Some Further Comments on the First Sectionof the Vigrahavyāvartanī
Claus Oetke
Published online: 18 May 2012
© Springer Science+Business Media B.V. 2012
Abstract The publication deals with topics concerning the interpretation of the
Vigrahavyāvartanī in as much as they are relevant for the understanding of (early)
Madhyamaka-philosophy in general. A major part of the article is dedicated to a
critical assessment of a number of views which have been propagated recently in a
paper by Sharma (In: Nagoya studies in Indian culture and Buddhism, Sam˙bhas
˙a,
2011). A primary goal of the present investigation consists in substantiating the
claim that early Madhyamaka represents a metaphysical teaching which stands in
sharp contrast not only to stances of common sense but also to tenets propagated in
Buddhist dogmatics.
Keywords History of philosophy · Textual exegesis · Madhyamaka
I
If the contention is true that the conciseness of many theoretical treatises written in
classical Sanskrit generates equivocations and vagueness of expression affecting
textual exegesis to a vital degree it is sensible to suppose that consideration of extra-
linguistic facts deserves to play a most important role for the interpretation of those
sources. Hence, in spite of being criticizable because of its excessive imprecision,
the slogan that philosophical works of the classical Indian tradition should be
understood against the background of ‘their own historical context’ should possess a
grain of truth. In fact the idea of bringing considerations of ‘historical context’ into
play can be implemented both with respect to the reconstruction of beliefs adopted
by an author of a text and with regard to the recognition of communicative
intentions connected with particular textual segments: Given that a particular
C. Oetke (&)
University of Stockholm, Stockholm, Sweden
e-mail: [email protected]
123
J Indian Philos (2012) 40:371–394
DOI 10.1007/s10781-012-9156-1
linguistic utterance can be regarded as a fairly suitable expression of some particular
belief both considerations pertaining to the coherence with other beliefs attributable
to the same author and considerations regarding the issue of whether the adoption of
the concerned belief appears intelligible against the backdrop of what we suppose to
know about views prevailing at the time and in the milieu in which a text has been
written open possibilities of effectively controlling the assessment of the probability
of pertinent belief-ascriptions. In a similar manner a hypothesis concerning
communicative intentions associated with some individual utterance or chain of
utterances1 can be assessed with respect to the probability of its correctness against
the background of either other communicative goals and intentions indicated by the
cotext2 of a textual unit or their relevance relative to what is supposed to be known
about attitudes prevailing in the milieu in which a work has been created. It should
not be difficult to discern that the criterion of coherence and relevance with respect
to biographical and historical data of the above indicated types possesses merely
limited efficiency: On the one hand it vitally relies on suppositions which are
themselves established as hypothetical assumptions by similar criteria and on the
other hand biographical and historical circumstances do not fully determine
individual actions and attitudes. For example, due to the possibility that some
individual adopted particular beliefs which no contemporary envisaged or accepted
can never be ruled out it would be an absolutely unacceptable methodological
maxim to assess the correctness of some belief-ascription solely against the
background of the question whether it is traceable elsewhere in its own ‘historical
context’. This deficiency is relevant because not only is uniqueness of a belief
compatible with its truth but history even offers examples showing that views
deviating from common stances can embody particularly deep insights.
For those reasons it is imperative to extend the range of considerations
controlling probability-assessments regarding specimens of textual exegesis. As far
as theoretical texts are concerned theoretical intelligibility constitutes a most
important aspect. This means in particular that suppositions about opinions or
communicative intentions deserve to be confronted with the question of whether or
not some relevant individual or tradition of thought was in the possession of
intelligible theoretical motivations for adopting pertinent opinions or attitudes
inasmuch as their content relates to theoretical, for example philosophical, issues.
This is presumably not a trifling maxim. It implies, for example, that if there should
be textual passages suggesting either rejections or admissions of the acceptability of
sentences and their negated counterparts it is illegitimate to draw the conclusion that
they vindicate exceptional opinions regarding logical matters, such as a rejection of
the law of non-contradiction, without taking the trouble of ascertaining whether or
not theoretical views have been held which at least psychologically sustain the
acceptance of the pertinent opinions. This principle is relevant in particular with
1 It should go without saying that the expression ‘utterance’ is understood here in a way which allows its
application to acts performed in the medium of writing.2 The word ‘cotext’ is employed following the example of its use in traditions of textual linguistics
differentiating between the linguistic environment of some item (‘cotext’) and other text-external factors
(‘context’), such as time and place of utterance etc.—There is no need to dispute the fact that this
characterization is fairly rough because the present connection does not necessitate further refinements.
372 C. Oetke
123
respect to the interpretation of early Madhyamaka-sources, like the Mūlama-dhyamakakārikās (MMK) and the Vigrahavyāvartanī (VV), because it appears thatthose texts have been interpreted in the tradition of Western scholarship in variousdivergent ways, such that some exegetes even attributed to their author philosophicaltheorems akin to tenets advocated by I. Kant or attitudes reminiscent of J. Derridaetc. The crux is that those and similar claims are affected by a methodological defectas long as they do not even attempt to identify in the pertinent works theoreticalelements which are suited to make the adoption of the attributed views plausible. Onthe other hand, it is highly important not to conflate the postulate of finding a basiswhich is connected with certain fundamental tenets by a link of intelligibility orplausibility with the contention that the concerned views or tenets must bethemselves plausible (in our eyes). Possibly a number of scholars were—and still are—reluctant to concede the possibility that both in the MMK and the VV tenets areadvocated which contradict views of common sense and therefore favour differentinterpretations not entailing similar consequences. It is my contention that not onlyvarious attempts of attributing to those works views and concerns which harmonizewith philosophical outlooks prevailing at present are mistaken but that there shouldnot be even any reluctance to acknowledge in those sources the existence of tenetswhich militate against commonly accepted views. We even conjecture that the veryexistence of the VV is due to the circumstance that its author was clearly aware of thefact that the tenets he was advocating are suited to evoke in the eyes of hiscontemporaries impressions of internal inconsistency or of incompatibility withindisputable truths.
II
Possibly the highly controversial nature of the propagated stance is indicated in the
VV at the beginning of the proponent’s reply to the objections formulated in the
section of the kārikā-s 1–20. Verse 22 reads:
yaś ca pratītyabhāvo bhāvānāṃ śūnyateti sā proktā /yaś ca pratītyabhāvo bhavati hi tasyāsvabhāvatvam //
which has been rendered in the translation of K. Bhattacharya as follows:
That nature of the things which is dependent is called voidness, for that nature
which is dependent is devoid of an intrinsic nature (yaś ca pratītyabhāvobhavati hi tasyāsvabhāvatvam). (Bhattacharya 1978, p. 17)
In a recently published article the verse has been rendered by R.K. Sharma likewise:
That nature of the things which is dependent is called voidness, for that nature
which is dependent is devoid of an intrinsic nature . (Sharma 2011, p. 3)
On the other hand an older translation into German given by E. Frauwallner reads as
follows:
The First Section of the Vigrahavyavartanı 373
123
Das abhangige Entstehen der Dinge wird namlich Leerheit genannt.
Denn ein Ding, das abhangig entsteht, ist wesenlos.
(Frauwallner 1994 (1956), p. 201)
The vital difference is that (a) Frauwallner employs different renderings for the two
occurrences of the expression pratītyabhāva- in the first and the second halves of theverse whereas Bhattacharya and Sharma employ the same translation-equivalent
and that (b) Frauwallner interprets the first occurrence of pratītyabhāva- in the senseof ‘dependent origination’ whereas Bhattacharya and Sharma understand it in thesense of ‘the dependent nature (of things’) suggesting the idea that the item denotedby the expression ‘voidness’ as well as the item exhibiting the characteristic of beingdevoid of an intrinsic nature is identifiable with a nature or essence of things orparticulars. Sharma even explicitly equates the proposition that the nature of things isdependent with the proposition that the things originate in dependence, for the authorformulates before the above quoted rendering the sentence:
What does a thing or an existent (bhāva) owe its emptiness to?, presuming
(with Nagarjuna) that it is indeed empty of svabhāva.
and adds immediately after his citation and rendering of the verse the following
remark:
So the existents’ lack of intrinsic nature is asserted by Nagarjuna to consist in
the fact or the consideration that they originate in dependence (ye hipratītyasamutpannā bhāvās te na sasvabhāvā bhavanti svabhāvābhāvāt (VV,vṛtti, k. 22), that they come into being dependent on causes and conditions:hetupratyayasāpekṣatvāt.
