Buddhist philosophy and psychology

♦ Nāgārjuna on “cause” and “condition”: pars construens (2)

If it is true that both the Pāli Canon and Nāgārjuna consider the conditional relations on the basis of the positive and negative (α and β) twofold formulae (αβ.1 and αβ.2) (see gārjuna on “cause” and “condition”: pars construens (1)), nonetheless between the two perspectives there is a crucial difference: although, according to SN I, 1, we noticed that the kind of relation existing between the conditioning factor and the conditioned one is monodirectional, we find that Nāgārjuna interprets such a conditional relation as being sometimes monodirectional – as in I, 48cd-49cd: asyotpādād udetīdaṃ dīpotpādād yathā prabhā || […] pradīpasyāpi anutpādāt prabhāyā apy asaṃbhavaḥ || («From the origination of this, that arises, like from the [previous] origination of the lantern, the light [arises]; […] moreover, from the non origination of the lantern, surely there is not appearance of light»; it is clear that, without a lamp, no flame is possible) – and sometimes bilateral – as in ŚS 11ab, where he writes: | ma rig ’du byed med mi ’byung | | de med ’du byed mi ’byung bas | («The ignorance does not originate without the pre-forming tendencies and the pre-forming tendencies do not originate without that [ignorance]»).1 We have to note that, in both the examples just quoted, Nāgārjuna treats the (αβ.2) kind of relation (indeed, the Tibetan verb ’byung ba is equivalent to Sanskrit ut√pad).

Now, as far as I, 48cd-49cd is concerned, we can observe that Nāgārjuna here speaks (i) of relations between “concrete” elements (such as the lamp and the flame),2 and admits the need for a pre-existing substratum (the lamp, i.e., the oil and a proper vessel) for the occurrence of a certain other existent (the flame, i.e., the oil burning on the wick), but we have also to note that the simple presence of this substratum does not entail the presence of that other existent, such as the presence of the lamp does not in itself entail the presence of the flame. Indeed, to have a flame from a lamp, one needs some other thing – for instance a match –, which light the oil. Thus we can say that if the lamp (oil, vessel, wick, etc.) plays the role of primary cause, the match plays the role of secondary condition: the cause is that which keeps in itself the possibility to bring forth a certain effect, the condition is that which allows that very effect to pass from potentiality to actuality. In this way, that which can be called effect – but «product» (P. uppanno/-ā/-aṃ, S. utpannaḥ/-ā/-aṃ, T. skye ba) would be a more proper word – arises in the concurrence of both hetuḥ and pratyayaḥ, where neither hetuḥ nor pratyayaḥ – exactly because they are productive of something different from themselves – are thought to be endowed with svabhāvaḥ.3

When Nāgārjuna speaks (ii) of those utpādaḥ-anutpādaḥ relations which take place between “psychological” elements – as in ŚS 11ab – he admits a particular kind of circularity. If, indeed, ŚS 11ab states that ignorance exists only in the presence of pre-forming tendencies and vice versa, in MMK XXVI, 11ab Nāgārjuna remind us that: avidyāyāṃ niruddhāyāṃ saṃskārāṇām asaṃbhavaḥ | («When ignorance is suppressed, there is no occasion for pre-forming tendencies»). This means that, although when origination (utpādaḥ) is considered, both the elements involved in it hang together, nonetheless, when cessation (nirodhaḥ) is considered – as I have explained here –, only from the elimination of the first conditioning factor, there is also the further elimination of the other. Thus, it appears that, according to Nāgārjuna, the utpādaḥ-anutpādaḥ relations are mutual (AB) only in the case of origination between psychological elements, but linear in the case of concrete elements (Ab, where the capital letter identifies the producer) and in the case of cessation between psychological factors (~A→~B). Notwithstanding that, the mutuality about which we have just spoken of is not so bilateral as it could appear because, from the comparison of ŚS 11ab with MMK XXVI, 11ab, we infer that, as far as psychological relations are concerned, the primary cause must be the element whose destruction involves the further dissolution of the other element, whereas the role of secondary condition must be played reciprocally by the two factors. In other words: saṃskhārāḥ represent the condition for further development of avidyā, but only avidyā is both primary cause and secondary condition of saṃskhārāḥ. In this relation we have to note that neither avidyā nor saṃskhārāḥ remain the same: in fact, avidyā is needed for producing saṃskhārāḥ, and those originated (utpannāḥ) saṃskhārāḥ, in their turn, influence that avidyā by modifying it. Furthermore, this modified avidyā, in its turn, produces other saṃskhārāḥ, and so on. We are, thus, in front of an imperfect bilateral relation, that is to say, a relation that we could exemplify with (Ab), but not with (AB), nor with (Ab). Of course, also this relation – like the lamp-flame one – can take place only in the absence of svabhāvaḥ. Nāgārjuna explains it as being a «production», making use of the example of the father and the son:

