yā samīhā tayā saha
ratrotpannasya vicchinnābhāsasya kāryakāraṇabhāvagrahaṇam eva
rāsiddhau na yuktam. “The very fact that [I] apprehend the relation of
use and effect between [my] will [to be heard while speaking – a will
at takes the form] ‘may a manifestation [of my action] separated [from
e subject, i.e. expressed as ‘he/she is speaking’,] arise in the other
ubject]’ – and the manifestation that occurs in the other [subject
d] is separated [from the subject, i.e. takes the form ‘he/she is
eaking’], is not possible if the other [subject] has not been
tablished [beforehand].”

See ĪPVV II: 124: tad anyena dhūmena yathāgnyantarasyāvagamas
thānyena vicchinnanr̥ttādyavabhāsenānyasya pramātur avasānaṁ ghaṭeta.
hen just as [we] learn [the existence] of fire, which is one thing,
rough [the knowledge of the existence of] smoke, which is [something]
se, in the same way, [we] could determine [the existence of] the
ther] subject, which is one [thing], through a manifestation which is
omething else] such as dancing and so on [since this manifestation
es not take the form ‘I am dancing’ characteristic of a conscious
ream, but rather, is] separated [from the subject in the form ‘he/she
nces’, etc.].”

See ĪPVV II: 124: anyata iti pratyakṣāt.

According to Abhinavagupta’s commentary (ĪPVV II: 124), Utpaladeva is
w showing that there is no invariable concomitance between the
bjective desire to be seen acting and the objective cognition “X is
ting” (so that the Vijñānavādin cannot rightly claim that the other
bject can be inferred on the basis of the logical pervasion between
ese two).

That is, my action might not be seen by anybody, whether I want
hers to see it or not; and conversely, my action might be seen by
hers whereas I do not have any particular intention of showing it to
hers: cf. Abhinavagupta’s example of the crying child in ĪPVV I: 124
ee Ratié 2007: 330-31; 2011a: 599).

See ĪPVV II: 125: yaḥ pramātā maitraḥ, tasmād avicchinna eva yo
āharāmīty ābhāsaḥ svātmani pratyakṣapratibhāsī. . . “The manifestation
am speaking’ that is not separated from the subject [called] Maitra
nd] involves a perceptual image (pratyakṣapratibhāsin) in myself
vātman) . . .”

The passage is difficult and I am not sure whether my interpretation
correct. At any rate here is what I understand: the Vijñānavādin’s
ponent has just explained that the intention (abhisandhi) of the
eaker willing to be heard by others is not enough to ascertain the
istence of other subjects, and he now explains that the existence of
her subjects must rather be established from the fact that the same
tion of speaking is perceived both by the speaker (in the subjective
rm “I am speaking”) and by the hearer (in the objective form “he/she
speaking”) and is determined (avadhārita) in both cases according to
e same spatio-temporal conditions – a fact that the Vijñānavādin is
ecisely incapable of justifying. Abhinavagupta considers that here the
iva non-dualist is still speaking and anticipating on his own
planation of intersubjectivity: according to him, the Śaiva
n-dualistic perspective justifies this unity of experiences belonging
several subjects in so far as it considers all these subjects as
mited aspects that the all-powerful and all-encompassing consciousness
eely takes on while playfully dividing and synthesizing its
numerable manifestations. See ĪPVV II: 125: tasminn eva yataś caitraḥ
amātā parameśvareṇābhāsāntaraikyād upasaṁhr̥tyaikībhāvena sr̥ṣṭaḥ . . .
tataḥ so ’pi sarvaṁ maitravad evāvabudhyate, paratas tu
rīrādyahaṁbhāvadārḍhyāt svaparavibhāgagaṇanānapagamād* vikalpayati
itro vyāharatīty evam īśvaravādisaṁmatam etat. [*-gaṇanānapagamād
nj.: -gaṇanāpagamād ĪPVV] “Because the Highest Lord, after withdrawing
he individual called Caitra] from the unity of [all] other
nifestations, has created the subject [called] Caitra as being one
ith the subject Maitra] with respect to this very [manifestation of
eech subjectively apprehended in Maitra as ‘I am speaking’], [. . .]
is [Caitra] too is aware of everything exactly as Maitra is; but on
e other hand, because [Caitra] does not cease taking into account the
stinction between himself and the other(s) due to the strength of his
lf-identification with his body [, intellect, vital breath and void
ich are mere objects that he wrongly considers as himself], he
nceptualizes [this manifestation] in this form: ‘Maitra is speaking’ –
is is the opinion of the proponent of the [non-duality with] the
rd.” However, Abhinavagupta also considers that up to this point, the
scourse of the protagonist fighting with the Vijñānavādin could be
terpreted as an expression of the Sautrāntika’s opinion
āhyavādimata, ibid.). I assume that this is the case because so far
is protagonist has not explained how this unity of experiences
rtaining to various subjects is possible (the Śaiva non-dualist
plains it as a result of the dynamism of consciousness, which is
pable of dividing itself without ceasing to be one, whereas the
utrāntika justifies it by having recourse to the existence of an
ternal world common to all subjects). But whatever the interpretation
osen, what matters here, according to Abhinavagupta, is the assertion
at the sphere of human practice (vyavahāra) has a fundamental unity
at must be accounted for. See ibid.: ekarūpatā tāvad
avahārasyānenoktā. “In any case (tāvat), the [passage] expresses the
ity of the sphere of mundane practice.”