Thus the import of the two occurrences of the expression pratītyabhāva- seems to
coincide with the import that has been attributed by Frauwallner to the second
occurrence in particular. It can be presumed that Sharma believes that his
understanding is supported by the prose-commentary which is usually attributed to
the same author as the kārikā-s—although deviant conjectures have been suggestedtoo. A closer look reveals, however, that even under the premise that the vṛttifaithfully reflects the intended import of the verse the contention that the last or evenboth occurrences of the term pratītyabhāva-designate nothing but the particularswhich allegedly exhibit dependent origination is not secure. The relevant portion ofthe commentary reads as follows:
iha hi yaḥ pratītyabhāvo bhāvānāṃ sā śūnyatā / kasmāt / niḥsvabhāvatvāt / yehi pratītyasamutpannā bhāvās te na sasvabhāvā bhavanti svabhāvābhāvāt /kasmāt / hetupratyayasāpekṣatvāt
The fact that Sharma quotes only a selected part of the commentary intimates a
delusive support for his understanding. For it is only in the section beginning with
ye hi that an attribution of the characteristic of lacking own-being (svabhāva) tothings or particulars is expressed. In contrast the first sentence suggests that the
equation with voidness or emptiness (śūnyatā) should not pertain to any items
originating in dependence but rather to the dependent origination of the things.
374 C. Oetke
123
It might be presumed that precisely for this reason Frauwallner adopted two
different renderings for the first and the second occurrence of the term pratītyasa-mutpāda-. Anyhow, as far as the expression niḥsvabhāvatvāt is concerned a linguisticequivocation is noteworthy. According to a natural understanding the characteriza-tion embodied by the term would relate to the denotation of the expressionpratītyabhāvo (bhāvānāṃ) in the immediately preceding phrase so that niḥsvabhā-vatvāt would be equivalent to tasya niḥsvabhāvatvāt conveying that the dependentorigination of things amounts to the same as emptiness (śūnyatā). But this reading isadmittedly not compelling. Understanding niḥsvabhāvatvāt as an elliptical variant ofteṣāṃ niḥsvabhāvatvāt the expression might convey the thought that emptinessamounts to the same as dependent origination because the things which allegedlyoriginate in dependence are devoid of any own-being (svabhāva). Hence the viewsuggested by Frauwallner’s translation could be brought in line with the twoalternative readings of the phrase niḥsvabhāvatvāt.
In spite of this the attribution of deviant imports concerning the two occurrences
of pratītyasamutpāda- in the verse can be put into question. Even if one took for
granted the authenticity of the prose-commentary and its faithfulness regarding the
interpretation of the global import of the verse the issue is not definitely settled. For
the question of how precisely parts of the verse are assignable to specific segments
of the vṛtti is vital. To be sure, the view suggested by Frauwallner would be
supported if alternatively either iha hi yaḥ pratītyabhāvo bhāvānāṃ sā śūnyatā hasto be correlated with the first half of the verse and the subsequent section with thesecond half or iha hi yaḥ pratītyabhāvo bhāvānāṃ sā śūnyatā + niḥsvabhāvatvātwith the first and the subsequent remarks with the second half. But the assumptionthat the assignments have to be construed in this manner is a mere hypothesis.Sharma—and perhaps also Bhattacharya—might be inclined to contend that only thesegments subsequent to either kasmāt or to niḥsvabhāvatvāt correlate to material inthe verse. On the other hand, the hypothesis that the two occurrences ofpratītyasamutpāda- should convey exactly the same import can be maintained evenunder the assumption that the two halves of the kārikā possess explicit correlates inthe commentary. This would hold good in particular if the first half correlated withyaḥ pratītyabhāvo bhāvānāṃ sā śūnyatā and the second half with niḥsvabhāvatvāt,read as equivalent to tasya niḥsvabhāvatvāt. To be sure the pertinent hypothesisentails that the subsequent remarks do not represent any paraphrases of elementsoccurring in the verse, but this circumstance should not constitute a decisive counter-argument because the remarks beginning with ye hi might be intended as presentingsupport for the contention expressed in the kārikā 22. A more serious obstacleagainst accepting this view could lie in the circumstance that under the envisagedassumption the verse would not convey any plausible import. It can be demonstrated,however, that this is by no means the case.
Given that the phrase yaś ca pratītyabhāvo bhāvānāṃ śūnyateti sā proktā isinterpreted in a manner suggested by Frauwallner’s translation in the sense of:
That which is [according to traditional Buddhist dogmatics] the dependent
origination of things, is designated [by us] by [the term] ‘voidness’
The First Section of the Vigrahavyavartanı 375
123
the following half of the verse should convey the corresponding reason, which lies
in the fact that that which is [in the view of Buddhist tradition] dependent
origination exhibits the feature of being devoid of a svabhāva. This in its turn is
justified by the circumstance that dependently originating particulars, being
dependent on causes and conditions, lack svabhāva. The underlying principle isobviously that if certain items lack svabhāva, then lack of svabhāva must equallycharacterize any purported item that could be designated by terms of the form ‘theorigination of x’ or ‘the dependent origination of x’, where the place of ‘x’ would besatisfied by terms designating entities lacking svabhāva. Given that being devoid ofsvabhāva is taken to entail some, possibly qualified, denial of existence, thisprinciple appears plausible. If it turns out that some item, contrary to firstappearances, does not exist, it seems to follow that even events like its origination,destruction, etc. cannot be existent against the backdrop of corresponding existence-criteria. It is true that neither in the kārikā nor in the corresponding prose-commentary any reasoning is presented attempting to demonstrate why particularscannot possess any svabhāva so that lack of svabhāva with respect to dependentorigination can be derived. But this is not inexplicable: The author of the VV canhave presumed that the task of demonstrating this tenet has been accomplished in theMMK and that a repetition of the pertinent pieces of reasoning is not called for.Instead the writer of the text considers it appropriate to highlight a pertinent pointrelating to the consequence of the theorem of universal lack of svabhāva, namely thefact that the assumption of dependent origination, lack of svabhāva and pervasiveemptiness does not rule out the possibility of causal efficiency at the phenomenallevel of ordinary experience. Against the background of the Buddhist tradition thisconsideration possesses weight which does not crucially depend on the validity ofreasoning against the existence of svabhāva at any level: The established tenet, thatobjects of ordinary experience, like persons, chariots, etc. are non-existent against thebackground of existence-criteria which could be considered as ultimately valid, is notregarded in the Buddhist tradition as a theorem establishing the ineffectiveness ofthose types of objects on the level of everyday practice. Hence on condition that theBuddhist tradition equally accepts that the concerned objects are devoid of svabhāva,they are committed to acknowledge that an objects’ lack of svabhāva does not entailits causal inefficiency on all levels, and in particular not on the level of ordinaryexperience and practice.
To be sure, we have propagated an interpretation according to which the position
of the writer of the VV crucially differs from the traditional outlook inasmuch it
involves the claim that—in contrast to assumptions made in Buddhist dogmatics—
no entities exist at any possible level which could be credited with a possession of
svabhāva. It could be therefore argued that the only explanation for the
circumstance that objects of ordinary experience are able to perform functions on
the level of phenomenal appearances lies in positing the existence of some entities
belonging to some underlying (theoretical) level that must be credited with
possession of svabhāva. But even if this reasoning were objectively valid, it wouldnot follow that the interpretation of the position advocated in the VV must beincorrect. Presumably the writer of the text would not have attributed to the indicatedcounter-argument decisive weight, assuming that the pieces of reasoning expounded
376 C. Oetke
123
elsewhere, in particular in the MMK, (allegedly) establishing that attribution ofsvabhāva to any entity necessitates inconsistencies are more forceful and invalidate theconsideration relating to the explanation of observable phenomena. Moreover, giventhat the concern of the author of the VV in the pertinent section of the text lies inestablishing that his own position is free of internal inconsistencies the achievement ofhis goal cannot be impaired by the possibility of a counterargument of the abovecharacterized sort. For if the counterargumentwere compelling it could not demonstratean immanent inconsistency affecting the Madhyamaka-position but at most thepossibility of questioning its truth against the background of external considerations ofplausibility. For these reasons no valid basis exists for rejecting the proposedinterpretation of kārikā 22 in view of the circumstance that the prose-commentary inthe section beginning with yathā ca pratītyasamutpannatvāt svabhāvaśūnyā apihighlights the circumstance that ordinary objects are able to perform practical functions.3
One could presume that an internal inconsistency jeopardizes the propounded
interpretation of verse 22 due to the circumstance that it has been asserted on the
one hand that according to the author of the VV lack of svabhāva characterizes
dependent origination and on the other hand that the writer of the text expresses the
proposition that dependent origination lacks svabhāva. But one might object thatdenial of svabhāva with respect to dependent origination is incompatible with theattribution of any characteristic to that item, and accordingly also with the attributionof the characteristic of lack of svabhāva. This objection can be easily countered,however, by pointing out that an attribution of a characteristic to an item allegedlydevoid of svabhāva is merely apparent. Given that in the theoretical context of earlyMadhyamaka attribution of lack of svabhāva amounts to a disclaimer of existence onany ultimate level of reality its nature does not equal the attribution of some propertyto some item whose existence is presupposed but rather to the sort of characterizationexpressed by statements to the effect that certain units do in fact not exist. The pointis connected with the issue that existence denials do not necessarily possess thelogical nature of property ascriptions. One can conjecture that in the historicalcontext in which the doctrine of emptiness has been originally advocated thispeculiar feature could evoke a considerable amount of puzzlement. But this is adifferent matter which neither establishes that the doctrine defended in the VV isobjectively inconsistent on that account nor that it cannot have been actuallyadvocated by the author of the text.