| pha bu ma yin bu pha min |
| de gnyis phan tshun med min la |
| de gnyis cig car yang min ltar |
| yan lag bcu gnyis de bzhin no ||

The father is not the son, the son is not the father; those two don’t exist the one without the other, moreover those two don’t exist simultaneously: the twelve factors are like them.4

This kārikā – which could be put in parallel with ŚS 11ab and MMK XXVI, 11ab – can be integrated by another one – which could be compared with I, 48cd-49cd –, where Nāgārjuna refers the same example according to another acceptation:

pitrā yadyutpādyaḥ putro yadi tena caiva putreṇa |
utpādyaḥ sa yadi pitā vada tatrotpādayati kaḥ kam ||

If by the father it is produced the son and if by that very son it is produced that [very] father; speak! There, who produces who?5

Only the father is the concrete producer of the son, because the son cannot exist without the father. Notwithstanding this, only a son is that which makes the father a father (of course the same argument can be applied to the mother). In the same way, only ignorance can produce pre-forming tendencies, but those tendencies are that which makes that ignorance to be an actual ignorance. To conclude, we can notice that, when he treats these particular kind of relations, Nāgārjuna refers not to the link itself, rather to the linked events (S. bhāvāḥ, T. dngos po rnams): in the presence of something (lamp, ignorance, father), some other thing (flame, tendencies, son) can originate, whereas in the absence of something some other thing can not originate (concrete relation) or is not further produced (psychological relation).

This last consideration opens the way to a further reflection: if, until now, we have dealt with the utpadaḥ/anutpādaḥ relations, now what can we say about the (αβ.1) model? Nāgārjuna, in I, 48ab-49ab, explains such a kind of relationship as follows: asmin satīdaṃ bhavati dīrghe hrasvaṃ yathā sati | […] hrasve’sati punar dīrghaṃ na bhavaty asvabhāvataḥ | («Being this, there is that, like when there is the long, there is [also] the short […]. Moreover, if there is not the short, there is not the long: [they] are not by [their] intrinsic nature»).6 It is clear, here, that the sort of link which relates the factors involved in it is exclusively logical: indeed, to say that A exists only when there is also B and that B exist only when there is also A is a perfect bilateral relation (AB) – note that no reference is made here to «origination», whereas the verb ’byung ba (S. ut√pad) is fundamental for a correct interpretation of ŚS 11ab.7 The relation described in I, 48ab-49ab is known in Sanskrit with the name of parasparāpekṣā that we can render with «mutual dependence». In Nāgārjuna’s works we meet with several of such cases;8 Sasaki focuses our attention in particular on one of them. He writes (1992: 68-69, note 15): «Although a direct argument of parasparāpekṣā is not found in Nāgārjuna’s treatises, the following verse could be of the foremost importance to prove the implication of it. It runs: pratītya kārakaḥ karma taṃ pratītya ca kārakaṃ | karma pravartate nānyatpaśyāmaḥ siddhikāraṇaṃ || […] It means: The doer is in relation to the deed (karma) and the deed to the doer. We cannot recognize any other cause for this establishment». The kārikā quoted by Sasaki is MMK VIII, 12. Here, the use of the term pratītya must be particularly underlined. If we read I, 48ab-49ab in the light of MMK VIII, 12, it has to be noted that although these kinds of relation are logical relations, nonetheless it appears that Nāgārjuna never considers “long” and “short”, or whatever else, as pure abstract factors. With “long”, he always means “something which is long”: thus, the grasping of its being long depends on a comparison with another thing which is shorter. In other words, what is long is long only in relation of something else that is short, and what is short is short only in relation of something else that is long, and no reference is made to “long” in itself or to “short” in itself.9 In the words of Seyfort Ruegg this sounds as follows (1981:24): «The canonical formula for conditionship “This being that is” (asmin satīdaṃ bhavati) does not then imply self-existence (svabhāva); it refers rather to conditioned relation». The term pratītya, which here corresponds to parasparāpekṣā and then more to Seyfort Ruegg’s «conditionship» than to Sasaki’s «is in relation to», entails simultaneity because a mutual dependence can subsist only if both the elements are existent and linked to one another in the same time. In this case the factor A does not produce the factor B: it simply makes the factor B assume some particular logical characteristic in respect to its relation to the factor A (like the example of the short and the long), without that any actual chagement of both factors occurs. In this case, the accent is not on the events involved – as it happened for the productive relation –, rather on the nature of the relation itself.