According to Abhinavagupta, from here on the Sautrāntika is speaking
ain. See ĪPVV II: 126: sā ca bāhyatāyām apīti labdhocchvāso bāhyavādy
a: evaṁ ceti . . . “And the externalist, who has regained faith [in
e possibility of his victory thanks to the intervention of the Śaiva
n-dualist and considers that] this [unity of the sphere of human
actice is possible] in the externalist perspective as well, says: ‘And
t is thus] . . .’.”

See ĪPVV II: 126: aparathā kampate maitra iti yaś caitrasyābhāsas
ddeśābhāsasahitaḥ, sa brahmalokasthamaitrābh-
hānapramātrantarajanito na bhavatīti kuto niścayaḥ. “Otherwise how
uld [we] be certain that the manifestation ‘Maitra is moving’ that
longs to Caitra[’s conscious stream and] is associated with the
nifestation of the place [where Caitra is perceiving Maitra] is not
n fact] produced by another subject called ‘Maitra’ who resides in
ahma’s world?” Abhinavagupta explains that the Sautrāntika is
sisting on the fact that the other subject must be associated with the
rticular time and place where I perceive him/her to exist because he
nsiders that the Vijñānavādin, according to whom there is no external
rld, is unable to account for this fact. See ĪPVV II: 126: evaṁ
ddeśatvādyanavadhāraṇadoṣam uktvā doṣāntaram api vijñānavādina
ādayaty evaṁ ceti. “Having thus formulated the fault [in the
jñānavādin’s thesis consisting in] the fact that the [other subject’s]
istence in that [very] place, etc., cannot be determined
navadhāraṇa), he brings about another fault of the Vijñānavādin [in
e sentence beginning with] ‘moreover, since it is so. . .’.” This
cusation is of course reminiscent of the externalist’s objection
dressed by Vasubandhu at the beginning of the Viṁśikā, and as a matter
fact Abhinavagupta quotes Viṁśikā 2 while commenting on this passage.
e ibid.: tat satyam uktam: na deśakālaniyamaḥ santānaniyamo na । na ca
̥tyakriyā yuktā vijñaptir yadi nārthataḥ ।। iti. “Therefore it has
ghtly been said that ‘If representation (vijñapti) does not [arise]
om an [external] object, there [can be] no restriction (niyama) [of
is representation to a particular] place and time, nor any restriction
f this representation to a particular conscious] stream; nor is
ficacy possible [, since dreamt or illusory objects lack efficacy].”