It is true, nevertheless, that even if no immanent inconsistency can be attributed
to the claim advanced in VV 22 it is highly controversial and bold. For by saying
that dependent origination lacks svabhāva implying its non-existence on any
ultimate level of truth (paramārtha) it gainsays the most central element of Buddhist
tradition, given that the expression pratītyabhāva- in the verse is merely an
alternative term for pratītyasamutpāda. But this consideration cannot vindicate the
correctness of the interpretation suggested by Sharma and insinuated by the
translation presented in Sharma (2011) and Bhattacharya (1978). For the expression
3 Since the tenets expressed in the pertinent segment of the VV according to the interpretation advocated
here do not possess any entailment pertaining to the ascribability or non-ascribability of dispositional
properties to ordinary objects, the deliberations formulated in Sharma (2011, pp. 7–11) are immaterial.
The First Section of the Vigrahavyavartanı 377
123
of an equation of pratītyabhāva with emptiness (śūnyatā) cannot be denied on anyaccount, and on account of purely linguistic reasons it could not be maintained thatthe expression pratītyabhāvo bhāvānānāṃ relates to the same range of items towhich the expression bhāva- would refer. Hence one cannot recognize any legitimatereason to re-interpret the verse in a way which eliminates its controversial aspects.
III
According to a view prevalent in Buddhist thought true assertions can be conveyed
by using formulations which would be bound to represent inconsistent statements if
they had to be interpreted in an ordinary sense and which convey an import that
presupposes a distinction between different levels of ‘reality’ or ‘truth’. Thus
pointing to a chariot one could be entitled to use the words
This does not exist.
for conveying a claim that represents a truth according to the pertinent outlook. The
idea is that—allegedly—ordinary objects cannot be considered as existing on a final
level of analysis because they exhibit a ‘composite nature’ requiring the existence of
(different) constituents. Given that somebody who formulates such a statement is a
psycho-physical entity exhibiting a similar composite nature it would be appropriate
on the basis of the concerned view to make analogous disclaimers of existence
regarding that item. It should be obvious that, if it is true that the tenet defended in the
VV embodies a radicalization of that outlook to the effect that the idea that objects
acknowledged on the level of everyday experience are based on objects which are
themselves free of any sort of dependence is necessarily inconsistent, it must be in
accordance with this teaching to affirm that all items acknowledged in ordinary
experience are neither existent nor based on entities whose existence can be admitted
in the light of rigorous criteria. This intimates the view that the existence of objects of
everyday experience can be acknowledged at best with respect to a merely
phenomenal level of reality which is, however, comparable to an illusion precisely
inasmuch as its constituents cannot be reduced to objects whose existence could be
vindicated in the final analysis. If, however, precisely this fact were asserted it should
follow from the pertinent premises that on a purely phenomenal level an assertion
would occur which conveys information about the status of the level on which it
(apparently) occurs to the effect that on no level of final analysis items exist which
could perform the role of constituting the particulars whose existence is ordinarily
taken for granted. One might say that the central tenet advocated in the VV represents
a, so to speak, ‘external’ theorem with respect to the world of ordinary experience
contrasting with statements which purport to communicate what is the case or what is
not the case in the realm to which those statements could be themselves allocated.We
presume that precisely this peculiarity is highlighted in verse 28 of the VV and in the
corresponding prose commentary—and specifically in the section which immediately
precedes the introduction of verse 29.
In Sharma (2011) the above presented considerations are dismissed. As a
consequence the author disclaims the possibility of any interpretation of kārikā 29
378 C. Oetke
123
of the VV which relies on a recognition of the peculiar nature of the advocated
tenet. Thus in Sharma (2011, pp. 20–21) we can read the following remarks:
Indeed, Oetke in an earlier analysis of Nagarjuna’s statement asserts that
Nagarjuna’s “denial of the existence of any assertion of his own should be
taken as pertaining to the paramārtha-level, because the main tenet logically
entails precisely this”
and
It is on this view of Oetke’s, that Nagarjuna’s denial that he has any pratijñā tomake should be taken as occurring at the paramārtha level, that I wish brieflyto comment.
As far as the second remark is concerned it should be evident that—irrespective of
what the author of the VV might have himself thought about the pertinent matter—
the claim that ‘Nagarjuna’s denial that he has any pratijñā to make should be taken
as occurring at the paramārtha level’ is a distortion of the concerned interpretationof the text. In fact it is even a relevant misrepresentation because the import whichthe remark attributes to it amounts to the contrary of its original sense. Against thebackground of the exposition delineated in the preceding section it is clear that onecannot assert that according to the writer of the VV a denial of making any assertionoccurs at the paramārtha-level, given that the term paramārtha relates to anyultimate level encompassing particulars on which ordinary phenomena might bebased. Certainly the phrase ‘Nāgārjuna’s denial of the existence of any assertion ofhis own should be taken as pertaining to the paramārtha-level’ ought not possess thesense of ‘Nāgārjuna’s denial of the existence of any assertion of his own should beconsidered as occurring on the paramārtha-level’ but rather convey the thought thatthe content of the pertinent assertion (apparently) occurring at the level of ordinaryexperience relates to the level of paramārtha, or more precisely: the tenet which is—on the level of ordinary experience—propagated and justified by the writer of the VVembodies a claim implying that on any level of ultimate analysis or reality(paramārtha) no particulars exist which could constitute a basis for objects ofeveryday experience. The significance of this theorem relies on the circumstance thatit deviates from common assumptions of Buddhist dogmatics and does not amount toa universally acknowledged commonplace. It might be correct to say that the tenetentails that viewed from a perspective of ultimate analysis the non-existence ofparticulars on a level of paramārtha cannot constitute any content of any assertionexisting from the viewpoint of that level. But one is entitled to presume that thewriter of the VV would not be impressed by the objection that he implicitly disclaimsthe existence of any true content of any existing assertion made by him. According tocustomary—and presumably sound—intuitions being a fact or a truth does notnecessitate being asserted as such: It is not evidently inconsistent to suppose that atlater periods of the history of the universe no consciousness will exist and that therewill be several truths connected with those future stages which are never recognizedas such and are never asserted. Accordingly it would not be absurd to contend that itis important not to confuse the question of appropriateness of describing the tenet ofthe non-existence of particulars as the content of an assertion with the question of its
The First Section of the Vigrahavyavartanı 379
123
truth and that even from a perspective of ultimate analysis it is true that no particulars(constituting apparent objects of ordinary experience) exist, although it should nothold true from that viewpoint that precisely this truth is ever asserted by somebody.
Against this background a definite reply can be provided to the following
deliberations propounded in Sharma (2011, p. 22):
Another question that arises with regard to Oetke’s paramārtha-view of
Nagarjuna’s denial is whether such a view would allow assertion of a thesis/
view on Nagarjuna’s part at the saṃvṛti level. If Oetke answers ‘yes’, then our
question would be whether dependent origination, with which (svabhāva-)śūnyatā is equated by Nāgārjuna, is something that is a feature of bhāvas fromthe paramārtha-viewpoint in such a way that at the ordinary saṃvṛti level thereis no dependent origination.