Moreover, if pratītya is the mark of a pure conditional relation (i.e., as we have said, of a relation which is perfectly bilateral), this means that also the concept of svabhāvaḥ involved in it has to be different from the concept of svabhāvaḥ involved in the (αβ.2) kind of relation. Indeed, if – as we have seen – in (αβ.2) there is no place for a svabhāvaḥ on account of change (because only without svabhāvaḥ there can be origination, destruction, etc.), in the (αβ.1) kind of relation there is not a real lack of svabhāvaḥ, rather svabhāvaḥ is not implied: in the first case, to admit relationship, we must negate the svabhāvaḥ, whereas in the second case we do not need to affirm any svabhāvaḥ. The conclusion is the same in both cases: the intrinsic nature is not justified.


(1) See Sasaki (1992:67-69, note 16).

(2) I use, here, the term «concrete» in a bare sense and mostly with the meaning of bāhyaḥ («external»), only to signify what is not «psychological» or adhyātmikaḥ («internal»).

(3) We could exemplify this kind of relation in the following way:


Generally, Nāgārjuna upholds that every kind of relation can exist only in the absence of svabhāvaḥ. See PSHK 4cd: stong pa kho na’i chos rnams las | stong pa’i chos rnams ’byung bar zad ||; rendered by Aiyaswami Sastri (1968:13) into Sanskrit as follows: śūnyebhya eva śūnyā dharmāḥ prabhavanti dharmebhyaḥ || («Void dharmāḥ arise only from void dharmāḥ»). Compare with BV 63ab: | mdor na stong pa’i chos rnams las | | chos rnams stong pa skye bar ’gyur |.

(4) ŚS 13.

(5) VV 49.

(6) See AS 13b (dīrghe hrasvaṃ tathā sati); I, 92, 95. Compare with Vṛttiḥ ad ŚS 23cd: […] skye ba la brten nas ’jig pa dang | ’jig pa la brtennasskye ba yin pa de’i phyir yang stong nyid kho na yin no || («[…] destruction is dependent on origination and origination is dependent on destruction; moreover, exactly for this [reason] there is voidness»). See also YṢ 1. In MV 3a (supposed that this work has been written by Nāgārjuna) a similar relation is said to exist between this shore (S. avāraṃ, T. tshu [rol]) and the other shore (S. paraṃ, T. pha rol) of a river (we find the same terms parāvaram in AS 11d). In AS 11ab we read: svatve sati paratve syāt paratve svatvam iṣyate | («If there is own-ness, there is other-ness, if [there is] other-ness, own-ness is accepted»).

(7) Taber (1998) defines such a kind of relation as «the principle of coexisting counterparts».