According to Abhinavagupta, the gist of the argument about to be
rmulated here is as follows: the existence of the other subject must
inferred on the grounds that a form which is manifest in something as
it were external to the nature of that thing must have been projected
to that thing by something else, just as the form of a pot in a mirror
st be the projection of a pot existing outside the mirror. See ibid.:
āvad eva hi tadā syād yad yatrābhāti* svarūpabahirbhūtam iva ca, tad
yataḥ saṁkrāntaṁ darpaṇa iva ghaṭādiḥ . . . caitrapramātari cābhāti
āhāraḥ svarūpabahirbhāvena maitro vyāharatīti, tad anenānyataḥ
ṁkrāntena bhavitavyam iti svabhāvahetuḥ. [*yad yatrābhāti corr.: yady
rābhāti ĪPVV] “For then this only can be [admitted by the
jñānavādin]: a [thing A] that is manifest in a [thing B] and [is
nifest there] as if it were external to the nature [of B] is [a form]
at has passed [onto B] (saṁkrānta) from [something] else, just as a
t in a mirror for instance [is the form of a pot external to the
rror that has passed onto the mirror]. [. . .] And the speech [of
itra] is manifest in the subject [called] Caitra as being external to
e nature [of Caitra, in the form] ‘Maitra is speaking’; therefore this
peech] must be [a form] passed onto [Caitra’s consciousness] from some
her [entity, i.e. another subject] – therefore the inferential reason
n which the inference of the other subjects’ existence is based] is a
ture [and not an effect].” This thesis is of course reminiscent of the
y in which the Sautrāntika demonstrates the existence of the external
rld by inferring the existence of a “reflected object” (bimba)
ojecting its form onto consciousness and thus accounting for the
riety and specificity of our perceptions (see Ratié 2011a: 368-76;
11b: 481-84). But as far as I understand, here the Sautrāntika is
inting out that the Vijñānavādin is compelled to accept this
planation of the way we become aware of the other subjects (because he
s no other option left, and also because the thesis that several
reams of consciousness reflect their images onto each other is still
mpatible with his idealism and seems to have been held at least by
me Vijñānavādins: see above, nn 22-23). And the Sautrāntika is
sisting on this point because he wants to show (as he will in the next
ragraph) that this explanation of our awareness that other subjects
ist necessarily leads to his own thesis, namely, that of the existence
an external world.