As far as the question raised in the first sentence of the quoted passage is concerned
a straightforward answer is to be given: Yes, the thesis propounded in the VV allows
for the assumption that on the level or from the perspective of saṃvṛti an assertion
of a thesis performed by the writer of the VV exists, provided 1) that the import of
the expressions ‘assertion’ and ‘thesis’ is not artificially restricted and 2) that the
view concerned is not taken as implying an acknowledgment of the existence of a
saṃvṛti-level as a separate entity existing from the ultimate viewpoint. Since theimport of the constituent ‘in such a way that’ in the second sentence is notimmediately clear it should be first clarified that according to the view advocated herethe two component phrases, viz. ‘dependent origination is something that is a featureof bhāva-s from the paramārtha-viewpoint’ and ‘at the ordinary saṃvṛti level thereis no dependent origination’ would represent an untruth. A true proposition could beexpressed rather by the formulation: ‘Dependent origination is not a feature of bhāva-sfrom the paramārtha-viewpoint in such a way that at the ordinary saṃvṛti-level there isdependent origination’, given that the expression ‘in such a way that’ essentiallyamounts to a conjunction. The truth of the first phrase of the sentence-coordinationwould be even explicitly acknowledged in the section of verse 22 if the interpretationenvisaged in the preceding paragraph should be correct. The second phrase at leastpermits a reading according towhich it expresses a true content, given that attribution ofexistence to dependent origination at the ordinary saṃvṛti-level is not taken as entailingan acknowledgment of dependent origination as an entity existing at the level ofultimate analysis.
The mere circumstance that this position does not reflect common intuitions
cannot discredit the interpretation entailing the attribution of this view as long as the
maxim is regarded as valid that the history of philosophy is not necessarily restricted
to the propagation of tenets which appear immediately plausible to ordinary
persons. It can appear that Sharma adopts a different stance in this regard. For the
following remarks occur in Sharma (2011, p. 22):
There is another paradox which ensues from Oetke’s view. If the non-existence
of all pratijñās (assertions)—whether one’s own or any other—is entailed or
“necessitated” (as Oetke says) by the fact that all things (bhāvas) are void, thenone’s own existence, whether as a momentary entity or as a series of
380 C. Oetke
123
momentary particulars, also cannot be asserted as something real; which means,Nāgārjuna has to deny that he (however this ‘he’ be viewed) himself exists. Butcan Nāgārjuna deny his own existence whether in thought or in words withoutrisking self-contradiction. What would be the status, or implication, I want toask of Oetke, of the assertion “I exist” or of “I do not exist”.
As far as the last question raised in the quoted passage is concerned a definite reply
can be provided with respect to the ‘status’ or ‘implication’ of assertions made by
the sentences ‘I exist’ and ‘I do not exist’, at least with respect to a view widely
accepted in Buddhism: The pronoun ‘I’ does, contrary to ordinary conceptions, not
refer to any entity existing on a level of final analysis. Hence, irrespective of
Sharma’s personal intuitions, the possibility of denying the existence of any item to
which personal pronouns might refer was not commonly regarded as an absurdity.
But why should we suppose that the writer of the VV had adopted a different
opinion in this respect?
A more pertinent difficulty could be discerned with respect to mental states or
states of consciousness. Here it is relevant that one must not take for granted the
absurdity of the contention that an impossibility of establishing something’s
objective non-occurrence does not vindicate its objective existence. If somebody
instead of saying ‘I am seeing a blue patch’ employs the formulation ‘It appears to
me as if I was seeing a blue patch’ there is apparently no way of establishing that an
objective untruth has been asserted because instead of making a claim about the way
things really are the concerned person merely communicates how things appear to
him to be. But couldn’t it be argued that precisely for this reason no objective fact
can be asserted in the envisaged situation because something’s being objective
presupposes the existence of a way of establishing that something is really the case
and does not merely appear to be so? If it were true that objective reality of an item
requires the existence of criteria permitting to decide, at least in principle, whether
something in fact exists or merely appears to exist, then the mere impossibility of
establishing that somebody’s ascription of psychological events or states of
consciousness to himself is the outcome of an error cannot be taken as a definite
vindication of their objective existence as particulars.
A related problem is surreptitiously indicated in Sharma (2011, pp. 8–9).
Apparently the author of the VV (and the MMK) does not regard as absurd the
practice of a ‘religious life’. But if the endeavor of being liberated from the
sufferings of life in the ordinary world (of appearances) is a reasonable goal, doesn’t
it follow that the existence of suffering, its origination and cessation or at least the
possibility of the real occurrence of those items need to be acknowledged?4 We
should presume that exactly this predicament has been recognized by the author of
the VV and—given that he is identical with the writer of the MMK—explicitly
4 Cf. Sharma (2011, p. 9): ‘… But the origin of sorrow cannot be denied, for with that goes the whole
idea of cessation of sorrow (duḥkhanirodhasya pratyākhyānaṃ bhavati). But if there is no cessation of
sorrow, that is, if its (sorrow’s) notion is rejected, the Way (mārga) gets rejected, or what is the same,
loses its meaning, for the whole raison d’être of the mārga consists in its leading to the termination of
suffering. The mārga however cannot be rejected according to Nagarjuna, for that will involve
repudiation of the four Noble Truths (āryasatyas) … I think, in the light of the above the following
proposition can be safely asserted: Sorrow exists; its facthood cannot be denied, whatever its cause be …’
The First Section of the Vigrahavyavartanı 381
123
discussed. Since precisely in this connection the significance of the distinction
between lokasaṃvṛttisatya- and paramārthasatya- (or satyaṃ paramārthataḥ inMMK 24, 8) is emphasized it can be supposed that according to the earliestproponent(s) of Madhyamaka-teaching a tenet has been advocated which in factentails a denial of the existence of items like suffering, the cessation of suffering insome qualified sense, i.e. a denial of their existence from an ultimate viewpoint.
In this place it might be objected that ascribing this view to the writer(s) of
Madhyamaka-texts amounts to attributing a completely incoherent view to him
(them). For isn’t it preposterous to even disclaim any difference between existence
and non-existence of suffering by disclaiming the possibility of the occurrence of
suffering from some point of view of ultimate analysis? Would not this supposition
contradict the very distinction between the dimensions of lokasaṃvṛtti(satya-) andparamārtha(satya-) which had been emphasized in the MMK? Against those
objections it can be pointed out first that at least some writer of a Madhyamaka text
did not hesitate to advocate those allegedly incongruous tenets. For we encounter in
MMK 25,19 the following remark:
na saṃsārasya nirvāṇāt kiṃcid asti viśeṣaṇaṃ /na nirvāṇasya saṃsārāt kiṃcid asti viśeṣaṇaṃ //
There is no difference at all between Sam˙sara and Nirvan
˙a /
There is no difference at all between Nirvan˙a and Sam
˙sara //5
In view of this one should be entitled to ask: Where is the proof establishing that the
writer of the VV must have advocated tenets which smoothly conform to Sharma’s
outlook? But we can readily admit that Sharma’s assessment of absurdity is by no
means unwarranted. The problem is that any doctrine entailing the impossibility of
change on some level of ultimate reality is confronted with the problem of
accounting for the occurrence of change at least on the level of appearance.
However, if this circumstance constitutes a serious problem for the teaching of
Madhyamaka and if it should render it incoherent or even absurd from an objective
point of view then the same predicament affects all the more other teachings that
have been propagated in Indian thought, and this holds good in particular of the
doctrine of Illusionary Monism attributed to the school of Sam˙kara. If any change is
merely illusory and does not pertain to the Being which is alone acknowledged as
ultimately real, how can the status of illusion and the (apparent) occurrence of
change as well as the difference between illusion and reality be accounted for? Is
there anybody who is willing to subscribe to the contention that no interpretationcan be correct entailing the ascribability of tenets which are possibly incoherent or
even absurd from an objective point of view? Then it should follow that the entire
tradition of scholarship must be dismissed which supposes that in the tradition of
Indian thought a doctrine of Illusionary Monism has been propagated. In this case
one would presumably not need to resort to further considerations questioning the
value of Indian Studies.