(8) See, for instance, the extreme case of AS 13a,c: astitve sati nāstitvaṃ […] | nāstitve sati cāstitvaṃ («If there is existence, [it follows that] there is non existence […], and if there is non existence, [it follows that] there is existence»).

(9) In the mutual reference needed for a consistent definition of these kind of related factors (long/short, far/near, etc.) dwell both the Nāgārjunian interpretation of logic (which represents, we can say so, the pure and/or astractive conceptual level of a relational link based on comparisons between effective existents), and its “atemporality”, as it is explained by Oetke (1990:98) who distinguishes between two kinds of condition: logical and real; between these two, only real conditions would have a subtype kind of condition which Oetke calls «causal condition». Furthermore, he writes (1990:99): «there are at least three possible senses in which one could speak of a condition for the applicability of a predicate to a subject or of a concept to a particular […][:] “logical condition”, “causal condition” and “foundational condition” […]. Whereas the kind of condition that was termed “logical condition” is “atemporal” with respect to the conditioning relation, “real conditions” are conceived as involving certain requirements regarding the temporal order». Moreover, the attention payed by Nāgārjuna in making reference (even if implicitly) always to effective long things, short things, and so on, keeps his logical perspective far from dangerous “essentialistic” implications as far as the relationship between a substance (dravyaṃ) and its characteristics (lakṣaṇāni) is concerned.

References 1 (texts):

AS=Acintya-stavaḥ, see Catuḥ-stavaḥ.

Catuḥ-stavaḥ=(1) “The Catustava of Nāgārjuna”, in On Voidness: A Study on Buddhist Nihilism, Tola, F., Dragonetti, C. (ed. and Engl. trans. by), Motilal Banarsidass, Delhi 1995, pp. 101-153; (2) Catuḥstavaḥ of Ācārya Nāgārjuna, Namdol, G. (Hindi trans. and critically ed. in Sanskrit and Tibetan), Central Institute of Higher Tibetan Studies, Sarnath (Varanasi) 2001.

MMK=(1) Mūlamadhyamakakārikā, de Jong, J.W. (ed. by), The Adyar Library and Research Center, Madras 1977; (2) Madhyamakaśāstram of Nāgārjuna. With the Commentaries Akutobhayā by Nāgārjuna, Madhyamakavṛtti by Buddhapālita, Prajñāpradīpavṛtti by Bhāvaviveka, Prasannapadāvṛtti by Candrakīrti, Critically Reconstructed (2 vols.), Pandeya, R. (ed. by), Motilal Banarsidass, Delhi 1988; (3) Mūlamadhyamakakārikā of Nāgārjuna, The Philosophy of the Middle Way, Kalupahana, D.J. (ed. and Engl. trans. by), Motilal Banarsidass, Delhi 1996 (rep.); (4) Candrakīrti Prasannapadā Madhyamakavrtti. Douze chapitres traduits du sanscrit et du tibétain, accompagnés d’une introduction, de notes, et d’une édition critique de la version tibétaine, May, J. (ed. and Franch trans. by), A. Maisonneuve, Paris 1959.

MV (Mahāyāna-viṃśikā)=(1) Minor Buddhist Texts (vol. 1), Tucci, G., ISMEO, Rome 1956, pp. 195-207; (2) Nāgārjuna’s Verses on the Great Vehicle and the Heart of Dependent Origination, Jamieson, R.C., D.K. Printworld, New Delhi 2001.

PSHK (Pratītya-samutpāda-hṛdaya-kārikā)=(1) Nāgārjuna’s Exposition of Twelve Causal Links, Aiyasvami Sastri, N., «Bulletin of Tibetology» 5.2 (1968), pp. 5-27; (2) Nāgārjuna’s Verses on the Great Vehicle and the Heart of Dependent Origination, Jamieson, R.C., D.K. Printworld, New Delhi 2001.