According to Abhinavagupta, the Sautrāntika is explaining here that
e Vijñānavādin has to admit defeat (by conceding the existence of
ternal objects) because the only idealistic option left to him in this
scussion leads to absurd consequences. This option consists in
nsidering that the phenomenal variety in my own conscious stream is
timately due to manifestations occurring in other conscious streams
d projecting their reflections onto my conscious stream, just as, when
other subject has the cognition “I am speaking”, his/her action is
flected in my conscious stream in the form “he/she is acting”. This
esis (on which see above, nn 22-23) has the advantage of enabling the
jñānavādin to acknowledge the existence of other subjects while not
ving recourse to any reality external to consciousness (since the
iverse is then nothing but a variety of consciousnesses reflecting
to each other their own manifestations). But it is problematic for two
asons. First, if it were the case then all the manifestations of
jects occurring in one conscious stream should occur in all the
hers, so that every conscious stream would be simultaneously conscious
all objects. Second, the manifestation of an object in a given
nscious stream, which supposedly projects its form onto other
nscious streams, must be considered itself as some sort of reflection
nce it appears objectively, in the form “this”, i.e. as separated from
e subject perceiving it, and since it has been established that an
jective thought such as “he/she is speaking” must be a reflection in
e conscious stream of a subjective thought “I am speaking” in some
her conscious stream. If any manifestation of an object in a given
nscious stream is to be considered as a reflection of a manifestation
some other conscious stream, then in order to avoid an infinite
gress (since my thought “the pot” must be a reflection of somebody
se’s thought “the pot”, which in turn is a reflection of somebody
se’s thought, etc.), we could consider that the objective thought “the
t” ultimately amounts to the subjective thought “I am the pot” or “I
manifest [as] the pot”, just as my objective thought “he/she is
eaking” is in fact a reflection of somebody else’s subjective thought
am speaking”. But then, as Abhinavagupta points out, we would have to
mit that all objects are ultimately one and the same thing, namely,
e “I”, and phenomenal variety could not be accounted for. The only way
escape the infinite regress is therefore to admit that this
enomenal variety has as its cause the existence of entities that are
ternal to consciousness. See ĪPVV II: 127: sarva eva
amātrantaravartina ābhāsāḥ samarpakāḥ prāptā viśeṣābhāvāt.
rpaṇadvaye’pi hy anyonyapratibimbārpaṇe’saṁkhyapratibimbodayaḥ, kim-
ga pramātr̥koṭiparārdhānām anyonyapratibimbārpaṇe. tad āhāparyavasānaṁ
ād iti. kiṁ ca samarpakasaṁmato’pi pramātrantaravarty asāv ābhāso
madarthaviśrāntivailakṣaṇyenedam iti nirbhāsamānaḥ
̥thagbhāvābhimānayogyatvāt pratibimbakalpa iti tasyāpy arpakeṇa
āvyam, tatrāpy evam ity anavasthā. tatparihārāya tu paryante ghaṭa
āsmadarthasaṁbhinnaḥ sakalapramātr̥vyatirikta
hāsanasāro’bhyupagantavyo’haṁ ghaṭo bhāsa ity evaṁrūpaḥ. tatrāpi ca
aṭatā yadi nādhikā tad ahaṁ ghaṭo ’haṁ paṭa iti nirbhāsamānayor
aṭapaṭayor aikyam eva syāt. adhikā cet tatrāpi pratibimbateti
mbenānyena bhavitavyam ābhāsarūpatāśūnyeneti bāhya evārthaḥ siddhaḥ.
In that case] all the manifestations without exception that exist in
e other subjects would as a consequence be projecting [their forms
to my conscious stream,] since there is no difference [between them
at would prevent some of them from thus projecting their form]. For
en when two mirrors project reflections onto each other, countless
flections arise – all the more if innumerable subjects project their
flections onto each other! This is what [the Sautrāntika portrayed by
paladeva] expresses [by saying that] ‘there would be an infinite
mber [of manifestations in each of the subjects’]. Moreover, although
is manifestation [of a pot] existing in another subject is considered
what projects [its reflection onto my consciousness, since it is]
nifest [in the form] ‘this’, [i.e.] as [something] that does not rest
the first [grammatical] person, it is similar to a reflection, since
can be associated with the belief that it exists separately [from the
bject, in the form ‘this’ and not in the form ‘I’, and since we are
suming for the sake of argumentation that all the objective
nifestations in my conscious stream are reflections of a subjective
nifestation in some other conscious stream]. Therefore this
eflection] too must have [something that produces it by] projecting
ts form, and] that [thing projecting its form too] requires [a third
ojecting entity in so far as it must be a reflection as well];
erefore there is an infinite regress. But in order to avoid this
nfinite regress, we] could acknowledge that in the end the pot itself,
ich is [apprehended as an object, i.e. as being] distinct from the
rst grammatical person [and] separated from all subjects, has as its
sence the [subjective conscious] manifestation which takes such a
rm: ‘I am manifest [as] the pot’. And in that case too, if being a pot
nothing more [than an internal, subjective manifestation,] then the
t and the cloth, which should be manifest [in the same subjective
rm] ‘I am the pot, I am the cloth’ would just be one and the same
hing]. But if [being a pot] is something more than [an internal aspect
consciousness,] in that case too [this manifestation] is a
flection; therefore there must be some other [entity] that is the
flected object (bimba) [projecting its form onto consciousness and]
at does not consist [in turn] in a [conscious] manifestation [but
ther, is external to it] – therefore the [existence] of the external
ject is established.”

See ĪPVV II: 127: aparyavasānadoṣaparijihīrṣāṁ parakīyāṁ
ṁkate’thāpīti. vāsanāprabodhamātre teṣāṁ vyāpāro na
apratibimbārpaṇe, vāsanā ca ghaṭasahasrair api pūrvaghaṭānubhavajanitā
atibodhitā tam evaikaṁ pūrvānubhūtaghaṭābhāsam unmeṣayati, tato
paryavasānātmābhāsānām atiprasaṅgaḥ. “[With the words] ‘But still’,
tpaladeva] states as an objection the opponent’s desire to avoid the
ult of an infinite number [of manifestations in each conscious stream.
re is what the opponent means:] the activity of the [manifestations
curring in other conscious streams] only concerns the awakening of
prints [in my conscious stream, but] not the projection of a
flection of their own [form into my conscious stream]; and the
print, which has been produced by previous experiences of pots, [i.e.]
innumerable pots, makes this manifestation of a previously
perienced pot arise [and] this [manifestation] only; therefore there
no absurd consequence [in our position that would] consist in an
finite number of manifestations.”

Sources & corrections: edit source →. source lives here.