5 Or:
Sam˙sara does not possess any factor which distinguishes it from Nirvan
˙a /
Nirvan˙a does not possess any factor which distinguishes it from Sam
˙sara //
382 C. Oetke
123
Regarding the alleged impasse of being committed to deny the ultimate existence
of any subject of assertion and experience by disclaiming the occurrence of events
and acts of assertion it deserves to be pointed out that any need to acknowledge the
existence of an entity to which personal pronouns refer has to rely on additional
assumptions which are at least not evidently implied by a recognition of mental
events or states of consciousness. To be sure, in the final analysis the mundane view
according to which personal pronouns, including ‘I’ and synonyms, refer to items
which are the same as those to which ordinary proper names refer is presumably
correct and the assumption that psychological events and states can occur without
the existence of an item possessing them might be indeed untenable.6 Accordingly
both the views propagated in the tradition of Buddhism before the writing of the
MMK and the radical tenets defended in the VV implying the impossibility of
attributing ultimate existence either to acts of asserting or to subjects of assertion
might be criticizable. But it is undeniable that in the Indian tradition the thesis also
has been advocated that ‘I’ refers to a purely spiritual entity. Not only is this tenet
presumably wrong but it has also been sustained by arguments which turn out to be
fallacious on a closer examination. What matters is, however, that notwithstanding
its presumable untruth the tenet of a spiritual self is theoretically motivated, and in a
considerable number of cases one can definitely identify objective sources of error
underlying the attribution of plausibility to the tenet as well as to the arguments
sustaining them. In a similar manner the circumstance that the central tenet of
Madhyamaka might be open to criticism does not militate against the supposition
that it is theoretically motivated and that one could even ascertain the crucial points
which render the argumentative support questionable. The verdict that in the MMK
a metaphysical thesis is based on an intelligible theoretical motivation which is,
nevertheless, inconclusive is appropriate if it holds true that the relevant pieces of
reasoning vitally rely on the hypothesis that all varieties of requirement-relation are
asymmetric like causal relations and the sort of connection designated by the term
pratītyasamutpāda. Whenever some interpretation of a philosophical tenet can be
connected with a pertinent theoretical motivation, the mere circumstance that it does
not comply with ordinary intuitions cannot establish its incorrectness.
IV
As consideration of details of wording occurring in a text provides a tool for
restricting the range of acceptable interpretations even in cases in which
formulations exhibit equivocations or vagueness, it provides a most important
criterion for deciding whether or not the attribution of unfamiliar views to some
author rests on a sound basis. For this reason certain considerations expressed in a
6 This does surely not imply that the appropriate import of the expression ‘possess’ or its synonyms is
immediately evident. It is probably true that any possessor of psychological—or other personal—states is
not linked to them by a relation in a genuine sense of the term. But this does not bestow validity on pieces
of reasoning suggested, among others, in the Abhidharmakośa that since—unlike in a number of other
cases—possession cannot be explicated as a proper relation the supposition of a subject of experience and
of an item to which personal pronouns refer is untenable.
The First Section of the Vigrahavyavartanı 383
123
footnote in Sharma (2011) are significant. The relevant segments of the footnote
read as follows:
… Oetke’s further contention (in reply to Ruegg) that ‘nāsti ca mama pratijñā’does not entail that Nagarjuna “has said something specific about himself or
about any assertion made by him …” (p. 457) is open to question. It is open to
question, for Nagarjuna is not merely saying, as Oetke seems to believe, that
his pratijñā too, like everything else, is without svabhāva and so without anydistinguishing mark, but also and perhaps more importantly, that he has, unlikeothers, no pratijñā at all to offer or make. In other words, Nāgārjuna seems tobe conveying that his so-called or apparent standpoint involves no (real)standpoint or position for the simple reason that he has nothing to assert, unlikeother people’s assertions which may often involve one or another view orstance; after all Nāgārjuna does not say “nāsti ca mamāpi pratijñā” (I too haveno assertion to make); and this is important. Indeed, the initial words in VV 29“yadi kācana pratijñā syān me” (If I have any assertion of mine) seem toconfirm what we are attributing to Nāgārjuna. And it is because he has noassertion to make that Nāgārjuna says that no defect attaches to any specificcharacter of his (alleged) proposition (mama pratijñālakṣaṇaprāptatvāt: VV,vṛtti, k. 29). It is also to be noted that subsequently, as we shall see in the lastpart of our paper, when Nāgārjuna rejects the pramāṇa doctrine, he does so orattempts to do so mostly or mainly on independent philosophical/logicalgrounds and not merely because, what someone like Oetke should inconsistency hold, pramāṇas too, like all other bhāvas, are devoid of self-existence (niḥsvabhāva). I recognize that certain of Nāgārjuna’s remarks in the(auto) commentary on VV 29—tasmāt sarvabhāveṣu śūnyeṣv (in the quotedtext: śunyeṣv) atyantopaśānteṣu prakṛtivivikteṣu kutaḥ pratijñā (in the quotedtext: pratijña) etc.—pose some problem for the interpretation we have given,but that we think is due to the fact that Nāgārjuna’s utterances involve twodistinct but only half-explicit standpoints: one that he, unlike others, has noassertion to make or view to advance, and second, that his so-called pratijñā isto be understood, much like other existents, as empty because of its beingdependently originated. (Sharma 2011, p. 11, footnote 10)
First it deserves to be pointed out that the contention that the advocated
interpretation is threatened by an inconsistency because, when the author of the
VV ‘rejects the pramāṇa doctrine, he does so or attempts to do so mostly or mainly
on independent philosophical/logical grounds and not merely because … pramāṇastoo, like all other bhāvas, are devoid of self-existence niḥsvabhāva)’ is absolutelyunjustified. The objection wantonly neglects the argumentative structure of the text.
Given that the VV is primarily designed to defend some basic tenet(s) against
possible objections it will need to perform at least the following tasks: (a) to
establish that the concerned tenet(s) are mutually consistent and that their content is
compatible with the attitudes which its (their) proponent adopts, (b) to demonstrate
that the pertinent theorem(s) is (are) not in conflict with true propositions which,
even if not explicitly advocated, are either implied by the content of the advocated
thesis (theses) or such that the attitudes adopted by its (their) proponent commit him
384 C. Oetke
123
to acknowledge their truth. In addition to this a defence of the position of an author
could comprise (c) the demonstration that the relevant tenet(s) are compatible with
facts which need to be accepted on independent grounds. We surmise that, whereas
the segment of the verses 21–29 is primarily concerned with (a), the section of the
kārikā-s 30 ff, and in particular the verses 30–51, are concerned with (b). Even if thisanalysis of the text were not accepted it needs to be acknowledged that if anopponent claims that some thesis A must be false because it is incompatible with B,supposing that B is true, then there are in principle different options available forcountering the objection, viz. (i) demonstrating that, contrary to the opponent’sassumption, A is not incompatible with B, (ii) pointing out that acknowledgment ofA does not involve the commitment to acknowledge the truth of B too with the effectthat the contention of the truth of B cannot be taken for granted (in the pertinentcontext of discourse), (iii) establishing that the opponent’s assumption of the truth ofB is untenable and that B is (presumably) false. Sharma’s argument rests on theunjustified assumption that in the section following VV 29 the author of the text isobliged to abstain from adopting the third option if he could also employ the secondalternative and if he has in fact resorted to (ii) in some other connection. But thisassumption is utterly implausible. Given that the remark of VV 29 manifests thatfrom the tenets advocated by the author the non-existence of acts of assertiontogether with the non-existence of all particulars (on a level of final analysis) follows,why should the writer of the text be obliged to be merely repetitive in the subsequentdiscussions by confining himself to the remark that, as indicated before, theproponent’s own tenets entail the (ultimate) non-existence of means of knowledge/cognition (pramāṇa) and objects of knowledge/cognition (prameya) instead ofresorting to the stronger strategy of establishing the non-existence of pramāṇa-s andprameya-s as ultimate particulars on independent grounds and thereby refuting theopponent’s attribution of truth to the pertinent propositions, provided the latteralternative is available? The section of the verses 30–51 reveals clearly that the writerof the text indeed assumed the availability of the latter alternative and implemented itby an attempt to demonstrate the impossibility of attributing existence to pramāṇa-sand prameya-s. To be sure, the employed arguments bear significant analogies topieces of reasoning occurring in the MMK and are presumably vulnerable due to thefact that they rely on the assumption that relations of requirement are necessarilyasymmetric. But it should be plausible to assume that the author of the text attributeda higher degree of conclusiveness to this sort of argument in comparison to merelypointing out that the untruth of the opponent’s assumptions follows from thedoctrinal tenets of his own teaching because it does not presuppose the truth of thedoctrinal tenets themselves but rather underlying principles on the basis of whichthe tenets had been derived. We should be entitled to discard Sharma’s contentioninasmuch as it relies on the supposition that the author of the VV can only haveadopted the weakest possible reply in the section dealing with means and objects ofknowledge/cognition.