RĀ=Nāgārjuna’s Ratnāvalī vol. I: The Basic Text (Sanskrit, Tibetan, Chinese), Hahn, M. (ed. by), Indica et Tibetica, Bonn 1982.

SN=(1) Saṃyutta Nikāya (5 vols.), Feer, L. (vols. 2-5 ed. by), Pali Text Society, Oxford 1975-1999 (rep.), Somaratne, G.A., (vol. 1 ed. by), Pali Text Society, Oxford 1998; (2) Rhys Davids, C.A.F., Woodward, F.L. (2005), The Book of the Kindred Sayings (5 vols.), Motilal Banarsidass, Delhi.; (3) The Connected Discourses of the Buddha. A New Translation of the Saṃyutta Nikāya (2 vols.), Bhikkhu Bodhi (Engl. trans. by), Wisdom Publication, Boston 2000.

ŚS=(1) Śūnyatāsaptatiḥ with Auto-Commentary of Ārya Nāgārjuna, Sempa Dorje (restored into Sanskrit, trans. into Hindi and ed. by), Central Institute of Higher Tibetan Studies, Sarnath (Varanasi) 1996 (rep); (2) Nāgārjuna’s “Seventy Stanzas”. A Buddhist Psychology of Emptiness, Komito, D.R. (ed and Enlg. trans. by) Snow Lion Publications, Ithaca (New York) 1987.

VV=The Dialectical Method of Nāgārjuna, Vigrahavyāvartanī, Bhattacharya, K., Lohnston, E.H., Kunst, A. (ed. and Engl. trans. by), Motilal Banarsidass, Delhi 1990 (rep.)

YṢ=(1) Nāgārjuniana: Studies in the Writings and Philosophy of Nāgārjuna, Lindtner, Chr., Motilal Banarsidass, Delhi 1987, pp. 100-119; (2) Yuktiṣaṣṭikavṛtti. Commentaire à la soixantaine sur le raisonnement ou Du vrai einsegnement de la causalité, par le maître indien Candrakīrti, Sherrer-Schaub, C.A. (éd. et trad. de la version tibétaine par), Institut Belge des Hautes Études Chinoises, «Mélanges chinois et bouddhique» 25, Bruxelles (1991).

References 2 (studies):

Oetke, C., “On Some Non-Formal Aspects of the Proofs of the Madhyamakakārikās”, in J. Bronkhorst (gen. ed. by), Panels of the VIIth World Sanskrit Conference: Kern Institute, Leiden, August 23-29, 1987 vol. 2, D. Seyfort Ruegg, L. Schithausen (ed. by) Earliest Buddhism and Madhyamaka, Brill, Leiden 1990, pp. 91-109.

Sasaki, G. H., Linguistic Approach to Buddhist Thought, Motilal Banarsidass, Delhi 1992 (rep.)

Seyfort Ruegg, D., The Literature of the Madhyamaka School of Philosophy in India, Otto Harrassowitz, Wiesbaden 1981.

Taber, J.A., On Nāgārjuna’s So-Called Fallacies: A Comparative Approach, «Indo-Iranian Journal» 41 (1998), pp. 213-144.


6 thoughts on “♦ Nāgārjuna on “cause” and “condition”: pars construens (2)

  1. Thanks for these interesting posts, Krishna. Did you investigate further on the consequences of that on Dharmakīrti’s understanding of the avinābhāva (fixed concomitance) between sādhya and sādhana in an inference? Can alpha-beta1 (sorry for the missing Greek characters) serve in svabhāvānumāna?