It might be noted in passing that the adoption of a weak strategy of pointing out
that the proponent’s own tenets entail the (ultimate) non-existence of pramāṇa-s andprameya-s would be even more implausible if in the segment beginning with verse
30 the author envisaged more ambitious goals hypothesized by several scholars,
The First Section of the Vigrahavyavartanı 385
123
apparently including Sharma himself. If one assumed that ‘Nagarjuna’s programme
is to show … that all knowledge-talk is inherently unstable in that such talk wholly
fails to justify its credentials as knowledge’ then the pertinent objective is definitely
unachievable by any contention to the effect that ‘all knowledge talk is inherently
instable’ and ‘wholly fails to justify its credentials as knowledge’7 (whatever this
may mean) if the tenets of Madhyamaka are correct. The same would hold good
with respect to certain other assumptions, e.g. that ‘Nagarjuna’s aim in repudiating
the pramāṇa theory is not so much to call into question the possibility of knowledge
itself as to show that the pramāṇas are by themselves insufficient to prove what they
are taken to be proving—namely metaphysical realism’.8 However, we do not
intend to resort to the supposition that the writer of the VV in fact pursued such
further-reaching aims hypothesized by various scholars because they might be idle
speculations ignoring the circumstance that the text itself explicitly represents its
objective as providing an answer to an objection.
The contention that ‘Nagarjuna’s utterances involve two distinct, but only half-
explicit standpoints: one that he, unlike others, has no assertion to make or view to
advance, and second, that his so-called pratijñā is to be understood, much like other
existents, as empty because of its being dependently originated’ appears to be
elaborated in Sharma (2011, pp. 21–22) where we can read:
It is true that in his vṛtti on VV, k.2 Nagarjuna says that his pratijñā too, like
other bhāvas, is empty and so devoid of own nature, but it is also true that hedenies that he has any pratijñā to make. In other words, his pratijñā rather thanbeing simply empty (śūnya) like other bhāvas is also non-existent in a veryspecial way.
Its special non-existence consists in the fundamental fact that while all other
entities are declared or intimated (jñāpita) as being empty because of being
devoid of an intrinsic nature, it (i.e. Nagarjuna’s pratijñā) is also denied
existence due to its non-assertion (na mama kācid asti pratijñā (VV, vṛtti, k. 29).In other words, Nagarjuna’s pratijñā is not only void of intrinsic nature, it is
also, as unasserted, void of existence in a minimal way, however that term
may be understood or interpreted.
One should be entitled to ask where evidence can be found for the contention that
Nagarjuna’s assertion (pratijñā)9 is ‘non-existent in a very special way’, which is
due to ‘its non-assertion’ on account of the circumstance that it is ‘unasserted’. If the
writer of the VV had anywhere clearly said that any utterance made by himself—
specifically the utterance of words expressing that all things are empty—deviates
from other utterances by not representing an assertion it would be most relevant to
quote the pertinent passage. But as far as one can see no such quotation is presented
in the article. Instead the expression na mama kācid asti pratijñā taken from the
7 Cf. Sharma (2011, p. 35).8 Cf. Sharma (2011, p. 34), where this view is attributed to ‘some writers’.9 Incidentally one might also ask, which particular assertion is at stake here. If one regards the vṛtti asauthentic, the term could only relate to the expression of the words śūnyāḥ sarvabhāvā(ḥ) (‘All things areempty/void) mentioned in the prose commentary on verse 4.
386 C. Oetke
123
prose commentary on verse 29 is mentioned. But referring to this sentence isabsolutely useless for the argumentation because the contentious issue lies exactly inthe question whether the occurrence of this sentence—as well as the expressionsoccurring in its context—really vindicate the contention that the author of the textthereby intends to express any thought implying a deviance between utterances madeby himself and assertive utterances made by other persons. Hence it appears that thecontention expressed by the last sentence in the footnote is based on an entirelycircular argument. It can be admitted, nonetheless, that, as a matter of fact, tenetspropounded in Madhyamaka-teaching and in particular the tenet expressed by thewords ‘All things are void/empty’ exhibit features due to which their contentessentially differs from the content of the vast majority of other assertions. Thosedifferences do not justify the claim that, in contrast to others, utterances made by theproponent of the VV do not exemplify the status of assertions, provided that theimport of the word ‘assertion’ (pratijñā) is not artificially or, as in later times,technically restricted. But given the content of early Madhyamaka teaching astatement of an essential deviance between the utterance of Madhyamaka tenets onthe one hand and the utterance of Non-Madhyamaka tenets as well as ordinaryassertions on the other hand could be made in compliance with Madhyamaka-doctrine. It might be even conceivable that a Madhyamaka-author signals thisdifference in a rhetorical fashion by using the words ‘I do not make any assertion (inthe ordinary sense)’ or similar formulations. The crucial question is, however,whether this difference, even if in certain circumstances it can motivate theemployment of linguistic items possessing the literal sense of the expressionsoccurring in VV 29, is really at stake in the concerned passage of the text. Sincethe argumentative context does not make a reference to differences pertaining to thecontent of Madhyamaka-tenets in contrast to other teachings obligatory thehypothesis of a twofold import embodied by the words na mama kācid asti pratijñāis completely unfounded. It does not help to resort to the consideration thatinterpreting those words in a sense implying a difference between the utterances of anadvocate of Madhyamaka and those made by other proponents could be brought inharmony with the teachings of (early) Madhyamaka. For this would amount to aconflation between questions pertaining to the meaning of linguistic items andquestions relating to properties of (philosophical) theories or of beliefs held byauthors propounding some theory. It is not our purpose to disclaim the possibility thata number of scholars in Indian Studies are prone to confuse the different issues, butin the present connection this question can be left open.
The previously quoted footnote contains another ingredient which appears in fact
significant and it is in the first place this point which makes a consideration of
Sharma’s remarks worthwhile. The writer of the footnote declares as important the
circumstance that neither in verse 29 nor in the related prose commentary an
expression possessing the sense of ‘I too have no assertion’ (nāsti ca mamāpipratijñā) occurs. At first glance one could be inclined to dismiss that remark as
completely irrelevant because from the circumstance that some conclusion is
derived from a premise that would equally warrant the derivation of some other
conclusion it does not follow that the alternative conclusion is explicitly derived.
This holds true even if the alternative conclusion is ‘stronger’ than the actually
The First Section of the Vigrahavyavartanı 387
123
derived one inasmuch as it entails it. To be sure, nobody can be accused of some
sort of linguistic misbehavior if he utters the sequence of phrases:
All small children are fond of teddybears and Natalie is a small child.
Therefore Natalie is fond of teddybears.
To be sure, the (truth of the) premise of the first line would equally license the
derivation of a (true) proposition expressible by the formulation:
Therefore also Natalie is fond of teddybears.
But should we suppose that the mere occurrence of the first instead of the second
formulation disproves the supposition that the sentence is asserted on the basis of its
derivability from a premise like the above mentioned one? In a similar manner one
should doubt that the mere non-occurrence of some sentence with the meaning of
‘I too have no assertion’ or rather ‘There is also no assertion of mine’ disproves the
possibility that it is presented as the consequence of a principle entailing that
nobody’s assertion exists (ultimately). After all, at least the prose commentary
appears to vindicate that the proposition expressed by the formulation na mamakācid asti pratijñā follows from a general proposition permitting the same
derivation with respect to any item denotable by the expression pratijñā.Despite of this it would be inappropriate to sweepingly dismiss the issue raised in
this segment of the footnote as irrelevant. First it deserves to be noted that the
consideration articulated in the pertinent passage is related to an important principle
which appears to be unduly neglected. The principle is that for assessing the
correctness of a particular interpretation one ought examine whether the formulation
one encounters is, against the background of certain contextual data, a most suitableexpression of the content which the interpretation entails, instead of verifying
whether the encountered formulation is a possible expression of that content.
Therefore any argument to the effect that some exegesis of VV 29 implies that some
alternative formulation would be more suitable for expressing the content which the
exegesis attributes to the text need to be taken seriously as a matter of principle.
Second, regarding the particular case that is at stake, one would be entitled to argue
that even if no compelling need exists to expect the occurrence of the formulation
nāsti ca mamāpi pratijñā, the mere circumstance that the text (apparently) refers to
some particular assertion or group of assertions and not to assertions in general
requires some explanation. For if the point of VV 29 should lie in the contention that
the writer’s own assertion is ultimately inexistent because according to Madhya-
maka teaching all assertions are ultimately inexistent the theoretically more
important aspect pertaining to assertions in general remains inexplicit at the expense
of the theoretically less significant aspect relating to assertions made by the
proponent of Madhyamaka. One cannot resort to the argument that making the more
general point explicit would require additional expenses in linguistic respect for in
the vṛtti, which alone contains explicit evidence for the derivation of the non-
existence of assertion from a general metaphysical principle, the formulation
na mama kācid asti pratijñā, allegedly meaning ‘there is no assertion of minewhatsoever’, occurs and the general point could be expressed by a mere omission ofthe constituent mama.
388 C. Oetke
123
In answer to this reasoning one can object that a clear distinction needs to be
drawn between general theoretical significance and relevance in some particular
context. For even if it were granted that the proposition that no assertions exist at an
ultimate level is theoretically more significant than the proposition that an assertion
made by a proponent of Madhyamaka is inexistent it does not follow that the latter
one possesses more argumentative relevance in the context in which it appears. This
is by no means a merely theoretical possibility because the prose commentary
contains an explicit statement to the effect that verse 29 is designed to provide an
answer to the objection which has been formulated in verse 4 of the same text. Now,
if one merely supposes that the prose-commentary provides a faithful account of the
intended import, the objection of verse 4 must be the following: If against the
criticism that any words on the part of a Madhyamika must be unable to perform
any function of preventing something because, due to his doctrinal tenet of the
universal non-existence of svabhāva, in his own opinion also word(s) need to be
non-existent,10 the proponent of Madhyamaka replied that on exactly the same
grounds also the attempt to refute the Madhyamaka-position is inappropriate, then
his answer is improper. For it is only the proponent of Madhyamaka who asserts that
all particulars are empty (śūnyāḥ sarvabhāvā iti) and not the opponent, and thereforethe fault applies to his assertion in particular. The point is that one mightacknowledge that presupposing the tenets of Madhyamaka both the wordspronounced by himself and those pronounced by the opponent are ultimatelyinexistent so that if ultimate non-existence entailed causal inefficiency thepredicament of engaging in a futile endeavor of refutation, either of the svabhāvaof things or of the denial of svabhāva, pertains to the proponent of Madhyamaka andits opponent in the same way; but this consideration is of no avail for theMādhyamika because he alone propagates a metaphysical thesis entailing the non-existence of any words, so that, given the truth of the conditional-relation betweennon-existence and lack of efficiency, in contrast to the tenet(s) advocated by theopponent the thesis of the Madhyamaka-proponent is affected by the flaw that itscontent militates against the supposition that employing words for some purpose, inparticular the purpose of convincing others of the truth of something, is a reasonableenterprise. The upshot is that a special defect pertains to the assertion of universallack of svabhāva due to allegedly preposterous consequences following from itscontent. The mere circumstance that an issue is at stake which pertains to theproclamation of the fundamental Madhyamaka-tenet enhances the intelligibility ofthe fact that in a reply to this criticism overriding relevance is attributed to thequestion of the appropriate verdict regarding the thesis of universal lack of svabhāvaat the expense of verdicts about making assertions in general.
However, this contemplation is not yet apt to solve pertinent difficulties. For even
against the above sketched background it is not fully plausible why in the segment
of VV 29 the contentious metaphysical theorem should be employed for deriving
the consequence that on an ultimate level no assertion of universal emptiness exists
10 It might be noted in passing that in the formulation tava hi matena vacanam apy asat,sarvabhāvasvabāvo ‘py asan = ‘According to your opinion both the word is non-existent and the
svabhāva of all particulars is non-existent’ occurring in the prose-commentary on verse 3, the theorem of
non-existence of words is mentioned in a general form.
The First Section of the Vigrahavyavartanı 389
123
so that seen from the viewpoint of ultimate reality no existent flaw can pertain to an
existent assertion of universal emptiness. It is not even clear why the claim of
ultimate non-existence should have been exclusively related to the assertion of the
fundamental Madhyamaka-tenet without at least indicating that the same status
pertains to assertions in general. After all, in the prose section on verse 3 a
formulation has been chosen suggesting that the non-existence of words in general
follows from the tenet of universal lack of svabhāva, whereas the entailment
between non-existence of words and non-existence of words uttered by a proponent
of Madhyamaka has been left implicit, obviously presupposing that the latter
relationship of entailment is so obvious that it need not be highlighted. What weighs
even more is the circumstance that the very reference to an assertion made by a
Madhyamika in combination with the lack of reference to assertions in general is
conducive to misunderstanding suggesting that the verdict should pertain to some
utterance made by a proponent of Madhyamaka in contrast to other utterances.
Hence the problem results that one cannot discern any intelligible reason for
adopting a potentially misleading formulation instead of a wording which prevents
misunderstanding by not referring to anything performed by a proponent of
Madhyamaka.
The crux of the preceding criticism lies in the circumstance that it relies on a
questionable premise. The crucial question is whether the assumption that the
formulations occurring in VV 29 and the prose commentary embody a reference to a
Madhyamika’s own pratijñā is really true. The only possible support for this
assumption lies in the occurrence of the expressions (a) yadi kācana pratijñā syānme, (b) nāsti ca mama pratijñā, (c) yadi ca kācin mama pratijñā syāt, (d) na mamakācid asti pratijñā. Whereas (a) and (b) occur in the verse, (c) and (d) areencountered in the prose commentary. It appears to be generally assumed that theunderlying constituent structure of na mama kācid asti pratijñā corresponds to
[na[kacit[mama pratijna]asti]]
and similarly for the other expressions. The decisive feature is that mama/me =‘my/of mine’ and pratijñā = ‘assertion’ form together a syntactic constituentpossessing a meaning equivalent to ‘assertion of mine’. If this is combined with theadditional assumption that the expression pratijñā does not relate to an utterance butto the content of an utterance the consequence is intimated that the author of the verseintends to communicate that no utterance of his possesses an asserted content or thathe does not assert any thesis. There is no need to argue over the contention that thoseways of understanding are linguistically possible. The decisive question is whetherthey are compelling and true.
Various textual passages vindicate that a Sanskrit genitive is able to convey the
import of the English preposition ‘for’ and among the occurrences in which a noun
or noun-phrase in the genitive can be rendered by using the preposition ‘for’ there
are cases in which ‘for’ conveys the import of ‘according to’, ‘in the opinion of’.
The following is a selection of pertinent instances:
Nyāyapraveśavṛtti (Dhruva 1968, 21,16) vaiśeṣikasya hi kāraṇatrayāt kāryaṃbhavati / (tadyathā …)
390 C. Oetke
123
‘Because for a Vaises˙ika an effect arises from a triplet of (alternative) causes.
(They are as follows: …)’
Nyāyamañjarī (Varadacharya 1989, 45,7) api ca kriyā’pi pratyakṣadravya-vartinī pratyakṣaiva, bhāṭṭānāṃ pratyakṣaś cātmā‘Moreover even a kriyā that resides in a perceptible substance is in fact
perceptible. But for / [in the opinion] of the Bhat˙t˙a-Mımam
˙sakas the soul is
perceptible.’
Prasannapadā (de La Vallee Poussin 1970, 29,1) tathā hi, yadi cātur-mahābhautikaḥ śabdo gṛhyate sa parasyāsiddhaḥ …(Tib.) … de pha rol po la ma grub paḥo‘For it is so that if sound is taken as something consisting of the four great
elements then it is unestablished for the adversary.’
Prasannapadā (de La Vallee Poussin 1970, 44,14) api ca, yena hisarvadharmāṇāṃ mṛṣātvaṃ parijñātaṃ, kiṃ tasya karmāṇi santi saṃsāro vāstiStcherbatsky (1978) (Part II): 134 ‘And when a man has thoroughly realized
the pluralistic illusion of all separate entities, there is for him no Moral Law.
How can there be any virtuous actions for him, or any phenomenal life?’11
As soon as we suppose that the constituent structure of na mama kācid asti pratijñāis:
[mama[na[kacit pratijna]asti]]
and similarly for the other expressions the conclusiveness of the assumption that the
section of VV 29 embodies a reference to a Madhyamika’s own assertions vanishes.