  2. Ciao Elisa,
    thanks for your comment. Your suggestion to investigate Dharmakīrti’s position on anumāna sounds good! Years ago I have studied the way in which Materialists (Cārvāka) criticize the intervention of vyāpti (or, as you call it, avinābhāva) in an inferential process, basically reading the Sarva-darśana-saṃgraha of Sayāṇa-Mādhava. Of course, for Cārvākas vyāpti do not exist because it cannot be proved by pratyakṣa.
    The αβ1 kind of relationship describes a positive/negative relation between two factors, concerning existence/non-existence, so it apparently fits for svabhāvānumāna, a sort of inference in which one element is so connected with another element that the former can be inferred from the latter. But, as far as my knowledge can go, I interpret the kind of relation depicted by αβ1 as concerning only links between “things” (concrete or psychological they be), not between ideas, judgments, etc. So, if the “logic” seems more or less the same, nonetheless the difference between svabhāvānumāna-perspective and αβ1-perspective lies exactly in the fact that anumāna involves concepts and judgments, which αβ1 appears not to include. In my opinion αβ1 is simply a descriptive – not inferential – statement (thus: we HAVE “A”, we HAVE “B”, but we have “B” only if there is “A”; and not: we SEE “B”, we know that if there is “B” there SHOULD BE “A” somewhere because “B” depends on “A”, so we INFER “A”). In any case, one should analyze with accuracy the Pāli Canon in search of passages – if any – in which the same kind of relation is described as applying to ideas and judgments.


  3. Thanks for the detailed answer. I see your point, but I guess the gap between things and concepts should be bridgeable, since after all, both are vikalpas. One might add that as long as the anumāna is correctly formulated, both are correct vikalpas, that is, not merely delusions (such as dreams).


  4. Ciao Elisa!
    I think that your position on vikalpa is good for a “yogācāric” approach to the problem. Nonetheless, if we take it according to the ancient Buddhist perspective, the “existing A, B exists” kind of relation involves the idea that A and B exist even outside a, so to speak, vikalpa-catergory (of course it is not possible to say that they exist “in sé e per sé”). As far as Nāgārjuna is concerned on this point, in my opinion he walks on the middle way (sorry for the word play!) between early Buddhism and Yogācāra: his treatment of kalpana, vikalpa, etc., appears to involve a strict link between things and ideas both in a prapañca/negative and in an aprapañca/positive way; but in the prapañca mode it appears that things are accompanied with our representation of them because vikalpa precedes things, whereas in the aprapañca mode he seems to be aware of a certain independence of things from mental vikalpa production (i.e., when things precede vikalpa).
    On account of dreams, there is a good example of erroneous vikalpa in Bhavya’s Tarkajvālā (3rd Ch… but at the moment I do not remember exactly where) in connection with saṃvṛti satya (this example can be taken as representative also of the position of Nāgārjuna). What is interesting, here, is not only the instance in itself but also the fascinating prose. The text describes a young man sleeping, and dreaming about beautiful women, castles, temples, armies, etc., and he is the king of all (this remembers some similar passages of the Laṅkāvatāra Sūtra); but when he wakes up, he becomes suddenly aware of the unreality of what he has dreamt about. It is a parable on the “before” and the “after” knowing by means of paramārtha satya. 🙂


  5. Dear Krishna,
    thanks for your prompt answer. Do you mean that Nāgārjuna’s approach is non-illusionistic (that is: he does not assume that the world is unreal)?
    The dream argument is typical as an example of the saṃvṛti/paramārtha argument. But, the point remains: how does one awaken from that dream which we call reality? Moreover, dreams are not completely unreal (so Kumārila…and Freud:-)) because they mix elements of real life. What about what we call ‘reality’? Whence its elements?


  6. Ciao Elisa,
    sorry but today I can answer in brief. In my opinion the approach of Nāgārjuna to reality is similar to that of Husserl: his “path” from saṃvṛti to paramārtha – if I have understood it – can indeed be put in parallel to the “going to the very things” of Husserl. This means that things in some way exist. That “very things” is, in any case, to be understood as “things exist in connection with other things”, they are pratītya-samutpanna, but this does not mean that they are illusions (see the “positive” use Nāgārjuna makes of the term relationalistic bhāva “against” the fixistic sat). The illusion is connected to OUR interpretation of things: it consists on fixing things as independent monades.
    Concerning dreams, I agree with you in thinking that they are part of our reality… (personally I prefer the examples of the magic man, the city of Gandharvas, etc., which refer to “real” unreal things 😉 )


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s