If the words mama and me were interpreted in a sense of ‘in my view’—
representing an alternative formulation of mama matena, reminiscent of tava (hi)matena in the commentary of VV 3—the thought would be conveyed in VV 29 that
in accordance with Madhyamaka-views there is no assertion (on the ultimate level
of analysis) leaving implicit the consequence that no assertion made by a
Madhyamika exists so that a fortiori no existing flaw can accrue to it. Before a
11 The circumstance that the preposition ‘for’ (or German ‘fur’) can be employed in English (or German)
translations of all the quoted passages does not rule out the possibility that they involve differences of
significance corresponding to different imports of the word. In the first two examples ‘for’ could be
paraphrased by ‘according to’ or ‘in the opinion of’, and apparently a replacement of vaiśeṣikasya by
vaiśeṣikasya matena or bhāṭṭānāṃ by bhāṭṭānāṃ matena would preserve the import. In the third example‘for’ could be substituted by ‘in the eyes of’, ‘given the presuppositions of’, whereas in the last instance anexplication in the sense of ‘in the view of’ represents at most a possible reading. It is conceivable that theremark to the effect that for somebody who has fully realized the illusory character of all dharmas noactions and no transmigration exist should convey the thought that such a person is not affected by (thekarmic effects of) actions and by transmigration, so that the significance of ‘for’ comes close to the importof the word in sentences like ‘there is no danger for him’. One could even surmise that the formulationshould impart several senses simultaneously, specifically that somebody who has realized the illusorynature of all particulars neither entertains the view that karma and transmigration exist nor suffers fromkarma and transmigration. As those exegetical issues do not affect the main issue which is at stake oneneed not pursue them further in the present connection.—For invalidating the claim of a specific referenceto assertions made by the Madhyamaka-proponent the assumption suffices that the genitive of the pertinentoccurrences of mama or me corresponds to ‘for’ in any significance of the word.
The First Section of the Vigrahavyavartanı 391
123
further discussion of the linguistic probability of this analysis let us first consider the
question of the argumentative coherence which it would entail.
It is pretty evident that by pointing out that by acknowledging that the theorem
that all things are empty implies the non-existence of any assertion and consequently
of an assertion made by a proponent of Madhyamaka a vital premise is granted to the
opponent given the way the adversary is represented in verse 4. The assumption that
the writer of the VV hypothesizes that universal denial of svavhāva with respect to
particulars implies their non-existence including non-existence of assertions is
however corroborated by the circumstance that the analogous derivation of the non-
existence of words from the thesis of universal emptiness is, as far as one can see,
nowhere questioned by the writer of the text. But now it could appear that from the
envisaged interpretation it would follow that the author of the VV grants excessively
much to the opponent because he would only corroborate the opponent’s claim that
a specific defect pertains to his thesis since its content implies the unreasonableness
of employing words for any purpose. In this connection it is vital that the opponent’s
criticism relied on the supposition of an implicational connection between non-
existence (on a level of ultimate analysis) and lack of causal efficiency (on a
phenomenal level) as an additional premise. Since exactly this assumption has been
explicitly discarded in the VV the charge of implying the illogicality of uttering
words or making assertions given the premises of Madhyamaka itself has been
effectively refuted. This poses, however, the problem of making the purpose of the
remarks encountered in the section of VV 29 intelligible. If internal consistency
concerning the tenets, personal opinions and attitudes of the proponent of
Madhyamaka has been safeguarded what else should be achieved? To this question
the following answer could be provided: The objections formulated in the section of
VV 1-4 rest on the tacit assumption that the thesis of universal emptiness possesses
a content which is (internally) related to the level of phenomenal reality. The mere
contention that the tenet of universal emptiness does not entail causal inefficiency
on the phenomenal level cannot remove this possible misunderstanding. Moreover,
the criticism advanced in VV 4 rests on the assumption that the premises of
Madhyamaka allow for the possibility that a real defect pertains to the assertion of
universal emptiness. Even the acknowledgment that the claim of universal
emptiness pertains only to some level of ultimate analysis (paramārtha) cannotdispel the hypothesis that apart from a realm of paramārtha a different level ofphenomenal reality exists and that at least on this level items like words or acts ofassertion exist. It is precisely by the unqualified character of the contention that inaccordance with Madhyamaka no assertions exist, in contrast to a statement to theeffect that assertions do not occur on some level of paramārtha, that thismisconception is averted. This means that the writer of the VV is disposed to disputeany type of (ultimate) reality to the world of phenomena. Couched in syntacticalterms it means that one cannot express a truth by embedding sentences like ‘there is alevel of paramārtha and there is also a level of saṃvṛti’ or ‘assertions exist on thelevel of phenomenal reality and assertions do not exist on the level of ultimatereality’ under the operator: ‘Under the viewpoint of ultimate truth it is the case that’.Presumably this contention relies on the same basis as statements to the effect that nodifference exists between Nirvāṇa and Saṃsāra or that no single dharma is ever
392 C. Oetke
123
imparted by the Buddha anywhere for anyone (cf MMK 25,24). If this is correct thenit could be at least in accordance with the outlook of the author of the VV to point outin the context of replying to the objections formulated in VV 1-4 that the charge thatgiven Madhyamaka premises a defect affects the proclamation of the thesis ofuniversal emptiness rests on a fundamentally wrong hypothesis because in realitythose premises rule out even the possibility of such a defect on a non-ultimate level.If this implies that the remarks of VV 29 exhibit a predominantly clarifying characterthat consequence could be accepted.
The assumption that the occurrence of genitive forms of the first personal
pronoun are meant to signal accordance with opinions could be assailed on account
of the actual word order. This possibility is specially virulent for (c) because here
the word order cannot be determined by metrical considerations and the occurrence
of mama in the context of yadi ca kācin … pratijñā syāt does not smoothly
harmonize with the assignment of a constituent structure which separates the
personal pronoun from the noun pratijñā. But this objection rests on the premise thatin the variety of Sanskrit which the writer of the VVemployed word order of surface-structure is, at least in prose, in agreement with underlying constituent structure. It isfar from clear that this supposition is correct. But even if it were conceded it must notfollow that in the section of VV 29 reference pertains exclusively to assertions on thepart of a Mādhyamika. Isn’t it conceivable that in some instances, for example in theconditional clause yadi ca kācin mama pratijñā syāt as well as in the correspondingsection of the kārikā which the expression paraphrases, reference to an assertion ofthe Madhyamaka-proponent is intended whereas in the negated sentence na mamakācid asti pratijñā a reading in the sense of ‘for me there is no assertion whatsoever’is appropriate? For it can be discerned that both a reference to assertions in generaland a reference to assertions made by a Mādhyamika possess contextual relevance. Itcould moreover be surmised that the writer of the text intentionally envisaged atwofold reading precisely on that account. Be that as it may, one should be entitled toassert that an assumption, which seems to be shared by all (Western) interpreters,namely that the combination of the expression pratijñā with a first personal pronounin the genitive must be interpreted as involving a reference to certain specificutterances or their content, is doubtful. If in the later tradition of exegesis inMadhyamaka itself the pertinent phrases of VV 29 had been interpreted in the wayusually assumed by scholars the pertinent assumption would not be vindicatedbecause it is beyond all question that at later times the expression pratijñā has beenassociated with more restricted and sometimes technical imports signifying conceptswhich are in fact inapplicable to the statement of the metaphysical thesis that allparticulars are empty. Thus the employment of arguments from later traditions ofexegesis would merely highlight the methodological faultiness of exploiting a latertradition for exegetical purposes.
V
If the claim is correct that a reading of the genitive form of the first personal
pronoun as signaling accordance with opinions is a realistic possibility regarding a
The First Section of the Vigrahavyavartanı 393
123
central and most debated passage of a text which affects the understanding of
fundamental issues pertaining to an important philosophical teaching, then it must
be surmised that subtle, and partly yet unsettled, linguistic matters possess utmost
potential relevance for the exegesis of theoretical doctrines. This in its turn would
imply that at least at the present stage of research analysis of philosophical thought
represented in classical Indian languages cannot be completely detached from
comprehensive knowledge about the linguistic varieties in which individual textual
sources have been composed.
References
Bhattacharya, K. (1978). The dialectical method of Nāgārjuna (Vigrahavyāvartanī). Translated from the
original Sanskrit with introduction and notes. Delhi: Motilal Banarsidass.
de La Vallee Poussin, L. (Ed.). (1970).Mūlamadhyamakakārikās (Mādhyamikasūtras) de Nāgārjuna avecla Prasannapadā Commentaire de Candrakīrti [Bibliotheca Buddhica 4]. St. Petersburg 1903–1913.
Repr. Osnabruck [Biblio Verlag].
Dhruva, A. B. (1968). The Nyāyapraveśa (Part I). Sanskrit text with commentaries. Baroda: Oriental
Institute.
Frauwallner, E. (1994). Die Philosophie des Buddhismus (4th ed.). Berlin: Akademie Verlag.
Sharma, R. K. (2011). Critical reflections on Nagarjuna’s Vigrahavyāvartanī. In Nagoya studies in Indianculture and Buddhism. Saṃbhāṣā (Vol. 29, pp. 1–38).
Stcherbatsky, T. (1978). The conception of Buddhist Nirvāṇa. (Reprint) Delhi.Varadacharya, V. K. S. (1989). Nyāyamañjarī of Jayantabhaṭṭa (Vol. I). Mysore: Oriental Research
Institute.
394 C. Oetke
123